diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -4,5789 +4,5831 @@
   "best_model_checkpoint": null,
   "epoch": 3.0,
   "eval_steps": 500,
-  "global_step": 825,
+  "global_step": 831,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.0036363636363636364,
-      "grad_norm": 47.79498291015625,
+      "epoch": 0.0036101083032490976,
+      "grad_norm": 53.0424919128418,
       "learning_rate": 5e-05,
-      "loss": 20.1422,
+      "loss": 20.1881,
       "step": 1
     },
     {
-      "epoch": 0.007272727272727273,
-      "grad_norm": 45.03392028808594,
-      "learning_rate": 4.993939393939394e-05,
-      "loss": 19.7063,
+      "epoch": 0.007220216606498195,
+      "grad_norm": 40.154483795166016,
+      "learning_rate": 4.9939831528279186e-05,
+      "loss": 18.4212,
       "step": 2
     },
     {
-      "epoch": 0.01090909090909091,
-      "grad_norm": 41.21010971069336,
-      "learning_rate": 4.987878787878788e-05,
-      "loss": 18.7882,
+      "epoch": 0.010830324909747292,
+      "grad_norm": 41.030975341796875,
+      "learning_rate": 4.987966305655837e-05,
+      "loss": 18.1343,
       "step": 3
     },
     {
-      "epoch": 0.014545454545454545,
-      "grad_norm": 41.21604919433594,
-      "learning_rate": 4.981818181818182e-05,
-      "loss": 18.4652,
+      "epoch": 0.01444043321299639,
+      "grad_norm": 38.574737548828125,
+      "learning_rate": 4.981949458483755e-05,
+      "loss": 17.831,
       "step": 4
     },
     {
-      "epoch": 0.01818181818181818,
-      "grad_norm": 36.21095657348633,
-      "learning_rate": 4.9757575757575756e-05,
-      "loss": 17.261,
+      "epoch": 0.018050541516245487,
+      "grad_norm": 34.99953079223633,
+      "learning_rate": 4.9759326113116724e-05,
+      "loss": 16.8969,
       "step": 5
     },
     {
-      "epoch": 0.02181818181818182,
-      "grad_norm": 33.17779541015625,
-      "learning_rate": 4.9696969696969694e-05,
-      "loss": 17.3998,
+      "epoch": 0.021660649819494584,
+      "grad_norm": 36.076698303222656,
+      "learning_rate": 4.969915764139591e-05,
+      "loss": 17.5522,
       "step": 6
     },
     {
-      "epoch": 0.025454545454545455,
-      "grad_norm": 27.718563079833984,
-      "learning_rate": 4.963636363636364e-05,
-      "loss": 16.7747,
+      "epoch": 0.02527075812274368,
+      "grad_norm": 40.10169982910156,
+      "learning_rate": 4.963898916967509e-05,
+      "loss": 17.2688,
       "step": 7
     },
     {
-      "epoch": 0.02909090909090909,
-      "grad_norm": 29.30256462097168,
-      "learning_rate": 4.957575757575758e-05,
-      "loss": 16.5601,
+      "epoch": 0.02888086642599278,
+      "grad_norm": 30.18881607055664,
+      "learning_rate": 4.9578820697954276e-05,
+      "loss": 16.2351,
       "step": 8
     },
     {
-      "epoch": 0.03272727272727273,
-      "grad_norm": 29.383529663085938,
-      "learning_rate": 4.951515151515152e-05,
-      "loss": 17.3504,
+      "epoch": 0.032490974729241874,
+      "grad_norm": 30.67852020263672,
+      "learning_rate": 4.951865222623346e-05,
+      "loss": 17.0894,
       "step": 9
     },
     {
-      "epoch": 0.03636363636363636,
-      "grad_norm": 23.729032516479492,
-      "learning_rate": 4.945454545454546e-05,
-      "loss": 16.3259,
+      "epoch": 0.036101083032490974,
+      "grad_norm": 27.552501678466797,
+      "learning_rate": 4.945848375451264e-05,
+      "loss": 16.7366,
       "step": 10
     },
     {
-      "epoch": 0.04,
-      "grad_norm": 25.69147300720215,
-      "learning_rate": 4.93939393939394e-05,
-      "loss": 17.0484,
+      "epoch": 0.039711191335740074,
+      "grad_norm": 26.315004348754883,
+      "learning_rate": 4.939831528279182e-05,
+      "loss": 16.6955,
       "step": 11
     },
     {
-      "epoch": 0.04363636363636364,
-      "grad_norm": 32.13129425048828,
-      "learning_rate": 4.933333333333334e-05,
-      "loss": 16.2835,
+      "epoch": 0.04332129963898917,
+      "grad_norm": 27.678667068481445,
+      "learning_rate": 4.9338146811071e-05,
+      "loss": 16.1629,
       "step": 12
     },
     {
-      "epoch": 0.04727272727272727,
-      "grad_norm": 33.38395690917969,
-      "learning_rate": 4.9272727272727276e-05,
-      "loss": 16.2865,
+      "epoch": 0.04693140794223827,
+      "grad_norm": 39.11885452270508,
+      "learning_rate": 4.927797833935018e-05,
+      "loss": 15.8583,
       "step": 13
     },
     {
-      "epoch": 0.05090909090909091,
-      "grad_norm": 47.96619415283203,
-      "learning_rate": 4.9212121212121214e-05,
-      "loss": 16.763,
+      "epoch": 0.05054151624548736,
+      "grad_norm": 44.302581787109375,
+      "learning_rate": 4.9217809867629366e-05,
+      "loss": 16.4332,
       "step": 14
     },
     {
-      "epoch": 0.05454545454545454,
-      "grad_norm": 40.4400520324707,
-      "learning_rate": 4.915151515151515e-05,
-      "loss": 15.6799,
+      "epoch": 0.05415162454873646,
+      "grad_norm": 48.368934631347656,
+      "learning_rate": 4.915764139590854e-05,
+      "loss": 15.6681,
       "step": 15
     },
     {
-      "epoch": 0.05818181818181818,
-      "grad_norm": 32.743865966796875,
-      "learning_rate": 4.909090909090909e-05,
-      "loss": 15.7916,
+      "epoch": 0.05776173285198556,
+      "grad_norm": 62.672847747802734,
+      "learning_rate": 4.909747292418773e-05,
+      "loss": 15.9232,
       "step": 16
     },
     {
-      "epoch": 0.06181818181818182,
-      "grad_norm": 56.1345329284668,
-      "learning_rate": 4.903030303030303e-05,
-      "loss": 15.7303,
+      "epoch": 0.061371841155234655,
+      "grad_norm": 56.098548889160156,
+      "learning_rate": 4.903730445246691e-05,
+      "loss": 15.6061,
       "step": 17
     },
     {
-      "epoch": 0.06545454545454546,
-      "grad_norm": 43.6757698059082,
-      "learning_rate": 4.896969696969697e-05,
-      "loss": 15.492,
+      "epoch": 0.06498194945848375,
+      "grad_norm": 47.98686218261719,
+      "learning_rate": 4.8977135980746095e-05,
+      "loss": 15.2327,
       "step": 18
     },
     {
-      "epoch": 0.06909090909090909,
-      "grad_norm": 31.908369064331055,
