diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,14074 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.2196126277260944,
+  "eval_steps": 500,
+  "global_step": 2000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval_loss": 0.3103026747703552,
+      "eval_runtime": 148.5907,
+      "eval_samples_per_second": 7.201,
+      "eval_steps_per_second": 0.229,
+      "step": 0
+    },
+    {
+      "epoch": 0.0006100350770169285,
+      "grad_norm": 0.8454470981023452,
+      "learning_rate": 0.0,
+      "loss": 0.3287,
+      "step": 1
+    },
+    {
+      "epoch": 0.001220070154033857,
+      "grad_norm": 0.7768184622203883,
+      "learning_rate": 3.0487804878048784e-08,
+      "loss": 0.3074,
+      "step": 2
+    },
+    {
+      "epoch": 0.0018301052310507853,
+      "grad_norm": 0.8507240324798782,
+      "learning_rate": 6.097560975609757e-08,
+      "loss": 0.3157,
+      "step": 3
+    },
+    {
+      "epoch": 0.002440140308067714,
+      "grad_norm": 0.7931819707357988,
+      "learning_rate": 9.146341463414634e-08,
+      "loss": 0.3158,
+      "step": 4
+    },
+    {
+      "epoch": 0.0030501753850846426,
+      "grad_norm": 0.7827519593120217,
+      "learning_rate": 1.2195121951219514e-07,
+      "loss": 0.3054,
+      "step": 5
+    },
+    {
+      "epoch": 0.0036602104621015707,
+      "grad_norm": 0.924981881710818,
+      "learning_rate": 1.5243902439024392e-07,
+      "loss": 0.319,
+      "step": 6
+    },
+    {
+      "epoch": 0.0042702455391185,
+      "grad_norm": 0.8094982837416114,
+      "learning_rate": 1.8292682926829268e-07,
+      "loss": 0.2729,
+      "step": 7
+    },
+    {
+      "epoch": 0.004880280616135428,
+      "grad_norm": 0.7880554417981912,
+      "learning_rate": 2.134146341463415e-07,
+      "loss": 0.3025,
+      "step": 8
+    },
+    {
+      "epoch": 0.005490315693152357,
+      "grad_norm": 0.7959447687740115,
+      "learning_rate": 2.439024390243903e-07,
+      "loss": 0.3053,
+      "step": 9
+    },
+    {
+      "epoch": 0.006100350770169285,
+      "grad_norm": 0.82247619574212,
+      "learning_rate": 2.7439024390243906e-07,
+      "loss": 0.3097,
+      "step": 10
+    },
+    {
+      "epoch": 0.006710385847186213,
+      "grad_norm": 0.8010327654330195,
+      "learning_rate": 3.0487804878048784e-07,
+      "loss": 0.3284,
+      "step": 11
+    },
+    {
+      "epoch": 0.007320420924203141,
+      "grad_norm": 0.867259966178027,
+      "learning_rate": 3.3536585365853663e-07,
+      "loss": 0.3263,
+      "step": 12
+    },
+    {
+      "epoch": 0.00793045600122007,
+      "grad_norm": 0.6858690138397889,
+      "learning_rate": 3.6585365853658536e-07,
+      "loss": 0.3062,
+      "step": 13
+    },
+    {
+      "epoch": 0.008540491078237,
+      "grad_norm": 0.6968955854258266,
+      "learning_rate": 3.963414634146342e-07,
+      "loss": 0.3118,
+      "step": 14
+    },
+    {
+      "epoch": 0.009150526155253927,
+      "grad_norm": 0.7149293379600384,
+      "learning_rate": 4.26829268292683e-07,
+      "loss": 0.2998,
+      "step": 15
+    },
+    {
+      "epoch": 0.009760561232270856,
+      "grad_norm": 0.8384371568701614,
+      "learning_rate": 4.573170731707317e-07,
+      "loss": 0.2989,
+      "step": 16
+    },
+    {
+      "epoch": 0.010370596309287784,
+      "grad_norm": 0.7959356057845837,
+      "learning_rate": 4.878048780487805e-07,
+      "loss": 0.3226,
+      "step": 17
+    },
+    {
+      "epoch": 0.010980631386304713,
+      "grad_norm": 0.5860521260046243,
+      "learning_rate": 5.182926829268293e-07,
+      "loss": 0.3293,
+      "step": 18
+    },
+    {
+      "epoch": 0.011590666463321641,
+      "grad_norm": 0.42435828625242183,
+      "learning_rate": 5.487804878048781e-07,
+      "loss": 0.2976,
+      "step": 19
+    },
+    {
+      "epoch": 0.01220070154033857,
+      "grad_norm": 0.4723062068417477,
+      "learning_rate": 5.79268292682927e-07,
+      "loss": 0.2906,
+      "step": 20
+    },
+    {
+      "epoch": 0.012810736617355498,
+      "grad_norm": 0.5311333611487327,
+      "learning_rate": 6.097560975609757e-07,
+      "loss": 0.3175,
+      "step": 21
+    },
+    {
+      "epoch": 0.013420771694372426,
+      "grad_norm": 0.43195173965189027,
+      "learning_rate": 6.402439024390244e-07,
+      "loss": 0.3326,
+      "step": 22
+    },
+    {
+      "epoch": 0.014030806771389355,
+      "grad_norm": 0.5039374654423954,
+      "learning_rate": 6.707317073170733e-07,
+      "loss": 0.3243,
+      "step": 23
+    },
+    {
+      "epoch": 0.014640841848406283,
+      "grad_norm": 0.48160299981914606,
+      "learning_rate": 7.012195121951221e-07,
+      "loss": 0.3107,
+      "step": 24
+    },
+    {
+      "epoch": 0.015250876925423212,
+      "grad_norm": 0.4076510164467301,
+      "learning_rate": 7.317073170731707e-07,
+      "loss": 0.3197,
+      "step": 25
+    },
+    {
+      "epoch": 0.01586091200244014,
+      "grad_norm": 0.3353897704259775,
+      "learning_rate": 7.621951219512196e-07,
+      "loss": 0.3119,
+      "step": 26
+    },
+    {
+      "epoch": 0.01647094707945707,
+      "grad_norm": 0.3150496553683514,
+      "learning_rate": 7.926829268292684e-07,
+      "loss": 0.3008,
+      "step": 27
+    },
+    {
+      "epoch": 0.017080982156474,
+      "grad_norm": 0.32105723872932035,
+      "learning_rate": 8.231707317073172e-07,
+      "loss": 0.2905,
+      "step": 28
+    },
+    {
+      "epoch": 0.017691017233490924,
+      "grad_norm": 0.3280306420088101,
+      "learning_rate": 8.53658536585366e-07,
+      "loss": 0.2707,
+      "step": 29
+    },
+    {
+      "epoch": 0.018301052310507854,
+      "grad_norm": 0.30340527994818084,
+      "learning_rate": 8.841463414634147e-07,
+      "loss": 0.311,
+      "step": 30
+    },
+    {
+      "epoch": 0.018911087387524783,
+      "grad_norm": 0.32281828347533853,
+      "learning_rate": 9.146341463414634e-07,
+      "loss": 0.3284,
+      "step": 31
+    },
+    {
+      "epoch": 0.019521122464541713,
+      "grad_norm": 0.28874292519953626,
+      "learning_rate": 9.451219512195123e-07,
+      "loss": 0.29,
+      "step": 32
+    },
+    {
+      "epoch": 0.02013115754155864,
+      "grad_norm": 0.3027310057466991,
+      "learning_rate": 9.75609756097561e-07,
+      "loss": 0.314,
+      "step": 33
+    },
+    {
+      "epoch": 0.020741192618575568,
+      "grad_norm": 0.3008812932109969,
+      "learning_rate": 1.0060975609756098e-06,
+      "loss": 0.3144,
+      "step": 34
+    },
+    {
+      "epoch": 0.021351227695592497,
+      "grad_norm": 0.27169807758622333,
+      "learning_rate": 1.0365853658536586e-06,
+      "loss": 0.2963,
+      "step": 35
+    },
+    {
+      "epoch": 0.021961262772609427,
+      "grad_norm": 0.2695339964388015,
+      "learning_rate": 1.0670731707317073e-06,
+      "loss": 0.2893,
+      "step": 36
+    },
+    {
+      "epoch": 0.022571297849626352,
+      "grad_norm": 0.2843145950012066,
+      "learning_rate": 1.0975609756097562e-06,
+      "loss": 0.323,
+      "step": 37
+    },
+    {
+      "epoch": 0.023181332926643282,
+      "grad_norm": 0.29602446856008774,
+      "learning_rate": 1.128048780487805e-06,
+      "loss": 0.3209,
+      "step": 38
+    },
+    {
+      "epoch": 0.02379136800366021,
+      "grad_norm": 0.3423946407175081,
+      "learning_rate": 1.158536585365854e-06,
+      "loss": 0.3061,
+      "step": 39
+    },
+    {
+      "epoch": 0.02440140308067714,
+      "grad_norm": 0.3612641386449874,
+      "learning_rate": 1.1890243902439024e-06,
+      "loss": 0.31,
+      "step": 40
+    },
+    {
+      "epoch": 0.025011438157694067,
+      "grad_norm": 0.3325638984346252,
+      "learning_rate": 1.2195121951219514e-06,
+      "loss": 0.3158,
+      "step": 41
+    },
+    {
+      "epoch": 0.025621473234710996,
+      "grad_norm": 0.31401563220686574,
+      "learning_rate": 1.25e-06,
+      "loss": 0.2922,
+      "step": 42
+    },
+    {
+      "epoch": 0.026231508311727925,
+      "grad_norm": 0.3170156202095663,
+      "learning_rate": 1.2804878048780488e-06,
+      "loss": 0.3141,
+      "step": 43
+    },
+    {
+      "epoch": 0.02684154338874485,
+      "grad_norm": 0.32428209772408506,
+      "learning_rate": 1.3109756097560978e-06,
+      "loss": 0.2877,
+      "step": 44
+    },
+    {
+      "epoch": 0.02745157846576178,
+      "grad_norm": 0.2937477314360728,
+      "learning_rate": 1.3414634146341465e-06,
+      "loss": 0.2825,
+      "step": 45
+    },
+    {
+      "epoch": 0.02806161354277871,
+      "grad_norm": 0.2958373552353005,
+      "learning_rate": 1.3719512195121952e-06,
+      "loss": 0.3008,
+      "step": 46
+    },
+    {
+      "epoch": 0.02867164861979564,
+      "grad_norm": 0.2880444673791773,
+      "learning_rate": 1.4024390243902442e-06,
+      "loss": 0.3016,
+      "step": 47
+    },
+    {
+      "epoch": 0.029281683696812565,
+      "grad_norm": 0.2789112072045315,
+      "learning_rate": 1.4329268292682927e-06,
+      "loss": 0.3025,
+      "step": 48
+    },
+    {
+      "epoch": 0.029891718773829495,
+      "grad_norm": 0.25724915007561794,
+      "learning_rate": 1.4634146341463414e-06,
+      "loss": 0.2747,
+      "step": 49
+    },
+    {
+      "epoch": 0.030501753850846424,
+      "grad_norm": 0.2635046087062574,
+      "learning_rate": 1.4939024390243904e-06,
+      "loss": 0.2984,
+      "step": 50
+    },
+    {
+      "epoch": 0.031111788927863353,
+      "grad_norm": 0.26015662456743477,
+      "learning_rate": 1.5243902439024391e-06,
+      "loss": 0.3243,
+      "step": 51
+    },
+    {
+      "epoch": 0.03172182400488028,
+      "grad_norm": 0.24509502760724197,
+      "learning_rate": 1.5548780487804878e-06,
+      "loss": 0.321,
+      "step": 52
+    },
+    {
+      "epoch": 0.03233185908189721,
+      "grad_norm": 0.2908236246505609,
+      "learning_rate": 1.5853658536585368e-06,
+      "loss": 0.2937,
+      "step": 53
+    },
+    {
+      "epoch": 0.03294189415891414,
+      "grad_norm": 0.2718558063137255,
+      "learning_rate": 1.6158536585365855e-06,
+      "loss": 0.3195,
+      "step": 54
+    },
+    {
+      "epoch": 0.03355192923593107,
+      "grad_norm": 0.22727302568286986,
+      "learning_rate": 1.6463414634146345e-06,
+      "loss": 0.2784,
+      "step": 55
+    },
+    {
+      "epoch": 0.034161964312948,
+      "grad_norm": 0.30238990509268004,
+      "learning_rate": 1.6768292682926832e-06,
+      "loss": 0.3057,
+      "step": 56
+    },
+    {
+      "epoch": 0.034771999389964926,
+      "grad_norm": 0.23644592363288297,
+      "learning_rate": 1.707317073170732e-06,
+      "loss": 0.2666,
+      "step": 57
+    },
+    {
+      "epoch": 0.03538203446698185,
+      "grad_norm": 0.24896950567681486,
+      "learning_rate": 1.7378048780487804e-06,
+      "loss": 0.3081,
+      "step": 58
+    },
+    {
+      "epoch": 0.03599206954399878,
+      "grad_norm": 0.24000954305492433,
+      "learning_rate": 1.7682926829268294e-06,
+      "loss": 0.2852,
+      "step": 59
+    },
+    {
+      "epoch": 0.03660210462101571,
+      "grad_norm": 0.23441121755488187,
+      "learning_rate": 1.7987804878048781e-06,
+      "loss": 0.297,
+      "step": 60
+    },
+    {
+      "epoch": 0.03721213969803264,
+      "grad_norm": 0.22232489603709266,
+      "learning_rate": 1.8292682926829268e-06,
+      "loss": 0.2634,
+      "step": 61
+    },
+    {
+      "epoch": 0.037822174775049566,
+      "grad_norm": 0.24052014046880021,
+      "learning_rate": 1.8597560975609758e-06,
+      "loss": 0.2977,
+      "step": 62
+    },
+    {
+      "epoch": 0.038432209852066496,
+      "grad_norm": 0.23679217772055688,
+      "learning_rate": 1.8902439024390245e-06,
+      "loss": 0.3012,
+      "step": 63
+    },
+    {
+      "epoch": 0.039042244929083425,
+      "grad_norm": 0.2223272566001243,
+      "learning_rate": 1.9207317073170733e-06,
+      "loss": 0.2738,
+      "step": 64
+    },
+    {
+      "epoch": 0.03965228000610035,
+      "grad_norm": 0.22669213310403505,
+      "learning_rate": 1.951219512195122e-06,
+      "loss": 0.2846,
+      "step": 65
+    },
+    {
+      "epoch": 0.04026231508311728,
+      "grad_norm": 0.2414151537454095,
+      "learning_rate": 1.981707317073171e-06,
+      "loss": 0.303,
+      "step": 66
+    },
+    {
+      "epoch": 0.040872350160134206,
+      "grad_norm": 0.2370672126840143,
+      "learning_rate": 2.0121951219512197e-06,
+      "loss": 0.2904,
+      "step": 67
+    },
+    {
+      "epoch": 0.041482385237151136,
+      "grad_norm": 0.22809840538333365,
+      "learning_rate": 2.042682926829268e-06,
+      "loss": 0.2721,
+      "step": 68
+    },
+    {
+      "epoch": 0.042092420314168065,
+      "grad_norm": 0.2272895442779121,
+      "learning_rate": 2.073170731707317e-06,
+      "loss": 0.2707,
+      "step": 69
+    },
+    {
+      "epoch": 0.042702455391184994,
+      "grad_norm": 0.24377585427116602,
+      "learning_rate": 2.103658536585366e-06,
+      "loss": 0.303,
+      "step": 70
+    },
+    {
+      "epoch": 0.043312490468201924,
+      "grad_norm": 0.2254822602630038,
+      "learning_rate": 2.1341463414634146e-06,
+      "loss": 0.2956,
+      "step": 71
+    },
+    {
+      "epoch": 0.04392252554521885,
+      "grad_norm": 0.23412386235006508,
+      "learning_rate": 2.1646341463414635e-06,
+      "loss": 0.2861,
+      "step": 72
+    },
+    {
+      "epoch": 0.044532560622235776,
+      "grad_norm": 0.224370421334331,
+      "learning_rate": 2.1951219512195125e-06,
+      "loss": 0.2911,
+      "step": 73
+    },
+    {
+      "epoch": 0.045142595699252705,
+      "grad_norm": 0.23593807167360495,
+      "learning_rate": 2.225609756097561e-06,
+      "loss": 0.2875,
+      "step": 74
+    },
+    {
+      "epoch": 0.045752630776269634,
+      "grad_norm": 0.22302811027781724,
+      "learning_rate": 2.25609756097561e-06,
+      "loss": 0.296,
+      "step": 75
+    },
+    {
+      "epoch": 0.046362665853286564,
+      "grad_norm": 0.22758322152961505,
+      "learning_rate": 2.286585365853659e-06,
+      "loss": 0.283,
+      "step": 76
+    },
+    {
+      "epoch": 0.04697270093030349,
+      "grad_norm": 0.21591645504761844,
+      "learning_rate": 2.317073170731708e-06,
+      "loss": 0.2882,
+      "step": 77
+    },
+    {
+      "epoch": 0.04758273600732042,
+      "grad_norm": 0.21879095336822568,
+      "learning_rate": 2.3475609756097563e-06,
+      "loss": 0.257,
+      "step": 78
+    },
+    {
+      "epoch": 0.04819277108433735,
+      "grad_norm": 0.22323134239569448,
+      "learning_rate": 2.378048780487805e-06,
+      "loss": 0.2727,
+      "step": 79
+    },
+    {
+      "epoch": 0.04880280616135428,
+      "grad_norm": 0.224943076374116,
+      "learning_rate": 2.408536585365854e-06,
+      "loss": 0.3018,
+      "step": 80
+    },
+    {
+      "epoch": 0.049412841238371204,
+      "grad_norm": 0.23309225935335282,
+      "learning_rate": 2.4390243902439027e-06,
+      "loss": 0.2896,
+      "step": 81
+    },
+    {
+      "epoch": 0.05002287631538813,
+      "grad_norm": 0.21422363034398334,
+      "learning_rate": 2.4695121951219513e-06,
+      "loss": 0.2568,
+      "step": 82
+    },
+    {
+      "epoch": 0.05063291139240506,
+      "grad_norm": 0.23006182370727415,
+      "learning_rate": 2.5e-06,
+      "loss": 0.2893,
+      "step": 83
+    },
+    {
+      "epoch": 0.05124294646942199,
+      "grad_norm": 0.23133424818577172,
+      "learning_rate": 2.530487804878049e-06,
+      "loss": 0.2776,
+      "step": 84
+    },
+    {
+      "epoch": 0.05185298154643892,
+      "grad_norm": 0.22007966657198397,
+      "learning_rate": 2.5609756097560977e-06,
+      "loss": 0.2941,
+      "step": 85
+    },
+    {
+      "epoch": 0.05246301662345585,
+      "grad_norm": 0.22950578198182395,
+      "learning_rate": 2.5914634146341466e-06,
+      "loss": 0.2903,
+      "step": 86
+    },
+    {
+      "epoch": 0.05307305170047278,
+      "grad_norm": 0.2119291405117646,
+      "learning_rate": 2.6219512195121956e-06,
+      "loss": 0.289,
+      "step": 87
+    },
+    {
+      "epoch": 0.0536830867774897,
+      "grad_norm": 0.23230876607332274,
+      "learning_rate": 2.652439024390244e-06,
+      "loss": 0.2827,
+      "step": 88
+    },
+    {
+      "epoch": 0.05429312185450663,
+      "grad_norm": 0.22839417035165244,
+      "learning_rate": 2.682926829268293e-06,
+      "loss": 0.3045,
+      "step": 89
+    },
+    {
+      "epoch": 0.05490315693152356,
+      "grad_norm": 0.22634052130525656,
+      "learning_rate": 2.713414634146342e-06,
+      "loss": 0.2724,
+      "step": 90
+    },
+    {
+      "epoch": 0.05551319200854049,
+      "grad_norm": 0.21975015253358074,
+      "learning_rate": 2.7439024390243905e-06,
+      "loss": 0.2723,
+      "step": 91
+    },
+    {
+      "epoch": 0.05612322708555742,
+      "grad_norm": 0.225336062841578,
+      "learning_rate": 2.7743902439024394e-06,
+      "loss": 0.2998,
+      "step": 92
+    },
+    {
+      "epoch": 0.05673326216257435,
+      "grad_norm": 0.21097673082672147,
+      "learning_rate": 2.8048780487804884e-06,
+      "loss": 0.2827,
+      "step": 93
+    },
+    {
+      "epoch": 0.05734329723959128,
+      "grad_norm": 0.21759463074705687,
+      "learning_rate": 2.8353658536585365e-06,
+      "loss": 0.2841,
+      "step": 94
+    },
+    {
+      "epoch": 0.05795333231660821,
+      "grad_norm": 0.23262389502742112,
+      "learning_rate": 2.8658536585365854e-06,
+      "loss": 0.3015,
+      "step": 95
+    },
+    {
+      "epoch": 0.05856336739362513,
+      "grad_norm": 0.21924384664120367,
+      "learning_rate": 2.8963414634146343e-06,
+      "loss": 0.2772,
+      "step": 96
+    },
+    {
+      "epoch": 0.05917340247064206,
+      "grad_norm": 0.2257203912109911,
+      "learning_rate": 2.926829268292683e-06,
+      "loss": 0.2727,
+      "step": 97
+    },
+    {
+      "epoch": 0.05978343754765899,
+      "grad_norm": 0.2271960932362661,
+      "learning_rate": 2.957317073170732e-06,
+      "loss": 0.2792,
+      "step": 98
+    },
+    {
+      "epoch": 0.06039347262467592,
+      "grad_norm": 0.21768279717594183,
+      "learning_rate": 2.9878048780487808e-06,
+      "loss": 0.2804,
+      "step": 99
+    },
+    {
+      "epoch": 0.06100350770169285,
+      "grad_norm": 0.2233152083910907,
+      "learning_rate": 3.0182926829268293e-06,
+      "loss": 0.2655,
+      "step": 100
+    },
+    {
+      "epoch": 0.06161354277870978,
+      "grad_norm": 0.22662101387006744,
+      "learning_rate": 3.0487804878048782e-06,
+      "loss": 0.2703,
+      "step": 101
+    },
+    {
+      "epoch": 0.06222357785572671,
+      "grad_norm": 0.22292006274531256,
+      "learning_rate": 3.079268292682927e-06,
+      "loss": 0.2954,
+      "step": 102
+    },
+    {
+      "epoch": 0.06283361293274363,
+      "grad_norm": 0.220071896219013,
+      "learning_rate": 3.1097560975609757e-06,
+      "loss": 0.2476,
+      "step": 103
+    },
+    {
+      "epoch": 0.06344364800976056,
+      "grad_norm": 0.22535333283145692,
+      "learning_rate": 3.1402439024390246e-06,
+      "loss": 0.2795,
+      "step": 104
+    },
+    {
+      "epoch": 0.06405368308677749,
+      "grad_norm": 0.22789495334564042,
+      "learning_rate": 3.1707317073170736e-06,
+      "loss": 0.3072,
+      "step": 105
+    },
+    {
+      "epoch": 0.06466371816379442,
+      "grad_norm": 0.22845970490853684,
+      "learning_rate": 3.201219512195122e-06,
+      "loss": 0.2928,
+      "step": 106
+    },
+    {
+      "epoch": 0.06527375324081135,
+      "grad_norm": 0.22721669813586723,
+      "learning_rate": 3.231707317073171e-06,
+      "loss": 0.2701,
+      "step": 107
+    },
+    {
+      "epoch": 0.06588378831782828,
+      "grad_norm": 0.22501246073953735,
+      "learning_rate": 3.26219512195122e-06,
+      "loss": 0.295,
+      "step": 108
+    },
+    {
+      "epoch": 0.0664938233948452,
+      "grad_norm": 0.23994963496074384,
+      "learning_rate": 3.292682926829269e-06,
+      "loss": 0.3106,
+      "step": 109
+    },
+    {
+      "epoch": 0.06710385847186214,
+      "grad_norm": 0.21902655359899867,
+      "learning_rate": 3.3231707317073174e-06,
+      "loss": 0.2812,
+      "step": 110
+    },
+    {
+      "epoch": 0.06771389354887906,
+      "grad_norm": 0.22318592923675146,
+      "learning_rate": 3.3536585365853664e-06,
+      "loss": 0.2835,
+      "step": 111
+    },
+    {
+      "epoch": 0.068323928625896,
+      "grad_norm": 0.26363594205403645,
+      "learning_rate": 3.3841463414634153e-06,
+      "loss": 0.2903,
+      "step": 112
+    },
+    {
+      "epoch": 0.06893396370291292,
+      "grad_norm": 0.22377433701651386,
+      "learning_rate": 3.414634146341464e-06,
+      "loss": 0.2743,
+      "step": 113
+    },
+    {
+      "epoch": 0.06954399877992985,
+      "grad_norm": 0.23626130625058447,
+      "learning_rate": 3.4451219512195124e-06,
+      "loss": 0.299,
+      "step": 114
+    },
+    {
+      "epoch": 0.07015403385694677,
+      "grad_norm": 0.223610053073903,
+      "learning_rate": 3.475609756097561e-06,
+      "loss": 0.2555,
+      "step": 115
+    },
+    {
+      "epoch": 0.0707640689339637,
+      "grad_norm": 0.21087836508889907,
+      "learning_rate": 3.50609756097561e-06,
+      "loss": 0.2649,
+      "step": 116
+    },
+    {
+      "epoch": 0.07137410401098063,
+      "grad_norm": 0.24915773429575874,
+      "learning_rate": 3.5365853658536588e-06,
+      "loss": 0.2936,
+      "step": 117
+    },
+    {
+      "epoch": 0.07198413908799756,
+      "grad_norm": 0.24129416709675272,
+      "learning_rate": 3.5670731707317073e-06,
+      "loss": 0.2862,
+      "step": 118
+    },
+    {
+      "epoch": 0.07259417416501449,
+      "grad_norm": 0.24255820924509042,
+      "learning_rate": 3.5975609756097562e-06,
+      "loss": 0.2816,
+      "step": 119
+    },
+    {
+      "epoch": 0.07320420924203141,
+      "grad_norm": 0.22760789782296764,
+      "learning_rate": 3.628048780487805e-06,
+      "loss": 0.281,
+      "step": 120
+    },
+    {
+      "epoch": 0.07381424431904834,
+      "grad_norm": 0.2435576310663892,
+      "learning_rate": 3.6585365853658537e-06,
+      "loss": 0.29,
+      "step": 121
+    },
+    {
+      "epoch": 0.07442427939606527,
+      "grad_norm": 0.23105027146598722,
+      "learning_rate": 3.6890243902439026e-06,
+      "loss": 0.2801,
+      "step": 122
+    },
+    {
+      "epoch": 0.0750343144730822,
+      "grad_norm": 0.23364310548660652,
+      "learning_rate": 3.7195121951219516e-06,
+      "loss": 0.266,
+      "step": 123
+    },
+    {
+      "epoch": 0.07564434955009913,
+      "grad_norm": 0.24832037252957492,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 0.306,
+      "step": 124
+    },
+    {
+      "epoch": 0.07625438462711606,
+      "grad_norm": 0.2357002851720913,
+      "learning_rate": 3.780487804878049e-06,
+      "loss": 0.2933,
+      "step": 125
+    },
+    {
+      "epoch": 0.07686441970413299,
+      "grad_norm": 0.22556207921999175,
+      "learning_rate": 3.810975609756098e-06,
+      "loss": 0.2841,
+      "step": 126
+    },
+    {
+      "epoch": 0.07747445478114992,
+      "grad_norm": 0.24161879175738435,
+      "learning_rate": 3.8414634146341465e-06,
+      "loss": 0.2844,
+      "step": 127
+    },
+    {
+      "epoch": 0.07808448985816685,
+      "grad_norm": 0.23782173864569195,
+      "learning_rate": 3.8719512195121954e-06,
+      "loss": 0.285,
+      "step": 128
+    },
+    {
+      "epoch": 0.07869452493518378,
+      "grad_norm": 0.22466061236443696,
+      "learning_rate": 3.902439024390244e-06,
+      "loss": 0.2658,
+      "step": 129
+    },
+    {
+      "epoch": 0.0793045600122007,
+      "grad_norm": 0.21752730672405302,
+      "learning_rate": 3.932926829268293e-06,
+      "loss": 0.2773,
+      "step": 130
+    },
+    {
+      "epoch": 0.07991459508921762,
+      "grad_norm": 0.23637204821920585,
+      "learning_rate": 3.963414634146342e-06,
+      "loss": 0.2775,
+      "step": 131
+    },
+    {
+      "epoch": 0.08052463016623455,
+      "grad_norm": 0.22799647607474074,
+      "learning_rate": 3.99390243902439e-06,
+      "loss": 0.2636,
+      "step": 132
+    },
+    {
+      "epoch": 0.08113466524325148,
+      "grad_norm": 0.22714228884278848,
+      "learning_rate": 4.024390243902439e-06,
+      "loss": 0.2615,
+      "step": 133
+    },
+    {
+      "epoch": 0.08174470032026841,
+      "grad_norm": 0.21398203363805257,
+      "learning_rate": 4.054878048780488e-06,
+      "loss": 0.2507,
+      "step": 134
+    },
+    {
+      "epoch": 0.08235473539728534,
+      "grad_norm": 0.2264462526855402,
+      "learning_rate": 4.085365853658536e-06,
+      "loss": 0.2691,
+      "step": 135
+    },
+    {
+      "epoch": 0.08296477047430227,
+      "grad_norm": 0.23753082090613722,
+      "learning_rate": 4.115853658536585e-06,
+      "loss": 0.2824,
+      "step": 136
+    },
+    {
+      "epoch": 0.0835748055513192,
+      "grad_norm": 0.23304073294964817,
+      "learning_rate": 4.146341463414634e-06,
+      "loss": 0.2826,
+      "step": 137
+    },
+    {
+      "epoch": 0.08418484062833613,
+      "grad_norm": 0.23595581075210828,
+      "learning_rate": 4.176829268292683e-06,
+      "loss": 0.2916,
+      "step": 138
+    },
+    {
+      "epoch": 0.08479487570535306,
+      "grad_norm": 0.22041886879240966,
+      "learning_rate": 4.207317073170732e-06,
+      "loss": 0.2369,
+      "step": 139
+    },
+    {
+      "epoch": 0.08540491078236999,
+      "grad_norm": 0.24206221610814158,
+      "learning_rate": 4.237804878048781e-06,
+      "loss": 0.2821,
+      "step": 140
+    },
+    {
+      "epoch": 0.08601494585938692,
+      "grad_norm": 0.2393670152036277,
+      "learning_rate": 4.268292682926829e-06,
+      "loss": 0.2791,
+      "step": 141
+    },
+    {
+      "epoch": 0.08662498093640385,
+      "grad_norm": 0.23189352334620775,
+      "learning_rate": 4.298780487804878e-06,
+      "loss": 0.2919,
+      "step": 142
+    },
+    {
+      "epoch": 0.08723501601342078,
+      "grad_norm": 0.242567561823643,
+      "learning_rate": 4.329268292682927e-06,
+      "loss": 0.2831,
+      "step": 143
+    },
+    {
+      "epoch": 0.0878450510904377,
+      "grad_norm": 0.23058885789044348,
+      "learning_rate": 4.359756097560976e-06,
+      "loss": 0.269,
+      "step": 144
+    },
+    {
+      "epoch": 0.08845508616745462,
+      "grad_norm": 0.22276270950122776,
+      "learning_rate": 4.390243902439025e-06,
+      "loss": 0.2517,
+      "step": 145
+    },
+    {
+      "epoch": 0.08906512124447155,
+      "grad_norm": 0.23742463465017807,
+      "learning_rate": 4.420731707317074e-06,
+      "loss": 0.284,
+      "step": 146
+    },
+    {
+      "epoch": 0.08967515632148848,
+      "grad_norm": 0.2331170754231291,
+      "learning_rate": 4.451219512195122e-06,
+      "loss": 0.2624,
+      "step": 147
+    },
+    {
+      "epoch": 0.09028519139850541,
+      "grad_norm": 0.2308037721395711,
+      "learning_rate": 4.481707317073171e-06,
+      "loss": 0.2539,
+      "step": 148
+    },
+    {
+      "epoch": 0.09089522647552234,
+      "grad_norm": 0.22849284994691954,
+      "learning_rate": 4.51219512195122e-06,
+      "loss": 0.2826,
+      "step": 149
+    },
+    {
+      "epoch": 0.09150526155253927,
+      "grad_norm": 0.22290125661574434,
+      "learning_rate": 4.542682926829269e-06,
+      "loss": 0.2834,
+      "step": 150
+    },
+    {
+      "epoch": 0.0921152966295562,
+      "grad_norm": 0.22294802055510163,
+      "learning_rate": 4.573170731707318e-06,
+      "loss": 0.2664,
+      "step": 151
+    },
+    {
+      "epoch": 0.09272533170657313,
+      "grad_norm": 0.22216693279188238,
+      "learning_rate": 4.603658536585367e-06,
+      "loss": 0.2722,
+      "step": 152
+    },
+    {
+      "epoch": 0.09333536678359006,
+      "grad_norm": 0.22847119099300714,
+      "learning_rate": 4.634146341463416e-06,
+      "loss": 0.3077,
+      "step": 153
+    },
+    {
+      "epoch": 0.09394540186060699,
+      "grad_norm": 0.22825237894656306,
+      "learning_rate": 4.664634146341464e-06,
+      "loss": 0.279,
+      "step": 154
+    },
+    {
+      "epoch": 0.09455543693762392,
+      "grad_norm": 0.2265738168314992,
+      "learning_rate": 4.695121951219513e-06,
+      "loss": 0.2743,
+      "step": 155
+    },
+    {
+      "epoch": 0.09516547201464085,
+      "grad_norm": 0.2379254364113949,
+      "learning_rate": 4.725609756097561e-06,
+      "loss": 0.2821,
+      "step": 156
+    },
+    {
+      "epoch": 0.09577550709165777,
+      "grad_norm": 0.2533303804327538,
+      "learning_rate": 4.75609756097561e-06,
+      "loss": 0.2812,
+      "step": 157
+    },
+    {
+      "epoch": 0.0963855421686747,
+      "grad_norm": 0.21582395227434673,
+      "learning_rate": 4.786585365853659e-06,
+      "loss": 0.2557,
+      "step": 158
+    },
+    {
+      "epoch": 0.09699557724569163,
+      "grad_norm": 0.23974503879223094,
+      "learning_rate": 4.817073170731708e-06,
+      "loss": 0.2842,
+      "step": 159
+    },
+    {
+      "epoch": 0.09760561232270856,
+      "grad_norm": 0.23764839363381726,
+      "learning_rate": 4.8475609756097565e-06,
+      "loss": 0.2903,
+      "step": 160
+    },
+    {
+      "epoch": 0.09821564739972548,
+      "grad_norm": 0.2340765128011826,
+      "learning_rate": 4.8780487804878055e-06,
+      "loss": 0.2749,
+      "step": 161
+    },
+    {
+      "epoch": 0.09882568247674241,
+      "grad_norm": 0.21460415213463546,
+      "learning_rate": 4.908536585365854e-06,
+      "loss": 0.2611,
+      "step": 162
+    },
+    {
+      "epoch": 0.09943571755375934,
+      "grad_norm": 0.24181561044684743,
+      "learning_rate": 4.9390243902439025e-06,
+      "loss": 0.2844,
+      "step": 163
+    },
+    {
+      "epoch": 0.10004575263077627,
+      "grad_norm": 0.2405356599063858,
+      "learning_rate": 4.9695121951219515e-06,
+      "loss": 0.2965,
+      "step": 164
+    },
+    {
+      "epoch": 0.1006557877077932,
+      "grad_norm": 0.21661402310556976,
+      "learning_rate": 5e-06,
+      "loss": 0.2464,
+      "step": 165
+    },
+    {
+      "epoch": 0.10126582278481013,
+      "grad_norm": 0.2262455083804811,
+      "learning_rate": 5.030487804878049e-06,
+      "loss": 0.273,
+      "step": 166
+    },
+    {
+      "epoch": 0.10187585786182705,
+      "grad_norm": 0.2365786862729776,
+      "learning_rate": 5.060975609756098e-06,
+      "loss": 0.3042,
+      "step": 167
+    },
+    {
+      "epoch": 0.10248589293884398,
+      "grad_norm": 0.23673959129795522,
+      "learning_rate": 5.091463414634147e-06,
+      "loss": 0.2742,
+      "step": 168
+    },
+    {
+      "epoch": 0.10309592801586091,
+      "grad_norm": 0.25419425400530005,
+      "learning_rate": 5.121951219512195e-06,
+      "loss": 0.3005,
+      "step": 169
+    },
+    {
+      "epoch": 0.10370596309287784,
+      "grad_norm": 0.2419749655946345,
+      "learning_rate": 5.152439024390244e-06,
+      "loss": 0.2885,
+      "step": 170
+    },
+    {
+      "epoch": 0.10431599816989477,
+      "grad_norm": 0.21393573151462877,
+      "learning_rate": 5.182926829268293e-06,
+      "loss": 0.264,
+      "step": 171
+    },
+    {
+      "epoch": 0.1049260332469117,
+      "grad_norm": 0.2362091062195042,
+      "learning_rate": 5.213414634146342e-06,
+      "loss": 0.2585,
+      "step": 172
+    },
+    {
+      "epoch": 0.10553606832392863,
+      "grad_norm": 0.2704586811251837,
+      "learning_rate": 5.243902439024391e-06,
+      "loss": 0.2778,
+      "step": 173
+    },
+    {
+      "epoch": 0.10614610340094556,
+      "grad_norm": 0.2307320156865811,
+      "learning_rate": 5.27439024390244e-06,
+      "loss": 0.2724,
+      "step": 174
+    },
+    {
+      "epoch": 0.10675613847796249,
+      "grad_norm": 0.24480944164566806,
+      "learning_rate": 5.304878048780488e-06,
+      "loss": 0.2793,
+      "step": 175
+    },
+    {
+      "epoch": 0.1073661735549794,
+      "grad_norm": 0.24120432626907268,
+      "learning_rate": 5.335365853658537e-06,
+      "loss": 0.2594,
+      "step": 176
+    },
+    {
+      "epoch": 0.10797620863199633,
+      "grad_norm": 0.24750812136821748,
+      "learning_rate": 5.365853658536586e-06,
+      "loss": 0.2827,
+      "step": 177
+    },
+    {
+      "epoch": 0.10858624370901326,
+      "grad_norm": 0.24013243345908472,
+      "learning_rate": 5.396341463414635e-06,
+      "loss": 0.2573,
+      "step": 178
+    },
+    {
+      "epoch": 0.1091962787860302,
+      "grad_norm": 0.2609299332154205,
+      "learning_rate": 5.426829268292684e-06,
+      "loss": 0.2874,
+      "step": 179
+    },
+    {
+      "epoch": 0.10980631386304712,
+      "grad_norm": 0.26281954073730773,
+      "learning_rate": 5.457317073170733e-06,
+      "loss": 0.3054,
+      "step": 180
+    },
+    {
+      "epoch": 0.11041634894006405,
+      "grad_norm": 0.24458036151237894,
+      "learning_rate": 5.487804878048781e-06,
+      "loss": 0.2805,
+      "step": 181
+    },
+    {
+      "epoch": 0.11102638401708098,
+      "grad_norm": 0.24512128267371772,
+      "learning_rate": 5.51829268292683e-06,
+      "loss": 0.2914,
+      "step": 182
+    },
+    {
+      "epoch": 0.11163641909409791,
+      "grad_norm": 0.23669169587728295,
+      "learning_rate": 5.548780487804879e-06,
+      "loss": 0.2786,
+      "step": 183
+    },
+    {
+      "epoch": 0.11224645417111484,
+      "grad_norm": 0.24697726917520454,
+      "learning_rate": 5.579268292682928e-06,
+      "loss": 0.2767,
+      "step": 184
+    },
+    {
+      "epoch": 0.11285648924813177,
+      "grad_norm": 0.249714640479248,
+      "learning_rate": 5.609756097560977e-06,
+      "loss": 0.2685,
+      "step": 185
+    },
+    {
+      "epoch": 0.1134665243251487,
+      "grad_norm": 0.24841021945031513,
+      "learning_rate": 5.640243902439024e-06,
+      "loss": 0.2649,
+      "step": 186
+    },
+    {
+      "epoch": 0.11407655940216563,
+      "grad_norm": 0.23859920700579967,
+      "learning_rate": 5.670731707317073e-06,
+      "loss": 0.2858,
+      "step": 187
+    },
+    {
+      "epoch": 0.11468659447918256,
+      "grad_norm": 0.2584753235599307,
+      "learning_rate": 5.701219512195122e-06,
+      "loss": 0.2747,
+      "step": 188
+    },
+    {
+      "epoch": 0.11529662955619949,
+      "grad_norm": 0.23156317539426982,
+      "learning_rate": 5.731707317073171e-06,
+      "loss": 0.2529,
+      "step": 189
+    },
+    {
+      "epoch": 0.11590666463321642,
+      "grad_norm": 0.22687147729532126,
+      "learning_rate": 5.76219512195122e-06,
+      "loss": 0.2826,
+      "step": 190
+    },
+    {
+      "epoch": 0.11651669971023333,
+      "grad_norm": 0.25517296147028024,
+      "learning_rate": 5.792682926829269e-06,
+      "loss": 0.2559,
+      "step": 191
+    },
+    {
+      "epoch": 0.11712673478725026,
+      "grad_norm": 0.2473059947585708,
+      "learning_rate": 5.823170731707317e-06,
+      "loss": 0.2617,
+      "step": 192
+    },
+    {
+      "epoch": 0.11773676986426719,
+      "grad_norm": 0.2475596917448232,
+      "learning_rate": 5.853658536585366e-06,
+      "loss": 0.2757,
+      "step": 193
+    },
+    {
+      "epoch": 0.11834680494128412,
+      "grad_norm": 0.24485621157389584,
+      "learning_rate": 5.884146341463415e-06,
+      "loss": 0.2801,
+      "step": 194
+    },
+    {
+      "epoch": 0.11895684001830105,
+      "grad_norm": 0.2407535109141067,
+      "learning_rate": 5.914634146341464e-06,
+      "loss": 0.2541,
+      "step": 195
+    },
+    {
+      "epoch": 0.11956687509531798,
+      "grad_norm": 0.29475882116862984,
+      "learning_rate": 5.9451219512195126e-06,
+      "loss": 0.2672,
+      "step": 196
+    },
+    {
+      "epoch": 0.12017691017233491,
+      "grad_norm": 0.24490394924599188,
+      "learning_rate": 5.9756097560975615e-06,
+      "loss": 0.265,
+      "step": 197
+    },
+    {
+      "epoch": 0.12078694524935184,
+      "grad_norm": 0.24048637934439834,
+      "learning_rate": 6.0060975609756104e-06,
+      "loss": 0.2616,
+      "step": 198
+    },
+    {
+      "epoch": 0.12139698032636877,
+      "grad_norm": 0.2449235795372263,
+      "learning_rate": 6.0365853658536585e-06,
+      "loss": 0.2771,
+      "step": 199
+    },
+    {
+      "epoch": 0.1220070154033857,
+      "grad_norm": 0.24010278455894474,
+      "learning_rate": 6.0670731707317075e-06,
+      "loss": 0.2487,
+      "step": 200
+    },
+    {
+      "epoch": 0.12261705048040263,
+      "grad_norm": 0.2501744987758223,
+      "learning_rate": 6.0975609756097564e-06,
+      "loss": 0.2769,
+      "step": 201
+    },
+    {
+      "epoch": 0.12322708555741956,
+      "grad_norm": 0.24813908759409767,
+      "learning_rate": 6.128048780487805e-06,
+      "loss": 0.2705,
+      "step": 202
+    },
+    {
+      "epoch": 0.12383712063443648,
+      "grad_norm": 0.24533160040379387,
+      "learning_rate": 6.158536585365854e-06,
+      "loss": 0.2815,
+      "step": 203
+    },
+    {
+      "epoch": 0.12444715571145341,
+      "grad_norm": 0.2437938947166487,
+      "learning_rate": 6.189024390243903e-06,
+      "loss": 0.2521,
+      "step": 204
+    },
+    {
+      "epoch": 0.12505719078847033,
+      "grad_norm": 0.26225596287555486,
+      "learning_rate": 6.219512195121951e-06,
+      "loss": 0.2793,
+      "step": 205
+    },
+    {
+      "epoch": 0.12566722586548726,
+      "grad_norm": 0.2617035107047438,
+      "learning_rate": 6.25e-06,
+      "loss": 0.2868,
+      "step": 206
+    },
+    {
+      "epoch": 0.1262772609425042,
+      "grad_norm": 0.22489028420899687,
+      "learning_rate": 6.280487804878049e-06,
+      "loss": 0.2502,
+      "step": 207
+    },
+    {
+      "epoch": 0.12688729601952112,
+      "grad_norm": 0.23908878727484426,
+      "learning_rate": 6.310975609756098e-06,
+      "loss": 0.2842,
+      "step": 208
+    },
+    {
+      "epoch": 0.12749733109653805,
+      "grad_norm": 0.24229667129434546,
+      "learning_rate": 6.341463414634147e-06,
+      "loss": 0.2729,
+      "step": 209
+    },
+    {
+      "epoch": 0.12810736617355498,
+      "grad_norm": 0.23997663988223025,
+      "learning_rate": 6.371951219512196e-06,
+      "loss": 0.2469,
+      "step": 210
+    },
+    {
+      "epoch": 0.1287174012505719,
+      "grad_norm": 0.24052895148755218,
+      "learning_rate": 6.402439024390244e-06,
+      "loss": 0.2767,
+      "step": 211
+    },
+    {
+      "epoch": 0.12932743632758884,
+      "grad_norm": 0.24249027732334763,
+      "learning_rate": 6.432926829268293e-06,
+      "loss": 0.2458,
+      "step": 212
+    },
+    {
+      "epoch": 0.12993747140460576,
+      "grad_norm": 0.25137563661292134,
+      "learning_rate": 6.463414634146342e-06,
+      "loss": 0.273,
+      "step": 213
+    },
+    {
+      "epoch": 0.1305475064816227,
+      "grad_norm": 0.2587508767905881,
+      "learning_rate": 6.493902439024391e-06,
+      "loss": 0.2677,
+      "step": 214
+    },
+    {
+      "epoch": 0.13115754155863962,
+      "grad_norm": 0.2436596949895466,
+      "learning_rate": 6.52439024390244e-06,
+      "loss": 0.2576,
+      "step": 215
+    },
+    {
+      "epoch": 0.13176757663565655,
+      "grad_norm": 0.24764204486776847,
+      "learning_rate": 6.554878048780489e-06,
+      "loss": 0.2621,
+      "step": 216
+    },
+    {
+      "epoch": 0.13237761171267348,
+      "grad_norm": 0.25726179008925426,
+      "learning_rate": 6.585365853658538e-06,
+      "loss": 0.2757,
+      "step": 217
+    },
+    {
+      "epoch": 0.1329876467896904,
+      "grad_norm": 0.2569737346876388,
+      "learning_rate": 6.615853658536586e-06,
+      "loss": 0.2686,
+      "step": 218
+    },
+    {
+      "epoch": 0.13359768186670734,
+      "grad_norm": 0.24488929677587262,
+      "learning_rate": 6.646341463414635e-06,
+      "loss": 0.2703,
+      "step": 219
+    },
+    {
+      "epoch": 0.13420771694372427,
+      "grad_norm": 0.2625226665018392,
+      "learning_rate": 6.676829268292684e-06,
+      "loss": 0.2536,
+      "step": 220
+    },
+    {
+      "epoch": 0.1348177520207412,
+      "grad_norm": 0.32829586478398526,
+      "learning_rate": 6.707317073170733e-06,
+      "loss": 0.2851,
+      "step": 221
+    },
+    {
+      "epoch": 0.13542778709775813,
+      "grad_norm": 0.24381114982471824,
+      "learning_rate": 6.737804878048782e-06,
+      "loss": 0.2725,
+      "step": 222
+    },
+    {
+      "epoch": 0.13603782217477506,
+      "grad_norm": 0.2537539468384488,
+      "learning_rate": 6.768292682926831e-06,
+      "loss": 0.2702,
+      "step": 223
+    },
+    {
+      "epoch": 0.136647857251792,
+      "grad_norm": 0.24029077547019123,
+      "learning_rate": 6.798780487804879e-06,
+      "loss": 0.2506,
+      "step": 224
+    },
+    {
+      "epoch": 0.13725789232880892,
+      "grad_norm": 0.25072886281019574,
+      "learning_rate": 6.829268292682928e-06,
+      "loss": 0.2754,
+      "step": 225
+    },
+    {
+      "epoch": 0.13786792740582585,
+      "grad_norm": 0.2622525846487408,
+      "learning_rate": 6.859756097560977e-06,
+      "loss": 0.275,
+      "step": 226
+    },
+    {
+      "epoch": 0.13847796248284278,
+      "grad_norm": 0.27373368686851934,
+      "learning_rate": 6.890243902439025e-06,
+      "loss": 0.2584,
+      "step": 227
+    },
+    {
+      "epoch": 0.1390879975598597,
+      "grad_norm": 0.28138028268480203,
+      "learning_rate": 6.920731707317073e-06,
+      "loss": 0.273,
+      "step": 228
+    },
+    {
+      "epoch": 0.1396980326368766,
+      "grad_norm": 0.2506330599580721,
+      "learning_rate": 6.951219512195122e-06,
+      "loss": 0.281,
+      "step": 229
+    },
+    {
+      "epoch": 0.14030806771389354,
+      "grad_norm": 0.26945292402465437,
+      "learning_rate": 6.981707317073171e-06,
+      "loss": 0.2584,
+      "step": 230
+    },
+    {
+      "epoch": 0.14091810279091047,
+      "grad_norm": 0.25764293041454384,
+      "learning_rate": 7.01219512195122e-06,
+      "loss": 0.2693,
+      "step": 231
+    },
+    {
+      "epoch": 0.1415281378679274,
+      "grad_norm": 0.24772244759620793,
+      "learning_rate": 7.042682926829269e-06,
+      "loss": 0.2831,
+      "step": 232
+    },
+    {
+      "epoch": 0.14213817294494432,
+      "grad_norm": 0.24429762825269066,
+      "learning_rate": 7.0731707317073175e-06,
+      "loss": 0.2669,
+      "step": 233
+    },
+    {
+      "epoch": 0.14274820802196125,
+      "grad_norm": 0.2341275505054125,
+      "learning_rate": 7.1036585365853665e-06,
+      "loss": 0.2627,
+      "step": 234
+    },
+    {
+      "epoch": 0.14335824309897818,
+      "grad_norm": 0.2516323459329928,
+      "learning_rate": 7.1341463414634146e-06,
+      "loss": 0.2486,
+      "step": 235
+    },
+    {
+      "epoch": 0.1439682781759951,
+      "grad_norm": 0.24792052063072995,
+      "learning_rate": 7.1646341463414635e-06,
+      "loss": 0.2571,
+      "step": 236
+    },
+    {
+      "epoch": 0.14457831325301204,
+      "grad_norm": 0.24838425917856113,
+      "learning_rate": 7.1951219512195125e-06,
+      "loss": 0.2565,
+      "step": 237
+    },
+    {
+      "epoch": 0.14518834833002897,
+      "grad_norm": 0.2558154886637839,
+      "learning_rate": 7.225609756097561e-06,
+      "loss": 0.2602,
+      "step": 238
+    },
+    {
+      "epoch": 0.1457983834070459,
+      "grad_norm": 0.24095801169637035,
+      "learning_rate": 7.25609756097561e-06,
+      "loss": 0.276,
+      "step": 239
+    },
+    {
+      "epoch": 0.14640841848406283,
+      "grad_norm": 0.24456191278684733,
+      "learning_rate": 7.286585365853659e-06,
+      "loss": 0.2612,
+      "step": 240
+    },
+    {
+      "epoch": 0.14701845356107976,
+      "grad_norm": 0.24059500548054316,
+      "learning_rate": 7.317073170731707e-06,
+      "loss": 0.2528,
+      "step": 241
+    },
+    {
+      "epoch": 0.1476284886380967,
+      "grad_norm": 0.24031870127070679,
+      "learning_rate": 7.347560975609756e-06,
+      "loss": 0.257,
+      "step": 242
+    },
+    {
+      "epoch": 0.14823852371511362,
+      "grad_norm": 0.2928644520428681,
+      "learning_rate": 7.378048780487805e-06,
+      "loss": 0.2973,
+      "step": 243
+    },
+    {
+      "epoch": 0.14884855879213055,
+      "grad_norm": 0.2436600942196456,
+      "learning_rate": 7.408536585365854e-06,
+      "loss": 0.2527,
+      "step": 244
+    },
+    {
+      "epoch": 0.14945859386914748,
+      "grad_norm": 0.24910229884619056,
+      "learning_rate": 7.439024390243903e-06,
+      "loss": 0.2573,
+      "step": 245
+    },
+    {
+      "epoch": 0.1500686289461644,
+      "grad_norm": 0.257984994933436,
+      "learning_rate": 7.469512195121952e-06,
+      "loss": 0.2854,
+      "step": 246
+    },
+    {
+      "epoch": 0.15067866402318134,
+      "grad_norm": 0.26224460707949426,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2462,
+      "step": 247
+    },
+    {
+      "epoch": 0.15128869910019827,
+      "grad_norm": 0.23911637788829992,
+      "learning_rate": 7.530487804878049e-06,
+      "loss": 0.258,
+      "step": 248
+    },
+    {
+      "epoch": 0.1518987341772152,
+      "grad_norm": 0.26106166538596853,
+      "learning_rate": 7.560975609756098e-06,
+      "loss": 0.2395,
