diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,7180 +1,5080 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 12.0,
+  "epoch": 11.899159663865547,
   "eval_steps": 500,
-  "global_step": 1008,
+  "global_step": 708,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.011904761904761904,
-      "grad_norm": 17.139904900733598,
-      "learning_rate": 6.451612903225807e-07,
-      "loss": 9.2899,
+      "epoch": 0.01680672268907563,
+      "grad_norm": 4.385648622708784,
+      "learning_rate": 9.090909090909091e-07,
+      "loss": 4.108,
       "step": 1
     },
     {
-      "epoch": 0.023809523809523808,
-      "grad_norm": 16.71486840638675,
-      "learning_rate": 1.2903225806451614e-06,
-      "loss": 9.1432,
+      "epoch": 0.03361344537815126,
+      "grad_norm": 4.427354400402251,
+      "learning_rate": 1.8181818181818183e-06,
+      "loss": 4.1105,
       "step": 2
     },
     {
-      "epoch": 0.03571428571428571,
-      "grad_norm": 17.28789649557261,
-      "learning_rate": 1.935483870967742e-06,
-      "loss": 9.5388,
+      "epoch": 0.05042016806722689,
+      "grad_norm": 3.8339824048548925,
+      "learning_rate": 2.7272727272727272e-06,
+      "loss": 3.8298,
       "step": 3
     },
     {
-      "epoch": 0.047619047619047616,
-      "grad_norm": 13.220304931739214,
-      "learning_rate": 2.580645161290323e-06,
-      "loss": 7.9169,
+      "epoch": 0.06722689075630252,
+      "grad_norm": 4.9404739040660814,
+      "learning_rate": 3.6363636363636366e-06,
+      "loss": 4.4266,
       "step": 4
     },
     {
-      "epoch": 0.05952380952380952,
-      "grad_norm": 13.609291935209646,
-      "learning_rate": 3.225806451612903e-06,
-      "loss": 7.8849,
+      "epoch": 0.08403361344537816,
+      "grad_norm": 3.723380522292492,
+      "learning_rate": 4.5454545454545455e-06,
+      "loss": 3.6814,
       "step": 5
     },
     {
-      "epoch": 0.07142857142857142,
-      "grad_norm": 6.3010952694138105,
-      "learning_rate": 3.870967741935484e-06,
-      "loss": 5.0183,
+      "epoch": 0.10084033613445378,
+      "grad_norm": 10.136378288136632,
+      "learning_rate": 5.4545454545454545e-06,
+      "loss": 6.7591,
       "step": 6
     },
     {
-      "epoch": 0.08333333333333333,
-      "grad_norm": 9.731002424537186,
-      "learning_rate": 4.516129032258065e-06,
-      "loss": 6.5099,
+      "epoch": 0.11764705882352941,
+      "grad_norm": 3.750584042652294,
+      "learning_rate": 6.363636363636364e-06,
+      "loss": 3.6869,
       "step": 7
     },
     {
-      "epoch": 0.09523809523809523,
-      "grad_norm": 16.946585239278296,
-      "learning_rate": 5.161290322580646e-06,
-      "loss": 9.0453,
+      "epoch": 0.13445378151260504,
+      "grad_norm": 10.256473970042155,
+      "learning_rate": 7.272727272727273e-06,
+      "loss": 6.8465,
       "step": 8
     },
     {
-      "epoch": 0.10714285714285714,
-      "grad_norm": 13.29361044735187,
-      "learning_rate": 5.806451612903226e-06,
-      "loss": 7.8482,
+      "epoch": 0.15126050420168066,
+      "grad_norm": 5.299551412905417,
+      "learning_rate": 8.181818181818183e-06,
+      "loss": 4.627,
       "step": 9
     },
     {
-      "epoch": 0.11904761904761904,
-      "grad_norm": 9.616005044641154,
-      "learning_rate": 6.451612903225806e-06,
-      "loss": 6.374,
+      "epoch": 0.16806722689075632,
+      "grad_norm": 4.073848965470436,
+      "learning_rate": 9.090909090909091e-06,
+      "loss": 3.7922,
       "step": 10
     },
     {
-      "epoch": 0.13095238095238096,
-      "grad_norm": 13.507401041826714,
-      "learning_rate": 7.096774193548388e-06,
-      "loss": 7.8621,
+      "epoch": 0.18487394957983194,
+      "grad_norm": 6.270475941131987,
+      "learning_rate": 1e-05,
+      "loss": 5.0956,
       "step": 11
     },
     {
-      "epoch": 0.14285714285714285,
-      "grad_norm": 10.212045595466936,
-      "learning_rate": 7.741935483870968e-06,
-      "loss": 6.4647,
+      "epoch": 0.20168067226890757,
+      "grad_norm": 5.6458603155827225,
+      "learning_rate": 1.0909090909090909e-05,
+      "loss": 4.6846,
       "step": 12
     },
     {
-      "epoch": 0.15476190476190477,
-      "grad_norm": 13.721855682512686,
-      "learning_rate": 8.387096774193549e-06,
-      "loss": 7.5479,
+      "epoch": 0.2184873949579832,
+      "grad_norm": 4.136016188838338,
+      "learning_rate": 1.181818181818182e-05,
+      "loss": 3.9917,
       "step": 13
     },
     {
-      "epoch": 0.16666666666666666,
-      "grad_norm": 13.558802891735498,
-      "learning_rate": 9.03225806451613e-06,
-      "loss": 7.8138,
+      "epoch": 0.23529411764705882,
+      "grad_norm": 3.6135748045353333,
+      "learning_rate": 1.2727272727272728e-05,
+      "loss": 3.6259,
       "step": 14
     },
     {
-      "epoch": 0.17857142857142858,
-      "grad_norm": 13.55353645782886,
-      "learning_rate": 9.67741935483871e-06,
-      "loss": 7.6195,
+      "epoch": 0.25210084033613445,
+      "grad_norm": 4.964675562915649,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 4.2151,
       "step": 15
     },
     {
-      "epoch": 0.19047619047619047,
-      "grad_norm": 13.221857295140415,
-      "learning_rate": 1.0322580645161291e-05,
-      "loss": 7.5711,
+      "epoch": 0.2689075630252101,
+      "grad_norm": 6.524726574282932,
+      "learning_rate": 1.4545454545454546e-05,
+      "loss": 5.0345,
       "step": 16
     },
     {
-      "epoch": 0.20238095238095238,
-      "grad_norm": 14.021266819970709,
-      "learning_rate": 1.096774193548387e-05,
-      "loss": 7.4885,
+      "epoch": 0.2857142857142857,
+      "grad_norm": 5.639830925355285,
+      "learning_rate": 1.5454545454545454e-05,
+      "loss": 4.4856,
       "step": 17
     },
     {
-      "epoch": 0.21428571428571427,
-      "grad_norm": 13.88113956103264,
-      "learning_rate": 1.1612903225806453e-05,
-      "loss": 7.3179,
+      "epoch": 0.3025210084033613,
+      "grad_norm": 3.8533215935243077,
+      "learning_rate": 1.6363636363636366e-05,
+      "loss": 3.7071,
       "step": 18
     },
     {
-      "epoch": 0.2261904761904762,
-      "grad_norm": 17.39044599261459,
-      "learning_rate": 1.2258064516129034e-05,
-      "loss": 8.622,
+      "epoch": 0.31932773109243695,
+      "grad_norm": 5.161749059636034,
+      "learning_rate": 1.7272727272727274e-05,
+      "loss": 4.2947,
       "step": 19
     },
     {
-      "epoch": 0.23809523809523808,
-      "grad_norm": 17.515984626148686,
-      "learning_rate": 1.2903225806451613e-05,
-      "loss": 8.5228,
+      "epoch": 0.33613445378151263,
+      "grad_norm": 3.6136568344159885,
+      "learning_rate": 1.8181818181818182e-05,
+      "loss": 3.4683,
       "step": 20
     },
     {
-      "epoch": 0.25,
-      "grad_norm": 17.82859899843793,
-      "learning_rate": 1.3548387096774194e-05,
-      "loss": 8.7837,
+      "epoch": 0.35294117647058826,
+      "grad_norm": 4.435498267084917,
+      "learning_rate": 1.9090909090909094e-05,
+      "loss": 3.859,
       "step": 21
     },
     {
-      "epoch": 0.2619047619047619,
-      "grad_norm": 17.429672226015118,
-      "learning_rate": 1.4193548387096776e-05,
-      "loss": 8.2317,
+      "epoch": 0.3697478991596639,
+      "grad_norm": 3.919091925431486,
+      "learning_rate": 2e-05,
+      "loss": 3.5794,
       "step": 22
     },
     {
-      "epoch": 0.27380952380952384,
-      "grad_norm": 17.276258653744623,
-      "learning_rate": 1.4838709677419357e-05,
-      "loss": 8.0633,
+      "epoch": 0.3865546218487395,
+      "grad_norm": 3.488028179094918,
+      "learning_rate": 1.9999895137366746e-05,
+      "loss": 3.2719,
       "step": 23
     },
     {
-      "epoch": 0.2857142857142857,
-      "grad_norm": 9.957462076322903,
-      "learning_rate": 1.5483870967741936e-05,
-      "loss": 5.7366,
+      "epoch": 0.40336134453781514,
+      "grad_norm": 10.328288130878647,
+      "learning_rate": 1.9999580551666215e-05,
+      "loss": 5.7591,
       "step": 24
     },
     {
-      "epoch": 0.2976190476190476,
-      "grad_norm": 13.620021596849393,
-      "learning_rate": 1.6129032258064517e-05,
-      "loss": 6.7714,
+      "epoch": 0.42016806722689076,
+      "grad_norm": 3.774723233998757,
+      "learning_rate": 1.9999056249496065e-05,
+      "loss": 3.3805,
       "step": 25
     },
     {
-      "epoch": 0.30952380952380953,
-      "grad_norm": 6.569900345372609,
-      "learning_rate": 1.6774193548387098e-05,
-      "loss": 4.4136,
+      "epoch": 0.4369747899159664,
+      "grad_norm": 6.265735848819648,
+      "learning_rate": 1.9998322241852238e-05,
+      "loss": 4.2421,
       "step": 26
     },
     {
-      "epoch": 0.32142857142857145,
-      "grad_norm": 14.289993829511333,
-      "learning_rate": 1.741935483870968e-05,
-      "loss": 6.1463,
+      "epoch": 0.453781512605042,
+      "grad_norm": 5.405371137822227,
+      "learning_rate": 1.999737854412873e-05,
+      "loss": 3.8983,
       "step": 27
     },
     {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 10.418999900918601,
-      "learning_rate": 1.806451612903226e-05,
-      "loss": 4.9552,
+      "epoch": 0.47058823529411764,
+      "grad_norm": 3.7669212684787263,
+      "learning_rate": 1.9996225176117264e-05,
+      "loss": 3.1467,
       "step": 28
     },
     {
-      "epoch": 0.34523809523809523,
-      "grad_norm": 17.940074442849337,
-      "learning_rate": 1.870967741935484e-05,
-      "loss": 6.7963,
+      "epoch": 0.48739495798319327,
+      "grad_norm": 5.5431024884923845,
+      "learning_rate": 1.999486216200688e-05,
+      "loss": 3.5476,
       "step": 29
     },
     {
-      "epoch": 0.35714285714285715,
-      "grad_norm": 10.183500386894858,
-      "learning_rate": 1.935483870967742e-05,
-      "loss": 4.7842,
+      "epoch": 0.5042016806722689,
+      "grad_norm": 5.394079800803299,
+      "learning_rate": 1.9993289530383433e-05,
+      "loss": 3.4722,
       "step": 30
     },
     {
-      "epoch": 0.36904761904761907,
-      "grad_norm": 14.500865445423567,
-      "learning_rate": 2e-05,
-      "loss": 5.5398,
+      "epoch": 0.5210084033613446,
+      "grad_norm": 5.971573494363461,
+      "learning_rate": 1.999150731422898e-05,
+      "loss": 3.6263,
       "step": 31
     },
     {
-      "epoch": 0.38095238095238093,
-      "grad_norm": 11.000017652493565,
-      "learning_rate": 1.9999948301225546e-05,
-      "loss": 4.5158,
+      "epoch": 0.5378151260504201,
+      "grad_norm": 7.378751228659765,
+      "learning_rate": 1.9989515550921088e-05,
+      "loss": 3.5546,
       "step": 32
     },
     {
-      "epoch": 0.39285714285714285,
-      "grad_norm": 16.727462615797,
-      "learning_rate": 1.999979320543672e-05,
-      "loss": 4.6378,
+      "epoch": 0.5546218487394958,
+      "grad_norm": 5.09952164148768,
+      "learning_rate": 1.998731428223208e-05,
+      "loss": 3.0747,
       "step": 33
     },
     {
-      "epoch": 0.40476190476190477,
-      "grad_norm": 11.599298962952556,
-      "learning_rate": 1.999953471423719e-05,
-      "loss": 4.1898,
+      "epoch": 0.5714285714285714,
+      "grad_norm": 4.765579668290796,
+      "learning_rate": 1.9984903554328116e-05,
+      "loss": 2.7589,
       "step": 34
     },
     {
-      "epoch": 0.4166666666666667,
-      "grad_norm": 12.887490770056774,
-      "learning_rate": 1.999917283029968e-05,
-      "loss": 3.9778,
+      "epoch": 0.5882352941176471,
+      "grad_norm": 7.224492836222785,
+      "learning_rate": 1.9982283417768247e-05,
+      "loss": 2.871,
       "step": 35
     },
     {
-      "epoch": 0.42857142857142855,
-      "grad_norm": 13.379378492479097,
-      "learning_rate": 1.9998707557365983e-05,
-      "loss": 3.3893,
+      "epoch": 0.6050420168067226,
+      "grad_norm": 12.931777704179952,
+      "learning_rate": 1.9979453927503366e-05,
+      "loss": 3.215,
       "step": 36
     },
     {
-      "epoch": 0.44047619047619047,
-      "grad_norm": 19.77665757370661,
-      "learning_rate": 1.9998138900246916e-05,
-      "loss": 3.134,
+      "epoch": 0.6218487394957983,
+      "grad_norm": 3.6837055989248717,
+      "learning_rate": 1.9976415142875022e-05,
+      "loss": 2.4582,
       "step": 37
     },
     {
-      "epoch": 0.4523809523809524,
-      "grad_norm": 13.041378224509636,
-      "learning_rate": 1.999746686482224e-05,
-      "loss": 3.0121,
+      "epoch": 0.6386554621848739,
+      "grad_norm": 3.163696044731996,
+      "learning_rate": 1.9973167127614218e-05,
+      "loss": 2.4035,
       "step": 38
     },
     {
-      "epoch": 0.4642857142857143,
-      "grad_norm": 10.907950909034863,
-      "learning_rate": 1.999669145804065e-05,
-      "loss": 2.8979,
+      "epoch": 0.6554621848739496,
+      "grad_norm": 2.6130379292490127,
+      "learning_rate": 1.9969709949840034e-05,
+      "loss": 2.27,
       "step": 39
     },
     {
-      "epoch": 0.47619047619047616,
-      "grad_norm": 7.708693945099904,
-      "learning_rate": 1.9995812687919653e-05,
-      "loss": 2.705,
+      "epoch": 0.6722689075630253,
+      "grad_norm": 3.0774669175573903,
+      "learning_rate": 1.9966043682058223e-05,
+      "loss": 2.2352,
       "step": 40
     },
     {
-      "epoch": 0.4880952380952381,
-      "grad_norm": 4.95653030392137,
-      "learning_rate": 1.9994830563545524e-05,
-      "loss": 2.5711,
+      "epoch": 0.6890756302521008,
+      "grad_norm": 3.3703328756445514,
+      "learning_rate": 1.9962168401159685e-05,
+      "loss": 2.3572,
       "step": 41
     },
     {
-      "epoch": 0.5,
-      "grad_norm": 8.083358218918184,
-      "learning_rate": 1.999374509507318e-05,
-      "loss": 2.5499,
+      "epoch": 0.7058823529411765,
+      "grad_norm": 6.674678012817149,
+      "learning_rate": 1.995808418841885e-05,
+      "loss": 2.5806,
       "step": 42
     },
     {
-      "epoch": 0.5119047619047619,
-      "grad_norm": 8.299946354650668,
-      "learning_rate": 1.999255629372611e-05,
-      "loss": 2.4961,
+      "epoch": 0.7226890756302521,
+      "grad_norm": 2.7692115403847932,
+      "learning_rate": 1.9953791129491985e-05,
+      "loss": 2.2257,
       "step": 43
     },
     {
-      "epoch": 0.5238095238095238,
-      "grad_norm": 17.357515973246574,
-      "learning_rate": 1.9991264171796213e-05,
-      "loss": 1.8883,
+      "epoch": 0.7394957983193278,
+      "grad_norm": 3.4385080781188013,
+      "learning_rate": 1.9949289314415373e-05,
+      "loss": 2.3251,
       "step": 44
     },
     {
-      "epoch": 0.5357142857142857,
-      "grad_norm": 8.01068189035409,
-      "learning_rate": 1.9989868742643725e-05,
-      "loss": 2.2778,
+      "epoch": 0.7563025210084033,
+      "grad_norm": 3.513307097744062,
+      "learning_rate": 1.994457883760346e-05,
+      "loss": 2.2311,
       "step": 45
     },
     {
-      "epoch": 0.5476190476190477,
-      "grad_norm": 8.493792830509063,
-      "learning_rate": 1.998837002069704e-05,
-      "loss": 2.1784,
+      "epoch": 0.773109243697479,
+      "grad_norm": 3.485460514370931,
+      "learning_rate": 1.993965979784684e-05,
+      "loss": 2.0451,
       "step": 46
     },
     {
-      "epoch": 0.5595238095238095,
-      "grad_norm": 8.346064217450877,
-      "learning_rate": 1.9986768021452575e-05,
-      "loss": 2.1233,
+      "epoch": 0.7899159663865546,
+      "grad_norm": 2.6480026940047967,
+      "learning_rate": 1.9934532298310206e-05,
+      "loss": 2.1224,
       "step": 47
     },
     {
-      "epoch": 0.5714285714285714,
-      "grad_norm": 11.141768579260864,
-      "learning_rate": 1.9985062761474605e-05,
-      "loss": 1.5602,
+      "epoch": 0.8067226890756303,
+      "grad_norm": 3.191875488563921,
+      "learning_rate": 1.9929196446530184e-05,
+      "loss": 2.1002,
       "step": 48
     },
     {
-      "epoch": 0.5833333333333334,
-      "grad_norm": 11.272582504816725,
-      "learning_rate": 1.9983254258395105e-05,
-      "loss": 1.4972,
+      "epoch": 0.8235294117647058,
+      "grad_norm": 3.3443355134212833,
+      "learning_rate": 1.992365235441306e-05,
+      "loss": 2.0233,
       "step": 49
     },
     {
-      "epoch": 0.5952380952380952,
-      "grad_norm": 4.0084233661961415,
-      "learning_rate": 1.9981342530913556e-05,
-      "loss": 2.4526,
+      "epoch": 0.8403361344537815,
+      "grad_norm": 2.5999640350239948,
+      "learning_rate": 1.991790013823246e-05,
+      "loss": 1.966,
       "step": 50
     },
     {
-      "epoch": 0.6071428571428571,
-      "grad_norm": 9.158504351728157,
-      "learning_rate": 1.997932759879674e-05,
-      "loss": 1.234,
+      "epoch": 0.8571428571428571,
+      "grad_norm": 6.089294042169637,
+      "learning_rate": 1.991193991862689e-05,
+      "loss": 1.8622,
       "step": 51
     },
     {
-      "epoch": 0.6190476190476191,
-      "grad_norm": 9.338811416846326,
-      "learning_rate": 1.9977209482878576e-05,
-      "loss": 1.1168,
+      "epoch": 0.8739495798319328,
+      "grad_norm": 3.2204266121939362,
+      "learning_rate": 1.9905771820597214e-05,
+      "loss": 1.8818,
       "step": 52
     },
     {
-      "epoch": 0.6309523809523809,
-      "grad_norm": 11.308701021449329,
-      "learning_rate": 1.9974988205059848e-05,
-      "loss": 1.0833,
+      "epoch": 0.8907563025210085,
+      "grad_norm": 3.3930839641806942,
+      "learning_rate": 1.989939597350403e-05,
+      "loss": 1.7539,
       "step": 53
     },
     {
-      "epoch": 0.6428571428571429,
-      "grad_norm": 5.618069930191442,
-      "learning_rate": 1.997266378830804e-05,
-      "loss": 2.1432,
+      "epoch": 0.907563025210084,
+      "grad_norm": 3.449905117774067,
+      "learning_rate": 1.9892812511064962e-05,
+      "loss": 1.7651,
       "step": 54
     },
     {
-      "epoch": 0.6547619047619048,
-      "grad_norm": 6.282292794573764,
-      "learning_rate": 1.9970236256657044e-05,
-      "loss": 0.9732,
+      "epoch": 0.9243697478991597,
+      "grad_norm": 3.3847593536709097,
+      "learning_rate": 1.9886021571351854e-05,
+      "loss": 1.6882,
       "step": 55
     },
     {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 4.707647398070364,
-      "learning_rate": 1.9967705635206937e-05,
-      "loss": 1.5313,
+      "epoch": 0.9411764705882353,
+      "grad_norm": 3.845735022445724,
+      "learning_rate": 1.9879023296787866e-05,
+      "loss": 1.719,
       "step": 56
     },
     {
-      "epoch": 0.6785714285714286,
-      "grad_norm": 6.472249482990302,
-      "learning_rate": 1.9965071950123732e-05,
-      "loss": 0.8117,
+      "epoch": 0.957983193277311,
+      "grad_norm": 2.8219721101939386,
+      "learning_rate": 1.9871817834144506e-05,
+      "loss": 1.7362,
       "step": 57
     },
     {
-      "epoch": 0.6904761904761905,
-      "grad_norm": 4.307567663459201,
-      "learning_rate": 1.9962335228639084e-05,
-      "loss": 0.8454,
+      "epoch": 0.9747899159663865,
+      "grad_norm": 2.499202155153014,
+      "learning_rate": 1.9864405334538518e-05,
+      "loss": 1.7014,
       "step": 58
     },
     {
-      "epoch": 0.7023809523809523,
-      "grad_norm": 3.076092506616235,
-      "learning_rate": 1.9959495499050027e-05,
-      "loss": 0.6874,
+      "epoch": 0.9915966386554622,
+      "grad_norm": 2.511155724416345,
+      "learning_rate": 1.9856785953428757e-05,
+      "loss": 1.7195,
+      "step": 59
+    },
+    {
+      "epoch": 0.9915966386554622,
+      "eval_loss": 1.2069061994552612,
+      "eval_runtime": 115.8201,
+      "eval_samples_per_second": 0.812,
+      "eval_steps_per_second": 0.812,
       "step": 59
     },
     {
-      "epoch": 0.7142857142857143,
-      "grad_norm": 4.764863570511213,
-      "learning_rate": 1.9956552790718663e-05,
-      "loss": 1.9864,
+      "epoch": 1.0084033613445378,
+      "grad_norm": 5.357847266371679,
+      "learning_rate": 1.9848959850612895e-05,
+      "loss": 1.1994,
       "step": 60
     },
     {
-      "epoch": 0.7261904761904762,
-      "grad_norm": 4.331608161030718,
-      "learning_rate": 1.995350713407188e-05,
-      "loss": 0.252,
+      "epoch": 1.0252100840336134,
+      "grad_norm": 2.5172607304803383,
+      "learning_rate": 1.9840927190224083e-05,
+      "loss": 1.5096,
       "step": 61
     },
     {
-      "epoch": 0.7380952380952381,
-      "grad_norm": 6.590145628301957,
-      "learning_rate": 1.9950358560601014e-05,
-      "loss": 1.3728,
+      "epoch": 1.0420168067226891,
+      "grad_norm": 2.383993749480795,
+      "learning_rate": 1.9832688140727502e-05,
+      "loss": 1.5336,
       "step": 62
     },
     {
-      "epoch": 0.75,
-      "grad_norm": 4.867440106476626,
-      "learning_rate": 1.994710710286155e-05,
-      "loss": 0.8254,
+      "epoch": 1.0588235294117647,
+      "grad_norm": 5.571125580296848,
+      "learning_rate": 1.982424287491684e-05,
+      "loss": 1.0043,
       "step": 63
     },
     {
-      "epoch": 0.7619047619047619,
-      "grad_norm": 2.4733606881650556,
-      "learning_rate": 1.994375279447276e-05,
-      "loss": 0.7131,
+      "epoch": 1.0756302521008403,
+      "grad_norm": 2.2388809453197083,
+      "learning_rate": 1.9815591569910654e-05,
+      "loss": 1.2905,
       "step": 64
     },
     {
-      "epoch": 0.7738095238095238,
-      "grad_norm": 4.226942008613792,
-      "learning_rate": 1.9940295670117373e-05,
-      "loss": 0.2012,
+      "epoch": 1.092436974789916,
+      "grad_norm": 2.4116257513181356,
+      "learning_rate": 1.9806734407148674e-05,
+      "loss": 1.5209,
       "step": 65
     },
     {
-      "epoch": 0.7857142857142857,
-      "grad_norm": 3.08282145132758,
-      "learning_rate": 1.9936735765541202e-05,
-      "loss": 1.3355,
+      "epoch": 1.1092436974789917,
+      "grad_norm": 2.268214660771143,
+      "learning_rate": 1.9797671572387985e-05,
+      "loss": 1.3414,
       "step": 66
     },
     {
-      "epoch": 0.7976190476190477,
-      "grad_norm": 2.625421213967175,
-      "learning_rate": 1.9933073117552798e-05,
-      "loss": 1.1274,
+      "epoch": 1.1260504201680672,
+      "grad_norm": 2.1242648315918036,
+      "learning_rate": 1.978840325569912e-05,
+      "loss": 1.3624,
       "step": 67
     },
     {
-      "epoch": 0.8095238095238095,
-      "grad_norm": 4.111053115808539,
-      "learning_rate": 1.9929307764023032e-05,
-      "loss": 1.2906,
+      "epoch": 1.1428571428571428,
+      "grad_norm": 6.45315294142602,
+      "learning_rate": 1.977892965146211e-05,
+      "loss": 0.7577,
       "step": 68
     },
     {
-      "epoch": 0.8214285714285714,
-      "grad_norm": 2.284397902515471,
-      "learning_rate": 1.992543974388475e-05,
-      "loss": 1.1372,
+      "epoch": 1.1596638655462184,
+      "grad_norm": 2.0834553976731596,
+      "learning_rate": 1.976925095836236e-05,
+      "loss": 1.242,
       "step": 69
     },
     {
-      "epoch": 0.8333333333333334,
-      "grad_norm": 4.81537458518723,
-      "learning_rate": 1.992146909713232e-05,
-      "loss": 0.7222,
+      "epoch": 1.1764705882352942,
+      "grad_norm": 2.1599404414408934,
+      "learning_rate": 1.975936737938653e-05,
+      "loss": 1.2737,
       "step": 70
     },
     {
-      "epoch": 0.8452380952380952,
-      "grad_norm": 4.635344777090823,
-      "learning_rate": 1.9917395864821262e-05,
-      "loss": 1.3808,
+      "epoch": 1.1932773109243697,
+      "grad_norm": 3.6995588982952263,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 0.7993,
       "step": 71
     },
     {
-      "epoch": 0.8571428571428571,
-      "grad_norm": 5.22342912644428,
-      "learning_rate": 1.9913220089067794e-05,
-      "loss": 2.2206,
+      "epoch": 1.2100840336134453,
+      "grad_norm": 2.352377611171032,
+      "learning_rate": 1.9738986397233736e-05,
+      "loss": 0.7334,
       "step": 72
     },
     {
-      "epoch": 0.8690476190476191,
-      "grad_norm": 4.03293534110668,
-      "learning_rate": 1.990894181304842e-05,
-      "loss": 1.8158,
+      "epoch": 1.226890756302521,
+      "grad_norm": 2.152202331744084,
+      "learning_rate": 1.9728489421497465e-05,
+      "loss": 1.3078,
       "step": 73
     },
     {
-      "epoch": 0.8809523809523809,
-      "grad_norm": 2.675790769254283,
-      "learning_rate": 1.9904561080999466e-05,
-      "loss": 0.6809,
+      "epoch": 1.2436974789915967,
+      "grad_norm": 2.5203922797690015,
+      "learning_rate": 1.9717788414757523e-05,
+      "loss": 1.1814,
       "step": 74
     },
     {
-      "epoch": 0.8928571428571429,
-      "grad_norm": 6.044524344029198,
-      "learning_rate": 1.9900077938216623e-05,
-      "loss": 1.2401,
+      "epoch": 1.2605042016806722,
+      "grad_norm": 3.4705153272314475,
+      "learning_rate": 1.9706883601441066e-05,
+      "loss": 1.3192,
       "step": 75
     },
     {
-      "epoch": 0.9047619047619048,
-      "grad_norm": 3.975480451115108,
-      "learning_rate": 1.9895492431054492e-05,
-      "loss": 1.2328,
+      "epoch": 1.2773109243697478,
+      "grad_norm": 1.8630027919414862,
+      "learning_rate": 1.969577521024958e-05,
+      "loss": 0.7795,
       "step": 76
     },
     {
-      "epoch": 0.9166666666666666,
-      "grad_norm": 3.7971112191541994,
-      "learning_rate": 1.989080460692609e-05,
-      "loss": 2.1511,
+      "epoch": 1.2941176470588236,
+      "grad_norm": 2.9345669701977912,
+      "learning_rate": 1.9684463474154095e-05,
+      "loss": 1.2665,
       "step": 77
     },
     {
-      "epoch": 0.9285714285714286,
-      "grad_norm": 4.980700467181216,
-      "learning_rate": 1.988601451430237e-05,
-      "loss": 0.6743,
+      "epoch": 1.3109243697478992,
+      "grad_norm": 1.4751109531953661,
+      "learning_rate": 1.9672948630390296e-05,
+      "loss": 0.6425,
       "step": 78
     },
     {
-      "epoch": 0.9404761904761905,
-      "grad_norm": 2.4914351987187953,
-      "learning_rate": 1.9881122202711717e-05,
-      "loss": 1.1282,
+      "epoch": 1.3277310924369747,
+      "grad_norm": 3.300164935057302,
+      "learning_rate": 1.9661230920453553e-05,
+      "loss": 1.2188,
       "step": 79
     },
     {
-      "epoch": 0.9523809523809523,
-      "grad_norm": 2.836683437613531,
-      "learning_rate": 1.9876127722739436e-05,
-      "loss": 0.5786,
+      "epoch": 1.3445378151260505,
+      "grad_norm": 2.835411015713917,
+      "learning_rate": 1.964931059009385e-05,
+      "loss": 1.263,
       "step": 80
     },
     {
-      "epoch": 0.9642857142857143,
-      "grad_norm": 3.038983217485337,
-      "learning_rate": 1.987103112602722e-05,
-      "loss": 1.1614,
+      "epoch": 1.361344537815126,
+      "grad_norm": 2.235484344805954,
+      "learning_rate": 1.9637187889310632e-05,
+      "loss": 1.0342,
       "step": 81
     },
     {
-      "epoch": 0.9761904761904762,
-      "grad_norm": 3.137725304157932,
-      "learning_rate": 1.9865832465272636e-05,
-      "loss": 1.0887,
+      "epoch": 1.3781512605042017,
+      "grad_norm": 1.8454261871804942,
+      "learning_rate": 1.9624863072347565e-05,
+      "loss": 1.0816,
       "step": 82
     },
     {
-      "epoch": 0.9880952380952381,
-      "grad_norm": 4.0982549880174055,
-      "learning_rate": 1.986053179422856e-05,
-      "loss": 0.6275,
+      "epoch": 1.3949579831932772,
+      "grad_norm": 1.8700087156945158,
+      "learning_rate": 1.96123363976872e-05,
+      "loss": 1.0611,
       "step": 83
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 4.228465015472441,
-      "learning_rate": 1.9855129167702625e-05,
-      "loss": 1.247,
+      "epoch": 1.4117647058823528,
+      "grad_norm": 1.776825090533238,
+      "learning_rate": 1.9599608128045554e-05,
+      "loss": 1.0875,
       "step": 84
     },
     {
-      "epoch": 1.0,
-      "eval_loss": 0.855706512928009,
-      "eval_runtime": 57.6965,
-      "eval_samples_per_second": 1.04,
-      "eval_steps_per_second": 1.04,
-      "step": 84
-    },
-    {
-      "epoch": 1.0119047619047619,
-      "grad_norm": 2.4225946522090993,
-      "learning_rate": 1.9849624641556676e-05,
-      "loss": 1.0853,
+      "epoch": 1.4285714285714286,
+      "grad_norm": 3.2507229249266074,
+      "learning_rate": 1.9586678530366607e-05,
+      "loss": 0.6209,
       "step": 85
     },
     {
-      "epoch": 1.0238095238095237,
-      "grad_norm": 3.7101089119391726,
-      "learning_rate": 1.9844018272706155e-05,
-      "loss": 1.1892,
+      "epoch": 1.4453781512605042,
+      "grad_norm": 2.0636370694690442,
+      "learning_rate": 1.9573547875816685e-05,
+      "loss": 1.0625,
       "step": 86
     },
     {
-      "epoch": 1.0357142857142858,
-      "grad_norm": 2.755180603183131,
-      "learning_rate": 1.9838310119119545e-05,
-      "loss": 1.1356,
+      "epoch": 1.46218487394958,
+      "grad_norm": 1.941254802441877,
+      "learning_rate": 1.9560216439778795e-05,
+      "loss": 1.086,
       "step": 87
     },
     {
-      "epoch": 1.0476190476190477,
-      "grad_norm": 2.2353314330160097,
-      "learning_rate": 1.983250023981776e-05,
-      "loss": 1.4771,
+      "epoch": 1.4789915966386555,
+      "grad_norm": 1.9461861491812582,
+      "learning_rate": 1.954668450184683e-05,
+      "loss": 0.9987,
       "step": 88
     },
     {
-      "epoch": 1.0595238095238095,
-      "grad_norm": 2.1469727041530002,
-      "learning_rate": 1.9826588694873515e-05,
-      "loss": 1.4555,
+      "epoch": 1.495798319327731,
+      "grad_norm": 2.064972172013034,
+      "learning_rate": 1.9532952345819723e-05,
+      "loss": 1.0663,
       "step": 89
     },
     {
-      "epoch": 1.0714285714285714,
-      "grad_norm": 2.669812575529475,
-      "learning_rate": 1.982057554541075e-05,
-      "loss": 0.2101,
+      "epoch": 1.5126050420168067,
+      "grad_norm": 1.778380925304329,
+      "learning_rate": 1.951902025969548e-05,
+      "loss": 0.9426,
       "step": 90
     },
     {
-      "epoch": 1.0833333333333333,
-      "grad_norm": 2.9321935610432384,
-      "learning_rate": 1.981446085360395e-05,
-      "loss": 1.0504,
+      "epoch": 1.5294117647058822,
+      "grad_norm": 2.361605116359202,
+      "learning_rate": 1.950488853566515e-05,
+      "loss": 0.9878,
       "step": 91
     },
     {
-      "epoch": 1.0952380952380953,
-      "grad_norm": 2.4895259465307613,
-      "learning_rate": 1.980824468267753e-05,
-      "loss": 0.6211,
+      "epoch": 1.5462184873949578,
+      "grad_norm": 1.7448015788038604,
+      "learning_rate": 1.949055747010669e-05,
+      "loss": 1.054,
       "step": 92
     },
     {
-      "epoch": 1.1071428571428572,
-      "grad_norm": 3.3343638502561483,
-      "learning_rate": 1.980192709690517e-05,
-      "loss": 1.0306,
+      "epoch": 1.5630252100840336,
+      "grad_norm": 2.411683085008472,
+      "learning_rate": 1.9476027363578754e-05,
+      "loss": 0.9234,
       "step": 93
     },
     {
-      "epoch": 1.119047619047619,
-      "grad_norm": 2.3278851742380944,
-      "learning_rate": 1.9795508161609166e-05,
-      "loss": 0.5428,
+      "epoch": 1.5798319327731094,
+      "grad_norm": 1.8173611017503086,
+      "learning_rate": 1.946129852081439e-05,
+      "loss": 0.9983,
       "step": 94
     },
     {
-      "epoch": 1.130952380952381,
-      "grad_norm": 1.482575910310281,
-      "learning_rate": 1.978898794315973e-05,
-      "loss": 1.0905,
+      "epoch": 1.596638655462185,
+      "grad_norm": 1.9295587283934226,
+      "learning_rate": 1.9446371250714645e-05,
+      "loss": 0.9747,
       "step": 95
     },
     {
-      "epoch": 1.1428571428571428,
-      "grad_norm": 2.0177449871722386,
-      "learning_rate": 1.9782366508974325e-05,
-      "loss": 0.6188,
+      "epoch": 1.6134453781512605,
+      "grad_norm": 1.772861667848063,
+      "learning_rate": 1.943124586634209e-05,
+      "loss": 0.8761,
       "step": 96
     },
     {
-      "epoch": 1.1547619047619047,
-      "grad_norm": 4.772581701887709,
-      "learning_rate": 1.9775643927516956e-05,
-      "loss": 1.4641,
+      "epoch": 1.6302521008403361,
+      "grad_norm": 4.814779501362561,
+      "learning_rate": 1.941592268491425e-05,
+      "loss": 0.6477,
       "step": 97
     },
     {
-      "epoch": 1.1666666666666667,
-      "grad_norm": 1.851753440929594,
-      "learning_rate": 1.9768820268297467e-05,
-      "loss": 0.5434,
+      "epoch": 1.6470588235294117,
+      "grad_norm": 1.836972617211195,
+      "learning_rate": 1.9400402027796955e-05,
+      "loss": 0.924,
       "step": 98
     },
     {
-      "epoch": 1.1785714285714286,
-      "grad_norm": 1.8955707469335776,
-      "learning_rate": 1.976189560187083e-05,
-      "loss": 0.6231,
+      "epoch": 1.6638655462184873,
+      "grad_norm": 2.3766647197122595,
+      "learning_rate": 1.9384684220497605e-05,
+      "loss": 0.7468,
       "step": 99
     },
     {
-      "epoch": 1.1904761904761905,
-      "grad_norm": 1.4609006468604573,
-      "learning_rate": 1.9754869999836385e-05,
-      "loss": 0.898,
+      "epoch": 1.680672268907563,
+      "grad_norm": 1.6832736777019341,
+      "learning_rate": 1.936876959265833e-05,
+      "loss": 0.8439,
       "step": 100
     },
     {
-      "epoch": 1.2023809523809523,
-      "grad_norm": 2.455697301128514,
-      "learning_rate": 1.9747743534837143e-05,
-      "loss": 0.5471,
+      "epoch": 1.6974789915966386,
+      "grad_norm": 1.819587589935648,
+      "learning_rate": 1.9352658478049085e-05,
+      "loss": 0.7831,
       "step": 101
     },
     {
-      "epoch": 1.2142857142857142,
-      "grad_norm": 1.7348261869056978,
-      "learning_rate": 1.9740516280559005e-05,
-      "loss": 0.5655,
+      "epoch": 1.7142857142857144,
+      "grad_norm": 6.083715872130075,
+      "learning_rate": 1.9336351214560648e-05,
+      "loss": 0.5828,
       "step": 102
     },
     {
-      "epoch": 1.2261904761904763,
-      "grad_norm": 5.178036354232653,
-      "learning_rate": 1.973318831173001e-05,
-      "loss": 0.6072,
+      "epoch": 1.73109243697479,
+      "grad_norm": 2.0455350806117827,
+      "learning_rate": 1.9319848144197543e-05,
+      "loss": 0.8423,
       "step": 103
     },
     {
-      "epoch": 1.2380952380952381,
-      "grad_norm": 1.6076518163637115,
-      "learning_rate": 1.972575970411955e-05,
-      "loss": 0.9794,
+      "epoch": 1.7478991596638656,
+      "grad_norm": 2.575532108553041,
+      "learning_rate": 1.9303149613070852e-05,
+      "loss": 0.8617,
       "step": 104
     },
     {
-      "epoch": 1.25,
-      "grad_norm": 2.723769398836049,
-      "learning_rate": 1.971823053453762e-05,
-      "loss": 1.3593,
+      "epoch": 1.7647058823529411,
+      "grad_norm": 1.698878909041017,
+      "learning_rate": 1.928625597139096e-05,
+      "loss": 0.7528,
       "step": 105
     },
     {
-      "epoch": 1.2619047619047619,
-      "grad_norm": 1.2127891225080338,
-      "learning_rate": 1.971060088083398e-05,
-      "loss": 0.4988,
+      "epoch": 1.7815126050420167,
+      "grad_norm": 4.480689810270073,
+      "learning_rate": 1.926916757346022e-05,
+      "loss": 0.4962,
       "step": 106
     },
     {
-      "epoch": 1.2738095238095237,
-      "grad_norm": 1.4952785968327083,
-      "learning_rate": 1.9702870821897385e-05,
-      "loss": 0.9151,
+      "epoch": 1.7983193277310925,
+      "grad_norm": 2.43491273709643,
+      "learning_rate": 1.9251884777665513e-05,
+      "loss": 0.6916,
       "step": 107
     },
     {
-      "epoch": 1.2857142857142856,
-      "grad_norm": 3.485596886623978,
-      "learning_rate": 1.9695040437654744e-05,
-      "loss": 0.5045,
+      "epoch": 1.815126050420168,
+      "grad_norm": 2.9374762528937945,
+      "learning_rate": 1.9234407946470735e-05,
+      "loss": 0.7218,
       "step": 108
     },
     {
-      "epoch": 1.2976190476190477,
-      "grad_norm": 6.271447145920052,
-      "learning_rate": 1.9687109809070314e-05,
-      "loss": 1.3797,
+      "epoch": 1.8319327731092439,
+      "grad_norm": 3.110996772971968,
+      "learning_rate": 1.9216737446409192e-05,
+      "loss": 0.3829,
       "step": 109
     },
     {
-      "epoch": 1.3095238095238095,
-      "grad_norm": 7.3279656814005385,
-      "learning_rate": 1.967907901814485e-05,