-      "learning_rate": 4.890909090909091e-05,
-      "loss": 15.177,
+      "epoch": 0.06859205776173286,
+      "grad_norm": 49.93760299682617,
+      "learning_rate": 4.891696750902527e-05,
+      "loss": 15.6251,
       "step": 19
     },
     {
-      "epoch": 0.07272727272727272,
-      "grad_norm": 29.7440185546875,
-      "learning_rate": 4.884848484848485e-05,
-      "loss": 15.2295,
+      "epoch": 0.07220216606498195,
+      "grad_norm": 34.30894470214844,
+      "learning_rate": 4.8856799037304456e-05,
+      "loss": 14.8196,
       "step": 20
     },
     {
-      "epoch": 0.07636363636363637,
-      "grad_norm": 24.505130767822266,
-      "learning_rate": 4.878787878787879e-05,
-      "loss": 14.9099,
+      "epoch": 0.07581227436823104,
+      "grad_norm": 33.243125915527344,
+      "learning_rate": 4.879663056558363e-05,
+      "loss": 14.5626,
       "step": 21
     },
     {
-      "epoch": 0.08,
-      "grad_norm": 23.924467086791992,
-      "learning_rate": 4.872727272727273e-05,
-      "loss": 15.062,
+      "epoch": 0.07942238267148015,
+      "grad_norm": 27.303117752075195,
+      "learning_rate": 4.873646209386282e-05,
+      "loss": 14.1914,
       "step": 22
     },
     {
-      "epoch": 0.08363636363636363,
-      "grad_norm": 25.439834594726562,
-      "learning_rate": 4.866666666666667e-05,
-      "loss": 14.1283,
+      "epoch": 0.08303249097472924,
+      "grad_norm": 30.314437866210938,
+      "learning_rate": 4.8676293622142e-05,
+      "loss": 15.2119,
       "step": 23
     },
     {
-      "epoch": 0.08727272727272728,
-      "grad_norm": 21.35331153869629,
-      "learning_rate": 4.860606060606061e-05,
-      "loss": 13.9236,
+      "epoch": 0.08664259927797834,
+      "grad_norm": 25.56382942199707,
+      "learning_rate": 4.8616125150421185e-05,
+      "loss": 14.5087,
       "step": 24
     },
     {
-      "epoch": 0.09090909090909091,
-      "grad_norm": 26.658445358276367,
-      "learning_rate": 4.854545454545455e-05,
-      "loss": 14.4044,
+      "epoch": 0.09025270758122744,
+      "grad_norm": 28.333757400512695,
+      "learning_rate": 4.855595667870036e-05,
+      "loss": 14.3963,
       "step": 25
     },
     {
-      "epoch": 0.09454545454545454,
-      "grad_norm": 25.977502822875977,
-      "learning_rate": 4.848484848484849e-05,
-      "loss": 13.7121,
+      "epoch": 0.09386281588447654,
+      "grad_norm": 22.039608001708984,
+      "learning_rate": 4.8495788206979546e-05,
+      "loss": 14.4985,
       "step": 26
     },
     {
-      "epoch": 0.09818181818181818,
-      "grad_norm": 32.09416580200195,
-      "learning_rate": 4.8424242424242426e-05,
-      "loss": 13.8631,
+      "epoch": 0.09747292418772563,
+      "grad_norm": 27.54795265197754,
+      "learning_rate": 4.843561973525872e-05,
+      "loss": 13.7732,
       "step": 27
     },
     {
-      "epoch": 0.10181818181818182,
-      "grad_norm": 23.397937774658203,
-      "learning_rate": 4.8363636363636364e-05,
-      "loss": 13.3994,
+      "epoch": 0.10108303249097472,
+      "grad_norm": 21.023685455322266,
+      "learning_rate": 4.837545126353791e-05,
+      "loss": 13.5031,
       "step": 28
     },
     {
-      "epoch": 0.10545454545454545,
-      "grad_norm": 28.557424545288086,
-      "learning_rate": 4.83030303030303e-05,
-      "loss": 13.83,
+      "epoch": 0.10469314079422383,
+      "grad_norm": 21.715377807617188,
+      "learning_rate": 4.831528279181709e-05,
+      "loss": 13.9961,
       "step": 29
     },
     {
-      "epoch": 0.10909090909090909,
-      "grad_norm": 17.376995086669922,
-      "learning_rate": 4.824242424242425e-05,
-      "loss": 13.0998,
+      "epoch": 0.10830324909747292,
+      "grad_norm": 18.735675811767578,
+      "learning_rate": 4.8255114320096275e-05,
+      "loss": 13.2735,
       "step": 30
     },
     {
-      "epoch": 0.11272727272727273,
-      "grad_norm": 26.38998031616211,
-      "learning_rate": 4.8181818181818186e-05,
-      "loss": 13.134,
+      "epoch": 0.11191335740072202,
+      "grad_norm": 19.733570098876953,
+      "learning_rate": 4.819494584837546e-05,
+      "loss": 13.2835,
       "step": 31
     },
     {
-      "epoch": 0.11636363636363636,
-      "grad_norm": 17.54689598083496,
-      "learning_rate": 4.8121212121212125e-05,
-      "loss": 12.8039,
+      "epoch": 0.11552346570397112,
+      "grad_norm": 18.386388778686523,
+      "learning_rate": 4.8134777376654636e-05,
+      "loss": 13.3851,
       "step": 32
     },
     {
-      "epoch": 0.12,
-      "grad_norm": 18.970720291137695,
-      "learning_rate": 4.806060606060606e-05,
-      "loss": 13.0586,
+      "epoch": 0.11913357400722022,
+      "grad_norm": 18.647968292236328,
+      "learning_rate": 4.807460890493381e-05,
+      "loss": 12.7988,
       "step": 33
     },
     {
-      "epoch": 0.12363636363636364,
-      "grad_norm": 19.773618698120117,
-      "learning_rate": 4.8e-05,
-      "loss": 13.0412,
+      "epoch": 0.12274368231046931,
+      "grad_norm": 17.047182083129883,
+      "learning_rate": 4.8014440433213e-05,
+      "loss": 12.8917,
       "step": 34
     },
     {
-      "epoch": 0.12727272727272726,
-      "grad_norm": 22.805259704589844,
-      "learning_rate": 4.793939393939394e-05,
-      "loss": 12.8489,
+      "epoch": 0.1263537906137184,
+      "grad_norm": 17.250154495239258,
+      "learning_rate": 4.795427196149218e-05,
+      "loss": 12.9744,
       "step": 35
     },
     {
-      "epoch": 0.13090909090909092,
-      "grad_norm": 23.569717407226562,
-      "learning_rate": 4.787878787878788e-05,
-      "loss": 12.616,
+      "epoch": 0.1299638989169675,
+      "grad_norm": 19.833730697631836,
+      "learning_rate": 4.7894103489771365e-05,
+      "loss": 13.001,
       "step": 36
     },
     {
-      "epoch": 0.13454545454545455,
-      "grad_norm": 22.83664321899414,
-      "learning_rate": 4.781818181818182e-05,
-      "loss": 11.8258,
+      "epoch": 0.13357400722021662,
+      "grad_norm": 16.90082550048828,
+      "learning_rate": 4.783393501805055e-05,
+      "loss": 12.4627,
       "step": 37
     },
     {
-      "epoch": 0.13818181818181818,