+      "step": 249
+    },
+    {
+      "epoch": 0.15250876925423212,
+      "grad_norm": 0.2628905138359254,
+      "learning_rate": 7.591463414634147e-06,
+      "loss": 0.2557,
+      "step": 250
+    },
+    {
+      "epoch": 0.15311880433124905,
+      "grad_norm": 0.24772668368639703,
+      "learning_rate": 7.621951219512196e-06,
+      "loss": 0.2612,
+      "step": 251
+    },
+    {
+      "epoch": 0.15372883940826598,
+      "grad_norm": 0.2614036863610298,
+      "learning_rate": 7.652439024390244e-06,
+      "loss": 0.2418,
+      "step": 252
+    },
+    {
+      "epoch": 0.1543388744852829,
+      "grad_norm": 0.2533666402036696,
+      "learning_rate": 7.682926829268293e-06,
+      "loss": 0.2676,
+      "step": 253
+    },
+    {
+      "epoch": 0.15494890956229984,
+      "grad_norm": 0.2646349439683426,
+      "learning_rate": 7.713414634146342e-06,
+      "loss": 0.2807,
+      "step": 254
+    },
+    {
+      "epoch": 0.15555894463931677,
+      "grad_norm": 0.25217251940804086,
+      "learning_rate": 7.743902439024391e-06,
+      "loss": 0.2648,
+      "step": 255
+    },
+    {
+      "epoch": 0.1561689797163337,
+      "grad_norm": 0.2934490177685301,
+      "learning_rate": 7.77439024390244e-06,
+      "loss": 0.242,
+      "step": 256
+    },
+    {
+      "epoch": 0.15677901479335063,
+      "grad_norm": 0.23774700056796438,
+      "learning_rate": 7.804878048780489e-06,
+      "loss": 0.2628,
+      "step": 257
+    },
+    {
+      "epoch": 0.15738904987036756,
+      "grad_norm": 0.2794132704089381,
+      "learning_rate": 7.835365853658538e-06,
+      "loss": 0.2794,
+      "step": 258
+    },
+    {
+      "epoch": 0.1579990849473845,
+      "grad_norm": 0.24213388227268773,
+      "learning_rate": 7.865853658536587e-06,
+      "loss": 0.2647,
+      "step": 259
+    },
+    {
+      "epoch": 0.1586091200244014,
+      "grad_norm": 0.25082026136954216,
+      "learning_rate": 7.896341463414636e-06,
+      "loss": 0.2637,
+      "step": 260
+    },
+    {
+      "epoch": 0.15921915510141832,
+      "grad_norm": 0.2431946921818769,
+      "learning_rate": 7.926829268292685e-06,
+      "loss": 0.2626,
+      "step": 261
+    },
+    {
+      "epoch": 0.15982919017843525,
+      "grad_norm": 0.2442925275620386,
+      "learning_rate": 7.957317073170733e-06,
+      "loss": 0.2614,
+      "step": 262
+    },
+    {
+      "epoch": 0.16043922525545218,
+      "grad_norm": 0.258570314023015,
+      "learning_rate": 7.98780487804878e-06,
+      "loss": 0.2598,
+      "step": 263
+    },
+    {
+      "epoch": 0.1610492603324691,
+      "grad_norm": 0.24980544348638753,
+      "learning_rate": 8.01829268292683e-06,
+      "loss": 0.2682,
+      "step": 264
+    },
+    {
+      "epoch": 0.16165929540948604,
+      "grad_norm": 0.22851755945890959,
+      "learning_rate": 8.048780487804879e-06,
+      "loss": 0.2383,
+      "step": 265
+    },
+    {
+      "epoch": 0.16226933048650297,
+      "grad_norm": 0.2742225825226428,
+      "learning_rate": 8.079268292682928e-06,
+      "loss": 0.2613,
+      "step": 266
+    },
+    {
+      "epoch": 0.1628793655635199,
+      "grad_norm": 0.2363713257999274,
+      "learning_rate": 8.109756097560977e-06,
+      "loss": 0.2444,
+      "step": 267
+    },
+    {
+      "epoch": 0.16348940064053682,
+      "grad_norm": 0.28248352582246355,
+      "learning_rate": 8.140243902439024e-06,
+      "loss": 0.2696,
+      "step": 268
+    },
+    {
+      "epoch": 0.16409943571755375,
+      "grad_norm": 0.25728717998996453,
+      "learning_rate": 8.170731707317073e-06,
+      "loss": 0.2773,
+      "step": 269
+    },
+    {
+      "epoch": 0.16470947079457068,
+      "grad_norm": 0.3007522019243012,
+      "learning_rate": 8.201219512195122e-06,
+      "loss": 0.2508,
+      "step": 270
+    },
+    {
+      "epoch": 0.1653195058715876,
+      "grad_norm": 0.2640301074207759,
+      "learning_rate": 8.23170731707317e-06,
+      "loss": 0.2726,
+      "step": 271
+    },
+    {
+      "epoch": 0.16592954094860454,
+      "grad_norm": 0.27589684620981636,
+      "learning_rate": 8.26219512195122e-06,
+      "loss": 0.2594,
+      "step": 272
+    },
+    {
+      "epoch": 0.16653957602562147,
+      "grad_norm": 0.2603256028043959,
+      "learning_rate": 8.292682926829268e-06,
+      "loss": 0.2548,
+      "step": 273
+    },
+    {
+      "epoch": 0.1671496111026384,
+      "grad_norm": 0.2543287160421388,
+      "learning_rate": 8.323170731707317e-06,
+      "loss": 0.2639,
+      "step": 274
+    },
+    {
+      "epoch": 0.16775964617965533,
+      "grad_norm": 0.2663572684486127,
+      "learning_rate": 8.353658536585366e-06,
+      "loss": 0.2734,
+      "step": 275
+    },
+    {
+      "epoch": 0.16836968125667226,
+      "grad_norm": 0.24382846053208523,
+      "learning_rate": 8.384146341463415e-06,
+      "loss": 0.2567,
+      "step": 276
+    },
+    {
+      "epoch": 0.1689797163336892,
+      "grad_norm": 0.31604403568590417,
+      "learning_rate": 8.414634146341464e-06,
+      "loss": 0.2672,
+      "step": 277
+    },
+    {
+      "epoch": 0.16958975141070612,
+      "grad_norm": 0.27720443335337075,
+      "learning_rate": 8.445121951219513e-06,
+      "loss": 0.2584,
+      "step": 278
+    },
+    {
+      "epoch": 0.17019978648772305,
+      "grad_norm": 0.24158731216102847,
+      "learning_rate": 8.475609756097562e-06,
+      "loss": 0.2545,
+      "step": 279
+    },
+    {
+      "epoch": 0.17080982156473998,
+      "grad_norm": 0.26559843371987457,
+      "learning_rate": 8.506097560975611e-06,
+      "loss": 0.2686,
+      "step": 280
+    },
+    {
+      "epoch": 0.1714198566417569,
+      "grad_norm": 0.25167456958331147,
+      "learning_rate": 8.536585365853658e-06,
+      "loss": 0.2669,
+      "step": 281
+    },
+    {
+      "epoch": 0.17202989171877384,
+      "grad_norm": 0.24990090478397206,
+      "learning_rate": 8.567073170731707e-06,
+      "loss": 0.252,
+      "step": 282
+    },
+    {
+      "epoch": 0.17263992679579077,
+      "grad_norm": 0.26565282242517324,
+      "learning_rate": 8.597560975609756e-06,
+      "loss": 0.2612,
+      "step": 283
+    },
+    {
+      "epoch": 0.1732499618728077,
+      "grad_norm": 0.2612900682088091,
+      "learning_rate": 8.628048780487805e-06,
+      "loss": 0.2693,
+      "step": 284
+    },
+    {
+      "epoch": 0.17385999694982462,
+      "grad_norm": 0.2542685727592583,
+      "learning_rate": 8.658536585365854e-06,
+      "loss": 0.2497,
+      "step": 285
+    },
+    {
+      "epoch": 0.17447003202684155,
+      "grad_norm": 0.2581637078007705,
+      "learning_rate": 8.689024390243903e-06,
+      "loss": 0.2625,
+      "step": 286
+    },
+    {
+      "epoch": 0.17508006710385848,
+      "grad_norm": 0.26869785766327753,
+      "learning_rate": 8.719512195121952e-06,
+      "loss": 0.2854,
+      "step": 287
+    },
+    {
+      "epoch": 0.1756901021808754,
+      "grad_norm": 0.2522640376457165,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 0.2517,
+      "step": 288
+    },
+    {
+      "epoch": 0.17630013725789234,
+      "grad_norm": 0.2514912970501586,
+      "learning_rate": 8.78048780487805e-06,
+      "loss": 0.2643,
+      "step": 289
+    },
+    {
+      "epoch": 0.17691017233490924,
+      "grad_norm": 0.2670649726886704,
+      "learning_rate": 8.810975609756099e-06,
+      "loss": 0.2691,
+      "step": 290
+    },
+    {
+      "epoch": 0.17752020741192617,
+      "grad_norm": 0.2532280359730121,
+      "learning_rate": 8.841463414634148e-06,
+      "loss": 0.2766,
+      "step": 291
+    },
+    {
+      "epoch": 0.1781302424889431,
+      "grad_norm": 0.23982871280272544,
+      "learning_rate": 8.871951219512197e-06,
+      "loss": 0.2623,
+      "step": 292
+    },
+    {
+      "epoch": 0.17874027756596003,
+      "grad_norm": 0.22291820269074192,
+      "learning_rate": 8.902439024390244e-06,
+      "loss": 0.2542,
+      "step": 293
+    },
+    {
+      "epoch": 0.17935031264297696,
+      "grad_norm": 0.24373539859994214,
+      "learning_rate": 8.932926829268293e-06,
+      "loss": 0.275,
+      "step": 294
+    },
+    {
+      "epoch": 0.1799603477199939,
+      "grad_norm": 0.251820372702038,
+      "learning_rate": 8.963414634146342e-06,
+      "loss": 0.2699,
+      "step": 295
+    },
+    {
+      "epoch": 0.18057038279701082,
+      "grad_norm": 0.2341623068040451,
+      "learning_rate": 8.99390243902439e-06,
+      "loss": 0.2503,
+      "step": 296
+    },
+    {
+      "epoch": 0.18118041787402775,
+      "grad_norm": 0.23554175402707625,
+      "learning_rate": 9.02439024390244e-06,
+      "loss": 0.2423,
+      "step": 297
+    },
+    {
+      "epoch": 0.18179045295104468,
+      "grad_norm": 0.24651028549190948,
+      "learning_rate": 9.054878048780489e-06,
+      "loss": 0.2501,
+      "step": 298
+    },
+    {
+      "epoch": 0.1824004880280616,
+      "grad_norm": 0.23298368631390087,
+      "learning_rate": 9.085365853658538e-06,
+      "loss": 0.2486,
+      "step": 299
+    },
+    {
+      "epoch": 0.18301052310507854,
+      "grad_norm": 0.25387542394056234,
+      "learning_rate": 9.115853658536587e-06,
+      "loss": 0.2554,
+      "step": 300
+    },
+    {
+      "epoch": 0.18362055818209547,
+      "grad_norm": 0.2743582744551358,
+      "learning_rate": 9.146341463414635e-06,
+      "loss": 0.3009,
+      "step": 301
+    },
+    {
+      "epoch": 0.1842305932591124,
+      "grad_norm": 0.23267894105241863,
+      "learning_rate": 9.176829268292684e-06,
+      "loss": 0.2578,
+      "step": 302
+    },
+    {
+      "epoch": 0.18484062833612933,
+      "grad_norm": 0.23211732643003208,
+      "learning_rate": 9.207317073170733e-06,
+      "loss": 0.245,
+      "step": 303
+    },
+    {
+      "epoch": 0.18545066341314626,
+      "grad_norm": 0.27622864545287684,
+      "learning_rate": 9.237804878048782e-06,
+      "loss": 0.2789,
+      "step": 304
+    },
+    {
+      "epoch": 0.18606069849016318,
+      "grad_norm": 0.24220813262242386,
+      "learning_rate": 9.268292682926831e-06,
+      "loss": 0.2517,
+      "step": 305
+    },
+    {
+      "epoch": 0.1866707335671801,
+      "grad_norm": 0.24023476810215502,
+      "learning_rate": 9.298780487804879e-06,
+      "loss": 0.2671,
+      "step": 306
+    },
+    {
+      "epoch": 0.18728076864419704,
+      "grad_norm": 0.25961081713489753,
+      "learning_rate": 9.329268292682927e-06,
+      "loss": 0.2639,
+      "step": 307
+    },
+    {
+      "epoch": 0.18789080372121397,
+      "grad_norm": 0.2561706447722481,
+      "learning_rate": 9.359756097560976e-06,
+      "loss": 0.264,
+      "step": 308
+    },
+    {
+      "epoch": 0.1885008387982309,
+      "grad_norm": 0.24437473491122833,
+      "learning_rate": 9.390243902439025e-06,
+      "loss": 0.2604,
+      "step": 309
+    },
+    {
+      "epoch": 0.18911087387524783,
+      "grad_norm": 0.2582368491714471,
+      "learning_rate": 9.420731707317073e-06,
+      "loss": 0.2438,
+      "step": 310
+    },
+    {
+      "epoch": 0.18972090895226476,
+      "grad_norm": 0.24897032920997206,
+      "learning_rate": 9.451219512195122e-06,
+      "loss": 0.2579,
+      "step": 311
+    },
+    {
+      "epoch": 0.1903309440292817,
+      "grad_norm": 0.26649069100002015,
+      "learning_rate": 9.48170731707317e-06,
+      "loss": 0.2669,
+      "step": 312
+    },
+    {
+      "epoch": 0.19094097910629862,
+      "grad_norm": 0.2648892505808558,
+      "learning_rate": 9.51219512195122e-06,
+      "loss": 0.2696,
+      "step": 313
+    },
+    {
+      "epoch": 0.19155101418331555,
+      "grad_norm": 0.23334740105373933,
+      "learning_rate": 9.542682926829268e-06,
+      "loss": 0.2448,
+      "step": 314
+    },
+    {
+      "epoch": 0.19216104926033248,
+      "grad_norm": 0.246437572595513,
+      "learning_rate": 9.573170731707317e-06,
+      "loss": 0.2495,
+      "step": 315
+    },
+    {
+      "epoch": 0.1927710843373494,
+      "grad_norm": 0.2544520516547441,
+      "learning_rate": 9.603658536585366e-06,
+      "loss": 0.2731,
+      "step": 316
+    },
+    {
+      "epoch": 0.19338111941436634,
+      "grad_norm": 0.2588261171385072,
+      "learning_rate": 9.634146341463415e-06,
+      "loss": 0.2502,
+      "step": 317
+    },
+    {
+      "epoch": 0.19399115449138327,
+      "grad_norm": 0.247020197460797,
+      "learning_rate": 9.664634146341464e-06,
+      "loss": 0.2431,
+      "step": 318
+    },
+    {
+      "epoch": 0.1946011895684002,
+      "grad_norm": 0.2697722255574314,
+      "learning_rate": 9.695121951219513e-06,
+      "loss": 0.2861,
+      "step": 319
+    },
+    {
+      "epoch": 0.19521122464541713,
+      "grad_norm": 0.25055251104844295,
+      "learning_rate": 9.725609756097562e-06,
+      "loss": 0.2634,
+      "step": 320
+    },
+    {
+      "epoch": 0.19582125972243403,
+      "grad_norm": 0.25664030673072813,
+      "learning_rate": 9.756097560975611e-06,
+      "loss": 0.2594,
+      "step": 321
+    },
+    {
+      "epoch": 0.19643129479945096,
+      "grad_norm": 0.2568807633504417,
+      "learning_rate": 9.78658536585366e-06,
+      "loss": 0.2602,
+      "step": 322
+    },
+    {
+      "epoch": 0.19704132987646789,
+      "grad_norm": 0.30938855639166934,
+      "learning_rate": 9.817073170731707e-06,
+      "loss": 0.2646,
+      "step": 323
+    },
+    {
+      "epoch": 0.19765136495348481,
+      "grad_norm": 0.25640607299564616,
+      "learning_rate": 9.847560975609756e-06,
+      "loss": 0.2557,
+      "step": 324
+    },
+    {
+      "epoch": 0.19826140003050174,
+      "grad_norm": 0.2755277016357419,
+      "learning_rate": 9.878048780487805e-06,
+      "loss": 0.2676,
+      "step": 325
+    },
+    {
+      "epoch": 0.19887143510751867,
+      "grad_norm": 0.2828434225373175,
+      "learning_rate": 9.908536585365854e-06,
+      "loss": 0.278,
+      "step": 326
+    },
+    {
+      "epoch": 0.1994814701845356,
+      "grad_norm": 0.25703933013169117,
+      "learning_rate": 9.939024390243903e-06,
+      "loss": 0.252,
+      "step": 327
+    },
+    {
+      "epoch": 0.20009150526155253,
+      "grad_norm": 0.23354984579219248,
+      "learning_rate": 9.969512195121952e-06,
+      "loss": 0.2464,
+      "step": 328
+    },
+    {
+      "epoch": 0.20070154033856946,
+      "grad_norm": 0.26361369945265395,
+      "learning_rate": 1e-05,
+      "loss": 0.2502,
+      "step": 329
+    },
+    {
+      "epoch": 0.2013115754155864,
+      "grad_norm": 0.2507652167428992,
+      "learning_rate": 9.999997168562324e-06,
+      "loss": 0.2673,
+      "step": 330
+    },
+    {
+      "epoch": 0.20192161049260332,
+      "grad_norm": 0.27304236135561094,
+      "learning_rate": 9.9999886742525e-06,
+      "loss": 0.2598,
+      "step": 331
+    },
+    {
+      "epoch": 0.20253164556962025,
+      "grad_norm": 0.2287359091408107,
+      "learning_rate": 9.99997451708015e-06,
+      "loss": 0.2249,
+      "step": 332
+    },
+    {
+      "epoch": 0.20314168064663718,
+      "grad_norm": 0.2673246266757542,
+      "learning_rate": 9.999954697061305e-06,
+      "loss": 0.2487,
+      "step": 333
+    },
+    {
+      "epoch": 0.2037517157236541,
+      "grad_norm": 0.25713057306638515,
+      "learning_rate": 9.999929214218419e-06,
+      "loss": 0.2596,
+      "step": 334
+    },
+    {
+      "epoch": 0.20436175080067104,
+      "grad_norm": 0.24997141485448354,
+      "learning_rate": 9.999898068580346e-06,
+      "loss": 0.2632,
+      "step": 335
+    },
+    {
+      "epoch": 0.20497178587768797,
+      "grad_norm": 0.26986376068528667,
+      "learning_rate": 9.999861260182366e-06,
+      "loss": 0.2508,
+      "step": 336
+    },
+    {
+      "epoch": 0.2055818209547049,
+      "grad_norm": 0.25089494428271947,
+      "learning_rate": 9.999818789066164e-06,
+      "loss": 0.262,
+      "step": 337
+    },
+    {
+      "epoch": 0.20619185603172183,
+      "grad_norm": 0.26370819630042563,
+      "learning_rate": 9.999770655279843e-06,
+      "loss": 0.2579,
+      "step": 338
+    },
+    {
+      "epoch": 0.20680189110873876,
+      "grad_norm": 0.2951528596227915,
+      "learning_rate": 9.99971685887792e-06,
+      "loss": 0.2843,
+      "step": 339
+    },
+    {
+      "epoch": 0.20741192618575569,
+      "grad_norm": 0.25114828422920793,
+      "learning_rate": 9.999657399921321e-06,
+      "loss": 0.2701,
+      "step": 340
+    },
+    {
+      "epoch": 0.20802196126277261,
+      "grad_norm": 0.2542038881881685,
+      "learning_rate": 9.999592278477389e-06,
+      "loss": 0.264,
+      "step": 341
+    },
+    {
+      "epoch": 0.20863199633978954,
+      "grad_norm": 0.2783939826492031,
+      "learning_rate": 9.999521494619876e-06,
+      "loss": 0.2705,
+      "step": 342
+    },
+    {
+      "epoch": 0.20924203141680647,
+      "grad_norm": 0.2387066722281608,
+      "learning_rate": 9.999445048428952e-06,
+      "loss": 0.261,
+      "step": 343
+    },
+    {
+      "epoch": 0.2098520664938234,
+      "grad_norm": 0.2577640282603642,
+      "learning_rate": 9.999362939991202e-06,
+      "loss": 0.2521,
+      "step": 344
+    },
+    {
+      "epoch": 0.21046210157084033,
+      "grad_norm": 0.2277437965064566,
+      "learning_rate": 9.999275169399614e-06,
+      "loss": 0.2401,
+      "step": 345
+    },
+    {
+      "epoch": 0.21107213664785726,
+      "grad_norm": 0.2607184798714529,
+      "learning_rate": 9.999181736753598e-06,
+      "loss": 0.2368,
+      "step": 346
+    },
+    {
+      "epoch": 0.2116821717248742,
+      "grad_norm": 0.2743794302014217,
+      "learning_rate": 9.999082642158972e-06,
+      "loss": 0.25,
+      "step": 347
+    },
+    {
+      "epoch": 0.21229220680189112,
+      "grad_norm": 0.27233210453368745,
+      "learning_rate": 9.99897788572797e-06,
+      "loss": 0.2836,
+      "step": 348
+    },
+    {
+      "epoch": 0.21290224187890805,
+      "grad_norm": 0.25517501025766626,
+      "learning_rate": 9.998867467579234e-06,
+      "loss": 0.253,
+      "step": 349
+    },
+    {
+      "epoch": 0.21351227695592498,
+      "grad_norm": 0.2579378404488621,
+      "learning_rate": 9.998751387837822e-06,
+      "loss": 0.2616,
+      "step": 350
+    },
+    {
+      "epoch": 0.21412231203294188,
+      "grad_norm": 0.2812216022431848,
+      "learning_rate": 9.998629646635203e-06,
+      "loss": 0.2586,
+      "step": 351
+    },
+    {
+      "epoch": 0.2147323471099588,
+      "grad_norm": 0.24197851638553763,
+      "learning_rate": 9.99850224410926e-06,
+      "loss": 0.2422,
+      "step": 352
+    },
+    {
+      "epoch": 0.21534238218697574,
+      "grad_norm": 0.27306080649708614,
+      "learning_rate": 9.998369180404283e-06,
+      "loss": 0.2671,
+      "step": 353
+    },
+    {
+      "epoch": 0.21595241726399267,
+      "grad_norm": 0.357489524633953,
+      "learning_rate": 9.998230455670978e-06,
+      "loss": 0.2522,
+      "step": 354
+    },
+    {
+      "epoch": 0.2165624523410096,
+      "grad_norm": 0.2697247457552434,
+      "learning_rate": 9.998086070066459e-06,
+      "loss": 0.2809,
+      "step": 355
+    },
+    {
+      "epoch": 0.21717248741802653,
+      "grad_norm": 0.2620363759776813,
+      "learning_rate": 9.997936023754258e-06,
+      "loss": 0.26,
+      "step": 356
+    },
+    {
+      "epoch": 0.21778252249504346,
+      "grad_norm": 0.25870971628871203,
+      "learning_rate": 9.99778031690431e-06,
+      "loss": 0.2767,
+      "step": 357
+    },
+    {
+      "epoch": 0.2183925575720604,
+      "grad_norm": 0.2312512617517111,
+      "learning_rate": 9.997618949692966e-06,
+      "loss": 0.2467,
+      "step": 358
+    },
+    {
+      "epoch": 0.21900259264907732,
+      "grad_norm": 0.2731114305931224,
+      "learning_rate": 9.997451922302987e-06,
+      "loss": 0.2346,
+      "step": 359
+    },
+    {
+      "epoch": 0.21961262772609424,
+      "grad_norm": 0.25657499501781883,
+      "learning_rate": 9.997279234923542e-06,
+      "loss": 0.2504,
+      "step": 360
+    },
+    {
+      "epoch": 0.22022266280311117,
+      "grad_norm": 0.2510163321878893,
+      "learning_rate": 9.997100887750215e-06,
+      "loss": 0.2517,
+      "step": 361
+    },
+    {
+      "epoch": 0.2208326978801281,
+      "grad_norm": 0.25272708866757615,
+      "learning_rate": 9.996916880984995e-06,
+      "loss": 0.2477,
+      "step": 362
+    },
+    {
+      "epoch": 0.22144273295714503,
+      "grad_norm": 0.2584798059801417,
+      "learning_rate": 9.996727214836286e-06,
+      "loss": 0.2509,
+      "step": 363
+    },
+    {
+      "epoch": 0.22205276803416196,
+      "grad_norm": 0.27124501953300556,
+      "learning_rate": 9.996531889518898e-06,
+      "loss": 0.2635,
+      "step": 364
+    },
+    {
+      "epoch": 0.2226628031111789,
+      "grad_norm": 0.2807475690600024,
+      "learning_rate": 9.99633090525405e-06,
+      "loss": 0.2658,
+      "step": 365
+    },
+    {
+      "epoch": 0.22327283818819582,
+      "grad_norm": 0.26904250476142233,
+      "learning_rate": 9.996124262269376e-06,
+      "loss": 0.2446,
+      "step": 366
+    },
+    {
+      "epoch": 0.22388287326521275,
+      "grad_norm": 0.2680294619993861,
+      "learning_rate": 9.99591196079891e-06,
+      "loss": 0.2595,
+      "step": 367
+    },
+    {
+      "epoch": 0.22449290834222968,
+      "grad_norm": 0.285142520335724,
+      "learning_rate": 9.995694001083103e-06,
+      "loss": 0.2638,
+      "step": 368
+    },
+    {
+      "epoch": 0.2251029434192466,
+      "grad_norm": 0.27597769384607446,
+      "learning_rate": 9.995470383368808e-06,
+      "loss": 0.2659,
+      "step": 369
+    },
+    {
+      "epoch": 0.22571297849626354,
+      "grad_norm": 0.30424817017406613,
+      "learning_rate": 9.99524110790929e-06,
+      "loss": 0.2714,
+      "step": 370
+    },
+    {
+      "epoch": 0.22632301357328047,
+      "grad_norm": 0.25721127229254476,
+      "learning_rate": 9.99500617496422e-06,
+      "loss": 0.2611,
+      "step": 371
+    },
+    {
+      "epoch": 0.2269330486502974,
+      "grad_norm": 0.2751301645393888,
+      "learning_rate": 9.994765584799679e-06,
+      "loss": 0.2618,
+      "step": 372
+    },
+    {
+      "epoch": 0.22754308372731433,
+      "grad_norm": 0.2692736088441834,
+      "learning_rate": 9.994519337688152e-06,
+      "loss": 0.2492,
+      "step": 373
+    },
+    {
+      "epoch": 0.22815311880433126,
+      "grad_norm": 0.25304783010432796,
+      "learning_rate": 9.994267433908533e-06,
+      "loss": 0.2527,
+      "step": 374
+    },
+    {
+      "epoch": 0.22876315388134819,
+      "grad_norm": 0.2863607765873851,
+      "learning_rate": 9.994009873746121e-06,
+      "loss": 0.2608,
+      "step": 375
+    },
+    {
+      "epoch": 0.22937318895836512,
+      "grad_norm": 0.26194638397821157,
+      "learning_rate": 9.993746657492622e-06,
+      "loss": 0.2673,
+      "step": 376
+    },
+    {
+      "epoch": 0.22998322403538204,
+      "grad_norm": 0.26236574139222973,
+      "learning_rate": 9.993477785446151e-06,
+      "loss": 0.2449,
+      "step": 377
+    },
+    {
+      "epoch": 0.23059325911239897,
+      "grad_norm": 0.2750923719520727,
+      "learning_rate": 9.993203257911222e-06,
+      "loss": 0.2683,
+      "step": 378
+    },
+    {
+      "epoch": 0.2312032941894159,
+      "grad_norm": 0.28560193546700424,
+      "learning_rate": 9.99292307519876e-06,
+      "loss": 0.254,
+      "step": 379
+    },
+    {
+      "epoch": 0.23181332926643283,
+      "grad_norm": 0.3132035003182686,
+      "learning_rate": 9.992637237626092e-06,
+      "loss": 0.2523,
+      "step": 380
+    },
+    {
+      "epoch": 0.23242336434344976,
+      "grad_norm": 0.2825046469057102,
+      "learning_rate": 9.992345745516954e-06,
+      "loss": 0.2815,
+      "step": 381
+    },
+    {
+      "epoch": 0.23303339942046666,
+      "grad_norm": 0.28466890282763896,
+      "learning_rate": 9.992048599201478e-06,
+      "loss": 0.2652,
+      "step": 382
+    },
+    {
+      "epoch": 0.2336434344974836,
+      "grad_norm": 0.26975445758231287,
+      "learning_rate": 9.991745799016206e-06,
+      "loss": 0.2612,
+      "step": 383
+    },
+    {
+      "epoch": 0.23425346957450052,
+      "grad_norm": 0.23586273367056518,
+      "learning_rate": 9.991437345304084e-06,
+      "loss": 0.2596,
+      "step": 384
+    },
+    {
+      "epoch": 0.23486350465151745,
+      "grad_norm": 0.30573359370561365,
+      "learning_rate": 9.991123238414455e-06,
+      "loss": 0.2524,
+      "step": 385
+    },
+    {
+      "epoch": 0.23547353972853438,
+      "grad_norm": 0.2630216344835691,
+      "learning_rate": 9.990803478703073e-06,
+      "loss": 0.2446,
+      "step": 386
+    },
+    {
+      "epoch": 0.2360835748055513,
+      "grad_norm": 0.25756493585689366,
+      "learning_rate": 9.990478066532088e-06,
+      "loss": 0.2303,
+      "step": 387
+    },
+    {
+      "epoch": 0.23669360988256824,
+      "grad_norm": 0.2694584250938353,
+      "learning_rate": 9.990147002270051e-06,
+      "loss": 0.2526,
+      "step": 388
+    },
+    {
+      "epoch": 0.23730364495958517,
+      "grad_norm": 0.28404877974160536,
+      "learning_rate": 9.989810286291923e-06,
+      "loss": 0.2578,
+      "step": 389
+    },
+    {
+      "epoch": 0.2379136800366021,
+      "grad_norm": 0.24306934901870883,
+      "learning_rate": 9.989467918979055e-06,
+      "loss": 0.2508,
+      "step": 390
+    },
+    {
+      "epoch": 0.23852371511361903,
+      "grad_norm": 0.27162306691858007,
+      "learning_rate": 9.989119900719206e-06,
+      "loss": 0.2464,
+      "step": 391
+    },
+    {
+      "epoch": 0.23913375019063596,
+      "grad_norm": 0.267591647087372,
+      "learning_rate": 9.988766231906532e-06,
+      "loss": 0.2669,
+      "step": 392
+    },
+    {
+      "epoch": 0.2397437852676529,
+      "grad_norm": 0.250629654500331,
+      "learning_rate": 9.988406912941591e-06,
+      "loss": 0.2599,
+      "step": 393
+    },
+    {
+      "epoch": 0.24035382034466982,
+      "grad_norm": 0.27800678255553674,
+      "learning_rate": 9.988041944231335e-06,
+      "loss": 0.266,
+      "step": 394
+    },
+    {
+      "epoch": 0.24096385542168675,
+      "grad_norm": 0.2646178524474114,
+      "learning_rate": 9.987671326189123e-06,
+      "loss": 0.2297,
+      "step": 395
+    },
+    {
+      "epoch": 0.24157389049870368,
+      "grad_norm": 0.2618422966309834,
+      "learning_rate": 9.987295059234704e-06,
+      "loss": 0.2593,
+      "step": 396
+    },
+    {
+      "epoch": 0.2421839255757206,
+      "grad_norm": 0.26039061767925137,
+      "learning_rate": 9.986913143794232e-06,
+      "loss": 0.2594,
+      "step": 397
+    },
+    {
+      "epoch": 0.24279396065273753,
+      "grad_norm": 0.43661380805158756,
+      "learning_rate": 9.986525580300253e-06,
+      "loss": 0.2553,
+      "step": 398
+    },
+    {
+      "epoch": 0.24340399572975446,
+      "grad_norm": 0.2521963798608539,
+      "learning_rate": 9.986132369191712e-06,
+      "loss": 0.2386,
+      "step": 399
+    },
+    {
+      "epoch": 0.2440140308067714,
+      "grad_norm": 0.2522149023961132,
+      "learning_rate": 9.98573351091395e-06,
+      "loss": 0.2489,
+      "step": 400
+    },
+    {
+      "epoch": 0.24462406588378832,
+      "grad_norm": 0.24778130062040454,
+      "learning_rate": 9.985329005918702e-06,
+      "loss": 0.2291,
+      "step": 401
+    },
+    {
+      "epoch": 0.24523410096080525,
+      "grad_norm": 0.2569038342607788,
+      "learning_rate": 9.984918854664105e-06,
+      "loss": 0.2612,
+      "step": 402
+    },
+    {
+      "epoch": 0.24584413603782218,
+      "grad_norm": 0.25041191695544,
+      "learning_rate": 9.984503057614684e-06,
+      "loss": 0.2576,
+      "step": 403
+    },
+    {
+      "epoch": 0.2464541711148391,
+      "grad_norm": 0.24717652132981327,
+      "learning_rate": 9.984081615241356e-06,
+      "loss": 0.2495,
+      "step": 404
+    },
+    {
+      "epoch": 0.24706420619185604,
+      "grad_norm": 0.26673537646912776,
+      "learning_rate": 9.983654528021442e-06,
+      "loss": 0.2689,
+      "step": 405
+    },
+    {
+      "epoch": 0.24767424126887297,
+      "grad_norm": 0.2465342997361249,
+      "learning_rate": 9.98322179643865e-06,
+      "loss": 0.2616,
+      "step": 406
+    },
+    {
+      "epoch": 0.2482842763458899,
+      "grad_norm": 0.28245378040509006,
+      "learning_rate": 9.982783420983075e-06,
+      "loss": 0.2511,
+      "step": 407
+    },
+    {
+      "epoch": 0.24889431142290683,
+      "grad_norm": 0.26880312709115456,
+      "learning_rate": 9.982339402151217e-06,
+      "loss": 0.2327,
+      "step": 408
+    },
+    {
+      "epoch": 0.24950434649992376,
+      "grad_norm": 0.254261386844422,
+      "learning_rate": 9.981889740445958e-06,
+      "loss": 0.2388,
+      "step": 409
+    },
+    {
+      "epoch": 0.25011438157694066,
+      "grad_norm": 0.26583061665924723,
+      "learning_rate": 9.981434436376572e-06,
+      "loss": 0.2735,
+      "step": 410
+    },
+    {
+      "epoch": 0.2507244166539576,
+      "grad_norm": 0.2537203710329665,
+      "learning_rate": 9.980973490458728e-06,
+      "loss": 0.2802,
+      "step": 411
+    },
+    {
+      "epoch": 0.2513344517309745,
+      "grad_norm": 0.2669060501581036,
+      "learning_rate": 9.980506903214481e-06,
+      "loss": 0.2564,
+      "step": 412
+    },
+    {
+      "epoch": 0.2519444868079915,
+      "grad_norm": 0.2559541236925637,
+      "learning_rate": 9.980034675172274e-06,
+      "loss": 0.2432,
+      "step": 413
+    },
+    {
+      "epoch": 0.2525545218850084,
+      "grad_norm": 0.2558690804133373,
+      "learning_rate": 9.979556806866943e-06,
+      "loss": 0.2535,
+      "step": 414
+    },
+    {
+      "epoch": 0.25316455696202533,
+      "grad_norm": 0.25826409987008647,
+      "learning_rate": 9.97907329883971e-06,
+      "loss": 0.2762,
+      "step": 415
+    },
+    {
+      "epoch": 0.25377459203904223,
+      "grad_norm": 0.26642742809887227,
+      "learning_rate": 9.978584151638182e-06,
+      "loss": 0.2697,
+      "step": 416
+    },
+    {
+      "epoch": 0.2543846271160592,
+      "grad_norm": 0.23619289173506056,
+      "learning_rate": 9.978089365816357e-06,
+      "loss": 0.24,
+      "step": 417
+    },
+    {
+      "epoch": 0.2549946621930761,
+      "grad_norm": 0.2540500448608682,
+      "learning_rate": 9.977588941934615e-06,
+      "loss": 0.2762,
+      "step": 418
+    },
+    {
+      "epoch": 0.25560469727009305,
+      "grad_norm": 0.2565053314392929,
+      "learning_rate": 9.977082880559725e-06,
+      "loss": 0.27,
+      "step": 419
+    },
+    {
+      "epoch": 0.25621473234710995,
+      "grad_norm": 0.24569647149613266,
+      "learning_rate": 9.97657118226484e-06,
+      "loss": 0.2471,
+      "step": 420
+    },
+    {
+      "epoch": 0.2568247674241269,
+      "grad_norm": 0.24132230841255492,
+      "learning_rate": 9.976053847629496e-06,
+      "loss": 0.2463,
+      "step": 421
+    },
+    {
+      "epoch": 0.2574348025011438,
+      "grad_norm": 0.23527712284133528,
+      "learning_rate": 9.975530877239613e-06,
+      "loss": 0.2345,
+      "step": 422
+    },
+    {
+      "epoch": 0.25804483757816077,
+      "grad_norm": 0.2769563848341972,
+      "learning_rate": 9.975002271687496e-06,
+      "loss": 0.2664,
+      "step": 423
+    },
+    {
+      "epoch": 0.25865487265517767,
+      "grad_norm": 0.3293393338967704,
+      "learning_rate": 9.974468031571825e-06,
+      "loss": 0.2674,
+      "step": 424
+    },
+    {
+      "epoch": 0.2592649077321946,
+      "grad_norm": 0.2602552909440054,
+      "learning_rate": 9.973928157497675e-06,
+      "loss": 0.2832,
+      "step": 425
+    },
+    {
+      "epoch": 0.25987494280921153,
+      "grad_norm": 0.2580342895295799,
+      "learning_rate": 9.973382650076488e-06,
+      "loss": 0.2629,
+      "step": 426
+    },
+    {
+      "epoch": 0.26048497788622843,
+      "grad_norm": 0.26747490402513346,
+      "learning_rate": 9.972831509926094e-06,
+      "loss": 0.2719,
+      "step": 427
+    },
+    {
+      "epoch": 0.2610950129632454,
+      "grad_norm": 0.23993540925898904,
+      "learning_rate": 9.972274737670702e-06,
+      "loss": 0.2363,
+      "step": 428
+    },
+    {
+      "epoch": 0.2617050480402623,
+      "grad_norm": 0.30395384010333787,
+      "learning_rate": 9.971712333940896e-06,
+      "loss": 0.271,
+      "step": 429
+    },
+    {
+      "epoch": 0.26231508311727925,
+      "grad_norm": 0.2584379936178291,
+      "learning_rate": 9.971144299373643e-06,
+      "loss": 0.2482,
+      "step": 430
+    },
+    {
+      "epoch": 0.26292511819429615,
+      "grad_norm": 0.24603617492477198,
+      "learning_rate": 9.970570634612282e-06,
+      "loss": 0.2542,
+      "step": 431
+    },
+    {
+      "epoch": 0.2635351532713131,
+      "grad_norm": 0.2505894181965195,
+      "learning_rate": 9.969991340306533e-06,
+      "loss": 0.2404,
+      "step": 432
+    },
+    {
+      "epoch": 0.26414518834833,
+      "grad_norm": 0.24674783311739618,
+      "learning_rate": 9.969406417112489e-06,
+      "loss": 0.2363,
+      "step": 433
+    },
+    {
+      "epoch": 0.26475522342534696,
+      "grad_norm": 0.35440716800891336,
+      "learning_rate": 9.968815865692622e-06,
+      "loss": 0.2475,
+      "step": 434
+    },
+    {
+      "epoch": 0.26536525850236387,
+      "grad_norm": 0.2744766335288482,
+      "learning_rate": 9.968219686715773e-06,
+      "loss": 0.2581,
+      "step": 435
+    },
+    {
+      "epoch": 0.2659752935793808,
+      "grad_norm": 0.2674838161349361,
+      "learning_rate": 9.96761788085716e-06,
+      "loss": 0.2638,
+      "step": 436
+    },
+    {
+      "epoch": 0.2665853286563977,
+      "grad_norm": 0.2520490283184246,
+      "learning_rate": 9.967010448798376e-06,
+      "loss": 0.2529,
+      "step": 437
+    },
+    {
+      "epoch": 0.2671953637334147,
+      "grad_norm": 0.31522100232197137,
+      "learning_rate": 9.96639739122738e-06,
+      "loss": 0.2696,
+      "step": 438
+    },
+    {
+      "epoch": 0.2678053988104316,
+      "grad_norm": 0.26977483001310687,
+      "learning_rate": 9.965778708838509e-06,
+      "loss": 0.2372,
+      "step": 439
+    },
+    {
+      "epoch": 0.26841543388744854,
+      "grad_norm": 0.24287141515808836,
+      "learning_rate": 9.965154402332464e-06,
+      "loss": 0.2399,
+      "step": 440
+    },
+    {
+      "epoch": 0.26902546896446544,
+      "grad_norm": 0.2739660973493756,
+      "learning_rate": 9.964524472416319e-06,
+      "loss": 0.2529,
+      "step": 441
+    },
+    {
+      "epoch": 0.2696355040414824,
+      "grad_norm": 0.26189624333130374,
+      "learning_rate": 9.96388891980352e-06,
+      "loss": 0.2549,
+      "step": 442
+    },
+    {
+      "epoch": 0.2702455391184993,
+      "grad_norm": 0.26116140101705765,
+      "learning_rate": 9.963247745213876e-06,
+      "loss": 0.2613,
+      "step": 443
+    },
+    {
+      "epoch": 0.27085557419551626,
+      "grad_norm": 0.2698487180887434,
+      "learning_rate": 9.962600949373567e-06,
+      "loss": 0.2635,
+      "step": 444
+    },
+    {
+      "epoch": 0.27146560927253316,
+      "grad_norm": 0.24553015492634403,
+      "learning_rate": 9.961948533015135e-06,
+      "loss": 0.2442,
+      "step": 445
+    },
+    {
+      "epoch": 0.2720756443495501,
+      "grad_norm": 0.23926699877002677,
+      "learning_rate": 9.961290496877492e-06,
+      "loss": 0.2598,
+      "step": 446
+    },
+    {
+      "epoch": 0.272685679426567,
+      "grad_norm": 0.2765901598887178,
+      "learning_rate": 9.960626841705913e-06,
+      "loss": 0.2768,
+      "step": 447
+    },
+    {
+      "epoch": 0.273295714503584,
+      "grad_norm": 0.24629883798169278,
+      "learning_rate": 9.95995756825204e-06,
+      "loss": 0.2649,
+      "step": 448
+    },
+    {
+      "epoch": 0.2739057495806009,
+      "grad_norm": 0.26061938760384895,
+      "learning_rate": 9.959282677273869e-06,
+      "loss": 0.2599,
+      "step": 449
+    },
+    {
+      "epoch": 0.27451578465761783,
+      "grad_norm": 0.25357892001252424,
+      "learning_rate": 9.95860216953577e-06,
+      "loss": 0.2277,
+      "step": 450
+    },
+    {
+      "epoch": 0.27512581973463474,
+      "grad_norm": 0.24102293841362316,
+      "learning_rate": 9.95791604580847e-06,
+      "loss": 0.2368,
+      "step": 451
+    },
+    {
+      "epoch": 0.2757358548116517,
+      "grad_norm": 0.2656199764240277,
+      "learning_rate": 9.957224306869053e-06,
+      "loss": 0.2403,
+      "step": 452
+    },
+    {
+      "epoch": 0.2763458898886686,
+      "grad_norm": 0.26931339295947415,
+      "learning_rate": 9.956526953500965e-06,
+      "loss": 0.2528,
+      "step": 453
+    },
+    {
+      "epoch": 0.27695592496568555,
+      "grad_norm": 0.25037961889937077,
+      "learning_rate": 9.955823986494012e-06,
+      "loss": 0.2625,
+      "step": 454
+    },
+    {
+      "epoch": 0.27756596004270245,
+      "grad_norm": 0.25439066559296586,
+      "learning_rate": 9.955115406644357e-06,
+      "loss": 0.2665,
+      "step": 455
+    },
+    {
+      "epoch": 0.2781759951197194,
+      "grad_norm": 0.25659990784327696,
+      "learning_rate": 9.95440121475452e-06,
+      "loss": 0.2415,
+      "step": 456
+    },
+    {
+      "epoch": 0.2787860301967363,
+      "grad_norm": 0.24662984651313433,
+      "learning_rate": 9.953681411633376e-06,
+      "loss": 0.2545,
+      "step": 457
+    },
+    {
+      "epoch": 0.2793960652737532,
+      "grad_norm": 0.2558250167454128,
+      "learning_rate": 9.952955998096155e-06,
+      "loss": 0.2488,
+      "step": 458
+    },
+    {
+      "epoch": 0.28000610035077017,
+      "grad_norm": 0.23806375709432726,
+      "learning_rate": 9.952224974964446e-06,
+      "loss": 0.2284,
+      "step": 459
+    },
+    {
+      "epoch": 0.2806161354277871,
+      "grad_norm": 0.24600626960166394,
+      "learning_rate": 9.951488343066184e-06,
+      "loss": 0.2335,
+      "step": 460
+    },
+    {
+      "epoch": 0.28122617050480403,
+      "grad_norm": 0.2811619586861078,
+      "learning_rate": 9.950746103235663e-06,
+      "loss": 0.2628,
+      "step": 461
+    },
+    {
+      "epoch": 0.28183620558182093,
+      "grad_norm": 0.24366421930517498,
+      "learning_rate": 9.949998256313523e-06,
+      "loss": 0.2796,
+      "step": 462
+    },
+    {
+      "epoch": 0.2824462406588379,
+      "grad_norm": 0.25171729952336713,
+      "learning_rate": 9.949244803146757e-06,
+      "loss": 0.2493,
+      "step": 463
+    },
+    {
+      "epoch": 0.2830562757358548,
+      "grad_norm": 0.25433563615871724,
+      "learning_rate": 9.948485744588709e-06,
+      "loss": 0.2588,
+      "step": 464
+    },
+    {
+      "epoch": 0.28366631081287175,
+      "grad_norm": 0.2282071902405683,
+      "learning_rate": 9.947721081499068e-06,
+      "loss": 0.2464,
+      "step": 465
+    },
+    {
+      "epoch": 0.28427634588988865,
+      "grad_norm": 0.24596986616467853,
+      "learning_rate": 9.946950814743871e-06,
+      "loss": 0.244,
+      "step": 466
+    },
+    {
+      "epoch": 0.2848863809669056,
+      "grad_norm": 0.25314080481909895,
+      "learning_rate": 9.946174945195508e-06,
+      "loss": 0.2368,
+      "step": 467
+    },
+    {
+      "epoch": 0.2854964160439225,
+      "grad_norm": 0.2480664113962836,
+      "learning_rate": 9.945393473732706e-06,
+      "loss": 0.2467,
+      "step": 468
+    },
+    {
+      "epoch": 0.28610645112093946,
+      "grad_norm": 0.24597627536896932,
+      "learning_rate": 9.944606401240538e-06,
+      "loss": 0.2499,
+      "step": 469
+    },
+    {
+      "epoch": 0.28671648619795637,
+      "grad_norm": 0.26505498442178216,
+      "learning_rate": 9.943813728610428e-06,
+      "loss": 0.253,
+      "step": 470
+    },
+    {
+      "epoch": 0.2873265212749733,
+      "grad_norm": 0.2590414417352618,
+      "learning_rate": 9.943015456740132e-06,
+      "loss": 0.2534,
+      "step": 471
+    },
+    {
+      "epoch": 0.2879365563519902,
+      "grad_norm": 0.25994353284833804,
+      "learning_rate": 9.942211586533756e-06,
+      "loss": 0.269,
+      "step": 472
+    },
+    {
+      "epoch": 0.2885465914290072,
+      "grad_norm": 0.2757214458321111,
+      "learning_rate": 9.941402118901743e-06,
+      "loss": 0.2538,
+      "step": 473
+    },
+    {
+      "epoch": 0.2891566265060241,
+      "grad_norm": 0.22785777363847964,
+      "learning_rate": 9.940587054760875e-06,
+      "loss": 0.2449,
+      "step": 474
+    },
+    {
+      "epoch": 0.28976666158304104,
+      "grad_norm": 0.2977878583545315,
+      "learning_rate": 9.939766395034275e-06,
+      "loss": 0.2436,
+      "step": 475
+    },
+    {
+      "epoch": 0.29037669666005794,
+      "grad_norm": 0.25374903168798246,
+      "learning_rate": 9.938940140651398e-06,
+      "loss": 0.2551,
+      "step": 476
+    },
+    {
+      "epoch": 0.2909867317370749,
+      "grad_norm": 0.26923026331434907,
+      "learning_rate": 9.938108292548044e-06,
+      "loss": 0.2356,
+      "step": 477
+    },
+    {
+      "epoch": 0.2915967668140918,
+      "grad_norm": 0.2750449727972376,
+      "learning_rate": 9.937270851666341e-06,
+      "loss": 0.2638,
+      "step": 478
+    },
+    {
+      "epoch": 0.29220680189110876,
+      "grad_norm": 0.2533222033571772,
+      "learning_rate": 9.936427818954753e-06,
+      "loss": 0.2432,
+      "step": 479
+    },
+    {
+      "epoch": 0.29281683696812566,
+      "grad_norm": 0.2506513318638803,
+      "learning_rate": 9.935579195368078e-06,
+      "loss": 0.2677,
+      "step": 480
+    },
+    {
+      "epoch": 0.2934268720451426,
+      "grad_norm": 0.2631288863531059,
+      "learning_rate": 9.934724981867447e-06,
+      "loss": 0.2528,
+      "step": 481
+    },
+    {
+      "epoch": 0.2940369071221595,
+      "grad_norm": 0.23937145957999464,
+      "learning_rate": 9.93386517942032e-06,
+      "loss": 0.2478,
+      "step": 482
+    },
+    {
+      "epoch": 0.2946469421991765,
+      "grad_norm": 0.25924435777122534,
+      "learning_rate": 9.93299978900049e-06,
+      "loss": 0.2504,
+      "step": 483
+    },
+    {
+      "epoch": 0.2952569772761934,
+      "grad_norm": 0.26376643896473395,
+      "learning_rate": 9.932128811588074e-06,
+      "loss": 0.2467,
+      "step": 484
+    },
+    {
+      "epoch": 0.29586701235321033,
+      "grad_norm": 0.27081772073605664,
+      "learning_rate": 9.931252248169518e-06,
+      "loss": 0.2649,
+      "step": 485
+    },
+    {
+      "epoch": 0.29647704743022724,
+      "grad_norm": 0.2572481234239358,
+      "learning_rate": 9.9303700997376e-06,
+      "loss": 0.237,
+      "step": 486
+    },
+    {
+      "epoch": 0.2970870825072442,
+      "grad_norm": 0.263952425140558,
+      "learning_rate": 9.929482367291417e-06,
+      "loss": 0.2566,
+      "step": 487
+    },
+    {
+      "epoch": 0.2976971175842611,
+      "grad_norm": 0.25097369019657156,
+      "learning_rate": 9.928589051836392e-06,
+      "loss": 0.2513,
+      "step": 488
+    },
+    {
+      "epoch": 0.298307152661278,
+      "grad_norm": 0.25162784794934706,
+      "learning_rate": 9.927690154384273e-06,
+      "loss": 0.2493,
+      "step": 489
+    },
+    {
+      "epoch": 0.29891718773829495,
+      "grad_norm": 0.29235771663739313,
+      "learning_rate": 9.92678567595313e-06,
+      "loss": 0.2734,
+      "step": 490
+    },
+    {
+      "epoch": 0.29952722281531186,
+      "grad_norm": 0.24856494040312965,
+      "learning_rate": 9.92587561756735e-06,
+      "loss": 0.2632,
+      "step": 491
+    },
+    {
+      "epoch": 0.3001372578923288,
+      "grad_norm": 0.25410549867586324,
+      "learning_rate": 9.924959980257645e-06,
+      "loss": 0.2384,
+      "step": 492
+    },
+    {
+      "epoch": 0.3007472929693457,
+      "grad_norm": 0.2532511684622894,
+      "learning_rate": 9.924038765061042e-06,
+      "loss": 0.2358,
+      "step": 493
+    },
+    {
+      "epoch": 0.30135732804636267,
+      "grad_norm": 0.24565076144059494,
+      "learning_rate": 9.923111973020885e-06,
+      "loss": 0.257,
+      "step": 494
+    },
+    {
+      "epoch": 0.3019673631233796,
+      "grad_norm": 0.2583658220557501,
+      "learning_rate": 9.922179605186837e-06,
+      "loss": 0.2553,
+      "step": 495
+    },
+    {
+      "epoch": 0.30257739820039653,
+      "grad_norm": 0.2482486410141977,
+      "learning_rate": 9.921241662614874e-06,
+      "loss": 0.2399,
+      "step": 496
+    },
+    {
+      "epoch": 0.30318743327741343,
+      "grad_norm": 0.25631202391417307,
+      "learning_rate": 9.920298146367287e-06,
+      "loss": 0.2649,
+      "step": 497
+    },
+    {
+      "epoch": 0.3037974683544304,
+      "grad_norm": 0.24307313332469035,
+      "learning_rate": 9.919349057512679e-06,
+      "loss": 0.242,
+      "step": 498
+    },
+    {
+      "epoch": 0.3044075034314473,
+      "grad_norm": 0.22637978283295307,
+      "learning_rate": 9.918394397125963e-06,
+      "loss": 0.2278,
+      "step": 499
+    },
+    {
+      "epoch": 0.30501753850846425,
+      "grad_norm": 0.25163879424818747,
+      "learning_rate": 9.917434166288364e-06,
+      "loss": 0.2265,
+      "step": 500
+    },
+    {