-      "loss": 0.1904,
+      "epoch": 1.8487394957983194,
+      "grad_norm": 1.963122392871237,
+      "learning_rate": 1.919887364807592e-05,
+      "loss": 0.7205,
       "step": 110
     },
     {
-      "epoch": 1.3214285714285714,
-      "grad_norm": 1.0931520962056926,
-      "learning_rate": 1.9670948147914758e-05,
-      "loss": 0.4405,
+      "epoch": 1.865546218487395,
+      "grad_norm": 3.066257574584323,
+      "learning_rate": 1.9180816926119903e-05,
+      "loss": 0.5874,
       "step": 111
     },
     {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 1.4586197259444877,
-      "learning_rate": 1.9662717282451248e-05,
-      "loss": 0.5086,
+      "epoch": 1.8823529411764706,
+      "grad_norm": 2.2609617726833946,
+      "learning_rate": 1.9162567659236227e-05,
+      "loss": 0.5984,
       "step": 112
     },
     {
-      "epoch": 1.3452380952380953,
-      "grad_norm": 1.5506740361035394,
-      "learning_rate": 1.965438650685945e-05,
-      "loss": 0.9044,
+      "epoch": 1.8991596638655461,
+      "grad_norm": 2.6110497038118727,
+      "learning_rate": 1.9144126230158127e-05,
+      "loss": 0.6068,
       "step": 113
     },
     {
-      "epoch": 1.3571428571428572,
-      "grad_norm": 1.943168328137838,
-      "learning_rate": 1.9645955907277544e-05,
-      "loss": 0.072,
+      "epoch": 1.9159663865546217,
+      "grad_norm": 2.2601235737134195,
+      "learning_rate": 1.9125493025648963e-05,
+      "loss": 0.6273,
       "step": 114
     },
     {
-      "epoch": 1.369047619047619,
-      "grad_norm": 5.388416018455866,
-      "learning_rate": 1.9637425570875857e-05,
-      "loss": 0.9327,
+      "epoch": 1.9327731092436975,
+      "grad_norm": 2.271778447040359,
+      "learning_rate": 1.9106668436494113e-05,
+      "loss": 0.6891,
       "step": 115
     },
     {
-      "epoch": 1.380952380952381,
-      "grad_norm": 6.028065865896953,
-      "learning_rate": 1.9628795585855986e-05,
-      "loss": 0.6004,
+      "epoch": 1.949579831932773,
+      "grad_norm": 2.0592194165628275,
+      "learning_rate": 1.908765285749278e-05,
+      "loss": 0.6283,
       "step": 116
     },
     {
-      "epoch": 1.3928571428571428,
-      "grad_norm": 2.3104045537885853,
-      "learning_rate": 1.9620066041449854e-05,
-      "loss": 1.2527,
+      "epoch": 1.9663865546218489,
+      "grad_norm": 2.096696308042481,
+      "learning_rate": 1.9068446687449698e-05,
+      "loss": 0.5497,
       "step": 117
     },
     {
-      "epoch": 1.4047619047619047,
-      "grad_norm": 3.8366407382177226,
-      "learning_rate": 1.961123702791881e-05,
-      "loss": 0.5362,
+      "epoch": 1.9831932773109244,
+      "grad_norm": 2.4126456796916744,
+      "learning_rate": 1.9049050329166778e-05,
+      "loss": 0.5112,
       "step": 118
     },
     {
-      "epoch": 1.4166666666666667,
-      "grad_norm": 1.6321334992158414,
-      "learning_rate": 1.9602308636552692e-05,
-      "loss": 0.9381,
+      "epoch": 2.0,
+      "grad_norm": 7.4662803136656954,
+      "learning_rate": 1.9029464189434663e-05,
+      "loss": 0.5049,
       "step": 119
     },
     {
-      "epoch": 1.4285714285714286,
-      "grad_norm": 2.6046556479389738,
-      "learning_rate": 1.9593280959668878e-05,
-      "loss": 0.4521,
+      "epoch": 2.0,
+      "eval_loss": 0.4023877680301666,
+      "eval_runtime": 94.9843,
+      "eval_samples_per_second": 0.99,
+      "eval_steps_per_second": 0.99,
+      "step": 119
+    },
+    {
+      "epoch": 2.0168067226890756,
+      "grad_norm": 2.736890471961608,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.4793,
       "step": 120
     },
     {
-      "epoch": 1.4404761904761905,
-      "grad_norm": 2.426572164115562,
-      "learning_rate": 1.9584154090611335e-05,
-      "loss": 0.4931,
+      "epoch": 2.033613445378151,
+      "grad_norm": 2.030226042772956,
+      "learning_rate": 1.8989724212677784e-05,
+      "loss": 0.4857,
       "step": 121
     },
     {
-      "epoch": 1.4523809523809523,
-      "grad_norm": 5.329850166878175,
-      "learning_rate": 1.957492812374965e-05,
-      "loss": 0.6056,
+      "epoch": 2.0504201680672267,
+      "grad_norm": 2.308935605416641,
+      "learning_rate": 1.8969571209100738e-05,
+      "loss": 0.3551,
       "step": 122
     },
     {
-      "epoch": 1.4642857142857144,
-      "grad_norm": 4.396192643131844,
-      "learning_rate": 1.956560315447806e-05,
-      "loss": 0.8952,
+      "epoch": 2.0672268907563027,
+      "grad_norm": 1.9661040290247784,
+      "learning_rate": 1.8949230090952463e-05,
+      "loss": 0.4291,
       "step": 123
     },
     {
-      "epoch": 1.4761904761904763,
-      "grad_norm": 5.802283102298066,
-      "learning_rate": 1.9556179279214464e-05,
-      "loss": 0.1045,
+      "epoch": 2.0840336134453783,
+      "grad_norm": 2.0802991306768854,
+      "learning_rate": 1.8928701284837597e-05,
+      "loss": 0.4988,
       "step": 124
     },
     {
-      "epoch": 1.4880952380952381,
-      "grad_norm": 1.7429659702829874,
-      "learning_rate": 1.9546656595399418e-05,
-      "loss": 0.1404,
+      "epoch": 2.100840336134454,
+      "grad_norm": 2.0446185078523804,
+      "learning_rate": 1.890798522129708e-05,
+      "loss": 0.3199,
       "step": 125
     },
     {
-      "epoch": 1.5,
-      "grad_norm": 4.908284880057786,
-      "learning_rate": 1.9537035201495144e-05,
-      "loss": 0.5325,
+      "epoch": 2.1176470588235294,
+      "grad_norm": 2.403214003507413,
+      "learning_rate": 1.8887082334799098e-05,
+      "loss": 0.4125,
       "step": 126
     },
     {
-      "epoch": 1.5119047619047619,
-      "grad_norm": 1.3604274167302775,
-      "learning_rate": 1.9527315196984488e-05,
-      "loss": 0.8657,
+      "epoch": 2.134453781512605,
+      "grad_norm": 1.7254339304792963,
+      "learning_rate": 1.8865993063730003e-05,
+      "loss": 0.3017,
       "step": 127
     },
     {
-      "epoch": 1.5238095238095237,
-      "grad_norm": 4.411139336383562,
-      "learning_rate": 1.9517496682369925e-05,
-      "loss": 0.4741,
+      "epoch": 2.1512605042016806,
+      "grad_norm": 2.3578101043917643,
+      "learning_rate": 1.884471785038509e-05,
+      "loss": 0.228,
       "step": 128
     },
     {
-      "epoch": 1.5357142857142856,
-      "grad_norm": 3.2777311634635056,
-      "learning_rate": 1.950757975917248e-05,
-      "loss": 0.0753,
+      "epoch": 2.168067226890756,
+      "grad_norm": 1.6260308846139646,
+      "learning_rate": 1.882325714095934e-05,
+      "loss": 0.1768,
       "step": 129
     },
     {
-      "epoch": 1.5476190476190477,
-      "grad_norm": 2.1472145073542084,
-      "learning_rate": 1.9497564529930716e-05,
-      "loss": 0.1772,
+      "epoch": 2.184873949579832,
+      "grad_norm": 1.9572713847794219,
+      "learning_rate": 1.8801611385538047e-05,
+      "loss": 0.3387,
       "step": 130
     },
     {
-      "epoch": 1.5595238095238095,
-      "grad_norm": 1.5922012254987603,
-      "learning_rate": 1.9487451098199642e-05,
-      "loss": 0.921,
+      "epoch": 2.2016806722689077,
+      "grad_norm": 2.0610001563380127,
+      "learning_rate": 1.8779781038087406e-05,
+      "loss": 0.3104,
       "step": 131
     },
     {
-      "epoch": 1.5714285714285714,
-      "grad_norm": 2.6652448912819633,
-      "learning_rate": 1.9477239568549667e-05,
-      "loss": 0.8894,
+      "epoch": 2.2184873949579833,
+      "grad_norm": 2.770005426395026,
+      "learning_rate": 1.875776655644495e-05,
+      "loss": 0.3574,
       "step": 132
     },
     {
-      "epoch": 1.5833333333333335,
-      "grad_norm": 7.012220577504933,
-      "learning_rate": 1.9466930046565504e-05,
-      "loss": 0.5622,
+      "epoch": 2.235294117647059,
+      "grad_norm": 3.4273616297199796,
+      "learning_rate": 1.8735568402309987e-05,
+      "loss": 0.3576,
       "step": 133
     },
     {
-      "epoch": 1.5952380952380953,
-      "grad_norm": 14.136593122149362,
-      "learning_rate": 1.9456522638845082e-05,
-      "loss": 0.2466,
+      "epoch": 2.2521008403361344,
+      "grad_norm": 2.0797522448134242,
+      "learning_rate": 1.8713187041233896e-05,
+      "loss": 0.289,
       "step": 134
     },
     {
-      "epoch": 1.6071428571428572,
-      "grad_norm": 4.579182399679084,
-      "learning_rate": 1.9446017452998448e-05,
-      "loss": 0.0872,
+      "epoch": 2.26890756302521,
+      "grad_norm": 5.8224416033882065,
+      "learning_rate": 1.8690622942610367e-05,
+      "loss": 0.231,
       "step": 135
     },
     {
-      "epoch": 1.619047619047619,
-      "grad_norm": 5.159279776844667,
-      "learning_rate": 1.943541459764665e-05,
-      "loss": 0.0711,
+      "epoch": 2.2857142857142856,
+      "grad_norm": 1.805558912475057,
+      "learning_rate": 1.8667876579665556e-05,
+      "loss": 0.2549,
       "step": 136
     },
     {
-      "epoch": 1.630952380952381,
-      "grad_norm": 1.6198531580129043,
-      "learning_rate": 1.9424714182420606e-05,
-      "loss": 0.9309,
+      "epoch": 2.302521008403361,
+      "grad_norm": 2.1386013924513407,
+      "learning_rate": 1.8644948429448174e-05,
+      "loss": 0.3056,
       "step": 137
     },
     {
-      "epoch": 1.6428571428571428,
-      "grad_norm": 1.91584402163816,
-      "learning_rate": 1.9413916317959994e-05,
-      "loss": 0.1586,
+      "epoch": 2.3193277310924367,
+      "grad_norm": 2.891103001226989,
+      "learning_rate": 1.862183897281946e-05,
+      "loss": 0.2917,
       "step": 138
     },
     {
-      "epoch": 1.6547619047619047,
-      "grad_norm": 1.5986601917930672,
-      "learning_rate": 1.9403021115912083e-05,
-      "loss": 0.7846,
+      "epoch": 2.3361344537815127,
+      "grad_norm": 1.6366849032923414,
+      "learning_rate": 1.8598548694443102e-05,
+      "loss": 0.185,
       "step": 139
     },
     {
-      "epoch": 1.6666666666666665,
-      "grad_norm": 12.580007166459941,
-      "learning_rate": 1.93920286889306e-05,
-      "loss": 0.7669,
+      "epoch": 2.3529411764705883,
+      "grad_norm": 3.3075808645565825,
+      "learning_rate": 1.8575078082775096e-05,
+      "loss": 0.272,
       "step": 140
     },
     {
-      "epoch": 1.6785714285714286,
-      "grad_norm": 1.4187041663248747,
-      "learning_rate": 1.938093915067453e-05,
-      "loss": 0.8152,
+      "epoch": 2.369747899159664,
+      "grad_norm": 2.271589816301807,
+      "learning_rate": 1.8551427630053464e-05,
+      "loss": 0.245,
       "step": 141
     },
     {
-      "epoch": 1.6904761904761905,
-      "grad_norm": 5.695537042553588,
-      "learning_rate": 1.936975261580699e-05,
-      "loss": 1.0575,
+      "epoch": 2.3865546218487395,
+      "grad_norm": 1.84180801145786,
+      "learning_rate": 1.8527597832287954e-05,
+      "loss": 0.2091,
       "step": 142
     },
     {
-      "epoch": 1.7023809523809523,
-      "grad_norm": 6.521896789730948,
-      "learning_rate": 1.9358469199994006e-05,
-      "loss": 0.5823,
+      "epoch": 2.403361344537815,
+      "grad_norm": 3.766252312196367,
+      "learning_rate": 1.8503589189249637e-05,
+      "loss": 0.2895,
       "step": 143
     },
     {
-      "epoch": 1.7142857142857144,
-      "grad_norm": 3.577556847930223,
-      "learning_rate": 1.9347089019903333e-05,
-      "loss": 0.8156,
+      "epoch": 2.4201680672268906,
+      "grad_norm": 1.873377028811922,
+      "learning_rate": 1.847940220446042e-05,
+      "loss": 0.2368,
       "step": 144
     },
     {
-      "epoch": 1.7261904761904763,
-      "grad_norm": 8.100269368760127,
-      "learning_rate": 1.9335612193203243e-05,
-      "loss": 0.1723,
+      "epoch": 2.4369747899159666,
+      "grad_norm": 4.545298766002122,
+      "learning_rate": 1.845503738518249e-05,
+      "loss": 0.1958,
       "step": 145
     },
     {
-      "epoch": 1.7380952380952381,
-      "grad_norm": 2.6780547969473156,
-      "learning_rate": 1.9324038838561317e-05,
-      "loss": 1.2078,
+      "epoch": 2.453781512605042,
+      "grad_norm": 3.1348145302552903,
+      "learning_rate": 1.843049524240766e-05,
+      "loss": 0.2098,
       "step": 146
     },
     {
-      "epoch": 1.75,
-      "grad_norm": 1.8870873473190792,
-      "learning_rate": 1.9312369075643197e-05,
-      "loss": 0.541,
+      "epoch": 2.4705882352941178,
+      "grad_norm": 1.6186560450788932,
+      "learning_rate": 1.8405776290846672e-05,
+      "loss": 0.1754,
       "step": 147
     },
     {
-      "epoch": 1.7619047619047619,
-      "grad_norm": 3.457663458870956,
-      "learning_rate": 1.9300603025111374e-05,
-      "loss": 0.8025,
+      "epoch": 2.4873949579831933,
+      "grad_norm": 3.382962591579758,
+      "learning_rate": 1.8380881048918406e-05,
+      "loss": 0.2035,
       "step": 148
     },
     {
-      "epoch": 1.7738095238095237,
-      "grad_norm": 2.63314035325873,
-      "learning_rate": 1.9288740808623923e-05,
-      "loss": 0.8417,
+      "epoch": 2.504201680672269,
+      "grad_norm": 2.623712266453748,
+      "learning_rate": 1.8355810038738986e-05,
+      "loss": 0.2092,
       "step": 149
     },
     {
-      "epoch": 1.7857142857142856,
-      "grad_norm": 4.406400740021301,
-      "learning_rate": 1.9276782548833262e-05,
-      "loss": 0.5055,
+      "epoch": 2.5210084033613445,
+      "grad_norm": 2.2054319523334676,
+      "learning_rate": 1.8330563786110837e-05,
+      "loss": 0.1898,
       "step": 150
     },
     {
-      "epoch": 1.7976190476190477,
-      "grad_norm": 2.400162986097046,
-      "learning_rate": 1.9264728369384867e-05,
-      "loss": 0.4728,
+      "epoch": 2.53781512605042,
+      "grad_norm": 4.2861344256493785,
+      "learning_rate": 1.830514282051166e-05,
+      "loss": 0.1886,
       "step": 151
     },
     {
-      "epoch": 1.8095238095238095,
-      "grad_norm": 2.097382923710725,
-      "learning_rate": 1.925257839491599e-05,
-      "loss": 0.8526,
+      "epoch": 2.5546218487394956,
+      "grad_norm": 2.651873913211827,
+      "learning_rate": 1.8279547675083343e-05,
+      "loss": 0.1824,
       "step": 152
     },
     {
-      "epoch": 1.8214285714285714,
-      "grad_norm": 3.554097093286924,
-      "learning_rate": 1.9240332751054397e-05,
-      "loss": 1.2327,
+      "epoch": 2.571428571428571,
+      "grad_norm": 1.8547906501685159,
+      "learning_rate": 1.8253778886620754e-05,
+      "loss": 0.1705,
       "step": 153
     },
     {
-      "epoch": 1.8333333333333335,
-      "grad_norm": 1.596640390294483,
-      "learning_rate": 1.9227991564417038e-05,
-      "loss": 0.73,
+      "epoch": 2.588235294117647,
+      "grad_norm": 3.9635058525876277,
+      "learning_rate": 1.822783699556049e-05,
+      "loss": 0.1838,
       "step": 154
     },
     {
-      "epoch": 1.8452380952380953,
-      "grad_norm": 7.712486381371494,
-      "learning_rate": 1.9215554962608757e-05,
-      "loss": 0.6713,
+      "epoch": 2.6050420168067228,
+      "grad_norm": 2.3948916184720606,
+      "learning_rate": 1.820172254596956e-05,
+      "loss": 0.1778,
       "step": 155
     },
     {
-      "epoch": 1.8571428571428572,
-      "grad_norm": 0.9406036356596837,
-      "learning_rate": 1.9203023074220976e-05,
-      "loss": 0.4108,
+      "epoch": 2.6218487394957983,
+      "grad_norm": 2.3071047956425264,
+      "learning_rate": 1.817543608553395e-05,
+      "loss": 0.1708,
       "step": 156
     },
     {
-      "epoch": 1.869047619047619,
-      "grad_norm": 1.2783324468370174,
-      "learning_rate": 1.919039602883035e-05,
-      "loss": 0.7852,
+      "epoch": 2.638655462184874,
+      "grad_norm": 2.108074459212226,
+      "learning_rate": 1.814897816554715e-05,
+      "loss": 0.1749,
       "step": 157
     },
     {
-      "epoch": 1.880952380952381,
-      "grad_norm": 1.9809209865395272,
-      "learning_rate": 1.917767395699742e-05,
-      "loss": 0.8275,
+      "epoch": 2.6554621848739495,
+      "grad_norm": 1.0692114014911545,
+      "learning_rate": 1.8122349340898596e-05,
+      "loss": 0.1373,
       "step": 158
     },
     {
-      "epoch": 1.8928571428571428,
-      "grad_norm": 3.256407393144481,
-      "learning_rate": 1.9164856990265304e-05,
-      "loss": 0.4972,
+      "epoch": 2.6722689075630255,
+      "grad_norm": 2.0009322090894592,
+      "learning_rate": 1.809555017006202e-05,
+      "loss": 0.1709,
       "step": 159
     },
     {
-      "epoch": 1.9047619047619047,
-      "grad_norm": 1.7699190444451984,
-      "learning_rate": 1.9151945261158295e-05,
-      "loss": 0.7818,
+      "epoch": 2.689075630252101,
+      "grad_norm": 2.1909981915140375,
+      "learning_rate": 1.8068581215083752e-05,
+      "loss": 0.1484,
       "step": 160
     },
     {
-      "epoch": 1.9166666666666665,
-      "grad_norm": 2.215298903394841,
-      "learning_rate": 1.9138938903180496e-05,
-      "loss": 1.2303,
+      "epoch": 2.7058823529411766,
+      "grad_norm": 2.103149621290993,
+      "learning_rate": 1.804144304157091e-05,
+      "loss": 0.1482,
       "step": 161
     },
     {
-      "epoch": 1.9285714285714286,
-      "grad_norm": 1.8042937103793464,
-      "learning_rate": 1.9125838050814472e-05,
-      "loss": 0.3921,
+      "epoch": 2.722689075630252,
+      "grad_norm": 1.5470172068234862,
+      "learning_rate": 1.8014136218679566e-05,
+      "loss": 0.1489,
       "step": 162
     },
     {
-      "epoch": 1.9404761904761905,
-      "grad_norm": 2.0556006635467066,
-      "learning_rate": 1.911264283951982e-05,
-      "loss": 1.1463,
+      "epoch": 2.7394957983193278,
+      "grad_norm": 1.8588243552375616,
+      "learning_rate": 1.7986661319102795e-05,
+      "loss": 0.1526,
       "step": 163
     },
     {
-      "epoch": 1.9523809523809523,
-      "grad_norm": 2.05906081878954,
-      "learning_rate": 1.9099353405731794e-05,
-      "loss": 0.7994,
+      "epoch": 2.7563025210084033,
+      "grad_norm": 1.1673797887420725,
+      "learning_rate": 1.7959018919058654e-05,
+      "loss": 0.1361,
       "step": 164
     },
     {
-      "epoch": 1.9642857142857144,
-      "grad_norm": 2.8775318413447897,
-      "learning_rate": 1.908596988685988e-05,
-      "loss": 0.0429,
+      "epoch": 2.773109243697479,
+      "grad_norm": 0.9871009564387679,
+      "learning_rate": 1.7931209598278117e-05,
+      "loss": 0.1168,
       "step": 165
     },
     {
-      "epoch": 1.9761904761904763,
-      "grad_norm": 2.19883503502942,
-      "learning_rate": 1.9072492421286384e-05,
-      "loss": 0.7907,
+      "epoch": 2.7899159663865545,
+      "grad_norm": 1.375240831135664,
+      "learning_rate": 1.7903233939992904e-05,
+      "loss": 0.1271,
       "step": 166
     },
     {
-      "epoch": 1.9880952380952381,
-      "grad_norm": 1.5000744356797195,
-      "learning_rate": 1.9058921148364996e-05,
-      "loss": 0.3933,
+      "epoch": 2.80672268907563,
+      "grad_norm": 1.1795213655735046,
+      "learning_rate": 1.787509253092326e-05,
+      "loss": 0.1092,
       "step": 167
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 2.211057744430661,
-      "learning_rate": 1.904525620841935e-05,
-      "loss": 1.1364,
-      "step": 168
-    },
-    {
-      "epoch": 2.0,
-      "eval_loss": 0.6083582043647766,
-      "eval_runtime": 38.5831,
-      "eval_samples_per_second": 1.555,
-      "eval_steps_per_second": 1.555,
+      "epoch": 2.8235294117647056,
+      "grad_norm": 3.071368428139141,
+      "learning_rate": 1.784678596126563e-05,
+      "loss": 0.286,
       "step": 168
     },
     {
-      "epoch": 2.011904761904762,
-      "grad_norm": 1.2506294363329695,
-      "learning_rate": 1.9031497742741573e-05,
-      "loss": 0.3685,
+      "epoch": 2.8403361344537816,
+      "grad_norm": 1.0579777160189892,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.1197,
       "step": 169
     },
     {
-      "epoch": 2.0238095238095237,
-      "grad_norm": 2.1294740967375776,
-      "learning_rate": 1.9017645893590834e-05,
-      "loss": 0.4106,
+      "epoch": 2.857142857142857,
+      "grad_norm": 1.6336798015628837,
+      "learning_rate": 1.7789679718278944e-05,
+      "loss": 0.1282,
       "step": 170
     },
     {
-      "epoch": 2.0357142857142856,
-      "grad_norm": 2.195919937729915,
-      "learning_rate": 1.900370080419186e-05,
-      "loss": 1.0409,
+      "epoch": 2.8739495798319328,
+      "grad_norm": 1.8315662561658703,
+      "learning_rate": 1.7760881242612096e-05,
+      "loss": 0.1429,
       "step": 171
     },
     {
-      "epoch": 2.0476190476190474,
-      "grad_norm": 2.137987938056288,
-      "learning_rate": 1.8989662618733446e-05,
-      "loss": 1.0374,
+      "epoch": 2.8907563025210083,
+      "grad_norm": 2.9860725552192404,
+      "learning_rate": 1.773192000165655e-05,
+      "loss": 0.2833,
       "step": 172
     },
     {
-      "epoch": 2.0595238095238093,
-      "grad_norm": 2.321052769415701,
-      "learning_rate": 1.8975531482366998e-05,
-      "loss": 1.0739,
+      "epoch": 2.907563025210084,
+      "grad_norm": 1.04161581184864,
+      "learning_rate": 1.7702796602802705e-05,
+      "loss": 0.1042,
       "step": 173
     },
     {
-      "epoch": 2.0714285714285716,
-      "grad_norm": 4.17435976154088,
-      "learning_rate": 1.8961307541205003e-05,
-      "loss": 0.7877,
+      "epoch": 2.92436974789916,
+      "grad_norm": 1.2933586996454134,
+      "learning_rate": 1.7673511656841822e-05,
+      "loss": 0.1382,
       "step": 174
     },
     {
-      "epoch": 2.0833333333333335,
-      "grad_norm": 1.389768088500958,
-      "learning_rate": 1.8946990942319518e-05,
-      "loss": 0.7138,
+      "epoch": 2.9411764705882355,
+      "grad_norm": 1.8373790721021168,
+      "learning_rate": 1.7644065777953206e-05,
+      "loss": 0.2055,
       "step": 175
     },
     {
-      "epoch": 2.0952380952380953,
-      "grad_norm": 2.506177283623637,
-      "learning_rate": 1.8932581833740676e-05,
-      "loss": 0.3942,
+      "epoch": 2.957983193277311,
+      "grad_norm": 1.695542614817302,
+      "learning_rate": 1.7614459583691346e-05,
+      "loss": 0.144,
       "step": 176
     },
     {
-      "epoch": 2.107142857142857,
-      "grad_norm": 3.325446937404553,
-      "learning_rate": 1.8918080364455122e-05,
-      "loss": 0.7675,
+      "epoch": 2.9747899159663866,
+      "grad_norm": 1.2079374495539366,
+      "learning_rate": 1.758469369497293e-05,
+      "loss": 0.1286,
       "step": 177
     },
     {
-      "epoch": 2.119047619047619,
-      "grad_norm": 2.873764603659821,
-      "learning_rate": 1.8903486684404496e-05,
-      "loss": 0.4016,
+      "epoch": 2.991596638655462,
+      "grad_norm": 7.433145063552697,
+      "learning_rate": 1.7554768736063858e-05,
+      "loss": 0.2421,
       "step": 178
     },
     {
-      "epoch": 2.130952380952381,
-      "grad_norm": 2.081810368542138,
-      "learning_rate": 1.888880094448387e-05,
-      "loss": 0.9637,
+      "epoch": 2.991596638655462,
+      "eval_loss": 0.18344487249851227,
+      "eval_runtime": 97.2976,
+      "eval_samples_per_second": 0.966,
+      "eval_steps_per_second": 0.966,
+      "step": 178
+    },
+    {
+      "epoch": 3.008403361344538,
+      "grad_norm": 1.936304205658518,
+      "learning_rate": 1.7524685334566126e-05,
+      "loss": 0.0924,
       "step": 179
     },
     {
-      "epoch": 2.142857142857143,
-      "grad_norm": 4.506719205876918,
-      "learning_rate": 1.88740232965402e-05,
-      "loss": 0.1205,
+      "epoch": 3.0252100840336134,
+      "grad_norm": 3.65130263061498,
+      "learning_rate": 1.7494444121404673e-05,
+      "loss": 0.168,
       "step": 180
     },
     {
-      "epoch": 2.1547619047619047,
-      "grad_norm": 3.840347092360756,
-      "learning_rate": 1.8859153893370737e-05,
-      "loss": 0.0537,
+      "epoch": 3.042016806722689,
+      "grad_norm": 6.416735934730296,
+      "learning_rate": 1.746404573081415e-05,
+      "loss": 0.1856,
       "step": 181
     },
     {
-      "epoch": 2.1666666666666665,
-      "grad_norm": 2.620898271441886,
-      "learning_rate": 1.8844192888721473e-05,
-      "loss": 0.686,
+      "epoch": 3.0588235294117645,
+      "grad_norm": 2.2813578076866623,
+      "learning_rate": 1.7433490800325614e-05,
+      "loss": 0.1393,
       "step": 182
     },
     {
-      "epoch": 2.1785714285714284,
-      "grad_norm": 3.1614575918085066,
-      "learning_rate": 1.8829140437285525e-05,
-      "loss": 0.6877,
+      "epoch": 3.0756302521008405,
+      "grad_norm": 2.595639669498612,
+      "learning_rate": 1.7402779970753156e-05,
+      "loss": 0.1428,
       "step": 183
     },
     {
-      "epoch": 2.1904761904761907,
-      "grad_norm": 3.357401090981342,
-      "learning_rate": 1.8813996694701548e-05,
-      "loss": 0.106,
+      "epoch": 3.092436974789916,
+      "grad_norm": 1.767980230153596,
+      "learning_rate": 1.7371913886180473e-05,
+      "loss": 0.1331,
       "step": 184
     },
     {
-      "epoch": 2.2023809523809526,
-      "grad_norm": 2.641859959945962,
-      "learning_rate": 1.8798761817552135e-05,
-      "loss": 0.1825,
+      "epoch": 3.1092436974789917,
+      "grad_norm": 2.5213252449194252,
+      "learning_rate": 1.7340893193947342e-05,
+      "loss": 0.1252,
       "step": 185
     },
     {
-      "epoch": 2.2142857142857144,
-      "grad_norm": 1.3075201309355102,
-      "learning_rate": 1.8783435963362178e-05,
-      "loss": 0.6945,
+      "epoch": 3.1260504201680672,
+      "grad_norm": 2.1170186194258167,
+      "learning_rate": 1.7309718544636057e-05,
+      "loss": 0.1108,
       "step": 186
     },
     {
-      "epoch": 2.2261904761904763,
-      "grad_norm": 1.7495575930553038,
-      "learning_rate": 1.8768019290597254e-05,
-      "loss": 0.4308,
+      "epoch": 3.142857142857143,
+      "grad_norm": 2.0268322211129592,
+      "learning_rate": 1.7278390592057785e-05,
+      "loss": 0.1561,
       "step": 187
     },
     {
-      "epoch": 2.238095238095238,
-      "grad_norm": 3.735783992873633,
-      "learning_rate": 1.8752511958661977e-05,
-      "loss": 0.0707,
+      "epoch": 3.1596638655462184,
+      "grad_norm": 1.7261621681645014,
+      "learning_rate": 1.7246909993238844e-05,
+      "loss": 0.1246,
       "step": 188
     },
     {
-      "epoch": 2.25,
-      "grad_norm": 4.3626751100504455,
-      "learning_rate": 1.873691412789836e-05,
-      "loss": 0.6226,
+      "epoch": 3.176470588235294,
+      "grad_norm": 1.3950437881445872,
+      "learning_rate": 1.7215277408406932e-05,
+      "loss": 0.0998,
       "step": 189
     },
     {
-      "epoch": 2.261904761904762,
-      "grad_norm": 3.030076532866424,
-      "learning_rate": 1.8721225959584156e-05,
-      "loss": 0.3429,
+      "epoch": 3.19327731092437,
+      "grad_norm": 1.793336566476121,
+      "learning_rate": 1.7183493500977277e-05,
+      "loss": 0.128,
       "step": 190
     },
     {
-      "epoch": 2.2738095238095237,
-      "grad_norm": 4.25577167516674,
-      "learning_rate": 1.8705447615931172e-05,
-      "loss": 0.6827,
+      "epoch": 3.2100840336134455,
+      "grad_norm": 1.35402748785192,
+      "learning_rate": 1.7151558937538725e-05,
+      "loss": 0.0979,
       "step": 191
     },
     {
-      "epoch": 2.2857142857142856,
-      "grad_norm": 2.623136916804772,
-      "learning_rate": 1.8689579260083622e-05,
-      "loss": 1.0001,
+      "epoch": 3.226890756302521,
+      "grad_norm": 1.9875540888658763,
+      "learning_rate": 1.7119474387839764e-05,
+      "loss": 0.1477,
       "step": 192
     },
     {
-      "epoch": 2.2976190476190474,
-      "grad_norm": 4.114146115626149,
-      "learning_rate": 1.8673621056116405e-05,
-      "loss": 0.0817,
+      "epoch": 3.2436974789915967,
+      "grad_norm": 3.4010639220257755,
+      "learning_rate": 1.708724052477446e-05,
+      "loss": 0.2346,
       "step": 193
     },
     {
-      "epoch": 2.3095238095238093,
-      "grad_norm": 3.6811417292623867,
-      "learning_rate": 1.865757316903345e-05,
-      "loss": 0.4183,
+      "epoch": 3.2605042016806722,
+      "grad_norm": 1.1526809505800015,
+      "learning_rate": 1.7054858024368365e-05,
+      "loss": 0.1189,
       "step": 194
     },
     {
-      "epoch": 2.3214285714285716,
-      "grad_norm": 2.8700111469727236,
-      "learning_rate": 1.8641435764765966e-05,
-      "loss": 0.6386,
+      "epoch": 3.277310924369748,
+      "grad_norm": 1.8401498885713787,
+      "learning_rate": 1.7022327565764336e-05,
+      "loss": 0.1321,
       "step": 195
     },
     {
-      "epoch": 2.3333333333333335,
-      "grad_norm": 2.147973753973312,
-      "learning_rate": 1.8625209010170766e-05,
-      "loss": 0.7956,
+      "epoch": 3.2941176470588234,
+      "grad_norm": 1.823800189990081,
+      "learning_rate": 1.6989649831208286e-05,
+      "loss": 0.1276,
       "step": 196
     },
     {
-      "epoch": 2.3452380952380953,
-      "grad_norm": 2.1147626975453884,
-      "learning_rate": 1.8608893073028506e-05,
-      "loss": 0.0699,
+      "epoch": 3.310924369747899,
+      "grad_norm": 2.177879884931455,
+      "learning_rate": 1.6956825506034866e-05,
+      "loss": 0.1367,
       "step": 197
     },
     {
-      "epoch": 2.357142857142857,
-      "grad_norm": 4.759143787399854,
-      "learning_rate": 1.8592488122041987e-05,
-      "loss": 0.3981,
+      "epoch": 3.327731092436975,
+      "grad_norm": 4.377960471738762,
+      "learning_rate": 1.6923855278653114e-05,
+      "loss": 0.177,
       "step": 198
     },
     {
-      "epoch": 2.369047619047619,
-      "grad_norm": 3.446773840305837,
-      "learning_rate": 1.8575994326834378e-05,
-      "loss": 0.0768,
+      "epoch": 3.3445378151260505,
+      "grad_norm": 1.530590314148621,
+      "learning_rate": 1.6890739840532004e-05,
+      "loss": 0.1344,
       "step": 199
     },
     {
-      "epoch": 2.380952380952381,
-      "grad_norm": 1.7576609471238682,
-      "learning_rate": 1.8559411857947472e-05,
-      "loss": 0.6315,
+      "epoch": 3.361344537815126,
+      "grad_norm": 1.3469335679962628,
+      "learning_rate": 1.6857479886185942e-05,
+      "loss": 0.1052,
       "step": 200
     },
     {
-      "epoch": 2.392857142857143,
-      "grad_norm": 3.70255268224624,
-      "learning_rate": 1.8542740886839942e-05,
-      "loss": 0.4377,
+      "epoch": 3.3781512605042017,
+      "grad_norm": 1.6139041164207482,
+      "learning_rate": 1.682407611316021e-05,
+      "loss": 0.1475,
       "step": 201
     },
     {
-      "epoch": 2.4047619047619047,
-      "grad_norm": 4.380222821407496,
-      "learning_rate": 1.8525981585885537e-05,
-      "loss": 0.703,
+      "epoch": 3.3949579831932772,
+      "grad_norm": 1.6495915167747617,
+      "learning_rate": 1.6790529222016328e-05,
+      "loss": 0.0972,
       "step": 202
     },
     {
-      "epoch": 2.4166666666666665,
-      "grad_norm": 4.357939060601177,
-      "learning_rate": 1.8509134128371325e-05,
-      "loss": 0.8375,
+      "epoch": 3.411764705882353,
+      "grad_norm": 1.8398639057645052,
+      "learning_rate": 1.6756839916317358e-05,
+      "loss": 0.1084,
       "step": 203
     },
     {
-      "epoch": 2.4285714285714284,
-      "grad_norm": 5.09156824997242,
-      "learning_rate": 1.8492198688495884e-05,
-      "loss": 0.483,
+      "epoch": 3.4285714285714284,
+      "grad_norm": 2.0107317404674236,
+      "learning_rate": 1.672300890261317e-05,
+      "loss": 0.1102,
       "step": 204
     },
     {
-      "epoch": 2.4404761904761907,
-      "grad_norm": 1.2517867091759434,
-      "learning_rate": 1.8475175441367514e-05,
-      "loss": 0.5436,
+      "epoch": 3.4453781512605044,
+      "grad_norm": 1.9565606900612817,
+      "learning_rate": 1.6689036890425596e-05,
+      "loss": 0.1139,
       "step": 205
     },
     {
-      "epoch": 2.4523809523809526,
-      "grad_norm": 2.2431411923712408,
-      "learning_rate": 1.8458064563002417e-05,
-      "loss": 0.6378,
+      "epoch": 3.46218487394958,
+      "grad_norm": 1.639534786311468,
+      "learning_rate": 1.665492459223357e-05,
+      "loss": 0.0918,
       "step": 206
     },
     {
-      "epoch": 2.4642857142857144,
-      "grad_norm": 1.1954820415052247,
-      "learning_rate": 1.8440866230322877e-05,
-      "loss": 0.2742,
+      "epoch": 3.4789915966386555,
+      "grad_norm": 5.45699346508005,
+      "learning_rate": 1.6620672723458167e-05,
+      "loss": 0.1591,
       "step": 207
     },
     {
-      "epoch": 2.4761904761904763,
-      "grad_norm": 2.9995391650371923,
-      "learning_rate": 1.8423580621155447e-05,
-      "loss": 0.0413,
+      "epoch": 3.495798319327731,
+      "grad_norm": 2.0885082670769965,
+      "learning_rate": 1.658628200244763e-05,
+      "loss": 0.1118,
       "step": 208
     },
     {
-      "epoch": 2.488095238095238,
-      "grad_norm": 1.2324176262010005,
-      "learning_rate": 1.8406207914229083e-05,
-      "loss": 0.5359,
+      "epoch": 3.5126050420168067,
+      "grad_norm": 1.972708418564533,
+      "learning_rate": 1.6551753150462258e-05,
+      "loss": 0.1533,
       "step": 209
     },
     {
-      "epoch": 2.5,
-      "grad_norm": 2.299451434662873,
-      "learning_rate": 1.8388748289173315e-05,
-      "loss": 0.7862,
+      "epoch": 3.5294117647058822,
+      "grad_norm": 1.3414901525438894,
+      "learning_rate": 1.6517086891659335e-05,
+      "loss": 0.0973,
       "step": 210
     },
     {
-      "epoch": 2.511904761904762,
-      "grad_norm": 2.3447443905706438,
-      "learning_rate": 1.837120192651639e-05,
-      "loss": 0.7311,
+      "epoch": 3.546218487394958,
+      "grad_norm": 2.260816117595264,
+      "learning_rate": 1.6482283953077887e-05,
+      "loss": 0.1141,
       "step": 211
     },
     {
-      "epoch": 2.5238095238095237,
-      "grad_norm": 1.4847792611991646,
-      "learning_rate": 1.8353569007683397e-05,
-      "loss": 0.5814,
+      "epoch": 3.5630252100840334,
+      "grad_norm": 1.3728281964965443,
+      "learning_rate": 1.644734506462347e-05,
+      "loss": 0.1069,
       "step": 212
     },
     {
-      "epoch": 2.5357142857142856,
-      "grad_norm": 1.5842760285101158,
-      "learning_rate": 1.8335849714994394e-05,
-      "loss": 0.6283,
+      "epoch": 3.5798319327731094,
+      "grad_norm": 1.647487452421848,
+      "learning_rate": 1.641227095905286e-05,
+      "loss": 0.1317,
       "step": 213
     },
     {
-      "epoch": 2.5476190476190474,
-      "grad_norm": 1.8493596748650025,
-      "learning_rate": 1.8318044231662525e-05,
-      "loss": 0.3096,
+      "epoch": 3.596638655462185,
+      "grad_norm": 1.555286447904776,
+      "learning_rate": 1.637706237195867e-05,
+      "loss": 0.0997,
       "step": 214
     },
     {
-      "epoch": 2.5595238095238093,
-      "grad_norm": 2.0830391218860442,
-      "learning_rate": 1.8300152741792122e-05,
-      "loss": 0.7391,
+      "epoch": 3.6134453781512605,
+      "grad_norm": 1.4295747968285732,
+      "learning_rate": 1.6341720041753924e-05,
+      "loss": 0.1116,
       "step": 215
     },
     {
-      "epoch": 2.571428571428571,
-      "grad_norm": 1.532373441358143,
-      "learning_rate": 1.8282175430376804e-05,
-      "loss": 0.5466,
+      "epoch": 3.630252100840336,
+      "grad_norm": 1.4982199659580737,
+      "learning_rate": 1.6306244709656597e-05,
+      "loss": 0.1014,
       "step": 216
     },
     {
-      "epoch": 2.5833333333333335,
-      "grad_norm": 5.731342011284239,
-      "learning_rate": 1.826411248329757e-05,
-      "loss": 0.6475,
+      "epoch": 3.6470588235294117,
+      "grad_norm": 3.3567838619912513,
+      "learning_rate": 1.6270637119674023e-05,
+      "loss": 0.1758,
       "step": 217
     },
     {
-      "epoch": 2.5952380952380953,
-      "grad_norm": 12.481835165886968,
-      "learning_rate": 1.8245964087320856e-05,
-      "loss": 0.2665,
+      "epoch": 3.6638655462184873,
+      "grad_norm": 1.549085403651707,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.0874,
       "step": 218
     },
     {
-      "epoch": 2.607142857142857,
-      "grad_norm": 3.513207130809279,
-      "learning_rate": 1.822773043009664e-05,
-      "loss": 0.3703,
+      "epoch": 3.6806722689075633,
+      "grad_norm": 2.1336886853349295,
+      "learning_rate": 1.6199028155935793e-05,