-      "grad_norm": 26.738494873046875,
-      "learning_rate": 4.775757575757576e-05,
-      "loss": 12.1916,
+      "epoch": 0.1371841155234657,
+      "grad_norm": 15.739155769348145,
+      "learning_rate": 4.7773766546329726e-05,
+      "loss": 12.4672,
       "step": 38
     },
     {
-      "epoch": 0.14181818181818182,
-      "grad_norm": 18.36771583557129,
-      "learning_rate": 4.76969696969697e-05,
-      "loss": 12.2002,
+      "epoch": 0.1407942238267148,
+      "grad_norm": 16.412572860717773,
+      "learning_rate": 4.771359807460891e-05,
+      "loss": 11.8982,
       "step": 39
     },
     {
-      "epoch": 0.14545454545454545,
-      "grad_norm": 15.442994117736816,
-      "learning_rate": 4.763636363636364e-05,
-      "loss": 11.7265,
+      "epoch": 0.1444043321299639,
+      "grad_norm": 15.304023742675781,
+      "learning_rate": 4.765342960288809e-05,
+      "loss": 12.2821,
       "step": 40
     },
     {
-      "epoch": 0.14909090909090908,
-      "grad_norm": 17.741811752319336,
-      "learning_rate": 4.7575757575757576e-05,
-      "loss": 12.4528,
+      "epoch": 0.148014440433213,
+      "grad_norm": 15.000272750854492,
+      "learning_rate": 4.759326113116727e-05,
+      "loss": 11.3867,
       "step": 41
     },
     {
-      "epoch": 0.15272727272727274,
-      "grad_norm": 14.73928451538086,
-      "learning_rate": 4.751515151515152e-05,
-      "loss": 11.5017,
+      "epoch": 0.15162454873646208,
+      "grad_norm": 14.771124839782715,
+      "learning_rate": 4.7533092659446455e-05,
+      "loss": 11.7266,
       "step": 42
     },
     {
-      "epoch": 0.15636363636363637,
-      "grad_norm": 17.848480224609375,
-      "learning_rate": 4.745454545454546e-05,
-      "loss": 11.8774,
+      "epoch": 0.1552346570397112,
+      "grad_norm": 15.111937522888184,
+      "learning_rate": 4.747292418772563e-05,
+      "loss": 11.378,
       "step": 43
     },
     {
-      "epoch": 0.16,
-      "grad_norm": 17.191818237304688,
-      "learning_rate": 4.73939393939394e-05,
-      "loss": 11.0546,
+      "epoch": 0.1588447653429603,
+      "grad_norm": 14.241510391235352,
+      "learning_rate": 4.7412755716004816e-05,
+      "loss": 11.4672,
       "step": 44
     },
     {
-      "epoch": 0.16363636363636364,
-      "grad_norm": 15.763983726501465,
-      "learning_rate": 4.7333333333333336e-05,
-      "loss": 11.2831,
+      "epoch": 0.1624548736462094,
+      "grad_norm": 14.913339614868164,
+      "learning_rate": 4.7352587244284e-05,
+      "loss": 11.2229,
       "step": 45
     },
     {
-      "epoch": 0.16727272727272727,
-      "grad_norm": 16.031883239746094,
-      "learning_rate": 4.7272727272727275e-05,
-      "loss": 11.0432,
+      "epoch": 0.16606498194945848,
+      "grad_norm": 14.72385025024414,
+      "learning_rate": 4.7292418772563177e-05,
+      "loss": 10.8122,
       "step": 46
     },
     {
-      "epoch": 0.1709090909090909,
-      "grad_norm": 14.979501724243164,
-      "learning_rate": 4.721212121212121e-05,
-      "loss": 11.0545,
+      "epoch": 0.16967509025270758,
+      "grad_norm": 13.52721881866455,
+      "learning_rate": 4.723225030084236e-05,
+      "loss": 10.9708,
       "step": 47
     },
     {
-      "epoch": 0.17454545454545456,
-      "grad_norm": 15.489818572998047,
-      "learning_rate": 4.715151515151515e-05,
-      "loss": 11.3819,
+      "epoch": 0.17328519855595667,
+      "grad_norm": 13.60007095336914,
+      "learning_rate": 4.7172081829121544e-05,
+      "loss": 10.807,
       "step": 48
     },
     {
-      "epoch": 0.1781818181818182,
-      "grad_norm": 14.921260833740234,
-      "learning_rate": 4.709090909090909e-05,
-      "loss": 11.057,
+      "epoch": 0.17689530685920576,
+      "grad_norm": 13.957290649414062,
+      "learning_rate": 4.711191335740072e-05,
+      "loss": 10.7637,
       "step": 49
     },
     {
-      "epoch": 0.18181818181818182,
-      "grad_norm": 14.218314170837402,
-      "learning_rate": 4.703030303030303e-05,
-      "loss": 10.7186,
+      "epoch": 0.18050541516245489,
+      "grad_norm": 14.956080436706543,
+      "learning_rate": 4.7051744885679905e-05,
+      "loss": 10.9521,
       "step": 50
     },
     {
-      "epoch": 0.18545454545454546,
-      "grad_norm": 13.922761917114258,
-      "learning_rate": 4.696969696969697e-05,
-      "loss": 10.7164,
+      "epoch": 0.18411552346570398,
+      "grad_norm": 13.179266929626465,
+      "learning_rate": 4.699157641395909e-05,
+      "loss": 10.5118,
       "step": 51
     },
     {
-      "epoch": 0.1890909090909091,
-      "grad_norm": 14.378653526306152,
-      "learning_rate": 4.690909090909091e-05,
-      "loss": 10.6261,
+      "epoch": 0.18772563176895307,
+      "grad_norm": 16.000741958618164,
+      "learning_rate": 4.693140794223827e-05,
+      "loss": 10.7588,
       "step": 52
     },
     {
-      "epoch": 0.19272727272727272,
-      "grad_norm": 13.454336166381836,
-      "learning_rate": 4.684848484848485e-05,
-      "loss": 10.3784,
+      "epoch": 0.19133574007220217,
+      "grad_norm": 21.888368606567383,
+      "learning_rate": 4.687123947051745e-05,
+      "loss": 10.6924,
       "step": 53
     },
     {
-      "epoch": 0.19636363636363635,
-      "grad_norm": 14.680737495422363,
-      "learning_rate": 4.6787878787878795e-05,
-      "loss": 10.3173,
+      "epoch": 0.19494584837545126,
+      "grad_norm": 14.187573432922363,
+      "learning_rate": 4.681107099879663e-05,
+      "loss": 10.3625,
       "step": 54
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 15.582632064819336,
-      "learning_rate": 4.672727272727273e-05,
-      "loss": 10.1981,
+      "epoch": 0.19855595667870035,
+      "grad_norm": 13.989165306091309,
+      "learning_rate": 4.675090252707581e-05,
+      "loss": 9.9838,
       "step": 55
     },
     {
-      "epoch": 0.20363636363636364,
-      "grad_norm": 13.507526397705078,
-      "learning_rate": 4.666666666666667e-05,
-      "loss": 10.1057,
+      "epoch": 0.20216606498194944,
+      "grad_norm": 17.044803619384766,
+      "learning_rate": 4.6690734055354995e-05,
+      "loss": 10.3739,
       "step": 56
     },
     {