+      "epoch": 0.30501753850846425,
+      "eval_loss": 0.2490791380405426,
+      "eval_runtime": 260.635,
+      "eval_samples_per_second": 4.105,
+      "eval_steps_per_second": 0.13,
+      "step": 500
+    },
+    {
+      "epoch": 0.30562757358548115,
+      "grad_norm": 0.24358156361203398,
+      "learning_rate": 9.916468366087418e-06,
+      "loss": 0.2425,
+      "step": 501
+    },
+    {
+      "epoch": 0.3062376086624981,
+      "grad_norm": 0.33977096706149845,
+      "learning_rate": 9.915496997616963e-06,
+      "loss": 0.2468,
+      "step": 502
+    },
+    {
+      "epoch": 0.306847643739515,
+      "grad_norm": 0.23025095643742358,
+      "learning_rate": 9.91452006197715e-06,
+      "loss": 0.2266,
+      "step": 503
+    },
+    {
+      "epoch": 0.30745767881653197,
+      "grad_norm": 0.2600726676831414,
+      "learning_rate": 9.913537560274426e-06,
+      "loss": 0.2571,
+      "step": 504
+    },
+    {
+      "epoch": 0.30806771389354887,
+      "grad_norm": 0.2228281275268031,
+      "learning_rate": 9.912549493621555e-06,
+      "loss": 0.2368,
+      "step": 505
+    },
+    {
+      "epoch": 0.3086777489705658,
+      "grad_norm": 0.24438283754941054,
+      "learning_rate": 9.911555863137593e-06,
+      "loss": 0.2625,
+      "step": 506
+    },
+    {
+      "epoch": 0.3092877840475827,
+      "grad_norm": 0.23383710835984622,
+      "learning_rate": 9.910556669947902e-06,
+      "loss": 0.2471,
+      "step": 507
+    },
+    {
+      "epoch": 0.3098978191245997,
+      "grad_norm": 0.24743991801197227,
+      "learning_rate": 9.909551915184144e-06,
+      "loss": 0.2635,
+      "step": 508
+    },
+    {
+      "epoch": 0.3105078542016166,
+      "grad_norm": 0.2794396361771197,
+      "learning_rate": 9.908541599984276e-06,
+      "loss": 0.276,
+      "step": 509
+    },
+    {
+      "epoch": 0.31111788927863354,
+      "grad_norm": 0.2814300827849775,
+      "learning_rate": 9.907525725492559e-06,
+      "loss": 0.2646,
+      "step": 510
+    },
+    {
+      "epoch": 0.31172792435565044,
+      "grad_norm": 0.24066164246061003,
+      "learning_rate": 9.906504292859544e-06,
+      "loss": 0.2815,
+      "step": 511
+    },
+    {
+      "epoch": 0.3123379594326674,
+      "grad_norm": 0.2720619923340415,
+      "learning_rate": 9.905477303242085e-06,
+      "loss": 0.2475,
+      "step": 512
+    },
+    {
+      "epoch": 0.3129479945096843,
+      "grad_norm": 0.266396569737668,
+      "learning_rate": 9.904444757803322e-06,
+      "loss": 0.2143,
+      "step": 513
+    },
+    {
+      "epoch": 0.31355802958670126,
+      "grad_norm": 0.26843992179019743,
+      "learning_rate": 9.903406657712688e-06,
+      "loss": 0.258,
+      "step": 514
+    },
+    {
+      "epoch": 0.31416806466371816,
+      "grad_norm": 0.24052398634290234,
+      "learning_rate": 9.902363004145914e-06,
+      "loss": 0.2387,
+      "step": 515
+    },
+    {
+      "epoch": 0.3147780997407351,
+      "grad_norm": 0.2523106598314947,
+      "learning_rate": 9.901313798285012e-06,
+      "loss": 0.2466,
+      "step": 516
+    },
+    {
+      "epoch": 0.315388134817752,
+      "grad_norm": 0.22329761292848602,
+      "learning_rate": 9.90025904131829e-06,
+      "loss": 0.2187,
+      "step": 517
+    },
+    {
+      "epoch": 0.315998169894769,
+      "grad_norm": 0.22947361076735034,
+      "learning_rate": 9.899198734440335e-06,
+      "loss": 0.2544,
+      "step": 518
+    },
+    {
+      "epoch": 0.3166082049717859,
+      "grad_norm": 0.2645678387368134,
+      "learning_rate": 9.898132878852026e-06,
+      "loss": 0.2703,
+      "step": 519
+    },
+    {
+      "epoch": 0.3172182400488028,
+      "grad_norm": 0.22881394529975585,
+      "learning_rate": 9.897061475760528e-06,
+      "loss": 0.2188,
+      "step": 520
+    },
+    {
+      "epoch": 0.31782827512581974,
+      "grad_norm": 0.23413558660635153,
+      "learning_rate": 9.895984526379282e-06,
+      "loss": 0.2121,
+      "step": 521
+    },
+    {
+      "epoch": 0.31843831020283664,
+      "grad_norm": 0.24515654764657868,
+      "learning_rate": 9.894902031928014e-06,
+      "loss": 0.2582,
+      "step": 522
+    },
+    {
+      "epoch": 0.3190483452798536,
+      "grad_norm": 0.24783751437653048,
+      "learning_rate": 9.89381399363273e-06,
+      "loss": 0.2454,
+      "step": 523
+    },
+    {
+      "epoch": 0.3196583803568705,
+      "grad_norm": 0.25944084766719894,
+      "learning_rate": 9.892720412725717e-06,
+      "loss": 0.2339,
+      "step": 524
+    },
+    {
+      "epoch": 0.32026841543388745,
+      "grad_norm": 0.24221200657802805,
+      "learning_rate": 9.891621290445534e-06,
+      "loss": 0.2415,
+      "step": 525
+    },
+    {
+      "epoch": 0.32087845051090436,
+      "grad_norm": 0.26721221201594864,
+      "learning_rate": 9.890516628037024e-06,
+      "loss": 0.2467,
+      "step": 526
+    },
+    {
+      "epoch": 0.3214884855879213,
+      "grad_norm": 0.2661462723846872,
+      "learning_rate": 9.889406426751296e-06,
+      "loss": 0.2544,
+      "step": 527
+    },
+    {
+      "epoch": 0.3220985206649382,
+      "grad_norm": 0.25737118244462615,
+      "learning_rate": 9.88829068784574e-06,
+      "loss": 0.2519,
+      "step": 528
+    },
+    {
+      "epoch": 0.32270855574195517,
+      "grad_norm": 0.2566760568897278,
+      "learning_rate": 9.887169412584012e-06,
+      "loss": 0.248,
+      "step": 529
+    },
+    {
+      "epoch": 0.3233185908189721,
+      "grad_norm": 0.27835987527905887,
+      "learning_rate": 9.88604260223604e-06,
+      "loss": 0.2469,
+      "step": 530
+    },
+    {
+      "epoch": 0.32392862589598903,
+      "grad_norm": 0.2655282154010336,
+      "learning_rate": 9.884910258078022e-06,
+      "loss": 0.2599,
+      "step": 531
+    },
+    {
+      "epoch": 0.32453866097300593,
+      "grad_norm": 0.2667148794264865,
+      "learning_rate": 9.883772381392423e-06,
+      "loss": 0.2412,
+      "step": 532
+    },
+    {
+      "epoch": 0.3251486960500229,
+      "grad_norm": 0.2623581156385756,
+      "learning_rate": 9.882628973467972e-06,
+      "loss": 0.2618,
+      "step": 533
+    },
+    {
+      "epoch": 0.3257587311270398,
+      "grad_norm": 0.24318620380902956,
+      "learning_rate": 9.881480035599667e-06,
+      "loss": 0.2376,
+      "step": 534
+    },
+    {
+      "epoch": 0.32636876620405675,
+      "grad_norm": 0.26633961251095956,
+      "learning_rate": 9.880325569088765e-06,
+      "loss": 0.2503,
+      "step": 535
+    },
+    {
+      "epoch": 0.32697880128107365,
+      "grad_norm": 0.2630382405080692,
+      "learning_rate": 9.879165575242788e-06,
+      "loss": 0.2567,
+      "step": 536
+    },
+    {
+      "epoch": 0.3275888363580906,
+      "grad_norm": 0.2541644574129942,
+      "learning_rate": 9.878000055375512e-06,
+      "loss": 0.2474,
+      "step": 537
+    },
+    {
+      "epoch": 0.3281988714351075,
+      "grad_norm": 0.30155457807091895,
+      "learning_rate": 9.87682901080698e-06,
+      "loss": 0.2464,
+      "step": 538
+    },
+    {
+      "epoch": 0.32880890651212447,
+      "grad_norm": 0.2660733739753395,
+      "learning_rate": 9.875652442863483e-06,
+      "loss": 0.2494,
+      "step": 539
+    },
+    {
+      "epoch": 0.32941894158914137,
+      "grad_norm": 0.24132541271471367,
+      "learning_rate": 9.874470352877576e-06,
+      "loss": 0.2653,
+      "step": 540
+    },
+    {
+      "epoch": 0.3300289766661583,
+      "grad_norm": 0.2315138925207163,
+      "learning_rate": 9.873282742188066e-06,
+      "loss": 0.2417,
+      "step": 541
+    },
+    {
+      "epoch": 0.3306390117431752,
+      "grad_norm": 0.31417481601320424,
+      "learning_rate": 9.87208961214001e-06,
+      "loss": 0.2766,
+      "step": 542
+    },
+    {
+      "epoch": 0.3312490468201922,
+      "grad_norm": 0.24645469040878668,
+      "learning_rate": 9.870890964084713e-06,
+      "loss": 0.2439,
+      "step": 543
+    },
+    {
+      "epoch": 0.3318590818972091,
+      "grad_norm": 0.2363463068965621,
+      "learning_rate": 9.86968679937974e-06,
+      "loss": 0.2534,
+      "step": 544
+    },
+    {
+      "epoch": 0.33246911697422604,
+      "grad_norm": 0.2692926962955553,
+      "learning_rate": 9.868477119388897e-06,
+      "loss": 0.2481,
+      "step": 545
+    },
+    {
+      "epoch": 0.33307915205124294,
+      "grad_norm": 0.2470726477470715,
+      "learning_rate": 9.867261925482233e-06,
+      "loss": 0.2376,
+      "step": 546
+    },
+    {
+      "epoch": 0.3336891871282599,
+      "grad_norm": 0.26915801212713547,
+      "learning_rate": 9.866041219036051e-06,
+      "loss": 0.2567,
+      "step": 547
+    },
+    {
+      "epoch": 0.3342992222052768,
+      "grad_norm": 0.23819617510254015,
+      "learning_rate": 9.86481500143289e-06,
+      "loss": 0.2287,
+      "step": 548
+    },
+    {
+      "epoch": 0.3349092572822937,
+      "grad_norm": 0.23372133613640378,
+      "learning_rate": 9.863583274061535e-06,
+      "loss": 0.2132,
+      "step": 549
+    },
+    {
+      "epoch": 0.33551929235931066,
+      "grad_norm": 0.2560238801091462,
+      "learning_rate": 9.862346038317009e-06,
+      "loss": 0.2379,
+      "step": 550
+    },
+    {
+      "epoch": 0.33612932743632756,
+      "grad_norm": 0.2577909389450448,
+      "learning_rate": 9.861103295600574e-06,
+      "loss": 0.2468,
+      "step": 551
+    },
+    {
+      "epoch": 0.3367393625133445,
+      "grad_norm": 0.36804965214464147,
+      "learning_rate": 9.859855047319732e-06,
+      "loss": 0.2553,
+      "step": 552
+    },
+    {
+      "epoch": 0.3373493975903614,
+      "grad_norm": 0.25636657805348956,
+      "learning_rate": 9.858601294888212e-06,
+      "loss": 0.2595,
+      "step": 553
+    },
+    {
+      "epoch": 0.3379594326673784,
+      "grad_norm": 0.25072273193978134,
+      "learning_rate": 9.85734203972599e-06,
+      "loss": 0.2472,
+      "step": 554
+    },
+    {
+      "epoch": 0.3385694677443953,
+      "grad_norm": 0.2721479101551272,
+      "learning_rate": 9.856077283259262e-06,
+      "loss": 0.2394,
+      "step": 555
+    },
+    {
+      "epoch": 0.33917950282141224,
+      "grad_norm": 0.26469111104432536,
+      "learning_rate": 9.85480702692046e-06,
+      "loss": 0.2499,
+      "step": 556
+    },
+    {
+      "epoch": 0.33978953789842914,
+      "grad_norm": 0.24605892848863048,
+      "learning_rate": 9.853531272148248e-06,
+      "loss": 0.2402,
+      "step": 557
+    },
+    {
+      "epoch": 0.3403995729754461,
+      "grad_norm": 0.2470370356204931,
+      "learning_rate": 9.852250020387513e-06,
+      "loss": 0.2396,
+      "step": 558
+    },
+    {
+      "epoch": 0.341009608052463,
+      "grad_norm": 0.27967529556696236,
+      "learning_rate": 9.850963273089365e-06,
+      "loss": 0.2592,
+      "step": 559
+    },
+    {
+      "epoch": 0.34161964312947996,
+      "grad_norm": 0.2550975214568574,
+      "learning_rate": 9.849671031711146e-06,
+      "loss": 0.2357,
+      "step": 560
+    },
+    {
+      "epoch": 0.34222967820649686,
+      "grad_norm": 0.23782560164722819,
+      "learning_rate": 9.848373297716414e-06,
+      "loss": 0.2302,
+      "step": 561
+    },
+    {
+      "epoch": 0.3428397132835138,
+      "grad_norm": 0.26172256616517786,
+      "learning_rate": 9.847070072574952e-06,
+      "loss": 0.261,
+      "step": 562
+    },
+    {
+      "epoch": 0.3434497483605307,
+      "grad_norm": 0.25376783721614055,
+      "learning_rate": 9.84576135776276e-06,
+      "loss": 0.2525,
+      "step": 563
+    },
+    {
+      "epoch": 0.3440597834375477,
+      "grad_norm": 0.25534231185707695,
+      "learning_rate": 9.844447154762054e-06,
+      "loss": 0.2581,
+      "step": 564
+    },
+    {
+      "epoch": 0.3446698185145646,
+      "grad_norm": 0.2514937599287347,
+      "learning_rate": 9.84312746506127e-06,
+      "loss": 0.2408,
+      "step": 565
+    },
+    {
+      "epoch": 0.34527985359158153,
+      "grad_norm": 0.23793239330047294,
+      "learning_rate": 9.841802290155054e-06,
+      "loss": 0.2457,
+      "step": 566
+    },
+    {
+      "epoch": 0.34588988866859843,
+      "grad_norm": 0.2498278422776726,
+      "learning_rate": 9.840471631544266e-06,
+      "loss": 0.2472,
+      "step": 567
+    },
+    {
+      "epoch": 0.3464999237456154,
+      "grad_norm": 0.25133312384225875,
+      "learning_rate": 9.839135490735978e-06,
+      "loss": 0.2525,
+      "step": 568
+    },
+    {
+      "epoch": 0.3471099588226323,
+      "grad_norm": 0.23975828367424365,
+      "learning_rate": 9.837793869243468e-06,
+      "loss": 0.2483,
+      "step": 569
+    },
+    {
+      "epoch": 0.34771999389964925,
+      "grad_norm": 0.23902175350837487,
+      "learning_rate": 9.836446768586225e-06,
+      "loss": 0.2414,
+      "step": 570
+    },
+    {
+      "epoch": 0.34833002897666615,
+      "grad_norm": 0.23786061397646546,
+      "learning_rate": 9.835094190289941e-06,
+      "loss": 0.2419,
+      "step": 571
+    },
+    {
+      "epoch": 0.3489400640536831,
+      "grad_norm": 0.23971167681501773,
+      "learning_rate": 9.833736135886513e-06,
+      "loss": 0.2432,
+      "step": 572
+    },
+    {
+      "epoch": 0.3495500991307,
+      "grad_norm": 0.2469864119024162,
+      "learning_rate": 9.832372606914038e-06,
+      "loss": 0.2495,
+      "step": 573
+    },
+    {
+      "epoch": 0.35016013420771697,
+      "grad_norm": 0.2516517284719589,
+      "learning_rate": 9.831003604916815e-06,
+      "loss": 0.2301,
+      "step": 574
+    },
+    {
+      "epoch": 0.35077016928473387,
+      "grad_norm": 0.2604419831002641,
+      "learning_rate": 9.829629131445342e-06,
+      "loss": 0.2583,
+      "step": 575
+    },
+    {
+      "epoch": 0.3513802043617508,
+      "grad_norm": 0.24036837488264234,
+      "learning_rate": 9.828249188056314e-06,
+      "loss": 0.2423,
+      "step": 576
+    },
+    {
+      "epoch": 0.3519902394387677,
+      "grad_norm": 0.2693073521625091,
+      "learning_rate": 9.826863776312621e-06,
+      "loss": 0.2515,
+      "step": 577
+    },
+    {
+      "epoch": 0.3526002745157847,
+      "grad_norm": 0.24574119339988224,
+      "learning_rate": 9.825472897783344e-06,
+      "loss": 0.2328,
+      "step": 578
+    },
+    {
+      "epoch": 0.3532103095928016,
+      "grad_norm": 0.2422586260824654,
+      "learning_rate": 9.824076554043759e-06,
+      "loss": 0.23,
+      "step": 579
+    },
+    {
+      "epoch": 0.3538203446698185,
+      "grad_norm": 0.24616937036065017,
+      "learning_rate": 9.822674746675329e-06,
+      "loss": 0.216,
+      "step": 580
+    },
+    {
+      "epoch": 0.35443037974683544,
+      "grad_norm": 0.2619722786538483,
+      "learning_rate": 9.821267477265705e-06,
+      "loss": 0.2529,
+      "step": 581
+    },
+    {
+      "epoch": 0.35504041482385235,
+      "grad_norm": 0.23474122964467006,
+      "learning_rate": 9.819854747408728e-06,
+      "loss": 0.2459,
+      "step": 582
+    },
+    {
+      "epoch": 0.3556504499008693,
+      "grad_norm": 0.25658578266252213,
+      "learning_rate": 9.81843655870442e-06,
+      "loss": 0.2407,
+      "step": 583
+    },
+    {
+      "epoch": 0.3562604849778862,
+      "grad_norm": 0.2543668048771176,
+      "learning_rate": 9.817012912758986e-06,
+      "loss": 0.2607,
+      "step": 584
+    },
+    {
+      "epoch": 0.35687052005490316,
+      "grad_norm": 0.24916680040867104,
+      "learning_rate": 9.815583811184809e-06,
+      "loss": 0.2371,
+      "step": 585
+    },
+    {
+      "epoch": 0.35748055513192006,
+      "grad_norm": 0.24253556661774894,
+      "learning_rate": 9.814149255600458e-06,
+      "loss": 0.2505,
+      "step": 586
+    },
+    {
+      "epoch": 0.358090590208937,
+      "grad_norm": 0.25098425841167166,
+      "learning_rate": 9.812709247630671e-06,
+      "loss": 0.2719,
+      "step": 587
+    },
+    {
+      "epoch": 0.3587006252859539,
+      "grad_norm": 0.2410722695307963,
+      "learning_rate": 9.81126378890637e-06,
+      "loss": 0.2537,
+      "step": 588
+    },
+    {
+      "epoch": 0.3593106603629709,
+      "grad_norm": 0.23987057965373262,
+      "learning_rate": 9.80981288106464e-06,
+      "loss": 0.2652,
+      "step": 589
+    },
+    {
+      "epoch": 0.3599206954399878,
+      "grad_norm": 0.24963404051174887,
+      "learning_rate": 9.808356525748748e-06,
+      "loss": 0.236,
+      "step": 590
+    },
+    {
+      "epoch": 0.36053073051700474,
+      "grad_norm": 0.2484565381470736,
+      "learning_rate": 9.806894724608122e-06,
+      "loss": 0.2628,
+      "step": 591
+    },
+    {
+      "epoch": 0.36114076559402164,
+      "grad_norm": 0.23634017497508783,
+      "learning_rate": 9.805427479298365e-06,
+      "loss": 0.2463,
+      "step": 592
+    },
+    {
+      "epoch": 0.3617508006710386,
+      "grad_norm": 0.2790490778566573,
+      "learning_rate": 9.803954791481239e-06,
+      "loss": 0.2687,
+      "step": 593
+    },
+    {
+      "epoch": 0.3623608357480555,
+      "grad_norm": 0.25749534513801714,
+      "learning_rate": 9.802476662824676e-06,
+      "loss": 0.2622,
+      "step": 594
+    },
+    {
+      "epoch": 0.36297087082507246,
+      "grad_norm": 0.24333551374135254,
+      "learning_rate": 9.800993095002767e-06,
+      "loss": 0.247,
+      "step": 595
+    },
+    {
+      "epoch": 0.36358090590208936,
+      "grad_norm": 0.24467156233879625,
+      "learning_rate": 9.799504089695762e-06,
+      "loss": 0.2414,
+      "step": 596
+    },
+    {
+      "epoch": 0.3641909409791063,
+      "grad_norm": 0.24529249172357293,
+      "learning_rate": 9.798009648590073e-06,
+      "loss": 0.2644,
+      "step": 597
+    },
+    {
+      "epoch": 0.3648009760561232,
+      "grad_norm": 0.281323360902479,
+      "learning_rate": 9.796509773378267e-06,
+      "loss": 0.2503,
+      "step": 598
+    },
+    {
+      "epoch": 0.3654110111331402,
+      "grad_norm": 0.24713120414672382,
+      "learning_rate": 9.795004465759067e-06,
+      "loss": 0.2581,
+      "step": 599
+    },
+    {
+      "epoch": 0.3660210462101571,
+      "grad_norm": 0.244640088774755,
+      "learning_rate": 9.793493727437343e-06,
+      "loss": 0.2415,
+      "step": 600
+    },
+    {
+      "epoch": 0.36663108128717403,
+      "grad_norm": 0.24741333198569834,
+      "learning_rate": 9.79197756012412e-06,
+      "loss": 0.2553,
+      "step": 601
+    },
+    {
+      "epoch": 0.36724111636419093,
+      "grad_norm": 0.2486624353021153,
+      "learning_rate": 9.790455965536574e-06,
+      "loss": 0.2425,
+      "step": 602
+    },
+    {
+      "epoch": 0.3678511514412079,
+      "grad_norm": 0.25239998228300475,
+      "learning_rate": 9.788928945398025e-06,
+      "loss": 0.2555,
+      "step": 603
+    },
+    {
+      "epoch": 0.3684611865182248,
+      "grad_norm": 0.2651978042658129,
+      "learning_rate": 9.787396501437934e-06,
+      "loss": 0.2413,
+      "step": 604
+    },
+    {
+      "epoch": 0.36907122159524175,
+      "grad_norm": 0.2404173861055526,
+      "learning_rate": 9.785858635391913e-06,
+      "loss": 0.2475,
+      "step": 605
+    },
+    {
+      "epoch": 0.36968125667225865,
+      "grad_norm": 0.23571440170965438,
+      "learning_rate": 9.78431534900171e-06,
+      "loss": 0.2533,
+      "step": 606
+    },
+    {
+      "epoch": 0.3702912917492756,
+      "grad_norm": 0.24233704202409498,
+      "learning_rate": 9.782766644015212e-06,
+      "loss": 0.2397,
+      "step": 607
+    },
+    {
+      "epoch": 0.3709013268262925,
+      "grad_norm": 0.2610196149576871,
+      "learning_rate": 9.781212522186442e-06,
+      "loss": 0.2421,
+      "step": 608
+    },
+    {
+      "epoch": 0.37151136190330947,
+      "grad_norm": 0.23835119864458468,
+      "learning_rate": 9.779652985275562e-06,
+      "loss": 0.2245,
+      "step": 609
+    },
+    {
+      "epoch": 0.37212139698032637,
+      "grad_norm": 0.23920196375213928,
+      "learning_rate": 9.778088035048866e-06,
+      "loss": 0.2632,
+      "step": 610
+    },
+    {
+      "epoch": 0.37273143205734327,
+      "grad_norm": 0.2649601180748752,
+      "learning_rate": 9.776517673278772e-06,
+      "loss": 0.263,
+      "step": 611
+    },
+    {
+      "epoch": 0.3733414671343602,
+      "grad_norm": 0.2542987280812582,
+      "learning_rate": 9.774941901743838e-06,
+      "loss": 0.2374,
+      "step": 612
+    },
+    {
+      "epoch": 0.37395150221137713,
+      "grad_norm": 0.24882298018650365,
+      "learning_rate": 9.773360722228742e-06,
+      "loss": 0.2457,
+      "step": 613
+    },
+    {
+      "epoch": 0.3745615372883941,
+      "grad_norm": 0.23412848045448256,
+      "learning_rate": 9.771774136524287e-06,
+      "loss": 0.2483,
+      "step": 614
+    },
+    {
+      "epoch": 0.375171572365411,
+      "grad_norm": 0.25754931185275737,
+      "learning_rate": 9.770182146427403e-06,
+      "loss": 0.2501,
+      "step": 615
+    },
+    {
+      "epoch": 0.37578160744242795,
+      "grad_norm": 0.24720643726302463,
+      "learning_rate": 9.768584753741134e-06,
+      "loss": 0.2336,
+      "step": 616
+    },
+    {
+      "epoch": 0.37639164251944485,
+      "grad_norm": 0.23429526920676347,
+      "learning_rate": 9.766981960274653e-06,
+      "loss": 0.2523,
+      "step": 617
+    },
+    {
+      "epoch": 0.3770016775964618,
+      "grad_norm": 0.23286582855824883,
+      "learning_rate": 9.76537376784324e-06,
+      "loss": 0.2507,
+      "step": 618
+    },
+    {
+      "epoch": 0.3776117126734787,
+      "grad_norm": 0.24213790165188145,
+      "learning_rate": 9.763760178268296e-06,
+      "loss": 0.2422,
+      "step": 619
+    },
+    {
+      "epoch": 0.37822174775049566,
+      "grad_norm": 0.23510019087838768,
+      "learning_rate": 9.762141193377329e-06,
+      "loss": 0.2207,
+      "step": 620
+    },
+    {
+      "epoch": 0.37883178282751256,
+      "grad_norm": 0.2468787558422412,
+      "learning_rate": 9.760516815003965e-06,
+      "loss": 0.2523,
+      "step": 621
+    },
+    {
+      "epoch": 0.3794418179045295,
+      "grad_norm": 0.3058221224160152,
+      "learning_rate": 9.758887044987929e-06,
+      "loss": 0.2443,
+      "step": 622
+    },
+    {
+      "epoch": 0.3800518529815464,
+      "grad_norm": 0.26093269918959,
+      "learning_rate": 9.757251885175063e-06,
+      "loss": 0.2678,
+      "step": 623
+    },
+    {
+      "epoch": 0.3806618880585634,
+      "grad_norm": 0.22969564204204285,
+      "learning_rate": 9.755611337417306e-06,
+      "loss": 0.2374,
+      "step": 624
+    },
+    {
+      "epoch": 0.3812719231355803,
+      "grad_norm": 0.23089944539257504,
+      "learning_rate": 9.753965403572703e-06,
+      "loss": 0.231,
+      "step": 625
+    },
+    {
+      "epoch": 0.38188195821259724,
+      "grad_norm": 0.2597564856286656,
+      "learning_rate": 9.752314085505396e-06,
+      "loss": 0.2576,
+      "step": 626
+    },
+    {
+      "epoch": 0.38249199328961414,
+      "grad_norm": 0.2531822089084447,
+      "learning_rate": 9.750657385085627e-06,
+      "loss": 0.2298,
+      "step": 627
+    },
+    {
+      "epoch": 0.3831020283666311,
+      "grad_norm": 0.2848691745062799,
+      "learning_rate": 9.748995304189734e-06,
+      "loss": 0.2518,
+      "step": 628
+    },
+    {
+      "epoch": 0.383712063443648,
+      "grad_norm": 0.24751375845503512,
+      "learning_rate": 9.747327844700147e-06,
+      "loss": 0.242,
+      "step": 629
+    },
+    {
+      "epoch": 0.38432209852066496,
+      "grad_norm": 0.2311387087802481,
+      "learning_rate": 9.745655008505392e-06,
+      "loss": 0.2535,
+      "step": 630
+    },
+    {
+      "epoch": 0.38493213359768186,
+      "grad_norm": 0.2412556673680029,
+      "learning_rate": 9.74397679750008e-06,
+      "loss": 0.2412,
+      "step": 631
+    },
+    {
+      "epoch": 0.3855421686746988,
+      "grad_norm": 0.24858957090381073,
+      "learning_rate": 9.74229321358491e-06,
+      "loss": 0.2623,
+      "step": 632
+    },
+    {
+      "epoch": 0.3861522037517157,
+      "grad_norm": 0.233987379984257,
+      "learning_rate": 9.740604258666668e-06,
+      "loss": 0.2426,
+      "step": 633
+    },
+    {
+      "epoch": 0.3867622388287327,
+      "grad_norm": 0.2529694629667342,
+      "learning_rate": 9.738909934658223e-06,
+      "loss": 0.251,
+      "step": 634
+    },
+    {
+      "epoch": 0.3873722739057496,
+      "grad_norm": 0.2508951547032724,
+      "learning_rate": 9.737210243478522e-06,
+      "loss": 0.254,
+      "step": 635
+    },
+    {
+      "epoch": 0.38798230898276653,
+      "grad_norm": 0.2356895929288683,
+      "learning_rate": 9.735505187052595e-06,
+      "loss": 0.2533,
+      "step": 636
+    },
+    {
+      "epoch": 0.38859234405978343,
+      "grad_norm": 0.27045617478946093,
+      "learning_rate": 9.733794767311545e-06,
+      "loss": 0.242,
+      "step": 637
+    },
+    {
+      "epoch": 0.3892023791368004,
+      "grad_norm": 0.2549042475520284,
+      "learning_rate": 9.732078986192552e-06,
+      "loss": 0.2471,
+      "step": 638
+    },
+    {
+      "epoch": 0.3898124142138173,
+      "grad_norm": 0.23978833581681586,
+      "learning_rate": 9.730357845638866e-06,
+      "loss": 0.2437,
+      "step": 639
+    },
+    {
+      "epoch": 0.39042244929083425,
+      "grad_norm": 0.2377076563680043,
+      "learning_rate": 9.72863134759981e-06,
+      "loss": 0.2443,
+      "step": 640
+    },
+    {
+      "epoch": 0.39103248436785115,
+      "grad_norm": 0.24192790263242697,
+      "learning_rate": 9.726899494030768e-06,
+      "loss": 0.2199,
+      "step": 641
+    },
+    {
+      "epoch": 0.39164251944486805,
+      "grad_norm": 0.2292328096544485,
+      "learning_rate": 9.725162286893197e-06,
+      "loss": 0.2402,
+      "step": 642
+    },
+    {
+      "epoch": 0.392252554521885,
+      "grad_norm": 0.2339555266898248,
+      "learning_rate": 9.723419728154617e-06,
+      "loss": 0.2467,
+      "step": 643
+    },
+    {
+      "epoch": 0.3928625895989019,
+      "grad_norm": 0.2542148373647154,
+      "learning_rate": 9.721671819788603e-06,
+      "loss": 0.2633,
+      "step": 644
+    },
+    {
+      "epoch": 0.39347262467591887,
+      "grad_norm": 0.24151750283902107,
+      "learning_rate": 9.719918563774793e-06,
+      "loss": 0.2212,
+      "step": 645
+    },
+    {
+      "epoch": 0.39408265975293577,
+      "grad_norm": 0.2430793021843514,
+      "learning_rate": 9.71815996209888e-06,
+      "loss": 0.2498,
+      "step": 646
+    },
+    {
+      "epoch": 0.39469269482995273,
+      "grad_norm": 0.252445657472582,
+      "learning_rate": 9.716396016752616e-06,
+      "loss": 0.2512,
+      "step": 647
+    },
+    {
+      "epoch": 0.39530272990696963,
+      "grad_norm": 0.2429734159537907,
+      "learning_rate": 9.7146267297338e-06,
+      "loss": 0.2601,
+      "step": 648
+    },
+    {
+      "epoch": 0.3959127649839866,
+      "grad_norm": 0.2582242931327393,
+      "learning_rate": 9.712852103046281e-06,
+      "loss": 0.2587,
+      "step": 649
+    },
+    {
+      "epoch": 0.3965228000610035,
+      "grad_norm": 0.26067550631747444,
+      "learning_rate": 9.711072138699957e-06,
+      "loss": 0.2501,
+      "step": 650
+    },
+    {
+      "epoch": 0.39713283513802045,
+      "grad_norm": 0.24554660526811284,
+      "learning_rate": 9.709286838710774e-06,
+      "loss": 0.2731,
+      "step": 651
+    },
+    {
+      "epoch": 0.39774287021503735,
+      "grad_norm": 0.23889263674607647,
+      "learning_rate": 9.707496205100714e-06,
+      "loss": 0.2485,
+      "step": 652
+    },
+    {
+      "epoch": 0.3983529052920543,
+      "grad_norm": 0.25991253213394694,
+      "learning_rate": 9.705700239897809e-06,
+      "loss": 0.2524,
+      "step": 653
+    },
+    {
+      "epoch": 0.3989629403690712,
+      "grad_norm": 0.2459360727540675,
+      "learning_rate": 9.70389894513612e-06,
+      "loss": 0.2523,
+      "step": 654
+    },
+    {
+      "epoch": 0.39957297544608816,
+      "grad_norm": 0.25005925280691177,
+      "learning_rate": 9.70209232285575e-06,
+      "loss": 0.2494,
+      "step": 655
+    },
+    {
+      "epoch": 0.40018301052310506,
+      "grad_norm": 0.2675096301368609,
+      "learning_rate": 9.700280375102835e-06,
+      "loss": 0.2611,
+      "step": 656
+    },
+    {
+      "epoch": 0.400793045600122,
+      "grad_norm": 0.2406185003357182,
+      "learning_rate": 9.698463103929542e-06,
+      "loss": 0.2437,
+      "step": 657
+    },
+    {
+      "epoch": 0.4014030806771389,
+      "grad_norm": 0.2641251121809679,
+      "learning_rate": 9.696640511394066e-06,
+      "loss": 0.2444,
+      "step": 658
+    },
+    {
+      "epoch": 0.4020131157541559,
+      "grad_norm": 0.24262195457977112,
+      "learning_rate": 9.694812599560632e-06,
+      "loss": 0.2508,
+      "step": 659
+    },
+    {
+      "epoch": 0.4026231508311728,
+      "grad_norm": 0.23038210084607058,
+      "learning_rate": 9.692979370499485e-06,
+      "loss": 0.2411,
+      "step": 660
+    },
+    {
+      "epoch": 0.40323318590818974,
+      "grad_norm": 0.24981753963500378,
+      "learning_rate": 9.691140826286893e-06,
+      "loss": 0.2684,
+      "step": 661
+    },
+    {
+      "epoch": 0.40384322098520664,
+      "grad_norm": 0.23408140036354103,
+      "learning_rate": 9.689296969005151e-06,
+      "loss": 0.2554,
+      "step": 662
+    },
+    {
+      "epoch": 0.4044532560622236,
+      "grad_norm": 0.24536784626888322,
+      "learning_rate": 9.68744780074256e-06,
+      "loss": 0.247,
+      "step": 663
+    },
+    {
+      "epoch": 0.4050632911392405,
+      "grad_norm": 0.24786305639021228,
+      "learning_rate": 9.685593323593445e-06,
+      "loss": 0.241,
+      "step": 664
+    },
+    {
+      "epoch": 0.40567332621625746,
+      "grad_norm": 0.2437211055920838,
+      "learning_rate": 9.68373353965814e-06,
+      "loss": 0.2563,
+      "step": 665
+    },
+    {
+      "epoch": 0.40628336129327436,
+      "grad_norm": 0.23494770231046405,
+      "learning_rate": 9.68186845104299e-06,
+      "loss": 0.2622,
+      "step": 666
+    },
+    {
+      "epoch": 0.4068933963702913,
+      "grad_norm": 0.2561543984617132,
+      "learning_rate": 9.679998059860347e-06,
+      "loss": 0.2498,
+      "step": 667
+    },
+    {
+      "epoch": 0.4075034314473082,
+      "grad_norm": 0.23142595032383642,
+      "learning_rate": 9.678122368228571e-06,
+      "loss": 0.2503,
+      "step": 668
+    },
+    {
+      "epoch": 0.4081134665243252,
+      "grad_norm": 0.23388256298097315,
+      "learning_rate": 9.676241378272022e-06,
+      "loss": 0.2611,
+      "step": 669
+    },
+    {
+      "epoch": 0.4087235016013421,
+      "grad_norm": 0.2171665951648235,
+      "learning_rate": 9.674355092121064e-06,
+      "loss": 0.2339,
+      "step": 670
+    },
+    {
+      "epoch": 0.409333536678359,
+      "grad_norm": 0.23523716756609286,
+      "learning_rate": 9.672463511912056e-06,
+      "loss": 0.2349,
+      "step": 671
+    },
+    {
+      "epoch": 0.40994357175537594,
+      "grad_norm": 0.26050667381938597,
+      "learning_rate": 9.670566639787355e-06,
+      "loss": 0.2588,
+      "step": 672
+    },
+    {
+      "epoch": 0.41055360683239284,
+      "grad_norm": 0.2568032147090647,
+      "learning_rate": 9.66866447789531e-06,
+      "loss": 0.2666,
+      "step": 673
+    },
+    {
+      "epoch": 0.4111636419094098,
+      "grad_norm": 0.23258591576948226,
+      "learning_rate": 9.666757028390267e-06,
+      "loss": 0.2453,
+      "step": 674
+    },
+    {
+      "epoch": 0.4117736769864267,
+      "grad_norm": 0.2532173686218808,
+      "learning_rate": 9.66484429343255e-06,
+      "loss": 0.2507,
+      "step": 675
+    },
+    {
+      "epoch": 0.41238371206344365,
+      "grad_norm": 0.24742853540044907,
+      "learning_rate": 9.662926275188478e-06,
+      "loss": 0.2469,
+      "step": 676
+    },
+    {
+      "epoch": 0.41299374714046055,
+      "grad_norm": 0.26148385144664343,
+      "learning_rate": 9.66100297583035e-06,
+      "loss": 0.2606,
+      "step": 677
+    },
+    {
+      "epoch": 0.4136037822174775,
+      "grad_norm": 0.242123360508457,
+      "learning_rate": 9.659074397536446e-06,
+      "loss": 0.2413,
+      "step": 678
+    },
+    {
+      "epoch": 0.4142138172944944,
+      "grad_norm": 0.25053386707994624,
+      "learning_rate": 9.657140542491025e-06,
+      "loss": 0.2598,
+      "step": 679
+    },
+    {
+      "epoch": 0.41482385237151137,
+      "grad_norm": 0.2266347754111311,
+      "learning_rate": 9.655201412884328e-06,
+      "loss": 0.2333,
+      "step": 680
+    },
+    {
+      "epoch": 0.41543388744852827,
+      "grad_norm": 0.29818873435724125,
+      "learning_rate": 9.65325701091256e-06,
+      "loss": 0.2582,
+      "step": 681
+    },
+    {
+      "epoch": 0.41604392252554523,
+      "grad_norm": 0.2658125962674758,
+      "learning_rate": 9.651307338777903e-06,
+      "loss": 0.2581,
+      "step": 682
+    },
+    {
+      "epoch": 0.41665395760256213,
+      "grad_norm": 0.2554674490027522,
+      "learning_rate": 9.649352398688506e-06,
+      "loss": 0.2738,
+      "step": 683
+    },
+    {
+      "epoch": 0.4172639926795791,
+      "grad_norm": 0.25453506037265533,
+      "learning_rate": 9.647392192858489e-06,
+      "loss": 0.234,
+      "step": 684
+    },
+    {
+      "epoch": 0.417874027756596,
+      "grad_norm": 0.25461734067333075,
+      "learning_rate": 9.645426723507929e-06,
+      "loss": 0.2424,
+      "step": 685
+    },
+    {
+      "epoch": 0.41848406283361295,
+      "grad_norm": 0.23888484597715387,
+      "learning_rate": 9.64345599286287e-06,
+      "loss": 0.2428,
+      "step": 686
+    },
+    {
+      "epoch": 0.41909409791062985,
+      "grad_norm": 0.2489790124226144,
+      "learning_rate": 9.64148000315531e-06,
+      "loss": 0.2418,
+      "step": 687
+    },
+    {
+      "epoch": 0.4197041329876468,
+      "grad_norm": 0.2678885780331205,
+      "learning_rate": 9.63949875662321e-06,
+      "loss": 0.2616,
+      "step": 688
+    },
+    {
+      "epoch": 0.4203141680646637,
+      "grad_norm": 0.2538936493974687,
+      "learning_rate": 9.637512255510475e-06,
+      "loss": 0.2534,
+      "step": 689
+    },
+    {
+      "epoch": 0.42092420314168066,
+      "grad_norm": 0.23909031877638567,
+      "learning_rate": 9.635520502066968e-06,
+      "loss": 0.2389,
+      "step": 690
+    },
+    {
+      "epoch": 0.42153423821869757,
+      "grad_norm": 0.25332515366600283,
+      "learning_rate": 9.633523498548502e-06,
+      "loss": 0.2502,
+      "step": 691
+    },
+    {
+      "epoch": 0.4221442732957145,
+      "grad_norm": 0.2391451036740086,
+      "learning_rate": 9.63152124721683e-06,
+      "loss": 0.2659,
+      "step": 692
+    },
+    {
+      "epoch": 0.4227543083727314,
+      "grad_norm": 0.2576670329623059,
+      "learning_rate": 9.629513750339656e-06,
+      "loss": 0.2621,
+      "step": 693
+    },
+    {
+      "epoch": 0.4233643434497484,
+      "grad_norm": 0.25480315338019477,
+      "learning_rate": 9.627501010190616e-06,
+      "loss": 0.2515,
+      "step": 694
+    },
+    {
+      "epoch": 0.4239743785267653,
+      "grad_norm": 0.23458259674144666,
+      "learning_rate": 9.625483029049295e-06,
+      "loss": 0.2643,
+      "step": 695
+    },
+    {
+      "epoch": 0.42458441360378224,
+      "grad_norm": 0.24782100843445132,
+      "learning_rate": 9.623459809201201e-06,
+      "loss": 0.2412,
+      "step": 696
+    },
+    {
+      "epoch": 0.42519444868079914,
+      "grad_norm": 0.2629037328093925,
+      "learning_rate": 9.62143135293779e-06,
+      "loss": 0.2435,
+      "step": 697
+    },
+    {
+      "epoch": 0.4258044837578161,
+      "grad_norm": 0.23441780135241397,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 0.2275,
+      "step": 698
+    },
+    {
+      "epoch": 0.426414518834833,
+      "grad_norm": 0.25570881804535683,
+      "learning_rate": 9.617358740360446e-06,
+      "loss": 0.2376,
+      "step": 699
+    },
+    {
+      "epoch": 0.42702455391184996,
+      "grad_norm": 0.2575592928975001,
+      "learning_rate": 9.615314588659054e-06,
+      "loss": 0.2375,
+      "step": 700
+    },
+    {
+      "epoch": 0.42763458898886686,
+      "grad_norm": 0.26485082967353496,
+      "learning_rate": 9.613265209767417e-06,
+      "loss": 0.218,
+      "step": 701
+    },
+    {
+      "epoch": 0.42824462406588376,
+      "grad_norm": 0.26790743064582434,
+      "learning_rate": 9.611210606006606e-06,
+      "loss": 0.2398,
+      "step": 702
+    },
+    {
+      "epoch": 0.4288546591429007,
+      "grad_norm": 0.2327699012795595,
+      "learning_rate": 9.60915077970362e-06,
+      "loss": 0.2458,
+      "step": 703
+    },
+    {
+      "epoch": 0.4294646942199176,
+      "grad_norm": 0.25209047510843435,
+      "learning_rate": 9.607085733191362e-06,
+      "loss": 0.2484,
+      "step": 704
+    },
+    {
+      "epoch": 0.4300747292969346,
+      "grad_norm": 0.24532855214020663,
+      "learning_rate": 9.605015468808651e-06,
+      "loss": 0.2241,
+      "step": 705
+    },
+    {
+      "epoch": 0.4306847643739515,
+      "grad_norm": 0.23707628308383044,
+      "learning_rate": 9.602939988900223e-06,
+      "loss": 0.2118,
+      "step": 706
+    },
+    {
+      "epoch": 0.43129479945096844,
+      "grad_norm": 0.2717265067792356,
+      "learning_rate": 9.600859295816708e-06,
+      "loss": 0.2365,
+      "step": 707
+    },
+    {
+      "epoch": 0.43190483452798534,
+      "grad_norm": 0.24187592973175528,
+      "learning_rate": 9.598773391914655e-06,
+      "loss": 0.2528,
+      "step": 708
+    },
+    {
+      "epoch": 0.4325148696050023,
+      "grad_norm": 0.24095158717754256,
+      "learning_rate": 9.596682279556499e-06,
+      "loss": 0.2316,
+      "step": 709
+    },
+    {
+      "epoch": 0.4331249046820192,
+      "grad_norm": 0.2384804389224989,
+      "learning_rate": 9.594585961110586e-06,
+      "loss": 0.2195,
+      "step": 710
+    },
+    {
+      "epoch": 0.43373493975903615,
+      "grad_norm": 0.22980754276143053,
+      "learning_rate": 9.59248443895115e-06,
+      "loss": 0.2197,
+      "step": 711
+    },
+    {
+      "epoch": 0.43434497483605305,
+      "grad_norm": 0.23000195088140646,
+      "learning_rate": 9.590377715458328e-06,
+      "loss": 0.2132,
+      "step": 712
+    },
+    {
+      "epoch": 0.43495500991307,
+      "grad_norm": 0.24348636008374336,
+      "learning_rate": 9.588265793018141e-06,
+      "loss": 0.2598,
+      "step": 713
+    },
+    {
+      "epoch": 0.4355650449900869,
+      "grad_norm": 0.24462607007817821,
+      "learning_rate": 9.586148674022498e-06,
+      "loss": 0.2716,
+      "step": 714
+    },
+    {
+      "epoch": 0.43617508006710387,
+      "grad_norm": 0.24199303136022748,
+      "learning_rate": 9.584026360869195e-06,
+      "loss": 0.2577,
+      "step": 715
+    },
+    {
+      "epoch": 0.4367851151441208,
+      "grad_norm": 0.2444394229369308,
+      "learning_rate": 9.581898855961911e-06,
+      "loss": 0.2409,
+      "step": 716
+    },
+    {
+      "epoch": 0.43739515022113773,
+      "grad_norm": 0.22421862207213614,
+      "learning_rate": 9.579766161710209e-06,
+      "loss": 0.2193,
+      "step": 717
+    },
+    {
+      "epoch": 0.43800518529815463,
+      "grad_norm": 0.25355224565574364,
+      "learning_rate": 9.577628280529519e-06,
+      "loss": 0.2433,
+      "step": 718
+    },
+    {
+      "epoch": 0.4386152203751716,
+      "grad_norm": 0.24034777143737782,
+      "learning_rate": 9.575485214841158e-06,
+      "loss": 0.2358,
+      "step": 719
+    },
+    {
+      "epoch": 0.4392252554521885,
+      "grad_norm": 0.23465493475078136,
+      "learning_rate": 9.573336967072304e-06,
+      "loss": 0.2584,
+      "step": 720
+    },
+    {
+      "epoch": 0.43983529052920545,
+      "grad_norm": 0.22722796480040341,
+      "learning_rate": 9.571183539656011e-06,
+      "loss": 0.2401,
+      "step": 721
+    },
+    {
+      "epoch": 0.44044532560622235,
+      "grad_norm": 0.2627819359298546,
+      "learning_rate": 9.569024935031198e-06,
+      "loss": 0.2439,
+      "step": 722
+    },
+    {
+      "epoch": 0.4410553606832393,
+      "grad_norm": 0.24326571420142928,
+      "learning_rate": 9.566861155642646e-06,
+      "loss": 0.2381,
+      "step": 723
+    },
+    {
+      "epoch": 0.4416653957602562,
+      "grad_norm": 0.23545903061020426,
+      "learning_rate": 9.564692203940997e-06,
+      "loss": 0.2525,
+      "step": 724
+    },
+    {
+      "epoch": 0.44227543083727316,
+      "grad_norm": 0.24436023364810228,
+      "learning_rate": 9.562518082382751e-06,
+      "loss": 0.2426,
+      "step": 725
+    },
+    {
+      "epoch": 0.44288546591429007,
+      "grad_norm": 0.24538160588010707,
+      "learning_rate": 9.560338793430266e-06,
+      "loss": 0.2364,
+      "step": 726
+    },
+    {
+      "epoch": 0.443495500991307,
+      "grad_norm": 0.2275693745015377,
+      "learning_rate": 9.558154339551748e-06,
+      "loss": 0.2425,
+      "step": 727
+    },
+    {
+      "epoch": 0.4441055360683239,
+      "grad_norm": 0.256020016141183,
+      "learning_rate": 9.555964723221258e-06,
+      "loss": 0.27,
+      "step": 728
+    },
+    {
+      "epoch": 0.4447155711453409,
+      "grad_norm": 0.27300315086725535,
+      "learning_rate": 9.553769946918698e-06,
+      "loss": 0.2575,
+      "step": 729
+    },
+    {
+      "epoch": 0.4453256062223578,
+      "grad_norm": 0.23385109804562323,
+      "learning_rate": 9.551570013129819e-06,
+      "loss": 0.2208,
+      "step": 730
+    },
+    {
+      "epoch": 0.44593564129937474,
+      "grad_norm": 0.23386938978555916,
+      "learning_rate": 9.54936492434621e-06,
+      "loss": 0.2359,
+      "step": 731
+    },
+    {
+      "epoch": 0.44654567637639164,
+      "grad_norm": 0.2404558436293708,
+      "learning_rate": 9.547154683065298e-06,
+      "loss": 0.2435,
+      "step": 732
+    },
+    {
+      "epoch": 0.44715571145340854,
+      "grad_norm": 0.24500388350627236,
+      "learning_rate": 9.544939291790352e-06,
+      "loss": 0.2415,
+      "step": 733
+    },
+    {
+      "epoch": 0.4477657465304255,
+      "grad_norm": 0.23691618297887454,
+      "learning_rate": 9.542718753030463e-06,
+      "loss": 0.2236,
+      "step": 734
+    },
+    {
+      "epoch": 0.4483757816074424,
+      "grad_norm": 0.24857806187672893,
+      "learning_rate": 9.540493069300563e-06,
+      "loss": 0.2485,
+      "step": 735
+    },
+    {
+      "epoch": 0.44898581668445936,
+      "grad_norm": 0.2382926637523894,
+      "learning_rate": 9.538262243121403e-06,
+      "loss": 0.2386,
+      "step": 736
+    },
+    {
+      "epoch": 0.44959585176147626,
+      "grad_norm": 0.24030488684741594,
+      "learning_rate": 9.536026277019562e-06,
+      "loss": 0.2353,
+      "step": 737
+    },
+    {
+      "epoch": 0.4502058868384932,
+      "grad_norm": 0.23339236344005018,
+      "learning_rate": 9.533785173527438e-06,
+      "loss": 0.2347,
+      "step": 738
+    },
+    {
+      "epoch": 0.4508159219155101,
+      "grad_norm": 0.2299604201218471,
+      "learning_rate": 9.531538935183252e-06,
+      "loss": 0.2274,
+      "step": 739
+    },
+    {
+      "epoch": 0.4514259569925271,
+      "grad_norm": 0.2301101735723042,
+      "learning_rate": 9.529287564531034e-06,
+      "loss": 0.2379,
+      "step": 740
+    },
+    {
+      "epoch": 0.452035992069544,
+      "grad_norm": 0.24777348892591716,
+      "learning_rate": 9.527031064120632e-06,
+      "loss": 0.2158,
+      "step": 741
+    },
+    {
+      "epoch": 0.45264602714656094,
+      "grad_norm": 0.2382250577730699,
+      "learning_rate": 9.524769436507703e-06,
+      "loss": 0.257,
+      "step": 742
+    },
+    {
+      "epoch": 0.45325606222357784,
+      "grad_norm": 0.23491855965377412,
+      "learning_rate": 9.522502684253709e-06,
+      "loss": 0.2374,
+      "step": 743
+    },
+    {
+      "epoch": 0.4538660973005948,
+      "grad_norm": 0.2381867196722309,
+      "learning_rate": 9.520230809925917e-06,
+      "loss": 0.2415,
+      "step": 744
+    },
+    {
+      "epoch": 0.4544761323776117,
+      "grad_norm": 0.23464500865994656,
+      "learning_rate": 9.517953816097396e-06,
+      "loss": 0.252,
+      "step": 745
+    },
+    {
+      "epoch": 0.45508616745462865,
+      "grad_norm": 0.23333387158301092,
+      "learning_rate": 9.515671705347012e-06,
+      "loss": 0.2315,
+      "step": 746
+    },
+    {
+      "epoch": 0.45569620253164556,
+      "grad_norm": 0.2511717945420412,
+      "learning_rate": 9.513384480259427e-06,
+      "loss": 0.2442,
+      "step": 747
+    },
+    {
+      "epoch": 0.4563062376086625,
+      "grad_norm": 0.2476626774783739,
+      "learning_rate": 9.511092143425093e-06,
+      "loss": 0.2407,
+      "step": 748
+    },
+    {
+      "epoch": 0.4569162726856794,
+      "grad_norm": 0.25129087071494044,
+      "learning_rate": 9.508794697440257e-06,
+      "loss": 0.2677,
+      "step": 749
+    },
+    {
+      "epoch": 0.45752630776269637,
+      "grad_norm": 0.25628919436974645,
+      "learning_rate": 9.506492144906949e-06,
+      "loss": 0.2609,
+      "step": 750
+    },
+    {
+      "epoch": 0.4581363428397133,