+      "loss": 0.1467,
       "step": 219
     },
     {
-      "epoch": 2.619047619047619,
-      "grad_norm": 7.210938305978828,
-      "learning_rate": 1.8209411700156457e-05,
-      "loss": 0.3699,
+      "epoch": 3.697478991596639,
+      "grad_norm": 2.306782696878204,
+      "learning_rate": 1.6163028284001034e-05,
+      "loss": 0.1186,
       "step": 220
     },
     {
-      "epoch": 2.630952380952381,
-      "grad_norm": 4.21198369122006,
-      "learning_rate": 1.8191008086911494e-05,
-      "loss": 0.3044,
+      "epoch": 3.7142857142857144,
+      "grad_norm": 1.520623845541843,
+      "learning_rate": 1.612689915779134e-05,
+      "loss": 0.0923,
       "step": 221
     },
     {
-      "epoch": 2.642857142857143,
-      "grad_norm": 8.206928947078168,
-      "learning_rate": 1.8172519780650594e-05,
-      "loss": 0.406,
+      "epoch": 3.73109243697479,
+      "grad_norm": 1.147756639869472,
+      "learning_rate": 1.6090641535025773e-05,
+      "loss": 0.0854,
       "step": 222
     },
     {
-      "epoch": 2.6547619047619047,
-      "grad_norm": 1.4773204257836823,
-      "learning_rate": 1.8153946972538315e-05,
-      "loss": 0.4871,
+      "epoch": 3.7478991596638656,
+      "grad_norm": 1.2212881766323158,
+      "learning_rate": 1.605425617611829e-05,
+      "loss": 0.0868,
       "step": 223
     },
     {
-      "epoch": 2.6666666666666665,
-      "grad_norm": 2.461150337006145,
-      "learning_rate": 1.8135289854612942e-05,
-      "loss": 0.6975,
+      "epoch": 3.764705882352941,
+      "grad_norm": 1.374752115412969,
+      "learning_rate": 1.6017743844161802e-05,
+      "loss": 0.0857,
       "step": 224
     },
     {
-      "epoch": 2.678571428571429,
-      "grad_norm": 1.5762960515761975,
-      "learning_rate": 1.8116548619784497e-05,
-      "loss": 0.4559,
+      "epoch": 3.7815126050420167,
+      "grad_norm": 1.0655866618063339,
+      "learning_rate": 1.598110530491216e-05,
+      "loss": 0.0915,
       "step": 225
     },
     {
-      "epoch": 2.6904761904761907,
-      "grad_norm": 8.566362905559712,
-      "learning_rate": 1.8097723461832762e-05,
-      "loss": 0.4719,
+      "epoch": 3.7983193277310923,
+      "grad_norm": 2.573005569504097,
+      "learning_rate": 1.5944341326772112e-05,
+      "loss": 0.0879,
       "step": 226
     },
     {
-      "epoch": 2.7023809523809526,
-      "grad_norm": 1.5057487839402208,
-      "learning_rate": 1.8078814575405245e-05,
-      "loss": 0.4647,
+      "epoch": 3.815126050420168,
+      "grad_norm": 7.560760061208962,
+      "learning_rate": 1.5907452680775164e-05,
+      "loss": 0.2092,
       "step": 227
     },
     {
-      "epoch": 2.7142857142857144,
-      "grad_norm": 1.4988719026370554,
-      "learning_rate": 1.8059822156015208e-05,
-      "loss": 0.4375,
+      "epoch": 3.831932773109244,
+      "grad_norm": 1.9398630844655902,
+      "learning_rate": 1.587044014056943e-05,
+      "loss": 0.1255,
       "step": 228
     },
     {
-      "epoch": 2.7261904761904763,
-      "grad_norm": 8.453323259002874,
-      "learning_rate": 1.8040746400039604e-05,
-      "loss": 0.4722,
+      "epoch": 3.8487394957983194,
+      "grad_norm": 1.6555395105411446,
+      "learning_rate": 1.583330448240139e-05,
+      "loss": 0.104,
       "step": 229
     },
     {
-      "epoch": 2.738095238095238,
-      "grad_norm": 0.9029433541228935,
-      "learning_rate": 1.802158750471708e-05,
-      "loss": 0.2602,
+      "epoch": 3.865546218487395,
+      "grad_norm": 2.7598515858574775,
+      "learning_rate": 1.5796046485099633e-05,
+      "loss": 0.1696,
       "step": 230
     },
     {
-      "epoch": 2.75,
-      "grad_norm": 7.26322817577555,
-      "learning_rate": 1.8002345668145916e-05,
-      "loss": 0.4212,
+      "epoch": 3.8823529411764706,
+      "grad_norm": 2.3775725159616377,
+      "learning_rate": 1.57586669300585e-05,
+      "loss": 0.1144,
       "step": 231
     },
     {
-      "epoch": 2.761904761904762,
-      "grad_norm": 1.82321572656156,
-      "learning_rate": 1.7983021089281983e-05,
-      "loss": 0.0316,
+      "epoch": 3.899159663865546,
+      "grad_norm": 7.982198513719052,
+      "learning_rate": 1.5721166601221697e-05,
+      "loss": 0.1838,
       "step": 232
     },
     {
-      "epoch": 2.7738095238095237,
-      "grad_norm": 1.7568134018279924,
-      "learning_rate": 1.7963613967936693e-05,
-      "loss": 0.2427,
+      "epoch": 3.9159663865546217,
+      "grad_norm": 2.9062227838430847,
+      "learning_rate": 1.5683546285065878e-05,
+      "loss": 0.1136,
       "step": 233
     },
     {
-      "epoch": 2.7857142857142856,
-      "grad_norm": 4.452715421557287,
-      "learning_rate": 1.794412450477492e-05,
-      "loss": 0.1355,
+      "epoch": 3.9327731092436977,
+      "grad_norm": 2.4135228480944035,
+      "learning_rate": 1.5645806770584122e-05,
+      "loss": 0.0948,
       "step": 234
     },
     {
-      "epoch": 2.7976190476190474,
-      "grad_norm": 1.9661146108992258,
-      "learning_rate": 1.7924552901312943e-05,
-      "loss": 0.4154,
+      "epoch": 3.9495798319327733,
+      "grad_norm": 3.0055227991300333,
+      "learning_rate": 1.5607948849269404e-05,
+      "loss": 0.1268,
       "step": 235
     },
     {
-      "epoch": 2.8095238095238093,
-      "grad_norm": 3.0643788562396588,
-      "learning_rate": 1.7904899359916337e-05,
-      "loss": 0.2914,
+      "epoch": 3.966386554621849,
+      "grad_norm": 5.10158046985291,
+      "learning_rate": 1.5569973315097985e-05,
+      "loss": 0.152,
       "step": 236
     },
     {
-      "epoch": 2.821428571428571,
-      "grad_norm": 1.4486771783750234,
-      "learning_rate": 1.788516408379791e-05,
-      "loss": 0.3255,
+      "epoch": 3.9831932773109244,
+      "grad_norm": 2.436291138010729,
+      "learning_rate": 1.5531880964512773e-05,
+      "loss": 0.1158,
       "step": 237
     },
     {
-      "epoch": 2.8333333333333335,
-      "grad_norm": 3.6986784345856796,
-      "learning_rate": 1.7865347277015575e-05,
-      "loss": 0.0578,
+      "epoch": 4.0,
+      "grad_norm": 4.281661543875596,
+      "learning_rate": 1.54936725964066e-05,
+      "loss": 0.1045,
+      "step": 238
+    },
+    {
+      "epoch": 4.0,
+      "eval_loss": 0.12479228526353836,
+      "eval_runtime": 94.0271,
+      "eval_samples_per_second": 1.0,
+      "eval_steps_per_second": 1.0,
       "step": 238
     },
     {
-      "epoch": 2.8452380952380953,
-      "grad_norm": 2.445003348278355,
-      "learning_rate": 1.7845449144470256e-05,
-      "loss": 0.5247,
+      "epoch": 4.016806722689076,
+      "grad_norm": 1.912176029492126,
+      "learning_rate": 1.5455349012105488e-05,
+      "loss": 0.135,
       "step": 239
     },
     {
-      "epoch": 2.857142857142857,
-      "grad_norm": 1.9974207748375905,
-      "learning_rate": 1.7825469891903768e-05,
-      "loss": 0.6317,
+      "epoch": 4.033613445378151,
+      "grad_norm": 2.215120261832349,
+      "learning_rate": 1.5416911015351827e-05,
+      "loss": 0.127,
       "step": 240
     },
     {
-      "epoch": 2.869047619047619,
-      "grad_norm": 3.4932838771747425,
-      "learning_rate": 1.7805409725896687e-05,
-      "loss": 0.0541,
+      "epoch": 4.050420168067227,
+      "grad_norm": 1.52744144609313,
+      "learning_rate": 1.5378359412287537e-05,
+      "loss": 0.1167,
       "step": 241
     },
     {
-      "epoch": 2.880952380952381,
-      "grad_norm": 2.823086359189426,
-      "learning_rate": 1.7785268853866213e-05,
-      "loss": 0.34,
+      "epoch": 4.067226890756302,
+      "grad_norm": 1.9530294241000907,
+      "learning_rate": 1.5339695011437128e-05,
+      "loss": 0.1071,
       "step": 242
     },
     {
-      "epoch": 2.892857142857143,
-      "grad_norm": 1.834893296841028,
-      "learning_rate": 1.776504748406402e-05,
-      "loss": 0.074,
+      "epoch": 4.084033613445378,
+      "grad_norm": 2.215735360838886,
+      "learning_rate": 1.530091862369078e-05,
+      "loss": 0.1498,
       "step": 243
     },
     {
-      "epoch": 2.9047619047619047,
-      "grad_norm": 1.5422190408413234,
-      "learning_rate": 1.7744745825574123e-05,
-      "loss": 0.2529,
+      "epoch": 4.100840336134453,
+      "grad_norm": 2.2899623389238246,
+      "learning_rate": 1.526203106228733e-05,
+      "loss": 0.1452,
       "step": 244
     },
     {
-      "epoch": 2.9166666666666665,
-      "grad_norm": 2.126497256827208,
-      "learning_rate": 1.772436408831069e-05,
-      "loss": 0.414,
+      "epoch": 4.117647058823529,
+      "grad_norm": 2.366322124903928,
+      "learning_rate": 1.5223033142797183e-05,
+      "loss": 0.0997,
       "step": 245
     },
     {
-      "epoch": 2.928571428571429,
-      "grad_norm": 1.6106901139166283,
-      "learning_rate": 1.770390248301589e-05,
-      "loss": 0.2728,
+      "epoch": 4.1344537815126055,
+      "grad_norm": 2.180155804240841,
+      "learning_rate": 1.5183925683105254e-05,
+      "loss": 0.1103,
       "step": 246
     },
     {
-      "epoch": 2.9404761904761907,
-      "grad_norm": 2.544109776535844,
-      "learning_rate": 1.7683361221257705e-05,
-      "loss": 0.6377,
+      "epoch": 4.151260504201681,
+      "grad_norm": 1.9915581272469869,
+      "learning_rate": 1.5144709503393773e-05,
+      "loss": 0.1178,
       "step": 247
     },
     {
-      "epoch": 2.9523809523809526,
-      "grad_norm": 3.931188529393615,
-      "learning_rate": 1.7662740515427753e-05,
-      "loss": 0.5146,
+      "epoch": 4.168067226890757,
+      "grad_norm": 2.0239687647120905,
+      "learning_rate": 1.5105385426125123e-05,
+      "loss": 0.0996,
       "step": 248
     },
     {
-      "epoch": 2.9642857142857144,
-      "grad_norm": 20.724460916802276,
-      "learning_rate": 1.764204057873907e-05,
-      "loss": 0.8433,
+      "epoch": 4.184873949579832,
+      "grad_norm": 1.86641722627037,
+      "learning_rate": 1.5065954276024561e-05,
+      "loss": 0.1104,
       "step": 249
     },
     {
-      "epoch": 2.9761904761904763,
-      "grad_norm": 3.8203367792590024,
-      "learning_rate": 1.762126162522393e-05,
-      "loss": 0.0734,
+      "epoch": 4.201680672268908,
+      "grad_norm": 1.6792110703483962,
+      "learning_rate": 1.5026416880062932e-05,
+      "loss": 0.1237,
       "step": 250
     },
     {
-      "epoch": 2.988095238095238,
-      "grad_norm": 1.4135137345241042,
-      "learning_rate": 1.760040386973162e-05,
-      "loss": 0.3819,
+      "epoch": 4.218487394957983,
+      "grad_norm": 1.2608521044993408,
+      "learning_rate": 1.4986774067439327e-05,
+      "loss": 0.1011,
       "step": 251
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 3.288964040896793,
-      "learning_rate": 1.7579467527926223e-05,
-      "loss": 0.513,
-      "step": 252
-    },
-    {
-      "epoch": 3.0,
-      "eval_loss": 0.37594377994537354,
-      "eval_runtime": 38.8272,
-      "eval_samples_per_second": 1.545,
-      "eval_steps_per_second": 1.545,
+      "epoch": 4.235294117647059,
+      "grad_norm": 2.5144065390038617,
+      "learning_rate": 1.4947026669563687e-05,
+      "loss": 0.1515,
       "step": 252
     },
     {
-      "epoch": 3.011904761904762,
-      "grad_norm": 2.2658309314703837,
-      "learning_rate": 1.7558452816284374e-05,
-      "loss": 0.2167,
+      "epoch": 4.2521008403361344,
+      "grad_norm": 1.5804219840273723,
+      "learning_rate": 1.4907175520039381e-05,
+      "loss": 0.1163,
       "step": 253
     },
     {
-      "epoch": 3.0238095238095237,
-      "grad_norm": 2.171321063645936,
-      "learning_rate": 1.7537359952093046e-05,
-      "loss": 0.2767,
+      "epoch": 4.26890756302521,
+      "grad_norm": 2.2244097182086273,
+      "learning_rate": 1.4867221454645696e-05,
+      "loss": 0.1188,
       "step": 254
     },
     {
-      "epoch": 3.0357142857142856,
-      "grad_norm": 3.224696150470818,
-      "learning_rate": 1.7516189153447283e-05,
-      "loss": 0.3851,
+      "epoch": 4.285714285714286,
+      "grad_norm": 1.64104407678126,
+      "learning_rate": 1.482716531132034e-05,
+      "loss": 0.0965,
       "step": 255
     },
     {
-      "epoch": 3.0476190476190474,
-      "grad_norm": 3.3132846666505413,
-      "learning_rate": 1.7494940639247953e-05,
-      "loss": 0.0783,
+      "epoch": 4.302521008403361,
+      "grad_norm": 0.8551932780454022,
+      "learning_rate": 1.4787007930141841e-05,
+      "loss": 0.0655,
       "step": 256
     },
     {
-      "epoch": 3.0595238095238093,
-      "grad_norm": 3.282432290241069,
-      "learning_rate": 1.747361462919949e-05,
-      "loss": 0.5529,
+      "epoch": 4.319327731092437,
+      "grad_norm": 1.752648892273715,
+      "learning_rate": 1.4746750153311951e-05,
+      "loss": 0.0946,
       "step": 257
     },
     {
-      "epoch": 3.0714285714285716,
-      "grad_norm": 4.303685952728001,
-      "learning_rate": 1.74522113438076e-05,
-      "loss": 0.2379,
+      "epoch": 4.336134453781512,
+      "grad_norm": 5.829984337336428,
+      "learning_rate": 1.4706392825137962e-05,
+      "loss": 0.1826,
       "step": 258
     },
     {
-      "epoch": 3.0833333333333335,
-      "grad_norm": 3.1849988371458466,
-      "learning_rate": 1.7430731004377016e-05,
-      "loss": 0.3984,
+      "epoch": 4.352941176470588,
+      "grad_norm": 2.302171396641537,
+      "learning_rate": 1.4665936792015021e-05,
+      "loss": 0.0909,
       "step": 259
     },
     {
-      "epoch": 3.0952380952380953,
-      "grad_norm": 6.847041494024382,
-      "learning_rate": 1.7409173833009184e-05,
-      "loss": 0.5108,
+      "epoch": 4.369747899159664,
+      "grad_norm": 1.2086828157662006,
+      "learning_rate": 1.4625382902408356e-05,
+      "loss": 0.0972,
       "step": 260
     },
     {
-      "epoch": 3.107142857142857,
-      "grad_norm": 2.758331616979815,
-      "learning_rate": 1.738754005259997e-05,
-      "loss": 0.0378,
+      "epoch": 4.38655462184874,
+      "grad_norm": 1.3490901275333183,
+      "learning_rate": 1.4584732006835495e-05,
+      "loss": 0.0823,
       "step": 261
     },
     {
-      "epoch": 3.119047619047619,
-      "grad_norm": 4.627462090057032,
-      "learning_rate": 1.7365829886837355e-05,
-      "loss": 0.0942,
+      "epoch": 4.4033613445378155,
+      "grad_norm": 3.3951376429875393,
+      "learning_rate": 1.4543984957848438e-05,
+      "loss": 0.1492,
       "step": 262
     },
     {
-      "epoch": 3.130952380952381,
-      "grad_norm": 4.2867703712422305,
-      "learning_rate": 1.7344043560199137e-05,
-      "loss": 0.4127,
+      "epoch": 4.420168067226891,
+      "grad_norm": 2.4333139248557165,
+      "learning_rate": 1.4503142610015751e-05,
+      "loss": 0.0974,
       "step": 263
     },
     {
-      "epoch": 3.142857142857143,
-      "grad_norm": 3.324654892695759,
-      "learning_rate": 1.7322181297950595e-05,
-      "loss": 0.4679,
+      "epoch": 4.436974789915967,
+      "grad_norm": 1.818247800901397,
+      "learning_rate": 1.4462205819904658e-05,
+      "loss": 0.0813,
       "step": 264
     },
     {
-      "epoch": 3.1547619047619047,
-      "grad_norm": 1.501271013583071,
-      "learning_rate": 1.7300243326142157e-05,
-      "loss": 0.3555,
+      "epoch": 4.453781512605042,
+      "grad_norm": 2.761939597967942,
+      "learning_rate": 1.4421175446063086e-05,
+      "loss": 0.1356,
       "step": 265
     },
     {
-      "epoch": 3.1666666666666665,
-      "grad_norm": 2.9950158939240867,
-      "learning_rate": 1.7278229871607083e-05,
-      "loss": 0.3781,
+      "epoch": 4.470588235294118,
+      "grad_norm": 2.3235891826621895,
+      "learning_rate": 1.4380052349001647e-05,
+      "loss": 0.1393,
       "step": 266
     },
     {
-      "epoch": 3.1785714285714284,
-      "grad_norm": 1.4907573955706444,
-      "learning_rate": 1.7256141161959087e-05,
-      "loss": 0.3505,
+      "epoch": 4.487394957983193,
+      "grad_norm": 1.4778394471726972,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.1124,
       "step": 267
     },
     {
-      "epoch": 3.1904761904761907,
-      "grad_norm": 2.66786279641303,
-      "learning_rate": 1.7233977425590015e-05,
-      "loss": 0.3757,
+      "epoch": 4.504201680672269,
+      "grad_norm": 1.6212161850248412,
+      "learning_rate": 1.42975314369667e-05,
+      "loss": 0.0834,
       "step": 268
     },
     {
-      "epoch": 3.2023809523809526,
-      "grad_norm": 3.7324135967195082,
-      "learning_rate": 1.7211738891667474e-05,
-      "loss": 0.2186,
+      "epoch": 4.5210084033613445,
+      "grad_norm": 1.513176697844455,
+      "learning_rate": 1.4256135352665217e-05,
+      "loss": 0.1053,
       "step": 269
     },
     {
-      "epoch": 3.2142857142857144,
-      "grad_norm": 0.8930591045310717,
-      "learning_rate": 1.7189425790132452e-05,
-      "loss": 0.1605,
+      "epoch": 4.53781512605042,
+      "grad_norm": 1.5842489277719052,
+      "learning_rate": 1.4214650006451622e-05,
+      "loss": 0.0793,
       "step": 270
     },
     {
-      "epoch": 3.2261904761904763,
-      "grad_norm": 2.5813547947324644,
-      "learning_rate": 1.7167038351696947e-05,
-      "loss": 0.5126,
+      "epoch": 4.554621848739496,
+      "grad_norm": 2.882811290847085,
+      "learning_rate": 1.4173076268378443e-05,
+      "loss": 0.0944,
       "step": 271
     },
     {
-      "epoch": 3.238095238095238,
-      "grad_norm": 0.9752237664543323,
-      "learning_rate": 1.7144576807841583e-05,
-      "loss": 0.1811,
+      "epoch": 4.571428571428571,
+      "grad_norm": 2.540707077443368,
+      "learning_rate": 1.4131415010352007e-05,
+      "loss": 0.0929,
       "step": 272
     },
     {
-      "epoch": 3.25,
-      "grad_norm": 0.9647196518957709,
-      "learning_rate": 1.7122041390813228e-05,
-      "loss": 0.0154,
+      "epoch": 4.588235294117647,
+      "grad_norm": 2.224835930797884,
+      "learning_rate": 1.408966710611416e-05,
+      "loss": 0.0965,
       "step": 273
     },
     {
-      "epoch": 3.261904761904762,
-      "grad_norm": 2.550395344630908,
-      "learning_rate": 1.709943233362256e-05,
-      "loss": 0.5381,
+      "epoch": 4.605042016806722,
+      "grad_norm": 1.603233390085239,
+      "learning_rate": 1.4047833431223938e-05,
+      "loss": 0.1082,
       "step": 274
     },
     {
-      "epoch": 3.2738095238095237,
-      "grad_norm": 0.9538094415143878,
-      "learning_rate": 1.7076749870041692e-05,
-      "loss": 0.1486,
+      "epoch": 4.621848739495798,
+      "grad_norm": 1.7397244673026468,
+      "learning_rate": 1.4005914863039203e-05,
+      "loss": 0.0765,
       "step": 275
     },
     {
-      "epoch": 3.2857142857142856,
-      "grad_norm": 5.604529448927459,
-      "learning_rate": 1.7053994234601736e-05,
-      "loss": 0.4003,
+      "epoch": 4.6386554621848735,
+      "grad_norm": 1.9121905502244794,
+      "learning_rate": 1.3963912280698238e-05,
+      "loss": 0.0852,
       "step": 276
     },
     {
-      "epoch": 3.2976190476190474,
-      "grad_norm": 9.077045722590805,
-      "learning_rate": 1.703116566259039e-05,
-      "loss": 0.408,
+      "epoch": 4.65546218487395,
+      "grad_norm": 1.8233199956684913,
+      "learning_rate": 1.3921826565101325e-05,
+      "loss": 0.104,
       "step": 277
     },
     {
-      "epoch": 3.3095238095238093,
-      "grad_norm": 3.300132707260137,
-      "learning_rate": 1.7008264390049493e-05,
-      "loss": 0.2426,
+      "epoch": 4.6722689075630255,
+      "grad_norm": 0.7406392209229766,
+      "learning_rate": 1.3879658598892254e-05,
+      "loss": 0.0597,
       "step": 278
     },
     {
-      "epoch": 3.3214285714285716,
-      "grad_norm": 2.0242722270180527,
-      "learning_rate": 1.6985290653772583e-05,
-      "loss": 0.1482,
+      "epoch": 4.689075630252101,
+      "grad_norm": 7.419162565626874,
+      "learning_rate": 1.3837409266439818e-05,
+      "loss": 0.209,
       "step": 279
     },
     {
-      "epoch": 3.3333333333333335,
-      "grad_norm": 2.8369343983065294,
-      "learning_rate": 1.6962244691302466e-05,
-      "loss": 0.1434,
+      "epoch": 4.705882352941177,
+      "grad_norm": 1.1738246179937548,
+      "learning_rate": 1.3795079453819276e-05,
+      "loss": 0.0996,
       "step": 280
     },
     {
-      "epoch": 3.3452380952380953,
-      "grad_norm": 7.081419291772338,
-      "learning_rate": 1.6939126740928745e-05,
-      "loss": 0.1271,
+      "epoch": 4.722689075630252,
+      "grad_norm": 0.7316310386069824,
+      "learning_rate": 1.3752670048793744e-05,
+      "loss": 0.0573,
       "step": 281
     },
     {
-      "epoch": 3.357142857142857,
-      "grad_norm": 2.3722193904935716,
-      "learning_rate": 1.691593704168536e-05,
-      "loss": 0.2892,
+      "epoch": 4.739495798319328,
+      "grad_norm": 1.0245543216679647,
+      "learning_rate": 1.37101819407956e-05,
+      "loss": 0.0798,
       "step": 282
     },
     {
-      "epoch": 3.369047619047619,
-      "grad_norm": 6.960729490454029,
-      "learning_rate": 1.6892675833348124e-05,
-      "loss": 0.1294,
+      "epoch": 4.756302521008403,
+      "grad_norm": 1.2604913385475165,
+      "learning_rate": 1.366761602090782e-05,
+      "loss": 0.082,
       "step": 283
     },
     {
-      "epoch": 3.380952380952381,
-      "grad_norm": 1.1500777231564219,
-      "learning_rate": 1.686934335643222e-05,
-      "loss": 0.2509,
+      "epoch": 4.773109243697479,
+      "grad_norm": 1.6057488873958579,
+      "learning_rate": 1.3624973181845302e-05,
+      "loss": 0.0765,
       "step": 284
     },
     {
-      "epoch": 3.392857142857143,
-      "grad_norm": 1.5870567460181373,
-      "learning_rate": 1.684593985218974e-05,
-      "loss": 0.2563,
+      "epoch": 4.7899159663865545,
+      "grad_norm": 2.0379166749602895,
+      "learning_rate": 1.3582254317936117e-05,
+      "loss": 0.1001,
       "step": 285
     },
     {
-      "epoch": 3.4047619047619047,
-      "grad_norm": 5.098839245282105,
-      "learning_rate": 1.6822465562607186e-05,
-      "loss": 0.095,
+      "epoch": 4.80672268907563,
+      "grad_norm": 1.7603012476865616,
+      "learning_rate": 1.3539460325102779e-05,
+      "loss": 0.1126,
       "step": 286
     },
     {
-      "epoch": 3.4166666666666665,
-      "grad_norm": 4.33924245316836,
-      "learning_rate": 1.6798920730402962e-05,
-      "loss": 0.1761,
+      "epoch": 4.823529411764706,
+      "grad_norm": 1.4253082853875376,
+      "learning_rate": 1.349659210084344e-05,
+      "loss": 0.0978,
       "step": 287
     },
     {
-      "epoch": 3.4285714285714284,
-      "grad_norm": 1.4460187648094192,
-      "learning_rate": 1.6775305599024853e-05,
-      "loss": 0.2282,
+      "epoch": 4.840336134453781,
+      "grad_norm": 1.1284525438619117,
+      "learning_rate": 1.3453650544213078e-05,
+      "loss": 0.0695,
       "step": 288
     },
     {
-      "epoch": 3.4404761904761907,
-      "grad_norm": 2.984916897434454,
-      "learning_rate": 1.675162041264754e-05,
-      "loss": 0.0698,
+      "epoch": 4.857142857142857,
+      "grad_norm": 0.9802953497383712,
+      "learning_rate": 1.3410636555804634e-05,
+      "loss": 0.0751,
       "step": 289
     },
     {
-      "epoch": 3.4523809523809526,
-      "grad_norm": 5.808631040149753,
-      "learning_rate": 1.6727865416170032e-05,
-      "loss": 0.0843,
+      "epoch": 4.873949579831933,
+      "grad_norm": 1.3612858881668406,
+      "learning_rate": 1.3367551037730129e-05,
+      "loss": 0.0966,
       "step": 290
     },
     {
-      "epoch": 3.4642857142857144,
-      "grad_norm": 2.666294292390359,
-      "learning_rate": 1.6704040855213182e-05,
-      "loss": 0.0391,
+      "epoch": 4.890756302521009,
+      "grad_norm": 1.0157890782271555,
+      "learning_rate": 1.3324394893601734e-05,
+      "loss": 0.059,
       "step": 291
     },
     {
-      "epoch": 3.4761904761904763,
-      "grad_norm": 2.4526472332810165,
-      "learning_rate": 1.6680146976117105e-05,
-      "loss": 0.393,
+      "epoch": 4.907563025210084,
+      "grad_norm": 1.4054910887816698,
+      "learning_rate": 1.3281169028512838e-05,
+      "loss": 0.0709,
       "step": 292
     },
     {
-      "epoch": 3.488095238095238,
-      "grad_norm": 5.063252926047181,
-      "learning_rate": 1.6656184025938654e-05,
-      "loss": 0.3265,
+      "epoch": 4.92436974789916,
+      "grad_norm": 0.8662398448794504,
+      "learning_rate": 1.3237874349019041e-05,
+      "loss": 0.0756,
       "step": 293
     },
     {
-      "epoch": 3.5,
-      "grad_norm": 0.7680043862411393,
-      "learning_rate": 1.663215225244886e-05,
-      "loss": 0.0138,
+      "epoch": 4.9411764705882355,
+      "grad_norm": 1.3483246839258962,
+      "learning_rate": 1.319451176311917e-05,
+      "loss": 0.0769,
       "step": 294
     },
     {
-      "epoch": 3.511904761904762,
-      "grad_norm": 2.3680085112002307,
-      "learning_rate": 1.6608051904130375e-05,
-      "loss": 0.4016,
+      "epoch": 4.957983193277311,
+      "grad_norm": 2.4153558320453614,
+      "learning_rate": 1.315108218023621e-05,
+      "loss": 0.1131,
       "step": 295
     },
     {
-      "epoch": 3.5238095238095237,
-      "grad_norm": 1.9514276941132964,
-      "learning_rate": 1.6583883230174884e-05,
-      "loss": 0.1216,
+      "epoch": 4.974789915966387,
+      "grad_norm": 2.555834294764848,
+      "learning_rate": 1.3107586511198243e-05,
+      "loss": 0.141,
       "step": 296
     },
     {
-      "epoch": 3.5357142857142856,
-      "grad_norm": 1.9177647290193789,
-      "learning_rate": 1.6559646480480563e-05,
-      "loss": 0.2042,
+      "epoch": 4.991596638655462,
+      "grad_norm": 4.147522093497237,
+      "learning_rate": 1.306402566821935e-05,
+      "loss": 0.1456,
+      "step": 297
+    },
+    {
+      "epoch": 4.991596638655462,
+      "eval_loss": 0.1449918895959854,
+      "eval_runtime": 97.1487,
+      "eval_samples_per_second": 0.968,
+      "eval_steps_per_second": 0.968,
       "step": 297
     },
     {
-      "epoch": 3.5476190476190474,
-      "grad_norm": 11.831680317369488,
-      "learning_rate": 1.6535341905649453e-05,
-      "loss": 0.3236,
+      "epoch": 5.008403361344538,
+      "grad_norm": 1.6155708244562734,
+      "learning_rate": 1.302040056488047e-05,
+      "loss": 0.0717,
       "step": 298
     },
     {
-      "epoch": 3.5595238095238093,
-      "grad_norm": 0.854538769203965,
-      "learning_rate": 1.651096975698491e-05,
-      "loss": 0.1621,
+      "epoch": 5.025210084033613,
+      "grad_norm": 1.144238074710071,
+      "learning_rate": 1.297671211611025e-05,
+      "loss": 0.0719,
       "step": 299
     },
     {
-      "epoch": 3.571428571428571,
-      "grad_norm": 1.7325718951049507,
-      "learning_rate": 1.648653028648897e-05,
-      "loss": 0.3236,
+      "epoch": 5.042016806722689,
+      "grad_norm": 2.0576085887744795,
+      "learning_rate": 1.2932961238165837e-05,
+      "loss": 0.1152,
       "step": 300
     },
     {
-      "epoch": 3.5833333333333335,
-      "grad_norm": 1.6802180110167104,
-      "learning_rate": 1.6462023746859774e-05,
-      "loss": 0.1567,
+      "epoch": 5.0588235294117645,
+      "grad_norm": 1.0604218669017003,
+      "learning_rate": 1.2889148848613695e-05,
+      "loss": 0.0712,
       "step": 301
     },
     {
-      "epoch": 3.5952380952380953,
-      "grad_norm": 3.275665572161553,
-      "learning_rate": 1.6437450391488928e-05,
-      "loss": 0.3051,
+      "epoch": 5.07563025210084,
+      "grad_norm": 1.1699347626659127,
+      "learning_rate": 1.2845275866310325e-05,
+      "loss": 0.061,
       "step": 302
     },
     {
-      "epoch": 3.607142857142857,
-      "grad_norm": 0.8259080343432141,
-      "learning_rate": 1.6412810474458906e-05,
-      "loss": 0.1376,
+      "epoch": 5.092436974789916,
+      "grad_norm": 1.1371190586967697,
+      "learning_rate": 1.2801343211383021e-05,
+      "loss": 0.0747,
       "step": 303
     },
     {
-      "epoch": 3.619047619047619,
-      "grad_norm": 3.0586732720031526,
-      "learning_rate": 1.6388104250540414e-05,
-      "loss": 0.121,
+      "epoch": 5.109243697478991,
+      "grad_norm": 1.8857383359198918,
+      "learning_rate": 1.2757351805210557e-05,
+      "loss": 0.1447,
       "step": 304
     },
     {
-      "epoch": 3.630952380952381,
-      "grad_norm": 2.473548405954027,
-      "learning_rate": 1.6363331975189748e-05,
-      "loss": 0.2078,
+      "epoch": 5.126050420168067,
+      "grad_norm": 1.0626942422062413,
+      "learning_rate": 1.2713302570403872e-05,
+      "loss": 0.0657,
       "step": 305
     },
     {
-      "epoch": 3.642857142857143,
-      "grad_norm": 3.2034115805081163,
-      "learning_rate": 1.6338493904546163e-05,
-      "loss": 0.1119,
+      "epoch": 5.142857142857143,
+      "grad_norm": 1.3367342967159013,
+      "learning_rate": 1.2669196430786715e-05,
+      "loss": 0.086,
       "step": 306
     },
     {
-      "epoch": 3.6547619047619047,
-      "grad_norm": 1.735238565417923,
-      "learning_rate": 1.6313590295429224e-05,
-      "loss": 0.0482,
+      "epoch": 5.159663865546219,
+      "grad_norm": 4.036414956554653,
+      "learning_rate": 1.2625034311376276e-05,
+      "loss": 0.1317,
       "step": 307
     },
     {
-      "epoch": 3.6666666666666665,
-      "grad_norm": 1.7705110125878167,
-      "learning_rate": 1.6288621405336144e-05,
-      "loss": 0.167,
+      "epoch": 5.176470588235294,
+      "grad_norm": 2.092631890687451,
+      "learning_rate": 1.258081713836378e-05,
+      "loss": 0.0958,
       "step": 308
     },
     {
-      "epoch": 3.678571428571429,
-      "grad_norm": 1.350904255609373,
-      "learning_rate": 1.6263587492439127e-05,
-      "loss": 0.1935,
+      "epoch": 5.19327731092437,
+      "grad_norm": 2.470224875937704,
+      "learning_rate": 1.2536545839095074e-05,
+      "loss": 0.0764,
       "step": 309
     },
     {
-      "epoch": 3.6904761904761907,
-      "grad_norm": 5.116050315328356,
-      "learning_rate": 1.6238488815582693e-05,
-      "loss": 0.1613,
+      "epoch": 5.2100840336134455,
+      "grad_norm": 2.6817499179267483,
+      "learning_rate": 1.2492221342051153e-05,
+      "loss": 0.1124,
       "step": 310
     },
     {
-      "epoch": 3.7023809523809526,
-      "grad_norm": 4.879662048621329,
-      "learning_rate": 1.6213325634281017e-05,
-      "loss": 0.2637,
+      "epoch": 5.226890756302521,
+      "grad_norm": 2.000873702339644,
+      "learning_rate": 1.2447844576828719e-05,
+      "loss": 0.1067,
       "step": 311
     },
     {
-      "epoch": 3.7142857142857144,
-      "grad_norm": 8.743930661693144,
-      "learning_rate": 1.6188098208715216e-05,
-      "loss": 0.3958,
+      "epoch": 5.243697478991597,
+      "grad_norm": 2.039025163817311,
+      "learning_rate": 1.2403416474120657e-05,
+      "loss": 0.0703,
       "step": 312
     },
     {
-      "epoch": 3.7261904761904763,
-      "grad_norm": 14.84650411290692,
-      "learning_rate": 1.6162806799730694e-05,
-      "loss": 0.3667,
+      "epoch": 5.260504201680672,
+      "grad_norm": 1.2573161613968584,
+      "learning_rate": 1.2358937965696538e-05,
+      "loss": 0.0694,
       "step": 313
     },
     {
-      "epoch": 3.738095238095238,
-      "grad_norm": 3.343827891489822,
-      "learning_rate": 1.6137451668834415e-05,
-      "loss": 0.1576,
+      "epoch": 5.277310924369748,
+      "grad_norm": 1.6533379980735152,
+      "learning_rate": 1.2314409984383066e-05,
+      "loss": 0.0828,
       "step": 314
     },
     {
-      "epoch": 3.75,
-      "grad_norm": 2.1179290591609656,
-      "learning_rate": 1.6112033078192225e-05,
-      "loss": 0.1657,
+      "epoch": 5.294117647058823,
+      "grad_norm": 2.0132269342271467,
+      "learning_rate": 1.2269833464044514e-05,
+      "loss": 0.064,
       "step": 315
     },
     {
-      "epoch": 3.761904761904762,
-      "grad_norm": 9.300028025529468,
-      "learning_rate": 1.6086551290626117e-05,
-      "loss": 0.162,
+      "epoch": 5.310924369747899,
+      "grad_norm": 1.6870312479105358,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.1093,
       "step": 316
     },
     {
-      "epoch": 3.7738095238095237,
-      "grad_norm": 2.974584952079088,
-      "learning_rate": 1.6061006569611524e-05,
-      "loss": 0.1714,
+      "epoch": 5.3277310924369745,
+      "grad_norm": 1.2264520173137188,
+      "learning_rate": 1.2180538546819595e-05,
+      "loss": 0.0844,
       "step": 317
     },
     {
-      "epoch": 3.7857142857142856,
-      "grad_norm": 1.3270086319514083,
-      "learning_rate": 1.6035399179274604e-05,
-      "loss": 0.2454,
+      "epoch": 5.34453781512605,
+      "grad_norm": 1.6926310927686044,
+      "learning_rate": 1.2135822022673263e-05,
+      "loss": 0.0663,
       "step": 318
     },
     {
-      "epoch": 3.7976190476190474,
-      "grad_norm": 2.239433984527888,
-      "learning_rate": 1.60097293843895e-05,
-      "loss": 0.1539,
+      "epoch": 5.361344537815126,
+      "grad_norm": 4.790942950250594,
+      "learning_rate": 1.2091060704942636e-05,
+      "loss": 0.1585,
       "step": 319
     },
     {
-      "epoch": 3.8095238095238093,
-      "grad_norm": 1.0309175730109765,
-      "learning_rate": 1.5983997450375594e-05,
-      "loss": 0.1329,
+      "epoch": 5.378151260504202,
+      "grad_norm": 1.0161253142944433,
+      "learning_rate": 1.204625553238565e-05,
+      "loss": 0.0723,
       "step": 320
     },
     {
-      "epoch": 3.821428571428571,
-      "grad_norm": 8.007832806905567,
-      "learning_rate": 1.595820364329478e-05,
-      "loss": 0.2519,
+      "epoch": 5.394957983193278,
+      "grad_norm": 1.1558751249048937,
+      "learning_rate": 1.200140744467997e-05,
+      "loss": 0.084,
       "step": 321
     },
     {
-      "epoch": 3.8333333333333335,
-      "grad_norm": 4.178320830537582,
-      "learning_rate": 1.5932348229848704e-05,
-      "loss": 0.2863,
+      "epoch": 5.411764705882353,
+      "grad_norm": 0.9419572615825091,
+      "learning_rate": 1.195651738240332e-05,
+      "loss": 0.0571,
       "step": 322
     },
     {
-      "epoch": 3.8452380952380953,
-      "grad_norm": 7.344798396674696,
-      "learning_rate": 1.5906431477375994e-05,
-      "loss": 0.2325,
+      "epoch": 5.428571428571429,
+      "grad_norm": 1.622044146409352,
+      "learning_rate": 1.1911586287013726e-05,
+      "loss": 0.0698,
       "step": 323
     },
     {
-      "epoch": 3.857142857142857,
-      "grad_norm": 2.728295310180829,
-      "learning_rate": 1.5880453653849523e-05,
-      "loss": 0.2494,
+      "epoch": 5.445378151260504,
+      "grad_norm": 1.7252519658503152,
+      "learning_rate": 1.1866615100829777e-05,
+      "loss": 0.0727,
       "step": 324
     },
     {
-      "epoch": 3.869047619047619,
-      "grad_norm": 1.5354481844217602,
-      "learning_rate": 1.5854415027873624e-05,
-      "loss": 0.1782,
+      "epoch": 5.46218487394958,
+      "grad_norm": 1.2769460474277,
+      "learning_rate": 1.1821604767010883e-05,
+      "loss": 0.0728,
       "step": 325
     },
     {
-      "epoch": 3.880952380952381,
-      "grad_norm": 2.826240940303501,
-      "learning_rate": 1.58283158686813e-05,
-      "loss": 0.2712,
+      "epoch": 5.4789915966386555,
+      "grad_norm": 1.6489519703945064,
+      "learning_rate": 1.1776556229537461e-05,
+      "loss": 0.0712,
       "step": 326
     },
     {
-      "epoch": 3.892857142857143,
-      "grad_norm": 5.714545377682789,
-      "learning_rate": 1.5802156446131463e-05,
-      "loss": 0.3086,
+      "epoch": 5.495798319327731,
+      "grad_norm": 1.9144641595351362,
+      "learning_rate": 1.1731470433191173e-05,
+      "loss": 0.0932,
       "step": 327
     },
     {
-      "epoch": 3.9047619047619047,
-      "grad_norm": 3.037951498821358,
-      "learning_rate": 1.577593703070613e-05,
-      "loss": 0.1702,
+      "epoch": 5.512605042016807,
+      "grad_norm": 1.0334398447359328,
+      "learning_rate": 1.1686348323535078e-05,
+      "loss": 0.0391,
       "step": 328
     },
     {
-      "epoch": 3.9166666666666665,
-      "grad_norm": 5.415513375394119,