-      "epoch": 0.20727272727272728,
-      "grad_norm": 12.818306922912598,
-      "learning_rate": 4.660606060606061e-05,
-      "loss": 9.9707,
+      "epoch": 0.20577617328519857,
+      "grad_norm": 14.635570526123047,
+      "learning_rate": 4.663056558363418e-05,
+      "loss": 9.9174,
       "step": 57
     },
     {
-      "epoch": 0.2109090909090909,
-      "grad_norm": 21.36762809753418,
-      "learning_rate": 4.654545454545455e-05,
-      "loss": 10.374,
+      "epoch": 0.20938628158844766,
+      "grad_norm": 13.500565528869629,
+      "learning_rate": 4.657039711191336e-05,
+      "loss": 9.8747,
       "step": 58
     },
     {
-      "epoch": 0.21454545454545454,
-      "grad_norm": 12.67024040222168,
-      "learning_rate": 4.6484848484848487e-05,
-      "loss": 9.8168,
+      "epoch": 0.21299638989169675,
+      "grad_norm": 13.646149635314941,
+      "learning_rate": 4.651022864019254e-05,
+      "loss": 9.7799,
       "step": 59
     },
     {
-      "epoch": 0.21818181818181817,
-      "grad_norm": 17.809871673583984,
-      "learning_rate": 4.6424242424242425e-05,
-      "loss": 9.5406,
+      "epoch": 0.21660649819494585,
+      "grad_norm": 13.92541217803955,
+      "learning_rate": 4.645006016847172e-05,
+      "loss": 9.5702,
       "step": 60
     },
     {
-      "epoch": 0.22181818181818183,
-      "grad_norm": 16.511991500854492,
-      "learning_rate": 4.636363636363636e-05,
-      "loss": 10.0524,
+      "epoch": 0.22021660649819494,
+      "grad_norm": 12.247454643249512,
+      "learning_rate": 4.63898916967509e-05,
+      "loss": 9.5437,
       "step": 61
     },
     {
-      "epoch": 0.22545454545454546,
-      "grad_norm": 14.556917190551758,
-      "learning_rate": 4.63030303030303e-05,
-      "loss": 9.8776,
+      "epoch": 0.22382671480144403,
+      "grad_norm": 13.187448501586914,
+      "learning_rate": 4.6329723225030085e-05,
+      "loss": 9.3673,
       "step": 62
     },
     {
-      "epoch": 0.2290909090909091,
-      "grad_norm": 21.906808853149414,
-      "learning_rate": 4.624242424242424e-05,
-      "loss": 9.2781,
+      "epoch": 0.22743682310469315,
+      "grad_norm": 16.612987518310547,
+      "learning_rate": 4.626955475330927e-05,
+      "loss": 9.4453,
       "step": 63
     },
     {
-      "epoch": 0.23272727272727273,
-      "grad_norm": 13.430219650268555,
-      "learning_rate": 4.618181818181818e-05,
-      "loss": 9.3972,
+      "epoch": 0.23104693140794225,
+      "grad_norm": 12.24003791809082,
+      "learning_rate": 4.620938628158845e-05,
+      "loss": 9.2878,
       "step": 64
     },
     {
-      "epoch": 0.23636363636363636,
-      "grad_norm": 16.371122360229492,
-      "learning_rate": 4.612121212121212e-05,
-      "loss": 9.2046,
+      "epoch": 0.23465703971119134,
+      "grad_norm": 12.949747085571289,
+      "learning_rate": 4.614921780986764e-05,
+      "loss": 9.2656,
       "step": 65
     },
     {
-      "epoch": 0.24,
-      "grad_norm": 17.033885955810547,
-      "learning_rate": 4.606060606060607e-05,
-      "loss": 9.4568,
+      "epoch": 0.23826714801444043,
+      "grad_norm": 15.98535442352295,
+      "learning_rate": 4.6089049338146814e-05,
+      "loss": 9.1888,
       "step": 66
     },
     {
-      "epoch": 0.24363636363636362,
-      "grad_norm": 19.499509811401367,
-      "learning_rate": 4.600000000000001e-05,
-      "loss": 9.3437,
+      "epoch": 0.24187725631768953,
+      "grad_norm": 12.989156723022461,
+      "learning_rate": 4.602888086642599e-05,
+      "loss": 8.9948,
       "step": 67
     },
     {
-      "epoch": 0.24727272727272728,
-      "grad_norm": 13.871404647827148,
-      "learning_rate": 4.5939393939393945e-05,
-      "loss": 9.2115,
+      "epoch": 0.24548736462093862,
+      "grad_norm": 19.851245880126953,
+      "learning_rate": 4.5968712394705175e-05,
+      "loss": 9.4233,
       "step": 68
     },
     {
-      "epoch": 0.2509090909090909,
-      "grad_norm": 20.010080337524414,
-      "learning_rate": 4.5878787878787883e-05,
-      "loss": 9.2831,
+      "epoch": 0.2490974729241877,
+      "grad_norm": 12.913300514221191,
+      "learning_rate": 4.590854392298436e-05,
+      "loss": 9.0834,
       "step": 69
     },
     {
-      "epoch": 0.2545454545454545,
-      "grad_norm": 17.872783660888672,
-      "learning_rate": 4.581818181818182e-05,
-      "loss": 9.3362,
+      "epoch": 0.2527075812274368,
+      "grad_norm": 14.333419799804688,
+      "learning_rate": 4.584837545126354e-05,
+      "loss": 9.1562,
       "step": 70
     },
     {
-      "epoch": 0.2581818181818182,
-      "grad_norm": 18.30128288269043,
-      "learning_rate": 4.575757575757576e-05,
-      "loss": 8.7413,
+      "epoch": 0.2563176895306859,
+      "grad_norm": 13.390801429748535,
+      "learning_rate": 4.578820697954272e-05,
+      "loss": 9.0329,
       "step": 71
     },
     {
-      "epoch": 0.26181818181818184,
-      "grad_norm": 21.37891387939453,
-      "learning_rate": 4.56969696969697e-05,
-      "loss": 8.6968,
+      "epoch": 0.259927797833935,
+      "grad_norm": 26.999013900756836,
+      "learning_rate": 4.5728038507821904e-05,
+      "loss": 8.702,
       "step": 72
     },
     {
-      "epoch": 0.26545454545454544,
-      "grad_norm": 20.191150665283203,
-      "learning_rate": 4.563636363636364e-05,
-      "loss": 8.7501,
+      "epoch": 0.26353790613718414,
+      "grad_norm": 15.850932121276855,
+      "learning_rate": 4.566787003610109e-05,
+      "loss": 8.9099,
       "step": 73
     },
     {
-      "epoch": 0.2690909090909091,
-      "grad_norm": 17.527233123779297,
-      "learning_rate": 4.5575757575757575e-05,
-      "loss": 8.589,
+      "epoch": 0.26714801444043323,
+      "grad_norm": 22.490720748901367,
+      "learning_rate": 4.5607701564380265e-05,
+      "loss": 8.6662,
       "step": 74
     },
     {
-      "epoch": 0.2727272727272727,
-      "grad_norm": 12.172165870666504,
-      "learning_rate": 4.5515151515151513e-05,
-      "loss": 8.5487,
+      "epoch": 0.27075812274368233,
+      "grad_norm": 21.35521125793457,
+      "learning_rate": 4.554753309265945e-05,
+      "loss": 8.3598,