+      "grad_norm": 0.24629875943849752,
+      "learning_rate": 9.50418448843298e-06,
+      "loss": 0.2318,
+      "step": 751
+    },
+    {
+      "epoch": 0.45874637791673023,
+      "grad_norm": 0.2291062788929306,
+      "learning_rate": 9.501871730631944e-06,
+      "loss": 0.2361,
+      "step": 752
+    },
+    {
+      "epoch": 0.45935641299374713,
+      "grad_norm": 0.22973777787593416,
+      "learning_rate": 9.499553874123213e-06,
+      "loss": 0.2352,
+      "step": 753
+    },
+    {
+      "epoch": 0.4599664480707641,
+      "grad_norm": 0.23346050639344187,
+      "learning_rate": 9.497230921531938e-06,
+      "loss": 0.246,
+      "step": 754
+    },
+    {
+      "epoch": 0.460576483147781,
+      "grad_norm": 0.23581281213008645,
+      "learning_rate": 9.494902875489031e-06,
+      "loss": 0.2576,
+      "step": 755
+    },
+    {
+      "epoch": 0.46118651822479795,
+      "grad_norm": 0.2496519595616449,
+      "learning_rate": 9.492569738631182e-06,
+      "loss": 0.2328,
+      "step": 756
+    },
+    {
+      "epoch": 0.46179655330181485,
+      "grad_norm": 0.24711405854515872,
+      "learning_rate": 9.490231513600842e-06,
+      "loss": 0.2448,
+      "step": 757
+    },
+    {
+      "epoch": 0.4624065883788318,
+      "grad_norm": 0.2629979912664831,
+      "learning_rate": 9.487888203046232e-06,
+      "loss": 0.2282,
+      "step": 758
+    },
+    {
+      "epoch": 0.4630166234558487,
+      "grad_norm": 0.23229398082709177,
+      "learning_rate": 9.485539809621319e-06,
+      "loss": 0.2529,
+      "step": 759
+    },
+    {
+      "epoch": 0.46362665853286567,
+      "grad_norm": 0.2351650163696824,
+      "learning_rate": 9.48318633598584e-06,
+      "loss": 0.2423,
+      "step": 760
+    },
+    {
+      "epoch": 0.46423669360988257,
+      "grad_norm": 0.22334068920688743,
+      "learning_rate": 9.480827784805278e-06,
+      "loss": 0.2503,
+      "step": 761
+    },
+    {
+      "epoch": 0.4648467286868995,
+      "grad_norm": 0.23324497333424316,
+      "learning_rate": 9.478464158750873e-06,
+      "loss": 0.2598,
+      "step": 762
+    },
+    {
+      "epoch": 0.4654567637639164,
+      "grad_norm": 0.2376429511505979,
+      "learning_rate": 9.476095460499604e-06,
+      "loss": 0.2426,
+      "step": 763
+    },
+    {
+      "epoch": 0.4660667988409333,
+      "grad_norm": 0.23127830303431715,
+      "learning_rate": 9.473721692734203e-06,
+      "loss": 0.2293,
+      "step": 764
+    },
+    {
+      "epoch": 0.4666768339179503,
+      "grad_norm": 0.2279735382560775,
+      "learning_rate": 9.47134285814314e-06,
+      "loss": 0.232,
+      "step": 765
+    },
+    {
+      "epoch": 0.4672868689949672,
+      "grad_norm": 0.24060612490706318,
+      "learning_rate": 9.468958959420622e-06,
+      "loss": 0.2395,
+      "step": 766
+    },
+    {
+      "epoch": 0.46789690407198414,
+      "grad_norm": 0.24073558221650648,
+      "learning_rate": 9.466569999266595e-06,
+      "loss": 0.2362,
+      "step": 767
+    },
+    {
+      "epoch": 0.46850693914900104,
+      "grad_norm": 0.25234148310686183,
+      "learning_rate": 9.464175980386735e-06,
+      "loss": 0.2364,
+      "step": 768
+    },
+    {
+      "epoch": 0.469116974226018,
+      "grad_norm": 0.24301381464309002,
+      "learning_rate": 9.461776905492446e-06,
+      "loss": 0.2541,
+      "step": 769
+    },
+    {
+      "epoch": 0.4697270093030349,
+      "grad_norm": 0.2181646957236186,
+      "learning_rate": 9.459372777300863e-06,
+      "loss": 0.2315,
+      "step": 770
+    },
+    {
+      "epoch": 0.47033704438005186,
+      "grad_norm": 0.24408126431197386,
+      "learning_rate": 9.456963598534843e-06,
+      "loss": 0.2635,
+      "step": 771
+    },
+    {
+      "epoch": 0.47094707945706876,
+      "grad_norm": 0.2536793988988464,
+      "learning_rate": 9.454549371922958e-06,
+      "loss": 0.245,
+      "step": 772
+    },
+    {
+      "epoch": 0.4715571145340857,
+      "grad_norm": 0.23467096465196274,
+      "learning_rate": 9.452130100199504e-06,
+      "loss": 0.2282,
+      "step": 773
+    },
+    {
+      "epoch": 0.4721671496111026,
+      "grad_norm": 0.23975568079817508,
+      "learning_rate": 9.449705786104486e-06,
+      "loss": 0.2628,
+      "step": 774
+    },
+    {
+      "epoch": 0.4727771846881196,
+      "grad_norm": 0.2531867022216451,
+      "learning_rate": 9.447276432383622e-06,
+      "loss": 0.2539,
+      "step": 775
+    },
+    {
+      "epoch": 0.4733872197651365,
+      "grad_norm": 0.23729965993164157,
+      "learning_rate": 9.44484204178834e-06,
+      "loss": 0.2431,
+      "step": 776
+    },
+    {
+      "epoch": 0.47399725484215344,
+      "grad_norm": 0.23266240192528784,
+      "learning_rate": 9.442402617075765e-06,
+      "loss": 0.2485,
+      "step": 777
+    },
+    {
+      "epoch": 0.47460728991917034,
+      "grad_norm": 0.4593727203802515,
+      "learning_rate": 9.439958161008733e-06,
+      "loss": 0.2467,
+      "step": 778
+    },
+    {
+      "epoch": 0.4752173249961873,
+      "grad_norm": 0.2423521825841295,
+      "learning_rate": 9.437508676355774e-06,
+      "loss": 0.2515,
+      "step": 779
+    },
+    {
+      "epoch": 0.4758273600732042,
+      "grad_norm": 0.23786940493723296,
+      "learning_rate": 9.43505416589111e-06,
+      "loss": 0.2582,
+      "step": 780
+    },
+    {
+      "epoch": 0.47643739515022115,
+      "grad_norm": 0.24386754510833997,
+      "learning_rate": 9.43259463239466e-06,
+      "loss": 0.2364,
+      "step": 781
+    },
+    {
+      "epoch": 0.47704743022723806,
+      "grad_norm": 0.24902097785836147,
+      "learning_rate": 9.43013007865203e-06,
+      "loss": 0.214,
+      "step": 782
+    },
+    {
+      "epoch": 0.477657465304255,
+      "grad_norm": 0.23911865820557984,
+      "learning_rate": 9.427660507454515e-06,
+      "loss": 0.2443,
+      "step": 783
+    },
+    {
+      "epoch": 0.4782675003812719,
+      "grad_norm": 0.2557095629339037,
+      "learning_rate": 9.425185921599085e-06,
+      "loss": 0.2402,
+      "step": 784
+    },
+    {
+      "epoch": 0.47887753545828887,
+      "grad_norm": 0.2521433109274837,
+      "learning_rate": 9.422706323888398e-06,
+      "loss": 0.2578,
+      "step": 785
+    },
+    {
+      "epoch": 0.4794875705353058,
+      "grad_norm": 0.2365601659243881,
+      "learning_rate": 9.420221717130783e-06,
+      "loss": 0.2491,
+      "step": 786
+    },
+    {
+      "epoch": 0.48009760561232273,
+      "grad_norm": 0.24468779134408927,
+      "learning_rate": 9.417732104140243e-06,
+      "loss": 0.2361,
+      "step": 787
+    },
+    {
+      "epoch": 0.48070764068933963,
+      "grad_norm": 0.25848472600184347,
+      "learning_rate": 9.415237487736452e-06,
+      "loss": 0.2455,
+      "step": 788
+    },
+    {
+      "epoch": 0.4813176757663566,
+      "grad_norm": 0.24659632129132983,
+      "learning_rate": 9.412737870744752e-06,
+      "loss": 0.259,
+      "step": 789
+    },
+    {
+      "epoch": 0.4819277108433735,
+      "grad_norm": 0.2376635659932075,
+      "learning_rate": 9.410233255996146e-06,
+      "loss": 0.2486,
+      "step": 790
+    },
+    {
+      "epoch": 0.48253774592039045,
+      "grad_norm": 0.26330890674932006,
+      "learning_rate": 9.407723646327299e-06,
+      "loss": 0.2724,
+      "step": 791
+    },
+    {
+      "epoch": 0.48314778099740735,
+      "grad_norm": 0.2965239250982198,
+      "learning_rate": 9.40520904458053e-06,
+      "loss": 0.2246,
+      "step": 792
+    },
+    {
+      "epoch": 0.48375781607442425,
+      "grad_norm": 0.2462468999949858,
+      "learning_rate": 9.402689453603815e-06,
+      "loss": 0.2877,
+      "step": 793
+    },
+    {
+      "epoch": 0.4843678511514412,
+      "grad_norm": 0.390541205009224,
+      "learning_rate": 9.400164876250781e-06,
+      "loss": 0.2237,
+      "step": 794
+    },
+    {
+      "epoch": 0.4849778862284581,
+      "grad_norm": 0.2487914992956595,
+      "learning_rate": 9.3976353153807e-06,
+      "loss": 0.2456,
+      "step": 795
+    },
+    {
+      "epoch": 0.48558792130547507,
+      "grad_norm": 0.22828415306478353,
+      "learning_rate": 9.395100773858492e-06,
+      "loss": 0.24,
+      "step": 796
+    },
+    {
+      "epoch": 0.48619795638249197,
+      "grad_norm": 0.2651952079542631,
+      "learning_rate": 9.392561254554712e-06,
+      "loss": 0.2479,
+      "step": 797
+    },
+    {
+      "epoch": 0.4868079914595089,
+      "grad_norm": 0.2186420215795485,
+      "learning_rate": 9.39001676034556e-06,
+      "loss": 0.2062,
+      "step": 798
+    },
+    {
+      "epoch": 0.48741802653652583,
+      "grad_norm": 0.23488406613309,
+      "learning_rate": 9.387467294112864e-06,
+      "loss": 0.2374,
+      "step": 799
+    },
+    {
+      "epoch": 0.4880280616135428,
+      "grad_norm": 0.2335215343559726,
+      "learning_rate": 9.384912858744088e-06,
+      "loss": 0.252,
+      "step": 800
+    },
+    {
+      "epoch": 0.4886380966905597,
+      "grad_norm": 0.24953746869431825,
+      "learning_rate": 9.382353457132318e-06,
+      "loss": 0.2427,
+      "step": 801
+    },
+    {
+      "epoch": 0.48924813176757664,
+      "grad_norm": 0.2328812255084704,
+      "learning_rate": 9.379789092176275e-06,
+      "loss": 0.229,
+      "step": 802
+    },
+    {
+      "epoch": 0.48985816684459355,
+      "grad_norm": 0.22737943989850504,
+      "learning_rate": 9.377219766780288e-06,
+      "loss": 0.2239,
+      "step": 803
+    },
+    {
+      "epoch": 0.4904682019216105,
+      "grad_norm": 0.23159191713799232,
+      "learning_rate": 9.374645483854315e-06,
+      "loss": 0.2438,
+      "step": 804
+    },
+    {
+      "epoch": 0.4910782369986274,
+      "grad_norm": 0.22823543143777628,
+      "learning_rate": 9.372066246313922e-06,
+      "loss": 0.2145,
+      "step": 805
+    },
+    {
+      "epoch": 0.49168827207564436,
+      "grad_norm": 0.24969394403631143,
+      "learning_rate": 9.369482057080293e-06,
+      "loss": 0.2089,
+      "step": 806
+    },
+    {
+      "epoch": 0.49229830715266126,
+      "grad_norm": 0.23891653993332385,
+      "learning_rate": 9.366892919080213e-06,
+      "loss": 0.2488,
+      "step": 807
+    },
+    {
+      "epoch": 0.4929083422296782,
+      "grad_norm": 0.24089267103927142,
+      "learning_rate": 9.364298835246074e-06,
+      "loss": 0.2459,
+      "step": 808
+    },
+    {
+      "epoch": 0.4935183773066951,
+      "grad_norm": 0.2489514639584053,
+      "learning_rate": 9.361699808515877e-06,
+      "loss": 0.2434,
+      "step": 809
+    },
+    {
+      "epoch": 0.4941284123837121,
+      "grad_norm": 0.2184131072697668,
+      "learning_rate": 9.359095841833206e-06,
+      "loss": 0.2084,
+      "step": 810
+    },
+    {
+      "epoch": 0.494738447460729,
+      "grad_norm": 0.23816642896153215,
+      "learning_rate": 9.356486938147256e-06,
+      "loss": 0.2271,
+      "step": 811
+    },
+    {
+      "epoch": 0.49534848253774594,
+      "grad_norm": 0.23819781990228103,
+      "learning_rate": 9.353873100412805e-06,
+      "loss": 0.2371,
+      "step": 812
+    },
+    {
+      "epoch": 0.49595851761476284,
+      "grad_norm": 0.23618169911622136,
+      "learning_rate": 9.351254331590216e-06,
+      "loss": 0.2276,
+      "step": 813
+    },
+    {
+      "epoch": 0.4965685526917798,
+      "grad_norm": 0.26104490763327737,
+      "learning_rate": 9.348630634645446e-06,
+      "loss": 0.2497,
+      "step": 814
+    },
+    {
+      "epoch": 0.4971785877687967,
+      "grad_norm": 0.2271620479721002,
+      "learning_rate": 9.346002012550027e-06,
+      "loss": 0.2464,
+      "step": 815
+    },
+    {
+      "epoch": 0.49778862284581366,
+      "grad_norm": 0.2323705973707938,
+      "learning_rate": 9.34336846828107e-06,
+      "loss": 0.2491,
+      "step": 816
+    },
+    {
+      "epoch": 0.49839865792283056,
+      "grad_norm": 0.21723754684916077,
+      "learning_rate": 9.340730004821266e-06,
+      "loss": 0.224,
+      "step": 817
+    },
+    {
+      "epoch": 0.4990086929998475,
+      "grad_norm": 0.23734482056770834,
+      "learning_rate": 9.338086625158867e-06,
+      "loss": 0.2391,
+      "step": 818
+    },
+    {
+      "epoch": 0.4996187280768644,
+      "grad_norm": 0.22493990824961632,
+      "learning_rate": 9.3354383322877e-06,
+      "loss": 0.2414,
+      "step": 819
+    },
+    {
+      "epoch": 0.5002287631538813,
+      "grad_norm": 0.23763953888935077,
+      "learning_rate": 9.33278512920716e-06,
+      "loss": 0.257,
+      "step": 820
+    },
+    {
+      "epoch": 0.5008387982308983,
+      "grad_norm": 0.23206692778749136,
+      "learning_rate": 9.330127018922195e-06,
+      "loss": 0.2389,
+      "step": 821
+    },
+    {
+      "epoch": 0.5014488333079152,
+      "grad_norm": 0.22541539391517743,
+      "learning_rate": 9.327464004443315e-06,
+      "loss": 0.2336,
+      "step": 822
+    },
+    {
+      "epoch": 0.5020588683849322,
+      "grad_norm": 0.24256078896560065,
+      "learning_rate": 9.324796088786581e-06,
+      "loss": 0.2429,
+      "step": 823
+    },
+    {
+      "epoch": 0.502668903461949,
+      "grad_norm": 0.2468862949593829,
+      "learning_rate": 9.322123274973613e-06,
+      "loss": 0.2355,
+      "step": 824
+    },
+    {
+      "epoch": 0.503278938538966,
+      "grad_norm": 0.22384411091337553,
+      "learning_rate": 9.31944556603157e-06,
+      "loss": 0.2213,
+      "step": 825
+    },
+    {
+      "epoch": 0.503888973615983,
+      "grad_norm": 0.2514036453108826,
+      "learning_rate": 9.31676296499316e-06,
+      "loss": 0.2463,
+      "step": 826
+    },
+    {
+      "epoch": 0.5044990086929998,
+      "grad_norm": 0.24925114054301487,
+      "learning_rate": 9.314075474896631e-06,
+      "loss": 0.2551,
+      "step": 827
+    },
+    {
+      "epoch": 0.5051090437700168,
+      "grad_norm": 0.23276254658106843,
+      "learning_rate": 9.311383098785765e-06,
+      "loss": 0.2332,
+      "step": 828
+    },
+    {
+      "epoch": 0.5057190788470337,
+      "grad_norm": 0.2370474417278097,
+      "learning_rate": 9.308685839709878e-06,
+      "loss": 0.2613,
+      "step": 829
+    },
+    {
+      "epoch": 0.5063291139240507,
+      "grad_norm": 0.22872370038160483,
+      "learning_rate": 9.305983700723824e-06,
+      "loss": 0.2372,
+      "step": 830
+    },
+    {
+      "epoch": 0.5069391490010675,
+      "grad_norm": 0.24716515266866768,
+      "learning_rate": 9.303276684887973e-06,
+      "loss": 0.2552,
+      "step": 831
+    },
+    {
+      "epoch": 0.5075491840780845,
+      "grad_norm": 0.2265318455600586,
+      "learning_rate": 9.30056479526823e-06,
+      "loss": 0.2368,
+      "step": 832
+    },
+    {
+      "epoch": 0.5081592191551014,
+      "grad_norm": 0.23946821858214107,
+      "learning_rate": 9.297848034936007e-06,
+      "loss": 0.2269,
+      "step": 833
+    },
+    {
+      "epoch": 0.5087692542321184,
+      "grad_norm": 0.26397442672414717,
+      "learning_rate": 9.295126406968241e-06,
+      "loss": 0.2524,
+      "step": 834
+    },
+    {
+      "epoch": 0.5093792893091352,
+      "grad_norm": 0.21612815204717434,
+      "learning_rate": 9.292399914447381e-06,
+      "loss": 0.2177,
+      "step": 835
+    },
+    {
+      "epoch": 0.5099893243861522,
+      "grad_norm": 0.23050449682089552,
+      "learning_rate": 9.289668560461385e-06,
+      "loss": 0.2382,
+      "step": 836
+    },
+    {
+      "epoch": 0.5105993594631691,
+      "grad_norm": 0.25879227486567125,
+      "learning_rate": 9.286932348103716e-06,
+      "loss": 0.2786,
+      "step": 837
+    },
+    {
+      "epoch": 0.5112093945401861,
+      "grad_norm": 0.23581274681256434,
+      "learning_rate": 9.284191280473338e-06,
+      "loss": 0.2475,
+      "step": 838
+    },
+    {
+      "epoch": 0.511819429617203,
+      "grad_norm": 0.23997499104246758,
+      "learning_rate": 9.281445360674717e-06,
+      "loss": 0.2421,
+      "step": 839
+    },
+    {
+      "epoch": 0.5124294646942199,
+      "grad_norm": 0.23196299343759688,
+      "learning_rate": 9.278694591817814e-06,
+      "loss": 0.2537,
+      "step": 840
+    },
+    {
+      "epoch": 0.5130394997712369,
+      "grad_norm": 0.23865479598143985,
+      "learning_rate": 9.275938977018082e-06,
+      "loss": 0.2429,
+      "step": 841
+    },
+    {
+      "epoch": 0.5136495348482538,
+      "grad_norm": 0.22539671688187207,
+      "learning_rate": 9.273178519396459e-06,
+      "loss": 0.235,
+      "step": 842
+    },
+    {
+      "epoch": 0.5142595699252707,
+      "grad_norm": 0.2529599698610769,
+      "learning_rate": 9.270413222079373e-06,
+      "loss": 0.2516,
+      "step": 843
+    },
+    {
+      "epoch": 0.5148696050022876,
+      "grad_norm": 0.257432749177655,
+      "learning_rate": 9.26764308819873e-06,
+      "loss": 0.251,
+      "step": 844
+    },
+    {
+      "epoch": 0.5154796400793046,
+      "grad_norm": 0.23477208951640044,
+      "learning_rate": 9.264868120891913e-06,
+      "loss": 0.2448,
+      "step": 845
+    },
+    {
+      "epoch": 0.5160896751563215,
+      "grad_norm": 0.2599692203117331,
+      "learning_rate": 9.262088323301782e-06,
+      "loss": 0.2395,
+      "step": 846
+    },
+    {
+      "epoch": 0.5166997102333384,
+      "grad_norm": 0.2391225469089055,
+      "learning_rate": 9.259303698576669e-06,
+      "loss": 0.2352,
+      "step": 847
+    },
+    {
+      "epoch": 0.5173097453103553,
+      "grad_norm": 0.2610324632293431,
+      "learning_rate": 9.256514249870366e-06,
+      "loss": 0.242,
+      "step": 848
+    },
+    {
+      "epoch": 0.5179197803873723,
+      "grad_norm": 0.22258363931378589,
+      "learning_rate": 9.253719980342134e-06,
+      "loss": 0.2384,
+      "step": 849
+    },
+    {
+      "epoch": 0.5185298154643893,
+      "grad_norm": 0.32551326135551306,
+      "learning_rate": 9.250920893156696e-06,
+      "loss": 0.2443,
+      "step": 850
+    },
+    {
+      "epoch": 0.5191398505414061,
+      "grad_norm": 0.23679924925359405,
+      "learning_rate": 9.24811699148423e-06,
+      "loss": 0.23,
+      "step": 851
+    },
+    {
+      "epoch": 0.5197498856184231,
+      "grad_norm": 0.22408751438114305,
+      "learning_rate": 9.245308278500355e-06,
+      "loss": 0.2163,
+      "step": 852
+    },
+    {
+      "epoch": 0.52035992069544,
+      "grad_norm": 0.21958835705262147,
+      "learning_rate": 9.24249475738616e-06,
+      "loss": 0.2357,
+      "step": 853
+    },
+    {
+      "epoch": 0.5209699557724569,
+      "grad_norm": 0.23911714756619634,
+      "learning_rate": 9.239676431328164e-06,
+      "loss": 0.2427,
+      "step": 854
+    },
+    {
+      "epoch": 0.5215799908494738,
+      "grad_norm": 0.22342842695193937,
+      "learning_rate": 9.236853303518333e-06,
+      "loss": 0.2375,
+      "step": 855
+    },
+    {
+      "epoch": 0.5221900259264908,
+      "grad_norm": 0.27681328656512555,
+      "learning_rate": 9.234025377154073e-06,
+      "loss": 0.2572,
+      "step": 856
+    },
+    {
+      "epoch": 0.5228000610035077,
+      "grad_norm": 0.25175601110086143,
+      "learning_rate": 9.231192655438222e-06,
+      "loss": 0.2491,
+      "step": 857
+    },
+    {
+      "epoch": 0.5234100960805246,
+      "grad_norm": 0.2637545918682464,
+      "learning_rate": 9.228355141579048e-06,
+      "loss": 0.2643,
+      "step": 858
+    },
+    {
+      "epoch": 0.5240201311575415,
+      "grad_norm": 0.24678037303000705,
+      "learning_rate": 9.225512838790252e-06,
+      "loss": 0.2295,
+      "step": 859
+    },
+    {
+      "epoch": 0.5246301662345585,
+      "grad_norm": 0.25271929900504264,
+      "learning_rate": 9.222665750290953e-06,
+      "loss": 0.2332,
+      "step": 860
+    },
+    {
+      "epoch": 0.5252402013115754,
+      "grad_norm": 0.2346919500276136,
+      "learning_rate": 9.219813879305692e-06,
+      "loss": 0.2612,
+      "step": 861
+    },
+    {
+      "epoch": 0.5258502363885923,
+      "grad_norm": 0.25570687887348353,
+      "learning_rate": 9.21695722906443e-06,
+      "loss": 0.2395,
+      "step": 862
+    },
+    {
+      "epoch": 0.5264602714656093,
+      "grad_norm": 0.27891063472335015,
+      "learning_rate": 9.214095802802533e-06,
+      "loss": 0.25,
+      "step": 863
+    },
+    {
+      "epoch": 0.5270703065426262,
+      "grad_norm": 0.23748539072986685,
+      "learning_rate": 9.211229603760787e-06,
+      "loss": 0.2344,
+      "step": 864
+    },
+    {
+      "epoch": 0.5276803416196432,
+      "grad_norm": 0.22278988238334246,
+      "learning_rate": 9.208358635185372e-06,
+      "loss": 0.2276,
+      "step": 865
+    },
+    {
+      "epoch": 0.52829037669666,
+      "grad_norm": 0.2661064237307424,
+      "learning_rate": 9.20548290032788e-06,
+      "loss": 0.2404,
+      "step": 866
+    },
+    {
+      "epoch": 0.528900411773677,
+      "grad_norm": 0.2482715675689319,
+      "learning_rate": 9.202602402445294e-06,
+      "loss": 0.231,
+      "step": 867
+    },
+    {
+      "epoch": 0.5295104468506939,
+      "grad_norm": 0.23627964937286555,
+      "learning_rate": 9.199717144799994e-06,
+      "loss": 0.2182,
+      "step": 868
+    },
+    {
+      "epoch": 0.5301204819277109,
+      "grad_norm": 0.23312268728117894,
+      "learning_rate": 9.196827130659752e-06,
+      "loss": 0.231,
+      "step": 869
+    },
+    {
+      "epoch": 0.5307305170047277,
+      "grad_norm": 0.2521885753183132,
+      "learning_rate": 9.193932363297724e-06,
+      "loss": 0.2152,
+      "step": 870
+    },
+    {
+      "epoch": 0.5313405520817447,
+      "grad_norm": 0.2572040177255539,
+      "learning_rate": 9.191032845992453e-06,
+      "loss": 0.231,
+      "step": 871
+    },
+    {
+      "epoch": 0.5319505871587616,
+      "grad_norm": 0.24710288355302004,
+      "learning_rate": 9.18812858202786e-06,
+      "loss": 0.2555,
+      "step": 872
+    },
+    {
+      "epoch": 0.5325606222357786,
+      "grad_norm": 0.2486931954461719,
+      "learning_rate": 9.185219574693242e-06,
+      "loss": 0.2306,
+      "step": 873
+    },
+    {
+      "epoch": 0.5331706573127954,
+      "grad_norm": 0.2542190718905183,
+      "learning_rate": 9.18230582728327e-06,
+      "loss": 0.271,
+      "step": 874
+    },
+    {
+      "epoch": 0.5337806923898124,
+      "grad_norm": 0.24410915290177013,
+      "learning_rate": 9.179387343097978e-06,
+      "loss": 0.2576,
+      "step": 875
+    },
+    {
+      "epoch": 0.5343907274668294,
+      "grad_norm": 0.22953871125837036,
+      "learning_rate": 9.17646412544277e-06,
+      "loss": 0.2488,
+      "step": 876
+    },
+    {
+      "epoch": 0.5350007625438463,
+      "grad_norm": 0.25812105965097,
+      "learning_rate": 9.17353617762841e-06,
+      "loss": 0.2596,
+      "step": 877
+    },
+    {
+      "epoch": 0.5356107976208632,
+      "grad_norm": 0.26246862819412475,
+      "learning_rate": 9.170603502971017e-06,
+      "loss": 0.2496,
+      "step": 878
+    },
+    {
+      "epoch": 0.5362208326978801,
+      "grad_norm": 0.22330376240227373,
+      "learning_rate": 9.167666104792067e-06,
+      "loss": 0.2296,
+      "step": 879
+    },
+    {
+      "epoch": 0.5368308677748971,
+      "grad_norm": 0.23681969799356747,
+      "learning_rate": 9.164723986418385e-06,
+      "loss": 0.2319,
+      "step": 880
+    },
+    {
+      "epoch": 0.537440902851914,
+      "grad_norm": 0.2445777883564648,
+      "learning_rate": 9.161777151182137e-06,
+      "loss": 0.2382,
+      "step": 881
+    },
+    {
+      "epoch": 0.5380509379289309,
+      "grad_norm": 0.240934911695689,
+      "learning_rate": 9.158825602420838e-06,
+      "loss": 0.2514,
+      "step": 882
+    },
+    {
+      "epoch": 0.5386609730059478,
+      "grad_norm": 0.23990828086840085,
+      "learning_rate": 9.155869343477339e-06,
+      "loss": 0.2412,
+      "step": 883
+    },
+    {
+      "epoch": 0.5392710080829648,
+      "grad_norm": 0.22730222183887502,
+      "learning_rate": 9.152908377699823e-06,
+      "loss": 0.2334,
+      "step": 884
+    },
+    {
+      "epoch": 0.5398810431599816,
+      "grad_norm": 0.22530806978627918,
+      "learning_rate": 9.149942708441808e-06,
+      "loss": 0.2341,
+      "step": 885
+    },
+    {
+      "epoch": 0.5404910782369986,
+      "grad_norm": 0.24120868240608,
+      "learning_rate": 9.146972339062136e-06,
+      "loss": 0.2541,
+      "step": 886
+    },
+    {
+      "epoch": 0.5411011133140156,
+      "grad_norm": 0.308352315358742,
+      "learning_rate": 9.143997272924974e-06,
+      "loss": 0.2423,
+      "step": 887
+    },
+    {
+      "epoch": 0.5417111483910325,
+      "grad_norm": 0.25467840429327515,
+      "learning_rate": 9.141017513399806e-06,
+      "loss": 0.2497,
+      "step": 888
+    },
+    {
+      "epoch": 0.5423211834680494,
+      "grad_norm": 0.2422144432264907,
+      "learning_rate": 9.138033063861436e-06,
+      "loss": 0.2685,
+      "step": 889
+    },
+    {
+      "epoch": 0.5429312185450663,
+      "grad_norm": 0.25084723479258747,
+      "learning_rate": 9.135043927689975e-06,
+      "loss": 0.2681,
+      "step": 890
+    },
+    {
+      "epoch": 0.5435412536220833,
+      "grad_norm": 0.2496126574784138,
+      "learning_rate": 9.132050108270845e-06,
+      "loss": 0.2412,
+      "step": 891
+    },
+    {
+      "epoch": 0.5441512886991002,
+      "grad_norm": 0.24937237576609259,
+      "learning_rate": 9.129051608994773e-06,
+      "loss": 0.2455,
+      "step": 892
+    },
+    {
+      "epoch": 0.5447613237761171,
+      "grad_norm": 0.24410548774952506,
+      "learning_rate": 9.12604843325778e-06,
+      "loss": 0.2531,
+      "step": 893
+    },
+    {
+      "epoch": 0.545371358853134,
+      "grad_norm": 0.25039944675489567,
+      "learning_rate": 9.123040584461192e-06,
+      "loss": 0.2581,
+      "step": 894
+    },
+    {
+      "epoch": 0.545981393930151,
+      "grad_norm": 0.24217578431818285,
+      "learning_rate": 9.120028066011621e-06,
+      "loss": 0.2642,
+      "step": 895
+    },
+    {
+      "epoch": 0.546591429007168,
+      "grad_norm": 0.23799501463139952,
+      "learning_rate": 9.117010881320973e-06,
+      "loss": 0.2445,
+      "step": 896
+    },
+    {
+      "epoch": 0.5472014640841848,
+      "grad_norm": 0.2328066882782319,
+      "learning_rate": 9.113989033806434e-06,
+      "loss": 0.2446,
+      "step": 897
+    },
+    {
+      "epoch": 0.5478114991612018,
+      "grad_norm": 0.2256128476716643,
+      "learning_rate": 9.110962526890474e-06,
+      "loss": 0.2261,
+      "step": 898
+    },
+    {
+      "epoch": 0.5484215342382187,
+      "grad_norm": 0.24277658311529213,
+      "learning_rate": 9.10793136400084e-06,
+      "loss": 0.2533,
+      "step": 899
+    },
+    {
+      "epoch": 0.5490315693152357,
+      "grad_norm": 0.2244964492427359,
+      "learning_rate": 9.104895548570549e-06,
+      "loss": 0.2417,
+      "step": 900
+    },
+    {
+      "epoch": 0.5496416043922525,
+      "grad_norm": 0.23817530380584634,
+      "learning_rate": 9.101855084037893e-06,
+      "loss": 0.2299,
+      "step": 901
+    },
+    {
+      "epoch": 0.5502516394692695,
+      "grad_norm": 0.21462378141591873,
+      "learning_rate": 9.098809973846425e-06,
+      "loss": 0.2282,
+      "step": 902
+    },
+    {
+      "epoch": 0.5508616745462864,
+      "grad_norm": 0.22887797228368942,
+      "learning_rate": 9.09576022144496e-06,
+      "loss": 0.2188,
+      "step": 903
+    },
+    {
+      "epoch": 0.5514717096233034,
+      "grad_norm": 0.2317966772564381,
+      "learning_rate": 9.092705830287572e-06,
+      "loss": 0.2339,
+      "step": 904
+    },
+    {
+      "epoch": 0.5520817447003202,
+      "grad_norm": 0.23477031220665048,
+      "learning_rate": 9.089646803833589e-06,
+      "loss": 0.2334,
+      "step": 905
+    },
+    {
+      "epoch": 0.5526917797773372,
+      "grad_norm": 0.24063497257307356,
+      "learning_rate": 9.086583145547588e-06,
+      "loss": 0.238,
+      "step": 906
+    },
+    {
+      "epoch": 0.5533018148543541,
+      "grad_norm": 0.24278984054747293,
+      "learning_rate": 9.083514858899391e-06,
+      "loss": 0.2422,
+      "step": 907
+    },
+    {
+      "epoch": 0.5539118499313711,
+      "grad_norm": 0.24698518564289468,
+      "learning_rate": 9.080441947364065e-06,
+      "loss": 0.2354,
+      "step": 908
+    },
+    {
+      "epoch": 0.554521885008388,
+      "grad_norm": 0.2573974858694405,
+      "learning_rate": 9.07736441442191e-06,
+      "loss": 0.2366,
+      "step": 909
+    },
+    {
+      "epoch": 0.5551319200854049,
+      "grad_norm": 0.23686076398195488,
+      "learning_rate": 9.074282263558465e-06,
+      "loss": 0.2332,
+      "step": 910
+    },
+    {
+      "epoch": 0.5557419551624219,
+      "grad_norm": 0.26716589782054984,
+      "learning_rate": 9.071195498264497e-06,
+      "loss": 0.23,
+      "step": 911
+    },
+    {
+      "epoch": 0.5563519902394388,
+      "grad_norm": 0.23254352826520663,
+      "learning_rate": 9.068104122036e-06,
+      "loss": 0.2261,
+      "step": 912
+    },
+    {
+      "epoch": 0.5569620253164557,
+      "grad_norm": 0.23774506033550974,
+      "learning_rate": 9.06500813837419e-06,
+      "loss": 0.2538,
+      "step": 913
+    },
+    {
+      "epoch": 0.5575720603934726,
+      "grad_norm": 0.24578732684387927,
+      "learning_rate": 9.061907550785498e-06,
+      "loss": 0.2267,
+      "step": 914
+    },
+    {
+      "epoch": 0.5581820954704896,
+      "grad_norm": 0.24273173792130215,
+      "learning_rate": 9.058802362781576e-06,
+      "loss": 0.2429,
+      "step": 915
+    },
+    {
+      "epoch": 0.5587921305475064,
+      "grad_norm": 0.24871249511209675,
+      "learning_rate": 9.055692577879279e-06,
+      "loss": 0.2645,
+      "step": 916
+    },
+    {
+      "epoch": 0.5594021656245234,
+      "grad_norm": 0.2489675357704421,
+      "learning_rate": 9.052578199600675e-06,
+      "loss": 0.2641,
+      "step": 917
+    },
+    {
+      "epoch": 0.5600122007015403,
+      "grad_norm": 0.2348175485183629,
+      "learning_rate": 9.049459231473028e-06,
+      "loss": 0.2179,
+      "step": 918
+    },
+    {
+      "epoch": 0.5606222357785573,
+      "grad_norm": 0.24516386553768676,
+      "learning_rate": 9.046335677028806e-06,
+      "loss": 0.2398,
+      "step": 919
+    },
+    {
+      "epoch": 0.5612322708555741,
+      "grad_norm": 0.2539618785828311,
+      "learning_rate": 9.04320753980567e-06,
+      "loss": 0.2443,
+      "step": 920
+    },
+    {
+      "epoch": 0.5618423059325911,
+      "grad_norm": 0.2375763291210803,
+      "learning_rate": 9.040074823346466e-06,
+      "loss": 0.2319,
+      "step": 921
+    },
+    {
+      "epoch": 0.5624523410096081,
+      "grad_norm": 0.2295019349307068,
+      "learning_rate": 9.036937531199235e-06,
+      "loss": 0.2174,
+      "step": 922
+    },
+    {
+      "epoch": 0.563062376086625,
+      "grad_norm": 0.23913641148482256,
+      "learning_rate": 9.033795666917191e-06,
+      "loss": 0.2336,
+      "step": 923
+    },
+    {
+      "epoch": 0.5636724111636419,
+      "grad_norm": 0.2365015694994553,
+      "learning_rate": 9.030649234058738e-06,
+      "loss": 0.2288,
+      "step": 924
+    },
+    {
+      "epoch": 0.5642824462406588,
+      "grad_norm": 0.2269035714506007,
+      "learning_rate": 9.02749823618744e-06,
+      "loss": 0.2359,
+      "step": 925
+    },
+    {
+      "epoch": 0.5648924813176758,
+      "grad_norm": 0.23862880675583445,
+      "learning_rate": 9.024342676872044e-06,
+      "loss": 0.2321,
+      "step": 926
+    },
+    {
+      "epoch": 0.5655025163946927,
+      "grad_norm": 0.24866509535913342,
+      "learning_rate": 9.021182559686454e-06,
+      "loss": 0.2517,
+      "step": 927
+    },
+    {
+      "epoch": 0.5661125514717096,
+      "grad_norm": 0.22871794221013542,
+      "learning_rate": 9.018017888209744e-06,
+      "loss": 0.2467,
+      "step": 928
+    },
+    {
+      "epoch": 0.5667225865487265,
+      "grad_norm": 0.25107482056399233,
+      "learning_rate": 9.014848666026138e-06,
+      "loss": 0.2661,
+      "step": 929
+    },
+    {
+      "epoch": 0.5673326216257435,
+      "grad_norm": 0.23633368721157266,
+      "learning_rate": 9.01167489672502e-06,
+      "loss": 0.2463,
+      "step": 930
+    },
+    {
+      "epoch": 0.5679426567027605,
+      "grad_norm": 0.2469957443857779,
+      "learning_rate": 9.008496583900925e-06,
+      "loss": 0.2485,
+      "step": 931
+    },
+    {
+      "epoch": 0.5685526917797773,
+      "grad_norm": 0.23023949616042658,
+      "learning_rate": 9.005313731153525e-06,
+      "loss": 0.2401,
+      "step": 932
+    },
+    {
+      "epoch": 0.5691627268567943,
+      "grad_norm": 0.2324877334752998,
+      "learning_rate": 9.002126342087643e-06,
+      "loss": 0.2297,
+      "step": 933
+    },
+    {
+      "epoch": 0.5697727619338112,
+      "grad_norm": 0.234321661561038,
+      "learning_rate": 8.998934420313236e-06,
+      "loss": 0.2384,
+      "step": 934
+    },
+    {
+      "epoch": 0.5703827970108282,
+      "grad_norm": 0.24123765281463405,
+      "learning_rate": 8.995737969445395e-06,
+      "loss": 0.2352,
+      "step": 935
+    },
+    {
+      "epoch": 0.570992832087845,
+      "grad_norm": 0.23544270953963511,
+      "learning_rate": 8.992536993104339e-06,
+      "loss": 0.2304,
+      "step": 936
+    },
+    {
+      "epoch": 0.571602867164862,
+      "grad_norm": 0.24273749530372107,
+      "learning_rate": 8.989331494915417e-06,
+      "loss": 0.2376,
+      "step": 937
+    },
+    {
+      "epoch": 0.5722129022418789,
+      "grad_norm": 0.23238771153358787,
+      "learning_rate": 8.986121478509096e-06,
+      "loss": 0.2325,
+      "step": 938
+    },
+    {
+      "epoch": 0.5728229373188959,
+      "grad_norm": 0.25536548436456125,
+      "learning_rate": 8.982906947520958e-06,
+      "loss": 0.2471,
+      "step": 939
+    },
+    {
+      "epoch": 0.5734329723959127,
+      "grad_norm": 0.2277984497148008,
+      "learning_rate": 8.9796879055917e-06,
+      "loss": 0.2329,
+      "step": 940
+    },
+    {
+      "epoch": 0.5740430074729297,
+      "grad_norm": 0.3157076986257636,
+      "learning_rate": 8.976464356367133e-06,
+      "loss": 0.2472,
+      "step": 941
+    },
+    {
+      "epoch": 0.5746530425499466,
+      "grad_norm": 0.2431944807189335,
+      "learning_rate": 8.973236303498165e-06,
+      "loss": 0.2265,
+      "step": 942
+    },
+    {
+      "epoch": 0.5752630776269636,
+      "grad_norm": 0.2415392811651247,
+      "learning_rate": 8.97000375064081e-06,
+      "loss": 0.2401,
+      "step": 943
+    },
+    {
+      "epoch": 0.5758731127039804,
+      "grad_norm": 0.24697514473461818,
+      "learning_rate": 8.966766701456177e-06,
+      "loss": 0.2481,
+      "step": 944
+    },
+    {
+      "epoch": 0.5764831477809974,
+      "grad_norm": 0.2310906025904227,
+      "learning_rate": 8.963525159610465e-06,
+      "loss": 0.2204,
+      "step": 945
+    },
+    {
+      "epoch": 0.5770931828580144,
+      "grad_norm": 0.3167379712129428,
+      "learning_rate": 8.960279128774965e-06,
+      "loss": 0.2319,
+      "step": 946
+    },
+    {
+      "epoch": 0.5777032179350312,
+      "grad_norm": 0.2532736663817971,
+      "learning_rate": 8.957028612626051e-06,
+      "loss": 0.2571,
+      "step": 947
+    },
+    {
+      "epoch": 0.5783132530120482,
+      "grad_norm": 0.2709052682297909,
+      "learning_rate": 8.953773614845175e-06,
+      "loss": 0.2432,
+      "step": 948
+    },
+    {
+      "epoch": 0.5789232880890651,
+      "grad_norm": 0.24875501207670353,
+      "learning_rate": 8.950514139118868e-06,
+      "loss": 0.2381,
+      "step": 949
+    },
+    {
+      "epoch": 0.5795333231660821,
+      "grad_norm": 0.2386346874806699,
+      "learning_rate": 8.947250189138732e-06,
+      "loss": 0.2541,
+      "step": 950
+    },
+    {
+      "epoch": 0.5801433582430989,
+      "grad_norm": 0.23503372535899722,
+      "learning_rate": 8.943981768601431e-06,
+      "loss": 0.2399,
+      "step": 951
+    },
+    {
+      "epoch": 0.5807533933201159,
+      "grad_norm": 0.23484091677557625,
+      "learning_rate": 8.9407088812087e-06,
+      "loss": 0.2385,
+      "step": 952
+    },
+    {
+      "epoch": 0.5813634283971328,
+      "grad_norm": 0.2786973828276919,
+      "learning_rate": 8.937431530667329e-06,
+      "loss": 0.2768,
+      "step": 953
+    },
+    {
+      "epoch": 0.5819734634741498,
+      "grad_norm": 0.24501647742704377,
+      "learning_rate": 8.934149720689164e-06,
+      "loss": 0.2405,
+      "step": 954
+    },
+    {
+      "epoch": 0.5825834985511666,
+      "grad_norm": 0.24403169887634135,
+      "learning_rate": 8.9308634549911e-06,
+      "loss": 0.2548,
+      "step": 955
+    },
+    {
+      "epoch": 0.5831935336281836,
+      "grad_norm": 0.2272567853866675,
+      "learning_rate": 8.927572737295081e-06,
+      "loss": 0.2418,
+      "step": 956
+    },
+    {
+      "epoch": 0.5838035687052006,
+      "grad_norm": 0.24563126522644446,
+      "learning_rate": 8.924277571328091e-06,
+      "loss": 0.2529,
+      "step": 957
+    },
+    {
+      "epoch": 0.5844136037822175,
+      "grad_norm": 0.2364803637356654,
+      "learning_rate": 8.92097796082215e-06,
+      "loss": 0.2267,
+      "step": 958
+    },
+    {
+      "epoch": 0.5850236388592344,
+      "grad_norm": 0.24044223758674313,
+      "learning_rate": 8.917673909514321e-06,
+      "loss": 0.2749,
+      "step": 959
+    },
+    {
+      "epoch": 0.5856336739362513,
+      "grad_norm": 0.2235920684260497,
+      "learning_rate": 8.914365421146685e-06,
+      "loss": 0.2375,
+      "step": 960
+    },
+    {
+      "epoch": 0.5862437090132683,
+      "grad_norm": 0.24233427434900542,
+      "learning_rate": 8.911052499466358e-06,
+      "loss": 0.2574,
+      "step": 961
+    },
+    {
+      "epoch": 0.5868537440902852,
+      "grad_norm": 0.2199417231731091,
+      "learning_rate": 8.907735148225465e-06,
+      "loss": 0.2321,
+      "step": 962
+    },
+    {
+      "epoch": 0.5874637791673021,
+      "grad_norm": 0.24156632473583037,
+      "learning_rate": 8.904413371181164e-06,
+      "loss": 0.2362,
+      "step": 963
+    },
+    {
+      "epoch": 0.588073814244319,
+      "grad_norm": 0.24319106385211212,
+      "learning_rate": 8.90108717209561e-06,
+      "loss": 0.2524,
+      "step": 964
+    },
+    {
+      "epoch": 0.588683849321336,
+      "grad_norm": 0.21780477785194519,
+      "learning_rate": 8.897756554735976e-06,
+      "loss": 0.2375,
+      "step": 965
+    },
+    {
+      "epoch": 0.589293884398353,
+      "grad_norm": 0.2631983155577083,
+      "learning_rate": 8.894421522874438e-06,
+      "loss": 0.2531,
+      "step": 966
+    },
+    {
+      "epoch": 0.5899039194753698,
+      "grad_norm": 0.24907326563896523,
+      "learning_rate": 8.891082080288167e-06,
+      "loss": 0.2501,
+      "step": 967
+    },
+    {
+      "epoch": 0.5905139545523868,
+      "grad_norm": 0.2183747192980442,
+      "learning_rate": 8.887738230759334e-06,
+      "loss": 0.2476,
+      "step": 968
+    },
+    {
+      "epoch": 0.5911239896294037,
+      "grad_norm": 0.2429805274522966,
+      "learning_rate": 8.884389978075098e-06,
+      "loss": 0.2394,
+      "step": 969
+    },
+    {
+      "epoch": 0.5917340247064207,
+      "grad_norm": 0.26143482773747845,
+      "learning_rate": 8.88103732602761e-06,
+      "loss": 0.2501,
+      "step": 970
+    },
+    {
+      "epoch": 0.5923440597834375,
+      "grad_norm": 0.24008352338552205,
+      "learning_rate": 8.877680278413995e-06,
+      "loss": 0.2301,
+      "step": 971
+    },
+    {
+      "epoch": 0.5929540948604545,
+      "grad_norm": 0.2504323744037465,
+      "learning_rate": 8.874318839036367e-06,
+      "loss": 0.2632,
+      "step": 972
+    },
+    {
+      "epoch": 0.5935641299374714,
+      "grad_norm": 0.22502293301141857,
+      "learning_rate": 8.870953011701804e-06,
+      "loss": 0.2138,
+      "step": 973
+    },
+    {
+      "epoch": 0.5941741650144884,
+      "grad_norm": 0.25060434823535316,
+      "learning_rate": 8.867582800222358e-06,
+      "loss": 0.216,
+      "step": 974
+    },
+    {
+      "epoch": 0.5947842000915052,
+      "grad_norm": 0.22294734491003537,
+      "learning_rate": 8.864208208415051e-06,
+      "loss": 0.2355,
+      "step": 975
+    },
+    {
+      "epoch": 0.5953942351685222,
+      "grad_norm": 0.23476612585041143,
+      "learning_rate": 8.860829240101858e-06,
+      "loss": 0.2454,
+      "step": 976
+    },
+    {
+      "epoch": 0.5960042702455391,
+      "grad_norm": 0.24478825993869993,
+      "learning_rate": 8.857445899109716e-06,
+      "loss": 0.2362,
+      "step": 977
+    },
+    {
+      "epoch": 0.596614305322556,
+      "grad_norm": 0.23675812972527574,
+      "learning_rate": 8.85405818927051e-06,
+      "loss": 0.2433,
+      "step": 978
+    },
+    {
+      "epoch": 0.597224340399573,
+      "grad_norm": 0.23289797925718586,
+      "learning_rate": 8.85066611442108e-06,
+      "loss": 0.2616,
+      "step": 979
+    },
+    {
+      "epoch": 0.5978343754765899,
+      "grad_norm": 0.22975309494536794,
+      "learning_rate": 8.847269678403202e-06,
+      "loss": 0.2372,
+      "step": 980
+    },
+    {
+      "epoch": 0.5984444105536069,
+      "grad_norm": 0.25014127869116004,
+      "learning_rate": 8.843868885063594e-06,
+      "loss": 0.2481,
+      "step": 981
+    },
+    {
+      "epoch": 0.5990544456306237,
+      "grad_norm": 0.235261338075581,
+      "learning_rate": 8.840463738253913e-06,
+      "loss": 0.224,
+      "step": 982
+    },
+    {
+      "epoch": 0.5996644807076407,
+      "grad_norm": 0.263864345048999,
+      "learning_rate": 8.83705424183074e-06,
+      "loss": 0.2315,
+      "step": 983
+    },
+    {
+      "epoch": 0.6002745157846576,
+      "grad_norm": 0.2313037025839392,
+      "learning_rate": 8.83364039965559e-06,
+      "loss": 0.2343,
+      "step": 984
+    },
+    {
+      "epoch": 0.6008845508616746,
+      "grad_norm": 0.2316483778451406,
+      "learning_rate": 8.83022221559489e-06,
+      "loss": 0.2446,
+      "step": 985
+    },
+    {
+      "epoch": 0.6014945859386914,
+      "grad_norm": 0.2635286913095841,
+      "learning_rate": 8.826799693519996e-06,
+      "loss": 0.2477,
+      "step": 986
+    },
+    {
+      "epoch": 0.6021046210157084,
+      "grad_norm": 0.2402482306391233,
+      "learning_rate": 8.823372837307167e-06,
+      "loss": 0.2513,
+      "step": 987
+    },
+    {
+      "epoch": 0.6027146560927253,
+      "grad_norm": 0.2402515762798396,
+      "learning_rate": 8.819941650837576e-06,
+      "loss": 0.2607,
+      "step": 988
+    },
+    {
+      "epoch": 0.6033246911697423,
+      "grad_norm": 0.24283381310704727,
+      "learning_rate": 8.8165061379973e-06,
+      "loss": 0.2366,
+      "step": 989
+    },
+    {
+      "epoch": 0.6039347262467591,
+      "grad_norm": 0.25297933475508455,
+      "learning_rate": 8.813066302677314e-06,
+      "loss": 0.2363,
+      "step": 990
+    },
+    {
+      "epoch": 0.6045447613237761,
+      "grad_norm": 0.2449648651812992,
+      "learning_rate": 8.809622148773492e-06,
+      "loss": 0.2499,
+      "step": 991
+    },
+    {
+      "epoch": 0.6051547964007931,
+      "grad_norm": 0.26539602479579855,
+      "learning_rate": 8.806173680186594e-06,
+      "loss": 0.2663,
+      "step": 992
+    },
+    {
+      "epoch": 0.60576483147781,
+      "grad_norm": 0.2514174464993971,
+      "learning_rate": 8.80272090082227e-06,
+      "loss": 0.2418,
+      "step": 993
+    },
+    {
+      "epoch": 0.6063748665548269,
+      "grad_norm": 0.26721468418789796,
+      "learning_rate": 8.799263814591053e-06,
+      "loss": 0.2289,
+      "step": 994
+    },
+    {
+      "epoch": 0.6069849016318438,
+      "grad_norm": 0.25553432819619254,
+      "learning_rate": 8.795802425408352e-06,
+      "loss": 0.2732,
+      "step": 995
+    },
+    {
+      "epoch": 0.6075949367088608,
+      "grad_norm": 0.2376491834604695,
+      "learning_rate": 8.792336737194452e-06,
+      "loss": 0.2329,
+      "step": 996
+    },
+    {
+      "epoch": 0.6082049717858777,
+      "grad_norm": 0.234012654353243,
+      "learning_rate": 8.788866753874504e-06,
+      "loss": 0.2178,
+      "step": 997
+    },
+    {
+      "epoch": 0.6088150068628946,
+      "grad_norm": 0.23538003697799972,
+      "learning_rate": 8.785392479378522e-06,
+      "loss": 0.2324,
+      "step": 998
+    },
+    {
+      "epoch": 0.6094250419399115,
+      "grad_norm": 0.24857362136607877,
+      "learning_rate": 8.781913917641385e-06,
+      "loss": 0.2392,
+      "step": 999
+    },
+    {
+      "epoch": 0.6100350770169285,
+      "grad_norm": 0.2346747879400166,
+      "learning_rate": 8.778431072602825e-06,
+      "loss": 0.2271,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6100350770169285,
+      "eval_loss": 0.23594006896018982,
+      "eval_runtime": 267.1885,
+      "eval_samples_per_second": 4.005,
+      "eval_steps_per_second": 0.127,
+      "step": 1000
+    },
+    {
+      "epoch": 0.6106451120939455,
+      "grad_norm": 0.21228502344726596,
+      "learning_rate": 8.774943948207427e-06,
+      "loss": 0.2289,
+      "step": 1001
+    },
+    {