-      "learning_rate": 1.5749657893507627e-05,
-      "loss": 0.1378,
+      "epoch": 5.529411764705882,
+      "grad_norm": 2.4736197087343195,
+      "learning_rate": 1.1641190846893824e-05,
+      "loss": 0.0955,
       "step": 329
     },
     {
-      "epoch": 3.928571428571429,
-      "grad_norm": 2.616666785733442,
-      "learning_rate": 1.5723319306255798e-05,
-      "loss": 0.2401,
+      "epoch": 5.546218487394958,
+      "grad_norm": 1.3189455008208653,
+      "learning_rate": 1.1595998950333794e-05,
+      "loss": 0.0565,
       "step": 330
     },
     {
-      "epoch": 3.9404761904761907,
-      "grad_norm": 2.716518576090543,
-      "learning_rate": 1.5696921541285176e-05,
-      "loss": 0.2716,
+      "epoch": 5.563025210084033,
+      "grad_norm": 1.7006720766337666,
+      "learning_rate": 1.1550773581643245e-05,
+      "loss": 0.0786,
       "step": 331
     },
     {
-      "epoch": 3.9523809523809526,
-      "grad_norm": 6.0778719414757365,
-      "learning_rate": 1.5670464871542178e-05,
-      "loss": 0.1819,
+      "epoch": 5.579831932773109,
+      "grad_norm": 4.52735397655905,
+      "learning_rate": 1.1505515689312424e-05,
+      "loss": 0.1432,
       "step": 332
     },
     {
-      "epoch": 3.9642857142857144,
-      "grad_norm": 2.300785383942732,
-      "learning_rate": 1.564394957058229e-05,
-      "loss": 0.1137,
+      "epoch": 5.5966386554621845,
+      "grad_norm": 3.0704885963032953,
+      "learning_rate": 1.1460226222513682e-05,
+      "loss": 0.1099,
       "step": 333
     },
     {
-      "epoch": 3.9761904761904763,
-      "grad_norm": 1.4978039537439425,
-      "learning_rate": 1.5617375912567218e-05,
-      "loss": 0.2404,
+      "epoch": 5.61344537815126,
+      "grad_norm": 4.068391506810968,
+      "learning_rate": 1.1414906131081575e-05,
+      "loss": 0.1559,
       "step": 334
     },
     {
-      "epoch": 3.988095238095238,
-      "grad_norm": 1.8520009539803934,
-      "learning_rate": 1.5590744172262076e-05,
-      "loss": 0.1392,
+      "epoch": 5.630252100840336,
+      "grad_norm": 1.6178193211411984,
+      "learning_rate": 1.1369556365492924e-05,
+      "loss": 0.0881,
       "step": 335
     },
     {
-      "epoch": 4.0,
-      "grad_norm": 1.0739415756051076,
-      "learning_rate": 1.5564054625032534e-05,
-      "loss": 0.1541,
-      "step": 336
-    },
-    {
-      "epoch": 4.0,
-      "eval_loss": 0.2209147959947586,
-      "eval_runtime": 38.5178,
-      "eval_samples_per_second": 1.558,
-      "eval_steps_per_second": 1.558,
+      "epoch": 5.647058823529412,
+      "grad_norm": 1.472096896311054,
+      "learning_rate": 1.1324177876846897e-05,
+      "loss": 0.0668,
       "step": 336
     },
     {
-      "epoch": 4.011904761904762,
-      "grad_norm": 2.827870135987854,
-      "learning_rate": 1.553730754684196e-05,
-      "loss": 0.039,
+      "epoch": 5.663865546218488,
+      "grad_norm": 0.9377593241374005,
+      "learning_rate": 1.1278771616845061e-05,
+      "loss": 0.0528,
       "step": 337
     },
     {
-      "epoch": 4.023809523809524,
-      "grad_norm": 2.071434578952237,
-      "learning_rate": 1.5510503214248597e-05,
-      "loss": 0.1753,
+      "epoch": 5.680672268907563,
+      "grad_norm": 1.60191420846021,
+      "learning_rate": 1.1233338537771408e-05,
+      "loss": 0.0839,
       "step": 338
     },
     {
-      "epoch": 4.035714285714286,
-      "grad_norm": 1.7843811186555119,
-      "learning_rate": 1.5483641904402666e-05,
-      "loss": 0.106,
+      "epoch": 5.697478991596639,
+      "grad_norm": 2.4242623809016757,
+      "learning_rate": 1.1187879592472402e-05,
+      "loss": 0.0765,
       "step": 339
     },
     {
-      "epoch": 4.0476190476190474,
-      "grad_norm": 8.82649724600633,
-      "learning_rate": 1.545672389504353e-05,
-      "loss": 0.1881,
+      "epoch": 5.714285714285714,
+      "grad_norm": 3.3339459249935075,
+      "learning_rate": 1.1142395734336986e-05,
+      "loss": 0.1304,
       "step": 340
     },
     {
-      "epoch": 4.059523809523809,
-      "grad_norm": 6.291910460115927,
-      "learning_rate": 1.5429749464496808e-05,
-      "loss": 0.1765,
+      "epoch": 5.73109243697479,
+      "grad_norm": 2.7416841232755824,
+      "learning_rate": 1.1096887917276585e-05,
+      "loss": 0.0776,
       "step": 341
     },
     {
-      "epoch": 4.071428571428571,
-      "grad_norm": 2.643403230336219,
-      "learning_rate": 1.54027188916715e-05,
-      "loss": 0.0459,
+      "epoch": 5.7478991596638656,
+      "grad_norm": 1.8254292684914686,
+      "learning_rate": 1.1051357095705102e-05,
+      "loss": 0.0679,
       "step": 342
     },
     {
-      "epoch": 4.083333333333333,
-      "grad_norm": 3.497562450238014,
-      "learning_rate": 1.5375632456057104e-05,
-      "loss": 0.1232,
+      "epoch": 5.764705882352941,
+      "grad_norm": 1.807300374973592,
+      "learning_rate": 1.1005804224518912e-05,
+      "loss": 0.0769,
       "step": 343
     },
     {
-      "epoch": 4.095238095238095,
-      "grad_norm": 2.2512921097593592,
-      "learning_rate": 1.5348490437720725e-05,
-      "loss": 0.323,
+      "epoch": 5.781512605042017,
+      "grad_norm": 1.8476121472163518,
+      "learning_rate": 1.0960230259076819e-05,
+      "loss": 0.0777,
       "step": 344
     },
     {
-      "epoch": 4.107142857142857,
-      "grad_norm": 5.464247034665501,
-      "learning_rate": 1.532129311730418e-05,
-      "loss": 0.0886,
+      "epoch": 5.798319327731092,
+      "grad_norm": 2.2451749634606317,
+      "learning_rate": 1.0914636155180025e-05,
+      "loss": 0.0843,
       "step": 345
     },
     {
-      "epoch": 4.119047619047619,
-      "grad_norm": 1.4220143614732972,
-      "learning_rate": 1.5294040776021092e-05,
-      "loss": 0.1782,
+      "epoch": 5.815126050420168,
+      "grad_norm": 2.016326952399232,
+      "learning_rate": 1.0869022869052091e-05,
+      "loss": 0.0852,
       "step": 346
     },
     {
-      "epoch": 4.130952380952381,
-      "grad_norm": 2.175470210456222,
-      "learning_rate": 1.5266733695653998e-05,
-      "loss": 0.181,
+      "epoch": 5.831932773109243,
+      "grad_norm": 1.502892225736369,
+      "learning_rate": 1.0823391357318876e-05,
+      "loss": 0.0637,
       "step": 347
     },
     {
-      "epoch": 4.142857142857143,
-      "grad_norm": 2.3599094937304463,
-      "learning_rate": 1.523937215855141e-05,
-      "loss": 0.2284,
+      "epoch": 5.848739495798319,
+      "grad_norm": 2.2571728633731376,
+      "learning_rate": 1.0777742576988474e-05,
+      "loss": 0.0667,
       "step": 348
     },
     {
-      "epoch": 4.154761904761905,
-      "grad_norm": 3.1602264124779587,
-      "learning_rate": 1.5211956447624916e-05,
-      "loss": 0.3619,
+      "epoch": 5.865546218487395,
+      "grad_norm": 2.3431094429758264,
+      "learning_rate": 1.0732077485431152e-05,
+      "loss": 0.0984,
       "step": 349
     },
     {
-      "epoch": 4.166666666666667,
-      "grad_norm": 2.530357790936566,
-      "learning_rate": 1.5184486846346248e-05,
-      "loss": 0.0911,
+      "epoch": 5.882352941176471,
+      "grad_norm": 1.5381903109004802,
+      "learning_rate": 1.0686397040359253e-05,
+      "loss": 0.0712,
       "step": 350
     },
     {
-      "epoch": 4.178571428571429,
-      "grad_norm": 5.786049264827654,
-      "learning_rate": 1.5156963638744348e-05,
-      "loss": 0.2455,
+      "epoch": 5.899159663865547,
+      "grad_norm": 1.36348896527395,
+      "learning_rate": 1.064070219980713e-05,
+      "loss": 0.0736,
       "step": 351
     },
     {
-      "epoch": 4.190476190476191,
-      "grad_norm": 1.632349364082718,
-      "learning_rate": 1.512938710940244e-05,
-      "loss": 0.0994,
+      "epoch": 5.915966386554622,
+      "grad_norm": 2.8850647864954384,
+      "learning_rate": 1.059499392211105e-05,
+      "loss": 0.1959,
       "step": 352
     },
     {
-      "epoch": 4.2023809523809526,
-      "grad_norm": 3.180319196728077,
-      "learning_rate": 1.5101757543455074e-05,
-      "loss": 0.1107,
+      "epoch": 5.932773109243698,
+      "grad_norm": 2.2743426048039814,
+      "learning_rate": 1.0549273165889079e-05,
+      "loss": 0.0661,
       "step": 353
     },
     {
-      "epoch": 4.214285714285714,
-      "grad_norm": 1.4203335804196207,
-      "learning_rate": 1.5074075226585195e-05,
-      "loss": 0.1874,
+      "epoch": 5.949579831932773,
+      "grad_norm": 2.1602273939968106,
+      "learning_rate": 1.0503540890020997e-05,
+      "loss": 0.0978,
       "step": 354
     },
     {
-      "epoch": 4.226190476190476,
-      "grad_norm": 2.1086034749298084,
-      "learning_rate": 1.5046340445021171e-05,
-      "loss": 0.1564,
+      "epoch": 5.966386554621849,
+      "grad_norm": 1.6282000235425853,
+      "learning_rate": 1.0457798053628181e-05,
+      "loss": 0.0712,
       "step": 355
     },
     {
-      "epoch": 4.238095238095238,
-      "grad_norm": 2.3264426202911226,
-      "learning_rate": 1.5018553485533844e-05,
-      "loss": 0.2398,
+      "epoch": 5.983193277310924,
+      "grad_norm": 2.2108120483379072,
+      "learning_rate": 1.0412045616053486e-05,
+      "loss": 0.0959,
       "step": 356
     },
     {
-      "epoch": 4.25,
-      "grad_norm": 3.0488496848457394,
-      "learning_rate": 1.4990714635433568e-05,
-      "loss": 0.1081,
+      "epoch": 6.0,
+      "grad_norm": 6.96305244714079,
+      "learning_rate": 1.0366284536841124e-05,
+      "loss": 0.1132,
+      "step": 357
+    },
+    {
+      "epoch": 6.0,
+      "eval_loss": 0.1090986505150795,
+      "eval_runtime": 95.9195,
+      "eval_samples_per_second": 0.98,
+      "eval_steps_per_second": 0.98,
       "step": 357
     },
     {
-      "epoch": 4.261904761904762,
-      "grad_norm": 1.7394283118255627,
-      "learning_rate": 1.4962824182567226e-05,
-      "loss": 0.1901,
+      "epoch": 6.016806722689076,
+      "grad_norm": 1.4364004042145684,
+      "learning_rate": 1.0320515775716556e-05,
+      "loss": 0.0628,
       "step": 358
     },
     {
-      "epoch": 4.273809523809524,
-      "grad_norm": 3.3515867531674903,
-      "learning_rate": 1.4934882415315266e-05,
-      "loss": 0.185,
+      "epoch": 6.033613445378151,
+      "grad_norm": 1.9184329896408099,
+      "learning_rate": 1.0274740292566335e-05,
+      "loss": 0.085,
       "step": 359
     },
     {
-      "epoch": 4.285714285714286,
-      "grad_norm": 18.82745619545086,
-      "learning_rate": 1.4906889622588714e-05,
-      "loss": 0.2833,
+      "epoch": 6.050420168067227,
+      "grad_norm": 2.722228602616158,
+      "learning_rate": 1.0228959047418005e-05,
+      "loss": 0.0764,
       "step": 360
     },
     {
-      "epoch": 4.2976190476190474,
-      "grad_norm": 4.511322420570053,
-      "learning_rate": 1.487884609382618e-05,
-      "loss": 0.1722,
+      "epoch": 6.067226890756302,
+      "grad_norm": 1.9487119713360619,
+      "learning_rate": 1.0183173000419954e-05,
+      "loss": 0.0735,
       "step": 361
     },
     {
-      "epoch": 4.309523809523809,
-      "grad_norm": 1.533278706530758,
-      "learning_rate": 1.4850752118990884e-05,
-      "loss": 0.1364,
+      "epoch": 6.084033613445378,
+      "grad_norm": 2.4618304983550883,
+      "learning_rate": 1.0137383111821267e-05,
+      "loss": 0.0742,
       "step": 362
     },
     {
-      "epoch": 4.321428571428571,
-      "grad_norm": 8.629338884395395,
-      "learning_rate": 1.4822607988567637e-05,
-      "loss": 0.123,
+      "epoch": 6.100840336134453,
+      "grad_norm": 3.82982664079687,
+      "learning_rate": 1.009159034195161e-05,
+      "loss": 0.1549,
       "step": 363
     },
     {
-      "epoch": 4.333333333333333,
-      "grad_norm": 3.0735320587228085,
-      "learning_rate": 1.4794413993559851e-05,
-      "loss": 0.0625,
+      "epoch": 6.117647058823529,
+      "grad_norm": 2.5393832652005686,
+      "learning_rate": 1.0045795651201062e-05,
+      "loss": 0.072,
       "step": 364
     },
     {
-      "epoch": 4.345238095238095,
-      "grad_norm": 2.0992688835189472,
-      "learning_rate": 1.476617042548652e-05,
-      "loss": 0.119,
+      "epoch": 6.1344537815126055,
+      "grad_norm": 2.1988376418173305,
+      "learning_rate": 1e-05,
+      "loss": 0.0786,
       "step": 365
     },
     {
-      "epoch": 4.357142857142857,
-      "grad_norm": 6.415048214826861,
-      "learning_rate": 1.473787757637922e-05,
-      "loss": 0.2251,
+      "epoch": 6.151260504201681,
+      "grad_norm": 3.668071708848647,
+      "learning_rate": 9.954204348798938e-06,
+      "loss": 0.0928,
       "step": 366
     },
     {
-      "epoch": 4.369047619047619,
-      "grad_norm": 1.6932021512404514,
-      "learning_rate": 1.4709535738779075e-05,
-      "loss": 0.1117,
+      "epoch": 6.168067226890757,
+      "grad_norm": 2.378790242128915,
+      "learning_rate": 9.908409658048395e-06,
+      "loss": 0.0717,
       "step": 367
     },
     {
-      "epoch": 4.380952380952381,
-      "grad_norm": 2.9272562546997047,
-      "learning_rate": 1.4681145205733736e-05,
-      "loss": 0.0407,
+      "epoch": 6.184873949579832,
+      "grad_norm": 1.8836633579149347,
+      "learning_rate": 9.862616888178733e-06,
+      "loss": 0.0519,
       "step": 368
     },
     {
-      "epoch": 4.392857142857143,
-      "grad_norm": 0.8387686118428367,
-      "learning_rate": 1.4652706270794354e-05,
-      "loss": 0.068,
+      "epoch": 6.201680672268908,
+      "grad_norm": 2.181673653848983,
+      "learning_rate": 9.816826999580049e-06,
+      "loss": 0.0748,
       "step": 369
     },
     {
-      "epoch": 4.404761904761905,
-      "grad_norm": 0.9780904318278195,
-      "learning_rate": 1.4624219228012555e-05,
-      "loss": 0.0767,
+      "epoch": 6.218487394957983,
+      "grad_norm": 2.544264789630926,
+      "learning_rate": 9.771040952581998e-06,
+      "loss": 0.1029,
       "step": 370
     },
     {
-      "epoch": 4.416666666666667,
-      "grad_norm": 2.4997987914726894,
-      "learning_rate": 1.4595684371937369e-05,
-      "loss": 0.1666,
+      "epoch": 6.235294117647059,
+      "grad_norm": 2.673802179214641,
+      "learning_rate": 9.72525970743367e-06,
+      "loss": 0.0857,
       "step": 371
     },
     {
-      "epoch": 4.428571428571429,
-      "grad_norm": 7.25267339982332,
-      "learning_rate": 1.4567101997612216e-05,
-      "loss": 0.1855,
+      "epoch": 6.2521008403361344,
+      "grad_norm": 2.601200264033827,
+      "learning_rate": 9.67948422428345e-06,
+      "loss": 0.0886,
       "step": 372
     },
     {
-      "epoch": 4.440476190476191,
-      "grad_norm": 3.059862921630944,
-      "learning_rate": 1.453847240057185e-05,
-      "loss": 0.1155,
+      "epoch": 6.26890756302521,
+      "grad_norm": 1.9627943528177962,
+      "learning_rate": 9.633715463158881e-06,
+      "loss": 0.1016,
       "step": 373
     },
     {
-      "epoch": 4.4523809523809526,
-      "grad_norm": 0.940939205370957,
-      "learning_rate": 1.4509795876839274e-05,
-      "loss": 0.0152,
+      "epoch": 6.285714285714286,
+      "grad_norm": 1.0216877914513534,
+      "learning_rate": 9.587954383946518e-06,
+      "loss": 0.0672,
       "step": 374
     },
     {
-      "epoch": 4.464285714285714,
-      "grad_norm": 0.7709486540869183,
-      "learning_rate": 1.4481072722922721e-05,
-      "loss": 0.0702,
+      "epoch": 6.302521008403361,
+      "grad_norm": 1.8373707532976224,
+      "learning_rate": 9.542201946371819e-06,
+      "loss": 0.0754,
       "step": 375
     },
     {
-      "epoch": 4.476190476190476,
-      "grad_norm": 1.5326012961960709,
-      "learning_rate": 1.4452303235812562e-05,
-      "loss": 0.1224,
+      "epoch": 6.319327731092437,
+      "grad_norm": 0.9199770052216976,
+      "learning_rate": 9.496459109979004e-06,
+      "loss": 0.0655,
       "step": 376
     },
     {
-      "epoch": 4.488095238095238,
-      "grad_norm": 2.0531110608488072,
-      "learning_rate": 1.442348771297824e-05,
-      "loss": 0.2465,
+      "epoch": 6.336134453781512,
+      "grad_norm": 1.9486911830327862,
+      "learning_rate": 9.450726834110923e-06,
+      "loss": 0.0682,
       "step": 377
     },
     {
-      "epoch": 4.5,
-      "grad_norm": 0.9970189018607668,
-      "learning_rate": 1.4394626452365202e-05,
-      "loss": 0.0771,
+      "epoch": 6.352941176470588,
+      "grad_norm": 1.4805691280913602,
+      "learning_rate": 9.405006077888954e-06,
+      "loss": 0.0487,
       "step": 378
     },
     {
-      "epoch": 4.511904761904762,
-      "grad_norm": 1.1650153863777009,
-      "learning_rate": 1.4365719752391805e-05,
-      "loss": 0.0153,
+      "epoch": 6.369747899159664,
+      "grad_norm": 1.4042709384224348,
+      "learning_rate": 9.359297800192873e-06,
+      "loss": 0.0689,
       "step": 379
     },
     {
-      "epoch": 4.523809523809524,
-      "grad_norm": 1.79896147766944,
-      "learning_rate": 1.4336767911946238e-05,
-      "loss": 0.0211,
+      "epoch": 6.38655462184874,
+      "grad_norm": 3.6158792324022775,
+      "learning_rate": 9.313602959640754e-06,
+      "loss": 0.1121,
       "step": 380
     },
     {
-      "epoch": 4.535714285714286,
-      "grad_norm": 8.039019319027208,
-      "learning_rate": 1.4307771230383446e-05,
-      "loss": 0.2942,
+      "epoch": 6.4033613445378155,
+      "grad_norm": 1.5349914188342018,
+      "learning_rate": 9.267922514568853e-06,
+      "loss": 0.0548,
       "step": 381
     },
     {
-      "epoch": 4.5476190476190474,
-      "grad_norm": 2.2623133990723274,
-      "learning_rate": 1.4278730007521996e-05,
-      "loss": 0.2017,
+      "epoch": 6.420168067226891,
+      "grad_norm": 5.196823890519191,
+      "learning_rate": 9.22225742301153e-06,
+      "loss": 0.1468,
       "step": 382
     },
     {
-      "epoch": 4.559523809523809,
-      "grad_norm": 0.7549719818461573,
-      "learning_rate": 1.4249644543641026e-05,
-      "loss": 0.0121,
+      "epoch": 6.436974789915967,
+      "grad_norm": 2.4164444219519816,
+      "learning_rate": 9.176608642681127e-06,
+      "loss": 0.0689,
       "step": 383
     },
     {
-      "epoch": 4.571428571428571,
-      "grad_norm": 1.6291253889342527,
-      "learning_rate": 1.4220515139477099e-05,
-      "loss": 0.1301,
+      "epoch": 6.453781512605042,
+      "grad_norm": 1.0767734429156512,
+      "learning_rate": 9.13097713094791e-06,
+      "loss": 0.0537,
       "step": 384
     },
     {
-      "epoch": 4.583333333333333,
-      "grad_norm": 1.3753384069531984,
-      "learning_rate": 1.419134209622111e-05,
-      "loss": 0.0843,
+      "epoch": 6.470588235294118,
+      "grad_norm": 1.1364625205117547,
+      "learning_rate": 9.085363844819979e-06,
+      "loss": 0.0505,
       "step": 385
     },
     {
-      "epoch": 4.595238095238095,
-      "grad_norm": 16.083102720375773,
-      "learning_rate": 1.416212571551518e-05,
-      "loss": 0.2996,
+      "epoch": 6.487394957983193,
+      "grad_norm": 1.2535232324818428,
+      "learning_rate": 9.039769740923183e-06,
+      "loss": 0.0518,
       "step": 386
     },
     {
-      "epoch": 4.607142857142857,
-      "grad_norm": 2.4053795337835817,
-      "learning_rate": 1.4132866299449523e-05,
-      "loss": 0.2103,
+      "epoch": 6.504201680672269,
+      "grad_norm": 3.057282990063465,
+      "learning_rate": 8.99419577548109e-06,
+      "loss": 0.1124,
       "step": 387
     },
     {
-      "epoch": 4.619047619047619,
-      "grad_norm": 2.1608334893802668,
-      "learning_rate": 1.410356415055933e-05,
-      "loss": 0.1842,
+      "epoch": 6.5210084033613445,
+      "grad_norm": 1.9870672030743288,
+      "learning_rate": 8.948642904294901e-06,
+      "loss": 0.0534,
       "step": 388
     },
     {
-      "epoch": 4.630952380952381,
-      "grad_norm": 1.6675074064641335,
-      "learning_rate": 1.407421957182164e-05,
-      "loss": 0.1343,
+      "epoch": 6.53781512605042,
+      "grad_norm": 4.506177944111496,
+      "learning_rate": 8.90311208272342e-06,
+      "loss": 0.1007,
       "step": 389
     },
     {
-      "epoch": 4.642857142857143,
-      "grad_norm": 7.385252787773213,
-      "learning_rate": 1.4044832866652198e-05,
-      "loss": 0.0953,
+      "epoch": 6.554621848739496,
+      "grad_norm": 4.687056129955279,
+      "learning_rate": 8.857604265663016e-06,
+      "loss": 0.1015,
       "step": 390
     },
     {
-      "epoch": 4.654761904761905,
-      "grad_norm": 13.223392660691903,
-      "learning_rate": 1.4015404338902338e-05,
-      "loss": 0.2051,
+      "epoch": 6.571428571428571,
+      "grad_norm": 2.793771357084546,
+      "learning_rate": 8.812120407527603e-06,
+      "loss": 0.0791,
       "step": 391
     },
     {
-      "epoch": 4.666666666666667,
-      "grad_norm": 1.7997237114139486,
-      "learning_rate": 1.3985934292855827e-05,
-      "loss": 0.0745,
+      "epoch": 6.588235294117647,
+      "grad_norm": 1.5631834583542967,
+      "learning_rate": 8.766661462228593e-06,
+      "loss": 0.0583,
       "step": 392
     },
     {
-      "epoch": 4.678571428571429,
-      "grad_norm": 1.7124646994476471,
-      "learning_rate": 1.3956423033225713e-05,
-      "loss": 0.14,
+      "epoch": 6.605042016806722,
+      "grad_norm": 2.503384804703961,
+      "learning_rate": 8.721228383154939e-06,
+      "loss": 0.0819,
       "step": 393
     },
     {
-      "epoch": 4.690476190476191,
-      "grad_norm": 1.5626636243483676,
-      "learning_rate": 1.3926870865151185e-05,
-      "loss": 0.037,
+      "epoch": 6.621848739495798,
+      "grad_norm": 1.298763753508359,
+      "learning_rate": 8.675822123153104e-06,
+      "loss": 0.0452,
       "step": 394
     },
     {
-      "epoch": 4.7023809523809526,
-      "grad_norm": 1.8377267092001748,
-      "learning_rate": 1.3897278094194422e-05,
-      "loss": 0.149,
+      "epoch": 6.6386554621848735,
+      "grad_norm": 3.468345102178359,
+      "learning_rate": 8.630443634507077e-06,
+      "loss": 0.0893,
       "step": 395
     },
     {
-      "epoch": 4.714285714285714,
-      "grad_norm": 1.1393534152121498,
-      "learning_rate": 1.386764502633742e-05,
-      "loss": 0.1064,
+      "epoch": 6.65546218487395,
+      "grad_norm": 2.526122457983386,
+      "learning_rate": 8.585093868918426e-06,
+      "loss": 0.1011,
       "step": 396
     },
     {
-      "epoch": 4.726190476190476,
-      "grad_norm": 5.6011749497367855,
-      "learning_rate": 1.3837971967978836e-05,
-      "loss": 0.1539,
+      "epoch": 6.6722689075630255,
+      "grad_norm": 2.075764833700991,
+      "learning_rate": 8.539773777486321e-06,
+      "loss": 0.0752,
       "step": 397
     },
     {
-      "epoch": 4.738095238095238,
-      "grad_norm": 1.6182850109100042,
-      "learning_rate": 1.3808259225930825e-05,
-      "loss": 0.1354,
+      "epoch": 6.689075630252101,
+      "grad_norm": 1.3259731049756247,
+      "learning_rate": 8.494484310687581e-06,
+      "loss": 0.0536,
       "step": 398
     },
     {
-      "epoch": 4.75,
-      "grad_norm": 2.664012777423232,
-      "learning_rate": 1.3778507107415849e-05,
-      "loss": 0.0815,
+      "epoch": 6.705882352941177,
+      "grad_norm": 1.5262707963177293,
+      "learning_rate": 8.44922641835676e-06,
+      "loss": 0.0649,
       "step": 399
     },
     {
-      "epoch": 4.761904761904762,
-      "grad_norm": 6.413333811764225,
-      "learning_rate": 1.374871592006353e-05,
-      "loss": 0.1265,
+      "epoch": 6.722689075630252,
+      "grad_norm": 1.3923407644979977,
+      "learning_rate": 8.404001049666211e-06,
+      "loss": 0.0781,
       "step": 400
     },
     {
-      "epoch": 4.773809523809524,
-      "grad_norm": 1.5093904430192655,
-      "learning_rate": 1.371888597190744e-05,
-      "loss": 0.1129,
+      "epoch": 6.739495798319328,
+      "grad_norm": 1.474705597824637,
+      "learning_rate": 8.35880915310618e-06,
+      "loss": 0.0614,
       "step": 401
     },
     {
-      "epoch": 4.785714285714286,
-      "grad_norm": 1.1398541908667996,
-      "learning_rate": 1.3689017571381928e-05,
-      "loss": 0.1792,
+      "epoch": 6.756302521008403,
+      "grad_norm": 1.5133797172950414,
+      "learning_rate": 8.313651676464924e-06,
+      "loss": 0.0588,
       "step": 402
     },
     {
-      "epoch": 4.7976190476190474,
-      "grad_norm": 4.7568958126507495,
-      "learning_rate": 1.3659111027318936e-05,
-      "loss": 0.2022,
+      "epoch": 6.773109243697479,
+      "grad_norm": 1.5937067083804244,
+      "learning_rate": 8.26852956680883e-06,
+      "loss": 0.0658,
       "step": 403
     },
     {
-      "epoch": 4.809523809523809,
-      "grad_norm": 1.1634943614210518,
-      "learning_rate": 1.3629166648944803e-05,
-      "loss": 0.1091,
+      "epoch": 6.7899159663865545,
+      "grad_norm": 3.2146725324389447,
+      "learning_rate": 8.223443770462539e-06,
+      "loss": 0.0967,
       "step": 404
     },
     {
-      "epoch": 4.821428571428571,
-      "grad_norm": 2.0853097817874096,
-      "learning_rate": 1.3599184745877059e-05,
-      "loss": 0.125,
+      "epoch": 6.80672268907563,
+      "grad_norm": 1.6054189964924985,
+      "learning_rate": 8.17839523298912e-06,
+      "loss": 0.0676,
       "step": 405
     },
     {
-      "epoch": 4.833333333333333,
-      "grad_norm": 7.354405370327422,
-      "learning_rate": 1.356916562812123e-05,
-      "loss": 0.131,
+      "epoch": 6.823529411764706,
+      "grad_norm": 1.3282376180631894,
+      "learning_rate": 8.133384899170224e-06,
+      "loss": 0.0373,
       "step": 406
     },
     {
-      "epoch": 4.845238095238095,
-      "grad_norm": 2.2454476121472693,
-      "learning_rate": 1.3539109606067642e-05,
-      "loss": 0.1625,
+      "epoch": 6.840336134453781,
+      "grad_norm": 1.1953714752813585,
+      "learning_rate": 8.08841371298628e-06,
+      "loss": 0.0591,
       "step": 407
     },
     {
-      "epoch": 4.857142857142857,
-      "grad_norm": 1.0859281248071477,
-      "learning_rate": 1.3509016990488191e-05,
-      "loss": 0.1148,
+      "epoch": 6.857142857142857,
+      "grad_norm": 2.9661126814818393,
+      "learning_rate": 8.043482617596681e-06,
+      "loss": 0.141,
       "step": 408
     },
     {
-      "epoch": 4.869047619047619,
-      "grad_norm": 1.8112995115739032,
-      "learning_rate": 1.3478888092533145e-05,
-      "loss": 0.1333,
+      "epoch": 6.873949579831933,
+      "grad_norm": 1.3266256289537959,
+      "learning_rate": 7.99859255532003e-06,
+      "loss": 0.053,
       "step": 409
     },
     {
-      "epoch": 4.880952380952381,
-      "grad_norm": 6.093790266627286,
-      "learning_rate": 1.3448723223727929e-05,
-      "loss": 0.1594,
+      "epoch": 6.890756302521009,
+      "grad_norm": 2.1914466357984574,
+      "learning_rate": 7.953744467614356e-06,
+      "loss": 0.0778,
       "step": 410
     },
     {
-      "epoch": 4.892857142857143,
-      "grad_norm": 4.418732377785542,
-      "learning_rate": 1.3418522695969892e-05,
-      "loss": 0.1858,
+      "epoch": 6.907563025210084,
+      "grad_norm": 4.698036043159151,
+      "learning_rate": 7.908939295057362e-06,
+      "loss": 0.1673,
       "step": 411
     },
     {
-      "epoch": 4.904761904761905,
-      "grad_norm": 4.690321231618584,
-      "learning_rate": 1.3388286821525086e-05,
-      "loss": 0.0972,
+      "epoch": 6.92436974789916,
+      "grad_norm": 2.0220996669989555,
+      "learning_rate": 7.864177977326739e-06,
+      "loss": 0.0706,
       "step": 412
     },
     {
-      "epoch": 4.916666666666667,
-      "grad_norm": 2.671198917779827,
-      "learning_rate": 1.3358015913025045e-05,
-      "loss": 0.1115,
+      "epoch": 6.9411764705882355,
+      "grad_norm": 0.863174779051157,
+      "learning_rate": 7.819461453180403e-06,
+      "loss": 0.0536,
       "step": 413
     },
     {
-      "epoch": 4.928571428571429,
-      "grad_norm": 2.3833023595529723,
-      "learning_rate": 1.3327710283463535e-05,
-      "loss": 0.1534,
+      "epoch": 6.957983193277311,
+      "grad_norm": 1.162974910961646,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.053,
       "step": 414
     },
     {
-      "epoch": 4.940476190476191,
-      "grad_norm": 1.626651095857683,
-      "learning_rate": 1.3297370246193349e-05,
-      "loss": 0.0785,
+      "epoch": 6.974789915966387,
+      "grad_norm": 1.5401943180276223,
+      "learning_rate": 7.730166535955489e-06,
+      "loss": 0.0562,
       "step": 415
     },
     {
-      "epoch": 4.9523809523809526,
-      "grad_norm": 3.184821297784408,
-      "learning_rate": 1.3266996114923027e-05,
-      "loss": 0.1509,
+      "epoch": 6.991596638655462,
+      "grad_norm": 1.0451220578083906,
+      "learning_rate": 7.685590015616939e-06,
+      "loss": 0.061,
+      "step": 416
+    },
+    {
+      "epoch": 6.991596638655462,
+      "eval_loss": 0.11021654307842255,
+      "eval_runtime": 96.6092,
+      "eval_samples_per_second": 0.973,
+      "eval_steps_per_second": 0.973,
       "step": 416
     },
     {
-      "epoch": 4.964285714285714,
-      "grad_norm": 2.7515690992856445,
-      "learning_rate": 1.3236588203713645e-05,
-      "loss": 0.0581,
+      "epoch": 7.008403361344538,
+      "grad_norm": 1.9748689801680077,
+      "learning_rate": 7.641062034303464e-06,
+      "loss": 0.0744,
       "step": 417
     },
     {
-      "epoch": 4.976190476190476,
-      "grad_norm": 1.581027664457046,
-      "learning_rate": 1.3206146826975554e-05,
-      "loss": 0.1458,
+      "epoch": 7.025210084033613,
+      "grad_norm": 1.407098109780849,
+      "learning_rate": 7.596583525879344e-06,
+      "loss": 0.0575,
       "step": 418
     },
     {
-      "epoch": 4.988095238095238,
-      "grad_norm": 3.227099328398388,
-      "learning_rate": 1.3175672299465124e-05,
-      "loss": 0.0788,
+      "epoch": 7.042016806722689,
+      "grad_norm": 3.002224281961417,
+      "learning_rate": 7.5521554231712845e-06,
+      "loss": 0.0761,
       "step": 419
     },
     {
-      "epoch": 5.0,
-      "grad_norm": 2.307032199226437,
-      "learning_rate": 1.3145164936281501e-05,
-      "loss": 0.1162,
-      "step": 420
-    },
-    {
-      "epoch": 5.0,
-      "eval_loss": 0.16767826676368713,
-      "eval_runtime": 38.2511,
-      "eval_samples_per_second": 1.569,
-      "eval_steps_per_second": 1.569,
+      "epoch": 7.0588235294117645,
+      "grad_norm": 1.3836323412125109,
+      "learning_rate": 7.507778657948847e-06,
+      "loss": 0.0508,
       "step": 420
     },
     {
-      "epoch": 5.011904761904762,
-      "grad_norm": 2.480615008589385,
-      "learning_rate": 1.3114625052863343e-05,
-      "loss": 0.1365,
+      "epoch": 7.07563025210084,
+      "grad_norm": 1.5627121625333806,
+      "learning_rate": 7.463454160904928e-06,
+      "loss": 0.0619,
       "step": 421
     },
     {
-      "epoch": 5.023809523809524,
-      "grad_norm": 1.7186104572642102,
-      "learning_rate": 1.3084052964985559e-05,
-      "loss": 0.1,
+      "epoch": 7.092436974789916,
+      "grad_norm": 1.7887521098557067,
+      "learning_rate": 7.419182861636218e-06,
+      "loss": 0.0566,
       "step": 422
     },
     {
-      "epoch": 5.035714285714286,
-      "grad_norm": 1.0422770723266574,
-      "learning_rate": 1.3053448988756046e-05,
-      "loss": 0.1036,
+      "epoch": 7.109243697478991,
+      "grad_norm": 1.6937210033771022,
+      "learning_rate": 7.374965688623726e-06,
+      "loss": 0.0601,
       "step": 423
     },
     {
-      "epoch": 5.0476190476190474,
-      "grad_norm": 1.3346688546661123,
-      "learning_rate": 1.3022813440612411e-05,
-      "loss": 0.1213,
+      "epoch": 7.126050420168067,
+      "grad_norm": 2.35276516070813,
+      "learning_rate": 7.3308035692132896e-06,
+      "loss": 0.0778,
       "step": 424
     },
     {
-      "epoch": 5.059523809523809,
-      "grad_norm": 1.8542917766521358,
-      "learning_rate": 1.2992146637318721e-05,
-      "loss": 0.0678,
+      "epoch": 7.142857142857143,
+      "grad_norm": 1.7638699741053976,
+      "learning_rate": 7.286697429596135e-06,
+      "loss": 0.0622,
       "step": 425
     },
     {
-      "epoch": 5.071428571428571,
-      "grad_norm": 5.6008497642741855,
-      "learning_rate": 1.2961448895962199e-05,
-      "loss": 0.1197,
+      "epoch": 7.159663865546219,
+      "grad_norm": 2.20424078040461,
+      "learning_rate": 7.242648194789447e-06,
+      "loss": 0.0597,
       "step": 426
     },
     {
-      "epoch": 5.083333333333333,
-      "grad_norm": 3.4136608303525633,
-      "learning_rate": 1.2930720533949967e-05,
-      "loss": 0.2551,
+      "epoch": 7.176470588235294,
+      "grad_norm": 2.50814340199316,
+      "learning_rate": 7.19865678861698e-06,
+      "loss": 0.0686,
       "step": 427
     },
     {
-      "epoch": 5.095238095238095,
-      "grad_norm": 2.3745380363755166,
-      "learning_rate": 1.2899961869005761e-05,
-      "loss": 0.1615,
+      "epoch": 7.19327731092437,
+      "grad_norm": 1.448580373256141,
+      "learning_rate": 7.154724133689677e-06,
+      "loss": 0.0312,
       "step": 428
     },
     {
-      "epoch": 5.107142857142857,
-      "grad_norm": 1.8910210447295193,
-      "learning_rate": 1.2869173219166632e-05,
-      "loss": 0.0842,
+      "epoch": 7.2100840336134455,
+      "grad_norm": 2.1552454847239075,
+      "learning_rate": 7.110851151386306e-06,
+      "loss": 0.0547,
       "step": 429
     },
     {
-      "epoch": 5.119047619047619,
-      "grad_norm": 2.9450829681087742,
-      "learning_rate": 1.283835490277968e-05,
-      "loss": 0.1633,
+      "epoch": 7.226890756302521,
+      "grad_norm": 2.9974503334538602,
+      "learning_rate": 7.067038761834164e-06,
+      "loss": 0.0881,
       "step": 430
     },
     {
-      "epoch": 5.130952380952381,
-      "grad_norm": 6.731369409179587,
-      "learning_rate": 1.2807507238498733e-05,
-      "loss": 0.1296,
+      "epoch": 7.243697478991597,
+      "grad_norm": 0.9735826198652626,
+      "learning_rate": 7.023287883889753e-06,
+      "loss": 0.039,
       "step": 431
     },
     {
-      "epoch": 5.142857142857143,
-      "grad_norm": 2.9173972705274034,
-      "learning_rate": 1.2776630545281088e-05,
-      "loss": 0.156,
+      "epoch": 7.260504201680672,
+      "grad_norm": 2.5579376905692293,
+      "learning_rate": 6.979599435119531e-06,
+      "loss": 0.0919,
       "step": 432
     },
     {
-      "epoch": 5.154761904761905,
-      "grad_norm": 1.4766409198526602,
-      "learning_rate": 1.2745725142384176e-05,
-      "loss": 0.1113,
+      "epoch": 7.277310924369748,
+      "grad_norm": 2.4454794736196,
+      "learning_rate": 6.935974331780653e-06,
+      "loss": 0.0544,
       "step": 433
     },
     {
-      "epoch": 5.166666666666667,
-      "grad_norm": 5.5619564739728915,
-      "learning_rate": 1.2714791349362293e-05,
-      "loss": 0.0989,
+      "epoch": 7.294117647058823,
+      "grad_norm": 1.4023653427232834,
+      "learning_rate": 6.892413488801762e-06,
+      "loss": 0.0626,
       "step": 434
     },
     {
-      "epoch": 5.178571428571429,
-      "grad_norm": 4.883941398663358,
-      "learning_rate": 1.2683829486063275e-05,
-      "loss": 0.0922,
+      "epoch": 7.310924369747899,
+      "grad_norm": 1.36640843494829,
+      "learning_rate": 6.848917819763794e-06,
+      "loss": 0.061,
       "step": 435
     },
     {
-      "epoch": 5.190476190476191,
-      "grad_norm": 2.153822102950277,
-      "learning_rate": 1.2652839872625198e-05,
-      "loss": 0.1536,
+      "epoch": 7.3277310924369745,
+      "grad_norm": 1.6526923758851404,
+      "learning_rate": 6.805488236880831e-06,
+      "loss": 0.0628,
       "step": 436
     },
     {
-      "epoch": 5.2023809523809526,
-      "grad_norm": 1.550572046017145,
-      "learning_rate": 1.2621822829473074e-05,
-      "loss": 0.0688,
+      "epoch": 7.34453781512605,
+      "grad_norm": 1.5372290912683437,
+      "learning_rate": 6.76212565098096e-06,
+      "loss": 0.06,
       "step": 437
     },
     {
-      "epoch": 5.214285714285714,
-      "grad_norm": 2.7372539541607512,
-      "learning_rate": 1.2590778677315525e-05,
-      "loss": 0.0374,