       "step": 75
     },
     {
-      "epoch": 0.27636363636363637,
-      "grad_norm": 12.90199089050293,
-      "learning_rate": 4.545454545454546e-05,
-      "loss": 8.4269,
+      "epoch": 0.2743682310469314,
+      "grad_norm": 14.975168228149414,
+      "learning_rate": 4.548736462093863e-05,
+      "loss": 8.2639,
       "step": 76
     },
     {
-      "epoch": 0.28,
-      "grad_norm": 14.662657737731934,
-      "learning_rate": 4.53939393939394e-05,
-      "loss": 8.47,
+      "epoch": 0.2779783393501805,
+      "grad_norm": 14.411498069763184,
+      "learning_rate": 4.542719614921781e-05,
+      "loss": 8.2173,
       "step": 77
     },
     {
-      "epoch": 0.28363636363636363,
-      "grad_norm": 12.177051544189453,
-      "learning_rate": 4.5333333333333335e-05,
-      "loss": 8.1216,
+      "epoch": 0.2815884476534296,
+      "grad_norm": 15.177107810974121,
+      "learning_rate": 4.5367027677496994e-05,
+      "loss": 8.2124,
       "step": 78
     },
     {
-      "epoch": 0.2872727272727273,
-      "grad_norm": 15.531538009643555,
-      "learning_rate": 4.5272727272727274e-05,
-      "loss": 8.2657,
+      "epoch": 0.2851985559566787,
+      "grad_norm": 21.427085876464844,
+      "learning_rate": 4.530685920577618e-05,
+      "loss": 8.1488,
       "step": 79
     },
     {
-      "epoch": 0.2909090909090909,
-      "grad_norm": 42.798797607421875,
-      "learning_rate": 4.521212121212122e-05,
-      "loss": 8.5536,
+      "epoch": 0.2888086642599278,
+      "grad_norm": 18.792469024658203,
+      "learning_rate": 4.5246690734055355e-05,
+      "loss": 8.0504,
       "step": 80
     },
     {
-      "epoch": 0.29454545454545455,
-      "grad_norm": 14.968511581420898,
-      "learning_rate": 4.515151515151516e-05,
-      "loss": 8.0826,
+      "epoch": 0.2924187725631769,
+      "grad_norm": 13.447463989257812,
+      "learning_rate": 4.518652226233454e-05,
+      "loss": 7.7557,
       "step": 81
     },
     {
-      "epoch": 0.29818181818181816,
-      "grad_norm": 13.593378067016602,
-      "learning_rate": 4.5090909090909095e-05,
-      "loss": 8.0562,
+      "epoch": 0.296028880866426,
+      "grad_norm": 11.585929870605469,
+      "learning_rate": 4.5126353790613716e-05,
+      "loss": 7.7533,
       "step": 82
     },
     {
-      "epoch": 0.3018181818181818,
-      "grad_norm": 20.985275268554688,
-      "learning_rate": 4.5030303030303034e-05,
-      "loss": 7.7742,
+      "epoch": 0.2996389891696751,
+      "grad_norm": 16.441787719726562,
+      "learning_rate": 4.50661853188929e-05,
+      "loss": 7.6073,
       "step": 83
     },
     {
-      "epoch": 0.3054545454545455,
-      "grad_norm": 22.25298309326172,
-      "learning_rate": 4.496969696969697e-05,
-      "loss": 7.8028,
+      "epoch": 0.30324909747292417,
+      "grad_norm": 12.96184253692627,
+      "learning_rate": 4.5006016847172084e-05,
+      "loss": 7.7666,
       "step": 84
     },
     {
-      "epoch": 0.3090909090909091,
-      "grad_norm": 14.785983085632324,
-      "learning_rate": 4.490909090909091e-05,
-      "loss": 8.0702,
+      "epoch": 0.30685920577617326,
+      "grad_norm": 32.2669677734375,
+      "learning_rate": 4.494584837545127e-05,
+      "loss": 8.1308,
       "step": 85
     },
     {
-      "epoch": 0.31272727272727274,
-      "grad_norm": 19.781902313232422,
-      "learning_rate": 4.484848484848485e-05,
-      "loss": 7.6381,
+      "epoch": 0.3104693140794224,
+      "grad_norm": 18.520849227905273,
+      "learning_rate": 4.488567990373045e-05,
+      "loss": 7.8578,
       "step": 86
     },
     {
-      "epoch": 0.31636363636363635,
-      "grad_norm": 14.84749698638916,
-      "learning_rate": 4.478787878787879e-05,
-      "loss": 7.635,
+      "epoch": 0.3140794223826715,
+      "grad_norm": 18.274093627929688,
+      "learning_rate": 4.482551143200963e-05,
+      "loss": 7.8863,
       "step": 87
     },
     {
-      "epoch": 0.32,
-      "grad_norm": 15.319684982299805,
-      "learning_rate": 4.472727272727273e-05,
-      "loss": 7.3204,
+      "epoch": 0.3176895306859206,
+      "grad_norm": 20.193405151367188,
+      "learning_rate": 4.4765342960288806e-05,
+      "loss": 7.6916,
       "step": 88
     },
     {
-      "epoch": 0.3236363636363636,
-      "grad_norm": 17.83990478515625,
-      "learning_rate": 4.466666666666667e-05,
-      "loss": 7.4941,
+      "epoch": 0.3212996389891697,
+      "grad_norm": 27.587148666381836,
+      "learning_rate": 4.470517448856799e-05,
+      "loss": 7.3948,
       "step": 89
     },
     {
-      "epoch": 0.32727272727272727,
-      "grad_norm": 13.8068265914917,
-      "learning_rate": 4.460606060606061e-05,
-      "loss": 7.3864,
+      "epoch": 0.3249097472924188,
+      "grad_norm": 19.598400115966797,
+      "learning_rate": 4.4645006016847174e-05,
+      "loss": 7.6276,
       "step": 90
     },
     {
-      "epoch": 0.33090909090909093,
-      "grad_norm": 42.767478942871094,
-      "learning_rate": 4.454545454545455e-05,
-      "loss": 7.6903,
+      "epoch": 0.3285198555956679,
+      "grad_norm": 34.88701629638672,
+      "learning_rate": 4.458483754512636e-05,
+      "loss": 7.29,
       "step": 91
     },
     {
-      "epoch": 0.33454545454545453,
-      "grad_norm": 12.295734405517578,
-      "learning_rate": 4.4484848484848485e-05,
-      "loss": 7.14,
+      "epoch": 0.33212996389891697,
+      "grad_norm": 27.295541763305664,
+      "learning_rate": 4.452466907340554e-05,
+      "loss": 7.2514,
       "step": 92
     },
     {
-      "epoch": 0.3381818181818182,
-      "grad_norm": 13.037458419799805,
-      "learning_rate": 4.4424242424242424e-05,
-      "loss": 7.2247,
+      "epoch": 0.33574007220216606,
+      "grad_norm": 23.34226417541504,
+      "learning_rate": 4.4464500601684725e-05,
+      "loss": 7.0975,
       "step": 93
     },
     {
-      "epoch": 0.3418181818181818,
-      "grad_norm": 14.935813903808594,
-      "learning_rate": 4.436363636363637e-05,
-      "loss": 7.1593,
+      "epoch": 0.33935018050541516,
+      "grad_norm": 15.4843111038208,
+      "learning_rate": 4.44043321299639e-05,
+      "loss": 7.039,