+      "epoch": 0.6112551471709623,
+      "grad_norm": 0.2500646392480775,
+      "learning_rate": 8.771452548404618e-06,
+      "loss": 0.2441,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6118651822479793,
+      "grad_norm": 0.23169540639899572,
+      "learning_rate": 8.767956877148671e-06,
+      "loss": 0.242,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6124752173249962,
+      "grad_norm": 0.22760124382669475,
+      "learning_rate": 8.7644569383987e-06,
+      "loss": 0.2552,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6130852524020132,
+      "grad_norm": 0.2484290973427134,
+      "learning_rate": 8.760952736118645e-06,
+      "loss": 0.2324,
+      "step": 1005
+    },
+    {
+      "epoch": 0.61369528747903,
+      "grad_norm": 0.2359583380201497,
+      "learning_rate": 8.757444274277277e-06,
+      "loss": 0.2175,
+      "step": 1006
+    },
+    {
+      "epoch": 0.614305322556047,
+      "grad_norm": 0.23220814369598353,
+      "learning_rate": 8.753931556848195e-06,
+      "loss": 0.2334,
+      "step": 1007
+    },
+    {
+      "epoch": 0.6149153576330639,
+      "grad_norm": 0.27498536282455094,
+      "learning_rate": 8.750414587809815e-06,
+      "loss": 0.2582,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6155253927100808,
+      "grad_norm": 0.24078199124375543,
+      "learning_rate": 8.746893371145367e-06,
+      "loss": 0.2317,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6161354277870977,
+      "grad_norm": 0.23280560001762854,
+      "learning_rate": 8.743367910842895e-06,
+      "loss": 0.2349,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6167454628641147,
+      "grad_norm": 0.23708670544193472,
+      "learning_rate": 8.739838210895244e-06,
+      "loss": 0.2215,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6173554979411316,
+      "grad_norm": 0.236517827135338,
+      "learning_rate": 8.736304275300069e-06,
+      "loss": 0.2372,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6179655330181485,
+      "grad_norm": 0.22912668594274044,
+      "learning_rate": 8.732766108059814e-06,
+      "loss": 0.2253,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6185755680951655,
+      "grad_norm": 0.23350025433250618,
+      "learning_rate": 8.72922371318172e-06,
+      "loss": 0.2274,
+      "step": 1014
+    },
+    {
+      "epoch": 0.6191856031721824,
+      "grad_norm": 0.23095428571608784,
+      "learning_rate": 8.725677094677816e-06,
+      "loss": 0.2349,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6197956382491994,
+      "grad_norm": 0.2458511385558197,
+      "learning_rate": 8.722126256564911e-06,
+      "loss": 0.2405,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6204056733262162,
+      "grad_norm": 0.26209582180802815,
+      "learning_rate": 8.718571202864598e-06,
+      "loss": 0.2391,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6210157084032332,
+      "grad_norm": 0.23368392839276564,
+      "learning_rate": 8.715011937603243e-06,
+      "loss": 0.2296,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6216257434802501,
+      "grad_norm": 0.23329724132770577,
+      "learning_rate": 8.711448464811978e-06,
+      "loss": 0.2425,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6222357785572671,
+      "grad_norm": 0.23902811017652673,
+      "learning_rate": 8.707880788526708e-06,
+      "loss": 0.2478,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6228458136342839,
+      "grad_norm": 0.41056486667562464,
+      "learning_rate": 8.70430891278809e-06,
+      "loss": 0.2198,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6234558487113009,
+      "grad_norm": 0.22537846767846592,
+      "learning_rate": 8.700732841641542e-06,
+      "loss": 0.2273,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6240658837883178,
+      "grad_norm": 0.44369492215713807,
+      "learning_rate": 8.697152579137235e-06,
+      "loss": 0.235,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6246759188653348,
+      "grad_norm": 0.23115109526114427,
+      "learning_rate": 8.693568129330083e-06,
+      "loss": 0.2273,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6252859539423516,
+      "grad_norm": 0.23509942514935336,
+      "learning_rate": 8.689979496279747e-06,
+      "loss": 0.2222,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6258959890193686,
+      "grad_norm": 0.2368473427816096,
+      "learning_rate": 8.68638668405062e-06,
+      "loss": 0.227,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6265060240963856,
+      "grad_norm": 0.21340890219018352,
+      "learning_rate": 8.682789696711835e-06,
+      "loss": 0.222,
+      "step": 1027
+    },
+    {
+      "epoch": 0.6271160591734025,
+      "grad_norm": 0.26572592638064496,
+      "learning_rate": 8.679188538337248e-06,
+      "loss": 0.2567,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6277260942504194,
+      "grad_norm": 0.22809021260408102,
+      "learning_rate": 8.675583213005443e-06,
+      "loss": 0.2279,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6283361293274363,
+      "grad_norm": 0.23394941073505565,
+      "learning_rate": 8.671973724799719e-06,
+      "loss": 0.2365,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6289461644044533,
+      "grad_norm": 0.23887151195100773,
+      "learning_rate": 8.668360077808093e-06,
+      "loss": 0.2472,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6295561994814702,
+      "grad_norm": 0.22443946646077903,
+      "learning_rate": 8.664742276123293e-06,
+      "loss": 0.2356,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6301662345584871,
+      "grad_norm": 0.24621520519113224,
+      "learning_rate": 8.661120323842751e-06,
+      "loss": 0.2312,
+      "step": 1033
+    },
+    {
+      "epoch": 0.630776269635504,
+      "grad_norm": 0.2292352894257401,
+      "learning_rate": 8.657494225068599e-06,
+      "loss": 0.2375,
+      "step": 1034
+    },
+    {
+      "epoch": 0.631386304712521,
+      "grad_norm": 0.23841878871186253,
+      "learning_rate": 8.653863983907665e-06,
+      "loss": 0.2423,
+      "step": 1035
+    },
+    {
+      "epoch": 0.631996339789538,
+      "grad_norm": 0.23381301415513847,
+      "learning_rate": 8.650229604471471e-06,
+      "loss": 0.217,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6326063748665548,
+      "grad_norm": 0.2482976093582336,
+      "learning_rate": 8.646591090876225e-06,
+      "loss": 0.2387,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6332164099435718,
+      "grad_norm": 0.2403988608376904,
+      "learning_rate": 8.642948447242816e-06,
+      "loss": 0.2294,
+      "step": 1038
+    },
+    {
+      "epoch": 0.6338264450205887,
+      "grad_norm": 0.25498061501682,
+      "learning_rate": 8.639301677696812e-06,
+      "loss": 0.254,
+      "step": 1039
+    },
+    {
+      "epoch": 0.6344364800976056,
+      "grad_norm": 0.2589882916862488,
+      "learning_rate": 8.635650786368452e-06,
+      "loss": 0.2442,
+      "step": 1040
+    },
+    {
+      "epoch": 0.6350465151746225,
+      "grad_norm": 0.23064501898207432,
+      "learning_rate": 8.631995777392645e-06,
+      "loss": 0.2291,
+      "step": 1041
+    },
+    {
+      "epoch": 0.6356565502516395,
+      "grad_norm": 0.2275940945269994,
+      "learning_rate": 8.628336654908965e-06,
+      "loss": 0.2256,
+      "step": 1042
+    },
+    {
+      "epoch": 0.6362665853286564,
+      "grad_norm": 0.2329033335658531,
+      "learning_rate": 8.62467342306164e-06,
+      "loss": 0.2364,
+      "step": 1043
+    },
+    {
+      "epoch": 0.6368766204056733,
+      "grad_norm": 0.22329035062909994,
+      "learning_rate": 8.621006085999557e-06,
+      "loss": 0.2302,
+      "step": 1044
+    },
+    {
+      "epoch": 0.6374866554826902,
+      "grad_norm": 0.24244570767153245,
+      "learning_rate": 8.61733464787625e-06,
+      "loss": 0.2568,
+      "step": 1045
+    },
+    {
+      "epoch": 0.6380966905597072,
+      "grad_norm": 0.23192088392396648,
+      "learning_rate": 8.613659112849898e-06,
+      "loss": 0.2563,
+      "step": 1046
+    },
+    {
+      "epoch": 0.6387067256367241,
+      "grad_norm": 0.24548506765765826,
+      "learning_rate": 8.609979485083319e-06,
+      "loss": 0.2491,
+      "step": 1047
+    },
+    {
+      "epoch": 0.639316760713741,
+      "grad_norm": 0.22389446741910482,
+      "learning_rate": 8.606295768743971e-06,
+      "loss": 0.2235,
+      "step": 1048
+    },
+    {
+      "epoch": 0.639926795790758,
+      "grad_norm": 0.22417184612529703,
+      "learning_rate": 8.602607968003935e-06,
+      "loss": 0.2241,
+      "step": 1049
+    },
+    {
+      "epoch": 0.6405368308677749,
+      "grad_norm": 0.25746416694776264,
+      "learning_rate": 8.598916087039929e-06,
+      "loss": 0.2283,
+      "step": 1050
+    },
+    {
+      "epoch": 0.6411468659447919,
+      "grad_norm": 0.24729586756768165,
+      "learning_rate": 8.595220130033278e-06,
+      "loss": 0.225,
+      "step": 1051
+    },
+    {
+      "epoch": 0.6417569010218087,
+      "grad_norm": 0.22642669069332533,
+      "learning_rate": 8.591520101169932e-06,
+      "loss": 0.2142,
+      "step": 1052
+    },
+    {
+      "epoch": 0.6423669360988257,
+      "grad_norm": 0.23342351734889488,
+      "learning_rate": 8.587816004640456e-06,
+      "loss": 0.2225,
+      "step": 1053
+    },
+    {
+      "epoch": 0.6429769711758426,
+      "grad_norm": 0.23649016461111497,
+      "learning_rate": 8.584107844640015e-06,
+      "loss": 0.2223,
+      "step": 1054
+    },
+    {
+      "epoch": 0.6435870062528596,
+      "grad_norm": 0.23848683592171122,
+      "learning_rate": 8.580395625368377e-06,
+      "loss": 0.2374,
+      "step": 1055
+    },
+    {
+      "epoch": 0.6441970413298764,
+      "grad_norm": 0.2284818810785837,
+      "learning_rate": 8.57667935102991e-06,
+      "loss": 0.2325,
+      "step": 1056
+    },
+    {
+      "epoch": 0.6448070764068934,
+      "grad_norm": 0.23563332797270775,
+      "learning_rate": 8.572959025833573e-06,
+      "loss": 0.2266,
+      "step": 1057
+    },
+    {
+      "epoch": 0.6454171114839103,
+      "grad_norm": 0.2464668648207247,
+      "learning_rate": 8.569234653992916e-06,
+      "loss": 0.2311,
+      "step": 1058
+    },
+    {
+      "epoch": 0.6460271465609273,
+      "grad_norm": 0.23023004719076975,
+      "learning_rate": 8.565506239726068e-06,
+      "loss": 0.2202,
+      "step": 1059
+    },
+    {
+      "epoch": 0.6466371816379441,
+      "grad_norm": 0.2640229604622966,
+      "learning_rate": 8.561773787255738e-06,
+      "loss": 0.2353,
+      "step": 1060
+    },
+    {
+      "epoch": 0.6472472167149611,
+      "grad_norm": 0.23223099173660483,
+      "learning_rate": 8.558037300809209e-06,
+      "loss": 0.2291,
+      "step": 1061
+    },
+    {
+      "epoch": 0.6478572517919781,
+      "grad_norm": 0.22720435240012227,
+      "learning_rate": 8.554296784618331e-06,
+      "loss": 0.2359,
+      "step": 1062
+    },
+    {
+      "epoch": 0.648467286868995,
+      "grad_norm": 0.22326014158801946,
+      "learning_rate": 8.550552242919521e-06,
+      "loss": 0.2503,
+      "step": 1063
+    },
+    {
+      "epoch": 0.6490773219460119,
+      "grad_norm": 0.21990941872169437,
+      "learning_rate": 8.546803679953754e-06,
+      "loss": 0.2176,
+      "step": 1064
+    },
+    {
+      "epoch": 0.6496873570230288,
+      "grad_norm": 0.21398433020072413,
+      "learning_rate": 8.543051099966558e-06,
+      "loss": 0.209,
+      "step": 1065
+    },
+    {
+      "epoch": 0.6502973921000458,
+      "grad_norm": 0.262452859799478,
+      "learning_rate": 8.539294507208011e-06,
+      "loss": 0.2169,
+      "step": 1066
+    },
+    {
+      "epoch": 0.6509074271770627,
+      "grad_norm": 0.23857047674039852,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 0.2535,
+      "step": 1067
+    },
+    {
+      "epoch": 0.6515174622540796,
+      "grad_norm": 0.2290193027949121,
+      "learning_rate": 8.531769300399901e-06,
+      "loss": 0.2282,
+      "step": 1068
+    },
+    {
+      "epoch": 0.6521274973310965,
+      "grad_norm": 0.22083932716278717,
+      "learning_rate": 8.5280006948732e-06,
+      "loss": 0.2336,
+      "step": 1069
+    },
+    {
+      "epoch": 0.6527375324081135,
+      "grad_norm": 0.22246158483514522,
+      "learning_rate": 8.524228093620858e-06,
+      "loss": 0.2416,
+      "step": 1070
+    },
+    {
+      "epoch": 0.6533475674851303,
+      "grad_norm": 0.21561212780574032,
+      "learning_rate": 8.520451500915636e-06,
+      "loss": 0.2179,
+      "step": 1071
+    },
+    {
+      "epoch": 0.6539576025621473,
+      "grad_norm": 0.23427975137844478,
+      "learning_rate": 8.516670921034808e-06,
+      "loss": 0.2238,
+      "step": 1072
+    },
+    {
+      "epoch": 0.6545676376391643,
+      "grad_norm": 0.2424857066387141,
+      "learning_rate": 8.512886358260162e-06,
+      "loss": 0.2243,
+      "step": 1073
+    },
+    {
+      "epoch": 0.6551776727161812,
+      "grad_norm": 0.22388291176428463,
+      "learning_rate": 8.509097816877998e-06,
+      "loss": 0.2331,
+      "step": 1074
+    },
+    {
+      "epoch": 0.6557877077931981,
+      "grad_norm": 0.24291536165799196,
+      "learning_rate": 8.505305301179126e-06,
+      "loss": 0.2399,
+      "step": 1075
+    },
+    {
+      "epoch": 0.656397742870215,
+      "grad_norm": 0.21783349465923516,
+      "learning_rate": 8.501508815458856e-06,
+      "loss": 0.2105,
+      "step": 1076
+    },
+    {
+      "epoch": 0.657007777947232,
+      "grad_norm": 0.24460448320216843,
+      "learning_rate": 8.49770836401699e-06,
+      "loss": 0.2381,
+      "step": 1077
+    },
+    {
+      "epoch": 0.6576178130242489,
+      "grad_norm": 0.23595918167659072,
+      "learning_rate": 8.493903951157827e-06,
+      "loss": 0.2404,
+      "step": 1078
+    },
+    {
+      "epoch": 0.6582278481012658,
+      "grad_norm": 0.2463785035159111,
+      "learning_rate": 8.490095581190149e-06,
+      "loss": 0.2536,
+      "step": 1079
+    },
+    {
+      "epoch": 0.6588378831782827,
+      "grad_norm": 0.24510766049649824,
+      "learning_rate": 8.486283258427222e-06,
+      "loss": 0.2363,
+      "step": 1080
+    },
+    {
+      "epoch": 0.6594479182552997,
+      "grad_norm": 0.2217249773828561,
+      "learning_rate": 8.482466987186785e-06,
+      "loss": 0.2324,
+      "step": 1081
+    },
+    {
+      "epoch": 0.6600579533323166,
+      "grad_norm": 0.2765387808119517,
+      "learning_rate": 8.478646771791054e-06,
+      "loss": 0.2508,
+      "step": 1082
+    },
+    {
+      "epoch": 0.6606679884093335,
+      "grad_norm": 0.22300248422446325,
+      "learning_rate": 8.474822616566711e-06,
+      "loss": 0.2109,
+      "step": 1083
+    },
+    {
+      "epoch": 0.6612780234863505,
+      "grad_norm": 0.2173767023833337,
+      "learning_rate": 8.470994525844895e-06,
+      "loss": 0.2235,
+      "step": 1084
+    },
+    {
+      "epoch": 0.6618880585633674,
+      "grad_norm": 0.2352796493955204,
+      "learning_rate": 8.467162503961209e-06,
+      "loss": 0.2236,
+      "step": 1085
+    },
+    {
+      "epoch": 0.6624980936403844,
+      "grad_norm": 0.23816022190314465,
+      "learning_rate": 8.463326555255705e-06,
+      "loss": 0.2456,
+      "step": 1086
+    },
+    {
+      "epoch": 0.6631081287174012,
+      "grad_norm": 0.24662909644422232,
+      "learning_rate": 8.459486684072883e-06,
+      "loss": 0.2315,
+      "step": 1087
+    },
+    {
+      "epoch": 0.6637181637944182,
+      "grad_norm": 0.22918344342880975,
+      "learning_rate": 8.455642894761684e-06,
+      "loss": 0.2243,
+      "step": 1088
+    },
+    {
+      "epoch": 0.6643281988714351,
+      "grad_norm": 0.24701880968682127,
+      "learning_rate": 8.451795191675488e-06,
+      "loss": 0.2282,
+      "step": 1089
+    },
+    {
+      "epoch": 0.6649382339484521,
+      "grad_norm": 0.252271888950006,
+      "learning_rate": 8.44794357917211e-06,
+      "loss": 0.2249,
+      "step": 1090
+    },
+    {
+      "epoch": 0.6655482690254689,
+      "grad_norm": 0.25713164139831324,
+      "learning_rate": 8.444088061613788e-06,
+      "loss": 0.2483,
+      "step": 1091
+    },
+    {
+      "epoch": 0.6661583041024859,
+      "grad_norm": 0.23416139444019676,
+      "learning_rate": 8.440228643367188e-06,
+      "loss": 0.2362,
+      "step": 1092
+    },
+    {
+      "epoch": 0.6667683391795028,
+      "grad_norm": 0.22046541488555407,
+      "learning_rate": 8.436365328803386e-06,
+      "loss": 0.24,
+      "step": 1093
+    },
+    {
+      "epoch": 0.6673783742565198,
+      "grad_norm": 0.24786462384821747,
+      "learning_rate": 8.432498122297879e-06,
+      "loss": 0.2548,
+      "step": 1094
+    },
+    {
+      "epoch": 0.6679884093335366,
+      "grad_norm": 0.24671418121331395,
+      "learning_rate": 8.42862702823057e-06,
+      "loss": 0.2313,
+      "step": 1095
+    },
+    {
+      "epoch": 0.6685984444105536,
+      "grad_norm": 0.2386497978703496,
+      "learning_rate": 8.424752050985759e-06,
+      "loss": 0.2235,
+      "step": 1096
+    },
+    {
+      "epoch": 0.6692084794875706,
+      "grad_norm": 0.22968504637849413,
+      "learning_rate": 8.420873194952153e-06,
+      "loss": 0.2318,
+      "step": 1097
+    },
+    {
+      "epoch": 0.6698185145645874,
+      "grad_norm": 0.22870461975088988,
+      "learning_rate": 8.416990464522847e-06,
+      "loss": 0.213,
+      "step": 1098
+    },
+    {
+      "epoch": 0.6704285496416044,
+      "grad_norm": 0.24043487052170476,
+      "learning_rate": 8.413103864095322e-06,
+      "loss": 0.228,
+      "step": 1099
+    },
+    {
+      "epoch": 0.6710385847186213,
+      "grad_norm": 0.27758706857221244,
+      "learning_rate": 8.409213398071448e-06,
+      "loss": 0.229,
+      "step": 1100
+    },
+    {
+      "epoch": 0.6716486197956383,
+      "grad_norm": 0.2278187512949372,
+      "learning_rate": 8.405319070857466e-06,
+      "loss": 0.2099,
+      "step": 1101
+    },
+    {
+      "epoch": 0.6722586548726551,
+      "grad_norm": 0.21328727861837193,
+      "learning_rate": 8.401420886863998e-06,
+      "loss": 0.2176,
+      "step": 1102
+    },
+    {
+      "epoch": 0.6728686899496721,
+      "grad_norm": 0.2575211195112028,
+      "learning_rate": 8.39751885050603e-06,
+      "loss": 0.2354,
+      "step": 1103
+    },
+    {
+      "epoch": 0.673478725026689,
+      "grad_norm": 0.24555722594295123,
+      "learning_rate": 8.393612966202907e-06,
+      "loss": 0.2367,
+      "step": 1104
+    },
+    {
+      "epoch": 0.674088760103706,
+      "grad_norm": 0.22888826965165,
+      "learning_rate": 8.38970323837834e-06,
+      "loss": 0.2321,
+      "step": 1105
+    },
+    {
+      "epoch": 0.6746987951807228,
+      "grad_norm": 0.23363250380097145,
+      "learning_rate": 8.385789671460387e-06,
+      "loss": 0.2238,
+      "step": 1106
+    },
+    {
+      "epoch": 0.6753088302577398,
+      "grad_norm": 0.2274158118030155,
+      "learning_rate": 8.381872269881457e-06,
+      "loss": 0.2212,
+      "step": 1107
+    },
+    {
+      "epoch": 0.6759188653347568,
+      "grad_norm": 0.265507657795025,
+      "learning_rate": 8.377951038078303e-06,
+      "loss": 0.2848,
+      "step": 1108
+    },
+    {
+      "epoch": 0.6765289004117737,
+      "grad_norm": 0.23260057000097717,
+      "learning_rate": 8.37402598049201e-06,
+      "loss": 0.2232,
+      "step": 1109
+    },
+    {
+      "epoch": 0.6771389354887906,
+      "grad_norm": 0.240932782955572,
+      "learning_rate": 8.370097101568007e-06,
+      "loss": 0.2259,
+      "step": 1110
+    },
+    {
+      "epoch": 0.6777489705658075,
+      "grad_norm": 0.25307725757101474,
+      "learning_rate": 8.366164405756039e-06,
+      "loss": 0.249,
+      "step": 1111
+    },
+    {
+      "epoch": 0.6783590056428245,
+      "grad_norm": 0.2363549436507277,
+      "learning_rate": 8.36222789751018e-06,
+      "loss": 0.2317,
+      "step": 1112
+    },
+    {
+      "epoch": 0.6789690407198414,
+      "grad_norm": 0.283464866982075,
+      "learning_rate": 8.358287581288824e-06,
+      "loss": 0.2374,
+      "step": 1113
+    },
+    {
+      "epoch": 0.6795790757968583,
+      "grad_norm": 0.24552247588641626,
+      "learning_rate": 8.35434346155467e-06,
+      "loss": 0.2321,
+      "step": 1114
+    },
+    {
+      "epoch": 0.6801891108738752,
+      "grad_norm": 0.24535783592570692,
+      "learning_rate": 8.350395542774737e-06,
+      "loss": 0.2426,
+      "step": 1115
+    },
+    {
+      "epoch": 0.6807991459508922,
+      "grad_norm": 0.26345289189142446,
+      "learning_rate": 8.34644382942033e-06,
+      "loss": 0.2379,
+      "step": 1116
+    },
+    {
+      "epoch": 0.6814091810279091,
+      "grad_norm": 0.24433115654789947,
+      "learning_rate": 8.342488325967068e-06,
+      "loss": 0.2436,
+      "step": 1117
+    },
+    {
+      "epoch": 0.682019216104926,
+      "grad_norm": 0.24813412245663188,
+      "learning_rate": 8.338529036894855e-06,
+      "loss": 0.2427,
+      "step": 1118
+    },
+    {
+      "epoch": 0.682629251181943,
+      "grad_norm": 0.2246533356599739,
+      "learning_rate": 8.334565966687882e-06,
+      "loss": 0.2309,
+      "step": 1119
+    },
+    {
+      "epoch": 0.6832392862589599,
+      "grad_norm": 0.24678856538769056,
+      "learning_rate": 8.330599119834622e-06,
+      "loss": 0.2429,
+      "step": 1120
+    },
+    {
+      "epoch": 0.6838493213359769,
+      "grad_norm": 0.2750478481998693,
+      "learning_rate": 8.326628500827826e-06,
+      "loss": 0.2408,
+      "step": 1121
+    },
+    {
+      "epoch": 0.6844593564129937,
+      "grad_norm": 0.21598660094808886,
+      "learning_rate": 8.322654114164524e-06,
+      "loss": 0.2102,
+      "step": 1122
+    },
+    {
+      "epoch": 0.6850693914900107,
+      "grad_norm": 0.2343051371469007,
+      "learning_rate": 8.318675964346001e-06,
+      "loss": 0.2483,
+      "step": 1123
+    },
+    {
+      "epoch": 0.6856794265670276,
+      "grad_norm": 0.25297310363283115,
+      "learning_rate": 8.314694055877814e-06,
+      "loss": 0.2521,
+      "step": 1124
+    },
+    {
+      "epoch": 0.6862894616440446,
+      "grad_norm": 0.2337948691832747,
+      "learning_rate": 8.310708393269773e-06,
+      "loss": 0.2205,
+      "step": 1125
+    },
+    {
+      "epoch": 0.6868994967210614,
+      "grad_norm": 0.23465407792916493,
+      "learning_rate": 8.306718981035937e-06,
+      "loss": 0.2289,
+      "step": 1126
+    },
+    {
+      "epoch": 0.6875095317980784,
+      "grad_norm": 0.24891299806667935,
+      "learning_rate": 8.302725823694619e-06,
+      "loss": 0.2286,
+      "step": 1127
+    },
+    {
+      "epoch": 0.6881195668750953,
+      "grad_norm": 0.24328476121321269,
+      "learning_rate": 8.298728925768367e-06,
+      "loss": 0.224,
+      "step": 1128
+    },
+    {
+      "epoch": 0.6887296019521122,
+      "grad_norm": 0.2382382369544805,
+      "learning_rate": 8.294728291783967e-06,
+      "loss": 0.2386,
+      "step": 1129
+    },
+    {
+      "epoch": 0.6893396370291291,
+      "grad_norm": 0.2385597546167655,
+      "learning_rate": 8.290723926272439e-06,
+      "loss": 0.212,
+      "step": 1130
+    },
+    {
+      "epoch": 0.6899496721061461,
+      "grad_norm": 0.25577463335566336,
+      "learning_rate": 8.286715833769027e-06,
+      "loss": 0.2453,
+      "step": 1131
+    },
+    {
+      "epoch": 0.6905597071831631,
+      "grad_norm": 0.2560431692968465,
+      "learning_rate": 8.282704018813199e-06,
+      "loss": 0.2554,
+      "step": 1132
+    },
+    {
+      "epoch": 0.6911697422601799,
+      "grad_norm": 0.24340131037369614,
+      "learning_rate": 8.278688485948634e-06,
+      "loss": 0.2067,
+      "step": 1133
+    },
+    {
+      "epoch": 0.6917797773371969,
+      "grad_norm": 0.23092696341317517,
+      "learning_rate": 8.274669239723223e-06,
+      "loss": 0.2216,
+      "step": 1134
+    },
+    {
+      "epoch": 0.6923898124142138,
+      "grad_norm": 0.2554840520735112,
+      "learning_rate": 8.270646284689067e-06,
+      "loss": 0.2304,
+      "step": 1135
+    },
+    {
+      "epoch": 0.6929998474912308,
+      "grad_norm": 0.24958593690963987,
+      "learning_rate": 8.266619625402465e-06,
+      "loss": 0.231,
+      "step": 1136
+    },
+    {
+      "epoch": 0.6936098825682476,
+      "grad_norm": 0.2536907103950832,
+      "learning_rate": 8.262589266423908e-06,
+      "loss": 0.2377,
+      "step": 1137
+    },
+    {
+      "epoch": 0.6942199176452646,
+      "grad_norm": 0.2742601681630492,
+      "learning_rate": 8.258555212318083e-06,
+      "loss": 0.2636,
+      "step": 1138
+    },
+    {
+      "epoch": 0.6948299527222815,
+      "grad_norm": 0.24462935260812824,
+      "learning_rate": 8.254517467653858e-06,
+      "loss": 0.2355,
+      "step": 1139
+    },
+    {
+      "epoch": 0.6954399877992985,
+      "grad_norm": 0.23730909678619028,
+      "learning_rate": 8.25047603700428e-06,
+      "loss": 0.2246,
+      "step": 1140
+    },
+    {
+      "epoch": 0.6960500228763153,
+      "grad_norm": 0.2441627512964431,
+      "learning_rate": 8.246430924946575e-06,
+      "loss": 0.2242,
+      "step": 1141
+    },
+    {
+      "epoch": 0.6966600579533323,
+      "grad_norm": 0.23069368020362158,
+      "learning_rate": 8.242382136062135e-06,
+      "loss": 0.2344,
+      "step": 1142
+    },
+    {
+      "epoch": 0.6972700930303493,
+      "grad_norm": 0.26857725161103285,
+      "learning_rate": 8.238329674936518e-06,
+      "loss": 0.2517,
+      "step": 1143
+    },
+    {
+      "epoch": 0.6978801281073662,
+      "grad_norm": 0.2558117137259953,
+      "learning_rate": 8.234273546159441e-06,
+      "loss": 0.2293,
+      "step": 1144
+    },
+    {
+      "epoch": 0.6984901631843831,
+      "grad_norm": 0.2269265773408567,
+      "learning_rate": 8.230213754324773e-06,
+      "loss": 0.2158,
+      "step": 1145
+    },
+    {
+      "epoch": 0.6991001982614,
+      "grad_norm": 0.25171573657859114,
+      "learning_rate": 8.226150304030534e-06,
+      "loss": 0.2317,
+      "step": 1146
+    },
+    {
+      "epoch": 0.699710233338417,
+      "grad_norm": 0.23133479000353166,
+      "learning_rate": 8.222083199878885e-06,
+      "loss": 0.2099,
+      "step": 1147
+    },
+    {
+      "epoch": 0.7003202684154339,
+      "grad_norm": 0.25686510449037725,
+      "learning_rate": 8.218012446476128e-06,
+      "loss": 0.223,
+      "step": 1148
+    },
+    {
+      "epoch": 0.7009303034924508,
+      "grad_norm": 0.2456359179322101,
+      "learning_rate": 8.213938048432697e-06,
+      "loss": 0.2481,
+      "step": 1149
+    },
+    {
+      "epoch": 0.7015403385694677,
+      "grad_norm": 0.251265568965118,
+      "learning_rate": 8.209860010363153e-06,
+      "loss": 0.2336,
+      "step": 1150
+    },
+    {
+      "epoch": 0.7021503736464847,
+      "grad_norm": 0.25798870050684164,
+      "learning_rate": 8.205778336886182e-06,
+      "loss": 0.2155,
+      "step": 1151
+    },
+    {
+      "epoch": 0.7027604087235017,
+      "grad_norm": 0.2567842887016378,
+      "learning_rate": 8.201693032624584e-06,
+      "loss": 0.2299,
+      "step": 1152
+    },
+    {
+      "epoch": 0.7033704438005185,
+      "grad_norm": 0.2318668305771687,
+      "learning_rate": 8.19760410220527e-06,
+      "loss": 0.2303,
+      "step": 1153
+    },
+    {
+      "epoch": 0.7039804788775355,
+      "grad_norm": 0.22835210783508866,
+      "learning_rate": 8.193511550259268e-06,
+      "loss": 0.2392,
+      "step": 1154
+    },
+    {
+      "epoch": 0.7045905139545524,
+      "grad_norm": 0.2609582153500793,
+      "learning_rate": 8.189415381421693e-06,
+      "loss": 0.2466,
+      "step": 1155
+    },
+    {
+      "epoch": 0.7052005490315694,
+      "grad_norm": 0.22982680193286878,
+      "learning_rate": 8.185315600331768e-06,
+      "loss": 0.2321,
+      "step": 1156
+    },
+    {
+      "epoch": 0.7058105841085862,
+      "grad_norm": 0.2319979015513836,
+      "learning_rate": 8.1812122116328e-06,
+      "loss": 0.2271,
+      "step": 1157
+    },
+    {
+      "epoch": 0.7064206191856032,
+      "grad_norm": 0.25474865680783587,
+      "learning_rate": 8.177105219972187e-06,
+      "loss": 0.2326,
+      "step": 1158
+    },
+    {
+      "epoch": 0.7070306542626201,
+      "grad_norm": 0.22168370615095642,
+      "learning_rate": 8.172994630001405e-06,
+      "loss": 0.2037,
+      "step": 1159
+    },
+    {
+      "epoch": 0.707640689339637,
+      "grad_norm": 0.22219009238421022,
+      "learning_rate": 8.168880446376003e-06,
+      "loss": 0.2162,
+      "step": 1160
+    },
+    {
+      "epoch": 0.7082507244166539,
+      "grad_norm": 0.24093738372339843,
+      "learning_rate": 8.16476267375561e-06,
+      "loss": 0.2563,
+      "step": 1161
+    },
+    {
+      "epoch": 0.7088607594936709,
+      "grad_norm": 0.23841753177830316,
+      "learning_rate": 8.160641316803904e-06,
+      "loss": 0.2312,
+      "step": 1162
+    },
+    {
+      "epoch": 0.7094707945706878,
+      "grad_norm": 0.22861548309953486,
+      "learning_rate": 8.156516380188635e-06,
+      "loss": 0.2336,
+      "step": 1163
+    },
+    {
+      "epoch": 0.7100808296477047,
+      "grad_norm": 0.2360284376142477,
+      "learning_rate": 8.152387868581606e-06,
+      "loss": 0.2287,
+      "step": 1164
+    },
+    {
+      "epoch": 0.7106908647247216,
+      "grad_norm": 0.2520783115103026,
+      "learning_rate": 8.148255786658661e-06,
+      "loss": 0.2432,
+      "step": 1165
+    },
+    {
+      "epoch": 0.7113008998017386,
+      "grad_norm": 0.22384974462504423,
+      "learning_rate": 8.144120139099697e-06,
+      "loss": 0.2309,
+      "step": 1166
+    },
+    {
+      "epoch": 0.7119109348787556,
+      "grad_norm": 0.23670142303489072,
+      "learning_rate": 8.139980930588643e-06,
+      "loss": 0.2373,
+      "step": 1167
+    },
+    {
+      "epoch": 0.7125209699557724,
+      "grad_norm": 0.23092054782886554,
+      "learning_rate": 8.135838165813465e-06,
+      "loss": 0.2276,
+      "step": 1168
+    },
+    {
+      "epoch": 0.7131310050327894,
+      "grad_norm": 0.2271715206436096,
+      "learning_rate": 8.131691849466154e-06,
+      "loss": 0.2255,
+      "step": 1169
+    },
+    {
+      "epoch": 0.7137410401098063,
+      "grad_norm": 0.2229315252564618,
+      "learning_rate": 8.127541986242726e-06,
+      "loss": 0.2117,
+      "step": 1170
+    },
+    {
+      "epoch": 0.7143510751868233,
+      "grad_norm": 0.24446595930379747,
+      "learning_rate": 8.123388580843209e-06,
+      "loss": 0.2647,
+      "step": 1171
+    },
+    {
+      "epoch": 0.7149611102638401,
+      "grad_norm": 0.24913399767279057,
+      "learning_rate": 8.119231637971651e-06,
+      "loss": 0.2427,
+      "step": 1172
+    },
+    {
+      "epoch": 0.7155711453408571,
+      "grad_norm": 0.2981638083013536,
+      "learning_rate": 8.115071162336099e-06,
+      "loss": 0.2353,
+      "step": 1173
+    },
+    {
+      "epoch": 0.716181180417874,
+      "grad_norm": 0.24437466812048375,
+      "learning_rate": 8.110907158648605e-06,
+      "loss": 0.247,
+      "step": 1174
+    },
+    {
+      "epoch": 0.716791215494891,
+      "grad_norm": 0.2449625500398084,
+      "learning_rate": 8.106739631625216e-06,
+      "loss": 0.2364,
+      "step": 1175
+    },
+    {
+      "epoch": 0.7174012505719078,
+      "grad_norm": 0.2224944056103836,
+      "learning_rate": 8.102568585985968e-06,
+      "loss": 0.2266,
+      "step": 1176
+    },
+    {
+      "epoch": 0.7180112856489248,
+      "grad_norm": 0.24738870833311483,
+      "learning_rate": 8.098394026454886e-06,
+      "loss": 0.2283,
+      "step": 1177
+    },
+    {
+      "epoch": 0.7186213207259418,
+      "grad_norm": 0.24328672716739527,
+      "learning_rate": 8.09421595775997e-06,
+      "loss": 0.2401,
+      "step": 1178
+    },
+    {
+      "epoch": 0.7192313558029587,
+      "grad_norm": 0.22719417926256943,
+      "learning_rate": 8.090034384633194e-06,
+      "loss": 0.232,
+      "step": 1179
+    },
+    {
+      "epoch": 0.7198413908799756,
+      "grad_norm": 0.24324266904153669,
+      "learning_rate": 8.08584931181051e-06,
+      "loss": 0.2224,
+      "step": 1180
+    },
+    {
+      "epoch": 0.7204514259569925,
+      "grad_norm": 0.2199116279585472,
+      "learning_rate": 8.081660744031818e-06,
+      "loss": 0.222,
+      "step": 1181
+    },
+    {
+      "epoch": 0.7210614610340095,
+      "grad_norm": 0.23189331283667097,
+      "learning_rate": 8.077468686040994e-06,
+      "loss": 0.2347,
+      "step": 1182
+    },
+    {
+      "epoch": 0.7216714961110264,
+      "grad_norm": 0.23879528408430015,
+      "learning_rate": 8.073273142585856e-06,
+      "loss": 0.2254,
+      "step": 1183
+    },
+    {
+      "epoch": 0.7222815311880433,
+      "grad_norm": 0.22352638483489393,
+      "learning_rate": 8.06907411841817e-06,
+      "loss": 0.2106,
+      "step": 1184
+    },
+    {
+      "epoch": 0.7228915662650602,
+      "grad_norm": 0.23030909325093743,
+      "learning_rate": 8.064871618293647e-06,
+      "loss": 0.2455,
+      "step": 1185
+    },
+    {
+      "epoch": 0.7235016013420772,
+      "grad_norm": 0.22349269159069882,
+      "learning_rate": 8.060665646971934e-06,
+      "loss": 0.2243,
+      "step": 1186
+    },
+    {
+      "epoch": 0.7241116364190942,
+      "grad_norm": 0.2344008817329467,
+      "learning_rate": 8.056456209216609e-06,
+      "loss": 0.2466,
+      "step": 1187
+    },
+    {
+      "epoch": 0.724721671496111,
+      "grad_norm": 0.21862505201609747,
+      "learning_rate": 8.052243309795175e-06,
+      "loss": 0.2281,
+      "step": 1188
+    },
+    {
+      "epoch": 0.725331706573128,
+      "grad_norm": 0.22352934742727967,
+      "learning_rate": 8.048026953479062e-06,
+      "loss": 0.2253,
+      "step": 1189
+    },
+    {
+      "epoch": 0.7259417416501449,
+      "grad_norm": 0.22584075570112383,
+      "learning_rate": 8.043807145043604e-06,
+      "loss": 0.2437,
+      "step": 1190
+    },
+    {
+      "epoch": 0.7265517767271618,
+      "grad_norm": 0.22596357819552168,
+      "learning_rate": 8.039583889268055e-06,
+      "loss": 0.2533,
+      "step": 1191
+    },
+    {
+      "epoch": 0.7271618118041787,
+      "grad_norm": 0.2241405931562285,
+      "learning_rate": 8.035357190935565e-06,
+      "loss": 0.219,
+      "step": 1192
+    },
+    {
+      "epoch": 0.7277718468811957,
+      "grad_norm": 0.2407653255257801,
+      "learning_rate": 8.031127054833192e-06,
+      "loss": 0.2347,
+      "step": 1193
+    },
+    {
+      "epoch": 0.7283818819582126,
+      "grad_norm": 0.21279327555477656,
+      "learning_rate": 8.026893485751877e-06,
+      "loss": 0.2188,
+      "step": 1194
+    },
+    {
+      "epoch": 0.7289919170352295,
+      "grad_norm": 0.2261188217955493,
+      "learning_rate": 8.02265648848646e-06,
+      "loss": 0.228,
+      "step": 1195
+    },
+    {
+      "epoch": 0.7296019521122464,
+      "grad_norm": 0.21871855242114913,
+      "learning_rate": 8.01841606783566e-06,
+      "loss": 0.2244,
+      "step": 1196
+    },
+    {
+      "epoch": 0.7302119871892634,
+      "grad_norm": 0.26675126905336366,
+      "learning_rate": 8.014172228602063e-06,
+      "loss": 0.2218,
+      "step": 1197
+    },
+    {
+      "epoch": 0.7308220222662803,
+      "grad_norm": 0.2253793993000552,
+      "learning_rate": 8.009924975592145e-06,
+      "loss": 0.2284,
+      "step": 1198
+    },
+    {
+      "epoch": 0.7314320573432972,
+      "grad_norm": 0.21110478260357746,
+      "learning_rate": 8.005674313616231e-06,
+      "loss": 0.2224,
+      "step": 1199
+    },
+    {
+      "epoch": 0.7320420924203141,
+      "grad_norm": 0.2207365556711429,
+      "learning_rate": 8.00142024748852e-06,
+      "loss": 0.2363,
+      "step": 1200
+    },
+    {
+      "epoch": 0.7326521274973311,
+      "grad_norm": 0.2348452507736591,
+      "learning_rate": 7.997162782027061e-06,
+      "loss": 0.2345,
+      "step": 1201
+    },
+    {
+      "epoch": 0.7332621625743481,
+      "grad_norm": 0.22623405697304158,
+      "learning_rate": 7.992901922053751e-06,
+      "loss": 0.2309,
+      "step": 1202
+    },
+    {
+      "epoch": 0.7338721976513649,
+      "grad_norm": 0.22676453319507958,
+      "learning_rate": 7.988637672394337e-06,
+      "loss": 0.2179,
+      "step": 1203
+    },
+    {
+      "epoch": 0.7344822327283819,
+      "grad_norm": 0.21444773415921856,
+      "learning_rate": 7.984370037878396e-06,
+      "loss": 0.2302,
+      "step": 1204
+    },
+    {
+      "epoch": 0.7350922678053988,
+      "grad_norm": 0.24197036456944898,
+      "learning_rate": 7.98009902333935e-06,
+      "loss": 0.2344,
+      "step": 1205
+    },
+    {
+      "epoch": 0.7357023028824158,
+      "grad_norm": 0.241279921065408,
+      "learning_rate": 7.975824633614443e-06,
+      "loss": 0.2179,
+      "step": 1206
+    },
+    {
+      "epoch": 0.7363123379594326,
+      "grad_norm": 0.21156520913188215,
+      "learning_rate": 7.971546873544737e-06,
+      "loss": 0.2086,
+      "step": 1207
+    },
+    {
+      "epoch": 0.7369223730364496,
+      "grad_norm": 0.21162231842894277,
+      "learning_rate": 7.967265747975124e-06,
+      "loss": 0.2241,
+      "step": 1208
+    },
+    {
+      "epoch": 0.7375324081134665,
+      "grad_norm": 0.23532186361425042,
+      "learning_rate": 7.962981261754295e-06,
+      "loss": 0.2316,
+      "step": 1209
+    },
+    {
+      "epoch": 0.7381424431904835,
+      "grad_norm": 0.2568086699360337,
+      "learning_rate": 7.958693419734752e-06,
+      "loss": 0.2404,
+      "step": 1210
+    },
+    {
+      "epoch": 0.7387524782675003,
+      "grad_norm": 0.22705182738619106,
+      "learning_rate": 7.954402226772804e-06,
+      "loss": 0.2381,
+      "step": 1211
+    },
+    {
+      "epoch": 0.7393625133445173,
+      "grad_norm": 0.2261242827810141,
+      "learning_rate": 7.950107687728543e-06,
+      "loss": 0.2314,
+      "step": 1212
+    },
+    {
+      "epoch": 0.7399725484215343,
+      "grad_norm": 0.22796628523765586,
+      "learning_rate": 7.945809807465857e-06,
+      "loss": 0.2346,
+      "step": 1213
+    },
+    {
+      "epoch": 0.7405825834985512,
+      "grad_norm": 0.23528591338995508,
+      "learning_rate": 7.941508590852422e-06,
+      "loss": 0.2377,
+      "step": 1214
+    },
+    {
+      "epoch": 0.7411926185755681,
+      "grad_norm": 0.25031030450654246,
+      "learning_rate": 7.937204042759685e-06,
+      "loss": 0.2375,
+      "step": 1215
+    },
+    {
+      "epoch": 0.741802653652585,
+      "grad_norm": 0.24714501535227756,
+      "learning_rate": 7.932896168062874e-06,
+      "loss": 0.2343,
+      "step": 1216
+    },
+    {
+      "epoch": 0.742412688729602,
+      "grad_norm": 0.22388225074316453,
+      "learning_rate": 7.928584971640974e-06,
+      "loss": 0.2272,
+      "step": 1217
+    },
+    {
+      "epoch": 0.7430227238066189,
+      "grad_norm": 0.2330751968816673,
+      "learning_rate": 7.924270458376746e-06,
+      "loss": 0.2314,
+      "step": 1218
+    },
+    {
+      "epoch": 0.7436327588836358,
+      "grad_norm": 0.24644856853549163,
+      "learning_rate": 7.919952633156695e-06,
+      "loss": 0.2593,
+      "step": 1219
+    },
+    {
+      "epoch": 0.7442427939606527,
+      "grad_norm": 0.22879737615778123,
+      "learning_rate": 7.915631500871084e-06,
+      "loss": 0.2643,
+      "step": 1220
+    },
+    {
+      "epoch": 0.7448528290376697,
+      "grad_norm": 0.23142298859084215,
+      "learning_rate": 7.91130706641392e-06,
+      "loss": 0.2379,
+      "step": 1221
+    },
+    {
+      "epoch": 0.7454628641146865,
+      "grad_norm": 0.2524317523072115,
+      "learning_rate": 7.906979334682948e-06,
+      "loss": 0.23,
+      "step": 1222
+    },
+    {
+      "epoch": 0.7460728991917035,
+      "grad_norm": 0.23189754430124038,
+      "learning_rate": 7.90264831057965e-06,
+      "loss": 0.2304,
+      "step": 1223
+    },
+    {
+      "epoch": 0.7466829342687205,
+      "grad_norm": 0.2533738094883779,
+      "learning_rate": 7.898313999009238e-06,
+      "loss": 0.2385,
+      "step": 1224
+    },
+    {
+      "epoch": 0.7472929693457374,
+      "grad_norm": 0.2553489046355843,
+      "learning_rate": 7.893976404880643e-06,
+      "loss": 0.2507,
+      "step": 1225
+    },
+    {
+      "epoch": 0.7479030044227543,
+      "grad_norm": 0.25852436326492156,
+      "learning_rate": 7.889635533106515e-06,
+      "loss": 0.2351,
+      "step": 1226
+    },
+    {
+      "epoch": 0.7485130394997712,
+      "grad_norm": 0.22967301302314835,
+      "learning_rate": 7.88529138860322e-06,
+      "loss": 0.2266,
+      "step": 1227
+    },
+    {
+      "epoch": 0.7491230745767882,
+      "grad_norm": 0.26120335319102406,
+      "learning_rate": 7.880943976290826e-06,
+      "loss": 0.2486,
+      "step": 1228
+    },
+    {
+      "epoch": 0.7497331096538051,
+      "grad_norm": 0.25231195952917956,
+      "learning_rate": 7.876593301093104e-06,
+      "loss": 0.2263,
+      "step": 1229
+    },
+    {
+      "epoch": 0.750343144730822,
+      "grad_norm": 0.22763902531310243,
+      "learning_rate": 7.87223936793752e-06,
+      "loss": 0.2245,
+      "step": 1230
+    },
+    {
+      "epoch": 0.7509531798078389,
+      "grad_norm": 0.22658329796075055,
+      "learning_rate": 7.86788218175523e-06,
+      "loss": 0.2008,
+      "step": 1231
+    },
+    {
+      "epoch": 0.7515632148848559,
+      "grad_norm": 0.23368671212970132,
+      "learning_rate": 7.863521747481078e-06,
+      "loss": 0.2558,
+      "step": 1232
+    },
+    {
+      "epoch": 0.7521732499618728,
+      "grad_norm": 0.228740592341535,
+      "learning_rate": 7.859158070053578e-06,
+      "loss": 0.2369,
+      "step": 1233
+    },
+    {
+      "epoch": 0.7527832850388897,
+      "grad_norm": 0.21748172031890037,
+      "learning_rate": 7.854791154414925e-06,
+      "loss": 0.2074,
+      "step": 1234
+    },
+    {
+      "epoch": 0.7533933201159067,
+      "grad_norm": 0.21994817465646302,
+      "learning_rate": 7.850421005510977e-06,
+      "loss": 0.2332,
+      "step": 1235
+    },
+    {
+      "epoch": 0.7540033551929236,
+      "grad_norm": 0.23920796683942835,
+      "learning_rate": 7.846047628291259e-06,
+      "loss": 0.2242,
+      "step": 1236
+    },
+    {
+      "epoch": 0.7546133902699406,
+      "grad_norm": 0.2225627258100121,
+      "learning_rate": 7.841671027708945e-06,
+      "loss": 0.2407,
+      "step": 1237
+    },
+    {
+      "epoch": 0.7552234253469574,
+      "grad_norm": 0.23186328985692906,
+      "learning_rate": 7.837291208720867e-06,
+      "loss": 0.2325,
+      "step": 1238
+    },
+    {
+      "epoch": 0.7558334604239744,
+      "grad_norm": 0.21580905533020986,
+      "learning_rate": 7.832908176287497e-06,
+      "loss": 0.2167,
+      "step": 1239
+    },
+    {
+      "epoch": 0.7564434955009913,
+      "grad_norm": 0.22400244139158307,
+      "learning_rate": 7.828521935372948e-06,
+      "loss": 0.2289,
+      "step": 1240
+    },
+    {
+      "epoch": 0.7570535305780083,
+      "grad_norm": 0.22654448268365424,
+      "learning_rate": 7.824132490944968e-06,
+      "loss": 0.2203,
+      "step": 1241
+    },
+    {
+      "epoch": 0.7576635656550251,
+      "grad_norm": 0.23447871941330142,
+      "learning_rate": 7.819739847974932e-06,
+      "loss": 0.2479,
+      "step": 1242
+    },
+    {
+      "epoch": 0.7582736007320421,
+      "grad_norm": 0.2298061806062539,
+      "learning_rate": 7.81534401143784e-06,
+      "loss": 0.2247,
+      "step": 1243
+    },
+    {
+      "epoch": 0.758883635809059,
+      "grad_norm": 0.20948289511876436,
+      "learning_rate": 7.810944986312303e-06,
+      "loss": 0.229,
+      "step": 1244
+    },
+    {
+      "epoch": 0.759493670886076,
+      "grad_norm": 0.21746148602267854,
+      "learning_rate": 7.80654277758055e-06,
+      "loss": 0.2249,
+      "step": 1245
+    },
+    {
+      "epoch": 0.7601037059630928,
+      "grad_norm": 0.24122673913757323,
+      "learning_rate": 7.802137390228413e-06,
+      "loss": 0.2371,
+      "step": 1246
+    },
+    {
+      "epoch": 0.7607137410401098,
+      "grad_norm": 0.24321359321004912,
+      "learning_rate": 7.797728829245321e-06,
+      "loss": 0.2248,
+      "step": 1247
+    },
+    {
+      "epoch": 0.7613237761171268,
+      "grad_norm": 0.25483262345781993,
+      "learning_rate": 7.793317099624304e-06,
+      "loss": 0.2659,
+      "step": 1248
+    },
+    {
+      "epoch": 0.7619338111941437,
+      "grad_norm": 0.22017019551240405,
+      "learning_rate": 7.788902206361974e-06,
+      "loss": 0.2327,
+      "step": 1249
+    },
+    {
+      "epoch": 0.7625438462711606,
+      "grad_norm": 0.23290295174722392,
+      "learning_rate": 7.784484154458529e-06,
+      "loss": 0.2337,
+      "step": 1250
+    },
+    {
+      "epoch": 0.7631538813481775,
+      "grad_norm": 0.24425639545299802,
+      "learning_rate": 7.780062948917748e-06,
+      "loss": 0.253,
+      "step": 1251
+    },
+    {
+      "epoch": 0.7637639164251945,
+      "grad_norm": 0.25547736930193293,
+      "learning_rate": 7.775638594746975e-06,
+      "loss": 0.2324,