+      "epoch": 7.361344537815126,
+      "grad_norm": 1.1019297711397662,
+      "learning_rate": 6.718830971487165e-06,
+      "loss": 0.0534,
       "step": 438
     },
     {
-      "epoch": 5.226190476190476,
-      "grad_norm": 1.760106704960083,
-      "learning_rate": 1.2559707737141471e-05,
-      "loss": 0.1008,
+      "epoch": 7.378151260504202,
+      "grad_norm": 1.474792358247268,
+      "learning_rate": 6.675605106398269e-06,
+      "loss": 0.0498,
       "step": 439
     },
     {
-      "epoch": 5.238095238095238,
-      "grad_norm": 2.264297339685535,
-      "learning_rate": 1.2528610330216824e-05,
-      "loss": 0.1123,
+      "epoch": 7.394957983193278,
+      "grad_norm": 1.5934869270544132,
+      "learning_rate": 6.632448962269873e-06,
+      "loss": 0.0579,
       "step": 440
     },
     {
-      "epoch": 5.25,
-      "grad_norm": 1.9709076777699328,
-      "learning_rate": 1.2497486778081145e-05,
-      "loss": 0.0664,
+      "epoch": 7.411764705882353,
+      "grad_norm": 1.5423848913860474,
+      "learning_rate": 6.589363444195367e-06,
+      "loss": 0.0836,
       "step": 441
     },
     {
-      "epoch": 5.261904761904762,
-      "grad_norm": 1.9643428286751554,
-      "learning_rate": 1.2466337402544333e-05,
-      "loss": 0.0964,
+      "epoch": 7.428571428571429,
+      "grad_norm": 0.6052937489588177,
+      "learning_rate": 6.546349455786926e-06,
+      "loss": 0.0235,
       "step": 442
     },
     {
-      "epoch": 5.273809523809524,
-      "grad_norm": 1.300317642643842,
-      "learning_rate": 1.2435162525683303e-05,
-      "loss": 0.0798,
+      "epoch": 7.445378151260504,
+      "grad_norm": 0.8930183656715149,
+      "learning_rate": 6.503407899156565e-06,
+      "loss": 0.0536,
       "step": 443
     },
     {
-      "epoch": 5.285714285714286,
-      "grad_norm": 7.741590751845092,
-      "learning_rate": 1.2403962469838633e-05,
-      "loss": 0.1114,
+      "epoch": 7.46218487394958,
+      "grad_norm": 1.9378732272470842,
+      "learning_rate": 6.460539674897226e-06,
+      "loss": 0.0713,
       "step": 444
     },
     {
-      "epoch": 5.2976190476190474,
-      "grad_norm": 1.257936578683669,
-      "learning_rate": 1.2372737557611256e-05,
-      "loss": 0.0165,
+      "epoch": 7.4789915966386555,
+      "grad_norm": 0.9653273021580584,
+      "learning_rate": 6.417745682063884e-06,
+      "loss": 0.0572,
       "step": 445
     },
     {
-      "epoch": 5.309523809523809,
-      "grad_norm": 1.5441967809002295,
-      "learning_rate": 1.2341488111859111e-05,
-      "loss": 0.095,
+      "epoch": 7.495798319327731,
+      "grad_norm": 5.150704023745161,
+      "learning_rate": 6.3750268181547e-06,
+      "loss": 0.1011,
       "step": 446
     },
     {
-      "epoch": 5.321428571428571,
-      "grad_norm": 4.463551970591222,
-      "learning_rate": 1.2310214455693808e-05,
-      "loss": 0.0405,
+      "epoch": 7.512605042016807,
+      "grad_norm": 2.663053864822168,
+      "learning_rate": 6.3323839790921785e-06,
+      "loss": 0.1071,
       "step": 447
     },
     {
-      "epoch": 5.333333333333333,
-      "grad_norm": 3.676401909989546,
-      "learning_rate": 1.2278916912477285e-05,
-      "loss": 0.0297,
+      "epoch": 7.529411764705882,
+      "grad_norm": 1.8175181083869496,
+      "learning_rate": 6.289818059204404e-06,
+      "loss": 0.0409,
       "step": 448
     },
     {
-      "epoch": 5.345238095238095,
-      "grad_norm": 0.7008333575781895,
-      "learning_rate": 1.2247595805818471e-05,
-      "loss": 0.0599,
+      "epoch": 7.546218487394958,
+      "grad_norm": 1.2608121221023865,
+      "learning_rate": 6.24732995120626e-06,
+      "loss": 0.0447,
       "step": 449
     },
     {
-      "epoch": 5.357142857142857,
-      "grad_norm": 6.856328864129851,
-      "learning_rate": 1.2216251459569927e-05,
-      "loss": 0.1179,
+      "epoch": 7.563025210084033,
+      "grad_norm": 1.558905828362155,
+      "learning_rate": 6.204920546180728e-06,
+      "loss": 0.0629,
       "step": 450
     },
     {
-      "epoch": 5.369047619047619,
-      "grad_norm": 6.86386327644984,
-      "learning_rate": 1.2184884197824513e-05,
-      "loss": 0.0773,
+      "epoch": 7.579831932773109,
+      "grad_norm": 3.0836236832919774,
+      "learning_rate": 6.162590733560183e-06,
+      "loss": 0.0714,
       "step": 451
     },
     {
-      "epoch": 5.380952380952381,
-      "grad_norm": 2.646115956267491,
-      "learning_rate": 1.2153494344912031e-05,
-      "loss": 0.1069,
+      "epoch": 7.5966386554621845,
+      "grad_norm": 1.9457063009832318,
+      "learning_rate": 6.120341401107751e-06,
+      "loss": 0.0591,
       "step": 452
     },
     {
-      "epoch": 5.392857142857143,
-      "grad_norm": 1.8101984992707947,
-      "learning_rate": 1.212208222539586e-05,
-      "loss": 0.0745,
+      "epoch": 7.61344537815126,
+      "grad_norm": 2.8239440684779,
+      "learning_rate": 6.078173434898679e-06,
+      "loss": 0.0598,
       "step": 453
     },
     {
-      "epoch": 5.404761904761905,
-      "grad_norm": 6.744068306166054,
-      "learning_rate": 1.2090648164069613e-05,
-      "loss": 0.1462,
+      "epoch": 7.630252100840336,
+      "grad_norm": 1.6266933180384318,
+      "learning_rate": 6.036087719301763e-06,
+      "loss": 0.0622,
       "step": 454
     },
     {
-      "epoch": 5.416666666666667,
-      "grad_norm": 2.587664561358256,
-      "learning_rate": 1.205919248595379e-05,
-      "loss": 0.175,
+      "epoch": 7.647058823529412,
+      "grad_norm": 1.5079333156466859,
+      "learning_rate": 5.994085136960801e-06,
+      "loss": 0.0662,
       "step": 455
     },
     {
-      "epoch": 5.428571428571429,
-      "grad_norm": 1.0425686405747994,
-      "learning_rate": 1.2027715516292386e-05,
-      "loss": 0.0695,
+      "epoch": 7.663865546218488,
+      "grad_norm": 2.513503195610099,
+      "learning_rate": 5.952166568776062e-06,
+      "loss": 0.0741,
       "step": 456
     },
     {
-      "epoch": 5.440476190476191,
-      "grad_norm": 1.6830078270896764,
-      "learning_rate": 1.1996217580549557e-05,
-      "loss": 0.1039,
+      "epoch": 7.680672268907563,
+      "grad_norm": 2.942516455656699,
+      "learning_rate": 5.910332893885842e-06,
+      "loss": 0.0768,
       "step": 457
     },
     {
-      "epoch": 5.4523809523809526,
-      "grad_norm": 6.172195451892937,
-      "learning_rate": 1.1964699004406231e-05,
-      "loss": 0.0442,
+      "epoch": 7.697478991596639,
+      "grad_norm": 0.9029614933690709,
+      "learning_rate": 5.868584989647994e-06,
+      "loss": 0.0465,
       "step": 458
     },
     {
-      "epoch": 5.464285714285714,
-      "grad_norm": 0.9630413019254777,
-      "learning_rate": 1.1933160113756764e-05,
-      "loss": 0.062,
+      "epoch": 7.714285714285714,
+      "grad_norm": 1.188640664205059,
+      "learning_rate": 5.826923731621562e-06,
+      "loss": 0.0507,
       "step": 459
     },
     {
-      "epoch": 5.476190476190476,
-      "grad_norm": 4.906480139073013,
-      "learning_rate": 1.1901601234705556e-05,
-      "loss": 0.1649,
+      "epoch": 7.73109243697479,
+      "grad_norm": 1.2597887336311209,
+      "learning_rate": 5.785349993548382e-06,
+      "loss": 0.0442,
       "step": 460
     },
     {
-      "epoch": 5.488095238095238,
-      "grad_norm": 6.576239158062562,
-      "learning_rate": 1.187002269356368e-05,
-      "loss": 0.0965,
+      "epoch": 7.7478991596638656,
+      "grad_norm": 5.82455886129036,
+      "learning_rate": 5.743864647334789e-06,
+      "loss": 0.154,
       "step": 461
     },
     {
-      "epoch": 5.5,
-      "grad_norm": 1.6597413036008377,
-      "learning_rate": 1.1838424816845507e-05,
-      "loss": 0.0302,
+      "epoch": 7.764705882352941,
+      "grad_norm": 1.2355207857604258,
+      "learning_rate": 5.702468563033307e-06,
+      "loss": 0.0472,
       "step": 462
     },
     {
-      "epoch": 5.511904761904762,
-      "grad_norm": 3.7920628634602385,
-      "learning_rate": 1.1806807931265344e-05,
-      "loss": 0.1098,
+      "epoch": 7.781512605042017,
+      "grad_norm": 1.673091621193694,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.0714,
       "step": 463
     },
     {
-      "epoch": 5.523809523809524,
-      "grad_norm": 1.4114425562472657,
-      "learning_rate": 1.177517236373403e-05,
-      "loss": 0.0626,
+      "epoch": 7.798319327731092,
+      "grad_norm": 1.2037358486316267,
+      "learning_rate": 5.6199476509983546e-06,
+      "loss": 0.0576,
       "step": 464
     },
     {
-      "epoch": 5.535714285714286,
-      "grad_norm": 2.9416388986651296,
-      "learning_rate": 1.174351844135559e-05,
-      "loss": 0.1376,
+      "epoch": 7.815126050420168,
+      "grad_norm": 1.4923009661165723,
+      "learning_rate": 5.5788245539369144e-06,
+      "loss": 0.0546,
       "step": 465
     },
     {
-      "epoch": 5.5476190476190474,
-      "grad_norm": 6.283324412491074,
-      "learning_rate": 1.1711846491423818e-05,
-      "loss": 0.1851,
+      "epoch": 7.831932773109243,
+      "grad_norm": 7.004074578858643,
+      "learning_rate": 5.537794180095341e-06,
+      "loss": 0.0931,
       "step": 466
     },
     {
-      "epoch": 5.559523809523809,
-      "grad_norm": 15.186006477965242,
-      "learning_rate": 1.1680156841418911e-05,
-      "loss": 0.1297,
+      "epoch": 7.848739495798319,
+      "grad_norm": 1.9350012591991381,
+      "learning_rate": 5.496857389984251e-06,
+      "loss": 0.0501,
       "step": 467
     },
     {
-      "epoch": 5.571428571428571,
-      "grad_norm": 2.7222956541416945,
-      "learning_rate": 1.1648449819004084e-05,
-      "loss": 0.0728,
+      "epoch": 7.865546218487395,
+      "grad_norm": 2.0402004080487575,
+      "learning_rate": 5.456015042151563e-06,
+      "loss": 0.0548,
       "step": 468
     },
     {
-      "epoch": 5.583333333333333,
-      "grad_norm": 2.6958569293331958,
-      "learning_rate": 1.1616725752022178e-05,
-      "loss": 0.1425,
+      "epoch": 7.882352941176471,
+      "grad_norm": 1.5295516734084738,
+      "learning_rate": 5.415267993164504e-06,
+      "loss": 0.0529,
       "step": 469
     },
     {
-      "epoch": 5.595238095238095,
-      "grad_norm": 2.9947378953143864,
-      "learning_rate": 1.1584984968492267e-05,
-      "loss": 0.0379,
+      "epoch": 7.899159663865547,
+      "grad_norm": 1.8934498235843238,
+      "learning_rate": 5.37461709759165e-06,
+      "loss": 0.0541,
       "step": 470
     },
     {
-      "epoch": 5.607142857142857,
-      "grad_norm": 3.5361493539488844,
-      "learning_rate": 1.155322779660628e-05,
-      "loss": 0.2285,
+      "epoch": 7.915966386554622,
+      "grad_norm": 2.3477637506052935,
+      "learning_rate": 5.334063207984983e-06,
+      "loss": 0.091,
       "step": 471
     },
     {
-      "epoch": 5.619047619047619,
-      "grad_norm": 1.4994769810175483,
-      "learning_rate": 1.152145456472558e-05,
-      "loss": 0.1222,
+      "epoch": 7.932773109243698,
+      "grad_norm": 2.068442722717763,
+      "learning_rate": 5.2936071748620386e-06,
+      "loss": 0.0659,
       "step": 472
     },
     {
-      "epoch": 5.630952380952381,
-      "grad_norm": 1.751393424685345,
-      "learning_rate": 1.1489665601377603e-05,
-      "loss": 0.0892,
+      "epoch": 7.949579831932773,
+      "grad_norm": 1.5583233740963265,
+      "learning_rate": 5.253249846688053e-06,
+      "loss": 0.0582,
       "step": 473
     },
     {
-      "epoch": 5.642857142857143,
-      "grad_norm": 0.9029221853359657,
-      "learning_rate": 1.1457861235252436e-05,
-      "loss": 0.1103,
+      "epoch": 7.966386554621849,
+      "grad_norm": 1.3210668874752445,
+      "learning_rate": 5.21299206985816e-06,
+      "loss": 0.0457,
       "step": 474
     },
     {
-      "epoch": 5.654761904761905,
-      "grad_norm": 1.9581770337649325,
-      "learning_rate": 1.1426041795199434e-05,
-      "loss": 0.1159,
+      "epoch": 7.983193277310924,
+      "grad_norm": 1.8720045795583338,
+      "learning_rate": 5.172834688679665e-06,
+      "loss": 0.0485,
       "step": 475
     },
     {
-      "epoch": 5.666666666666667,
-      "grad_norm": 6.706135908789148,
-      "learning_rate": 1.1394207610223802e-05,
-      "loss": 0.1742,
+      "epoch": 8.0,
+      "grad_norm": 1.717240919628548,
+      "learning_rate": 5.132778545354305e-06,
+      "loss": 0.0528,
       "step": 476
     },
     {
-      "epoch": 5.678571428571429,
-      "grad_norm": 2.630766596113569,
-      "learning_rate": 1.1362359009483213e-05,
-      "loss": 0.0714,
-      "step": 477
+      "epoch": 8.0,
+      "eval_loss": 0.09206734597682953,
+      "eval_runtime": 95.9403,
+      "eval_samples_per_second": 0.98,
+      "eval_steps_per_second": 0.98,
+      "step": 476
     },
     {
-      "epoch": 5.690476190476191,
-      "grad_norm": 2.091897523457701,
-      "learning_rate": 1.1330496322284392e-05,
-      "loss": 0.0672,
+      "epoch": 8.016806722689076,
+      "grad_norm": 1.5169536755225592,
+      "learning_rate": 5.092824479960625e-06,
+      "loss": 0.0474,
+      "step": 477
+    },
+    {
+      "epoch": 8.033613445378151,
+      "grad_norm": 1.7357799184112437,
+      "learning_rate": 5.0529733304363145e-06,
+      "loss": 0.0431,
       "step": 478
     },
     {
-      "epoch": 5.7023809523809526,
-      "grad_norm": 2.1214408759543155,
-      "learning_rate": 1.1298619878079713e-05,
-      "loss": 0.0767,
+      "epoch": 8.050420168067227,
+      "grad_norm": 1.3870139040501723,
+      "learning_rate": 5.013225932560679e-06,
+      "loss": 0.0423,
       "step": 479
     },
     {
-      "epoch": 5.714285714285714,
-      "grad_norm": 5.6326738420170495,
-      "learning_rate": 1.1266730006463797e-05,
-      "loss": 0.0673,
+      "epoch": 8.067226890756302,
+      "grad_norm": 3.302734792491234,
+      "learning_rate": 4.973583119937072e-06,
+      "loss": 0.0655,
       "step": 480
     },
     {
-      "epoch": 5.726190476190476,
-      "grad_norm": 2.5553857823512955,
-      "learning_rate": 1.1234827037170101e-05,
-      "loss": 0.0799,
+      "epoch": 8.084033613445378,
+      "grad_norm": 1.8227280505702954,
+      "learning_rate": 4.934045723975441e-06,
+      "loss": 0.0731,
       "step": 481
     },
     {
-      "epoch": 5.738095238095238,
-      "grad_norm": 1.3221943520824615,
-      "learning_rate": 1.1202911300067508e-05,
-      "loss": 0.0491,
+      "epoch": 8.100840336134453,
+      "grad_norm": 1.3256962048722176,
+      "learning_rate": 4.894614573874877e-06,
+      "loss": 0.0432,
       "step": 482
     },
     {
-      "epoch": 5.75,
-      "grad_norm": 1.3244889662711417,
-      "learning_rate": 1.1170983125156913e-05,
-      "loss": 0.1125,
+      "epoch": 8.117647058823529,
+      "grad_norm": 1.427387472053757,
+      "learning_rate": 4.85529049660623e-06,
+      "loss": 0.0549,
       "step": 483
     },
     {
-      "epoch": 5.761904761904762,
-      "grad_norm": 1.6046163371636937,
-      "learning_rate": 1.1139042842567824e-05,
-      "loss": 0.07,
+      "epoch": 8.134453781512605,
+      "grad_norm": 1.722178796219511,
+      "learning_rate": 4.81607431689475e-06,
+      "loss": 0.057,
       "step": 484
     },
     {
-      "epoch": 5.773809523809524,
-      "grad_norm": 1.70434051577756,
-      "learning_rate": 1.110709078255493e-05,
-      "loss": 0.0972,
+      "epoch": 8.15126050420168,
+      "grad_norm": 1.4946268224043258,
+      "learning_rate": 4.776966857202816e-06,
+      "loss": 0.0501,
       "step": 485
     },
     {
-      "epoch": 5.785714285714286,
-      "grad_norm": 1.731299719780292,
-      "learning_rate": 1.1075127275494704e-05,
-      "loss": 0.0555,
+      "epoch": 8.168067226890756,
+      "grad_norm": 1.4410674537931873,
+      "learning_rate": 4.737968937712674e-06,
+      "loss": 0.0472,
       "step": 486
     },
     {
-      "epoch": 5.7976190476190474,
-      "grad_norm": 6.126806212414693,
-      "learning_rate": 1.1043152651881972e-05,
-      "loss": 0.0715,
+      "epoch": 8.184873949579831,
+      "grad_norm": 4.079218048537049,
+      "learning_rate": 4.699081376309218e-06,
+      "loss": 0.0669,
       "step": 487
     },
     {
-      "epoch": 5.809523809523809,
-      "grad_norm": 5.169619165945196,
-      "learning_rate": 1.1011167242326504e-05,
-      "loss": 0.0784,
+      "epoch": 8.201680672268907,
+      "grad_norm": 2.445073900654683,
+      "learning_rate": 4.660304988562877e-06,
+      "loss": 0.0642,
       "step": 488
     },
     {
-      "epoch": 5.821428571428571,
-      "grad_norm": 2.7914135139256384,
-      "learning_rate": 1.0979171377549595e-05,
-      "loss": 0.1272,
+      "epoch": 8.218487394957982,
+      "grad_norm": 6.718089966341145,
+      "learning_rate": 4.621640587712468e-06,
+      "loss": 0.1271,
       "step": 489
     },
     {
-      "epoch": 5.833333333333333,
-      "grad_norm": 2.3959930088185635,
-      "learning_rate": 1.0947165388380646e-05,
-      "loss": 0.0279,
+      "epoch": 8.235294117647058,
+      "grad_norm": 2.8570358473448256,
+      "learning_rate": 4.583088984648172e-06,
+      "loss": 0.0629,
       "step": 490
     },
     {
-      "epoch": 5.845238095238095,
-      "grad_norm": 1.035178074469445,
-      "learning_rate": 1.091514960575374e-05,
-      "loss": 0.0466,
+      "epoch": 8.252100840336134,
+      "grad_norm": 1.9274312241325455,
+      "learning_rate": 4.544650987894514e-06,
+      "loss": 0.0367,
       "step": 491
     },
     {
-      "epoch": 5.857142857142857,
-      "grad_norm": 3.400565752147332,
-      "learning_rate": 1.088312436070422e-05,
-      "loss": 0.1819,
+      "epoch": 8.268907563025211,
+      "grad_norm": 1.7166820491309733,
+      "learning_rate": 4.5063274035934016e-06,
+      "loss": 0.0451,
       "step": 492
     },
     {
-      "epoch": 5.869047619047619,
-      "grad_norm": 0.9053402187508278,
-      "learning_rate": 1.0851089984365272e-05,
-      "loss": 0.0552,
+      "epoch": 8.285714285714286,
+      "grad_norm": 1.3196462720335975,
+      "learning_rate": 4.468119035487231e-06,
+      "loss": 0.0451,
       "step": 493
     },
     {
-      "epoch": 5.880952380952381,
-      "grad_norm": 3.403235950315029,
-      "learning_rate": 1.0819046807964495e-05,
-      "loss": 0.1056,
+      "epoch": 8.302521008403362,
+      "grad_norm": 1.5329980369426424,
+      "learning_rate": 4.430026684902017e-06,
+      "loss": 0.0382,
       "step": 494
     },
     {
-      "epoch": 5.892857142857143,
-      "grad_norm": 1.8344860921030868,
-      "learning_rate": 1.078699516282048e-05,
-      "loss": 0.0865,
+      "epoch": 8.319327731092438,
+      "grad_norm": 1.7844728493839146,
+      "learning_rate": 4.392051150730602e-06,
+      "loss": 0.0346,
       "step": 495
     },
     {
-      "epoch": 5.904761904761905,
-      "grad_norm": 2.4533834665154806,
-      "learning_rate": 1.075493538033938e-05,
-      "loss": 0.0199,
+      "epoch": 8.336134453781513,
+      "grad_norm": 1.233877897725148,
+      "learning_rate": 4.354193229415882e-06,
+      "loss": 0.0347,
       "step": 496
     },
     {
-      "epoch": 5.916666666666667,
-      "grad_norm": 0.8203250859689373,
-      "learning_rate": 1.0722867792011486e-05,
-      "loss": 0.0477,
+      "epoch": 8.352941176470589,
+      "grad_norm": 1.0911759571068154,
+      "learning_rate": 4.3164537149341246e-06,
+      "loss": 0.0534,
       "step": 497
     },
     {
-      "epoch": 5.928571428571429,
-      "grad_norm": 1.8912690142583974,
-      "learning_rate": 1.0690792729407807e-05,
-      "loss": 0.1064,
+      "epoch": 8.369747899159664,
+      "grad_norm": 2.1095342838945093,
+      "learning_rate": 4.278833398778306e-06,
+      "loss": 0.0531,
       "step": 498
     },
     {
-      "epoch": 5.940476190476191,
-      "grad_norm": 2.468608525491094,
-      "learning_rate": 1.0658710524176623e-05,
-      "loss": 0.1368,
+      "epoch": 8.38655462184874,
+      "grad_norm": 2.3616676446482647,
+      "learning_rate": 4.241333069941503e-06,
+      "loss": 0.0569,
       "step": 499
     },
     {
-      "epoch": 5.9523809523809526,
-      "grad_norm": 1.6228727525493614,
-      "learning_rate": 1.0626621508040074e-05,
-      "loss": 0.1318,
+      "epoch": 8.403361344537815,
+      "grad_norm": 1.3977212089055078,
+      "learning_rate": 4.203953514900366e-06,
+      "loss": 0.0353,
       "step": 500
     },
     {
-      "epoch": 5.964285714285714,
-      "grad_norm": 3.5639251749513496,
-      "learning_rate": 1.059452601279072e-05,
-      "loss": 0.0895,
+      "epoch": 8.420168067226891,
+      "grad_norm": 2.6496004578015575,
+      "learning_rate": 4.166695517598611e-06,
+      "loss": 0.1042,
       "step": 501
     },
     {
-      "epoch": 5.976190476190476,
-      "grad_norm": 1.5178172676340558,
-      "learning_rate": 1.056242437028812e-05,
-      "loss": 0.1341,
+      "epoch": 8.436974789915967,
+      "grad_norm": 2.06089201000074,
+      "learning_rate": 4.129559859430573e-06,
+      "loss": 0.0968,
       "step": 502
     },
     {
-      "epoch": 5.988095238095238,
-      "grad_norm": 1.1893273945201521,
-      "learning_rate": 1.0530316912455387e-05,
-      "loss": 0.0256,
+      "epoch": 8.453781512605042,
+      "grad_norm": 1.826445704568566,
+      "learning_rate": 4.092547319224837e-06,
+      "loss": 0.0516,
       "step": 503
     },
     {
-      "epoch": 6.0,
-      "grad_norm": 3.9672848129665574,
-      "learning_rate": 1.0498203971275762e-05,
-      "loss": 0.1522,
-      "step": 504
-    },
-    {
-      "epoch": 6.0,
-      "eval_loss": 0.1508193165063858,
-      "eval_runtime": 38.2578,
-      "eval_samples_per_second": 1.568,
-      "eval_steps_per_second": 1.568,
+      "epoch": 8.470588235294118,
+      "grad_norm": 3.8412502857797195,
+      "learning_rate": 4.055658673227891e-06,
+      "loss": 0.0884,
       "step": 504
     },
     {
-      "epoch": 6.011904761904762,
-      "grad_norm": 1.6644304828026015,
-      "learning_rate": 1.046608587878919e-05,
-      "loss": 0.1135,
+      "epoch": 8.487394957983193,
+      "grad_norm": 0.9584236905314262,
+      "learning_rate": 4.01889469508784e-06,
+      "loss": 0.0249,
       "step": 505
     },
     {
-      "epoch": 6.023809523809524,
-      "grad_norm": 1.6379161599185448,
-      "learning_rate": 1.0433962967088871e-05,
-      "loss": 0.0236,
+      "epoch": 8.504201680672269,
+      "grad_norm": 3.291735033394008,
+      "learning_rate": 3.982256155838199e-06,
+      "loss": 0.0854,
       "step": 506
     },
     {
-      "epoch": 6.035714285714286,
-      "grad_norm": 2.632685664860106,
-      "learning_rate": 1.0401835568317842e-05,
-      "loss": 0.1443,
+      "epoch": 8.521008403361344,
+      "grad_norm": 4.532957288732598,
+      "learning_rate": 3.945743823881713e-06,
+      "loss": 0.0629,
       "step": 507
     },
     {
-      "epoch": 6.0476190476190474,
-      "grad_norm": 0.8860899184377556,
-      "learning_rate": 1.036970401466553e-05,
-      "loss": 0.0727,
+      "epoch": 8.53781512605042,
+      "grad_norm": 4.273353415832438,
+      "learning_rate": 3.909358464974228e-06,
+      "loss": 0.0624,
       "step": 508
     },
     {
-      "epoch": 6.059523809523809,
-      "grad_norm": 2.23073958875819,
-      "learning_rate": 1.0337568638364322e-05,
-      "loss": 0.0933,
+      "epoch": 8.554621848739496,
+      "grad_norm": 1.437659985078597,
+      "learning_rate": 3.873100842208661e-06,
+      "loss": 0.044,
       "step": 509
     },
     {
-      "epoch": 6.071428571428571,
-      "grad_norm": 5.244025160546965,
-      "learning_rate": 1.0305429771686135e-05,
-      "loss": 0.1129,
+      "epoch": 8.571428571428571,
+      "grad_norm": 2.6929955709108584,
+      "learning_rate": 3.836971715998968e-06,
+      "loss": 0.0598,
       "step": 510
     },
     {
-      "epoch": 6.083333333333333,
-      "grad_norm": 2.2806324713435027,
-      "learning_rate": 1.0273287746938974e-05,
-      "loss": 0.1352,
+      "epoch": 8.588235294117647,
+      "grad_norm": 1.9024448417705355,
+      "learning_rate": 3.8009718440642128e-06,
+      "loss": 0.0603,
       "step": 511
     },
     {
-      "epoch": 6.095238095238095,
-      "grad_norm": 3.734983785509321,
-      "learning_rate": 1.0241142896463492e-05,
-      "loss": 0.094,
+      "epoch": 8.605042016806722,
+      "grad_norm": 0.7656090871294385,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.0271,
       "step": 512
     },
     {
-      "epoch": 6.107142857142857,
-      "grad_norm": 1.488803778322093,
-      "learning_rate": 1.0208995552629568e-05,
-      "loss": 0.0475,
+      "epoch": 8.621848739495798,
+      "grad_norm": 1.498792756768189,
+      "learning_rate": 3.729362880325983e-06,
+      "loss": 0.0406,
       "step": 513
     },
     {
-      "epoch": 6.119047619047619,
-      "grad_norm": 2.402837670487696,
-      "learning_rate": 1.0176846047832855e-05,
-      "loss": 0.1298,
+      "epoch": 8.638655462184873,
+      "grad_norm": 1.914650360335185,
+      "learning_rate": 3.693755290343409e-06,
+      "loss": 0.0538,
       "step": 514
     },
     {
-      "epoch": 6.130952380952381,
-      "grad_norm": 2.0357587086107087,
-      "learning_rate": 1.0144694714491356e-05,
-      "loss": 0.0564,
+      "epoch": 8.655462184873949,
+      "grad_norm": 1.097214252517178,
+      "learning_rate": 3.658279958246075e-06,
+      "loss": 0.0381,
       "step": 515
     },
     {
-      "epoch": 6.142857142857143,
-      "grad_norm": 3.109809726164325,
-      "learning_rate": 1.0112541885041973e-05,
-      "loss": 0.068,
+      "epoch": 8.672268907563025,
+      "grad_norm": 2.861008119753485,
+      "learning_rate": 3.622937628041334e-06,
+      "loss": 0.0808,
       "step": 516
     },
     {
-      "epoch": 6.154761904761905,
-      "grad_norm": 2.0837285046565293,
-      "learning_rate": 1.0080387891937085e-05,
-      "loss": 0.1077,
+      "epoch": 8.6890756302521,
+      "grad_norm": 1.9341336610200206,
+      "learning_rate": 3.587729040947141e-06,
+      "loss": 0.0343,
       "step": 517
     },
     {
-      "epoch": 6.166666666666667,
-      "grad_norm": 1.6180323000868266,
-      "learning_rate": 1.0048233067641098e-05,
-      "loss": 0.0779,
+      "epoch": 8.705882352941176,
+      "grad_norm": 2.42481473153521,
+      "learning_rate": 3.5526549353765294e-06,
+      "loss": 0.1077,
       "step": 518
     },
     {
-      "epoch": 6.178571428571429,
-      "grad_norm": 1.9200465377373805,
-      "learning_rate": 1.0016077744627012e-05,
-      "loss": 0.0946,
+      "epoch": 8.722689075630251,
+      "grad_norm": 2.4065694655289973,
+      "learning_rate": 3.5177160469221184e-06,
+      "loss": 0.0641,
       "step": 519
     },
     {
-      "epoch": 6.190476190476191,
-      "grad_norm": 2.4780925220259737,
-      "learning_rate": 9.983922255372991e-06,
-      "loss": 0.0664,
+      "epoch": 8.739495798319329,
+      "grad_norm": 1.8534338905199441,
+      "learning_rate": 3.4829131083406684e-06,
+      "loss": 0.0529,
       "step": 520
     },
     {
-      "epoch": 6.2023809523809526,
-      "grad_norm": 1.269298365112061,
-      "learning_rate": 9.951766932358907e-06,
-      "loss": 0.1073,
+      "epoch": 8.756302521008404,
+      "grad_norm": 2.072202396225646,
+      "learning_rate": 3.448246849537741e-06,
+      "loss": 0.0565,
       "step": 521
     },
     {
-      "epoch": 6.214285714285714,
-      "grad_norm": 2.1626917897691436,
-      "learning_rate": 9.919612108062919e-06,
-      "loss": 0.0672,
+      "epoch": 8.77310924369748,
+      "grad_norm": 1.5479608735352894,
+      "learning_rate": 3.413717997552376e-06,
+      "loss": 0.0411,
       "step": 522
     },
     {
-      "epoch": 6.226190476190476,
-      "grad_norm": 3.8302257883114357,
-      "learning_rate": 9.887458114958032e-06,
-      "loss": 0.158,
+      "epoch": 8.789915966386555,
+      "grad_norm": 2.6939787148898016,
+      "learning_rate": 3.379327276541834e-06,
+      "loss": 0.0722,
       "step": 523
     },
     {
-      "epoch": 6.238095238095238,
-      "grad_norm": 1.469046690940951,
-      "learning_rate": 9.855305285508649e-06,
-      "loss": 0.0456,
+      "epoch": 8.806722689075631,
+      "grad_norm": 1.5450816630775992,
+      "learning_rate": 3.3450754077664337e-06,
+      "loss": 0.0544,
       "step": 524
     },
     {
-      "epoch": 6.25,
-      "grad_norm": 2.750101219167286,
-      "learning_rate": 9.823153952167148e-06,
-      "loss": 0.0828,
+      "epoch": 8.823529411764707,
+      "grad_norm": 1.5918689109996174,
+      "learning_rate": 3.310963109574408e-06,
+      "loss": 0.0407,
       "step": 525
     },
     {
-      "epoch": 6.261904761904762,
-      "grad_norm": 1.8402371967805808,
-      "learning_rate": 9.791004447370439e-06,
-      "loss": 0.0953,
+      "epoch": 8.840336134453782,
+      "grad_norm": 1.5593134309915293,
+      "learning_rate": 3.2769910973868314e-06,
+      "loss": 0.0418,
       "step": 526
     },
     {
-      "epoch": 6.273809523809524,
-      "grad_norm": 3.048460879374815,
-      "learning_rate": 9.758857103536513e-06,
-      "loss": 0.0729,
+      "epoch": 8.857142857142858,
+      "grad_norm": 1.9817294913123196,
+      "learning_rate": 3.243160083682645e-06,
+      "loss": 0.0529,
       "step": 527
     },
     {
-      "epoch": 6.285714285714286,
-      "grad_norm": 1.6638375884973138,
-      "learning_rate": 9.726712253061031e-06,
-      "loss": 0.0628,
+      "epoch": 8.873949579831933,
+      "grad_norm": 1.695431393346344,
+      "learning_rate": 3.209470777983675e-06,
+      "loss": 0.0488,
       "step": 528
     },
     {
-      "epoch": 6.2976190476190474,
-      "grad_norm": 1.6689888226784142,
-      "learning_rate": 9.694570228313868e-06,
-      "loss": 0.0589,
+      "epoch": 8.890756302521009,
+      "grad_norm": 1.8840574831179149,
+      "learning_rate": 3.1759238868397925e-06,
+      "loss": 0.0529,
       "step": 529
     },
     {
-      "epoch": 6.309523809523809,
-      "grad_norm": 1.6801934021210152,
-      "learning_rate": 9.662431361635681e-06,
-      "loss": 0.061,
+      "epoch": 8.907563025210084,
+      "grad_norm": 5.952373137687765,
+      "learning_rate": 3.1425201138140592e-06,
+      "loss": 0.0808,
       "step": 530
     },
     {
-      "epoch": 6.321428571428571,
-      "grad_norm": 1.2262908116444213,
-      "learning_rate": 9.630295985334474e-06,
-      "loss": 0.0916,
+      "epoch": 8.92436974789916,
+      "grad_norm": 1.290713251381066,
+      "learning_rate": 3.1092601594679993e-06,
+      "loss": 0.0439,
       "step": 531
     },
     {
-      "epoch": 6.333333333333333,
-      "grad_norm": 1.5047283215622451,
-      "learning_rate": 9.598164431682161e-06,
-      "loss": 0.0847,
+      "epoch": 8.941176470588236,
+      "grad_norm": 0.9948068099144867,
+      "learning_rate": 3.0761447213468888e-06,
+      "loss": 0.0405,
       "step": 532
     },
     {
-      "epoch": 6.345238095238095,
-      "grad_norm": 1.8569321486125632,
-      "learning_rate": 9.56603703291113e-06,
-      "loss": 0.0703,
+      "epoch": 8.957983193277311,
+      "grad_norm": 1.9536076330602832,
+      "learning_rate": 3.0431744939651365e-06,
+      "loss": 0.0538,
       "step": 533
     },
     {
-      "epoch": 6.357142857142857,
-      "grad_norm": 2.3223165705202584,
-      "learning_rate": 9.533914121210813e-06,
-      "loss": 0.0141,
+      "epoch": 8.974789915966387,
+      "grad_norm": 1.6523207102197872,
+      "learning_rate": 3.0103501687917192e-06,
+      "loss": 0.0544,
       "step": 534
     },
     {
-      "epoch": 6.369047619047619,
-      "grad_norm": 1.947523583239929,
-      "learning_rate": 9.501796028724241e-06,
-      "loss": 0.1206,
+      "epoch": 8.991596638655462,
+      "grad_norm": 0.9746183785405975,
+      "learning_rate": 2.9776724342356654e-06,
+      "loss": 0.0531,
+      "step": 535
+    },
+    {
+      "epoch": 8.991596638655462,
+      "eval_loss": 0.09362534433603287,
+      "eval_runtime": 84.973,
+      "eval_samples_per_second": 1.106,
+      "eval_steps_per_second": 1.106,
       "step": 535
     },
     {
-      "epoch": 6.380952380952381,
-      "grad_norm": 1.3333488170380448,
-      "learning_rate": 9.469683087544616e-06,
-      "loss": 0.071,
+      "epoch": 9.008403361344538,
+      "grad_norm": 3.1103066899739282,
+      "learning_rate": 2.945141975631637e-06,
+      "loss": 0.0553,
       "step": 536
     },
     {
-      "epoch": 6.392857142857143,
-      "grad_norm": 1.0684885808241553,
-      "learning_rate": 9.437575629711883e-06,
-      "loss": 0.0483,
+      "epoch": 9.025210084033613,
+      "grad_norm": 2.3282836310480435,
+      "learning_rate": 2.912759475225546e-06,
+      "loss": 0.0532,
       "step": 537
     },
     {
-      "epoch": 6.404761904761905,
-      "grad_norm": 1.5479791429558227,
-      "learning_rate": 9.405473987209283e-06,
-      "loss": 0.1061,
+      "epoch": 9.042016806722689,
+      "grad_norm": 1.1197327667128945,
+      "learning_rate": 2.88052561216024e-06,
+      "loss": 0.0308,
       "step": 538
     },
     {
-      "epoch": 6.416666666666667,
-      "grad_norm": 2.06718170775221,
-      "learning_rate": 9.37337849195993e-06,
-      "loss": 0.101,
+      "epoch": 9.058823529411764,
+      "grad_norm": 1.4098107198655643,
+      "learning_rate": 2.8484410624612744e-06,
+      "loss": 0.0364,
       "step": 539
     },
     {
-      "epoch": 6.428571428571429,
-      "grad_norm": 2.5758094309956165,
-      "learning_rate": 9.34128947582338e-06,
-      "loss": 0.0697,
+      "epoch": 9.07563025210084,
+      "grad_norm": 4.438658787202018,
+      "learning_rate": 2.8165064990227255e-06,
+      "loss": 0.0561,
       "step": 540
     },
     {
-      "epoch": 6.440476190476191,
-      "grad_norm": 1.585812534614295,
-      "learning_rate": 9.309207270592196e-06,
-      "loss": 0.0768,
+      "epoch": 9.092436974789916,
+      "grad_norm": 1.9493994527119616,
+      "learning_rate": 2.7847225915930697e-06,
+      "loss": 0.0601,
       "step": 541
     },
     {
-      "epoch": 6.4523809523809526,
-      "grad_norm": 1.9676744482847963,
-      "learning_rate": 9.277132207988516e-06,
-      "loss": 0.0691,
+      "epoch": 9.109243697478991,
+      "grad_norm": 1.8527667273461481,
+      "learning_rate": 2.7530900067611577e-06,
+      "loss": 0.0531,
       "step": 542
     },
     {
-      "epoch": 6.464285714285714,
-      "grad_norm": 1.5053512527538524,
-      "learning_rate": 9.245064619660624e-06,
-      "loss": 0.0825,
+      "epoch": 9.126050420168067,
+      "grad_norm": 1.701462207843824,
+      "learning_rate": 2.7216094079422185e-06,
+      "loss": 0.0472,
       "step": 543
     },
     {
-      "epoch": 6.476190476190476,
-      "grad_norm": 2.034981803501994,
-      "learning_rate": 9.213004837179524e-06,
-      "loss": 0.0244,
+      "epoch": 9.142857142857142,
+      "grad_norm": 1.8914078563116579,
+      "learning_rate": 2.6902814553639443e-06,
+      "loss": 0.0551,
       "step": 544
     },
     {
-      "epoch": 6.488095238095238,
-      "grad_norm": 0.9575306509567822,
-      "learning_rate": 9.180953192035507e-06,
-      "loss": 0.0468,
+      "epoch": 9.159663865546218,
+      "grad_norm": 2.591885452858992,
+      "learning_rate": 2.6591068060526626e-06,
+      "loss": 0.0395,
       "step": 545
     },
     {
-      "epoch": 6.5,
-      "grad_norm": 1.295046452458436,
-      "learning_rate": 9.14891001563473e-06,
-      "loss": 0.0384,
+      "epoch": 9.176470588235293,
+      "grad_norm": 3.074064642620058,
+      "learning_rate": 2.62808611381953e-06,
+      "loss": 0.0619,
       "step": 546
     },
     {
-      "epoch": 6.511904761904762,
-      "grad_norm": 2.0081622877090144,
-      "learning_rate": 9.116875639295783e-06,
-      "loss": 0.0507,
+      "epoch": 9.193277310924369,
+      "grad_norm": 1.6148671787143798,
+      "learning_rate": 2.597220029246846e-06,
+      "loss": 0.0369,
       "step": 547
     },
     {
-      "epoch": 6.523809523809524,
-      "grad_norm": 1.6714670350427359,
-      "learning_rate": 9.084850394246262e-06,