       "step": 94
     },
     {
-      "epoch": 0.34545454545454546,
-      "grad_norm": 23.534366607666016,
-      "learning_rate": 4.430303030303031e-05,
-      "loss": 7.4126,
+      "epoch": 0.34296028880866425,
+      "grad_norm": 33.904781341552734,
+      "learning_rate": 4.434416365824308e-05,
+      "loss": 7.1548,
       "step": 95
     },
     {
-      "epoch": 0.3490909090909091,
-      "grad_norm": 15.437704086303711,
-      "learning_rate": 4.4242424242424246e-05,
-      "loss": 7.2912,
+      "epoch": 0.34657039711191334,
+      "grad_norm": 18.602880477905273,
+      "learning_rate": 4.4283995186522264e-05,
+      "loss": 6.866,
       "step": 96
     },
     {
-      "epoch": 0.3527272727272727,
-      "grad_norm": 19.122468948364258,
-      "learning_rate": 4.4181818181818184e-05,
-      "loss": 7.2778,
+      "epoch": 0.35018050541516244,
+      "grad_norm": 68.61976623535156,
+      "learning_rate": 4.422382671480145e-05,
+      "loss": 7.4221,
       "step": 97
     },
     {
-      "epoch": 0.3563636363636364,
-      "grad_norm": 29.849777221679688,
-      "learning_rate": 4.412121212121212e-05,
-      "loss": 6.9932,
+      "epoch": 0.35379061371841153,
+      "grad_norm": 29.8818416595459,
+      "learning_rate": 4.416365824308063e-05,
+      "loss": 6.8978,
       "step": 98
     },
     {
-      "epoch": 0.36,
-      "grad_norm": 31.232065200805664,
-      "learning_rate": 4.406060606060606e-05,
-      "loss": 6.9073,
+      "epoch": 0.3574007220216607,
+      "grad_norm": 19.478824615478516,
+      "learning_rate": 4.410348977135981e-05,
+      "loss": 6.7587,
       "step": 99
     },
     {
-      "epoch": 0.36363636363636365,
-      "grad_norm": 20.299890518188477,
-      "learning_rate": 4.4000000000000006e-05,
-      "loss": 6.7348,
+      "epoch": 0.36101083032490977,
+      "grad_norm": 18.693496704101562,
+      "learning_rate": 4.404332129963899e-05,
+      "loss": 6.8979,
       "step": 100
     },
     {
-      "epoch": 0.36727272727272725,
-      "grad_norm": 12.967691421508789,
-      "learning_rate": 4.3939393939393944e-05,
-      "loss": 6.7341,
+      "epoch": 0.36462093862815886,
+      "grad_norm": 20.177623748779297,
+      "learning_rate": 4.398315282791817e-05,
+      "loss": 6.6204,
       "step": 101
     },
     {
-      "epoch": 0.3709090909090909,
-      "grad_norm": 33.69974899291992,
-      "learning_rate": 4.387878787878788e-05,
-      "loss": 7.1418,
+      "epoch": 0.36823104693140796,
+      "grad_norm": 21.361373901367188,
+      "learning_rate": 4.3922984356197353e-05,
+      "loss": 6.685,
       "step": 102
     },
     {
-      "epoch": 0.37454545454545457,
-      "grad_norm": 16.100889205932617,
-      "learning_rate": 4.381818181818182e-05,
-      "loss": 6.6484,
+      "epoch": 0.37184115523465705,
+      "grad_norm": 22.163894653320312,
+      "learning_rate": 4.386281588447654e-05,
+      "loss": 6.6413,
       "step": 103
     },
     {
-      "epoch": 0.3781818181818182,
-      "grad_norm": 16.84703254699707,
-      "learning_rate": 4.375757575757576e-05,
-      "loss": 6.7367,
+      "epoch": 0.37545126353790614,
+      "grad_norm": 25.493696212768555,
+      "learning_rate": 4.380264741275572e-05,
+      "loss": 6.5075,
       "step": 104
     },
     {
-      "epoch": 0.38181818181818183,
-      "grad_norm": 12.23398494720459,
-      "learning_rate": 4.36969696969697e-05,
-      "loss": 6.5239,
+      "epoch": 0.37906137184115524,
+      "grad_norm": 21.208959579467773,
+      "learning_rate": 4.37424789410349e-05,
+      "loss": 6.5093,
       "step": 105
     },
     {
-      "epoch": 0.38545454545454544,
-      "grad_norm": 16.255165100097656,
-      "learning_rate": 4.3636363636363636e-05,
-      "loss": 6.409,
+      "epoch": 0.38267148014440433,
+      "grad_norm": 16.638938903808594,
+      "learning_rate": 4.368231046931408e-05,
+      "loss": 6.853,
       "step": 106
     },
     {
-      "epoch": 0.3890909090909091,
-      "grad_norm": 14.847028732299805,
-      "learning_rate": 4.3575757575757574e-05,
-      "loss": 6.4569,
+      "epoch": 0.3862815884476534,
+      "grad_norm": 14.78966999053955,
+      "learning_rate": 4.3622141997593266e-05,
+      "loss": 6.601,
       "step": 107
     },
     {
-      "epoch": 0.3927272727272727,
-      "grad_norm": 13.885218620300293,
-      "learning_rate": 4.351515151515152e-05,
-      "loss": 6.3969,
+      "epoch": 0.3898916967509025,
+      "grad_norm": 11.214503288269043,
+      "learning_rate": 4.356197352587244e-05,
+      "loss": 6.47,
       "step": 108
     },
     {
-      "epoch": 0.39636363636363636,
-      "grad_norm": 9.663321495056152,
-      "learning_rate": 4.345454545454546e-05,
-      "loss": 6.2821,
+      "epoch": 0.3935018050541516,
+      "grad_norm": 11.047112464904785,
+      "learning_rate": 4.350180505415163e-05,
+      "loss": 6.3487,
       "step": 109
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 11.254612922668457,
-      "learning_rate": 4.3393939393939396e-05,
-      "loss": 6.2634,
+      "epoch": 0.3971119133574007,
+      "grad_norm": 10.862858772277832,
+      "learning_rate": 4.3441636582430804e-05,
+      "loss": 6.4521,
       "step": 110
     },
     {
-      "epoch": 0.4036363636363636,
-      "grad_norm": 30.378196716308594,
-      "learning_rate": 4.3333333333333334e-05,
-      "loss": 6.6234,
+      "epoch": 0.4007220216606498,
+      "grad_norm": 10.075233459472656,
+      "learning_rate": 4.338146811070999e-05,
+      "loss": 6.2256,
       "step": 111
     },
     {
-      "epoch": 0.4072727272727273,
-      "grad_norm": 16.96799087524414,
-      "learning_rate": 4.327272727272728e-05,
-      "loss": 6.3086,
+      "epoch": 0.4043321299638989,
+      "grad_norm": 8.835427284240723,
+      "learning_rate": 4.332129963898917e-05,
+      "loss": 6.252,
       "step": 112
     },
     {
-      "epoch": 0.4109090909090909,
-      "grad_norm": 8.183362007141113,
-      "learning_rate": 4.321212121212122e-05,
-      "loss": 6.1431,
+      "epoch": 0.40794223826714804,
+      "grad_norm": 8.498411178588867,
+      "learning_rate": 4.3261131167268356e-05,