+      "step": 1252
+    },
+    {
+      "epoch": 0.7643739515022113,
+      "grad_norm": 0.25797109907955645,
+      "learning_rate": 7.771211096957125e-06,
+      "loss": 0.2235,
+      "step": 1253
+    },
+    {
+      "epoch": 0.7649839865792283,
+      "grad_norm": 0.22509840516733273,
+      "learning_rate": 7.766780460562668e-06,
+      "loss": 0.2356,
+      "step": 1254
+    },
+    {
+      "epoch": 0.7655940216562452,
+      "grad_norm": 0.23188690975236592,
+      "learning_rate": 7.762346690581636e-06,
+      "loss": 0.2324,
+      "step": 1255
+    },
+    {
+      "epoch": 0.7662040567332622,
+      "grad_norm": 0.2649412152942703,
+      "learning_rate": 7.757909792035608e-06,
+      "loss": 0.2499,
+      "step": 1256
+    },
+    {
+      "epoch": 0.766814091810279,
+      "grad_norm": 0.22237757487770793,
+      "learning_rate": 7.753469769949701e-06,
+      "loss": 0.2205,
+      "step": 1257
+    },
+    {
+      "epoch": 0.767424126887296,
+      "grad_norm": 0.2307508237229708,
+      "learning_rate": 7.749026629352574e-06,
+      "loss": 0.2214,
+      "step": 1258
+    },
+    {
+      "epoch": 0.768034161964313,
+      "grad_norm": 0.24305561326911349,
+      "learning_rate": 7.744580375276416e-06,
+      "loss": 0.2255,
+      "step": 1259
+    },
+    {
+      "epoch": 0.7686441970413299,
+      "grad_norm": 0.23895046411395413,
+      "learning_rate": 7.740131012756949e-06,
+      "loss": 0.2494,
+      "step": 1260
+    },
+    {
+      "epoch": 0.7692542321183468,
+      "grad_norm": 0.21275999961893463,
+      "learning_rate": 7.735678546833403e-06,
+      "loss": 0.2257,
+      "step": 1261
+    },
+    {
+      "epoch": 0.7698642671953637,
+      "grad_norm": 0.2300948862077056,
+      "learning_rate": 7.731222982548534e-06,
+      "loss": 0.2123,
+      "step": 1262
+    },
+    {
+      "epoch": 0.7704743022723807,
+      "grad_norm": 0.22850187582777248,
+      "learning_rate": 7.726764324948603e-06,
+      "loss": 0.2463,
+      "step": 1263
+    },
+    {
+      "epoch": 0.7710843373493976,
+      "grad_norm": 0.2349740693352672,
+      "learning_rate": 7.722302579083372e-06,
+      "loss": 0.2151,
+      "step": 1264
+    },
+    {
+      "epoch": 0.7716943724264145,
+      "grad_norm": 0.2291372748905964,
+      "learning_rate": 7.717837750006106e-06,
+      "loss": 0.2314,
+      "step": 1265
+    },
+    {
+      "epoch": 0.7723044075034314,
+      "grad_norm": 0.23656682674024532,
+      "learning_rate": 7.713369842773559e-06,
+      "loss": 0.2319,
+      "step": 1266
+    },
+    {
+      "epoch": 0.7729144425804484,
+      "grad_norm": 0.23540280383471496,
+      "learning_rate": 7.708898862445968e-06,
+      "loss": 0.2279,
+      "step": 1267
+    },
+    {
+      "epoch": 0.7735244776574653,
+      "grad_norm": 0.22784114220338783,
+      "learning_rate": 7.704424814087056e-06,
+      "loss": 0.2436,
+      "step": 1268
+    },
+    {
+      "epoch": 0.7741345127344822,
+      "grad_norm": 0.2153530551088378,
+      "learning_rate": 7.699947702764021e-06,
+      "loss": 0.2231,
+      "step": 1269
+    },
+    {
+      "epoch": 0.7747445478114992,
+      "grad_norm": 0.22471820248006716,
+      "learning_rate": 7.695467533547526e-06,
+      "loss": 0.2285,
+      "step": 1270
+    },
+    {
+      "epoch": 0.7753545828885161,
+      "grad_norm": 0.29785558070116797,
+      "learning_rate": 7.690984311511695e-06,
+      "loss": 0.2428,
+      "step": 1271
+    },
+    {
+      "epoch": 0.7759646179655331,
+      "grad_norm": 0.2330391693953769,
+      "learning_rate": 7.686498041734121e-06,
+      "loss": 0.2484,
+      "step": 1272
+    },
+    {
+      "epoch": 0.7765746530425499,
+      "grad_norm": 0.232168542840295,
+      "learning_rate": 7.682008729295834e-06,
+      "loss": 0.2431,
+      "step": 1273
+    },
+    {
+      "epoch": 0.7771846881195669,
+      "grad_norm": 0.2343178698731269,
+      "learning_rate": 7.67751637928132e-06,
+      "loss": 0.2376,
+      "step": 1274
+    },
+    {
+      "epoch": 0.7777947231965838,
+      "grad_norm": 0.24092057852258092,
+      "learning_rate": 7.673020996778507e-06,
+      "loss": 0.2305,
+      "step": 1275
+    },
+    {
+      "epoch": 0.7784047582736008,
+      "grad_norm": 0.22934187162290043,
+      "learning_rate": 7.668522586878747e-06,
+      "loss": 0.2271,
+      "step": 1276
+    },
+    {
+      "epoch": 0.7790147933506176,
+      "grad_norm": 0.23320760506261856,
+      "learning_rate": 7.664021154676828e-06,
+      "loss": 0.2465,
+      "step": 1277
+    },
+    {
+      "epoch": 0.7796248284276346,
+      "grad_norm": 0.25335406369839425,
+      "learning_rate": 7.659516705270964e-06,
+      "loss": 0.2355,
+      "step": 1278
+    },
+    {
+      "epoch": 0.7802348635046515,
+      "grad_norm": 0.22804728767238044,
+      "learning_rate": 7.655009243762779e-06,
+      "loss": 0.2329,
+      "step": 1279
+    },
+    {
+      "epoch": 0.7808448985816685,
+      "grad_norm": 0.24357066588081885,
+      "learning_rate": 7.650498775257308e-06,
+      "loss": 0.2481,
+      "step": 1280
+    },
+    {
+      "epoch": 0.7814549336586853,
+      "grad_norm": 0.22792879051170845,
+      "learning_rate": 7.645985304863004e-06,
+      "loss": 0.2274,
+      "step": 1281
+    },
+    {
+      "epoch": 0.7820649687357023,
+      "grad_norm": 0.2427200250101618,
+      "learning_rate": 7.641468837691704e-06,
+      "loss": 0.249,
+      "step": 1282
+    },
+    {
+      "epoch": 0.7826750038127193,
+      "grad_norm": 0.24302283034392322,
+      "learning_rate": 7.636949378858647e-06,
+      "loss": 0.2484,
+      "step": 1283
+    },
+    {
+      "epoch": 0.7832850388897361,
+      "grad_norm": 0.24328162193435018,
+      "learning_rate": 7.632426933482463e-06,
+      "loss": 0.2345,
+      "step": 1284
+    },
+    {
+      "epoch": 0.7838950739667531,
+      "grad_norm": 0.23545752554754826,
+      "learning_rate": 7.627901506685157e-06,
+      "loss": 0.2267,
+      "step": 1285
+    },
+    {
+      "epoch": 0.78450510904377,
+      "grad_norm": 0.24166143220162728,
+      "learning_rate": 7.623373103592117e-06,
+      "loss": 0.2232,
+      "step": 1286
+    },
+    {
+      "epoch": 0.785115144120787,
+      "grad_norm": 0.22697451038495747,
+      "learning_rate": 7.618841729332096e-06,
+      "loss": 0.2197,
+      "step": 1287
+    },
+    {
+      "epoch": 0.7857251791978038,
+      "grad_norm": 0.24022269483018174,
+      "learning_rate": 7.614307389037221e-06,
+      "loss": 0.2229,
+      "step": 1288
+    },
+    {
+      "epoch": 0.7863352142748208,
+      "grad_norm": 0.2734279797575056,
+      "learning_rate": 7.609770087842969e-06,
+      "loss": 0.2358,
+      "step": 1289
+    },
+    {
+      "epoch": 0.7869452493518377,
+      "grad_norm": 0.22672940324544666,
+      "learning_rate": 7.605229830888175e-06,
+      "loss": 0.2501,
+      "step": 1290
+    },
+    {
+      "epoch": 0.7875552844288547,
+      "grad_norm": 0.2313163672526963,
+      "learning_rate": 7.6006866233150185e-06,
+      "loss": 0.2309,
+      "step": 1291
+    },
+    {
+      "epoch": 0.7881653195058715,
+      "grad_norm": 0.24972872083148773,
+      "learning_rate": 7.596140470269029e-06,
+      "loss": 0.2446,
+      "step": 1292
+    },
+    {
+      "epoch": 0.7887753545828885,
+      "grad_norm": 0.23952232980453123,
+      "learning_rate": 7.5915913768990615e-06,
+      "loss": 0.2335,
+      "step": 1293
+    },
+    {
+      "epoch": 0.7893853896599055,
+      "grad_norm": 0.22291765355920107,
+      "learning_rate": 7.587039348357306e-06,
+      "loss": 0.2221,
+      "step": 1294
+    },
+    {
+      "epoch": 0.7899954247369224,
+      "grad_norm": 0.22097300349547738,
+      "learning_rate": 7.582484389799279e-06,
+      "loss": 0.226,
+      "step": 1295
+    },
+    {
+      "epoch": 0.7906054598139393,
+      "grad_norm": 0.23934611021017924,
+      "learning_rate": 7.57792650638381e-06,
+      "loss": 0.2405,
+      "step": 1296
+    },
+    {
+      "epoch": 0.7912154948909562,
+      "grad_norm": 0.23655610661652388,
+      "learning_rate": 7.573365703273045e-06,
+      "loss": 0.2409,
+      "step": 1297
+    },
+    {
+      "epoch": 0.7918255299679732,
+      "grad_norm": 0.2322235300104474,
+      "learning_rate": 7.568801985632439e-06,
+      "loss": 0.2316,
+      "step": 1298
+    },
+    {
+      "epoch": 0.7924355650449901,
+      "grad_norm": 0.23263724623119364,
+      "learning_rate": 7.564235358630741e-06,
+      "loss": 0.2224,
+      "step": 1299
+    },
+    {
+      "epoch": 0.793045600122007,
+      "grad_norm": 0.24587977411446366,
+      "learning_rate": 7.559665827439999e-06,
+      "loss": 0.2279,
+      "step": 1300
+    },
+    {
+      "epoch": 0.7936556351990239,
+      "grad_norm": 0.23587703962080825,
+      "learning_rate": 7.555093397235553e-06,
+      "loss": 0.2108,
+      "step": 1301
+    },
+    {
+      "epoch": 0.7942656702760409,
+      "grad_norm": 0.2298321573596229,
+      "learning_rate": 7.55051807319602e-06,
+      "loss": 0.2231,
+      "step": 1302
+    },
+    {
+      "epoch": 0.7948757053530578,
+      "grad_norm": 0.2419845813611781,
+      "learning_rate": 7.545939860503302e-06,
+      "loss": 0.2276,
+      "step": 1303
+    },
+    {
+      "epoch": 0.7954857404300747,
+      "grad_norm": 0.2376519344621617,
+      "learning_rate": 7.541358764342566e-06,
+      "loss": 0.2195,
+      "step": 1304
+    },
+    {
+      "epoch": 0.7960957755070917,
+      "grad_norm": 0.24021446019082004,
+      "learning_rate": 7.536774789902246e-06,
+      "loss": 0.2294,
+      "step": 1305
+    },
+    {
+      "epoch": 0.7967058105841086,
+      "grad_norm": 0.3073505701919987,
+      "learning_rate": 7.532187942374039e-06,
+      "loss": 0.2234,
+      "step": 1306
+    },
+    {
+      "epoch": 0.7973158456611256,
+      "grad_norm": 0.2504795234075228,
+      "learning_rate": 7.527598226952895e-06,
+      "loss": 0.2426,
+      "step": 1307
+    },
+    {
+      "epoch": 0.7979258807381424,
+      "grad_norm": 0.25002423317398603,
+      "learning_rate": 7.52300564883701e-06,
+      "loss": 0.2494,
+      "step": 1308
+    },
+    {
+      "epoch": 0.7985359158151594,
+      "grad_norm": 0.22645304542766165,
+      "learning_rate": 7.518410213227823e-06,
+      "loss": 0.226,
+      "step": 1309
+    },
+    {
+      "epoch": 0.7991459508921763,
+      "grad_norm": 0.2339205517501626,
+      "learning_rate": 7.51381192533001e-06,
+      "loss": 0.229,
+      "step": 1310
+    },
+    {
+      "epoch": 0.7997559859691933,
+      "grad_norm": 0.25401845978653675,
+      "learning_rate": 7.50921079035148e-06,
+      "loss": 0.2394,
+      "step": 1311
+    },
+    {
+      "epoch": 0.8003660210462101,
+      "grad_norm": 0.22318589801172659,
+      "learning_rate": 7.504606813503359e-06,
+      "loss": 0.1975,
+      "step": 1312
+    },
+    {
+      "epoch": 0.8009760561232271,
+      "grad_norm": 0.23765624448180353,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 0.2428,
+      "step": 1313
+    },
+    {
+      "epoch": 0.801586091200244,
+      "grad_norm": 0.22809616997517404,
+      "learning_rate": 7.495390355058965e-06,
+      "loss": 0.2151,
+      "step": 1314
+    },
+    {
+      "epoch": 0.8021961262772609,
+      "grad_norm": 0.23805330961204915,
+      "learning_rate": 7.490777883901022e-06,
+      "loss": 0.198,
+      "step": 1315
+    },
+    {
+      "epoch": 0.8028061613542778,
+      "grad_norm": 0.23254499760309993,
+      "learning_rate": 7.486162591750139e-06,
+      "loss": 0.2341,
+      "step": 1316
+    },
+    {
+      "epoch": 0.8034161964312948,
+      "grad_norm": 0.2368902139011421,
+      "learning_rate": 7.481544483833485e-06,
+      "loss": 0.2471,
+      "step": 1317
+    },
+    {
+      "epoch": 0.8040262315083118,
+      "grad_norm": 0.23430686145673094,
+      "learning_rate": 7.476923565381409e-06,
+      "loss": 0.2435,
+      "step": 1318
+    },
+    {
+      "epoch": 0.8046362665853286,
+      "grad_norm": 0.2548091577394907,
+      "learning_rate": 7.472299841627452e-06,
+      "loss": 0.2323,
+      "step": 1319
+    },
+    {
+      "epoch": 0.8052463016623456,
+      "grad_norm": 0.24438720549939227,
+      "learning_rate": 7.467673317808326e-06,
+      "loss": 0.236,
+      "step": 1320
+    },
+    {
+      "epoch": 0.8058563367393625,
+      "grad_norm": 0.23249596660173585,
+      "learning_rate": 7.463043999163919e-06,
+      "loss": 0.2241,
+      "step": 1321
+    },
+    {
+      "epoch": 0.8064663718163795,
+      "grad_norm": 0.2552422967988723,
+      "learning_rate": 7.45841189093728e-06,
+      "loss": 0.2226,
+      "step": 1322
+    },
+    {
+      "epoch": 0.8070764068933963,
+      "grad_norm": 0.23619599542524092,
+      "learning_rate": 7.4537769983746176e-06,
+      "loss": 0.2368,
+      "step": 1323
+    },
+    {
+      "epoch": 0.8076864419704133,
+      "grad_norm": 0.2338399745898082,
+      "learning_rate": 7.4491393267253e-06,
+      "loss": 0.2464,
+      "step": 1324
+    },
+    {
+      "epoch": 0.8082964770474302,
+      "grad_norm": 0.238242750737787,
+      "learning_rate": 7.444498881241835e-06,
+      "loss": 0.2349,
+      "step": 1325
+    },
+    {
+      "epoch": 0.8089065121244472,
+      "grad_norm": 0.24692516610408854,
+      "learning_rate": 7.439855667179878e-06,
+      "loss": 0.2449,
+      "step": 1326
+    },
+    {
+      "epoch": 0.809516547201464,
+      "grad_norm": 0.22680401164554698,
+      "learning_rate": 7.435209689798214e-06,
+      "loss": 0.2178,
+      "step": 1327
+    },
+    {
+      "epoch": 0.810126582278481,
+      "grad_norm": 0.21885722102120825,
+      "learning_rate": 7.430560954358764e-06,
+      "loss": 0.2185,
+      "step": 1328
+    },
+    {
+      "epoch": 0.810736617355498,
+      "grad_norm": 0.22330802053954865,
+      "learning_rate": 7.4259094661265685e-06,
+      "loss": 0.2227,
+      "step": 1329
+    },
+    {
+      "epoch": 0.8113466524325149,
+      "grad_norm": 0.2277247401992638,
+      "learning_rate": 7.421255230369789e-06,
+      "loss": 0.2187,
+      "step": 1330
+    },
+    {
+      "epoch": 0.8119566875095318,
+      "grad_norm": 0.21704734934316533,
+      "learning_rate": 7.4165982523596945e-06,
+      "loss": 0.2145,
+      "step": 1331
+    },
+    {
+      "epoch": 0.8125667225865487,
+      "grad_norm": 0.23756550362674023,
+      "learning_rate": 7.411938537370665e-06,
+      "loss": 0.2351,
+      "step": 1332
+    },
+    {
+      "epoch": 0.8131767576635657,
+      "grad_norm": 0.21908133335811206,
+      "learning_rate": 7.407276090680173e-06,
+      "loss": 0.2256,
+      "step": 1333
+    },
+    {
+      "epoch": 0.8137867927405826,
+      "grad_norm": 0.23818954219879968,
+      "learning_rate": 7.402610917568794e-06,
+      "loss": 0.2353,
+      "step": 1334
+    },
+    {
+      "epoch": 0.8143968278175995,
+      "grad_norm": 0.22786836784604528,
+      "learning_rate": 7.397943023320185e-06,
+      "loss": 0.2096,
+      "step": 1335
+    },
+    {
+      "epoch": 0.8150068628946164,
+      "grad_norm": 0.24966631084866364,
+      "learning_rate": 7.393272413221087e-06,
+      "loss": 0.276,
+      "step": 1336
+    },
+    {
+      "epoch": 0.8156168979716334,
+      "grad_norm": 0.22642197259958527,
+      "learning_rate": 7.388599092561315e-06,
+      "loss": 0.226,
+      "step": 1337
+    },
+    {
+      "epoch": 0.8162269330486503,
+      "grad_norm": 0.23581467816794235,
+      "learning_rate": 7.383923066633757e-06,
+      "loss": 0.2403,
+      "step": 1338
+    },
+    {
+      "epoch": 0.8168369681256672,
+      "grad_norm": 0.24104276175337372,
+      "learning_rate": 7.379244340734364e-06,
+      "loss": 0.2392,
+      "step": 1339
+    },
+    {
+      "epoch": 0.8174470032026842,
+      "grad_norm": 0.2309568139148668,
+      "learning_rate": 7.374562920162143e-06,
+      "loss": 0.2483,
+      "step": 1340
+    },
+    {
+      "epoch": 0.8180570382797011,
+      "grad_norm": 0.24852331910971381,
+      "learning_rate": 7.369878810219154e-06,
+      "loss": 0.2224,
+      "step": 1341
+    },
+    {
+      "epoch": 0.818667073356718,
+      "grad_norm": 0.23163239528602403,
+      "learning_rate": 7.365192016210504e-06,
+      "loss": 0.2423,
+      "step": 1342
+    },
+    {
+      "epoch": 0.8192771084337349,
+      "grad_norm": 0.22613304330564601,
+      "learning_rate": 7.360502543444339e-06,
+      "loss": 0.2288,
+      "step": 1343
+    },
+    {
+      "epoch": 0.8198871435107519,
+      "grad_norm": 0.2316101029741555,
+      "learning_rate": 7.355810397231839e-06,
+      "loss": 0.2302,
+      "step": 1344
+    },
+    {
+      "epoch": 0.8204971785877688,
+      "grad_norm": 0.38057976097802504,
+      "learning_rate": 7.351115582887212e-06,
+      "loss": 0.2353,
+      "step": 1345
+    },
+    {
+      "epoch": 0.8211072136647857,
+      "grad_norm": 0.21811622063757893,
+      "learning_rate": 7.3464181057276864e-06,
+      "loss": 0.2135,
+      "step": 1346
+    },
+    {
+      "epoch": 0.8217172487418026,
+      "grad_norm": 0.22294497991648154,
+      "learning_rate": 7.341717971073508e-06,
+      "loss": 0.2314,
+      "step": 1347
+    },
+    {
+      "epoch": 0.8223272838188196,
+      "grad_norm": 0.2198410794160411,
+      "learning_rate": 7.337015184247934e-06,
+      "loss": 0.2286,
+      "step": 1348
+    },
+    {
+      "epoch": 0.8229373188958365,
+      "grad_norm": 0.23378756412976776,
+      "learning_rate": 7.3323097505772225e-06,
+      "loss": 0.2143,
+      "step": 1349
+    },
+    {
+      "epoch": 0.8235473539728534,
+      "grad_norm": 0.2472951356627941,
+      "learning_rate": 7.32760167539063e-06,
+      "loss": 0.2393,
+      "step": 1350
+    },
+    {
+      "epoch": 0.8241573890498703,
+      "grad_norm": 0.22148366770064548,
+      "learning_rate": 7.322890964020404e-06,
+      "loss": 0.2371,
+      "step": 1351
+    },
+    {
+      "epoch": 0.8247674241268873,
+      "grad_norm": 0.2400413207311959,
+      "learning_rate": 7.318177621801781e-06,
+      "loss": 0.2268,
+      "step": 1352
+    },
+    {
+      "epoch": 0.8253774592039043,
+      "grad_norm": 0.244478678935058,
+      "learning_rate": 7.313461654072974e-06,
+      "loss": 0.2285,
+      "step": 1353
+    },
+    {
+      "epoch": 0.8259874942809211,
+      "grad_norm": 0.2424947875863482,
+      "learning_rate": 7.308743066175172e-06,
+      "loss": 0.2486,
+      "step": 1354
+    },
+    {
+      "epoch": 0.8265975293579381,
+      "grad_norm": 0.24438448325110573,
+      "learning_rate": 7.304021863452525e-06,
+      "loss": 0.2267,
+      "step": 1355
+    },
+    {
+      "epoch": 0.827207564434955,
+      "grad_norm": 0.23288442093325284,
+      "learning_rate": 7.299298051252155e-06,
+      "loss": 0.2216,
+      "step": 1356
+    },
+    {
+      "epoch": 0.827817599511972,
+      "grad_norm": 0.26337805688342736,
+      "learning_rate": 7.2945716349241305e-06,
+      "loss": 0.2435,
+      "step": 1357
+    },
+    {
+      "epoch": 0.8284276345889888,
+      "grad_norm": 0.22968689259833996,
+      "learning_rate": 7.289842619821475e-06,
+      "loss": 0.2263,
+      "step": 1358
+    },
+    {
+      "epoch": 0.8290376696660058,
+      "grad_norm": 0.2403816832878819,
+      "learning_rate": 7.285111011300153e-06,
+      "loss": 0.2492,
+      "step": 1359
+    },
+    {
+      "epoch": 0.8296477047430227,
+      "grad_norm": 0.24875377601622348,
+      "learning_rate": 7.280376814719065e-06,
+      "loss": 0.2502,
+      "step": 1360
+    },
+    {
+      "epoch": 0.8302577398200397,
+      "grad_norm": 0.24312711346639268,
+      "learning_rate": 7.2756400354400445e-06,
+      "loss": 0.2428,
+      "step": 1361
+    },
+    {
+      "epoch": 0.8308677748970565,
+      "grad_norm": 0.22234901823763067,
+      "learning_rate": 7.2709006788278505e-06,
+      "loss": 0.2126,
+      "step": 1362
+    },
+    {
+      "epoch": 0.8314778099740735,
+      "grad_norm": 0.23753237391538568,
+      "learning_rate": 7.266158750250158e-06,
+      "loss": 0.2435,
+      "step": 1363
+    },
+    {
+      "epoch": 0.8320878450510905,
+      "grad_norm": 0.22756154601895012,
+      "learning_rate": 7.261414255077561e-06,
+      "loss": 0.2296,
+      "step": 1364
+    },
+    {
+      "epoch": 0.8326978801281074,
+      "grad_norm": 0.23538810665024323,
+      "learning_rate": 7.2566671986835515e-06,
+      "loss": 0.232,
+      "step": 1365
+    },
+    {
+      "epoch": 0.8333079152051243,
+      "grad_norm": 0.2420894512158573,
+      "learning_rate": 7.2519175864445305e-06,
+      "loss": 0.224,
+      "step": 1366
+    },
+    {
+      "epoch": 0.8339179502821412,
+      "grad_norm": 0.2447468401595778,
+      "learning_rate": 7.24716542373979e-06,
+      "loss": 0.2458,
+      "step": 1367
+    },
+    {
+      "epoch": 0.8345279853591582,
+      "grad_norm": 0.22222057490321495,
+      "learning_rate": 7.242410715951511e-06,
+      "loss": 0.2421,
+      "step": 1368
+    },
+    {
+      "epoch": 0.8351380204361751,
+      "grad_norm": 0.22700120289677556,
+      "learning_rate": 7.237653468464756e-06,
+      "loss": 0.1935,
+      "step": 1369
+    },
+    {
+      "epoch": 0.835748055513192,
+      "grad_norm": 0.24917410581205887,
+      "learning_rate": 7.232893686667466e-06,
+      "loss": 0.2266,
+      "step": 1370
+    },
+    {
+      "epoch": 0.8363580905902089,
+      "grad_norm": 0.2213826712700349,
+      "learning_rate": 7.22813137595045e-06,
+      "loss": 0.2278,
+      "step": 1371
+    },
+    {
+      "epoch": 0.8369681256672259,
+      "grad_norm": 0.2150592481787688,
+      "learning_rate": 7.223366541707384e-06,
+      "loss": 0.2234,
+      "step": 1372
+    },
+    {
+      "epoch": 0.8375781607442427,
+      "grad_norm": 0.24553766658182227,
+      "learning_rate": 7.218599189334799e-06,
+      "loss": 0.2251,
+      "step": 1373
+    },
+    {
+      "epoch": 0.8381881958212597,
+      "grad_norm": 0.250753725538139,
+      "learning_rate": 7.21382932423208e-06,
+      "loss": 0.2321,
+      "step": 1374
+    },
+    {
+      "epoch": 0.8387982308982767,
+      "grad_norm": 0.23356843887284268,
+      "learning_rate": 7.209056951801459e-06,
+      "loss": 0.2524,
+      "step": 1375
+    },
+    {
+      "epoch": 0.8394082659752936,
+      "grad_norm": 0.2405478444046161,
+      "learning_rate": 7.204282077448002e-06,
+      "loss": 0.2302,
+      "step": 1376
+    },
+    {
+      "epoch": 0.8400183010523105,
+      "grad_norm": 0.24711542527518374,
+      "learning_rate": 7.199504706579617e-06,
+      "loss": 0.225,
+      "step": 1377
+    },
+    {
+      "epoch": 0.8406283361293274,
+      "grad_norm": 0.2440331913107547,
+      "learning_rate": 7.194724844607033e-06,
+      "loss": 0.2272,
+      "step": 1378
+    },
+    {
+      "epoch": 0.8412383712063444,
+      "grad_norm": 0.24071419085405785,
+      "learning_rate": 7.189942496943803e-06,
+      "loss": 0.2205,
+      "step": 1379
+    },
+    {
+      "epoch": 0.8418484062833613,
+      "grad_norm": 0.2241466462729327,
+      "learning_rate": 7.185157669006294e-06,
+      "loss": 0.2469,
+      "step": 1380
+    },
+    {
+      "epoch": 0.8424584413603782,
+      "grad_norm": 0.2643101109270153,
+      "learning_rate": 7.180370366213684e-06,
+      "loss": 0.23,
+      "step": 1381
+    },
+    {
+      "epoch": 0.8430684764373951,
+      "grad_norm": 0.24502726261774888,
+      "learning_rate": 7.175580593987952e-06,
+      "loss": 0.2623,
+      "step": 1382
+    },
+    {
+      "epoch": 0.8436785115144121,
+      "grad_norm": 0.2439187420244412,
+      "learning_rate": 7.170788357753873e-06,
+      "loss": 0.2445,
+      "step": 1383
+    },
+    {
+      "epoch": 0.844288546591429,
+      "grad_norm": 0.2166186222289749,
+      "learning_rate": 7.165993662939018e-06,
+      "loss": 0.1993,
+      "step": 1384
+    },
+    {
+      "epoch": 0.8448985816684459,
+      "grad_norm": 0.23837694647294,
+      "learning_rate": 7.161196514973735e-06,
+      "loss": 0.2447,
+      "step": 1385
+    },
+    {
+      "epoch": 0.8455086167454628,
+      "grad_norm": 0.232204914110554,
+      "learning_rate": 7.1563969192911576e-06,
+      "loss": 0.2278,
+      "step": 1386
+    },
+    {
+      "epoch": 0.8461186518224798,
+      "grad_norm": 0.24331665799552643,
+      "learning_rate": 7.1515948813271875e-06,
+      "loss": 0.2459,
+      "step": 1387
+    },
+    {
+      "epoch": 0.8467286868994968,
+      "grad_norm": 0.22854293193671033,
+      "learning_rate": 7.146790406520491e-06,
+      "loss": 0.2589,
+      "step": 1388
+    },
+    {
+      "epoch": 0.8473387219765136,
+      "grad_norm": 0.2483900851541747,
+      "learning_rate": 7.141983500312498e-06,
+      "loss": 0.2509,
+      "step": 1389
+    },
+    {
+      "epoch": 0.8479487570535306,
+      "grad_norm": 0.22671118415346048,
+      "learning_rate": 7.137174168147392e-06,
+      "loss": 0.2328,
+      "step": 1390
+    },
+    {
+      "epoch": 0.8485587921305475,
+      "grad_norm": 0.22736887462102334,
+      "learning_rate": 7.132362415472099e-06,
+      "loss": 0.2428,
+      "step": 1391
+    },
+    {
+      "epoch": 0.8491688272075645,
+      "grad_norm": 0.23152556848811895,
+      "learning_rate": 7.1275482477362946e-06,
+      "loss": 0.2188,
+      "step": 1392
+    },
+    {
+      "epoch": 0.8497788622845813,
+      "grad_norm": 0.23959385612766362,
+      "learning_rate": 7.122731670392381e-06,
+      "loss": 0.2403,
+      "step": 1393
+    },
+    {
+      "epoch": 0.8503888973615983,
+      "grad_norm": 0.23126381072035201,
+      "learning_rate": 7.1179126888954954e-06,
+      "loss": 0.2276,
+      "step": 1394
+    },
+    {
+      "epoch": 0.8509989324386152,
+      "grad_norm": 0.2408726152963115,
+      "learning_rate": 7.113091308703498e-06,
+      "loss": 0.233,
+      "step": 1395
+    },
+    {
+      "epoch": 0.8516089675156322,
+      "grad_norm": 0.23586865768989854,
+      "learning_rate": 7.108267535276962e-06,
+      "loss": 0.2342,
+      "step": 1396
+    },
+    {
+      "epoch": 0.852219002592649,
+      "grad_norm": 0.23510387961362605,
+      "learning_rate": 7.1034413740791705e-06,
+      "loss": 0.2529,
+      "step": 1397
+    },
+    {
+      "epoch": 0.852829037669666,
+      "grad_norm": 0.22766231088115058,
+      "learning_rate": 7.098612830576118e-06,
+      "loss": 0.2396,
+      "step": 1398
+    },
+    {
+      "epoch": 0.853439072746683,
+      "grad_norm": 0.23117358845079555,
+      "learning_rate": 7.09378191023649e-06,
+      "loss": 0.2501,
+      "step": 1399
+    },
+    {
+      "epoch": 0.8540491078236999,
+      "grad_norm": 0.24728580197582728,
+      "learning_rate": 7.088948618531668e-06,
+      "loss": 0.2496,
+      "step": 1400
+    },
+    {
+      "epoch": 0.8546591429007168,
+      "grad_norm": 0.239872576193232,
+      "learning_rate": 7.0841129609357165e-06,
+      "loss": 0.21,
+      "step": 1401
+    },
+    {
+      "epoch": 0.8552691779777337,
+      "grad_norm": 0.21627847504408382,
+      "learning_rate": 7.0792749429253785e-06,
+      "loss": 0.2115,
+      "step": 1402
+    },
+    {
+      "epoch": 0.8558792130547507,
+      "grad_norm": 0.23893267429014625,
+      "learning_rate": 7.0744345699800755e-06,
+      "loss": 0.2397,
+      "step": 1403
+    },
+    {
+      "epoch": 0.8564892481317675,
+      "grad_norm": 0.23226050538125845,
+      "learning_rate": 7.069591847581894e-06,
+      "loss": 0.2274,
+      "step": 1404
+    },
+    {
+      "epoch": 0.8570992832087845,
+      "grad_norm": 0.34291310562756117,
+      "learning_rate": 7.064746781215578e-06,
+      "loss": 0.2378,
+      "step": 1405
+    },
+    {
+      "epoch": 0.8577093182858014,
+      "grad_norm": 0.2330305946371675,
+      "learning_rate": 7.059899376368531e-06,
+      "loss": 0.2393,
+      "step": 1406
+    },
+    {
+      "epoch": 0.8583193533628184,
+      "grad_norm": 0.23466165031403366,
+      "learning_rate": 7.0550496385308e-06,
+      "loss": 0.2286,
+      "step": 1407
+    },
+    {
+      "epoch": 0.8589293884398352,
+      "grad_norm": 0.21108347674892095,
+      "learning_rate": 7.050197573195081e-06,
+      "loss": 0.2238,
+      "step": 1408
+    },
+    {
+      "epoch": 0.8595394235168522,
+      "grad_norm": 0.2119361948395696,
+      "learning_rate": 7.045343185856701e-06,
+      "loss": 0.2187,
+      "step": 1409
+    },
+    {
+      "epoch": 0.8601494585938692,
+      "grad_norm": 0.2288603535650092,
+      "learning_rate": 7.0404864820136165e-06,
+      "loss": 0.2311,
+      "step": 1410
+    },
+    {
+      "epoch": 0.8607594936708861,
+      "grad_norm": 0.2294300887569088,
+      "learning_rate": 7.03562746716641e-06,
+      "loss": 0.2376,
+      "step": 1411
+    },
+    {
+      "epoch": 0.861369528747903,
+      "grad_norm": 0.22628168525352094,
+      "learning_rate": 7.030766146818281e-06,
+      "loss": 0.2425,
+      "step": 1412
+    },
+    {
+      "epoch": 0.8619795638249199,
+      "grad_norm": 0.23004126887608609,
+      "learning_rate": 7.025902526475039e-06,
+      "loss": 0.2351,
+      "step": 1413
+    },
+    {
+      "epoch": 0.8625895989019369,
+      "grad_norm": 0.2156346960003979,
+      "learning_rate": 7.0210366116451e-06,
+      "loss": 0.2184,
+      "step": 1414
+    },
+    {
+      "epoch": 0.8631996339789538,
+      "grad_norm": 0.23633568097737323,
+      "learning_rate": 7.016168407839478e-06,
+      "loss": 0.2421,
+      "step": 1415
+    },
+    {
+      "epoch": 0.8638096690559707,
+      "grad_norm": 0.7032365965820666,
+      "learning_rate": 7.011297920571779e-06,
+      "loss": 0.22,
+      "step": 1416
+    },
+    {
+      "epoch": 0.8644197041329876,
+      "grad_norm": 0.22139560165332492,
+      "learning_rate": 7.006425155358195e-06,
+      "loss": 0.2411,
+      "step": 1417
+    },
+    {
+      "epoch": 0.8650297392100046,
+      "grad_norm": 0.22921141024552555,
+      "learning_rate": 7.001550117717499e-06,
+      "loss": 0.2243,
+      "step": 1418
+    },
+    {
+      "epoch": 0.8656397742870215,
+      "grad_norm": 0.2148756902424555,
+      "learning_rate": 6.996672813171036e-06,
+      "loss": 0.2191,
+      "step": 1419
+    },
+    {
+      "epoch": 0.8662498093640384,
+      "grad_norm": 0.24364684199653186,
+      "learning_rate": 6.991793247242719e-06,
+      "loss": 0.2504,
+      "step": 1420
+    },
+    {
+      "epoch": 0.8668598444410553,
+      "grad_norm": 0.22724841878361154,
+      "learning_rate": 6.986911425459028e-06,
+      "loss": 0.2158,
+      "step": 1421
+    },
+    {
+      "epoch": 0.8674698795180723,
+      "grad_norm": 0.2298061255709973,
+      "learning_rate": 6.982027353348985e-06,
+      "loss": 0.2276,
+      "step": 1422
+    },
+    {
+      "epoch": 0.8680799145950893,
+      "grad_norm": 0.24823382304574634,
+      "learning_rate": 6.977141036444174e-06,
+      "loss": 0.2608,
+      "step": 1423
+    },
+    {
+      "epoch": 0.8686899496721061,
+      "grad_norm": 0.23330840784018506,
+      "learning_rate": 6.9722524802787125e-06,
+      "loss": 0.2384,
+      "step": 1424
+    },
+    {
+      "epoch": 0.8692999847491231,
+      "grad_norm": 0.23031863340182507,
+      "learning_rate": 6.967361690389258e-06,
+      "loss": 0.2384,
+      "step": 1425
+    },
+    {
+      "epoch": 0.86991001982614,
+      "grad_norm": 0.23343096638678282,
+      "learning_rate": 6.962468672314999e-06,
+      "loss": 0.2247,
+      "step": 1426
+    },
+    {
+      "epoch": 0.870520054903157,
+      "grad_norm": 0.23412336450252114,
+      "learning_rate": 6.957573431597646e-06,
+      "loss": 0.2439,
+      "step": 1427
+    },
+    {
+      "epoch": 0.8711300899801738,
+      "grad_norm": 0.21705008516947805,
+      "learning_rate": 6.952675973781423e-06,
+      "loss": 0.2289,
+      "step": 1428
+    },
+    {
+      "epoch": 0.8717401250571908,
+      "grad_norm": 0.23462263092212576,
+      "learning_rate": 6.947776304413072e-06,
+      "loss": 0.2386,
+      "step": 1429
+    },
+    {
+      "epoch": 0.8723501601342077,
+      "grad_norm": 0.22881311935709336,
+      "learning_rate": 6.942874429041834e-06,
+      "loss": 0.2257,
+      "step": 1430
+    },
+    {
+      "epoch": 0.8729601952112247,
+      "grad_norm": 0.2165771690749278,
+      "learning_rate": 6.937970353219453e-06,
+      "loss": 0.2272,
+      "step": 1431
+    },
+    {
+      "epoch": 0.8735702302882415,
+      "grad_norm": 0.21766263672481576,
+      "learning_rate": 6.933064082500161e-06,
+      "loss": 0.227,
+      "step": 1432
+    },
+    {
+      "epoch": 0.8741802653652585,
+      "grad_norm": 0.23341818589676822,
+      "learning_rate": 6.92815562244068e-06,
+      "loss": 0.2266,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8747903004422755,
+      "grad_norm": 0.23091764664489853,
+      "learning_rate": 6.923244978600206e-06,
+      "loss": 0.2372,
+      "step": 1434
+    },
+    {
+      "epoch": 0.8754003355192923,
+      "grad_norm": 0.22688425972610404,
+      "learning_rate": 6.918332156540417e-06,
+      "loss": 0.2426,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8760103705963093,
+      "grad_norm": 0.2670114310566748,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 0.2505,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8766204056733262,
+      "grad_norm": 0.22422672515482298,
+      "learning_rate": 6.908500000021905e-06,
+      "loss": 0.2322,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8772304407503432,
+      "grad_norm": 0.24003057489356389,
+      "learning_rate": 6.903580676698838e-06,
+      "loss": 0.25,
+      "step": 1438
+    },
+    {
+      "epoch": 0.87784047582736,
+      "grad_norm": 0.2317200045289886,
+      "learning_rate": 6.898659197427748e-06,
+      "loss": 0.2423,
+      "step": 1439
+    },
+    {
+      "epoch": 0.878450510904377,
+      "grad_norm": 0.22708239407650227,
+      "learning_rate": 6.893735567782587e-06,
+      "loss": 0.2427,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8790605459813939,
+      "grad_norm": 0.2366178727370385,
+      "learning_rate": 6.888809793339729e-06,
+      "loss": 0.2324,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8796705810584109,
+      "grad_norm": 0.2376731316602752,
+      "learning_rate": 6.8838818796779875e-06,
+      "loss": 0.2234,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8802806161354277,
+      "grad_norm": 0.24035037512939267,
+      "learning_rate": 6.878951832378591e-06,
+      "loss": 0.2295,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8808906512124447,
+      "grad_norm": 0.21746369005203028,
+      "learning_rate": 6.87401965702519e-06,
+      "loss": 0.2382,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8815006862894617,
+      "grad_norm": 0.23319099616460107,
+      "learning_rate": 6.869085359203844e-06,
+      "loss": 0.2195,
+      "step": 1445
+    },
+    {
+      "epoch": 0.8821107213664786,
+      "grad_norm": 0.2382960087546005,
+      "learning_rate": 6.864148944503016e-06,
+      "loss": 0.2263,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8827207564434955,
+      "grad_norm": 0.2311147461646032,
+      "learning_rate": 6.859210418513564e-06,
+      "loss": 0.2364,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8833307915205124,
+      "grad_norm": 0.2363109991798035,
+      "learning_rate": 6.854269786828741e-06,
+      "loss": 0.2346,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8839408265975294,
+      "grad_norm": 0.22484383826058832,
+      "learning_rate": 6.849327055044182e-06,
+      "loss": 0.2165,
+      "step": 1449
+    },
+    {
+      "epoch": 0.8845508616745463,
+      "grad_norm": 0.22793442853232132,
+      "learning_rate": 6.844382228757904e-06,
+      "loss": 0.2263,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8851608967515632,
+      "grad_norm": 0.2133071411846939,
+      "learning_rate": 6.839435313570293e-06,
+      "loss": 0.2143,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8857709318285801,
+      "grad_norm": 0.23989669045168097,
+      "learning_rate": 6.834486315084101e-06,
+      "loss": 0.2279,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8863809669055971,
+      "grad_norm": 0.28193635749388085,
+      "learning_rate": 6.82953523890444e-06,
+      "loss": 0.2474,
+      "step": 1453
+    },
+    {
+      "epoch": 0.886991001982614,
+      "grad_norm": 0.21892568723573827,
+      "learning_rate": 6.824582090638777e-06,
+      "loss": 0.236,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8876010370596309,
+      "grad_norm": 0.228144663125668,
+      "learning_rate": 6.819626875896924e-06,
+      "loss": 0.2194,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8882110721366479,
+      "grad_norm": 0.2336697639430183,
+      "learning_rate": 6.814669600291033e-06,
+      "loss": 0.2244,
+      "step": 1456
+    },
+    {
+      "epoch": 0.8888211072136648,
+      "grad_norm": 0.23064852314498838,
+      "learning_rate": 6.80971026943559e-06,
+      "loss": 0.2359,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8894311422906818,
+      "grad_norm": 0.21840239732400749,
+      "learning_rate": 6.804748888947412e-06,
+      "loss": 0.2246,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8900411773676986,
+      "grad_norm": 0.2348987589111878,
+      "learning_rate": 6.799785464445633e-06,
+      "loss": 0.2491,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8906512124447156,
+      "grad_norm": 0.2428043433600667,
+      "learning_rate": 6.794820001551703e-06,
+      "loss": 0.2621,
+      "step": 1460
+    },
+    {
+      "epoch": 0.8912612475217325,
+      "grad_norm": 0.22983238516221302,
+      "learning_rate": 6.789852505889384e-06,
+      "loss": 0.2426,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8918712825987495,
+      "grad_norm": 0.23325969571323438,
+      "learning_rate": 6.784882983084735e-06,
+      "loss": 0.2225,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8924813176757663,
+      "grad_norm": 0.21727207523138228,
+      "learning_rate": 6.779911438766117e-06,
+      "loss": 0.211,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8930913527527833,
+      "grad_norm": 0.21790563597926196,
+      "learning_rate": 6.774937878564175e-06,
+      "loss": 0.2296,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8937013878298002,
+      "grad_norm": 0.23237825441684618,
+      "learning_rate": 6.769962308111839e-06,
+      "loss": 0.2386,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8943114229068171,
+      "grad_norm": 0.2245292716074995,
+      "learning_rate": 6.764984733044316e-06,
+      "loss": 0.2427,
+      "step": 1466
+    },
+    {
+      "epoch": 0.894921457983834,
+      "grad_norm": 0.23316404802575824,
+      "learning_rate": 6.7600051589990855e-06,
+      "loss": 0.237,
+      "step": 1467
+    },
+    {
+      "epoch": 0.895531493060851,
+      "grad_norm": 0.23682470929517443,
+      "learning_rate": 6.755023591615887e-06,
+      "loss": 0.2473,
+      "step": 1468
+    },
+    {
+      "epoch": 0.896141528137868,
+      "grad_norm": 0.2289751975303715,
+      "learning_rate": 6.750040036536718e-06,
+      "loss": 0.2436,
+      "step": 1469
+    },
+    {
+      "epoch": 0.8967515632148848,
+      "grad_norm": 0.2410302502040547,
+      "learning_rate": 6.745054499405833e-06,
+      "loss": 0.2363,
+      "step": 1470
+    },
+    {
+      "epoch": 0.8973615982919018,
+      "grad_norm": 0.23665105798930544,
+      "learning_rate": 6.740066985869724e-06,
+      "loss": 0.2328,
+      "step": 1471
+    },
+    {
+      "epoch": 0.8979716333689187,
+      "grad_norm": 0.268089173782497,
+      "learning_rate": 6.735077501577126e-06,
+      "loss": 0.2226,
+      "step": 1472
+    },
+    {
+      "epoch": 0.8985816684459357,
+      "grad_norm": 0.23567208315168722,
+      "learning_rate": 6.7300860521790034e-06,
+      "loss": 0.2443,
+      "step": 1473
+    },
+    {
+      "epoch": 0.8991917035229525,
+      "grad_norm": 0.22552131923653507,
+      "learning_rate": 6.725092643328548e-06,
+      "loss": 0.2373,
+      "step": 1474
+    },
+    {
+      "epoch": 0.8998017385999695,
+      "grad_norm": 0.21706492744114989,
+      "learning_rate": 6.72009728068117e-06,
+      "loss": 0.225,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9004117736769864,
+      "grad_norm": 0.2327906203614174,
+      "learning_rate": 6.715099969894492e-06,
+      "loss": 0.2326,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9010218087540034,
+      "grad_norm": 0.23253407140894464,
+      "learning_rate": 6.710100716628345e-06,
+      "loss": 0.226,
+      "step": 1477
+    },
+    {
+      "epoch": 0.9016318438310202,
+      "grad_norm": 0.22830354557774632,
+      "learning_rate": 6.705099526544757e-06,
+      "loss": 0.2539,
+      "step": 1478
+    },
+    {
+      "epoch": 0.9022418789080372,
+      "grad_norm": 0.24215917120267302,
+      "learning_rate": 6.700096405307952e-06,
+      "loss": 0.2487,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9028519139850542,
+      "grad_norm": 0.24211803786412073,
+      "learning_rate": 6.69509135858434e-06,
+      "loss": 0.2492,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9034619490620711,
+      "grad_norm": 0.24812360058343688,
+      "learning_rate": 6.690084392042514e-06,
+      "loss": 0.2219,
+      "step": 1481
+    },
+    {
+      "epoch": 0.904071984139088,
+      "grad_norm": 0.21262275648517254,
+      "learning_rate": 6.6850755113532385e-06,
+      "loss": 0.2234,
+      "step": 1482
+    },
+    {
+      "epoch": 0.9046820192161049,
+      "grad_norm": 0.23237360487836642,
+      "learning_rate": 6.680064722189445e-06,
+      "loss": 0.2312,
+      "step": 1483
+    },
+    {
+      "epoch": 0.9052920542931219,
+      "grad_norm": 0.22699230989997096,
+      "learning_rate": 6.675052030226231e-06,
+      "loss": 0.2332,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9059020893701388,
+      "grad_norm": 0.2430314059683205,
+      "learning_rate": 6.670037441140844e-06,
+      "loss": 0.2449,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9065121244471557,
+      "grad_norm": 0.22098823674287454,
+      "learning_rate": 6.665020960612685e-06,
+      "loss": 0.2344,
+      "step": 1486
+    },
+    {
+      "epoch": 0.9071221595241726,
+      "grad_norm": 0.24069656974852452,
+      "learning_rate": 6.6600025943232935e-06,
+      "loss": 0.2579,
+      "step": 1487
+    },
+    {
+      "epoch": 0.9077321946011896,
+      "grad_norm": 0.22598312198368797,
+      "learning_rate": 6.6549823479563445e-06,
+      "loss": 0.2487,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9083422296782065,
+      "grad_norm": 0.22896335937970488,
+      "learning_rate": 6.649960227197648e-06,
+      "loss": 0.2288,
+      "step": 1489
+    },
+    {
+      "epoch": 0.9089522647552234,
+      "grad_norm": 0.23738911819834904,
+      "learning_rate": 6.644936237735129e-06,
+      "loss": 0.2515,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9095622998322404,
+      "grad_norm": 0.2185717569808725,
+      "learning_rate": 6.639910385258834e-06,
+      "loss": 0.2392,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9101723349092573,
+      "grad_norm": 0.2321320019483118,
+      "learning_rate": 6.6348826754609195e-06,
+      "loss": 0.2603,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9107823699862743,
+      "grad_norm": 0.22568275472973856,
+      "learning_rate": 6.629853114035643e-06,
+      "loss": 0.2394,
+      "step": 1493
+    },
+    {
+      "epoch": 0.9113924050632911,
+      "grad_norm": 0.21026955772320566,
+      "learning_rate": 6.62482170667936e-06,
+      "loss": 0.2264,
+      "step": 1494
+    },
+    {
+      "epoch": 0.9120024401403081,
+      "grad_norm": 0.23213943301628961,
+      "learning_rate": 6.619788459090517e-06,
+      "loss": 0.2349,
+      "step": 1495
+    },
+    {
+      "epoch": 0.912612475217325,
+      "grad_norm": 0.22915949764604762,
+      "learning_rate": 6.614753376969647e-06,
+      "loss": 0.2295,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9132225102943419,
+      "grad_norm": 0.23372792813000748,
+      "learning_rate": 6.609716466019356e-06,
+      "loss": 0.2209,
+      "step": 1497
+    },
+    {
+      "epoch": 0.9138325453713588,
+      "grad_norm": 0.2309167856093707,
+      "learning_rate": 6.604677731944323e-06,
+      "loss": 0.2267,
+      "step": 1498
+    },
+    {
+      "epoch": 0.9144425804483758,
+      "grad_norm": 0.2148406989040156,
+      "learning_rate": 6.599637180451295e-06,
+      "loss": 0.2114,
+      "step": 1499
+    },
+    {
+      "epoch": 0.9150526155253927,
+      "grad_norm": 0.2189166439537073,
+      "learning_rate": 6.594594817249075e-06,
+      "loss": 0.24,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9150526155253927,
+      "eval_loss": 0.22871258854866028,
+      "eval_runtime": 265.6371,
+      "eval_samples_per_second": 4.028,
+      "eval_steps_per_second": 0.128,
+      "step": 1500
+    },
+    {
+      "epoch": 0.9156626506024096,
+      "grad_norm": 0.2519964998379413,
+      "learning_rate": 6.589550648048517e-06,
+      "loss": 0.2157,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9162726856794265,