-      "loss": 0.0184,
+      "epoch": 9.210084033613445,
+      "grad_norm": 5.145830985825016,
+      "learning_rate": 2.5665091996743898e-06,
+      "loss": 0.0912,
       "step": 548
     },
     {
-      "epoch": 6.535714285714286,
-      "grad_norm": 1.3524423408754032,
-      "learning_rate": 9.052834611619354e-06,
-      "loss": 0.0917,
+      "epoch": 9.22689075630252,
+      "grad_norm": 1.1197523582174669,
+      "learning_rate": 2.5359542691858542e-06,
+      "loss": 0.0411,
       "step": 549
     },
     {
-      "epoch": 6.5476190476190474,
-      "grad_norm": 1.208211246351009,
-      "learning_rate": 9.020828622450406e-06,
-      "loss": 0.0188,
+      "epoch": 9.243697478991596,
+      "grad_norm": 1.5754449262763028,
+      "learning_rate": 2.5055558785953304e-06,
+      "loss": 0.0502,
       "step": 550
     },
     {
-      "epoch": 6.559523809523809,
-      "grad_norm": 0.810515144898437,
-      "learning_rate": 8.988832757673497e-06,
-      "loss": 0.0625,
+      "epoch": 9.260504201680673,
+      "grad_norm": 2.843721070657444,
+      "learning_rate": 2.4753146654338765e-06,
+      "loss": 0.0835,
       "step": 551
     },
     {
-      "epoch": 6.571428571428571,
-      "grad_norm": 1.4738861513691155,
-      "learning_rate": 8.95684734811803e-06,
-      "loss": 0.0905,
+      "epoch": 9.277310924369749,
+      "grad_norm": 2.2196607101882817,
+      "learning_rate": 2.4452312639361462e-06,
+      "loss": 0.0465,
       "step": 552
     },
     {
-      "epoch": 6.583333333333333,
-      "grad_norm": 1.8126992202021779,
-      "learning_rate": 8.924872724505296e-06,
-      "loss": 0.0939,
+      "epoch": 9.294117647058824,
+      "grad_norm": 2.166755095388008,
+      "learning_rate": 2.415306305027072e-06,
+      "loss": 0.0421,
       "step": 553
     },
     {
-      "epoch": 6.595238095238095,
-      "grad_norm": 1.5254885661369932,
-      "learning_rate": 8.892909217445069e-06,
-      "loss": 0.0644,
+      "epoch": 9.3109243697479,
+      "grad_norm": 3.8956621071776945,
+      "learning_rate": 2.3855404163086558e-06,
+      "loss": 0.0456,
       "step": 554
     },
     {
-      "epoch": 6.607142857142857,
-      "grad_norm": 0.9610399303180617,
-      "learning_rate": 8.860957157432177e-06,
-      "loss": 0.0541,
+      "epoch": 9.327731092436975,
+      "grad_norm": 1.198392921249873,
+      "learning_rate": 2.355934222046794e-06,
+      "loss": 0.0368,
       "step": 555
     },
     {
-      "epoch": 6.619047619047619,
-      "grad_norm": 1.0851393239445002,
-      "learning_rate": 8.829016874843089e-06,
-      "loss": 0.0593,
+      "epoch": 9.344537815126051,
+      "grad_norm": 2.081788407790618,
+      "learning_rate": 2.32648834315818e-06,
+      "loss": 0.0742,
       "step": 556
     },
     {
-      "epoch": 6.630952380952381,
-      "grad_norm": 1.7224098343060739,
-      "learning_rate": 8.797088699932494e-06,
-      "loss": 0.0164,
+      "epoch": 9.361344537815127,
+      "grad_norm": 1.8547943707777044,
+      "learning_rate": 2.2972033971972953e-06,
+      "loss": 0.0404,
       "step": 557
     },
     {
-      "epoch": 6.642857142857143,
-      "grad_norm": 0.9697818551837348,
-      "learning_rate": 8.7651729628299e-06,
-      "loss": 0.0422,
+      "epoch": 9.378151260504202,
+      "grad_norm": 2.06348184199221,
+      "learning_rate": 2.2680799983434532e-06,
+      "loss": 0.0362,
       "step": 558
     },
     {
-      "epoch": 6.654761904761905,
-      "grad_norm": 3.3017631401846903,
-      "learning_rate": 8.733269993536208e-06,
-      "loss": 0.1914,
+      "epoch": 9.394957983193278,
+      "grad_norm": 1.6015036795043573,
+      "learning_rate": 2.239118757387907e-06,
+      "loss": 0.0461,
       "step": 559
     },
     {
-      "epoch": 6.666666666666667,
-      "grad_norm": 2.032789558563978,
-      "learning_rate": 8.701380121920292e-06,
-      "loss": 0.043,
+      "epoch": 9.411764705882353,
+      "grad_norm": 2.500319456131113,
+      "learning_rate": 2.2103202817210555e-06,
+      "loss": 0.0436,
       "step": 560
     },
     {
-      "epoch": 6.678571428571429,
-      "grad_norm": 0.8515648745590835,
-      "learning_rate": 8.669503677715614e-06,
-      "loss": 0.0083,
+      "epoch": 9.428571428571429,
+      "grad_norm": 2.5704589340698036,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 0.0797,
       "step": 561
     },
     {
-      "epoch": 6.690476190476191,
-      "grad_norm": 0.671994580593515,
-      "learning_rate": 8.637640990516793e-06,
-      "loss": 0.0473,
+      "epoch": 9.445378151260504,
+      "grad_norm": 1.1623903030627807,
+      "learning_rate": 2.1532140387343736e-06,
+      "loss": 0.0434,
       "step": 562
     },
     {
-      "epoch": 6.7023809523809526,
-      "grad_norm": 2.666195404252126,
-      "learning_rate": 8.605792389776203e-06,
-      "loss": 0.1427,
+      "epoch": 9.46218487394958,
+      "grad_norm": 1.7348909728003972,
+      "learning_rate": 2.1249074690767434e-06,
+      "loss": 0.0513,
       "step": 563
     },
     {
-      "epoch": 6.714285714285714,
-      "grad_norm": 1.763396796049088,
-      "learning_rate": 8.573958204800572e-06,
-      "loss": 0.0838,
+      "epoch": 9.478991596638656,
+      "grad_norm": 1.443247675887106,
+      "learning_rate": 2.096766060007096e-06,
+      "loss": 0.0436,
       "step": 564
     },
     {
-      "epoch": 6.726190476190476,
-      "grad_norm": 1.246603953754217,
-      "learning_rate": 8.542138764747565e-06,
-      "loss": 0.0616,
+      "epoch": 9.495798319327731,
+      "grad_norm": 1.719154980437653,
+      "learning_rate": 2.068790401721886e-06,
+      "loss": 0.0473,
       "step": 565
     },
     {
-      "epoch": 6.738095238095238,
-      "grad_norm": 1.203406526414491,
-      "learning_rate": 8.5103343986224e-06,
-      "loss": 0.0148,
+      "epoch": 9.512605042016807,
+      "grad_norm": 1.3479219484263167,
+      "learning_rate": 2.040981080941349e-06,
+      "loss": 0.0403,
       "step": 566
     },
     {
-      "epoch": 6.75,
-      "grad_norm": 1.950964925182016,
-      "learning_rate": 8.478545435274424e-06,
-      "loss": 0.0868,
+      "epoch": 9.529411764705882,
+      "grad_norm": 1.7448556207352623,
+      "learning_rate": 2.013338680897209e-06,
+      "loss": 0.0374,
       "step": 567
     },
     {
-      "epoch": 6.761904761904762,
-      "grad_norm": 0.8066526898246336,
-      "learning_rate": 8.446772203393726e-06,
-      "loss": 0.0531,
+      "epoch": 9.546218487394958,
+      "grad_norm": 1.778610985288892,
+      "learning_rate": 1.9858637813204352e-06,
+      "loss": 0.0481,
       "step": 568
     },
     {
-      "epoch": 6.773809523809524,
-      "grad_norm": 1.7913521121040872,
-      "learning_rate": 8.415015031507734e-06,
-      "loss": 0.0714,
+      "epoch": 9.563025210084033,
+      "grad_norm": 1.023079090275057,
+      "learning_rate": 1.958556958429092e-06,
+      "loss": 0.043,
       "step": 569
     },
     {
-      "epoch": 6.785714285714286,
-      "grad_norm": 0.9075804833918106,
-      "learning_rate": 8.383274247977827e-06,
-      "loss": 0.0437,
+      "epoch": 9.579831932773109,
+      "grad_norm": 2.2504655920316927,
+      "learning_rate": 1.9314187849162523e-06,
+      "loss": 0.104,
       "step": 570
     },
     {
-      "epoch": 6.7976190476190474,
-      "grad_norm": 4.654542604347542,
-      "learning_rate": 8.35155018099592e-06,
-      "loss": 0.1748,
+      "epoch": 9.596638655462185,
+      "grad_norm": 4.623708198206026,
+      "learning_rate": 1.904449829937981e-06,
+      "loss": 0.0878,
       "step": 571
     },
     {
-      "epoch": 6.809523809523809,
-      "grad_norm": 3.177641851084243,
-      "learning_rate": 8.319843158581092e-06,
-      "loss": 0.1297,
+      "epoch": 9.61344537815126,
+      "grad_norm": 1.1241851785983472,
+      "learning_rate": 1.8776506591014054e-06,
+      "loss": 0.0358,
       "step": 572
     },
     {
-      "epoch": 6.821428571428571,
-      "grad_norm": 1.9316344815684265,
-      "learning_rate": 8.288153508576186e-06,
-      "loss": 0.0958,
+      "epoch": 9.630252100840336,
+      "grad_norm": 1.9153117494048388,
+      "learning_rate": 1.851021834452853e-06,
+      "loss": 0.0523,
       "step": 573
     },
     {
-      "epoch": 6.833333333333333,
-      "grad_norm": 4.63072501048228,
-      "learning_rate": 8.256481558644412e-06,
-      "loss": 0.0768,
+      "epoch": 9.647058823529411,
+      "grad_norm": 4.390748395794137,
+      "learning_rate": 1.8245639144660532e-06,
+      "loss": 0.0744,
       "step": 574
     },
     {
-      "epoch": 6.845238095238095,
-      "grad_norm": 8.04435552149917,
-      "learning_rate": 8.224827636265971e-06,
-      "loss": 0.0355,
+      "epoch": 9.663865546218487,
+      "grad_norm": 2.7441495241393783,
+      "learning_rate": 1.7982774540304404e-06,
+      "loss": 0.0492,
       "step": 575
     },
     {
-      "epoch": 6.857142857142857,
-      "grad_norm": 2.840529961154322,
-      "learning_rate": 8.19319206873466e-06,
-      "loss": 0.0673,
+      "epoch": 9.680672268907562,
+      "grad_norm": 1.3900528175419058,
+      "learning_rate": 1.772163004439511e-06,
+      "loss": 0.0479,
       "step": 576
     },
     {
-      "epoch": 6.869047619047619,
-      "grad_norm": 1.4798999016311165,
-      "learning_rate": 8.161575183154495e-06,
-      "loss": 0.0858,
+      "epoch": 9.697478991596638,
+      "grad_norm": 2.77436372781934,
+      "learning_rate": 1.7462211133792484e-06,
+      "loss": 0.0621,
       "step": 577
     },
     {
-      "epoch": 6.880952380952381,
-      "grad_norm": 2.273599989764983,
-      "learning_rate": 8.129977306436326e-06,
-      "loss": 0.0953,
+      "epoch": 9.714285714285714,
+      "grad_norm": 1.560534629765185,
+      "learning_rate": 1.720452324916656e-06,
+      "loss": 0.0357,
       "step": 578
     },
     {
-      "epoch": 6.892857142857143,
-      "grad_norm": 4.560000854260947,
-      "learning_rate": 8.098398765294447e-06,
-      "loss": 0.1315,
+      "epoch": 9.731092436974789,
+      "grad_norm": 2.4373786388347294,
+      "learning_rate": 1.6948571794883406e-06,
+      "loss": 0.0497,
       "step": 579
     },
     {
-      "epoch": 6.904761904761905,
-      "grad_norm": 2.7031944822443634,
-      "learning_rate": 8.066839886243238e-06,
-      "loss": 0.0603,
+      "epoch": 9.747899159663866,
+      "grad_norm": 1.3388696128320787,
+      "learning_rate": 1.6694362138891674e-06,
+      "loss": 0.0409,
       "step": 580
     },
     {
-      "epoch": 6.916666666666667,
-      "grad_norm": 2.1777978934683584,
-      "learning_rate": 8.035300995593772e-06,
-      "loss": 0.1096,
+      "epoch": 9.764705882352942,
+      "grad_norm": 2.602632435071536,
+      "learning_rate": 1.6441899612610178e-06,
+      "loss": 0.0557,
       "step": 581
     },
     {
-      "epoch": 6.928571428571429,
-      "grad_norm": 1.7370224989930014,
-      "learning_rate": 8.003782419450448e-06,
-      "loss": 0.0487,
+      "epoch": 9.781512605042018,
+      "grad_norm": 2.241392316929053,
+      "learning_rate": 1.6191189510815942e-06,
+      "loss": 0.0852,
       "step": 582
     },
     {
-      "epoch": 6.940476190476191,
-      "grad_norm": 2.0369881375248435,
-      "learning_rate": 7.972284483707615e-06,
-      "loss": 0.1216,
+      "epoch": 9.798319327731093,
+      "grad_norm": 1.2305602460104232,
+      "learning_rate": 1.5942237091533297e-06,
+      "loss": 0.0424,
       "step": 583
     },
     {
-      "epoch": 6.9523809523809526,
-      "grad_norm": 1.5695162807965415,
-      "learning_rate": 7.940807514046211e-06,
-      "loss": 0.1012,
+      "epoch": 9.815126050420169,
+      "grad_norm": 1.2682290196019605,
+      "learning_rate": 1.5695047575923462e-06,
+      "loss": 0.0497,
       "step": 584
     },
     {
-      "epoch": 6.964285714285714,
-      "grad_norm": 1.9962936354689058,
-      "learning_rate": 7.909351835930389e-06,
-      "loss": 0.0667,
+      "epoch": 9.831932773109244,
+      "grad_norm": 3.1108653972627924,
+      "learning_rate": 1.5449626148175144e-06,
+      "loss": 0.0943,
       "step": 585
     },
     {
-      "epoch": 6.976190476190476,
-      "grad_norm": 2.495887351780799,
-      "learning_rate": 7.877917774604144e-06,
-      "loss": 0.1092,
+      "epoch": 9.84873949579832,
+      "grad_norm": 1.29875959685456,
+      "learning_rate": 1.5205977955395812e-06,
+      "loss": 0.0239,
       "step": 586
     },
     {
-      "epoch": 6.988095238095238,
-      "grad_norm": 1.3893141441816679,
-      "learning_rate": 7.846505655087972e-06,
-      "loss": 0.0481,
+      "epoch": 9.865546218487395,
+      "grad_norm": 1.255438999538065,
+      "learning_rate": 1.4964108107503638e-06,
+      "loss": 0.0432,
       "step": 587
     },
     {
-      "epoch": 7.0,
-      "grad_norm": 2.4723106201742917,
-      "learning_rate": 7.815115802175485e-06,
-      "loss": 0.0279,
-      "step": 588
-    },
-    {
-      "epoch": 7.0,
-      "eval_loss": 0.16396619379520416,
-      "eval_runtime": 38.7412,
-      "eval_samples_per_second": 1.549,
-      "eval_steps_per_second": 1.549,
+      "epoch": 9.882352941176471,
+      "grad_norm": 1.6849212493860068,
+      "learning_rate": 1.4724021677120491e-06,
+      "loss": 0.0498,
       "step": 588
     },
     {
-      "epoch": 7.011904761904762,
-      "grad_norm": 2.2356269896066165,
-      "learning_rate": 7.783748540430074e-06,
-      "loss": 0.0637,
+      "epoch": 9.899159663865547,
+      "grad_norm": 1.543073795782045,
+      "learning_rate": 1.4485723699465392e-06,
+      "loss": 0.0436,
       "step": 589
     },
     {
-      "epoch": 7.023809523809524,
-      "grad_norm": 12.593626155384655,
-      "learning_rate": 7.75240419418153e-06,
-      "loss": 0.1319,
+      "epoch": 9.915966386554622,
+      "grad_norm": 1.6326111794747336,
+      "learning_rate": 1.4249219172249051e-06,
+      "loss": 0.0448,
       "step": 590
     },
     {
-      "epoch": 7.035714285714286,
-      "grad_norm": 7.517415321784336,
-      "learning_rate": 7.721083087522717e-06,
-      "loss": 0.0776,
+      "epoch": 9.932773109243698,
+      "grad_norm": 1.6795029810703082,
+      "learning_rate": 1.4014513055568978e-06,
+      "loss": 0.0483,
       "step": 591
     },
     {
-      "epoch": 7.0476190476190474,
-      "grad_norm": 4.188123414846697,
-      "learning_rate": 7.689785544306194e-06,
-      "loss": 0.0452,
+      "epoch": 9.949579831932773,
+      "grad_norm": 5.509418938412195,
+      "learning_rate": 1.3781610271805436e-06,
+      "loss": 0.1002,
       "step": 592
     },
     {
-      "epoch": 7.059523809523809,
-      "grad_norm": 1.7616711514678394,
-      "learning_rate": 7.65851188814089e-06,
-      "loss": 0.0852,
+      "epoch": 9.966386554621849,
+      "grad_norm": 1.9231485974986042,
+      "learning_rate": 1.3550515705518263e-06,
+      "loss": 0.0406,
       "step": 593
     },
     {
-      "epoch": 7.071428571428571,
-      "grad_norm": 1.8987651197623507,
-      "learning_rate": 7.627262442388747e-06,
-      "loss": 0.0851,
+      "epoch": 9.983193277310924,
+      "grad_norm": 2.6221792896880918,
+      "learning_rate": 1.3321234203344435e-06,
+      "loss": 0.0369,
       "step": 594
     },
     {
-      "epoch": 7.083333333333333,
-      "grad_norm": 2.2298478849854337,
-      "learning_rate": 7.596037530161371e-06,
-      "loss": 0.065,
+      "epoch": 10.0,
+      "grad_norm": 3.378390284106787,
+      "learning_rate": 1.3093770573896369e-06,
+      "loss": 0.0475,
+      "step": 595
+    },
+    {
+      "epoch": 10.0,
+      "eval_loss": 0.08747600764036179,
+      "eval_runtime": 83.5907,
+      "eval_samples_per_second": 1.125,
+      "eval_steps_per_second": 1.125,
       "step": 595
     },
     {
-      "epoch": 7.095238095238095,
-      "grad_norm": 2.736235330587379,
-      "learning_rate": 7.5648374743167e-06,
-      "loss": 0.0301,
+      "epoch": 10.016806722689076,
+      "grad_norm": 1.340634522580302,
+      "learning_rate": 1.286812958766106e-06,
+      "loss": 0.0302,
       "step": 596
     },
     {
-      "epoch": 7.107142857142857,
-      "grad_norm": 0.9799891778907858,
-      "learning_rate": 7.533662597455667e-06,
-      "loss": 0.0379,
+      "epoch": 10.033613445378151,
+      "grad_norm": 1.836375096978306,
+      "learning_rate": 1.2644315976900145e-06,
+      "loss": 0.0546,
       "step": 597
     },
     {
-      "epoch": 7.119047619047619,
-      "grad_norm": 1.1354342827683341,
-      "learning_rate": 7.5025132219188615e-06,
-      "loss": 0.0629,
+      "epoch": 10.050420168067227,
+      "grad_norm": 4.930603999286428,
+      "learning_rate": 1.242233443555051e-06,
+      "loss": 0.0534,
       "step": 598
     },
     {
-      "epoch": 7.130952380952381,
-      "grad_norm": 4.503381375894526,
-      "learning_rate": 7.471389669783183e-06,
-      "loss": 0.0348,
+      "epoch": 10.067226890756302,
+      "grad_norm": 2.204329623971633,
+      "learning_rate": 1.220218961912597e-06,
+      "loss": 0.0588,
       "step": 599
     },
     {
-      "epoch": 7.142857142857143,
-      "grad_norm": 4.170613919008802,
-      "learning_rate": 7.440292262858533e-06,
-      "loss": 0.1561,
+      "epoch": 10.084033613445378,
+      "grad_norm": 3.9225927136224987,
+      "learning_rate": 1.1983886144619527e-06,
+      "loss": 0.0427,
       "step": 600
     },
     {
-      "epoch": 7.154761904761905,
-      "grad_norm": 1.7324596706185202,
-      "learning_rate": 7.409221322684481e-06,
-      "loss": 0.0623,
+      "epoch": 10.100840336134453,
+      "grad_norm": 1.5324167733429486,
+      "learning_rate": 1.1767428590406648e-06,
+      "loss": 0.0269,
       "step": 601
     },
     {
-      "epoch": 7.166666666666667,
-      "grad_norm": 8.391742154063005,
-      "learning_rate": 7.378177170526928e-06,
-      "loss": 0.0505,
+      "epoch": 10.117647058823529,
+      "grad_norm": 1.7250937116057505,
+      "learning_rate": 1.1552821496149136e-06,
+      "loss": 0.0535,
       "step": 602
     },
     {
-      "epoch": 7.178571428571429,
-      "grad_norm": 2.0982830668406427,
-      "learning_rate": 7.347160127374804e-06,
-      "loss": 0.1156,
+      "epoch": 10.134453781512605,
+      "grad_norm": 0.7987608814501375,
+      "learning_rate": 1.134006936269999e-06,
+      "loss": 0.0173,
       "step": 603
     },
     {
-      "epoch": 7.190476190476191,
-      "grad_norm": 2.9010341886525337,
-      "learning_rate": 7.31617051393673e-06,
-      "loss": 0.0472,
+      "epoch": 10.15126050420168,
+      "grad_norm": 2.1053285256762733,
+      "learning_rate": 1.1129176652009043e-06,
+      "loss": 0.0406,
       "step": 604
     },
     {
-      "epoch": 7.2023809523809526,
-      "grad_norm": 6.688428366168943,
-      "learning_rate": 7.285208650637712e-06,
-      "loss": 0.0577,
+      "epoch": 10.168067226890756,
+      "grad_norm": 2.1300541811924094,
+      "learning_rate": 1.0920147787029233e-06,
+      "loss": 0.0461,
       "step": 605
     },
     {
-      "epoch": 7.214285714285714,
-      "grad_norm": 3.8530997432495844,
-      "learning_rate": 7.2542748576158284e-06,
-      "loss": 0.1247,
+      "epoch": 10.184873949579831,
+      "grad_norm": 1.5136074758546754,
+      "learning_rate": 1.0712987151624056e-06,
+      "loss": 0.0433,
       "step": 606
     },
     {
-      "epoch": 7.226190476190476,
-      "grad_norm": 2.2161474037024718,
-      "learning_rate": 7.223369454718918e-06,
-      "loss": 0.0544,
+      "epoch": 10.201680672268907,
+      "grad_norm": 1.6819131803547542,
+      "learning_rate": 1.05076990904754e-06,
+      "loss": 0.0529,
       "step": 607
     },
     {
-      "epoch": 7.238095238095238,
-      "grad_norm": 2.3619986375880067,
-      "learning_rate": 7.19249276150127e-06,
-      "loss": 0.0982,
+      "epoch": 10.218487394957982,
+      "grad_norm": 3.0644142026231727,
+      "learning_rate": 1.0304287908992626e-06,
+      "loss": 0.0532,
       "step": 608
     },
     {
-      "epoch": 7.25,
-      "grad_norm": 1.9276942223040003,
-      "learning_rate": 7.161645097220324e-06,
-      "loss": 0.0893,
+      "epoch": 10.235294117647058,
+      "grad_norm": 1.59650378921116,
+      "learning_rate": 1.010275787322219e-06,
+      "loss": 0.0538,
       "step": 609
     },
     {
-      "epoch": 7.261904761904762,
-      "grad_norm": 2.1139529848798833,
-      "learning_rate": 7.130826780833368e-06,
-      "loss": 0.0864,
+      "epoch": 10.252100840336134,
+      "grad_norm": 1.5363947337082766,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.0428,
       "step": 610
     },
     {
-      "epoch": 7.273809523809524,
-      "grad_norm": 1.571395158740333,
-      "learning_rate": 7.100038130994242e-06,
-      "loss": 0.0806,
+      "epoch": 10.268907563025211,
+      "grad_norm": 1.5003741433198152,
+      "learning_rate": 9.705358105653373e-07,
+      "loss": 0.0322,
       "step": 611
     },
     {
-      "epoch": 7.285714285714286,
-      "grad_norm": 1.4697576676883244,
-      "learning_rate": 7.069279466050035e-06,
-      "loss": 0.0702,
+      "epoch": 10.285714285714286,
+      "grad_norm": 2.0150139046727453,
+      "learning_rate": 9.509496708332233e-07,
+      "loss": 0.0395,
       "step": 612
     },
     {
-      "epoch": 7.2976190476190474,
-      "grad_norm": 1.343916000854979,
-      "learning_rate": 7.0385511040378026e-06,
-      "loss": 0.0589,
+      "epoch": 10.302521008403362,
+      "grad_norm": 2.2695792176582885,
+      "learning_rate": 9.315533125503051e-07,
+      "loss": 0.0755,
       "step": 613
     },
     {
-      "epoch": 7.309523809523809,
-      "grad_norm": 2.8165274419800013,
-      "learning_rate": 7.007853362681282e-06,
-      "loss": 0.0978,
+      "epoch": 10.319327731092438,
+      "grad_norm": 2.7525380883317134,
+      "learning_rate": 9.123471425072205e-07,
+      "loss": 0.0816,
       "step": 614
     },
     {
-      "epoch": 7.321428571428571,
-      "grad_norm": 0.8192824022307189,
-      "learning_rate": 6.9771865593875895e-06,
-      "loss": 0.0091,
+      "epoch": 10.336134453781513,
+      "grad_norm": 1.9575555600743102,
+      "learning_rate": 8.933315635058881e-07,
+      "loss": 0.0444,
       "step": 615
     },
     {
-      "epoch": 7.333333333333333,
-      "grad_norm": 1.7629740150881354,
-      "learning_rate": 6.946551011243958e-06,
-      "loss": 0.0594,
+      "epoch": 10.352941176470589,
+      "grad_norm": 1.3340430154540783,
+      "learning_rate": 8.745069743510393e-07,
+      "loss": 0.0289,
       "step": 616
     },
     {
-      "epoch": 7.345238095238095,
-      "grad_norm": 1.7603693751148732,
-      "learning_rate": 6.915947035014443e-06,
-      "loss": 0.1052,
+      "epoch": 10.369747899159664,
+      "grad_norm": 2.350943233326136,
+      "learning_rate": 8.558737698418762e-07,
+      "loss": 0.0557,
       "step": 617
     },
     {
-      "epoch": 7.357142857142857,
-      "grad_norm": 1.4991915006798426,
-      "learning_rate": 6.88537494713666e-06,
-      "loss": 0.018,
+      "epoch": 10.38655462184874,
+      "grad_norm": 1.3949514445639306,
+      "learning_rate": 8.374323407637741e-07,
+      "loss": 0.0327,
       "step": 618
     },
     {
-      "epoch": 7.369047619047619,
-      "grad_norm": 1.5015776512071202,
-      "learning_rate": 6.854835063718502e-06,
-      "loss": 0.0716,
+      "epoch": 10.403361344537815,
+      "grad_norm": 1.3745535851759583,
+      "learning_rate": 8.191830738800977e-07,
+      "loss": 0.0412,
       "step": 619
     },
     {
-      "epoch": 7.380952380952381,
-      "grad_norm": 1.3848929171774853,
-      "learning_rate": 6.8243277005348794e-06,
-      "loss": 0.062,
+      "epoch": 10.420168067226891,
+      "grad_norm": 2.797313621901212,
+      "learning_rate": 8.01126351924082e-07,
+      "loss": 0.0415,
       "step": 620
     },
     {
-      "epoch": 7.392857142857143,
-      "grad_norm": 0.8705318075848788,
-      "learning_rate": 6.793853173024449e-06,
-      "loss": 0.0392,
+      "epoch": 10.436974789915967,
+      "grad_norm": 2.0008993409572753,
+      "learning_rate": 7.83262553590809e-07,
+      "loss": 0.0473,
       "step": 621
     },
     {
-      "epoch": 7.404761904761905,
-      "grad_norm": 1.0304663794903173,
-      "learning_rate": 6.763411796286357e-06,
-      "loss": 0.0567,
+      "epoch": 10.453781512605042,
+      "grad_norm": 2.566675704823962,
+      "learning_rate": 7.655920535292682e-07,
+      "loss": 0.0342,
       "step": 622
     },
     {
-      "epoch": 7.416666666666667,
-      "grad_norm": 1.0436188142656049,
-      "learning_rate": 6.733003885076974e-06,
-      "loss": 0.0446,
+      "epoch": 10.470588235294118,
+      "grad_norm": 1.3899500927307256,
+      "learning_rate": 7.48115222334489e-07,
+      "loss": 0.0508,
       "step": 623
     },
     {
-      "epoch": 7.428571428571429,
-      "grad_norm": 1.9837389714659937,
-      "learning_rate": 6.702629753806653e-06,
-      "loss": 0.0752,
+      "epoch": 10.487394957983193,
+      "grad_norm": 2.160264462063064,
+      "learning_rate": 7.308324265397837e-07,
+      "loss": 0.0535,
       "step": 624
     },
     {
-      "epoch": 7.440476190476191,
-      "grad_norm": 2.5086887412568903,
-      "learning_rate": 6.672289716536467e-06,
-      "loss": 0.0896,
+      "epoch": 10.504201680672269,
+      "grad_norm": 1.6549288751982645,
+      "learning_rate": 7.137440286090436e-07,
+      "loss": 0.0527,
       "step": 625
     },
     {
-      "epoch": 7.4523809523809526,
-      "grad_norm": 2.1349931644939852,
-      "learning_rate": 6.64198408697496e-06,
-      "loss": 0.0893,
+      "epoch": 10.521008403361344,
+      "grad_norm": 1.7616896227790626,
+      "learning_rate": 6.968503869291521e-07,
+      "loss": 0.0463,
       "step": 626
     },
     {
-      "epoch": 7.464285714285714,
-      "grad_norm": 3.430269705220585,
-      "learning_rate": 6.611713178474917e-06,
-      "loss": 0.1424,
+      "epoch": 10.53781512605042,
+      "grad_norm": 1.882435848120671,
+      "learning_rate": 6.8015185580246e-07,
+      "loss": 0.0791,
       "step": 627
     },
     {
-      "epoch": 7.476190476190476,
-      "grad_norm": 1.9197254471391216,
-      "learning_rate": 6.581477304030111e-06,
-      "loss": 0.0187,
+      "epoch": 10.554621848739496,
+      "grad_norm": 2.1866189366482485,
+      "learning_rate": 6.636487854393536e-07,
+      "loss": 0.0595,
       "step": 628
     },
     {
-      "epoch": 7.488095238095238,
-      "grad_norm": 2.9945545535503313,
-      "learning_rate": 6.551276776272072e-06,
-      "loss": 0.1063,
+      "epoch": 10.571428571428571,
+      "grad_norm": 0.8846616976894125,
+      "learning_rate": 6.473415219509182e-07,
+      "loss": 0.0297,
       "step": 629
     },
     {
-      "epoch": 7.5,
-      "grad_norm": 2.1068433684122714,
-      "learning_rate": 6.521111907466856e-06,
-      "loss": 0.1139,
+      "epoch": 10.588235294117647,
+      "grad_norm": 2.3038569022667614,
+      "learning_rate": 6.31230407341672e-07,
+      "loss": 0.0679,
       "step": 630
     },
     {
-      "epoch": 7.511904761904762,
-      "grad_norm": 1.040378375451962,
-      "learning_rate": 6.490983009511813e-06,
-      "loss": 0.0594,
+      "epoch": 10.605042016806722,
+      "grad_norm": 1.0106801357197905,
+      "learning_rate": 6.153157795023956e-07,
+      "loss": 0.0271,
       "step": 631
     },
     {
-      "epoch": 7.523809523809524,
-      "grad_norm": 3.2385969175520164,
-      "learning_rate": 6.460890393932362e-06,
-      "loss": 0.0299,
+      "epoch": 10.621848739495798,
+      "grad_norm": 1.9606151371435605,
+      "learning_rate": 5.995979722030443e-07,
+      "loss": 0.0566,
       "step": 632
     },
     {
-      "epoch": 7.535714285714286,
-      "grad_norm": 1.1327987232905252,
-      "learning_rate": 6.430834371878769e-06,
-      "loss": 0.041,
+      "epoch": 10.638655462184873,
+      "grad_norm": 8.812038870074772,
+      "learning_rate": 5.840773150857526e-07,
+      "loss": 0.0936,
       "step": 633
     },
     {
-      "epoch": 7.5476190476190474,
-      "grad_norm": 3.2654270253868316,
-      "learning_rate": 6.400815254122943e-06,
-      "loss": 0.1143,
+      "epoch": 10.655462184873949,
+      "grad_norm": 2.37646496100217,
+      "learning_rate": 5.687541336579127e-07,
+      "loss": 0.0689,
       "step": 634
     },
     {
-      "epoch": 7.559523809523809,
-      "grad_norm": 1.384074755179732,
-      "learning_rate": 6.3708333510551965e-06,
-      "loss": 0.0449,
+      "epoch": 10.672268907563025,
+      "grad_norm": 1.722317758769946,
+      "learning_rate": 5.536287492853575e-07,
+      "loss": 0.0409,
       "step": 635
     },
     {
-      "epoch": 7.571428571428571,
-      "grad_norm": 1.1295585227576246,
-      "learning_rate": 6.340888972681063e-06,
-      "loss": 0.0336,
+      "epoch": 10.6890756302521,
+      "grad_norm": 1.7758828341496589,
+      "learning_rate": 5.387014791856127e-07,
+      "loss": 0.0501,
       "step": 636
     },
     {
-      "epoch": 7.583333333333333,
-      "grad_norm": 2.0470567579399543,
-      "learning_rate": 6.310982428618078e-06,
-      "loss": 0.0606,
+      "epoch": 10.705882352941176,
+      "grad_norm": 0.9972447472445873,
+      "learning_rate": 5.239726364212494e-07,
+      "loss": 0.0354,
       "step": 637
     },
     {
-      "epoch": 7.595238095238095,
-      "grad_norm": 1.743014394487192,
-      "learning_rate": 6.281114028092567e-06,
-      "loss": 0.0666,
+      "epoch": 10.722689075630251,
+      "grad_norm": 1.8890386119269527,
+      "learning_rate": 5.094425298933136e-07,
+      "loss": 0.0375,
       "step": 638
     },
     {
-      "epoch": 7.607142857142857,
-      "grad_norm": 2.248181677644741,
-      "learning_rate": 6.251284079936473e-06,
-      "loss": 0.0661,
+      "epoch": 10.739495798319329,
+      "grad_norm": 3.0981629606543426,
+      "learning_rate": 4.951114643348531e-07,
+      "loss": 0.0618,
       "step": 639
     },
     {
-      "epoch": 7.619047619047619,
-      "grad_norm": 2.8242756319277396,
-      "learning_rate": 6.221492892584153e-06,
-      "loss": 0.1071,
+      "epoch": 10.756302521008404,
+      "grad_norm": 3.3069946004768576,
+      "learning_rate": 4.809797403045224e-07,
+      "loss": 0.0738,
       "step": 640
     },
     {
-      "epoch": 7.630952380952381,
-      "grad_norm": 1.6137966144044473,
-      "learning_rate": 6.19174077406918e-06,
-      "loss": 0.0765,
+      "epoch": 10.77310924369748,
+      "grad_norm": 1.0677359119397434,
+      "learning_rate": 4.670476541802782e-07,
+      "loss": 0.0349,
       "step": 641
     },
     {
-      "epoch": 7.642857142857143,
-      "grad_norm": 1.6637953498949227,
-      "learning_rate": 6.162028032021168e-06,
-      "loss": 0.0386,
+      "epoch": 10.789915966386555,
+      "grad_norm": 1.8374696207253192,
+      "learning_rate": 4.533154981531718e-07,
+      "loss": 0.0431,
       "step": 642
     },
     {
-      "epoch": 7.654761904761905,
-      "grad_norm": 1.4989918311563193,
-      "learning_rate": 6.1323549736625856e-06,
-      "loss": 0.0627,
+      "epoch": 10.806722689075631,
+      "grad_norm": 3.948719353562884,
+      "learning_rate": 4.397835602212064e-07,
+      "loss": 0.0533,
       "step": 643
     },
     {
-      "epoch": 7.666666666666667,
-      "grad_norm": 1.8355150312652448,
-      "learning_rate": 6.102721905805583e-06,
-      "loss": 0.0824,
+      "epoch": 10.823529411764707,
+      "grad_norm": 1.572765088896342,
+      "learning_rate": 4.264521241833153e-07,
+      "loss": 0.0425,
       "step": 644
     },
     {
-      "epoch": 7.678571428571429,
-      "grad_norm": 2.2397838981140437,
-      "learning_rate": 6.073129134848819e-06,
-      "loss": 0.068,
+      "epoch": 10.840336134453782,
+      "grad_norm": 1.2552552948039755,
+      "learning_rate": 4.133214696333943e-07,
+      "loss": 0.0326,
       "step": 645
     },
     {
-      "epoch": 7.690476190476191,
-      "grad_norm": 1.539999994990287,
-      "learning_rate": 6.043576966774292e-06,
-      "loss": 0.0709,
+      "epoch": 10.857142857142858,
+      "grad_norm": 1.4869473389842671,
+      "learning_rate": 4.003918719544464e-07,
+      "loss": 0.0334,
       "step": 646
     },
     {
-      "epoch": 7.7023809523809526,
-      "grad_norm": 3.9837741827034776,
-      "learning_rate": 6.014065707144177e-06,
-      "loss": 0.0807,
+      "epoch": 10.873949579831933,
+      "grad_norm": 1.4103343408301399,
+      "learning_rate": 3.876636023128022e-07,
+      "loss": 0.035,
       "step": 647
     },
     {
-      "epoch": 7.714285714285714,
-      "grad_norm": 2.4912819002955846,
-      "learning_rate": 5.984595661097663e-06,
-      "loss": 0.0735,
+      "epoch": 10.890756302521009,
+      "grad_norm": 1.491957032559717,
+      "learning_rate": 3.7513692765243637e-07,
+      "loss": 0.0446,
       "step": 648
     },
     {
-      "epoch": 7.726190476190476,
-      "grad_norm": 3.2181157159265785,
-      "learning_rate": 5.9551671333478056e-06,
-      "loss": 0.0644,
+      "epoch": 10.907563025210084,
+      "grad_norm": 5.277585616799005,
+      "learning_rate": 3.628121106893701e-07,
+      "loss": 0.063,
       "step": 649
     },
     {
-      "epoch": 7.738095238095238,
-      "grad_norm": 1.5618524452415126,
-      "learning_rate": 5.925780428178365e-06,
-      "loss": 0.0825,
+      "epoch": 10.92436974789916,
+      "grad_norm": 2.3243782807441837,
+      "learning_rate": 3.50689409906152e-07,
+      "loss": 0.0408,
       "step": 650
     },
     {
-      "epoch": 7.75,
-      "grad_norm": 4.130128619978615,
-      "learning_rate": 5.896435849440671e-06,
-      "loss": 0.0621,
+      "epoch": 10.941176470588236,
+      "grad_norm": 5.5511386963840605,
+      "learning_rate": 3.3876907954644933e-07,
+      "loss": 0.0991,
       "step": 651
     },
     {
-      "epoch": 7.761904761904762,
-      "grad_norm": 2.069627837047999,
-      "learning_rate": 5.867133700550479e-06,
-      "loss": 0.086,
+      "epoch": 10.957983193277311,
+      "grad_norm": 0.9258571015003478,
+      "learning_rate": 3.2705136960970554e-07,
+      "loss": 0.0324,
       "step": 652
     },
     {
-      "epoch": 7.773809523809524,
-      "grad_norm": 2.2145675279733537,
-      "learning_rate": 5.837874284484825e-06,
-      "loss": 0.0814,
+      "epoch": 10.974789915966387,
+      "grad_norm": 3.7891311679004582,
+      "learning_rate": 3.1553652584590864e-07,
+      "loss": 0.0649,
       "step": 653
     },
     {
-      "epoch": 7.785714285714286,
-      "grad_norm": 3.6395035540854574,
-      "learning_rate": 5.808657903778894e-06,
-      "loss": 0.0271,
+      "epoch": 10.991596638655462,
+      "grad_norm": 1.3430987417489015,
+      "learning_rate": 3.0422478975042245e-07,
+      "loss": 0.0329,
       "step": 654
     },
     {
-      "epoch": 7.7976190476190474,
-      "grad_norm": 1.8760063884280842,
-      "learning_rate": 5.779484860522904e-06,
-      "loss": 0.0838,
+      "epoch": 10.991596638655462,
+      "eval_loss": 0.08618413656949997,
+      "eval_runtime": 84.2423,
+      "eval_samples_per_second": 1.116,
+      "eval_steps_per_second": 1.116,
+      "step": 654
+    },
+    {
+      "epoch": 11.008403361344538,
+      "grad_norm": 10.112136406292526,
+      "learning_rate": 2.931163985589369e-07,
+      "loss": 0.0792,
       "step": 655
     },
     {
-      "epoch": 7.809523809523809,
-      "grad_norm": 1.8187225689477617,
-      "learning_rate": 5.750355456358976e-06,
-      "loss": 0.0666,
+      "epoch": 11.025210084033613,
+      "grad_norm": 1.2237182031841543,
+      "learning_rate": 2.8221158524248003e-07,
+      "loss": 0.0375,
       "step": 656
     },
     {
-      "epoch": 7.821428571428571,
-      "grad_norm": 1.8701879030711632,
-      "learning_rate": 5.721269992478002e-06,
-      "loss": 0.0665,
+      "epoch": 11.042016806722689,
+      "grad_norm": 1.8203551353192673,
+      "learning_rate": 2.7151057850253957e-07,
+      "loss": 0.0529,
       "step": 657
     },
     {
-      "epoch": 7.833333333333333,