+      "loss": 6.2111,
       "step": 113
     },
     {
-      "epoch": 0.41454545454545455,
-      "grad_norm": 9.962987899780273,
-      "learning_rate": 4.3151515151515156e-05,
-      "loss": 6.1345,
+      "epoch": 0.41155234657039713,
+      "grad_norm": 12.734055519104004,
+      "learning_rate": 4.320096269554753e-05,
+      "loss": 6.3314,
       "step": 114
     },
     {
-      "epoch": 0.41818181818181815,
-      "grad_norm": 15.649056434631348,
-      "learning_rate": 4.3090909090909094e-05,
-      "loss": 6.0388,
+      "epoch": 0.4151624548736462,
+      "grad_norm": 10.362188339233398,
+      "learning_rate": 4.314079422382672e-05,
+      "loss": 6.0026,
       "step": 115
     },
     {
-      "epoch": 0.4218181818181818,
-      "grad_norm": 11.750924110412598,
-      "learning_rate": 4.303030303030303e-05,
-      "loss": 6.1091,
+      "epoch": 0.4187725631768953,
+      "grad_norm": 8.478679656982422,
+      "learning_rate": 4.3080625752105894e-05,
+      "loss": 6.1003,
       "step": 116
     },
     {
-      "epoch": 0.4254545454545455,
-      "grad_norm": 11.011019706726074,
-      "learning_rate": 4.296969696969697e-05,
-      "loss": 5.9637,
+      "epoch": 0.4223826714801444,
+      "grad_norm": 9.426121711730957,
+      "learning_rate": 4.302045728038508e-05,
+      "loss": 5.9875,
       "step": 117
     },
     {
-      "epoch": 0.4290909090909091,
-      "grad_norm": 7.509008884429932,
-      "learning_rate": 4.290909090909091e-05,
-      "loss": 5.819,
+      "epoch": 0.4259927797833935,
+      "grad_norm": 7.937860012054443,
+      "learning_rate": 4.296028880866426e-05,
+      "loss": 5.9543,
       "step": 118
     },
     {
-      "epoch": 0.43272727272727274,
-      "grad_norm": 8.163488388061523,
-      "learning_rate": 4.284848484848485e-05,
-      "loss": 5.8944,
+      "epoch": 0.4296028880866426,
+      "grad_norm": 9.517463684082031,
+      "learning_rate": 4.2900120336943446e-05,
+      "loss": 5.9709,
       "step": 119
     },
     {
-      "epoch": 0.43636363636363634,
-      "grad_norm": 12.121551513671875,
-      "learning_rate": 4.2787878787878786e-05,
-      "loss": 5.8427,
+      "epoch": 0.4332129963898917,
+      "grad_norm": 5.840395450592041,
+      "learning_rate": 4.283995186522263e-05,
+      "loss": 5.8178,
       "step": 120
     },
     {
-      "epoch": 0.44,
-      "grad_norm": 10.191195487976074,
-      "learning_rate": 4.2727272727272724e-05,
-      "loss": 5.8268,
+      "epoch": 0.4368231046931408,
+      "grad_norm": 5.626274108886719,
+      "learning_rate": 4.277978339350181e-05,
+      "loss": 5.9067,
       "step": 121
     },
     {
-      "epoch": 0.44363636363636366,
-      "grad_norm": 6.44727087020874,
-      "learning_rate": 4.266666666666667e-05,
-      "loss": 5.8514,
+      "epoch": 0.4404332129963899,
+      "grad_norm": 4.778440952301025,
+      "learning_rate": 4.2719614921780984e-05,
+      "loss": 5.8001,
       "step": 122
     },
     {
-      "epoch": 0.44727272727272727,
-      "grad_norm": 5.603975296020508,
-      "learning_rate": 4.260606060606061e-05,
-      "loss": 5.7627,
+      "epoch": 0.44404332129963897,
+      "grad_norm": 8.221015930175781,
+      "learning_rate": 4.265944645006017e-05,
+      "loss": 6.0161,
       "step": 123
     },
     {
-      "epoch": 0.4509090909090909,
-      "grad_norm": 8.263181686401367,
-      "learning_rate": 4.254545454545455e-05,
-      "loss": 5.9339,
+      "epoch": 0.44765342960288806,
+      "grad_norm": 8.236863136291504,
+      "learning_rate": 4.259927797833935e-05,
+      "loss": 5.7349,
       "step": 124
     },
     {
-      "epoch": 0.45454545454545453,
-      "grad_norm": 9.458388328552246,
-      "learning_rate": 4.248484848484849e-05,
-      "loss": 5.623,
+      "epoch": 0.45126353790613716,
+      "grad_norm": 5.051707744598389,
+      "learning_rate": 4.2539109506618536e-05,
+      "loss": 5.8968,
       "step": 125
     },
     {
-      "epoch": 0.4581818181818182,
-      "grad_norm": 6.523182392120361,
-      "learning_rate": 4.242424242424243e-05,
-      "loss": 5.7246,
+      "epoch": 0.4548736462093863,
+      "grad_norm": 6.383734703063965,
+      "learning_rate": 4.247894103489772e-05,
+      "loss": 6.0475,
       "step": 126
     },
     {
-      "epoch": 0.4618181818181818,
-      "grad_norm": 4.975604057312012,
-      "learning_rate": 4.236363636363637e-05,
-      "loss": 5.7018,
+      "epoch": 0.4584837545126354,
+      "grad_norm": 8.41263484954834,
+      "learning_rate": 4.24187725631769e-05,
+      "loss": 5.7269,
       "step": 127
     },
     {
-      "epoch": 0.46545454545454545,
-      "grad_norm": 6.818673133850098,
-      "learning_rate": 4.2303030303030306e-05,
-      "loss": 5.9547,
+      "epoch": 0.4620938628158845,
+      "grad_norm": 7.515089511871338,
+      "learning_rate": 4.235860409145608e-05,
+      "loss": 5.6593,
       "step": 128
     },
     {
-      "epoch": 0.4690909090909091,
-      "grad_norm": 6.9742536544799805,
-      "learning_rate": 4.2242424242424244e-05,
-      "loss": 5.9003,
+      "epoch": 0.4657039711191336,
+      "grad_norm": 4.513574123382568,
+      "learning_rate": 4.229843561973526e-05,
+      "loss": 5.7707,
       "step": 129
     },
     {
-      "epoch": 0.4727272727272727,
-      "grad_norm": 3.8385305404663086,
-      "learning_rate": 4.218181818181818e-05,
-      "loss": 5.6588,
+      "epoch": 0.4693140794223827,
+      "grad_norm": 3.846007823944092,
+      "learning_rate": 4.223826714801444e-05,
+      "loss": 5.6836,
       "step": 130
     },
     {
-      "epoch": 0.4763636363636364,
-      "grad_norm": 6.935608863830566,
-      "learning_rate": 4.212121212121212e-05,
-      "loss": 5.7309,
+      "epoch": 0.4729241877256318,
+      "grad_norm": 4.019323825836182,
+      "learning_rate": 4.2178098676293626e-05,
+      "loss": 5.6789,
       "step": 131
     },
     {
-      "epoch": 0.48,
-      "grad_norm": 4.395711898803711,
-      "learning_rate": 4.206060606060606e-05,
-      "loss": 5.7787,
+      "epoch": 0.47653429602888087,
+      "grad_norm": 4.250916957855225,
+      "learning_rate": 4.211793020457281e-05,