+      "grad_norm": 0.2513906292659038,
+      "learning_rate": 6.584504678562521e-06,
+      "loss": 0.2317,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9168827207564435,
+      "grad_norm": 0.23399050870483937,
+      "learning_rate": 6.5794569145060264e-06,
+      "loss": 0.2383,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9174927558334605,
+      "grad_norm": 0.2498121509819766,
+      "learning_rate": 6.574407361596005e-06,
+      "loss": 0.2305,
+      "step": 1504
+    },
+    {
+      "epoch": 0.9181027909104773,
+      "grad_norm": 0.2493812842752683,
+      "learning_rate": 6.569356025551454e-06,
+      "loss": 0.23,
+      "step": 1505
+    },
+    {
+      "epoch": 0.9187128259874943,
+      "grad_norm": 0.22780378406134064,
+      "learning_rate": 6.564302912093393e-06,
+      "loss": 0.2258,
+      "step": 1506
+    },
+    {
+      "epoch": 0.9193228610645112,
+      "grad_norm": 0.2325339328201163,
+      "learning_rate": 6.5592480269448485e-06,
+      "loss": 0.255,
+      "step": 1507
+    },
+    {
+      "epoch": 0.9199328961415282,
+      "grad_norm": 0.23399882400423566,
+      "learning_rate": 6.554191375830861e-06,
+      "loss": 0.2258,
+      "step": 1508
+    },
+    {
+      "epoch": 0.920542931218545,
+      "grad_norm": 0.22460401789572665,
+      "learning_rate": 6.5491329644784655e-06,
+      "loss": 0.214,
+      "step": 1509
+    },
+    {
+      "epoch": 0.921152966295562,
+      "grad_norm": 0.23146171330320892,
+      "learning_rate": 6.544072798616694e-06,
+      "loss": 0.2291,
+      "step": 1510
+    },
+    {
+      "epoch": 0.9217630013725789,
+      "grad_norm": 0.24084505508421777,
+      "learning_rate": 6.539010883976562e-06,
+      "loss": 0.2497,
+      "step": 1511
+    },
+    {
+      "epoch": 0.9223730364495959,
+      "grad_norm": 0.22304395266301494,
+      "learning_rate": 6.53394722629107e-06,
+      "loss": 0.2187,
+      "step": 1512
+    },
+    {
+      "epoch": 0.9229830715266127,
+      "grad_norm": 0.2387792130084646,
+      "learning_rate": 6.5288818312951886e-06,
+      "loss": 0.2357,
+      "step": 1513
+    },
+    {
+      "epoch": 0.9235931066036297,
+      "grad_norm": 0.2401822676734899,
+      "learning_rate": 6.523814704725861e-06,
+      "loss": 0.2242,
+      "step": 1514
+    },
+    {
+      "epoch": 0.9242031416806467,
+      "grad_norm": 0.23445665401381172,
+      "learning_rate": 6.518745852321985e-06,
+      "loss": 0.2237,
+      "step": 1515
+    },
+    {
+      "epoch": 0.9248131767576636,
+      "grad_norm": 0.2284841709830546,
+      "learning_rate": 6.513675279824416e-06,
+      "loss": 0.2267,
+      "step": 1516
+    },
+    {
+      "epoch": 0.9254232118346805,
+      "grad_norm": 0.2128083194347949,
+      "learning_rate": 6.508602992975963e-06,
+      "loss": 0.2048,
+      "step": 1517
+    },
+    {
+      "epoch": 0.9260332469116974,
+      "grad_norm": 0.2175291929915765,
+      "learning_rate": 6.503528997521365e-06,
+      "loss": 0.2317,
+      "step": 1518
+    },
+    {
+      "epoch": 0.9266432819887144,
+      "grad_norm": 0.21551752748043324,
+      "learning_rate": 6.4984532992073094e-06,
+      "loss": 0.216,
+      "step": 1519
+    },
+    {
+      "epoch": 0.9272533170657313,
+      "grad_norm": 0.22973453408383288,
+      "learning_rate": 6.493375903782402e-06,
+      "loss": 0.2357,
+      "step": 1520
+    },
+    {
+      "epoch": 0.9278633521427482,
+      "grad_norm": 0.23097559158343273,
+      "learning_rate": 6.4882968169971734e-06,
+      "loss": 0.2372,
+      "step": 1521
+    },
+    {
+      "epoch": 0.9284733872197651,
+      "grad_norm": 0.22636389176483931,
+      "learning_rate": 6.483216044604073e-06,
+      "loss": 0.213,
+      "step": 1522
+    },
+    {
+      "epoch": 0.9290834222967821,
+      "grad_norm": 0.24225414024833547,
+      "learning_rate": 6.478133592357455e-06,
+      "loss": 0.241,
+      "step": 1523
+    },
+    {
+      "epoch": 0.929693457373799,
+      "grad_norm": 0.22840372865504438,
+      "learning_rate": 6.473049466013579e-06,
+      "loss": 0.2399,
+      "step": 1524
+    },
+    {
+      "epoch": 0.9303034924508159,
+      "grad_norm": 0.21765688745780767,
+      "learning_rate": 6.467963671330602e-06,
+      "loss": 0.2231,
+      "step": 1525
+    },
+    {
+      "epoch": 0.9309135275278329,
+      "grad_norm": 0.22555043157288188,
+      "learning_rate": 6.462876214068563e-06,
+      "loss": 0.222,
+      "step": 1526
+    },
+    {
+      "epoch": 0.9315235626048498,
+      "grad_norm": 0.22859166767599118,
+      "learning_rate": 6.457787099989392e-06,
+      "loss": 0.2102,
+      "step": 1527
+    },
+    {
+      "epoch": 0.9321335976818667,
+      "grad_norm": 0.23530311284338487,
+      "learning_rate": 6.452696334856895e-06,
+      "loss": 0.2484,
+      "step": 1528
+    },
+    {
+      "epoch": 0.9327436327588836,
+      "grad_norm": 0.2146376266487681,
+      "learning_rate": 6.447603924436744e-06,
+      "loss": 0.2119,
+      "step": 1529
+    },
+    {
+      "epoch": 0.9333536678359006,
+      "grad_norm": 0.2283231397865005,
+      "learning_rate": 6.442509874496475e-06,
+      "loss": 0.2297,
+      "step": 1530
+    },
+    {
+      "epoch": 0.9339637029129175,
+      "grad_norm": 0.21933733150433,
+      "learning_rate": 6.437414190805482e-06,
+      "loss": 0.2268,
+      "step": 1531
+    },
+    {
+      "epoch": 0.9345737379899344,
+      "grad_norm": 0.22898995559601057,
+      "learning_rate": 6.432316879135012e-06,
+      "loss": 0.2293,
+      "step": 1532
+    },
+    {
+      "epoch": 0.9351837730669513,
+      "grad_norm": 0.23928333789077255,
+      "learning_rate": 6.4272179452581505e-06,
+      "loss": 0.2358,
+      "step": 1533
+    },
+    {
+      "epoch": 0.9357938081439683,
+      "grad_norm": 0.22245443866611708,
+      "learning_rate": 6.422117394949824e-06,
+      "loss": 0.2196,
+      "step": 1534
+    },
+    {
+      "epoch": 0.9364038432209852,
+      "grad_norm": 0.2271006986531639,
+      "learning_rate": 6.417015233986786e-06,
+      "loss": 0.2434,
+      "step": 1535
+    },
+    {
+      "epoch": 0.9370138782980021,
+      "grad_norm": 0.2592277983360455,
+      "learning_rate": 6.41191146814762e-06,
+      "loss": 0.2214,
+      "step": 1536
+    },
+    {
+      "epoch": 0.937623913375019,
+      "grad_norm": 0.228391871107343,
+      "learning_rate": 6.406806103212725e-06,
+      "loss": 0.2122,
+      "step": 1537
+    },
+    {
+      "epoch": 0.938233948452036,
+      "grad_norm": 0.21874020313662693,
+      "learning_rate": 6.401699144964306e-06,
+      "loss": 0.2196,
+      "step": 1538
+    },
+    {
+      "epoch": 0.938843983529053,
+      "grad_norm": 0.23244284042470695,
+      "learning_rate": 6.3965905991863805e-06,
+      "loss": 0.233,
+      "step": 1539
+    },
+    {
+      "epoch": 0.9394540186060698,
+      "grad_norm": 0.22457758396597863,
+      "learning_rate": 6.391480471664757e-06,
+      "loss": 0.2362,
+      "step": 1540
+    },
+    {
+      "epoch": 0.9400640536830868,
+      "grad_norm": 0.2336482299735738,
+      "learning_rate": 6.38636876818704e-06,
+      "loss": 0.2326,
+      "step": 1541
+    },
+    {
+      "epoch": 0.9406740887601037,
+      "grad_norm": 0.2189721244318323,
+      "learning_rate": 6.381255494542618e-06,
+      "loss": 0.2167,
+      "step": 1542
+    },
+    {
+      "epoch": 0.9412841238371207,
+      "grad_norm": 0.2994373279270973,
+      "learning_rate": 6.376140656522656e-06,
+      "loss": 0.2428,
+      "step": 1543
+    },
+    {
+      "epoch": 0.9418941589141375,
+      "grad_norm": 0.33778356800748166,
+      "learning_rate": 6.371024259920091e-06,
+      "loss": 0.2371,
+      "step": 1544
+    },
+    {
+      "epoch": 0.9425041939911545,
+      "grad_norm": 0.21829401941414528,
+      "learning_rate": 6.365906310529631e-06,
+      "loss": 0.2361,
+      "step": 1545
+    },
+    {
+      "epoch": 0.9431142290681714,
+      "grad_norm": 0.23289317417704491,
+      "learning_rate": 6.360786814147731e-06,
+      "loss": 0.2453,
+      "step": 1546
+    },
+    {
+      "epoch": 0.9437242641451884,
+      "grad_norm": 0.25579141903640873,
+      "learning_rate": 6.3556657765726116e-06,
+      "loss": 0.221,
+      "step": 1547
+    },
+    {
+      "epoch": 0.9443342992222052,
+      "grad_norm": 0.23020496289050363,
+      "learning_rate": 6.350543203604228e-06,
+      "loss": 0.2244,
+      "step": 1548
+    },
+    {
+      "epoch": 0.9449443342992222,
+      "grad_norm": 0.2211248442137757,
+      "learning_rate": 6.345419101044281e-06,
+      "loss": 0.2175,
+      "step": 1549
+    },
+    {
+      "epoch": 0.9455543693762392,
+      "grad_norm": 0.22858010072311402,
+      "learning_rate": 6.340293474696198e-06,
+      "loss": 0.2345,
+      "step": 1550
+    },
+    {
+      "epoch": 0.9461644044532561,
+      "grad_norm": 0.2374051390192958,
+      "learning_rate": 6.335166330365141e-06,
+      "loss": 0.2223,
+      "step": 1551
+    },
+    {
+      "epoch": 0.946774439530273,
+      "grad_norm": 0.22322603988882175,
+      "learning_rate": 6.330037673857982e-06,
+      "loss": 0.2309,
+      "step": 1552
+    },
+    {
+      "epoch": 0.9473844746072899,
+      "grad_norm": 0.23828346993481017,
+      "learning_rate": 6.32490751098331e-06,
+      "loss": 0.2245,
+      "step": 1553
+    },
+    {
+      "epoch": 0.9479945096843069,
+      "grad_norm": 0.25786837395107554,
+      "learning_rate": 6.3197758475514195e-06,
+      "loss": 0.2414,
+      "step": 1554
+    },
+    {
+      "epoch": 0.9486045447613238,
+      "grad_norm": 0.21358766242427954,
+      "learning_rate": 6.314642689374304e-06,
+      "loss": 0.2082,
+      "step": 1555
+    },
+    {
+      "epoch": 0.9492145798383407,
+      "grad_norm": 0.20849750694020697,
+      "learning_rate": 6.309508042265654e-06,
+      "loss": 0.2007,
+      "step": 1556
+    },
+    {
+      "epoch": 0.9498246149153576,
+      "grad_norm": 0.21895390787981686,
+      "learning_rate": 6.30437191204084e-06,
+      "loss": 0.2209,
+      "step": 1557
+    },
+    {
+      "epoch": 0.9504346499923746,
+      "grad_norm": 0.24064011815601305,
+      "learning_rate": 6.299234304516914e-06,
+      "loss": 0.2374,
+      "step": 1558
+    },
+    {
+      "epoch": 0.9510446850693914,
+      "grad_norm": 0.22325511306252963,
+      "learning_rate": 6.294095225512604e-06,
+      "loss": 0.2308,
+      "step": 1559
+    },
+    {
+      "epoch": 0.9516547201464084,
+      "grad_norm": 0.23327277715190625,
+      "learning_rate": 6.288954680848303e-06,
+      "loss": 0.2352,
+      "step": 1560
+    },
+    {
+      "epoch": 0.9522647552234254,
+      "grad_norm": 0.22259263112982883,
+      "learning_rate": 6.2838126763460635e-06,
+      "loss": 0.223,
+      "step": 1561
+    },
+    {
+      "epoch": 0.9528747903004423,
+      "grad_norm": 0.22604430426038336,
+      "learning_rate": 6.27866921782959e-06,
+      "loss": 0.2166,
+      "step": 1562
+    },
+    {
+      "epoch": 0.9534848253774592,
+      "grad_norm": 0.21851943667633414,
+      "learning_rate": 6.273524311124237e-06,
+      "loss": 0.2073,
+      "step": 1563
+    },
+    {
+      "epoch": 0.9540948604544761,
+      "grad_norm": 0.2548954994078049,
+      "learning_rate": 6.268377962056999e-06,
+      "loss": 0.2328,
+      "step": 1564
+    },
+    {
+      "epoch": 0.9547048955314931,
+      "grad_norm": 0.23804348920601198,
+      "learning_rate": 6.263230176456497e-06,
+      "loss": 0.2146,
+      "step": 1565
+    },
+    {
+      "epoch": 0.95531493060851,
+      "grad_norm": 0.2266315320160639,
+      "learning_rate": 6.258080960152991e-06,
+      "loss": 0.2181,
+      "step": 1566
+    },
+    {
+      "epoch": 0.9559249656855269,
+      "grad_norm": 0.23542998803866033,
+      "learning_rate": 6.252930318978353e-06,
+      "loss": 0.2385,
+      "step": 1567
+    },
+    {
+      "epoch": 0.9565350007625438,
+      "grad_norm": 0.21317840339021937,
+      "learning_rate": 6.247778258766069e-06,
+      "loss": 0.212,
+      "step": 1568
+    },
+    {
+      "epoch": 0.9571450358395608,
+      "grad_norm": 0.2355547595678835,
+      "learning_rate": 6.2426247853512355e-06,
+      "loss": 0.2493,
+      "step": 1569
+    },
+    {
+      "epoch": 0.9577550709165777,
+      "grad_norm": 0.2365566242506874,
+      "learning_rate": 6.237469904570549e-06,
+      "loss": 0.2394,
+      "step": 1570
+    },
+    {
+      "epoch": 0.9583651059935946,
+      "grad_norm": 0.2351549924941449,
+      "learning_rate": 6.232313622262297e-06,
+      "loss": 0.2349,
+      "step": 1571
+    },
+    {
+      "epoch": 0.9589751410706115,
+      "grad_norm": 0.2153592643562907,
+      "learning_rate": 6.227155944266358e-06,
+      "loss": 0.2272,
+      "step": 1572
+    },
+    {
+      "epoch": 0.9595851761476285,
+      "grad_norm": 0.22869892531056712,
+      "learning_rate": 6.221996876424186e-06,
+      "loss": 0.249,
+      "step": 1573
+    },
+    {
+      "epoch": 0.9601952112246455,
+      "grad_norm": 0.23218107560341958,
+      "learning_rate": 6.216836424578816e-06,
+      "loss": 0.2243,
+      "step": 1574
+    },
+    {
+      "epoch": 0.9608052463016623,
+      "grad_norm": 0.23492027645928484,
+      "learning_rate": 6.211674594574847e-06,
+      "loss": 0.207,
+      "step": 1575
+    },
+    {
+      "epoch": 0.9614152813786793,
+      "grad_norm": 0.22911404127241716,
+      "learning_rate": 6.206511392258439e-06,
+      "loss": 0.229,
+      "step": 1576
+    },
+    {
+      "epoch": 0.9620253164556962,
+      "grad_norm": 0.2479077538171469,
+      "learning_rate": 6.2013468234773034e-06,
+      "loss": 0.2342,
+      "step": 1577
+    },
+    {
+      "epoch": 0.9626353515327132,
+      "grad_norm": 0.24617532460859282,
+      "learning_rate": 6.196180894080707e-06,
+      "loss": 0.2491,
+      "step": 1578
+    },
+    {
+      "epoch": 0.96324538660973,
+      "grad_norm": 0.22360258704509595,
+      "learning_rate": 6.191013609919448e-06,
+      "loss": 0.2256,
+      "step": 1579
+    },
+    {
+      "epoch": 0.963855421686747,
+      "grad_norm": 0.22912652157745447,
+      "learning_rate": 6.185844976845866e-06,
+      "loss": 0.2266,
+      "step": 1580
+    },
+    {
+      "epoch": 0.9644654567637639,
+      "grad_norm": 0.23041424403533584,
+      "learning_rate": 6.180675000713825e-06,
+      "loss": 0.2244,
+      "step": 1581
+    },
+    {
+      "epoch": 0.9650754918407809,
+      "grad_norm": 0.3259350286108923,
+      "learning_rate": 6.175503687378711e-06,
+      "loss": 0.245,
+      "step": 1582
+    },
+    {
+      "epoch": 0.9656855269177977,
+      "grad_norm": 0.2213882091045891,
+      "learning_rate": 6.170331042697425e-06,
+      "loss": 0.2309,
+      "step": 1583
+    },
+    {
+      "epoch": 0.9662955619948147,
+      "grad_norm": 0.2309385259789968,
+      "learning_rate": 6.1651570725283774e-06,
+      "loss": 0.2197,
+      "step": 1584
+    },
+    {
+      "epoch": 0.9669055970718317,
+      "grad_norm": 0.2276259703395926,
+      "learning_rate": 6.1599817827314744e-06,
+      "loss": 0.2468,
+      "step": 1585
+    },
+    {
+      "epoch": 0.9675156321488485,
+      "grad_norm": 0.2173855057127278,
+      "learning_rate": 6.154805179168122e-06,
+      "loss": 0.2299,
+      "step": 1586
+    },
+    {
+      "epoch": 0.9681256672258655,
+      "grad_norm": 0.22955242541686519,
+      "learning_rate": 6.149627267701212e-06,
+      "loss": 0.2314,
+      "step": 1587
+    },
+    {
+      "epoch": 0.9687357023028824,
+      "grad_norm": 0.24358603947580434,
+      "learning_rate": 6.144448054195119e-06,
+      "loss": 0.2214,
+      "step": 1588
+    },
+    {
+      "epoch": 0.9693457373798994,
+      "grad_norm": 0.22127455539624566,
+      "learning_rate": 6.139267544515689e-06,
+      "loss": 0.2297,
+      "step": 1589
+    },
+    {
+      "epoch": 0.9699557724569162,
+      "grad_norm": 0.24106981499557115,
+      "learning_rate": 6.1340857445302395e-06,
+      "loss": 0.227,
+      "step": 1590
+    },
+    {
+      "epoch": 0.9705658075339332,
+      "grad_norm": 0.22531162899583024,
+      "learning_rate": 6.128902660107547e-06,
+      "loss": 0.23,
+      "step": 1591
+    },
+    {
+      "epoch": 0.9711758426109501,
+      "grad_norm": 0.21576028626875046,
+      "learning_rate": 6.123718297117844e-06,
+      "loss": 0.2182,
+      "step": 1592
+    },
+    {
+      "epoch": 0.9717858776879671,
+      "grad_norm": 0.22383451511812855,
+      "learning_rate": 6.118532661432812e-06,
+      "loss": 0.2192,
+      "step": 1593
+    },
+    {
+      "epoch": 0.9723959127649839,
+      "grad_norm": 0.22162444020782598,
+      "learning_rate": 6.113345758925572e-06,
+      "loss": 0.2337,
+      "step": 1594
+    },
+    {
+      "epoch": 0.9730059478420009,
+      "grad_norm": 0.23306688141707715,
+      "learning_rate": 6.10815759547068e-06,
+      "loss": 0.2327,
+      "step": 1595
+    },
+    {
+      "epoch": 0.9736159829190179,
+      "grad_norm": 0.24668096163914657,
+      "learning_rate": 6.1029681769441195e-06,
+      "loss": 0.2362,
+      "step": 1596
+    },
+    {
+      "epoch": 0.9742260179960348,
+      "grad_norm": 0.22771560133975774,
+      "learning_rate": 6.097777509223299e-06,
+      "loss": 0.2181,
+      "step": 1597
+    },
+    {
+      "epoch": 0.9748360530730517,
+      "grad_norm": 0.227152471220962,
+      "learning_rate": 6.09258559818704e-06,
+      "loss": 0.228,
+      "step": 1598
+    },
+    {
+      "epoch": 0.9754460881500686,
+      "grad_norm": 0.24505550426884296,
+      "learning_rate": 6.087392449715568e-06,
+      "loss": 0.2365,
+      "step": 1599
+    },
+    {
+      "epoch": 0.9760561232270856,
+      "grad_norm": 0.2594678512272388,
+      "learning_rate": 6.0821980696905145e-06,
+      "loss": 0.2347,
+      "step": 1600
+    },
+    {
+      "epoch": 0.9766661583041025,
+      "grad_norm": 0.22648860453973974,
+      "learning_rate": 6.077002463994908e-06,
+      "loss": 0.2292,
+      "step": 1601
+    },
+    {
+      "epoch": 0.9772761933811194,
+      "grad_norm": 0.217024573137216,
+      "learning_rate": 6.0718056385131575e-06,
+      "loss": 0.2097,
+      "step": 1602
+    },
+    {
+      "epoch": 0.9778862284581363,
+      "grad_norm": 0.23583942188243961,
+      "learning_rate": 6.066607599131061e-06,
+      "loss": 0.2061,
+      "step": 1603
+    },
+    {
+      "epoch": 0.9784962635351533,
+      "grad_norm": 0.22690683270385684,
+      "learning_rate": 6.061408351735787e-06,
+      "loss": 0.2376,
+      "step": 1604
+    },
+    {
+      "epoch": 0.9791062986121702,
+      "grad_norm": 0.22111549660672086,
+      "learning_rate": 6.056207902215874e-06,
+      "loss": 0.2213,
+      "step": 1605
+    },
+    {
+      "epoch": 0.9797163336891871,
+      "grad_norm": 0.26246371397002916,
+      "learning_rate": 6.051006256461221e-06,
+      "loss": 0.2174,
+      "step": 1606
+    },
+    {
+      "epoch": 0.980326368766204,
+      "grad_norm": 0.22829641873217374,
+      "learning_rate": 6.045803420363085e-06,
+      "loss": 0.2238,
+      "step": 1607
+    },
+    {
+      "epoch": 0.980936403843221,
+      "grad_norm": 0.22819806706388343,
+      "learning_rate": 6.040599399814064e-06,
+      "loss": 0.2078,
+      "step": 1608
+    },
+    {
+      "epoch": 0.981546438920238,
+      "grad_norm": 0.23728526726592925,
+      "learning_rate": 6.035394200708104e-06,
+      "loss": 0.2246,
+      "step": 1609
+    },
+    {
+      "epoch": 0.9821564739972548,
+      "grad_norm": 0.22654727333712824,
+      "learning_rate": 6.030187828940485e-06,
+      "loss": 0.2085,
+      "step": 1610
+    },
+    {
+      "epoch": 0.9827665090742718,
+      "grad_norm": 0.2181113150986954,
+      "learning_rate": 6.024980290407813e-06,
+      "loss": 0.2051,
+      "step": 1611
+    },
+    {
+      "epoch": 0.9833765441512887,
+      "grad_norm": 0.22462912016487835,
+      "learning_rate": 6.019771591008015e-06,
+      "loss": 0.2116,
+      "step": 1612
+    },
+    {
+      "epoch": 0.9839865792283057,
+      "grad_norm": 0.2179899542904998,
+      "learning_rate": 6.014561736640334e-06,
+      "loss": 0.2244,
+      "step": 1613
+    },
+    {
+      "epoch": 0.9845966143053225,
+      "grad_norm": 0.2286697668059702,
+      "learning_rate": 6.009350733205324e-06,
+      "loss": 0.2286,
+      "step": 1614
+    },
+    {
+      "epoch": 0.9852066493823395,
+      "grad_norm": 0.24353389205515807,
+      "learning_rate": 6.0041385866048356e-06,
+      "loss": 0.2314,
+      "step": 1615
+    },
+    {
+      "epoch": 0.9858166844593564,
+      "grad_norm": 0.217771023881922,
+      "learning_rate": 5.998925302742017e-06,
+      "loss": 0.2182,
+      "step": 1616
+    },
+    {
+      "epoch": 0.9864267195363733,
+      "grad_norm": 0.2475617013818729,
+      "learning_rate": 5.993710887521302e-06,
+      "loss": 0.2686,
+      "step": 1617
+    },
+    {
+      "epoch": 0.9870367546133902,
+      "grad_norm": 0.23044990847915867,
+      "learning_rate": 5.988495346848409e-06,
+      "loss": 0.2369,
+      "step": 1618
+    },
+    {
+      "epoch": 0.9876467896904072,
+      "grad_norm": 0.2545012934007029,
+      "learning_rate": 5.983278686630327e-06,
+      "loss": 0.2435,
+      "step": 1619
+    },
+    {
+      "epoch": 0.9882568247674242,
+      "grad_norm": 0.26432844279930834,
+      "learning_rate": 5.978060912775319e-06,
+      "loss": 0.2241,
+      "step": 1620
+    },
+    {
+      "epoch": 0.988866859844441,
+      "grad_norm": 0.22084458591049622,
+      "learning_rate": 5.972842031192901e-06,
+      "loss": 0.2143,
+      "step": 1621
+    },
+    {
+      "epoch": 0.989476894921458,
+      "grad_norm": 0.2170629136166457,
+      "learning_rate": 5.967622047793853e-06,
+      "loss": 0.2133,
+      "step": 1622
+    },
+    {
+      "epoch": 0.9900869299984749,
+      "grad_norm": 0.21954983443109286,
+      "learning_rate": 5.962400968490194e-06,
+      "loss": 0.2265,
+      "step": 1623
+    },
+    {
+      "epoch": 0.9906969650754919,
+      "grad_norm": 0.2305444478220999,
+      "learning_rate": 5.957178799195191e-06,
+      "loss": 0.2213,
+      "step": 1624
+    },
+    {
+      "epoch": 0.9913070001525087,
+      "grad_norm": 0.23003753615902042,
+      "learning_rate": 5.951955545823342e-06,
+      "loss": 0.2289,
+      "step": 1625
+    },
+    {
+      "epoch": 0.9919170352295257,
+      "grad_norm": 0.23140788001630902,
+      "learning_rate": 5.946731214290373e-06,
+      "loss": 0.246,
+      "step": 1626
+    },
+    {
+      "epoch": 0.9925270703065426,
+      "grad_norm": 0.21227594828029972,
+      "learning_rate": 5.941505810513233e-06,
+      "loss": 0.2031,
+      "step": 1627
+    },
+    {
+      "epoch": 0.9931371053835596,
+      "grad_norm": 0.22010918804701388,
+      "learning_rate": 5.936279340410082e-06,
+      "loss": 0.2288,
+      "step": 1628
+    },
+    {
+      "epoch": 0.9937471404605764,
+      "grad_norm": 0.23602106031097625,
+      "learning_rate": 5.93105180990029e-06,
+      "loss": 0.2179,
+      "step": 1629
+    },
+    {
+      "epoch": 0.9943571755375934,
+      "grad_norm": 0.2309016046215796,
+      "learning_rate": 5.925823224904429e-06,
+      "loss": 0.2104,
+      "step": 1630
+    },
+    {
+      "epoch": 0.9949672106146104,
+      "grad_norm": 0.23950678926346633,
+      "learning_rate": 5.920593591344264e-06,
+      "loss": 0.2311,
+      "step": 1631
+    },
+    {
+      "epoch": 0.9955772456916273,
+      "grad_norm": 0.22637480916172084,
+      "learning_rate": 5.9153629151427475e-06,
+      "loss": 0.2307,
+      "step": 1632
+    },
+    {
+      "epoch": 0.9961872807686442,
+      "grad_norm": 0.22041864861326502,
+      "learning_rate": 5.910131202224011e-06,
+      "loss": 0.2083,
+      "step": 1633
+    },
+    {
+      "epoch": 0.9967973158456611,
+      "grad_norm": 0.22650306744690762,
+      "learning_rate": 5.904898458513365e-06,
+      "loss": 0.2206,
+      "step": 1634
+    },
+    {
+      "epoch": 0.9974073509226781,
+      "grad_norm": 0.22749499229284315,
+      "learning_rate": 5.899664689937283e-06,
+      "loss": 0.2373,
+      "step": 1635
+    },
+    {
+      "epoch": 0.998017385999695,
+      "grad_norm": 0.21288210761993082,
+      "learning_rate": 5.894429902423402e-06,
+      "loss": 0.2173,
+      "step": 1636
+    },
+    {
+      "epoch": 0.9986274210767119,
+      "grad_norm": 0.2217043488048322,
+      "learning_rate": 5.8891941019005095e-06,
+      "loss": 0.2329,
+      "step": 1637
+    },
+    {
+      "epoch": 0.9992374561537288,
+      "grad_norm": 0.27082147748875746,
+      "learning_rate": 5.883957294298545e-06,
+      "loss": 0.237,
+      "step": 1638
+    },
+    {
+      "epoch": 0.9998474912307458,
+      "grad_norm": 0.2390081544955565,
+      "learning_rate": 5.8787194855485855e-06,
+      "loss": 0.2498,
+      "step": 1639
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4362468690070915,
+      "learning_rate": 5.87348068158284e-06,
+      "loss": 0.2575,
+      "step": 1640
+    },
+    {
+      "epoch": 1.0006100350770168,
+      "grad_norm": 0.24525124320565972,
+      "learning_rate": 5.8682408883346535e-06,
+      "loss": 0.2165,
+      "step": 1641
+    },
+    {
+      "epoch": 1.001220070154034,
+      "grad_norm": 0.22532467581428622,
+      "learning_rate": 5.863000111738479e-06,
+      "loss": 0.1919,
+      "step": 1642
+    },
+    {
+      "epoch": 1.0018301052310508,
+      "grad_norm": 0.25364441859572456,
+      "learning_rate": 5.857758357729892e-06,
+      "loss": 0.2281,
+      "step": 1643
+    },
+    {
+      "epoch": 1.0024401403080676,
+      "grad_norm": 0.23172737765478774,
+      "learning_rate": 5.852515632245574e-06,
+      "loss": 0.2036,
+      "step": 1644
+    },
+    {
+      "epoch": 1.0030501753850847,
+      "grad_norm": 0.22918099355509405,
+      "learning_rate": 5.847271941223301e-06,
+      "loss": 0.2194,
+      "step": 1645
+    },
+    {
+      "epoch": 1.0036602104621015,
+      "grad_norm": 0.2328036742592211,
+      "learning_rate": 5.8420272906019506e-06,
+      "loss": 0.2154,
+      "step": 1646
+    },
+    {
+      "epoch": 1.0042702455391186,
+      "grad_norm": 0.2528041564358456,
+      "learning_rate": 5.8367816863214825e-06,
+      "loss": 0.2251,
+      "step": 1647
+    },
+    {
+      "epoch": 1.0048802806161354,
+      "grad_norm": 0.24623727405835405,
+      "learning_rate": 5.831535134322935e-06,
+      "loss": 0.2192,
+      "step": 1648
+    },
+    {
+      "epoch": 1.0054903156931523,
+      "grad_norm": 0.23508962134126507,
+      "learning_rate": 5.826287640548425e-06,
+      "loss": 0.2014,
+      "step": 1649
+    },
+    {
+      "epoch": 1.0061003507701693,
+      "grad_norm": 0.2545397092728682,
+      "learning_rate": 5.821039210941133e-06,
+      "loss": 0.2247,
+      "step": 1650
+    },
+    {
+      "epoch": 1.0067103858471862,
+      "grad_norm": 0.2341766055191371,
+      "learning_rate": 5.815789851445296e-06,
+      "loss": 0.2184,
+      "step": 1651
+    },
+    {
+      "epoch": 1.007320420924203,
+      "grad_norm": 0.2304950075693666,
+      "learning_rate": 5.810539568006213e-06,
+      "loss": 0.1997,
+      "step": 1652
+    },
+    {
+      "epoch": 1.00793045600122,
+      "grad_norm": 0.23771967596273264,
+      "learning_rate": 5.80528836657022e-06,
+      "loss": 0.2124,
+      "step": 1653
+    },
+    {
+      "epoch": 1.008540491078237,
+      "grad_norm": 0.23310797936523897,
+      "learning_rate": 5.800036253084699e-06,
+      "loss": 0.2084,
+      "step": 1654
+    },
+    {
+      "epoch": 1.009150526155254,
+      "grad_norm": 0.24051112098620298,
+      "learning_rate": 5.794783233498062e-06,
+      "loss": 0.2118,
+      "step": 1655
+    },
+    {
+      "epoch": 1.0097605612322709,
+      "grad_norm": 0.22746424422800543,
+      "learning_rate": 5.789529313759746e-06,
+      "loss": 0.2139,
+      "step": 1656
+    },
+    {
+      "epoch": 1.0103705963092877,
+      "grad_norm": 0.24267923602790298,
+      "learning_rate": 5.784274499820214e-06,
+      "loss": 0.2071,
+      "step": 1657
+    },
+    {
+      "epoch": 1.0109806313863048,
+      "grad_norm": 0.24322898160992537,
+      "learning_rate": 5.779018797630934e-06,
+      "loss": 0.2041,
+      "step": 1658
+    },
+    {
+      "epoch": 1.0115906664633216,
+      "grad_norm": 0.23510639718676693,
+      "learning_rate": 5.773762213144384e-06,
+      "loss": 0.2333,
+      "step": 1659
+    },
+    {
+      "epoch": 1.0122007015403385,
+      "grad_norm": 0.2447061546009269,
+      "learning_rate": 5.76850475231404e-06,
+      "loss": 0.2167,
+      "step": 1660
+    },
+    {
+      "epoch": 1.0128107366173555,
+      "grad_norm": 0.2564959269452337,
+      "learning_rate": 5.763246421094373e-06,
+      "loss": 0.2134,
+      "step": 1661
+    },
+    {
+      "epoch": 1.0134207716943724,
+      "grad_norm": 0.2388834769977569,
+      "learning_rate": 5.757987225440836e-06,
+      "loss": 0.1941,
+      "step": 1662
+    },
+    {
+      "epoch": 1.0140308067713895,
+      "grad_norm": 0.2232628943076604,
+      "learning_rate": 5.7527271713098645e-06,
+      "loss": 0.2023,
+      "step": 1663
+    },
+    {
+      "epoch": 1.0146408418484063,
+      "grad_norm": 0.23877377779070566,
+      "learning_rate": 5.747466264658863e-06,
+      "loss": 0.2216,
+      "step": 1664
+    },
+    {
+      "epoch": 1.0152508769254232,
+      "grad_norm": 0.23347183808731123,
+      "learning_rate": 5.742204511446203e-06,
+      "loss": 0.2154,
+      "step": 1665
+    },
+    {
+      "epoch": 1.0158609120024402,
+      "grad_norm": 0.23259316764185303,
+      "learning_rate": 5.736941917631217e-06,
+      "loss": 0.214,
+      "step": 1666
+    },
+    {
+      "epoch": 1.016470947079457,
+      "grad_norm": 0.2524496772639378,
+      "learning_rate": 5.731678489174186e-06,
+      "loss": 0.2389,
+      "step": 1667
+    },
+    {
+      "epoch": 1.017080982156474,
+      "grad_norm": 0.22833440973811123,
+      "learning_rate": 5.7264142320363384e-06,
+      "loss": 0.213,
+      "step": 1668
+    },
+    {
+      "epoch": 1.017691017233491,
+      "grad_norm": 0.2361574490453357,
+      "learning_rate": 5.72114915217984e-06,
+      "loss": 0.219,
+      "step": 1669
+    },
+    {
+      "epoch": 1.0183010523105078,
+      "grad_norm": 0.2280786951276804,
+      "learning_rate": 5.7158832555677904e-06,
+      "loss": 0.2108,
+      "step": 1670
+    },
+    {
+      "epoch": 1.018911087387525,
+      "grad_norm": 0.23367357132543207,
+      "learning_rate": 5.710616548164212e-06,
+      "loss": 0.2061,
+      "step": 1671
+    },
+    {
+      "epoch": 1.0195211224645417,
+      "grad_norm": 0.236030915869193,
+      "learning_rate": 5.705349035934047e-06,
+      "loss": 0.2025,
+      "step": 1672
+    },
+    {
+      "epoch": 1.0201311575415586,
+      "grad_norm": 0.22653660507034204,
+      "learning_rate": 5.7000807248431466e-06,
+      "loss": 0.1959,
+      "step": 1673
+    },
+    {
+      "epoch": 1.0207411926185757,
+      "grad_norm": 0.24496796300881274,
+      "learning_rate": 5.694811620858269e-06,
+      "loss": 0.2132,
+      "step": 1674
+    },
+    {
+      "epoch": 1.0213512276955925,
+      "grad_norm": 0.23452084512565427,
+      "learning_rate": 5.689541729947071e-06,
+      "loss": 0.2209,
+      "step": 1675
+    },
+    {
+      "epoch": 1.0219612627726093,
+      "grad_norm": 0.25138673187324173,
+      "learning_rate": 5.684271058078101e-06,
+      "loss": 0.2199,
+      "step": 1676
+    },
+    {
+      "epoch": 1.0225712978496264,
+      "grad_norm": 0.24117248054105317,
+      "learning_rate": 5.6789996112207865e-06,
+      "loss": 0.2204,
+      "step": 1677
+    },
+    {
+      "epoch": 1.0231813329266433,
+      "grad_norm": 0.23694095756163946,
+      "learning_rate": 5.673727395345442e-06,
+      "loss": 0.2154,
+      "step": 1678
+    },
+    {
+      "epoch": 1.02379136800366,
+      "grad_norm": 0.23021351407090954,
+      "learning_rate": 5.668454416423243e-06,
+      "loss": 0.2344,
+      "step": 1679
+    },
+    {
+      "epoch": 1.0244014030806772,
+      "grad_norm": 0.24354787085020255,
+      "learning_rate": 5.663180680426237e-06,
+      "loss": 0.2215,
+      "step": 1680
+    },
+    {
+      "epoch": 1.025011438157694,
+      "grad_norm": 0.23398078469245429,
+      "learning_rate": 5.657906193327325e-06,
+      "loss": 0.1964,
+      "step": 1681
+    },
+    {
+      "epoch": 1.025621473234711,
+      "grad_norm": 0.23555210978513935,
+      "learning_rate": 5.65263096110026e-06,
+      "loss": 0.2113,
+      "step": 1682
+    },
+    {
+      "epoch": 1.026231508311728,
+      "grad_norm": 0.24099367744273603,
+      "learning_rate": 5.647354989719635e-06,
+      "loss": 0.2326,
+      "step": 1683
+    },
+    {
+      "epoch": 1.0268415433887448,
+      "grad_norm": 0.2364743426107808,
+      "learning_rate": 5.64207828516089e-06,
+      "loss": 0.213,
+      "step": 1684
+    },
+    {
+      "epoch": 1.0274515784657619,
+      "grad_norm": 0.24114895326828484,
+      "learning_rate": 5.636800853400285e-06,
+      "loss": 0.2187,
+      "step": 1685
+    },
+    {
+      "epoch": 1.0280616135427787,
+      "grad_norm": 0.22602307697039975,
+      "learning_rate": 5.631522700414909e-06,
+      "loss": 0.2138,
+      "step": 1686
+    },
+    {
+      "epoch": 1.0286716486197955,
+      "grad_norm": 0.22019981637444508,
+      "learning_rate": 5.626243832182663e-06,
+      "loss": 0.1822,
+      "step": 1687
+    },
+    {
+      "epoch": 1.0292816836968126,
+      "grad_norm": 0.24498869528300118,
+      "learning_rate": 5.620964254682267e-06,
+      "loss": 0.229,
+      "step": 1688
+    },
+    {
+      "epoch": 1.0298917187738295,
+      "grad_norm": 0.22457143436015983,
+      "learning_rate": 5.615683973893235e-06,
+      "loss": 0.2321,
+      "step": 1689
+    },
+    {
+      "epoch": 1.0305017538508465,
+      "grad_norm": 0.230241215035034,
+      "learning_rate": 5.610402995795884e-06,
+      "loss": 0.2092,
+      "step": 1690
+    },
+    {
+      "epoch": 1.0311117889278634,
+      "grad_norm": 0.23816566700218206,
+      "learning_rate": 5.605121326371316e-06,
+      "loss": 0.2201,
+      "step": 1691
+    },
+    {
+      "epoch": 1.0317218240048802,
+      "grad_norm": 0.2450343849705002,
+      "learning_rate": 5.599838971601418e-06,
+      "loss": 0.2182,
+      "step": 1692
+    },
+    {
+      "epoch": 1.0323318590818973,
+      "grad_norm": 0.24422927352433343,
+      "learning_rate": 5.594555937468856e-06,
+      "loss": 0.2252,
+      "step": 1693
+    },
+    {
+      "epoch": 1.0329418941589141,
+      "grad_norm": 0.2343514132682759,
+      "learning_rate": 5.589272229957061e-06,
+      "loss": 0.185,
+      "step": 1694
+    },
+    {
+      "epoch": 1.033551929235931,
+      "grad_norm": 0.2361203993950127,
+      "learning_rate": 5.583987855050228e-06,
+      "loss": 0.2063,
+      "step": 1695
+    },
+    {
+      "epoch": 1.034161964312948,
+      "grad_norm": 0.2565116992016366,
+      "learning_rate": 5.578702818733308e-06,
+      "loss": 0.2156,
+      "step": 1696
+    },
+    {
+      "epoch": 1.034771999389965,
+      "grad_norm": 0.2546156826740772,
+      "learning_rate": 5.573417126992004e-06,
+      "loss": 0.2415,
+      "step": 1697
+    },
+    {
+      "epoch": 1.035382034466982,
+      "grad_norm": 0.23336883260302368,
+      "learning_rate": 5.568130785812755e-06,
+      "loss": 0.222,
+      "step": 1698
+    },
+    {
+      "epoch": 1.0359920695439988,
+      "grad_norm": 0.25462328137310886,
+      "learning_rate": 5.562843801182741e-06,
+      "loss": 0.2129,
+      "step": 1699
+    },
+    {
+      "epoch": 1.0366021046210157,
+      "grad_norm": 0.24585946060544714,
+      "learning_rate": 5.5575561790898705e-06,
+      "loss": 0.235,
+      "step": 1700
+    },
+    {
+      "epoch": 1.0372121396980327,
+      "grad_norm": 0.24127705207317346,
+      "learning_rate": 5.55226792552277e-06,
+      "loss": 0.2231,
+      "step": 1701
+    },
+    {
+      "epoch": 1.0378221747750496,
+      "grad_norm": 0.24249186568336176,
+      "learning_rate": 5.546979046470784e-06,
+      "loss": 0.2198,
+      "step": 1702
+    },
+    {
+      "epoch": 1.0384322098520664,
+      "grad_norm": 0.2841185684942545,
+      "learning_rate": 5.5416895479239665e-06,
+      "loss": 0.2212,
+      "step": 1703
+    },
+    {
+      "epoch": 1.0390422449290835,
+      "grad_norm": 0.22862341833091082,
+      "learning_rate": 5.536399435873071e-06,
+      "loss": 0.2184,
+      "step": 1704
+    },
+    {
+      "epoch": 1.0396522800061003,
+      "grad_norm": 0.2511982074093489,
+      "learning_rate": 5.5311087163095475e-06,
+      "loss": 0.2066,
+      "step": 1705
+    },
+    {
+      "epoch": 1.0402623150831172,
+      "grad_norm": 0.2313982450490478,
+      "learning_rate": 5.52581739522553e-06,
+      "loss": 0.212,
+      "step": 1706
+    },
+    {
+      "epoch": 1.0408723501601342,
+      "grad_norm": 0.23697031192994897,
+      "learning_rate": 5.520525478613838e-06,
+      "loss": 0.2121,
+      "step": 1707
+    },
+    {
+      "epoch": 1.041482385237151,
+      "grad_norm": 0.2561892836965319,
+      "learning_rate": 5.515232972467969e-06,
+      "loss": 0.2199,
+      "step": 1708
+    },
+    {
+      "epoch": 1.0420924203141682,
+      "grad_norm": 0.244397771634611,
+      "learning_rate": 5.509939882782077e-06,
+      "loss": 0.2068,
+      "step": 1709
+    },
+    {
+      "epoch": 1.042702455391185,
+      "grad_norm": 0.23380824757735447,
+      "learning_rate": 5.504646215550988e-06,
+      "loss": 0.2211,
+      "step": 1710
+    },
+    {
+      "epoch": 1.0433124904682018,
+      "grad_norm": 0.2452556417197051,
+      "learning_rate": 5.4993519767701745e-06,
+      "loss": 0.2125,
+      "step": 1711
+    },
+    {
+      "epoch": 1.043922525545219,
+      "grad_norm": 0.2305296614582525,
+      "learning_rate": 5.494057172435761e-06,
+      "loss": 0.2102,
+      "step": 1712
+    },
+    {
+      "epoch": 1.0445325606222358,
+      "grad_norm": 0.24478341739491696,
+      "learning_rate": 5.4887618085445094e-06,
+      "loss": 0.2128,
+      "step": 1713
+    },
+    {
+      "epoch": 1.0451425956992526,
+      "grad_norm": 0.22777418813905734,
+      "learning_rate": 5.48346589109382e-06,
+      "loss": 0.1938,
+      "step": 1714
+    },
+    {
+      "epoch": 1.0457526307762697,
+      "grad_norm": 0.2596034979250406,
+      "learning_rate": 5.478169426081712e-06,
+      "loss": 0.2123,
+      "step": 1715
+    },
+    {
+      "epoch": 1.0463626658532865,
+      "grad_norm": 0.241016728765862,
+      "learning_rate": 5.472872419506833e-06,
+      "loss": 0.2092,
+      "step": 1716
+    },
+    {
+      "epoch": 1.0469727009303036,
+      "grad_norm": 0.21788730778084672,
+      "learning_rate": 5.467574877368441e-06,
+      "loss": 0.1909,
+      "step": 1717
+    },
+    {
+      "epoch": 1.0475827360073204,
+      "grad_norm": 0.24697089084624776,
+      "learning_rate": 5.4622768056664e-06,
+      "loss": 0.2083,
+      "step": 1718
+    },
+    {
+      "epoch": 1.0481927710843373,
+      "grad_norm": 0.25807379681608617,
+      "learning_rate": 5.4569782104011725e-06,
+      "loss": 0.2187,
+      "step": 1719
+    },
+    {
+      "epoch": 1.0488028061613544,
+      "grad_norm": 0.24314953105974776,
+      "learning_rate": 5.4516790975738146e-06,
+      "loss": 0.2248,
+      "step": 1720
+    },
+    {
+      "epoch": 1.0494128412383712,
+      "grad_norm": 0.23758838267771906,
+      "learning_rate": 5.446379473185972e-06,
+      "loss": 0.1975,
+      "step": 1721
+    },
+    {
+      "epoch": 1.050022876315388,
+      "grad_norm": 0.25168940382766586,
+      "learning_rate": 5.441079343239867e-06,
+      "loss": 0.2196,
+      "step": 1722
+    },
+    {
+      "epoch": 1.0506329113924051,
+      "grad_norm": 0.23974904667847913,
+      "learning_rate": 5.435778713738292e-06,
+      "loss": 0.2012,
+      "step": 1723
+    },
+    {
+      "epoch": 1.051242946469422,
+      "grad_norm": 0.225951684544416,
+      "learning_rate": 5.4304775906846095e-06,
+      "loss": 0.2175,
+      "step": 1724
+    },
+    {
+      "epoch": 1.051852981546439,
+      "grad_norm": 0.23899014950838993,
+      "learning_rate": 5.42517598008274e-06,
+      "loss": 0.206,
+      "step": 1725
+    },
+    {
+      "epoch": 1.0524630166234559,
+      "grad_norm": 0.21642954447018442,
+      "learning_rate": 5.419873887937154e-06,
+      "loss": 0.1942,
+      "step": 1726
+    },
+    {
+      "epoch": 1.0530730517004727,
+      "grad_norm": 0.2287863512942669,
+      "learning_rate": 5.41457132025287e-06,
+      "loss": 0.2156,
+      "step": 1727
+    },
+    {
+      "epoch": 1.0536830867774898,
+      "grad_norm": 0.23290535502397355,
+      "learning_rate": 5.4092682830354435e-06,
+      "loss": 0.2085,
+      "step": 1728
+    },
+    {
+      "epoch": 1.0542931218545066,
+      "grad_norm": 0.23491286063679967,
+      "learning_rate": 5.403964782290962e-06,
+      "loss": 0.224,
+      "step": 1729
+    },
+    {
+      "epoch": 1.0549031569315235,
+      "grad_norm": 0.22898564744066244,
+      "learning_rate": 5.398660824026039e-06,
+      "loss": 0.2147,
+      "step": 1730
+    },
+    {
+      "epoch": 1.0555131920085405,
+      "grad_norm": 0.23321342607229884,
+      "learning_rate": 5.393356414247806e-06,
+      "loss": 0.189,
+      "step": 1731
+    },
+    {
+      "epoch": 1.0561232270855574,
+      "grad_norm": 0.24018281207640957,
+      "learning_rate": 5.388051558963904e-06,
+      "loss": 0.208,
+      "step": 1732
+    },
+    {
+      "epoch": 1.0567332621625742,
+      "grad_norm": 0.22159660079783608,
+      "learning_rate": 5.38274626418248e-06,
+      "loss": 0.1988,
+      "step": 1733
+    },
+    {
+      "epoch": 1.0573432972395913,
+      "grad_norm": 0.23295975100519303,
+      "learning_rate": 5.377440535912177e-06,
+      "loss": 0.2113,
+      "step": 1734
+    },
+    {
+      "epoch": 1.0579533323166082,
+      "grad_norm": 0.23448234332441525,
+      "learning_rate": 5.372134380162133e-06,
+      "loss": 0.2192,
+      "step": 1735
+    },
+    {
+      "epoch": 1.0585633673936252,
+      "grad_norm": 0.24268607330157105,
+      "learning_rate": 5.366827802941968e-06,
+      "loss": 0.2068,
+      "step": 1736
+    },
+    {
+      "epoch": 1.059173402470642,
+      "grad_norm": 0.2279938492718881,
+      "learning_rate": 5.361520810261779e-06,
+      "loss": 0.2069,
+      "step": 1737
+    },
+    {
+      "epoch": 1.059783437547659,
+      "grad_norm": 0.23729437606366588,
+      "learning_rate": 5.356213408132131e-06,
+      "loss": 0.2152,
+      "step": 1738
+    },
+    {
+      "epoch": 1.060393472624676,
+      "grad_norm": 0.22743666468075752,
+      "learning_rate": 5.3509056025640575e-06,
+      "loss": 0.2154,
+      "step": 1739
+    },
+    {
+      "epoch": 1.0610035077016928,
+      "grad_norm": 0.22491802092028118,
+      "learning_rate": 5.345597399569047e-06,
+      "loss": 0.217,
+      "step": 1740
+    },
+    {
+      "epoch": 1.0616135427787097,
+      "grad_norm": 0.22367511029922768,
+      "learning_rate": 5.340288805159037e-06,
+      "loss": 0.2202,
+      "step": 1741
+    },
+    {
+      "epoch": 1.0622235778557267,
+      "grad_norm": 0.22623616900798266,
+      "learning_rate": 5.334979825346409e-06,
+      "loss": 0.201,
+      "step": 1742
+    },
+    {
+      "epoch": 1.0628336129327436,
+      "grad_norm": 0.22928684429359641,
+      "learning_rate": 5.329670466143981e-06,
+      "loss": 0.2082,
+      "step": 1743
+    },
+    {
+      "epoch": 1.0634436480097607,
+      "grad_norm": 0.25200001642666464,
+      "learning_rate": 5.3243607335650014e-06,
+      "loss": 0.2215,
+      "step": 1744
+    },
+    {
+      "epoch": 1.0640536830867775,
+      "grad_norm": 0.23013795665760164,
+      "learning_rate": 5.319050633623141e-06,
+      "loss": 0.2219,
+      "step": 1745
+    },
+    {
+      "epoch": 1.0646637181637943,
+      "grad_norm": 0.23397812159289222,
+      "learning_rate": 5.3137401723324885e-06,
+      "loss": 0.2324,
+      "step": 1746
+    },
+    {
+      "epoch": 1.0652737532408114,
+      "grad_norm": 0.22448906533672924,
+      "learning_rate": 5.308429355707538e-06,
+      "loss": 0.2243,
+      "step": 1747
+    },
+    {
+      "epoch": 1.0658837883178283,
+      "grad_norm": 0.2268116613353298,
+      "learning_rate": 5.303118189763187e-06,
+      "loss": 0.1987,
+      "step": 1748
+    },
+    {
+      "epoch": 1.066493823394845,
+      "grad_norm": 0.25510457913585216,
+      "learning_rate": 5.297806680514731e-06,
+      "loss": 0.207,
+      "step": 1749
+    },
+    {
+      "epoch": 1.0671038584718622,
+      "grad_norm": 0.23858305824485018,
+      "learning_rate": 5.292494833977853e-06,
+      "loss": 0.202,
+      "step": 1750
+    },
+    {
+      "epoch": 1.067713893548879,
+      "grad_norm": 0.22279142082214048,
+      "learning_rate": 5.287182656168618e-06,
+      "loss": 0.204,
+      "step": 1751
+    },
+    {
+      "epoch": 1.068323928625896,
+      "grad_norm": 0.22140767823445726,
+      "learning_rate": 5.281870153103464e-06,
+      "loss": 0.2168,