-      "grad_norm": 2.008421784606381,
-      "learning_rate": 5.692228769616559e-06,
-      "loss": 0.0183,
+      "epoch": 11.058823529411764,
+      "grad_norm": 1.6032849699615903,
+      "learning_rate": 2.6101360276626795e-07,
+      "loss": 0.0446,
       "step": 658
     },
     {
-      "epoch": 7.845238095238095,
-      "grad_norm": 1.1103842322825477,
-      "learning_rate": 5.663232088053763e-06,
-      "loss": 0.0469,
+      "epoch": 11.07563025210084,
+      "grad_norm": 1.545240215165262,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.0356,
       "step": 659
     },
     {
-      "epoch": 7.857142857142857,
-      "grad_norm": 2.9071912477561095,
-      "learning_rate": 5.6342802476082014e-06,
-      "loss": 0.0963,
+      "epoch": 11.092436974789916,
+      "grad_norm": 1.5501298244234751,
+      "learning_rate": 2.406326206134724e-07,
+      "loss": 0.0413,
       "step": 660
     },
     {
-      "epoch": 7.869047619047619,
-      "grad_norm": 1.4498002971746087,
-      "learning_rate": 5.6053735476348025e-06,
-      "loss": 0.0482,
+      "epoch": 11.109243697478991,
+      "grad_norm": 3.6911299360545278,
+      "learning_rate": 2.3074904163764012e-07,
+      "loss": 0.1112,
       "step": 661
     },
     {
-      "epoch": 7.880952380952381,
-      "grad_norm": 2.159820723697225,
-      "learning_rate": 5.576512287021765e-06,
-      "loss": 0.0876,
+      "epoch": 11.126050420168067,
+      "grad_norm": 1.668296969639716,
+      "learning_rate": 2.210703485378929e-07,
+      "loss": 0.0343,
       "step": 662
     },
     {
-      "epoch": 7.892857142857143,
-      "grad_norm": 3.065401466063626,
-      "learning_rate": 5.5476967641874416e-06,
-      "loss": 0.0932,
+      "epoch": 11.142857142857142,
+      "grad_norm": 1.464185072769606,
+      "learning_rate": 2.115967443008804e-07,
+      "loss": 0.042,
       "step": 663
     },
     {
-      "epoch": 7.904761904761905,
-      "grad_norm": 5.139368448283141,
-      "learning_rate": 5.518927277077284e-06,
-      "loss": 0.1097,
+      "epoch": 11.159663865546218,
+      "grad_norm": 3.21826813941365,
+      "learning_rate": 2.0232842761201854e-07,
+      "loss": 0.0393,
       "step": 664
     },
     {
-      "epoch": 7.916666666666667,
-      "grad_norm": 2.5405581307288605,
-      "learning_rate": 5.49020412316073e-06,
-      "loss": 0.0904,
+      "epoch": 11.176470588235293,
+      "grad_norm": 1.5268309356925824,
+      "learning_rate": 1.9326559285132495e-07,
+      "loss": 0.0396,
       "step": 665
     },
     {
-      "epoch": 7.928571428571429,
-      "grad_norm": 0.45052066247569683,
-      "learning_rate": 5.4615275994281514e-06,
-      "loss": 0.0051,
+      "epoch": 11.193277310924369,
+      "grad_norm": 2.7559911444343625,
+      "learning_rate": 1.844084300893456e-07,
+      "loss": 0.0378,
       "step": 666
     },
     {
-      "epoch": 7.940476190476191,
-      "grad_norm": 2.3387165641238234,
-      "learning_rate": 5.432898002387783e-06,
-      "loss": 0.051,
+      "epoch": 11.210084033613445,
+      "grad_norm": 2.064362647787835,
+      "learning_rate": 1.7575712508316244e-07,
+      "loss": 0.0692,
       "step": 667
     },
     {
-      "epoch": 7.9523809523809526,
-      "grad_norm": 1.6393139092131652,
-      "learning_rate": 5.404315628062631e-06,
-      "loss": 0.0681,
+      "epoch": 11.22689075630252,
+      "grad_norm": 2.5064352487208676,
+      "learning_rate": 1.673118592724987e-07,
+      "loss": 0.0625,
       "step": 668
     },
     {
-      "epoch": 7.964285714285714,
-      "grad_norm": 1.856864744878801,
-      "learning_rate": 5.375780771987449e-06,
-      "loss": 0.016,
+      "epoch": 11.243697478991596,
+      "grad_norm": 2.7390411371250774,
+      "learning_rate": 1.5907280977591866e-07,
+      "loss": 0.0737,
       "step": 669
     },
     {
-      "epoch": 7.976190476190476,
-      "grad_norm": 2.1504287060037397,
-      "learning_rate": 5.347293729205644e-06,
-      "loss": 0.0984,
+      "epoch": 11.260504201680673,
+      "grad_norm": 0.9309858940918065,
+      "learning_rate": 1.5104014938710498e-07,
+      "loss": 0.0266,
       "step": 670
     },
     {
-      "epoch": 7.988095238095238,
-      "grad_norm": 1.0739005598429683,
-      "learning_rate": 5.318854794266268e-06,
-      "loss": 0.0733,
+      "epoch": 11.277310924369749,
+      "grad_norm": 1.851032261256866,
+      "learning_rate": 1.4321404657124393e-07,
+      "loss": 0.0493,
       "step": 671
     },
     {
-      "epoch": 8.0,
-      "grad_norm": 13.629034996878698,
-      "learning_rate": 5.290464261220927e-06,
-      "loss": 0.0811,
-      "step": 672
-    },
-    {
-      "epoch": 8.0,
-      "eval_loss": 0.18508009612560272,
-      "eval_runtime": 38.0386,
-      "eval_samples_per_second": 1.577,
-      "eval_steps_per_second": 1.577,
+      "epoch": 11.294117647058824,
+      "grad_norm": 1.0812191705502274,
+      "learning_rate": 1.3559466546148369e-07,
+      "loss": 0.0434,
       "step": 672
     },
     {
-      "epoch": 8.011904761904763,
-      "grad_norm": 1.9110702680782508,
-      "learning_rate": 5.262122423620782e-06,
-      "loss": 0.0195,
+      "epoch": 11.3109243697479,
+      "grad_norm": 1.8327569259876093,
+      "learning_rate": 1.2818216585549824e-07,
+      "loss": 0.0449,
       "step": 673
     },
     {
-      "epoch": 8.023809523809524,
-      "grad_norm": 1.2576546338351324,
-      "learning_rate": 5.2338295745134795e-06,
-      "loss": 0.0727,
+      "epoch": 11.327731092436975,
+      "grad_norm": 1.160640896943556,
+      "learning_rate": 1.209767032121345e-07,
+      "loss": 0.0328,
       "step": 674
     },
     {
-      "epoch": 8.035714285714286,
-      "grad_norm": 3.5597917543308806,
-      "learning_rate": 5.205586006440149e-06,
-      "loss": 0.0967,
+      "epoch": 11.344537815126051,
+      "grad_norm": 2.0693866133338017,
+      "learning_rate": 1.1397842864814712e-07,
+      "loss": 0.0599,
       "step": 675
     },
     {
-      "epoch": 8.047619047619047,
-      "grad_norm": 1.6721296431336412,
-      "learning_rate": 5.177392011432368e-06,
-      "loss": 0.0339,
+      "epoch": 11.361344537815127,
+      "grad_norm": 1.8624412914202513,
+      "learning_rate": 1.0718748893503883e-07,
+      "loss": 0.0458,
       "step": 676
     },
     {
-      "epoch": 8.05952380952381,
-      "grad_norm": 1.2478157185355454,
-      "learning_rate": 5.1492478810091184e-06,
-      "loss": 0.0276,
+      "epoch": 11.378151260504202,
+      "grad_norm": 2.0035143994192377,
+      "learning_rate": 1.0060402649597178e-07,
+      "loss": 0.0472,
       "step": 677
     },
     {
-      "epoch": 8.071428571428571,
-      "grad_norm": 2.9389850134580446,
-      "learning_rate": 5.121153906173826e-06,
-      "loss": 0.0583,
+      "epoch": 11.394957983193278,
+      "grad_norm": 2.6901112576784447,
+      "learning_rate": 9.422817940278773e-08,
+      "loss": 0.0458,
       "step": 678
     },
     {
-      "epoch": 8.083333333333334,
-      "grad_norm": 2.6905956141154848,
-      "learning_rate": 5.093110377411292e-06,
-      "loss": 0.0216,
+      "epoch": 11.411764705882353,
+      "grad_norm": 1.4284628255192529,
+      "learning_rate": 8.806008137311028e-08,
+      "loss": 0.0327,
       "step": 679
     },
     {
-      "epoch": 8.095238095238095,
-      "grad_norm": 2.1426781345772508,
-      "learning_rate": 5.06511758468474e-06,
-      "loss": 0.0768,
+      "epoch": 11.428571428571429,
+      "grad_norm": 2.222615104628747,
+      "learning_rate": 8.209986176753947e-08,
+      "loss": 0.041,
       "step": 680
     },
     {
-      "epoch": 8.107142857142858,
-      "grad_norm": 3.5859888616619515,
-      "learning_rate": 5.037175817432779e-06,
-      "loss": 0.1097,
+      "epoch": 11.445378151260504,
+      "grad_norm": 3.9740102271718634,
+      "learning_rate": 7.634764558693941e-08,
+      "loss": 0.0729,
       "step": 681
     },
     {
-      "epoch": 8.119047619047619,
-      "grad_norm": 1.1020497087264263,
-      "learning_rate": 5.009285364566435e-06,
-      "loss": 0.0304,
+      "epoch": 11.46218487394958,
+      "grad_norm": 1.4918545836760815,
+      "learning_rate": 7.080355346981815e-08,
+      "loss": 0.0566,
       "step": 682
     },
     {
-      "epoch": 8.130952380952381,
-      "grad_norm": 1.3896812421288112,
-      "learning_rate": 4.98144651446616e-06,
-      "loss": 0.0491,
+      "epoch": 11.478991596638656,
+      "grad_norm": 2.521127424464659,
+      "learning_rate": 6.546770168979421e-08,
+      "loss": 0.0437,
       "step": 683
     },
     {
-      "epoch": 8.142857142857142,
-      "grad_norm": 1.2000724845656414,
-      "learning_rate": 4.953659554978831e-06,
-      "loss": 0.0409,
+      "epoch": 11.495798319327731,
+      "grad_norm": 1.2741993338648574,
+      "learning_rate": 6.034020215316184e-08,
+      "loss": 0.0349,
       "step": 684
     },
     {
-      "epoch": 8.154761904761905,
-      "grad_norm": 1.142623132745125,
-      "learning_rate": 4.925924773414809e-06,
-      "loss": 0.0371,
+      "epoch": 11.512605042016807,
+      "grad_norm": 1.296314142560303,
+      "learning_rate": 5.5421162396542824e-08,
+      "loss": 0.0356,
       "step": 685
     },
     {
-      "epoch": 8.166666666666666,
-      "grad_norm": 2.721525193056032,
-      "learning_rate": 4.8982424565449274e-06,
-      "loss": 0.06,
+      "epoch": 11.529411764705882,
+      "grad_norm": 1.8595898721766801,
+      "learning_rate": 5.071068558462733e-08,
+      "loss": 0.0559,
       "step": 686
     },
     {
-      "epoch": 8.178571428571429,
-      "grad_norm": 1.0659298645748037,
-      "learning_rate": 4.870612890597564e-06,
-      "loss": 0.0305,
+      "epoch": 11.546218487394958,
+      "grad_norm": 8.704714098773897,
+      "learning_rate": 4.6208870508017703e-08,
+      "loss": 0.101,
       "step": 687
     },
     {
-      "epoch": 8.19047619047619,
-      "grad_norm": 4.6764392493870295,
-      "learning_rate": 4.843036361255654e-06,
-      "loss": 0.1477,
+      "epoch": 11.563025210084033,
+      "grad_norm": 2.032442231084466,
+      "learning_rate": 4.191581158115021e-08,
+      "loss": 0.0478,
       "step": 688
     },
     {
-      "epoch": 8.202380952380953,
-      "grad_norm": 2.8740981280962097,
-      "learning_rate": 4.815513153653758e-06,
-      "loss": 0.0549,
+      "epoch": 11.579831932773109,
+      "grad_norm": 9.418069463610777,
+      "learning_rate": 3.783159884031773e-08,
+      "loss": 0.1048,
       "step": 689
     },
     {
-      "epoch": 8.214285714285714,
-      "grad_norm": 3.606494816871051,
-      "learning_rate": 4.788043552375087e-06,
-      "loss": 0.0506,
+      "epoch": 11.596638655462185,
+      "grad_norm": 1.0069915518246801,
+      "learning_rate": 3.3956317941779004e-08,
+      "loss": 0.032,
       "step": 690
     },
     {
-      "epoch": 8.226190476190476,
-      "grad_norm": 4.537509345710984,
-      "learning_rate": 4.760627841448595e-06,
-      "loss": 0.0381,
+      "epoch": 11.61344537815126,
+      "grad_norm": 1.3869570014003472,
+      "learning_rate": 3.029005015996789e-08,
+      "loss": 0.0276,
       "step": 691
     },
     {
-      "epoch": 8.238095238095237,
-      "grad_norm": 1.9487056080497047,
-      "learning_rate": 4.733266304346005e-06,
-      "loss": 0.0575,
+      "epoch": 11.630252100840336,
+      "grad_norm": 2.0275074366047128,
+      "learning_rate": 2.6832872385783583e-08,
+      "loss": 0.041,
       "step": 692
     },
     {
-      "epoch": 8.25,
-      "grad_norm": 1.372774061299278,
-      "learning_rate": 4.705959223978908e-06,
-      "loss": 0.0362,
+      "epoch": 11.647058823529411,
+      "grad_norm": 3.7065994847690407,
+      "learning_rate": 2.3584857124977488e-08,
+      "loss": 0.058,
       "step": 693
     },
     {
-      "epoch": 8.261904761904763,
-      "grad_norm": 2.4087403103866882,
-      "learning_rate": 4.678706882695824e-06,
-      "loss": 0.0822,
+      "epoch": 11.663865546218487,
+      "grad_norm": 1.1492957912068942,
+      "learning_rate": 2.054607249663665e-08,
+      "loss": 0.04,
       "step": 694
     },
     {
-      "epoch": 8.273809523809524,
-      "grad_norm": 2.414899294008852,
-      "learning_rate": 4.651509562279276e-06,
-      "loss": 0.0634,
+      "epoch": 11.680672268907562,
+      "grad_norm": 7.151682674273455,
+      "learning_rate": 1.7716582231752656e-08,
+      "loss": 0.077,
       "step": 695
     },
     {
-      "epoch": 8.285714285714286,
-      "grad_norm": 3.378577203088908,
-      "learning_rate": 4.624367543942899e-06,
-      "loss": 0.0647,
+      "epoch": 11.697478991596638,
+      "grad_norm": 1.5415699598561934,
+      "learning_rate": 1.509644567188717e-08,
+      "loss": 0.041,
       "step": 696
     },
     {
-      "epoch": 8.297619047619047,
-      "grad_norm": 5.150913422584667,
-      "learning_rate": 4.597281108328502e-06,
-      "loss": 0.0849,
+      "epoch": 11.714285714285714,
+      "grad_norm": 1.7646669493531948,
+      "learning_rate": 1.2685717767921823e-08,
+      "loss": 0.0518,
       "step": 697
     },
     {
-      "epoch": 8.30952380952381,
-      "grad_norm": 1.1399120801392448,
-      "learning_rate": 4.570250535503196e-06,
-      "loss": 0.0354,
+      "epoch": 11.731092436974789,
+      "grad_norm": 4.9082532974240465,
+      "learning_rate": 1.048444907891244e-08,
+      "loss": 0.1108,
       "step": 698
     },
     {
-      "epoch": 8.321428571428571,
-      "grad_norm": 1.4227847802019646,
-      "learning_rate": 4.543276104956472e-06,
-      "loss": 0.0587,
+      "epoch": 11.747899159663866,
+      "grad_norm": 3.994721765198761,
+      "learning_rate": 8.492685771025466e-09,
+      "loss": 0.0613,
       "step": 699
     },
     {
-      "epoch": 8.333333333333334,
-      "grad_norm": 1.464743918175674,
-      "learning_rate": 4.5163580955973384e-06,
-      "loss": 0.0812,
+      "epoch": 11.764705882352942,
+      "grad_norm": 2.688716545259077,
+      "learning_rate": 6.710469616569848e-09,
+      "loss": 0.0534,
       "step": 700
     },
     {
-      "epoch": 8.345238095238095,
-      "grad_norm": 1.3386170144483294,
-      "learning_rate": 4.489496785751407e-06,
-      "loss": 0.0644,
+      "epoch": 11.781512605042018,
+      "grad_norm": 2.097132472666691,
+      "learning_rate": 5.137837993121064e-09,
+      "loss": 0.0534,
       "step": 701
     },
     {
-      "epoch": 8.357142857142858,
-      "grad_norm": 1.127458731365185,
-      "learning_rate": 4.4626924531580395e-06,
-      "loss": 0.0308,
+      "epoch": 11.798319327731093,
+      "grad_norm": 1.036594664039175,
+      "learning_rate": 3.774823882738421e-09,
+      "loss": 0.0338,
       "step": 702
     },
     {
-      "epoch": 8.369047619047619,
-      "grad_norm": 2.1999026564423425,
-      "learning_rate": 4.435945374967471e-06,
-      "loss": 0.0993,
+      "epoch": 11.815126050420169,
+      "grad_norm": 2.2273501309394907,
+      "learning_rate": 2.6214558712722714e-09,
+      "loss": 0.0337,
       "step": 703
     },
     {
-      "epoch": 8.380952380952381,
-      "grad_norm": 1.364108162519796,
-      "learning_rate": 4.4092558277379235e-06,
-      "loss": 0.071,
+      "epoch": 11.831932773109244,
+      "grad_norm": 1.4175794307129967,
+      "learning_rate": 1.677758147762276e-09,
+      "loss": 0.0304,
       "step": 704
     },
     {
-      "epoch": 8.392857142857142,
-      "grad_norm": 1.204961209910452,
-      "learning_rate": 4.382624087432784e-06,
-      "loss": 0.0438,
+      "epoch": 11.84873949579832,
+      "grad_norm": 2.118091027580824,
+      "learning_rate": 9.43750503935581e-10,
+      "loss": 0.0483,
       "step": 705
     },
     {
-      "epoch": 8.404761904761905,
-      "grad_norm": 2.1404897481502503,
-      "learning_rate": 4.356050429417711e-06,
-      "loss": 0.0177,
+      "epoch": 11.865546218487395,
+      "grad_norm": 7.711002671886926,
+      "learning_rate": 4.1944833378604334e-10,
+      "loss": 0.1102,
       "step": 706
     },
     {
-      "epoch": 8.416666666666666,
-      "grad_norm": 2.248346002608694,
-      "learning_rate": 4.329535128457822e-06,
-      "loss": 0.0747,
+      "epoch": 11.882352941176471,
+      "grad_norm": 1.7414362067953533,
+      "learning_rate": 1.0486263325559798e-10,
+      "loss": 0.0357,
       "step": 707
     },
     {
-      "epoch": 8.428571428571429,
-      "grad_norm": 1.8017077511033046,
-      "learning_rate": 4.303078458714824e-06,
-      "loss": 0.055,
+      "epoch": 11.899159663865547,
+      "grad_norm": 1.2682884045214264,
+      "learning_rate": 0.0,
+      "loss": 0.0266,
       "step": 708
     },
     {
-      "epoch": 8.44047619047619,
-      "grad_norm": 3.1411734635110826,
-      "learning_rate": 4.2766806937442025e-06,
-      "loss": 0.1054,
-      "step": 709
-    },
-    {
-      "epoch": 8.452380952380953,
-      "grad_norm": 1.9831648552999164,
-      "learning_rate": 4.250342106492371e-06,
-      "loss": 0.0455,
-      "step": 710
-    },
-    {
-      "epoch": 8.464285714285714,
-      "grad_norm": 1.3064388791471728,
-      "learning_rate": 4.224062969293873e-06,
-      "loss": 0.0118,
-      "step": 711
-    },
-    {
-      "epoch": 8.476190476190476,
-      "grad_norm": 4.303236040730501,
-      "learning_rate": 4.197843553868538e-06,
-      "loss": 0.1303,
-      "step": 712
-    },
-    {
-      "epoch": 8.488095238095237,
-      "grad_norm": 3.2998036981279277,
-      "learning_rate": 4.1716841313187e-06,
-      "loss": 0.0898,
-      "step": 713
-    },
-    {
-      "epoch": 8.5,
-      "grad_norm": 2.531047135551161,
-      "learning_rate": 4.145584972126377e-06,
-      "loss": 0.0926,
-      "step": 714
-    },
-    {
-      "epoch": 8.511904761904763,
-      "grad_norm": 3.84439794247204,
-      "learning_rate": 4.119546346150478e-06,
-      "loss": 0.1078,
-      "step": 715
-    },
-    {
-      "epoch": 8.523809523809524,
-      "grad_norm": 1.4003583986318646,
-      "learning_rate": 4.093568522624012e-06,
-      "loss": 0.0529,
-      "step": 716
-    },
-    {
-      "epoch": 8.535714285714286,
-      "grad_norm": 1.5373028609830812,
-      "learning_rate": 4.0676517701513015e-06,
-      "loss": 0.0128,
-      "step": 717
-    },
-    {
-      "epoch": 8.547619047619047,
-      "grad_norm": 2.0378816437403047,
-      "learning_rate": 4.04179635670522e-06,
-      "loss": 0.1006,
-      "step": 718
-    },
-    {
-      "epoch": 8.55952380952381,
-      "grad_norm": 0.9253802190426968,
-      "learning_rate": 4.016002549624408e-06,
-      "loss": 0.0256,
-      "step": 719
-    },
-    {
-      "epoch": 8.571428571428571,
-      "grad_norm": 1.8878359466590395,
-      "learning_rate": 3.990270615610502e-06,
-      "loss": 0.0427,
-      "step": 720
-    },
-    {
-      "epoch": 8.583333333333334,
-      "grad_norm": 1.8998654829048824,
-      "learning_rate": 3.964600820725399e-06,
-      "loss": 0.0149,
-      "step": 721
-    },
-    {
-      "epoch": 8.595238095238095,
-      "grad_norm": 2.1785811911898043,
-      "learning_rate": 3.938993430388481e-06,
-      "loss": 0.1016,
-      "step": 722
-    },
-    {
-      "epoch": 8.607142857142858,
-      "grad_norm": 1.6005146404043937,
-      "learning_rate": 3.9134487093738906e-06,
-      "loss": 0.098,
-      "step": 723
-    },
-    {
-      "epoch": 8.619047619047619,
-      "grad_norm": 1.944745995853955,
-      "learning_rate": 3.8879669218077785e-06,
-      "loss": 0.0867,
-      "step": 724
-    },
-    {
-      "epoch": 8.630952380952381,
-      "grad_norm": 3.442868509135672,
-      "learning_rate": 3.862548331165589e-06,
-      "loss": 0.0251,
-      "step": 725
-    },
-    {
-      "epoch": 8.642857142857142,
-      "grad_norm": 1.8096775564399332,
-      "learning_rate": 3.837193200269309e-06,
-      "loss": 0.0788,
-      "step": 726
-    },
-    {
-      "epoch": 8.654761904761905,
-      "grad_norm": 3.475557028703863,
-      "learning_rate": 3.811901791284788e-06,
-      "loss": 0.0846,
-      "step": 727
-    },
-    {
-      "epoch": 8.666666666666666,
-      "grad_norm": 2.2234463246243465,
-      "learning_rate": 3.7866743657189863e-06,
-      "loss": 0.0575,
-      "step": 728
-    },
-    {
-      "epoch": 8.678571428571429,
-      "grad_norm": 4.393598853012906,
-      "learning_rate": 3.761511184417306e-06,
-      "loss": 0.171,
-      "step": 729
-    },
-    {
-      "epoch": 8.69047619047619,
-      "grad_norm": 1.780187643448173,
-      "learning_rate": 3.736412507560876e-06,
-      "loss": 0.0431,
-      "step": 730
-    },
-    {
-      "epoch": 8.702380952380953,
-      "grad_norm": 1.2226552976534588,
-      "learning_rate": 3.711378594663857e-06,
-      "loss": 0.032,
-      "step": 731
-    },
-    {
-      "epoch": 8.714285714285714,
-      "grad_norm": 1.9370245760802747,
-      "learning_rate": 3.6864097045707783e-06,
-      "loss": 0.0444,
-      "step": 732
-    },
-    {
-      "epoch": 8.726190476190476,
-      "grad_norm": 3.5771359305604826,
-      "learning_rate": 3.6615060954538383e-06,
-      "loss": 0.0935,
-      "step": 733
-    },
-    {
-      "epoch": 8.738095238095237,
-      "grad_norm": 1.0674492370205682,
-      "learning_rate": 3.636668024810256e-06,
-      "loss": 0.0561,
-      "step": 734
-    },
-    {
-      "epoch": 8.75,
-      "grad_norm": 1.5812111523180694,
-      "learning_rate": 3.6118957494595885e-06,
-      "loss": 0.0476,
-      "step": 735
-    },
-    {
-      "epoch": 8.761904761904763,
-      "grad_norm": 1.5340016226171458,
-      "learning_rate": 3.587189525541097e-06,
-      "loss": 0.0649,
-      "step": 736
-    },
-    {
-      "epoch": 8.773809523809524,
-      "grad_norm": 4.646161423857292,
-      "learning_rate": 3.5625496085110757e-06,
-      "loss": 0.0356,
-      "step": 737
-    },
-    {
-      "epoch": 8.785714285714286,
-      "grad_norm": 2.3591372490715248,
-      "learning_rate": 3.537976253140232e-06,
-      "loss": 0.0535,
-      "step": 738
-    },
-    {
-      "epoch": 8.797619047619047,
-      "grad_norm": 3.09871629569407,
-      "learning_rate": 3.513469713511032e-06,
-      "loss": 0.0999,
-      "step": 739
-    },
-    {
-      "epoch": 8.80952380952381,
-      "grad_norm": 3.317284922510724,
-      "learning_rate": 3.4890302430150913e-06,
-      "loss": 0.0616,
-      "step": 740
-    },
-    {
-      "epoch": 8.821428571428571,
-      "grad_norm": 3.4094509283498127,
-      "learning_rate": 3.4646580943505483e-06,
-      "loss": 0.1103,
-      "step": 741
-    },
-    {
-      "epoch": 8.833333333333334,
-      "grad_norm": 3.729118936004436,
-      "learning_rate": 3.4403535195194393e-06,
-      "loss": 0.0514,
-      "step": 742
-    },
-    {
-      "epoch": 8.845238095238095,
-      "grad_norm": 7.947111022757592,
-      "learning_rate": 3.4161167698251176e-06,
-      "loss": 0.1472,
-      "step": 743
-    },
-    {
-      "epoch": 8.857142857142858,
-      "grad_norm": 1.5891949285087759,
-      "learning_rate": 3.391948095869628e-06,
-      "loss": 0.0649,
-      "step": 744
-    },
-    {
-      "epoch": 8.869047619047619,
-      "grad_norm": 1.688044557138942,
-      "learning_rate": 3.367847747551143e-06,
-      "loss": 0.0369,
-      "step": 745
-    },
-    {
-      "epoch": 8.880952380952381,
-      "grad_norm": 2.1729484304523172,
-      "learning_rate": 3.3438159740613474e-06,
-      "loss": 0.0483,
-      "step": 746
-    },
-    {
-      "epoch": 8.892857142857142,
-      "grad_norm": 1.704323955661607,
-      "learning_rate": 3.3198530238828973e-06,
-      "loss": 0.0679,
-      "step": 747
-    },
-    {
-      "epoch": 8.904761904761905,
-      "grad_norm": 1.0039040994703539,
-      "learning_rate": 3.2959591447868177e-06,
-      "loss": 0.0479,
-      "step": 748
-    },
-    {
-      "epoch": 8.916666666666666,
-      "grad_norm": 2.3677908012699014,
-      "learning_rate": 3.272134583829966e-06,
-      "loss": 0.0857,
-      "step": 749
-    },
-    {
-      "epoch": 8.928571428571429,
-      "grad_norm": 2.7444347195294685,
-      "learning_rate": 3.2483795873524625e-06,
-      "loss": 0.1112,
-      "step": 750
-    },
-    {
-      "epoch": 8.94047619047619,
-      "grad_norm": 1.2943783268976876,
-      "learning_rate": 3.224694400975145e-06,
-      "loss": 0.0266,
-      "step": 751
-    },
-    {
-      "epoch": 8.952380952380953,
-      "grad_norm": 3.044581999259029,
-      "learning_rate": 3.2010792695970407e-06,
-      "loss": 0.1023,
-      "step": 752
-    },
-    {
-      "epoch": 8.964285714285714,
-      "grad_norm": 2.6509128482408557,
-      "learning_rate": 3.177534437392813e-06,
-      "loss": 0.0748,
-      "step": 753
-    },
-    {
-      "epoch": 8.976190476190476,
-      "grad_norm": 1.221983721384426,
-      "learning_rate": 3.154060147810266e-06,
-      "loss": 0.0421,
-      "step": 754
-    },
-    {
-      "epoch": 8.988095238095237,
-      "grad_norm": 2.3808451565996958,
-      "learning_rate": 3.1306566435677877e-06,
-      "loss": 0.0184,
-      "step": 755
-    },
-    {
-      "epoch": 9.0,
-      "grad_norm": 1.4618055755215726,
-      "learning_rate": 3.1073241666518817e-06,
-      "loss": 0.0376,
-      "step": 756
-    },
-    {
-      "epoch": 9.0,
-      "eval_loss": 0.22630615532398224,
-      "eval_runtime": 38.2176,
-      "eval_samples_per_second": 1.57,
-      "eval_steps_per_second": 1.57,
-      "step": 756
-    },
-    {
-      "epoch": 9.011904761904763,
-      "grad_norm": 1.2638660863464435,
-      "learning_rate": 3.084062958314642e-06,
-      "loss": 0.0358,
-      "step": 757
-    },
-    {
-      "epoch": 9.023809523809524,
-      "grad_norm": 2.0988738098189965,
-      "learning_rate": 3.0608732590712574e-06,
-      "loss": 0.0933,
-      "step": 758
-    },
-    {
-      "epoch": 9.035714285714286,
-      "grad_norm": 1.446300469130805,
-      "learning_rate": 3.0377553086975397e-06,
-      "loss": 0.0549,
-      "step": 759
-    },
-    {
-      "epoch": 9.047619047619047,
-      "grad_norm": 2.2735202083993604,
-      "learning_rate": 3.014709346227421e-06,
-      "loss": 0.1031,
-      "step": 760
-    },
-    {
-      "epoch": 9.05952380952381,
-      "grad_norm": 4.144205517980198,
-      "learning_rate": 2.9917356099505137e-06,
-      "loss": 0.1125,
-      "step": 761
-    },
-    {
-      "epoch": 9.071428571428571,
-      "grad_norm": 1.7207853036637966,
-      "learning_rate": 2.968834337409612e-06,
-      "loss": 0.0453,
-      "step": 762
-    },
-    {
-      "epoch": 9.083333333333334,
-      "grad_norm": 1.3808210070725824,
-      "learning_rate": 2.9460057653982676e-06,
-      "loss": 0.0364,
-      "step": 763
-    },
-    {
-      "epoch": 9.095238095238095,
-      "grad_norm": 1.0699563203244846,
-      "learning_rate": 2.9232501299583126e-06,
-      "loss": 0.043,
-      "step": 764
-    },
-    {
-      "epoch": 9.107142857142858,
-      "grad_norm": 3.1313459792573566,
-      "learning_rate": 2.9005676663774464e-06,
-      "loss": 0.0667,
-      "step": 765
-    },
-    {
-      "epoch": 9.119047619047619,
-      "grad_norm": 2.8345479036451002,
-      "learning_rate": 2.8779586091867774e-06,
-      "loss": 0.072,
-      "step": 766
-    },
-    {
-      "epoch": 9.130952380952381,
-      "grad_norm": 4.237900498613799,
-      "learning_rate": 2.8554231921584164e-06,
-      "loss": 0.0266,
-      "step": 767
-    },
-    {
-      "epoch": 9.142857142857142,
-      "grad_norm": 4.626766434495465,
-      "learning_rate": 2.8329616483030574e-06,
-      "loss": 0.0622,
-      "step": 768
-    },
-    {
-      "epoch": 9.154761904761905,
-      "grad_norm": 5.703966216520634,
-      "learning_rate": 2.8105742098675504e-06,
-      "loss": 0.0481,
-      "step": 769
-    },
-    {
-      "epoch": 9.166666666666666,
-      "grad_norm": 2.19463955783124,
-      "learning_rate": 2.788261108332528e-06,
-      "loss": 0.0707,
-      "step": 770
-    },
-    {
-      "epoch": 9.178571428571429,
-      "grad_norm": 1.2240191096067952,
-      "learning_rate": 2.7660225744099858e-06,
-      "loss": 0.0353,
-      "step": 771
-    },
-    {
-      "epoch": 9.19047619047619,
-      "grad_norm": 2.332773587240624,
-      "learning_rate": 2.743858838040918e-06,
-      "loss": 0.0495,
-      "step": 772
-    },
-    {
-      "epoch": 9.202380952380953,
-      "grad_norm": 2.039653385293655,
-      "learning_rate": 2.7217701283929208e-06,
-      "loss": 0.0608,
-      "step": 773
-    },
-    {
-      "epoch": 9.214285714285714,
-      "grad_norm": 2.761650563323467,
-      "learning_rate": 2.699756673857845e-06,
-      "loss": 0.037,
-      "step": 774
-    },
-    {
-      "epoch": 9.226190476190476,
-      "grad_norm": 1.6011776954420698,
-      "learning_rate": 2.6778187020494086e-06,
-      "loss": 0.0123,
-      "step": 775
-    },
-    {
-      "epoch": 9.238095238095237,
-      "grad_norm": 5.53779887401531,
-      "learning_rate": 2.6559564398008643e-06,
-      "loss": 0.0378,
-      "step": 776
-    },
-    {
-      "epoch": 9.25,
-      "grad_norm": 2.1627086005140095,
-      "learning_rate": 2.6341701131626486e-06,
-      "loss": 0.0534,
-      "step": 777
-    },
-    {
-      "epoch": 9.261904761904763,
-      "grad_norm": 1.3638880451491888,
-      "learning_rate": 2.6124599474000347e-06,
-      "loss": 0.0417,
-      "step": 778
-    },
-    {
-      "epoch": 9.273809523809524,
-      "grad_norm": 5.103361689747087,
-      "learning_rate": 2.59082616699082e-06,
-      "loss": 0.1109,
-      "step": 779
-    },
-    {
-      "epoch": 9.285714285714286,
-      "grad_norm": 2.0038811026150793,
-      "learning_rate": 2.5692689956229842e-06,
-      "loss": 0.0863,
-      "step": 780
-    },
-    {
-      "epoch": 9.297619047619047,
-      "grad_norm": 1.4236628067476185,
-      "learning_rate": 2.5477886561924026e-06,
-      "loss": 0.0481,
-      "step": 781
-    },
-    {
-      "epoch": 9.30952380952381,
-      "grad_norm": 2.288250827736941,
-      "learning_rate": 2.526385370800515e-06,
-      "loss": 0.0696,
-      "step": 782
-    },
-    {
-      "epoch": 9.321428571428571,
-      "grad_norm": 2.241009646194953,
-      "learning_rate": 2.505059360752049e-06,
-      "loss": 0.0947,
-      "step": 783
-    },
-    {
-      "epoch": 9.333333333333334,
-      "grad_norm": 1.1079581601237796,
-      "learning_rate": 2.483810846552719e-06,
-      "loss": 0.009,
-      "step": 784
-    },
-    {
-      "epoch": 9.345238095238095,
-      "grad_norm": 1.9521426450571557,
-      "learning_rate": 2.462640047906958e-06,
-      "loss": 0.0685,
-      "step": 785
-    },
-    {
-      "epoch": 9.357142857142858,
-      "grad_norm": 1.7190040999607883,
-      "learning_rate": 2.441547183715628e-06,
-      "loss": 0.0669,
-      "step": 786
-    },
-    {
-      "epoch": 9.369047619047619,
-      "grad_norm": 2.822373050402177,
-      "learning_rate": 2.4205324720737787e-06,
-      "loss": 0.0947,
-      "step": 787
-    },
-    {
-      "epoch": 9.380952380952381,
-      "grad_norm": 5.160967804300185,
-      "learning_rate": 2.3995961302683803e-06,
-      "loss": 0.1173,
-      "step": 788
-    },
-    {
-      "epoch": 9.392857142857142,
-      "grad_norm": 1.4211193630802048,
-      "learning_rate": 2.3787383747760696e-06,
-      "loss": 0.0335,
-      "step": 789
-    },
-    {
-      "epoch": 9.404761904761905,
-      "grad_norm": 1.6880430372088504,
-      "learning_rate": 2.3579594212609325e-06,
-      "loss": 0.0638,
-      "step": 790
-    },
-    {
-      "epoch": 9.416666666666666,
-      "grad_norm": 2.5607635616852544,
-      "learning_rate": 2.3372594845722483e-06,
-      "loss": 0.0712,
-      "step": 791
-    },
-    {
-      "epoch": 9.428571428571429,
-      "grad_norm": 5.135385738571878,
-      "learning_rate": 2.316638778742295e-06,
-      "loss": 0.099,
-      "step": 792
-    },
-    {
-      "epoch": 9.44047619047619,
-      "grad_norm": 0.9753249765155438,
-      "learning_rate": 2.2960975169841106e-06,
-      "loss": 0.0386,
-      "step": 793
-    },
-    {
-      "epoch": 9.452380952380953,
-      "grad_norm": 3.059992847066805,
-      "learning_rate": 2.2756359116893122e-06,
-      "loss": 0.0766,
-      "step": 794
-    },
-    {
-      "epoch": 9.464285714285714,
-      "grad_norm": 1.852868554324358,
-      "learning_rate": 2.2552541744258816e-06,
-      "loss": 0.0771,
-      "step": 795
-    },
-    {
-      "epoch": 9.476190476190476,
-      "grad_norm": 3.678855356877469,
-      "learning_rate": 2.234952515935982e-06,
-      "loss": 0.0672,
-      "step": 796
-    },
-    {
-      "epoch": 9.488095238095237,
-      "grad_norm": 2.259811511231432,
-      "learning_rate": 2.214731146133793e-06,
-      "loss": 0.0717,
-      "step": 797
-    },
-    {
-      "epoch": 9.5,
-      "grad_norm": 2.922269519237516,
-      "learning_rate": 2.1945902741033154e-06,
-      "loss": 0.1276,
-      "step": 798
-    },
-    {
-      "epoch": 9.511904761904763,
-      "grad_norm": 2.0143579615385003,
-      "learning_rate": 2.1745301080962357e-06,
-      "loss": 0.0631,
-      "step": 799
-    },
-    {
-      "epoch": 9.523809523809524,
-      "grad_norm": 1.342784078461839,
-      "learning_rate": 2.1545508555297478e-06,
-      "loss": 0.0106,
-      "step": 800
-    },
-    {
-      "epoch": 9.535714285714286,
-      "grad_norm": 3.1514364794761267,
-      "learning_rate": 2.1346527229844305e-06,
-      "loss": 0.0592,
-      "step": 801
-    },
-    {
-      "epoch": 9.547619047619047,
-      "grad_norm": 4.658579628367995,
-      "learning_rate": 2.114835916202094e-06,
-      "loss": 0.0786,
-      "step": 802
-    },
-    {
-      "epoch": 9.55952380952381,
-      "grad_norm": 8.72775474775178,
-      "learning_rate": 2.095100640083664e-06,
-      "loss": 0.0577,
-      "step": 803
-    },
-    {
-      "epoch": 9.571428571428571,
-      "grad_norm": 3.078873859871242,
-      "learning_rate": 2.0754470986870602e-06,
-      "loss": 0.0186,
-      "step": 804
-    },
-    {
-      "epoch": 9.583333333333334,
-      "grad_norm": 3.0642918358992404,
-      "learning_rate": 2.05587549522508e-06,
-      "loss": 0.0207,
-      "step": 805
-    },
-    {
-      "epoch": 9.595238095238095,
-      "grad_norm": 1.7483461506332738,
-      "learning_rate": 2.036386032063311e-06,
-      "loss": 0.0442,
-      "step": 806
-    },
-    {
-      "epoch": 9.607142857142858,
-      "grad_norm": 3.801439720554629,
-      "learning_rate": 2.0169789107180195e-06,
-      "loss": 0.028,
-      "step": 807
-    },
-    {
-      "epoch": 9.619047619047619,
-      "grad_norm": 2.3775230600293473,
-      "learning_rate": 1.9976543318540887e-06,
-      "loss": 0.0629,
-      "step": 808
-    },
-    {
-      "epoch": 9.630952380952381,
-      "grad_norm": 1.3725986057321287,
-      "learning_rate": 1.978412495282922e-06,
-      "loss": 0.0341,
-      "step": 809
-    },
-    {
-      "epoch": 9.642857142857142,
-      "grad_norm": 3.093713374546874,
-      "learning_rate": 1.959253599960399e-06,
-      "loss": 0.0794,
-      "step": 810
-    },
-    {
-      "epoch": 9.654761904761905,
-      "grad_norm": 1.9678116047928416,
-      "learning_rate": 1.940177843984795e-06,
-      "loss": 0.0267,
-      "step": 811
-    },
-    {
-      "epoch": 9.666666666666666,
-      "grad_norm": 1.7929950739709355,
-      "learning_rate": 1.921185424594758e-06,
-      "loss": 0.0667,
-      "step": 812
-    },
-    {
-      "epoch": 9.678571428571429,
-      "grad_norm": 1.3460670621170243,
-      "learning_rate": 1.9022765381672426e-06,
-      "loss": 0.04,
-      "step": 813
-    },
-    {
-      "epoch": 9.69047619047619,
-      "grad_norm": 1.5856379303348087,
-      "learning_rate": 1.883451380215503e-06,
-      "loss": 0.0108,
-      "step": 814
-    },
-    {
-      "epoch": 9.702380952380953,
-      "grad_norm": 1.7602024319170253,
-      "learning_rate": 1.8647101453870608e-06,
-      "loss": 0.0434,
-      "step": 815
-    },
-    {
-      "epoch": 9.714285714285714,
-      "grad_norm": 2.3435995853073504,
-      "learning_rate": 1.846053027461686e-06,
-      "loss": 0.0863,
-      "step": 816
-    },
-    {