+      "step": 1752
+    },
+    {
+      "epoch": 1.068933963702913,
+      "grad_norm": 0.24279225254617254,
+      "learning_rate": 5.276557330799203e-06,
+      "loss": 0.2199,
+      "step": 1753
+    },
+    {
+      "epoch": 1.0695439987799298,
+      "grad_norm": 0.22987657138259443,
+      "learning_rate": 5.271244195273002e-06,
+      "loss": 0.2179,
+      "step": 1754
+    },
+    {
+      "epoch": 1.0701540338569469,
+      "grad_norm": 0.2242483145934005,
+      "learning_rate": 5.2659307525423895e-06,
+      "loss": 0.1995,
+      "step": 1755
+    },
+    {
+      "epoch": 1.0707640689339637,
+      "grad_norm": 0.23595573283445498,
+      "learning_rate": 5.260617008625235e-06,
+      "loss": 0.2001,
+      "step": 1756
+    },
+    {
+      "epoch": 1.0713741040109805,
+      "grad_norm": 0.2291741674921307,
+      "learning_rate": 5.255302969539753e-06,
+      "loss": 0.2034,
+      "step": 1757
+    },
+    {
+      "epoch": 1.0719841390879976,
+      "grad_norm": 0.22887782966092776,
+      "learning_rate": 5.2499886413044934e-06,
+      "loss": 0.2059,
+      "step": 1758
+    },
+    {
+      "epoch": 1.0725941741650145,
+      "grad_norm": 0.22360670111873499,
+      "learning_rate": 5.24467402993833e-06,
+      "loss": 0.2058,
+      "step": 1759
+    },
+    {
+      "epoch": 1.0732042092420313,
+      "grad_norm": 0.24012499144382562,
+      "learning_rate": 5.239359141460461e-06,
+      "loss": 0.2179,
+      "step": 1760
+    },
+    {
+      "epoch": 1.0738142443190484,
+      "grad_norm": 0.31511244352811957,
+      "learning_rate": 5.234043981890395e-06,
+      "loss": 0.2167,
+      "step": 1761
+    },
+    {
+      "epoch": 1.0744242793960652,
+      "grad_norm": 0.2266221058824141,
+      "learning_rate": 5.2287285572479475e-06,
+      "loss": 0.2051,
+      "step": 1762
+    },
+    {
+      "epoch": 1.0750343144730823,
+      "grad_norm": 0.22440582251739094,
+      "learning_rate": 5.22341287355324e-06,
+      "loss": 0.2028,
+      "step": 1763
+    },
+    {
+      "epoch": 1.0756443495500991,
+      "grad_norm": 0.2157945368902979,
+      "learning_rate": 5.218096936826681e-06,
+      "loss": 0.1913,
+      "step": 1764
+    },
+    {
+      "epoch": 1.076254384627116,
+      "grad_norm": 0.22696069047564313,
+      "learning_rate": 5.212780753088968e-06,
+      "loss": 0.2161,
+      "step": 1765
+    },
+    {
+      "epoch": 1.076864419704133,
+      "grad_norm": 0.22411271328064739,
+      "learning_rate": 5.207464328361078e-06,
+      "loss": 0.2134,
+      "step": 1766
+    },
+    {
+      "epoch": 1.07747445478115,
+      "grad_norm": 0.2163020875646112,
+      "learning_rate": 5.202147668664264e-06,
+      "loss": 0.207,
+      "step": 1767
+    },
+    {
+      "epoch": 1.078084489858167,
+      "grad_norm": 0.25491316943478276,
+      "learning_rate": 5.196830780020038e-06,
+      "loss": 0.2255,
+      "step": 1768
+    },
+    {
+      "epoch": 1.0786945249351838,
+      "grad_norm": 0.24006939574592479,
+      "learning_rate": 5.191513668450178e-06,
+      "loss": 0.2065,
+      "step": 1769
+    },
+    {
+      "epoch": 1.0793045600122007,
+      "grad_norm": 0.22565303602624653,
+      "learning_rate": 5.186196339976711e-06,
+      "loss": 0.2142,
+      "step": 1770
+    },
+    {
+      "epoch": 1.0799145950892177,
+      "grad_norm": 0.23431241380313964,
+      "learning_rate": 5.180878800621911e-06,
+      "loss": 0.224,
+      "step": 1771
+    },
+    {
+      "epoch": 1.0805246301662346,
+      "grad_norm": 0.2368388255119797,
+      "learning_rate": 5.175561056408291e-06,
+      "loss": 0.2057,
+      "step": 1772
+    },
+    {
+      "epoch": 1.0811346652432514,
+      "grad_norm": 0.2286738585549779,
+      "learning_rate": 5.170243113358594e-06,
+      "loss": 0.1967,
+      "step": 1773
+    },
+    {
+      "epoch": 1.0817447003202685,
+      "grad_norm": 0.23447789272378972,
+      "learning_rate": 5.164924977495792e-06,
+      "loss": 0.2247,
+      "step": 1774
+    },
+    {
+      "epoch": 1.0823547353972853,
+      "grad_norm": 0.2287799320408433,
+      "learning_rate": 5.1596066548430725e-06,
+      "loss": 0.2125,
+      "step": 1775
+    },
+    {
+      "epoch": 1.0829647704743022,
+      "grad_norm": 0.22061561139206476,
+      "learning_rate": 5.154288151423833e-06,
+      "loss": 0.1934,
+      "step": 1776
+    },
+    {
+      "epoch": 1.0835748055513192,
+      "grad_norm": 0.23075046956961687,
+      "learning_rate": 5.1489694732616805e-06,
+      "loss": 0.2154,
+      "step": 1777
+    },
+    {
+      "epoch": 1.084184840628336,
+      "grad_norm": 0.22967293413928785,
+      "learning_rate": 5.143650626380417e-06,
+      "loss": 0.2212,
+      "step": 1778
+    },
+    {
+      "epoch": 1.0847948757053532,
+      "grad_norm": 0.22575905174950292,
+      "learning_rate": 5.138331616804034e-06,
+      "loss": 0.198,
+      "step": 1779
+    },
+    {
+      "epoch": 1.08540491078237,
+      "grad_norm": 0.2362401899413697,
+      "learning_rate": 5.133012450556709e-06,
+      "loss": 0.2041,
+      "step": 1780
+    },
+    {
+      "epoch": 1.0860149458593868,
+      "grad_norm": 0.22921901774320988,
+      "learning_rate": 5.127693133662801e-06,
+      "loss": 0.2207,
+      "step": 1781
+    },
+    {
+      "epoch": 1.086624980936404,
+      "grad_norm": 0.2397430325638046,
+      "learning_rate": 5.122373672146831e-06,
+      "loss": 0.2218,
+      "step": 1782
+    },
+    {
+      "epoch": 1.0872350160134208,
+      "grad_norm": 0.2441605988712309,
+      "learning_rate": 5.117054072033492e-06,
+      "loss": 0.2252,
+      "step": 1783
+    },
+    {
+      "epoch": 1.0878450510904376,
+      "grad_norm": 0.23041025603342433,
+      "learning_rate": 5.111734339347629e-06,
+      "loss": 0.2101,
+      "step": 1784
+    },
+    {
+      "epoch": 1.0884550861674547,
+      "grad_norm": 0.210872259108622,
+      "learning_rate": 5.106414480114238e-06,
+      "loss": 0.2044,
+      "step": 1785
+    },
+    {
+      "epoch": 1.0890651212444715,
+      "grad_norm": 0.2315607312022513,
+      "learning_rate": 5.10109450035846e-06,
+      "loss": 0.2214,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0896751563214884,
+      "grad_norm": 0.2208674173967462,
+      "learning_rate": 5.095774406105572e-06,
+      "loss": 0.2031,
+      "step": 1787
+    },
+    {
+      "epoch": 1.0902851913985054,
+      "grad_norm": 0.23095765951855177,
+      "learning_rate": 5.0904542033809774e-06,
+      "loss": 0.1979,
+      "step": 1788
+    },
+    {
+      "epoch": 1.0908952264755223,
+      "grad_norm": 0.21581001155371557,
+      "learning_rate": 5.085133898210208e-06,
+      "loss": 0.2006,
+      "step": 1789
+    },
+    {
+      "epoch": 1.0915052615525394,
+      "grad_norm": 0.23434786464295787,
+      "learning_rate": 5.079813496618908e-06,
+      "loss": 0.2127,
+      "step": 1790
+    },
+    {
+      "epoch": 1.0921152966295562,
+      "grad_norm": 0.23162166178276317,
+      "learning_rate": 5.07449300463283e-06,
+      "loss": 0.2193,
+      "step": 1791
+    },
+    {
+      "epoch": 1.092725331706573,
+      "grad_norm": 0.231793141185086,
+      "learning_rate": 5.0691724282778345e-06,
+      "loss": 0.2222,
+      "step": 1792
+    },
+    {
+      "epoch": 1.0933353667835901,
+      "grad_norm": 0.2374667122949609,
+      "learning_rate": 5.06385177357987e-06,
+      "loss": 0.2114,
+      "step": 1793
+    },
+    {
+      "epoch": 1.093945401860607,
+      "grad_norm": 0.22213391536139876,
+      "learning_rate": 5.058531046564977e-06,
+      "loss": 0.1942,
+      "step": 1794
+    },
+    {
+      "epoch": 1.094555436937624,
+      "grad_norm": 0.22243485467096535,
+      "learning_rate": 5.053210253259282e-06,
+      "loss": 0.214,
+      "step": 1795
+    },
+    {
+      "epoch": 1.0951654720146409,
+      "grad_norm": 0.24391573079049042,
+      "learning_rate": 5.04788939968898e-06,
+      "loss": 0.2349,
+      "step": 1796
+    },
+    {
+      "epoch": 1.0957755070916577,
+      "grad_norm": 0.2358921418298421,
+      "learning_rate": 5.042568491880338e-06,
+      "loss": 0.2295,
+      "step": 1797
+    },
+    {
+      "epoch": 1.0963855421686748,
+      "grad_norm": 0.2389177385955543,
+      "learning_rate": 5.0372475358596825e-06,
+      "loss": 0.2196,
+      "step": 1798
+    },
+    {
+      "epoch": 1.0969955772456916,
+      "grad_norm": 0.2323273476825324,
+      "learning_rate": 5.031926537653396e-06,
+      "loss": 0.1973,
+      "step": 1799
+    },
+    {
+      "epoch": 1.0976056123227085,
+      "grad_norm": 0.23721478174918753,
+      "learning_rate": 5.026605503287911e-06,
+      "loss": 0.1953,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0982156473997255,
+      "grad_norm": 0.2390291825850533,
+      "learning_rate": 5.021284438789694e-06,
+      "loss": 0.2119,
+      "step": 1801
+    },
+    {
+      "epoch": 1.0988256824767424,
+      "grad_norm": 0.23791508453378607,
+      "learning_rate": 5.015963350185253e-06,
+      "loss": 0.2154,
+      "step": 1802
+    },
+    {
+      "epoch": 1.0994357175537592,
+      "grad_norm": 0.23144393652698764,
+      "learning_rate": 5.010642243501119e-06,
+      "loss": 0.1874,
+      "step": 1803
+    },
+    {
+      "epoch": 1.1000457526307763,
+      "grad_norm": 0.21940967709391074,
+      "learning_rate": 5.005321124763847e-06,
+      "loss": 0.2089,
+      "step": 1804
+    },
+    {
+      "epoch": 1.1006557877077932,
+      "grad_norm": 0.21810324102556533,
+      "learning_rate": 5e-06,
+      "loss": 0.2133,
+      "step": 1805
+    },
+    {
+      "epoch": 1.1012658227848102,
+      "grad_norm": 0.24532717574886334,
+      "learning_rate": 4.994678875236155e-06,
+      "loss": 0.2157,
+      "step": 1806
+    },
+    {
+      "epoch": 1.101875857861827,
+      "grad_norm": 0.24625319723639874,
+      "learning_rate": 4.989357756498882e-06,
+      "loss": 0.2002,
+      "step": 1807
+    },
+    {
+      "epoch": 1.102485892938844,
+      "grad_norm": 0.23395800432045705,
+      "learning_rate": 4.9840366498147495e-06,
+      "loss": 0.2114,
+      "step": 1808
+    },
+    {
+      "epoch": 1.103095928015861,
+      "grad_norm": 0.23251422610518932,
+      "learning_rate": 4.9787155612103076e-06,
+      "loss": 0.1838,
+      "step": 1809
+    },
+    {
+      "epoch": 1.1037059630928778,
+      "grad_norm": 0.24423256577250954,
+      "learning_rate": 4.9733944967120905e-06,
+      "loss": 0.2311,
+      "step": 1810
+    },
+    {
+      "epoch": 1.1043159981698947,
+      "grad_norm": 0.24075558744754294,
+      "learning_rate": 4.968073462346605e-06,
+      "loss": 0.2249,
+      "step": 1811
+    },
+    {
+      "epoch": 1.1049260332469117,
+      "grad_norm": 0.2243831584670897,
+      "learning_rate": 4.962752464140318e-06,
+      "loss": 0.2164,
+      "step": 1812
+    },
+    {
+      "epoch": 1.1055360683239286,
+      "grad_norm": 0.21557211649475622,
+      "learning_rate": 4.9574315081196634e-06,
+      "loss": 0.1978,
+      "step": 1813
+    },
+    {
+      "epoch": 1.1061461034009457,
+      "grad_norm": 0.2254521241476352,
+      "learning_rate": 4.952110600311021e-06,
+      "loss": 0.198,
+      "step": 1814
+    },
+    {
+      "epoch": 1.1067561384779625,
+      "grad_norm": 0.22564870909496654,
+      "learning_rate": 4.94678974674072e-06,
+      "loss": 0.2162,
+      "step": 1815
+    },
+    {
+      "epoch": 1.1073661735549793,
+      "grad_norm": 0.23231721978554046,
+      "learning_rate": 4.941468953435024e-06,
+      "loss": 0.1975,
+      "step": 1816
+    },
+    {
+      "epoch": 1.1079762086319964,
+      "grad_norm": 0.22830426660121386,
+      "learning_rate": 4.936148226420133e-06,
+      "loss": 0.1927,
+      "step": 1817
+    },
+    {
+      "epoch": 1.1085862437090133,
+      "grad_norm": 0.21468125174644645,
+      "learning_rate": 4.930827571722167e-06,
+      "loss": 0.1869,
+      "step": 1818
+    },
+    {
+      "epoch": 1.10919627878603,
+      "grad_norm": 0.23170361384957827,
+      "learning_rate": 4.92550699536717e-06,
+      "loss": 0.2112,
+      "step": 1819
+    },
+    {
+      "epoch": 1.1098063138630472,
+      "grad_norm": 0.22508120264253859,
+      "learning_rate": 4.9201865033810934e-06,
+      "loss": 0.205,
+      "step": 1820
+    },
+    {
+      "epoch": 1.110416348940064,
+      "grad_norm": 0.2277555513898493,
+      "learning_rate": 4.914866101789793e-06,
+      "loss": 0.2276,
+      "step": 1821
+    },
+    {
+      "epoch": 1.111026384017081,
+      "grad_norm": 0.21958491724982818,
+      "learning_rate": 4.909545796619023e-06,
+      "loss": 0.2094,
+      "step": 1822
+    },
+    {
+      "epoch": 1.111636419094098,
+      "grad_norm": 0.2281758250224125,
+      "learning_rate": 4.90422559389443e-06,
+      "loss": 0.2064,
+      "step": 1823
+    },
+    {
+      "epoch": 1.1122464541711148,
+      "grad_norm": 0.22903591968619846,
+      "learning_rate": 4.8989054996415414e-06,
+      "loss": 0.1996,
+      "step": 1824
+    },
+    {
+      "epoch": 1.1128564892481319,
+      "grad_norm": 0.24375202304106894,
+      "learning_rate": 4.893585519885764e-06,
+      "loss": 0.2167,
+      "step": 1825
+    },
+    {
+      "epoch": 1.1134665243251487,
+      "grad_norm": 0.2414970821532803,
+      "learning_rate": 4.8882656606523736e-06,
+      "loss": 0.2258,
+      "step": 1826
+    },
+    {
+      "epoch": 1.1140765594021655,
+      "grad_norm": 0.25006632120661687,
+      "learning_rate": 4.88294592796651e-06,
+      "loss": 0.229,
+      "step": 1827
+    },
+    {
+      "epoch": 1.1146865944791826,
+      "grad_norm": 0.2482403602836149,
+      "learning_rate": 4.8776263278531705e-06,
+      "loss": 0.2231,
+      "step": 1828
+    },
+    {
+      "epoch": 1.1152966295561995,
+      "grad_norm": 0.22840236275307774,
+      "learning_rate": 4.8723068663372005e-06,
+      "loss": 0.2209,
+      "step": 1829
+    },
+    {
+      "epoch": 1.1159066646332163,
+      "grad_norm": 0.23156238236177693,
+      "learning_rate": 4.866987549443291e-06,
+      "loss": 0.2178,
+      "step": 1830
+    },
+    {
+      "epoch": 1.1165166997102334,
+      "grad_norm": 0.2479495313346978,
+      "learning_rate": 4.861668383195968e-06,
+      "loss": 0.2329,
+      "step": 1831
+    },
+    {
+      "epoch": 1.1171267347872502,
+      "grad_norm": 0.23715382066159454,
+      "learning_rate": 4.856349373619586e-06,
+      "loss": 0.1867,
+      "step": 1832
+    },
+    {
+      "epoch": 1.1177367698642673,
+      "grad_norm": 0.2175436696513431,
+      "learning_rate": 4.851030526738321e-06,
+      "loss": 0.201,
+      "step": 1833
+    },
+    {
+      "epoch": 1.1183468049412841,
+      "grad_norm": 0.22644845105026176,
+      "learning_rate": 4.8457118485761686e-06,
+      "loss": 0.1985,
+      "step": 1834
+    },
+    {
+      "epoch": 1.118956840018301,
+      "grad_norm": 0.23796566819781487,
+      "learning_rate": 4.84039334515693e-06,
+      "loss": 0.2155,
+      "step": 1835
+    },
+    {
+      "epoch": 1.119566875095318,
+      "grad_norm": 0.20778757880164975,
+      "learning_rate": 4.835075022504211e-06,
+      "loss": 0.1967,
+      "step": 1836
+    },
+    {
+      "epoch": 1.120176910172335,
+      "grad_norm": 0.2261588006945494,
+      "learning_rate": 4.829756886641408e-06,
+      "loss": 0.2077,
+      "step": 1837
+    },
+    {
+      "epoch": 1.1207869452493517,
+      "grad_norm": 0.23049171062789117,
+      "learning_rate": 4.82443894359171e-06,
+      "loss": 0.2111,
+      "step": 1838
+    },
+    {
+      "epoch": 1.1213969803263688,
+      "grad_norm": 0.2372692688283098,
+      "learning_rate": 4.819121199378091e-06,
+      "loss": 0.2097,
+      "step": 1839
+    },
+    {
+      "epoch": 1.1220070154033857,
+      "grad_norm": 0.2137083466863717,
+      "learning_rate": 4.81380366002329e-06,
+      "loss": 0.1939,
+      "step": 1840
+    },
+    {
+      "epoch": 1.1226170504804027,
+      "grad_norm": 0.23852721077040967,
+      "learning_rate": 4.808486331549824e-06,
+      "loss": 0.2352,
+      "step": 1841
+    },
+    {
+      "epoch": 1.1232270855574196,
+      "grad_norm": 0.22148119954075798,
+      "learning_rate": 4.803169219979963e-06,
+      "loss": 0.2183,
+      "step": 1842
+    },
+    {
+      "epoch": 1.1238371206344364,
+      "grad_norm": 0.22592558525037146,
+      "learning_rate": 4.797852331335739e-06,
+      "loss": 0.1963,
+      "step": 1843
+    },
+    {
+      "epoch": 1.1244471557114535,
+      "grad_norm": 0.2353605626350167,
+      "learning_rate": 4.7925356716389235e-06,
+      "loss": 0.2207,
+      "step": 1844
+    },
+    {
+      "epoch": 1.1250571907884703,
+      "grad_norm": 0.2297212807928882,
+      "learning_rate": 4.787219246911034e-06,
+      "loss": 0.2177,
+      "step": 1845
+    },
+    {
+      "epoch": 1.1256672258654872,
+      "grad_norm": 0.2176316617932621,
+      "learning_rate": 4.781903063173321e-06,
+      "loss": 0.1829,
+      "step": 1846
+    },
+    {
+      "epoch": 1.1262772609425042,
+      "grad_norm": 0.22958634224819682,
+      "learning_rate": 4.776587126446761e-06,
+      "loss": 0.2042,
+      "step": 1847
+    },
+    {
+      "epoch": 1.126887296019521,
+      "grad_norm": 0.23047338285747526,
+      "learning_rate": 4.771271442752053e-06,
+      "loss": 0.2153,
+      "step": 1848
+    },
+    {
+      "epoch": 1.1274973310965382,
+      "grad_norm": 0.24143457652149827,
+      "learning_rate": 4.765956018109607e-06,
+      "loss": 0.2189,
+      "step": 1849
+    },
+    {
+      "epoch": 1.128107366173555,
+      "grad_norm": 0.21811634648251788,
+      "learning_rate": 4.760640858539541e-06,
+      "loss": 0.2038,
+      "step": 1850
+    },
+    {
+      "epoch": 1.1287174012505719,
+      "grad_norm": 0.22559255230976394,
+      "learning_rate": 4.755325970061671e-06,
+      "loss": 0.2027,
+      "step": 1851
+    },
+    {
+      "epoch": 1.129327436327589,
+      "grad_norm": 0.22981316462642773,
+      "learning_rate": 4.750011358695508e-06,
+      "loss": 0.2089,
+      "step": 1852
+    },
+    {
+      "epoch": 1.1299374714046058,
+      "grad_norm": 0.22879068330224503,
+      "learning_rate": 4.744697030460248e-06,
+      "loss": 0.2044,
+      "step": 1853
+    },
+    {
+      "epoch": 1.1305475064816226,
+      "grad_norm": 0.2072099344556385,
+      "learning_rate": 4.739382991374767e-06,
+      "loss": 0.1959,
+      "step": 1854
+    },
+    {
+      "epoch": 1.1311575415586397,
+      "grad_norm": 0.23776541076645313,
+      "learning_rate": 4.734069247457613e-06,
+      "loss": 0.1942,
+      "step": 1855
+    },
+    {
+      "epoch": 1.1317675766356565,
+      "grad_norm": 0.23016709695257465,
+      "learning_rate": 4.728755804726998e-06,
+      "loss": 0.2249,
+      "step": 1856
+    },
+    {
+      "epoch": 1.1323776117126734,
+      "grad_norm": 0.22533949944245424,
+      "learning_rate": 4.7234426692007985e-06,
+      "loss": 0.2008,
+      "step": 1857
+    },
+    {
+      "epoch": 1.1329876467896904,
+      "grad_norm": 0.21749457157915061,
+      "learning_rate": 4.718129846896538e-06,
+      "loss": 0.1991,
+      "step": 1858
+    },
+    {
+      "epoch": 1.1335976818667073,
+      "grad_norm": 0.22038974282764345,
+      "learning_rate": 4.712817343831384e-06,
+      "loss": 0.1961,
+      "step": 1859
+    },
+    {
+      "epoch": 1.1342077169437244,
+      "grad_norm": 0.2386468105906248,
+      "learning_rate": 4.707505166022149e-06,
+      "loss": 0.2112,
+      "step": 1860
+    },
+    {
+      "epoch": 1.1348177520207412,
+      "grad_norm": 0.24323963813817923,
+      "learning_rate": 4.702193319485271e-06,
+      "loss": 0.2062,
+      "step": 1861
+    },
+    {
+      "epoch": 1.135427787097758,
+      "grad_norm": 0.24281596413325793,
+      "learning_rate": 4.696881810236815e-06,
+      "loss": 0.2237,
+      "step": 1862
+    },
+    {
+      "epoch": 1.1360378221747751,
+      "grad_norm": 0.24077278804132538,
+      "learning_rate": 4.691570644292464e-06,
+      "loss": 0.2183,
+      "step": 1863
+    },
+    {
+      "epoch": 1.136647857251792,
+      "grad_norm": 0.22606116032964318,
+      "learning_rate": 4.686259827667514e-06,
+      "loss": 0.2098,
+      "step": 1864
+    },
+    {
+      "epoch": 1.137257892328809,
+      "grad_norm": 0.22888289623334254,
+      "learning_rate": 4.680949366376858e-06,
+      "loss": 0.2041,
+      "step": 1865
+    },
+    {
+      "epoch": 1.1378679274058259,
+      "grad_norm": 0.23305603580105133,
+      "learning_rate": 4.6756392664349985e-06,
+      "loss": 0.2023,
+      "step": 1866
+    },
+    {
+      "epoch": 1.1384779624828427,
+      "grad_norm": 0.24371372538251743,
+      "learning_rate": 4.67032953385602e-06,
+      "loss": 0.2082,
+      "step": 1867
+    },
+    {
+      "epoch": 1.1390879975598598,
+      "grad_norm": 0.22396952559232805,
+      "learning_rate": 4.6650201746535926e-06,
+      "loss": 0.2068,
+      "step": 1868
+    },
+    {
+      "epoch": 1.1396980326368766,
+      "grad_norm": 0.22634464794146542,
+      "learning_rate": 4.659711194840964e-06,
+      "loss": 0.2028,
+      "step": 1869
+    },
+    {
+      "epoch": 1.1403080677138935,
+      "grad_norm": 0.2240492344171807,
+      "learning_rate": 4.654402600430955e-06,
+      "loss": 0.2071,
+      "step": 1870
+    },
+    {
+      "epoch": 1.1409181027909105,
+      "grad_norm": 0.23557956266916885,
+      "learning_rate": 4.649094397435944e-06,
+      "loss": 0.2007,
+      "step": 1871
+    },
+    {
+      "epoch": 1.1415281378679274,
+      "grad_norm": 0.22438100866093574,
+      "learning_rate": 4.643786591867871e-06,
+      "loss": 0.209,
+      "step": 1872
+    },
+    {
+      "epoch": 1.1421381729449442,
+      "grad_norm": 0.24906239130894264,
+      "learning_rate": 4.638479189738224e-06,
+      "loss": 0.2175,
+      "step": 1873
+    },
+    {
+      "epoch": 1.1427482080219613,
+      "grad_norm": 0.24332170113088764,
+      "learning_rate": 4.633172197058034e-06,
+      "loss": 0.2238,
+      "step": 1874
+    },
+    {
+      "epoch": 1.1433582430989782,
+      "grad_norm": 0.24344778268910833,
+      "learning_rate": 4.6278656198378665e-06,
+      "loss": 0.2241,
+      "step": 1875
+    },
+    {
+      "epoch": 1.1439682781759952,
+      "grad_norm": 0.2383508238526649,
+      "learning_rate": 4.622559464087824e-06,
+      "loss": 0.2297,
+      "step": 1876
+    },
+    {
+      "epoch": 1.144578313253012,
+      "grad_norm": 0.22412248975169005,
+      "learning_rate": 4.617253735817522e-06,
+      "loss": 0.2078,
+      "step": 1877
+    },
+    {
+      "epoch": 1.145188348330029,
+      "grad_norm": 0.23701589723417668,
+      "learning_rate": 4.611948441036098e-06,
+      "loss": 0.2321,
+      "step": 1878
+    },
+    {
+      "epoch": 1.145798383407046,
+      "grad_norm": 0.2340242843088781,
+      "learning_rate": 4.606643585752195e-06,
+      "loss": 0.2103,
+      "step": 1879
+    },
+    {
+      "epoch": 1.1464084184840628,
+      "grad_norm": 0.23920144889311484,
+      "learning_rate": 4.6013391759739615e-06,
+      "loss": 0.2456,
+      "step": 1880
+    },
+    {
+      "epoch": 1.1470184535610797,
+      "grad_norm": 0.21459534836538638,
+      "learning_rate": 4.596035217709039e-06,
+      "loss": 0.1895,
+      "step": 1881
+    },
+    {
+      "epoch": 1.1476284886380967,
+      "grad_norm": 0.22716162499913228,
+      "learning_rate": 4.590731716964559e-06,
+      "loss": 0.2271,
+      "step": 1882
+    },
+    {
+      "epoch": 1.1482385237151136,
+      "grad_norm": 0.244394988171428,
+      "learning_rate": 4.585428679747133e-06,
+      "loss": 0.2111,
+      "step": 1883
+    },
+    {
+      "epoch": 1.1488485587921304,
+      "grad_norm": 0.21851983087595445,
+      "learning_rate": 4.580126112062847e-06,
+      "loss": 0.192,
+      "step": 1884
+    },
+    {
+      "epoch": 1.1494585938691475,
+      "grad_norm": 0.23185076923355083,
+      "learning_rate": 4.574824019917262e-06,
+      "loss": 0.2131,
+      "step": 1885
+    },
+    {
+      "epoch": 1.1500686289461644,
+      "grad_norm": 0.22895978103203585,
+      "learning_rate": 4.569522409315392e-06,
+      "loss": 0.2078,
+      "step": 1886
+    },
+    {
+      "epoch": 1.1506786640231814,
+      "grad_norm": 0.23953049315851418,
+      "learning_rate": 4.564221286261709e-06,
+      "loss": 0.2145,
+      "step": 1887
+    },
+    {
+      "epoch": 1.1512886991001983,
+      "grad_norm": 0.22458440085000345,
+      "learning_rate": 4.558920656760135e-06,
+      "loss": 0.2118,
+      "step": 1888
+    },
+    {
+      "epoch": 1.1518987341772151,
+      "grad_norm": 0.22196229519951427,
+      "learning_rate": 4.553620526814029e-06,
+      "loss": 0.2158,
+      "step": 1889
+    },
+    {
+      "epoch": 1.1525087692542322,
+      "grad_norm": 0.2926714766657743,
+      "learning_rate": 4.548320902426186e-06,
+      "loss": 0.2081,
+      "step": 1890
+    },
+    {
+      "epoch": 1.153118804331249,
+      "grad_norm": 0.2637006579000516,
+      "learning_rate": 4.543021789598831e-06,
+      "loss": 0.2046,
+      "step": 1891
+    },
+    {
+      "epoch": 1.153728839408266,
+      "grad_norm": 0.2303464141889435,
+      "learning_rate": 4.537723194333603e-06,
+      "loss": 0.2096,
+      "step": 1892
+    },
+    {
+      "epoch": 1.154338874485283,
+      "grad_norm": 0.22616168232534192,
+      "learning_rate": 4.532425122631559e-06,
+      "loss": 0.2195,
+      "step": 1893
+    },
+    {
+      "epoch": 1.1549489095622998,
+      "grad_norm": 0.22958441424908974,
+      "learning_rate": 4.527127580493167e-06,
+      "loss": 0.2152,
+      "step": 1894
+    },
+    {
+      "epoch": 1.1555589446393169,
+      "grad_norm": 0.23206621665711868,
+      "learning_rate": 4.521830573918289e-06,
+      "loss": 0.2056,
+      "step": 1895
+    },
+    {
+      "epoch": 1.1561689797163337,
+      "grad_norm": 0.24162175497638366,
+      "learning_rate": 4.5165341089061825e-06,
+      "loss": 0.2112,
+      "step": 1896
+    },
+    {
+      "epoch": 1.1567790147933505,
+      "grad_norm": 0.2236351442320407,
+      "learning_rate": 4.511238191455491e-06,
+      "loss": 0.1935,
+      "step": 1897
+    },
+    {
+      "epoch": 1.1573890498703676,
+      "grad_norm": 0.2454020762399105,
+      "learning_rate": 4.505942827564242e-06,
+      "loss": 0.2265,
+      "step": 1898
+    },
+    {
+      "epoch": 1.1579990849473845,
+      "grad_norm": 0.26020678242989564,
+      "learning_rate": 4.500648023229828e-06,
+      "loss": 0.211,
+      "step": 1899
+    },
+    {
+      "epoch": 1.1586091200244013,
+      "grad_norm": 0.22806777598404518,
+      "learning_rate": 4.495353784449015e-06,
+      "loss": 0.1971,
+      "step": 1900
+    },
+    {
+      "epoch": 1.1592191551014184,
+      "grad_norm": 0.22375142681862864,
+      "learning_rate": 4.490060117217925e-06,
+      "loss": 0.2164,
+      "step": 1901
+    },
+    {
+      "epoch": 1.1598291901784352,
+      "grad_norm": 0.23047793015134796,
+      "learning_rate": 4.484767027532032e-06,
+      "loss": 0.2272,
+      "step": 1902
+    },
+    {
+      "epoch": 1.1604392252554523,
+      "grad_norm": 0.22402678078073376,
+      "learning_rate": 4.479474521386161e-06,
+      "loss": 0.2024,
+      "step": 1903
+    },
+    {
+      "epoch": 1.1610492603324691,
+      "grad_norm": 0.24312729205941291,
+      "learning_rate": 4.474182604774471e-06,
+      "loss": 0.2037,
+      "step": 1904
+    },
+    {
+      "epoch": 1.161659295409486,
+      "grad_norm": 0.22900584787625933,
+      "learning_rate": 4.468891283690454e-06,
+      "loss": 0.2062,
+      "step": 1905
+    },
+    {
+      "epoch": 1.162269330486503,
+      "grad_norm": 0.23554230639630583,
+      "learning_rate": 4.4636005641269294e-06,
+      "loss": 0.2238,
+      "step": 1906
+    },
+    {
+      "epoch": 1.16287936556352,
+      "grad_norm": 0.24073626227280503,
+      "learning_rate": 4.458310452076034e-06,
+      "loss": 0.2442,
+      "step": 1907
+    },
+    {
+      "epoch": 1.1634894006405367,
+      "grad_norm": 0.2356536112560819,
+      "learning_rate": 4.453020953529217e-06,
+      "loss": 0.1972,
+      "step": 1908
+    },
+    {
+      "epoch": 1.1640994357175538,
+      "grad_norm": 0.24903944162427105,
+      "learning_rate": 4.447732074477233e-06,
+      "loss": 0.2303,
+      "step": 1909
+    },
+    {
+      "epoch": 1.1647094707945707,
+      "grad_norm": 0.24095256035225399,
+      "learning_rate": 4.442443820910133e-06,
+      "loss": 0.2214,
+      "step": 1910
+    },
+    {
+      "epoch": 1.1653195058715875,
+      "grad_norm": 0.2390760009355516,
+      "learning_rate": 4.437156198817262e-06,
+      "loss": 0.2072,
+      "step": 1911
+    },
+    {
+      "epoch": 1.1659295409486046,
+      "grad_norm": 0.24010310251085357,
+      "learning_rate": 4.431869214187246e-06,
+      "loss": 0.2162,
+      "step": 1912
+    },
+    {
+      "epoch": 1.1665395760256214,
+      "grad_norm": 0.2300662707483189,
+      "learning_rate": 4.426582873007999e-06,
+      "loss": 0.2002,
+      "step": 1913
+    },
+    {
+      "epoch": 1.1671496111026385,
+      "grad_norm": 0.22383983720254905,
+      "learning_rate": 4.421297181266694e-06,
+      "loss": 0.1819,
+      "step": 1914
+    },
+    {
+      "epoch": 1.1677596461796553,
+      "grad_norm": 0.2414352617015131,
+      "learning_rate": 4.4160121449497735e-06,
+      "loss": 0.2185,
+      "step": 1915
+    },
+    {
+      "epoch": 1.1683696812566722,
+      "grad_norm": 0.2539875282741942,
+      "learning_rate": 4.410727770042941e-06,
+      "loss": 0.2226,
+      "step": 1916
+    },
+    {
+      "epoch": 1.1689797163336892,
+      "grad_norm": 0.24420213569141258,
+      "learning_rate": 4.405444062531145e-06,
+      "loss": 0.2001,
+      "step": 1917
+    },
+    {
+      "epoch": 1.169589751410706,
+      "grad_norm": 0.23888115158233167,
+      "learning_rate": 4.400161028398583e-06,
+      "loss": 0.2013,
+      "step": 1918
+    },
+    {
+      "epoch": 1.1701997864877232,
+      "grad_norm": 0.24835651886999338,
+      "learning_rate": 4.3948786736286866e-06,
+      "loss": 0.199,
+      "step": 1919
+    },
+    {
+      "epoch": 1.17080982156474,
+      "grad_norm": 0.24692728404753747,
+      "learning_rate": 4.389597004204119e-06,
+      "loss": 0.2233,
+      "step": 1920
+    },
+    {
+      "epoch": 1.1714198566417569,
+      "grad_norm": 0.23560264343648826,
+      "learning_rate": 4.384316026106766e-06,
+      "loss": 0.2128,
+      "step": 1921
+    },
+    {
+      "epoch": 1.172029891718774,
+      "grad_norm": 0.2265026408766131,
+      "learning_rate": 4.379035745317734e-06,
+      "loss": 0.1931,
+      "step": 1922
+    },
+    {
+      "epoch": 1.1726399267957908,
+      "grad_norm": 0.24621590868324938,
+      "learning_rate": 4.373756167817338e-06,
+      "loss": 0.2067,
+      "step": 1923
+    },
+    {
+      "epoch": 1.1732499618728076,
+      "grad_norm": 0.24098881344443027,
+      "learning_rate": 4.368477299585094e-06,
+      "loss": 0.2099,
+      "step": 1924
+    },
+    {
+      "epoch": 1.1738599969498247,
+      "grad_norm": 0.2268182569936142,
+      "learning_rate": 4.363199146599717e-06,
+      "loss": 0.1841,
+      "step": 1925
+    },
+    {
+      "epoch": 1.1744700320268415,
+      "grad_norm": 0.23457894427475676,
+      "learning_rate": 4.3579217148391115e-06,
+      "loss": 0.2074,
+      "step": 1926
+    },
+    {
+      "epoch": 1.1750800671038584,
+      "grad_norm": 0.2538176166214765,
+      "learning_rate": 4.3526450102803654e-06,
+      "loss": 0.2043,
+      "step": 1927
+    },
+    {
+      "epoch": 1.1756901021808754,
+      "grad_norm": 0.298440743554529,
+      "learning_rate": 4.347369038899744e-06,
+      "loss": 0.2217,
+      "step": 1928
+    },
+    {
+      "epoch": 1.1763001372578923,
+      "grad_norm": 0.22397262376763313,
+      "learning_rate": 4.342093806672678e-06,
+      "loss": 0.2075,
+      "step": 1929
+    },
+    {
+      "epoch": 1.1769101723349094,
+      "grad_norm": 0.2497064497376091,
+      "learning_rate": 4.336819319573764e-06,
+      "loss": 0.2336,
+      "step": 1930
+    },
+    {
+      "epoch": 1.1775202074119262,
+      "grad_norm": 0.2171700331359755,
+      "learning_rate": 4.331545583576758e-06,
+      "loss": 0.1868,
+      "step": 1931
+    },
+    {
+      "epoch": 1.178130242488943,
+      "grad_norm": 0.2381864903430019,
+      "learning_rate": 4.32627260465456e-06,
+      "loss": 0.1959,
+      "step": 1932
+    },
+    {
+      "epoch": 1.1787402775659601,
+      "grad_norm": 0.2420621944986559,
+      "learning_rate": 4.321000388779214e-06,
+      "loss": 0.1993,
+      "step": 1933
+    },
+    {
+      "epoch": 1.179350312642977,
+      "grad_norm": 0.23433046760619705,
+      "learning_rate": 4.3157289419219e-06,
+      "loss": 0.2066,
+      "step": 1934
+    },
+    {
+      "epoch": 1.1799603477199938,
+      "grad_norm": 0.24468272597979573,
+      "learning_rate": 4.3104582700529295e-06,
+      "loss": 0.2074,
+      "step": 1935
+    },
+    {
+      "epoch": 1.1805703827970109,
+      "grad_norm": 0.23079154266496227,
+      "learning_rate": 4.3051883791417325e-06,
+      "loss": 0.2132,
+      "step": 1936
+    },
+    {
+      "epoch": 1.1811804178740277,
+      "grad_norm": 0.22032298652505528,
+      "learning_rate": 4.299919275156857e-06,
+      "loss": 0.2085,
+      "step": 1937
+    },
+    {
+      "epoch": 1.1817904529510446,
+      "grad_norm": 0.23205968573496338,
+      "learning_rate": 4.294650964065956e-06,
+      "loss": 0.2225,
+      "step": 1938
+    },
+    {
+      "epoch": 1.1824004880280616,
+      "grad_norm": 0.24676759799484482,
+      "learning_rate": 4.289383451835789e-06,
+      "loss": 0.2192,
+      "step": 1939
+    },
+    {
+      "epoch": 1.1830105231050785,
+      "grad_norm": 0.22110343486421205,
+      "learning_rate": 4.28411674443221e-06,
+      "loss": 0.2005,
+      "step": 1940
+    },
+    {
+      "epoch": 1.1836205581820956,
+      "grad_norm": 0.23647728822340291,
+      "learning_rate": 4.278850847820161e-06,
+      "loss": 0.2132,
+      "step": 1941
+    },
+    {
+      "epoch": 1.1842305932591124,
+      "grad_norm": 0.23142771149859084,
+      "learning_rate": 4.273585767963662e-06,
+      "loss": 0.2076,
+      "step": 1942
+    },
+    {
+      "epoch": 1.1848406283361292,
+      "grad_norm": 0.22278997597375186,
+      "learning_rate": 4.2683215108258145e-06,
+      "loss": 0.1969,
+      "step": 1943
+    },
+    {
+      "epoch": 1.1854506634131463,
+      "grad_norm": 0.2264849055331267,
+      "learning_rate": 4.263058082368785e-06,
+      "loss": 0.2093,
+      "step": 1944
+    },
+    {
+      "epoch": 1.1860606984901632,
+      "grad_norm": 0.23289169571659174,
+      "learning_rate": 4.2577954885537985e-06,
+      "loss": 0.2299,
+      "step": 1945
+    },
+    {
+      "epoch": 1.1866707335671802,
+      "grad_norm": 0.2319473383776966,
+      "learning_rate": 4.25253373534114e-06,
+      "loss": 0.2295,
+      "step": 1946
+    },
+    {
+      "epoch": 1.187280768644197,
+      "grad_norm": 0.24206179694564797,
+      "learning_rate": 4.247272828690138e-06,
+      "loss": 0.1919,
+      "step": 1947
+    },
+    {
+      "epoch": 1.187890803721214,
+      "grad_norm": 0.22777022144636788,
+      "learning_rate": 4.242012774559164e-06,
+      "loss": 0.1958,
+      "step": 1948
+    },
+    {
+      "epoch": 1.188500838798231,
+      "grad_norm": 0.24028328921186182,
+      "learning_rate": 4.236753578905627e-06,
+      "loss": 0.2146,
+      "step": 1949
+    },
+    {
+      "epoch": 1.1891108738752478,
+      "grad_norm": 0.23051792511004326,
+      "learning_rate": 4.231495247685961e-06,
+      "loss": 0.2044,
+      "step": 1950
+    },
+    {
+      "epoch": 1.1897209089522647,
+      "grad_norm": 0.22677821945909582,
+      "learning_rate": 4.2262377868556176e-06,
+      "loss": 0.196,
+      "step": 1951
+    },
+    {
+      "epoch": 1.1903309440292817,
+      "grad_norm": 0.23956247319151058,
+      "learning_rate": 4.220981202369067e-06,
+      "loss": 0.2218,
+      "step": 1952
+    },
+    {
+      "epoch": 1.1909409791062986,
+      "grad_norm": 0.23056731679274411,
+      "learning_rate": 4.215725500179788e-06,
+      "loss": 0.1964,
+      "step": 1953
+    },
+    {
+      "epoch": 1.1915510141833154,
+      "grad_norm": 0.22290865769000323,
+      "learning_rate": 4.210470686240255e-06,
+      "loss": 0.2004,
+      "step": 1954
+    },
+    {
+      "epoch": 1.1921610492603325,
+      "grad_norm": 0.2370506789467733,
+      "learning_rate": 4.205216766501941e-06,
+      "loss": 0.2192,
+      "step": 1955
+    },
+    {
+      "epoch": 1.1927710843373494,
+      "grad_norm": 0.24199429030566116,
+      "learning_rate": 4.199963746915304e-06,
+      "loss": 0.2111,
+      "step": 1956
+    },
+    {
+      "epoch": 1.1933811194143664,
+      "grad_norm": 0.24928723046464263,
+      "learning_rate": 4.194711633429782e-06,
+      "loss": 0.2232,
+      "step": 1957
+    },
+    {
+      "epoch": 1.1939911544913833,
+      "grad_norm": 0.2261523161741276,
+      "learning_rate": 4.189460431993788e-06,
+      "loss": 0.2226,
+      "step": 1958
+    },
+    {
+      "epoch": 1.1946011895684001,
+      "grad_norm": 0.21711436235680392,
+      "learning_rate": 4.184210148554704e-06,
+      "loss": 0.2173,
+      "step": 1959
+    },
+    {
+      "epoch": 1.1952112246454172,
+      "grad_norm": 0.2335041594132127,
+      "learning_rate": 4.178960789058869e-06,
+      "loss": 0.2259,
+      "step": 1960
+    },
+    {
+      "epoch": 1.195821259722434,
+      "grad_norm": 0.22420508190499794,
+      "learning_rate": 4.173712359451576e-06,
+      "loss": 0.2126,
+      "step": 1961
+    },
+    {
+      "epoch": 1.1964312947994509,
+      "grad_norm": 0.21841033430406878,
+      "learning_rate": 4.1684648656770655e-06,
+      "loss": 0.2166,
+      "step": 1962
+    },
+    {
+      "epoch": 1.197041329876468,
+      "grad_norm": 0.21237709660456103,
+      "learning_rate": 4.16321831367852e-06,
+      "loss": 0.1958,
+      "step": 1963
+    },
+    {
+      "epoch": 1.1976513649534848,
+      "grad_norm": 0.2427307708445763,
+      "learning_rate": 4.157972709398051e-06,
+      "loss": 0.2288,
+      "step": 1964
+    },
+    {
+      "epoch": 1.1982614000305016,
+      "grad_norm": 0.21122796020184387,
+      "learning_rate": 4.152728058776701e-06,
+      "loss": 0.2015,
+      "step": 1965
+    },
+    {
+      "epoch": 1.1988714351075187,
+      "grad_norm": 0.23690829355786672,
+      "learning_rate": 4.14748436775443e-06,
+      "loss": 0.2272,
+      "step": 1966
+    },
+    {
+      "epoch": 1.1994814701845355,
+      "grad_norm": 0.22601930005669627,
+      "learning_rate": 4.142241642270109e-06,
+      "loss": 0.2161,
+      "step": 1967
+    },
+    {
+      "epoch": 1.2000915052615526,
+      "grad_norm": 0.21905788470891707,
+      "learning_rate": 4.136999888261522e-06,
+      "loss": 0.207,
+      "step": 1968
+    },
+    {
+      "epoch": 1.2007015403385695,
+      "grad_norm": 0.23098794767319236,
+      "learning_rate": 4.131759111665349e-06,
+      "loss": 0.2265,
+      "step": 1969
+    },
+    {
+      "epoch": 1.2013115754155863,
+      "grad_norm": 0.24001379202195916,
+      "learning_rate": 4.1265193184171605e-06,
+      "loss": 0.2123,
+      "step": 1970
+    },
+    {
+      "epoch": 1.2019216104926034,
+      "grad_norm": 0.22602298697793693,
+      "learning_rate": 4.121280514451417e-06,
+      "loss": 0.2112,
+      "step": 1971
+    },
+    {
+      "epoch": 1.2025316455696202,
+      "grad_norm": 0.22965358583286585,
+      "learning_rate": 4.116042705701457e-06,
+      "loss": 0.2158,
+      "step": 1972
+    },
+    {
+      "epoch": 1.2031416806466373,
+      "grad_norm": 0.22836624548685788,
+      "learning_rate": 4.110805898099492e-06,
+      "loss": 0.207,
+      "step": 1973
+    },
+    {
+      "epoch": 1.2037517157236541,
+      "grad_norm": 0.20894006566169948,
+      "learning_rate": 4.105570097576601e-06,
+      "loss": 0.1953,
+      "step": 1974
+    },
+    {
+      "epoch": 1.204361750800671,
+      "grad_norm": 0.2333801850349286,
+      "learning_rate": 4.100335310062719e-06,
+      "loss": 0.2217,
+      "step": 1975
+    },
+    {
+      "epoch": 1.204971785877688,
+      "grad_norm": 0.2314869312834669,
+      "learning_rate": 4.095101541486636e-06,
+      "loss": 0.2302,
+      "step": 1976
+    },
+    {
+      "epoch": 1.205581820954705,
+      "grad_norm": 0.2178532501103331,
+      "learning_rate": 4.0898687977759895e-06,
+      "loss": 0.1874,
+      "step": 1977
+    },
+    {
+      "epoch": 1.2061918560317217,
+      "grad_norm": 0.22091585188727192,
+      "learning_rate": 4.084637084857254e-06,
+      "loss": 0.2074,
+      "step": 1978
+    },
+    {
+      "epoch": 1.2068018911087388,
+      "grad_norm": 0.21637361310776265,
+      "learning_rate": 4.079406408655737e-06,
+      "loss": 0.1906,
+      "step": 1979
+    },
+    {
+      "epoch": 1.2074119261857557,
+      "grad_norm": 0.22093702909221938,
+      "learning_rate": 4.0741767750955724e-06,
+      "loss": 0.2205,
+      "step": 1980
+    },
+    {
+      "epoch": 1.2080219612627725,
+      "grad_norm": 0.2212934988688623,
+      "learning_rate": 4.068948190099711e-06,
+      "loss": 0.2322,
+      "step": 1981
+    },
+    {
+      "epoch": 1.2086319963397896,
+      "grad_norm": 0.2365991900597681,
+      "learning_rate": 4.0637206595899206e-06,
+      "loss": 0.2376,
+      "step": 1982
+    },
+    {
+      "epoch": 1.2092420314168064,
+      "grad_norm": 0.23556079186727855,
+      "learning_rate": 4.058494189486769e-06,
+      "loss": 0.198,
+      "step": 1983
+    },
+    {
+      "epoch": 1.2098520664938235,
+      "grad_norm": 0.21780322654855477,
+      "learning_rate": 4.0532687857096285e-06,
+      "loss": 0.2245,
+      "step": 1984
+    },
+    {
+      "epoch": 1.2104621015708403,
+      "grad_norm": 0.2287696706059261,
+      "learning_rate": 4.048044454176658e-06,
+      "loss": 0.209,
+      "step": 1985
+    },
+    {
+      "epoch": 1.2110721366478572,
+      "grad_norm": 0.2610764735198003,
+      "learning_rate": 4.042821200804809e-06,
+      "loss": 0.2427,
+      "step": 1986
+    },
+    {
+      "epoch": 1.2116821717248742,
+      "grad_norm": 0.22708291802359176,
+      "learning_rate": 4.037599031509806e-06,
+      "loss": 0.2005,
+      "step": 1987
+    },
+    {
+      "epoch": 1.212292206801891,
+      "grad_norm": 0.2415172990342089,
+      "learning_rate": 4.032377952206148e-06,
+      "loss": 0.2256,
+      "step": 1988
+    },
+    {
+      "epoch": 1.2129022418789082,
+      "grad_norm": 0.24699658239416358,
+      "learning_rate": 4.0271579688071e-06,
+      "loss": 0.2141,
+      "step": 1989
+    },
+    {
+      "epoch": 1.213512276955925,
+      "grad_norm": 0.2399879409915332,
+      "learning_rate": 4.021939087224682e-06,
+      "loss": 0.2115,
+      "step": 1990
+    },
+    {
+      "epoch": 1.2141223120329419,
+      "grad_norm": 0.23415516645716586,
+      "learning_rate": 4.016721313369674e-06,
+      "loss": 0.2271,
+      "step": 1991
+    },
+    {
+      "epoch": 1.2147323471099587,
+      "grad_norm": 0.22806628490981498,
+      "learning_rate": 4.011504653151593e-06,
+      "loss": 0.2204,
+      "step": 1992
+    },
+    {
+      "epoch": 1.2153423821869758,
+      "grad_norm": 0.2595253670667039,
+      "learning_rate": 4.0062891124787e-06,
+      "loss": 0.2427,
+      "step": 1993
+    },
+    {
+      "epoch": 1.2159524172639926,
+      "grad_norm": 0.226343672771304,
+      "learning_rate": 4.0010746972579865e-06,
+      "loss": 0.2081,
+      "step": 1994
+    },
+    {
+      "epoch": 1.2165624523410097,
+      "grad_norm": 0.22187314948561185,
+      "learning_rate": 3.995861413395164e-06,
+      "loss": 0.2095,
+      "step": 1995
+    },
+    {
+      "epoch": 1.2171724874180265,
+      "grad_norm": 0.2408609051467771,
+      "learning_rate": 3.990649266794676e-06,
+      "loss": 0.2048,
+      "step": 1996
+    },
+    {
+      "epoch": 1.2177825224950434,
+      "grad_norm": 0.23466609594422677,
+      "learning_rate": 3.985438263359667e-06,
+      "loss": 0.2155,
+      "step": 1997
+    },
+    {
+      "epoch": 1.2183925575720604,
+      "grad_norm": 0.2194120357344308,
+      "learning_rate": 3.9802284089919876e-06,
+      "loss": 0.1925,
+      "step": 1998
+    },
+    {
+      "epoch": 1.2190025926490773,
+      "grad_norm": 0.22978200474534835,
+      "learning_rate": 3.975019709592189e-06,
+      "loss": 0.2111,
+      "step": 1999
+    },
+    {
+      "epoch": 1.2196126277260944,
+      "grad_norm": 0.22765530069297446,
+      "learning_rate": 3.969812171059516e-06,
+      "loss": 0.1976,
+      "step": 2000
+    },
+    {
+      "epoch": 1.2196126277260944,
+      "eval_loss": 0.22462303936481476,
+      "eval_runtime": 265.1969,
+      "eval_samples_per_second": 4.035,
+      "eval_steps_per_second": 0.128,
+      "step": 2000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3280,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2422457185075200.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}