-      "epoch": 9.726190476190476,
-      "grad_norm": 0.920694195118384,
-      "learning_rate": 1.827480219349409e-06,
-      "loss": 0.0082,
-      "step": 817
-    },
-    {
-      "epoch": 9.738095238095237,
-      "grad_norm": 1.94811857435231,
-      "learning_rate": 1.8089919130885081e-06,
-      "loss": 0.0144,
-      "step": 818
-    },
-    {
-      "epoch": 9.75,
-      "grad_norm": 2.0473503573165504,
-      "learning_rate": 1.7905882998435443e-06,
-      "loss": 0.0423,
-      "step": 819
-    },
-    {
-      "epoch": 9.761904761904763,
-      "grad_norm": 2.6541768645283983,
-      "learning_rate": 1.7722695699033632e-06,
-      "loss": 0.0192,
-      "step": 820
-    },
-    {
-      "epoch": 9.773809523809524,
-      "grad_norm": 2.3050964074362534,
-      "learning_rate": 1.754035912679145e-06,
-      "loss": 0.0409,
-      "step": 821
-    },
-    {
-      "epoch": 9.785714285714286,
-      "grad_norm": 2.8122349110450062,
-      "learning_rate": 1.7358875167024336e-06,
-      "loss": 0.0655,
-      "step": 822
-    },
-    {
-      "epoch": 9.797619047619047,
-      "grad_norm": 3.6832643159394554,
-      "learning_rate": 1.7178245696231953e-06,
-      "loss": 0.1298,
-      "step": 823
-    },
-    {
-      "epoch": 9.80952380952381,
-      "grad_norm": 1.3306255786886887,
-      "learning_rate": 1.6998472582078806e-06,
-      "loss": 0.0506,
-      "step": 824
-    },
-    {
-      "epoch": 9.821428571428571,
-      "grad_norm": 3.315709760719103,
-      "learning_rate": 1.6819557683374766e-06,
-      "loss": 0.0232,
-      "step": 825
-    },
-    {
-      "epoch": 9.833333333333334,
-      "grad_norm": 2.6658995622840576,
-      "learning_rate": 1.6641502850056078e-06,
-      "loss": 0.0858,
-      "step": 826
-    },
-    {
-      "epoch": 9.845238095238095,
-      "grad_norm": 1.9573291379645015,
-      "learning_rate": 1.6464309923166033e-06,
-      "loss": 0.0649,
-      "step": 827
-    },
-    {
-      "epoch": 9.857142857142858,
-      "grad_norm": 2.9219885547313984,
-      "learning_rate": 1.6287980734836118e-06,
-      "loss": 0.0637,
-      "step": 828
-    },
-    {
-      "epoch": 9.869047619047619,
-      "grad_norm": 1.7737112336849967,
-      "learning_rate": 1.6112517108266867e-06,
-      "loss": 0.0597,
-      "step": 829
-    },
-    {
-      "epoch": 9.880952380952381,
-      "grad_norm": 4.779901240190352,
-      "learning_rate": 1.593792085770921e-06,
-      "loss": 0.137,
-      "step": 830
-    },
-    {
-      "epoch": 9.892857142857142,
-      "grad_norm": 3.801958131422506,
-      "learning_rate": 1.5764193788445548e-06,
-      "loss": 0.1197,
-      "step": 831
-    },
-    {
-      "epoch": 9.904761904761905,
-      "grad_norm": 2.2120118800441984,
-      "learning_rate": 1.5591337696771247e-06,
-      "loss": 0.0728,
-      "step": 832
-    },
-    {
-      "epoch": 9.916666666666666,
-      "grad_norm": 1.4958592670875228,
-      "learning_rate": 1.541935436997588e-06,
-      "loss": 0.0481,
-      "step": 833
-    },
-    {
-      "epoch": 9.928571428571429,
-      "grad_norm": 2.8834804691841964,
-      "learning_rate": 1.5248245586324883e-06,
-      "loss": 0.0175,
-      "step": 834
-    },
-    {
-      "epoch": 9.94047619047619,
-      "grad_norm": 2.0804272562966135,
-      "learning_rate": 1.507801311504119e-06,
-      "loss": 0.0772,
-      "step": 835
-    },
-    {
-      "epoch": 9.952380952380953,
-      "grad_norm": 2.2745884531321514,
-      "learning_rate": 1.4908658716286784e-06,
-      "loss": 0.0617,
-      "step": 836
-    },
-    {
-      "epoch": 9.964285714285714,
-      "grad_norm": 1.526962860362152,
-      "learning_rate": 1.4740184141144664e-06,
-      "loss": 0.0281,
-      "step": 837
-    },
-    {
-      "epoch": 9.976190476190476,
-      "grad_norm": 4.076270588481699,
-      "learning_rate": 1.457259113160061e-06,
-      "loss": 0.0757,
-      "step": 838
-    },
-    {
-      "epoch": 9.988095238095237,
-      "grad_norm": 2.054731057370207,
-      "learning_rate": 1.4405881420525315e-06,
-      "loss": 0.0746,
-      "step": 839
-    },
-    {
-      "epoch": 10.0,
-      "grad_norm": 2.4967134409140845,
-      "learning_rate": 1.4240056731656271e-06,
-      "loss": 0.0998,
-      "step": 840
-    },
-    {
-      "epoch": 10.0,
-      "eval_loss": 0.29546400904655457,
-      "eval_runtime": 38.3401,
-      "eval_samples_per_second": 1.565,
-      "eval_steps_per_second": 1.565,
-      "step": 840
-    },
-    {
-      "epoch": 10.011904761904763,
-      "grad_norm": 1.939367732693264,
-      "learning_rate": 1.4075118779580155e-06,
-      "loss": 0.0772,
-      "step": 841
-    },
-    {
-      "epoch": 10.023809523809524,
-      "grad_norm": 5.127527273619253,
-      "learning_rate": 1.391106926971496e-06,
-      "loss": 0.1178,
-      "step": 842
-    },
-    {
-      "epoch": 10.035714285714286,
-      "grad_norm": 4.574690636051704,
-      "learning_rate": 1.3747909898292389e-06,
-      "loss": 0.03,
-      "step": 843
-    },
-    {
-      "epoch": 10.047619047619047,
-      "grad_norm": 2.7132236248947894,
-      "learning_rate": 1.3585642352340377e-06,
-      "loss": 0.0788,
-      "step": 844
-    },
-    {
-      "epoch": 10.05952380952381,
-      "grad_norm": 1.9056680106061799,
-      "learning_rate": 1.3424268309665546e-06,
-      "loss": 0.073,
-      "step": 845
-    },
-    {
-      "epoch": 10.071428571428571,
-      "grad_norm": 1.2385633126556088,
-      "learning_rate": 1.3263789438835973e-06,
-      "loss": 0.0297,
-      "step": 846
-    },
-    {
-      "epoch": 10.083333333333334,
-      "grad_norm": 2.944507356001365,
-      "learning_rate": 1.3104207399163816e-06,
-      "loss": 0.0199,
-      "step": 847
-    },
-    {
-      "epoch": 10.095238095238095,
-      "grad_norm": 1.7161167095821772,
-      "learning_rate": 1.294552384068829e-06,
-      "loss": 0.0627,
-      "step": 848
-    },
-    {
-      "epoch": 10.107142857142858,
-      "grad_norm": 4.6380379114120265,
-      "learning_rate": 1.2787740404158455e-06,
-      "loss": 0.0709,
-      "step": 849
-    },
-    {
-      "epoch": 10.119047619047619,
-      "grad_norm": 1.634226568884785,
-      "learning_rate": 1.263085872101638e-06,
-      "loss": 0.0333,
-      "step": 850
-    },
-    {
-      "epoch": 10.130952380952381,
-      "grad_norm": 1.4985265598741437,
-      "learning_rate": 1.2474880413380253e-06,
-      "loss": 0.0443,
-      "step": 851
-    },
-    {
-      "epoch": 10.142857142857142,
-      "grad_norm": 1.3371053714757155,
-      "learning_rate": 1.2319807094027492e-06,
-      "loss": 0.0362,
-      "step": 852
-    },
-    {
-      "epoch": 10.154761904761905,
-      "grad_norm": 2.994449032456818,
-      "learning_rate": 1.216564036637825e-06,
-      "loss": 0.0406,
-      "step": 853
-    },
-    {
-      "epoch": 10.166666666666666,
-      "grad_norm": 2.2449860040319245,
-      "learning_rate": 1.201238182447867e-06,
-      "loss": 0.0379,
-      "step": 854
-    },
-    {
-      "epoch": 10.178571428571429,
-      "grad_norm": 1.7187436793514488,
-      "learning_rate": 1.1860033052984544e-06,
-      "loss": 0.0571,
-      "step": 855
-    },
-    {
-      "epoch": 10.19047619047619,
-      "grad_norm": 4.381492396088356,
-      "learning_rate": 1.1708595627144782e-06,
-      "loss": 0.0284,
-      "step": 856
-    },
-    {
-      "epoch": 10.202380952380953,
-      "grad_norm": 3.8586763241442013,
-      "learning_rate": 1.1558071112785297e-06,
-      "loss": 0.0771,
-      "step": 857
-    },
-    {
-      "epoch": 10.214285714285714,
-      "grad_norm": 2.337542758928375,
-      "learning_rate": 1.1408461066292643e-06,
-      "loss": 0.0666,
-      "step": 858
-    },
-    {
-      "epoch": 10.226190476190476,
-      "grad_norm": 1.929433338882153,
-      "learning_rate": 1.1259767034598046e-06,
-      "loss": 0.0441,
-      "step": 859
-    },
-    {
-      "epoch": 10.238095238095237,
-      "grad_norm": 4.750689941855376,
-      "learning_rate": 1.1111990555161322e-06,
-      "loss": 0.1109,
-      "step": 860
-    },
-    {
-      "epoch": 10.25,
-      "grad_norm": 2.8991103758900914,
-      "learning_rate": 1.0965133155955066e-06,
-      "loss": 0.0216,
-      "step": 861
-    },
-    {
-      "epoch": 10.261904761904763,
-      "grad_norm": 1.7271815063958462,
-      "learning_rate": 1.0819196355448801e-06,
-      "loss": 0.0667,
-      "step": 862
-    },
-    {
-      "epoch": 10.273809523809524,
-      "grad_norm": 3.3616262268804293,
-      "learning_rate": 1.0674181662593253e-06,
-      "loss": 0.0788,
-      "step": 863
-    },
-    {
-      "epoch": 10.285714285714286,
-      "grad_norm": 3.5482395867466283,
-      "learning_rate": 1.053009057680483e-06,
-      "loss": 0.0221,
-      "step": 864
-    },
-    {
-      "epoch": 10.297619047619047,
-      "grad_norm": 2.28948006890838,
-      "learning_rate": 1.0386924587949998e-06,
-      "loss": 0.0662,
-      "step": 865
-    },
-    {
-      "epoch": 10.30952380952381,
-      "grad_norm": 2.8494445710276914,
-      "learning_rate": 1.0244685176330027e-06,
-      "loss": 0.0205,
-      "step": 866
-    },
-    {
-      "epoch": 10.321428571428571,
-      "grad_norm": 1.577698024983074,
-      "learning_rate": 1.0103373812665552e-06,
-      "loss": 0.0688,
-      "step": 867
-    },
-    {
-      "epoch": 10.333333333333334,
-      "grad_norm": 2.1289638850579284,
-      "learning_rate": 9.962991958081446e-07,
-      "loss": 0.0158,
-      "step": 868
-    },
-    {
-      "epoch": 10.345238095238095,
-      "grad_norm": 3.460213528935419,
-      "learning_rate": 9.823541064091657e-07,
-      "loss": 0.1106,
-      "step": 869
-    },
-    {
-      "epoch": 10.357142857142858,
-      "grad_norm": 3.481972900368815,
-      "learning_rate": 9.68502257258428e-07,
-      "loss": 0.091,
-      "step": 870
-    },
-    {
-      "epoch": 10.369047619047619,
-      "grad_norm": 2.0142509797571337,
-      "learning_rate": 9.547437915806534e-07,
-      "loss": 0.0608,
-      "step": 871
-    },
-    {
-      "epoch": 10.380952380952381,
-      "grad_norm": 2.2583132229155405,
-      "learning_rate": 9.410788516350078e-07,
-      "loss": 0.0839,
-      "step": 872
-    },
-    {
-      "epoch": 10.392857142857142,
-      "grad_norm": 2.570112503291791,
-      "learning_rate": 9.27507578713619e-07,
-      "loss": 0.0535,
-      "step": 873
-    },
-    {
-      "epoch": 10.404761904761905,
-      "grad_norm": 2.170958789831051,
-      "learning_rate": 9.140301131401208e-07,
-      "loss": 0.0804,
-      "step": 874
-    },
-    {
-      "epoch": 10.416666666666666,
-      "grad_norm": 2.8444438323548136,
-      "learning_rate": 9.006465942682074e-07,
-      "loss": 0.0207,
-      "step": 875
-    },
-    {
-      "epoch": 10.428571428571429,
-      "grad_norm": 4.208193249007848,
-      "learning_rate": 8.873571604801812e-07,
-      "loss": 0.1133,
-      "step": 876
-    },
-    {
-      "epoch": 10.44047619047619,
-      "grad_norm": 3.80053895933604,
-      "learning_rate": 8.741619491855291e-07,
-      "loss": 0.1123,
-      "step": 877
-    },
-    {
-      "epoch": 10.452380952380953,
-      "grad_norm": 3.8249855056475948,
-      "learning_rate": 8.610610968195065e-07,
-      "loss": 0.0262,
-      "step": 878
-    },
-    {
-      "epoch": 10.464285714285714,
-      "grad_norm": 2.1591266760594583,
-      "learning_rate": 8.480547388417104e-07,
-      "loss": 0.0689,
-      "step": 879
-    },
-    {
-      "epoch": 10.476190476190476,
-      "grad_norm": 2.2449611487373615,
-      "learning_rate": 8.35143009734698e-07,
-      "loss": 0.0536,
-      "step": 880
-    },
-    {
-      "epoch": 10.488095238095237,
-      "grad_norm": 1.945141389129938,
-      "learning_rate": 8.223260430025804e-07,
-      "loss": 0.0489,
-      "step": 881
-    },
-    {
-      "epoch": 10.5,
-      "grad_norm": 2.859367384367929,
-      "learning_rate": 8.096039711696546e-07,
-      "loss": 0.0779,
-      "step": 882
-    },
-    {
-      "epoch": 10.511904761904763,
-      "grad_norm": 2.257419186913638,
-      "learning_rate": 7.969769257790238e-07,
-      "loss": 0.0159,
-      "step": 883
-    },
-    {
-      "epoch": 10.523809523809524,
-      "grad_norm": 1.9636635789211396,
-      "learning_rate": 7.844450373912437e-07,
-      "loss": 0.0616,
-      "step": 884
-    },
-    {
-      "epoch": 10.535714285714286,
-      "grad_norm": 2.4048703336882733,
-      "learning_rate": 7.720084355829661e-07,
-      "loss": 0.0399,
-      "step": 885
-    },
-    {
-      "epoch": 10.547619047619047,
-      "grad_norm": 2.516393022925152,
-      "learning_rate": 7.596672489456092e-07,
-      "loss": 0.0173,
-      "step": 886
-    },
-    {
-      "epoch": 10.55952380952381,
-      "grad_norm": 3.5055609696337373,
-      "learning_rate": 7.474216050840122e-07,
-      "loss": 0.0772,
-      "step": 887
-    },
-    {
-      "epoch": 10.571428571428571,
-      "grad_norm": 1.1195443659303188,
-      "learning_rate": 7.352716306151353e-07,
-      "loss": 0.0465,
-      "step": 888
-    },
-    {
-      "epoch": 10.583333333333334,
-      "grad_norm": 1.433395584509959,
-      "learning_rate": 7.232174511667378e-07,
-      "loss": 0.0316,
-      "step": 889
-    },
-    {
-      "epoch": 10.595238095238095,
-      "grad_norm": 1.4384479578684133,
-      "learning_rate": 7.112591913760769e-07,
-      "loss": 0.0377,
-      "step": 890
-    },
-    {
-      "epoch": 10.607142857142858,
-      "grad_norm": 3.633376265932014,
-      "learning_rate": 6.993969748886297e-07,
-      "loss": 0.0432,
-      "step": 891
-    },
-    {
-      "epoch": 10.619047619047619,
-      "grad_norm": 1.4047327300385846,
-      "learning_rate": 6.876309243568058e-07,
-      "loss": 0.0097,
-      "step": 892
-    },
-    {
-      "epoch": 10.630952380952381,
-      "grad_norm": 1.9199258882874348,
-      "learning_rate": 6.759611614386863e-07,
-      "loss": 0.0569,
-      "step": 893
-    },
-    {
-      "epoch": 10.642857142857142,
-      "grad_norm": 4.306204615416398,
-      "learning_rate": 6.643878067967568e-07,
-      "loss": 0.1107,
-      "step": 894
-    },
-    {
-      "epoch": 10.654761904761905,
-      "grad_norm": 2.297829748178259,
-      "learning_rate": 6.529109800966693e-07,
-      "loss": 0.0561,
-      "step": 895
-    },
-    {
-      "epoch": 10.666666666666666,
-      "grad_norm": 3.275946348662118,
-      "learning_rate": 6.415308000059961e-07,
-      "loss": 0.0245,
-      "step": 896
-    },
-    {
-      "epoch": 10.678571428571429,
-      "grad_norm": 2.2844223874918783,
-      "learning_rate": 6.302473841930135e-07,
-      "loss": 0.0521,
-      "step": 897
-    },
-    {
-      "epoch": 10.69047619047619,
-      "grad_norm": 4.300283035486153,
-      "learning_rate": 6.190608493254746e-07,
-      "loss": 0.0922,
-      "step": 898
-    },
-    {
-      "epoch": 10.702380952380953,
-      "grad_norm": 2.828737463417787,
-      "learning_rate": 6.079713110694052e-07,
-      "loss": 0.044,
-      "step": 899
-    },
-    {
-      "epoch": 10.714285714285714,
-      "grad_norm": 1.5456393145327858,
-      "learning_rate": 5.969788840879165e-07,
-      "loss": 0.0536,
-      "step": 900
-    },
-    {
-      "epoch": 10.726190476190476,
-      "grad_norm": 1.5850187544297383,
-      "learning_rate": 5.860836820400062e-07,
-      "loss": 0.0404,
-      "step": 901
-    },
-    {
-      "epoch": 10.738095238095237,
-      "grad_norm": 3.148235644738001,
-      "learning_rate": 5.752858175793951e-07,
-      "loss": 0.1086,
-      "step": 902
-    },
-    {
-      "epoch": 10.75,
-      "grad_norm": 2.393998297233833,
-      "learning_rate": 5.645854023533537e-07,
-      "loss": 0.0632,
-      "step": 903
-    },
-    {
-      "epoch": 10.761904761904763,
-      "grad_norm": 2.407523392793869,
-      "learning_rate": 5.539825470015536e-07,
-      "loss": 0.0792,
-      "step": 904
-    },
-    {
-      "epoch": 10.773809523809524,
-      "grad_norm": 2.0772622564808945,
-      "learning_rate": 5.434773611549182e-07,
-      "loss": 0.0641,
-      "step": 905
-    },
-    {
-      "epoch": 10.785714285714286,
-      "grad_norm": 2.8317777886655,
-      "learning_rate": 5.330699534344986e-07,
-      "loss": 0.0725,
-      "step": 906
-    },
-    {
-      "epoch": 10.797619047619047,
-      "grad_norm": 3.3935608020239942,
-      "learning_rate": 5.227604314503343e-07,
-      "loss": 0.0232,
-      "step": 907
-    },
-    {
-      "epoch": 10.80952380952381,
-      "grad_norm": 3.689532389180214,
-      "learning_rate": 5.125489018003583e-07,
-      "loss": 0.0225,
-      "step": 908
-    },
-    {
-      "epoch": 10.821428571428571,
-      "grad_norm": 5.726396718481587,
-      "learning_rate": 5.024354700692868e-07,
-      "loss": 0.1026,
-      "step": 909
-    },
-    {
-      "epoch": 10.833333333333334,
-      "grad_norm": 2.7925292834564104,
-      "learning_rate": 4.924202408275203e-07,
-      "loss": 0.0819,
-      "step": 910
-    },
-    {
-      "epoch": 10.845238095238095,
-      "grad_norm": 1.6524737914806937,
-      "learning_rate": 4.825033176300786e-07,
-      "loss": 0.0103,
-      "step": 911
-    },
-    {
-      "epoch": 10.857142857142858,
-      "grad_norm": 2.135124230264784,
-      "learning_rate": 4.726848030155129e-07,
-      "loss": 0.0459,
-      "step": 912
-    },
-    {
-      "epoch": 10.869047619047619,
-      "grad_norm": 2.179953610716799,
-      "learning_rate": 4.6296479850486133e-07,
-      "loss": 0.0492,
-      "step": 913
-    },
-    {
-      "epoch": 10.880952380952381,
-      "grad_norm": 2.0409112787447827,
-      "learning_rate": 4.5334340460058354e-07,
-      "loss": 0.0595,
-      "step": 914
-    },
-    {
-      "epoch": 10.892857142857142,
-      "grad_norm": 1.592173470715575,
-      "learning_rate": 4.438207207855383e-07,
-      "loss": 0.0293,
-      "step": 915
-    },
-    {
-      "epoch": 10.904761904761905,
-      "grad_norm": 2.4703577660931177,
-      "learning_rate": 4.343968455219416e-07,
-      "loss": 0.0176,
-      "step": 916
-    },
-    {
-      "epoch": 10.916666666666666,
-      "grad_norm": 3.3207335636677318,
-      "learning_rate": 4.250718762503514e-07,
-      "loss": 0.04,
-      "step": 917
-    },
-    {
-      "epoch": 10.928571428571429,
-      "grad_norm": 1.649892185275184,
-      "learning_rate": 4.15845909388668e-07,
-      "loss": 0.0519,
-      "step": 918
-    },
-    {
-      "epoch": 10.94047619047619,
-      "grad_norm": 2.34723989626408,
-      "learning_rate": 4.0671904033112386e-07,
-      "loss": 0.0616,
-      "step": 919
-    },
-    {
-      "epoch": 10.952380952380953,
-      "grad_norm": 1.5884046811323744,
-      "learning_rate": 3.976913634473112e-07,
-      "loss": 0.01,
-      "step": 920
-    },
-    {
-      "epoch": 10.964285714285714,
-      "grad_norm": 3.0053504070381427,
-      "learning_rate": 3.8876297208119253e-07,
-      "loss": 0.0613,
-      "step": 921
-    },
-    {
-      "epoch": 10.976190476190476,
-      "grad_norm": 2.083830931442132,
-      "learning_rate": 3.799339585501494e-07,
-      "loss": 0.0547,
-      "step": 922
-    },
-    {
-      "epoch": 10.988095238095237,
-      "grad_norm": 4.938614455336452,
-      "learning_rate": 3.712044141440152e-07,
-      "loss": 0.1153,
-      "step": 923
-    },
-    {
-      "epoch": 11.0,
-      "grad_norm": 1.6270011686782586,
-      "learning_rate": 3.6257442912414175e-07,
-      "loss": 0.0587,
-      "step": 924
-    },
-    {
-      "epoch": 11.0,
-      "eval_loss": 0.29636117815971375,
-      "eval_runtime": 38.491,
-      "eval_samples_per_second": 1.559,
-      "eval_steps_per_second": 1.559,
-      "step": 924
-    },
-    {
-      "epoch": 11.011904761904763,
-      "grad_norm": 2.986931803591163,
-      "learning_rate": 3.5404409272245757e-07,
-      "loss": 0.092,
-      "step": 925
-    },
-    {
-      "epoch": 11.023809523809524,
-      "grad_norm": 1.3347524168100455,
-      "learning_rate": 3.456134931405497e-07,
-      "loss": 0.0418,
-      "step": 926
-    },
-    {
-      "epoch": 11.035714285714286,
-      "grad_norm": 5.5212254683737925,
-      "learning_rate": 3.3728271754875365e-07,
-      "loss": 0.0339,
-      "step": 927
-    },
-    {
-      "epoch": 11.047619047619047,
-      "grad_norm": 2.7505612562930226,
-      "learning_rate": 3.290518520852437e-07,
-      "loss": 0.0583,
-      "step": 928
-    },
-    {
-      "epoch": 11.05952380952381,
-      "grad_norm": 2.456582011903621,
-      "learning_rate": 3.20920981855154e-07,
-      "loss": 0.0799,
-      "step": 929
-    },
-    {
-      "epoch": 11.071428571428571,
-      "grad_norm": 1.420816754610895,
-      "learning_rate": 3.1289019092968774e-07,
-      "loss": 0.0393,
-      "step": 930
-    },
-    {
-      "epoch": 11.083333333333334,
-      "grad_norm": 1.579732141095948,
-      "learning_rate": 3.0495956234525836e-07,
-      "loss": 0.0543,
-      "step": 931
-    },
-    {
-      "epoch": 11.095238095238095,
-      "grad_norm": 1.3126593889196536,
-      "learning_rate": 2.9712917810261645e-07,
-      "loss": 0.0241,
-      "step": 932
-    },
-    {
-      "epoch": 11.107142857142858,
-      "grad_norm": 2.949990528078044,
-      "learning_rate": 2.893991191660195e-07,
-      "loss": 0.0984,
-      "step": 933
-    },
-    {
-      "epoch": 11.119047619047619,
-      "grad_norm": 2.120636416281821,
-      "learning_rate": 2.817694654623804e-07,
-      "loss": 0.0576,
-      "step": 934
-    },
-    {
-      "epoch": 11.130952380952381,
-      "grad_norm": 1.9186338861805565,
-      "learning_rate": 2.742402958804491e-07,
-      "loss": 0.0621,
-      "step": 935
-    },
-    {
-      "epoch": 11.142857142857142,
-      "grad_norm": 1.0631990758619283,
-      "learning_rate": 2.668116882699956e-07,
-      "loss": 0.0287,
-      "step": 936
-    },
-    {
-      "epoch": 11.154761904761905,
-      "grad_norm": 1.4820135737077909,
-      "learning_rate": 2.5948371944099715e-07,
-      "loss": 0.0307,
-      "step": 937
-    },
-    {
-      "epoch": 11.166666666666666,
-      "grad_norm": 4.437779501180707,
-      "learning_rate": 2.5225646516285897e-07,
-      "loss": 0.0996,
-      "step": 938
-    },
-    {
-      "epoch": 11.178571428571429,
-      "grad_norm": 1.8449980165305102,
-      "learning_rate": 2.45130000163617e-07,
-      "loss": 0.048,
-      "step": 939
-    },
-    {
-      "epoch": 11.19047619047619,
-      "grad_norm": 5.919702050199371,
-      "learning_rate": 2.381043981291753e-07,
-      "loss": 0.0688,
-      "step": 940
-    },
-    {
-      "epoch": 11.202380952380953,
-      "grad_norm": 2.825856353607067,
-      "learning_rate": 2.3117973170253216e-07,
-      "loss": 0.0402,
-      "step": 941
-    },
-    {
-      "epoch": 11.214285714285714,
-      "grad_norm": 2.7231510331845943,
-      "learning_rate": 2.2435607248304624e-07,
-      "loss": 0.0183,
-      "step": 942
-    },
-    {
-      "epoch": 11.226190476190476,
-      "grad_norm": 2.4767739656112497,
-      "learning_rate": 2.176334910256772e-07,
-      "loss": 0.0762,
-      "step": 943
-    },
-    {
-      "epoch": 11.238095238095237,
-      "grad_norm": 3.8407293601387416,
-      "learning_rate": 2.1101205684027294e-07,
-      "loss": 0.0448,
-      "step": 944
-    },
-    {
-      "epoch": 11.25,
-      "grad_norm": 1.3667000249854184,
-      "learning_rate": 2.0449183839083675e-07,
-      "loss": 0.0101,
-      "step": 945
-    },
-    {
-      "epoch": 11.261904761904763,
-      "grad_norm": 2.8261958005353924,
-      "learning_rate": 1.980729030948314e-07,
-      "loss": 0.0173,
-      "step": 946
-    },
-    {
-      "epoch": 11.273809523809524,
-      "grad_norm": 2.407670745363879,
-      "learning_rate": 1.91755317322474e-07,
-      "loss": 0.0418,
-      "step": 947
-    },
-    {
-      "epoch": 11.285714285714286,
-      "grad_norm": 2.288457051038679,
-      "learning_rate": 1.8553914639605207e-07,
-      "loss": 0.0691,
-      "step": 948
-    },
-    {
-      "epoch": 11.297619047619047,
-      "grad_norm": 1.3870092710553388,
-      "learning_rate": 1.7942445458925206e-07,
-      "loss": 0.0093,
-      "step": 949
-    },
-    {
-      "epoch": 11.30952380952381,
-      "grad_norm": 1.6211467655272436,
-      "learning_rate": 1.734113051264852e-07,
-      "loss": 0.0493,
-      "step": 950
-    },
-    {
-      "epoch": 11.321428571428571,
-      "grad_norm": 2.7633334414007518,
-      "learning_rate": 1.674997601822448e-07,
-      "loss": 0.0237,
-      "step": 951
-    },
-    {
-      "epoch": 11.333333333333334,
-      "grad_norm": 2.132739304201599,
-      "learning_rate": 1.616898808804568e-07,
-      "loss": 0.0702,
-      "step": 952
-    },
-    {
-      "epoch": 11.345238095238095,
-      "grad_norm": 5.177102761658638,
-      "learning_rate": 1.5598172729384686e-07,
-      "loss": 0.1246,
-      "step": 953
-    },
-    {
-      "epoch": 11.357142857142858,
-      "grad_norm": 2.469101298855786,
-      "learning_rate": 1.5037535844332762e-07,
-      "loss": 0.0643,
-      "step": 954
-    },
-    {
-      "epoch": 11.369047619047619,
-      "grad_norm": 1.0161239922881402,
-      "learning_rate": 1.4487083229737574e-07,
-      "loss": 0.0515,
-      "step": 955
-    },
-    {
-      "epoch": 11.380952380952381,
-      "grad_norm": 2.6222860821706564,
-      "learning_rate": 1.3946820577144359e-07,
-      "loss": 0.0812,
-      "step": 956
-    },
-    {
-      "epoch": 11.392857142857142,
-      "grad_norm": 1.0962889011973982,
-      "learning_rate": 1.341675347273652e-07,
-      "loss": 0.008,
-      "step": 957
-    },
-    {
-      "epoch": 11.404761904761905,
-      "grad_norm": 1.886834476713489,
-      "learning_rate": 1.2896887397278125e-07,
-      "loss": 0.047,
-      "step": 958
-    },
-    {
-      "epoch": 11.416666666666666,
-      "grad_norm": 2.739136030244547,
-      "learning_rate": 1.2387227726056606e-07,
-      "loss": 0.0757,
-      "step": 959
-    },
-    {
-      "epoch": 11.428571428571429,
-      "grad_norm": 5.288321746083624,
-      "learning_rate": 1.1887779728828486e-07,
-      "loss": 0.1198,
-      "step": 960
-    },
-    {
-      "epoch": 11.44047619047619,
-      "grad_norm": 2.104498900191149,
-      "learning_rate": 1.1398548569763191e-07,
-      "loss": 0.0718,
-      "step": 961
-    },
-    {
-      "epoch": 11.452380952380953,
-      "grad_norm": 2.306227037555575,
-      "learning_rate": 1.0919539307391314e-07,
-      "loss": 0.0477,
-      "step": 962
-    },
-    {
-      "epoch": 11.464285714285714,
-      "grad_norm": 3.1821099458410713,
-      "learning_rate": 1.0450756894551217e-07,
-      "loss": 0.0431,
-      "step": 963
-    },
-    {
-      "epoch": 11.476190476190476,
-      "grad_norm": 1.2721737087858365,
-      "learning_rate": 9.99220617833796e-08,
-      "loss": 0.0292,
-      "step": 964
-    },
-    {
-      "epoch": 11.488095238095237,
-      "grad_norm": 3.2490088015741097,
-      "learning_rate": 9.54389190005367e-08,
-      "loss": 0.0226,
-      "step": 965
-    },
-    {
-      "epoch": 11.5,
-      "grad_norm": 7.477486139245656,
-      "learning_rate": 9.105818695158031e-08,
-      "loss": 0.0483,
-      "step": 966
-    },
-    {
-      "epoch": 11.511904761904763,
-      "grad_norm": 4.16925857937326,
-      "learning_rate": 8.677991093220761e-08,
-      "loss": 0.1144,
-      "step": 967
-    },
-    {
-      "epoch": 11.523809523809524,
-      "grad_norm": 5.42535254900298,
-      "learning_rate": 8.26041351787421e-08,
-      "loss": 0.1407,
-      "step": 968
-    },
-    {
-      "epoch": 11.535714285714286,
-      "grad_norm": 2.0057715771874207,
-      "learning_rate": 7.853090286768284e-08,
-      "loss": 0.0648,
-      "step": 969
-    },
-    {
-      "epoch": 11.547619047619047,
-      "grad_norm": 1.7657521223378563,
-      "learning_rate": 7.456025611525253e-08,
-      "loss": 0.0442,
-      "step": 970
-    },
-    {
-      "epoch": 11.55952380952381,
-      "grad_norm": 1.568683495757951,
-      "learning_rate": 7.069223597696572e-08,
-      "loss": 0.0473,
-      "step": 971
-    },
-    {
-      "epoch": 11.571428571428571,
-      "grad_norm": 4.441369110805783,
-      "learning_rate": 6.692688244720357e-08,
-      "loss": 0.0279,
-      "step": 972
-    },
-    {
-      "epoch": 11.583333333333334,
-      "grad_norm": 2.9301820879708758,
-      "learning_rate": 6.326423445879748e-08,
-      "loss": 0.04,
-      "step": 973
-    },
-    {
-      "epoch": 11.595238095238095,
-      "grad_norm": 2.9514527443649117,
-      "learning_rate": 5.970432988263053e-08,
-      "loss": 0.0393,
-      "step": 974
-    },
-    {
-      "epoch": 11.607142857142858,
-      "grad_norm": 2.6193009344389733,
-      "learning_rate": 5.624720552724228e-08,
-      "loss": 0.0756,
-      "step": 975
-    },
-    {
-      "epoch": 11.619047619047619,
-      "grad_norm": 1.858288104051874,
-      "learning_rate": 5.289289713845347e-08,
-      "loss": 0.0866,
-      "step": 976
-    },
-    {
-      "epoch": 11.630952380952381,
-      "grad_norm": 1.3996258099249799,
-      "learning_rate": 4.964143939898747e-08,
-      "loss": 0.0504,
-      "step": 977
-    },
-    {
-      "epoch": 11.642857142857142,
-      "grad_norm": 1.4934729697666005,
-      "learning_rate": 4.6492865928123856e-08,
-      "loss": 0.0494,
-      "step": 978
-    },
-    {
-      "epoch": 11.654761904761905,
-      "grad_norm": 2.019745524010841,
-      "learning_rate": 4.344720928133761e-08,
-      "loss": 0.0724,
-      "step": 979
-    },
-    {
-      "epoch": 11.666666666666666,
-      "grad_norm": 0.8817349997020998,
-      "learning_rate": 4.0504500949974936e-08,
-      "loss": 0.0249,
-      "step": 980
-    },
-    {
-      "epoch": 11.678571428571429,
-      "grad_norm": 3.135857451457003,
-      "learning_rate": 3.766477136091573e-08,
-      "loss": 0.0205,
-      "step": 981
-    },
-    {
-      "epoch": 11.69047619047619,
-      "grad_norm": 4.821187540685422,
-      "learning_rate": 3.49280498762683e-08,
-      "loss": 0.0948,
-      "step": 982
-    },
-    {
-      "epoch": 11.702380952380953,
-      "grad_norm": 2.155760067566603,
-      "learning_rate": 3.2294364793064026e-08,
-      "loss": 0.0773,
-      "step": 983
-    },
-    {
-      "epoch": 11.714285714285714,
-      "grad_norm": 1.2790261765367164,
-      "learning_rate": 2.976374334295984e-08,
-      "loss": 0.0394,
-      "step": 984
-    },
-    {
-      "epoch": 11.726190476190476,
-      "grad_norm": 1.9945206184899933,
-      "learning_rate": 2.7336211691961767e-08,
-      "loss": 0.0607,
-      "step": 985
-    },
-    {
-      "epoch": 11.738095238095237,
-      "grad_norm": 2.2954595943776375,
-      "learning_rate": 2.5011794940150713e-08,
-      "loss": 0.0685,
-      "step": 986
-    },
-    {
-      "epoch": 11.75,
-      "grad_norm": 3.889746627099117,
-      "learning_rate": 2.2790517121428212e-08,
-      "loss": 0.092,
-      "step": 987
-    },
-    {
-      "epoch": 11.761904761904763,
-      "grad_norm": 1.9305228670767907,
-      "learning_rate": 2.0672401203261084e-08,
-      "loss": 0.0608,
-      "step": 988
-    },
-    {
-      "epoch": 11.773809523809524,
-      "grad_norm": 1.49732917316992,
-      "learning_rate": 1.8657469086448275e-08,
-      "loss": 0.0587,
-      "step": 989
-    },
-    {
-      "epoch": 11.785714285714286,
-      "grad_norm": 3.209573738861191,
-      "learning_rate": 1.6745741604895503e-08,
-      "loss": 0.0634,
-      "step": 990
-    },
-    {
-      "epoch": 11.797619047619047,
-      "grad_norm": 0.9302644870602609,
-      "learning_rate": 1.4937238525395416e-08,
-      "loss": 0.021,
-      "step": 991
-    },
-    {
-      "epoch": 11.80952380952381,
-      "grad_norm": 1.7458424653110582,
-      "learning_rate": 1.3231978547427749e-08,
-      "loss": 0.05,
-      "step": 992
-    },
-    {
-      "epoch": 11.821428571428571,
-      "grad_norm": 2.4382393205857316,
-      "learning_rate": 1.1629979302960614e-08,
-      "loss": 0.0457,
-      "step": 993
-    },
-    {
-      "epoch": 11.833333333333334,
-      "grad_norm": 0.17241553083238312,
-      "learning_rate": 1.013125735627507e-08,
-      "loss": 0.0031,
-      "step": 994
-    },
-    {
-      "epoch": 11.845238095238095,
-      "grad_norm": 2.1464199664004395,
-      "learning_rate": 8.735828203787489e-09,
-      "loss": 0.0836,
-      "step": 995
-    },
-    {
-      "epoch": 11.857142857142858,
-      "grad_norm": 2.8983917842024045,
-      "learning_rate": 7.443706273895235e-09,
-      "loss": 0.0511,
-      "step": 996
-    },
-    {
-      "epoch": 11.869047619047619,
-      "grad_norm": 5.046944747259621,
-      "learning_rate": 6.254904926820127e-09,
-      "loss": 0.1221,
-      "step": 997
-    },
-    {
-      "epoch": 11.880952380952381,
-      "grad_norm": 1.087638003537602,
-      "learning_rate": 5.169436454478538e-09,
-      "loss": 0.0359,
-      "step": 998
-    },
-    {
-      "epoch": 11.892857142857142,
-      "grad_norm": 2.6782948586641786,
-      "learning_rate": 4.187312080347061e-09,
-      "loss": 0.0578,
-      "step": 999
-    },
-    {
-      "epoch": 11.904761904761905,
-      "grad_norm": 1.4941663210879166,
-      "learning_rate": 3.3085419593514857e-09,
-      "loss": 0.0301,
-      "step": 1000
-    },
-    {
-      "epoch": 11.916666666666666,
-      "grad_norm": 2.03536965767824,
-      "learning_rate": 2.533135177760215e-09,
-      "loss": 0.0353,
-      "step": 1001
-    },
-    {
-      "epoch": 11.928571428571429,
-      "grad_norm": 2.870140395549658,
-      "learning_rate": 1.8610997530876806e-09,
-      "loss": 0.0526,
-      "step": 1002
-    },
-    {
-      "epoch": 11.94047619047619,
-      "grad_norm": 1.9821994732842032,
-      "learning_rate": 1.2924426340177321e-09,
-      "loss": 0.0368,
-      "step": 1003
-    },
-    {
-      "epoch": 11.952380952380953,
-      "grad_norm": 2.570985111435327,
-      "learning_rate": 8.271697003237045e-10,
-      "loss": 0.0624,
-      "step": 1004
-    },
-    {
-      "epoch": 11.964285714285714,
-      "grad_norm": 1.3718413803749656,
-      "learning_rate": 4.6528576281401616e-10,
-      "loss": 0.0661,
-      "step": 1005
-    },
-    {
-      "epoch": 11.976190476190476,
-      "grad_norm": 2.371522463710688,
-      "learning_rate": 2.0679456327998838e-10,
-      "loss": 0.062,
-      "step": 1006
-    },
-    {
-      "epoch": 11.988095238095237,
-      "grad_norm": 4.216784477215788,
-      "learning_rate": 5.169877445809768e-11,
-      "loss": 0.0267,
-      "step": 1007
-    },
-    {
-      "epoch": 12.0,
-      "grad_norm": 3.6030455433551953,
-      "learning_rate": 0.0,
-      "loss": 0.0259,
-      "step": 1008
-    },
-    {
-      "epoch": 12.0,
-      "eval_loss": 0.29545679688453674,
-      "eval_runtime": 38.8506,
-      "eval_samples_per_second": 1.544,
-      "eval_steps_per_second": 1.544,
-      "step": 1008
+      "epoch": 11.899159663865547,
+      "eval_loss": 0.08401793986558914,
+      "eval_runtime": 85.0205,
+      "eval_samples_per_second": 1.106,
+      "eval_steps_per_second": 1.106,
+      "step": 708
     },
     {
-      "epoch": 12.0,
-      "step": 1008,
-      "total_flos": 280627377340416.0,
-      "train_loss": 0.48460018415431794,
-      "train_runtime": 6901.6294,
-      "train_samples_per_second": 0.584,
-      "train_steps_per_second": 0.146
+      "epoch": 11.899159663865547,
+      "step": 708,
+      "total_flos": 356037728403456.0,
+      "train_loss": 0.4260152625183096,
+      "train_runtime": 5257.2185,
+      "train_samples_per_second": 1.08,
+      "train_steps_per_second": 0.135
     }
   ],
   "logging_steps": 1.0,
-  "max_steps": 1008,
+  "max_steps": 708,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 12,
-  "save_steps": 500,
+  "save_steps": 200.0,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -7187,8 +5087,8 @@
       "attributes": {}
     }
   },
-  "total_flos": 280627377340416.0,
-  "train_batch_size": 1,
+  "total_flos": 356037728403456.0,
+  "train_batch_size": 4,
   "trial_name": null,
   "trial_params": null
 }