+      "loss": 5.7052,
       "step": 132
     },
     {
-      "epoch": 0.48363636363636364,
-      "grad_norm": 6.357467174530029,
-      "learning_rate": 4.2e-05,
-      "loss": 5.5588,
+      "epoch": 0.48014440433212996,
+      "grad_norm": 3.1647579669952393,
+      "learning_rate": 4.205776173285199e-05,
+      "loss": 5.5919,
       "step": 133
     },
     {
-      "epoch": 0.48727272727272725,
-      "grad_norm": 8.066727638244629,
-      "learning_rate": 4.193939393939394e-05,
-      "loss": 5.5408,
+      "epoch": 0.48375451263537905,
+      "grad_norm": 3.269533157348633,
+      "learning_rate": 4.199759326113117e-05,
+      "loss": 5.7144,
       "step": 134
     },
     {
-      "epoch": 0.4909090909090909,
-      "grad_norm": 7.221047401428223,
-      "learning_rate": 4.187878787878788e-05,
-      "loss": 5.5937,
+      "epoch": 0.48736462093862815,
+      "grad_norm": 3.1203269958496094,
+      "learning_rate": 4.193742478941035e-05,
+      "loss": 5.6578,
       "step": 135
     },
     {
-      "epoch": 0.49454545454545457,
-      "grad_norm": 5.110551357269287,
-      "learning_rate": 4.181818181818182e-05,
-      "loss": 5.5971,
+      "epoch": 0.49097472924187724,
+      "grad_norm": 4.004848003387451,
+      "learning_rate": 4.187725631768953e-05,
+      "loss": 5.851,
       "step": 136
     },
     {
-      "epoch": 0.49818181818181817,
-      "grad_norm": 2.816383123397827,
-      "learning_rate": 4.1757575757575765e-05,
-      "loss": 5.563,
+      "epoch": 0.49458483754512633,
+      "grad_norm": 2.584901809692383,
+      "learning_rate": 4.1817087845968716e-05,
+      "loss": 5.5345,
       "step": 137
     },
     {
-      "epoch": 0.5018181818181818,
-      "grad_norm": 2.900914430618286,
-      "learning_rate": 4.16969696969697e-05,
-      "loss": 5.5362,
+      "epoch": 0.4981949458483754,
+      "grad_norm": 2.390103340148926,
+      "learning_rate": 4.175691937424789e-05,
+      "loss": 5.5518,
       "step": 138
     },
     {
-      "epoch": 0.5054545454545455,
-      "grad_norm": 4.6505560874938965,
-      "learning_rate": 4.163636363636364e-05,
-      "loss": 5.4366,
+      "epoch": 0.5018050541516246,
+      "grad_norm": 2.405592441558838,
+      "learning_rate": 4.169675090252708e-05,
+      "loss": 5.4639,
       "step": 139
     },
     {
-      "epoch": 0.509090909090909,
-      "grad_norm": 4.222754001617432,
-      "learning_rate": 4.157575757575758e-05,
-      "loss": 5.555,
+      "epoch": 0.5054151624548736,
+      "grad_norm": 2.404343605041504,
+      "learning_rate": 4.163658243080626e-05,
+      "loss": 5.3935,
       "step": 140
     },
     {
-      "epoch": 0.5127272727272727,
-      "grad_norm": 3.447328805923462,
-      "learning_rate": 4.151515151515152e-05,
-      "loss": 5.6128,
+      "epoch": 0.5090252707581228,
+      "grad_norm": 2.314335584640503,
+      "learning_rate": 4.1576413959085445e-05,
+      "loss": 5.4667,
       "step": 141
     },
     {
-      "epoch": 0.5163636363636364,
-      "grad_norm": 2.5096065998077393,
-      "learning_rate": 4.1454545454545456e-05,
-      "loss": 5.3437,
+      "epoch": 0.5126353790613718,
+      "grad_norm": 5.040092468261719,
+      "learning_rate": 4.151624548736462e-05,
+      "loss": 5.7844,
       "step": 142
     },
     {
-      "epoch": 0.52,
-      "grad_norm": 3.392411231994629,
-      "learning_rate": 4.1393939393939395e-05,
-      "loss": 5.4153,
+      "epoch": 0.516245487364621,
+      "grad_norm": 2.73350191116333,
+      "learning_rate": 4.1456077015643806e-05,
+      "loss": 5.581,
       "step": 143
     },
     {
-      "epoch": 0.5236363636363637,
-      "grad_norm": 3.3837954998016357,
-      "learning_rate": 4.133333333333333e-05,
-      "loss": 5.5665,
+      "epoch": 0.51985559566787,
+      "grad_norm": 3.540738821029663,
+      "learning_rate": 4.139590854392298e-05,
+      "loss": 5.433,
       "step": 144
     },
     {
-      "epoch": 0.5272727272727272,
-      "grad_norm": 3.1204018592834473,
-      "learning_rate": 4.127272727272727e-05,
-      "loss": 5.3836,
+      "epoch": 0.5234657039711191,
+      "grad_norm": 2.8726115226745605,
+      "learning_rate": 4.1335740072202167e-05,
+      "loss": 5.4571,
       "step": 145
     },
     {
-      "epoch": 0.5309090909090909,
-      "grad_norm": 2.5219132900238037,
-      "learning_rate": 4.1212121212121216e-05,
-      "loss": 5.5756,
+      "epoch": 0.5270758122743683,
+      "grad_norm": 2.460242509841919,
+      "learning_rate": 4.127557160048135e-05,
+      "loss": 5.3747,
       "step": 146
     },
     {
-      "epoch": 0.5345454545454545,
-      "grad_norm": 4.97948694229126,
-      "learning_rate": 4.1151515151515155e-05,
-      "loss": 5.583,
+      "epoch": 0.5306859205776173,
+      "grad_norm": 2.26070499420166,
+      "learning_rate": 4.1215403128760534e-05,
+      "loss": 5.337,
       "step": 147
     },
     {
-      "epoch": 0.5381818181818182,
-      "grad_norm": 4.947720050811768,
-      "learning_rate": 4.109090909090909e-05,
-      "loss": 5.4612,
+      "epoch": 0.5342960288808665,
+      "grad_norm": 2.353187084197998,
+      "learning_rate": 4.115523465703972e-05,
+      "loss": 5.3909,
       "step": 148
     },
     {
-      "epoch": 0.5418181818181819,
-      "grad_norm": 3.0988149642944336,
-      "learning_rate": 4.103030303030303e-05,
-      "loss": 5.4593,
+      "epoch": 0.5379061371841155,
+      "grad_norm": 3.6039485931396484,
+      "learning_rate": 4.1095066185318895e-05,
+      "loss": 5.4302,
       "step": 149
     },
     {
-      "epoch": 0.5454545454545454,
-      "grad_norm": 2.3757989406585693,
-      "learning_rate": 4.096969696969697e-05,
-      "loss": 5.3723,
+      "epoch": 0.5415162454873647,
+      "grad_norm": 2.220996856689453,
+      "learning_rate": 4.103489771359807e-05,
+      "loss": 5.3922,
       "step": 150
     },
     {
-      "epoch": 0.5490909090909091,
-      "grad_norm": 2.393914222717285,
-      "learning_rate": 4.0909090909090915e-05,
-      "loss": 5.4201,
+      "epoch": 0.5451263537906137,
+      "grad_norm": 2.450697660446167,
+      "learning_rate": 4.0974729241877256e-05,
+      "loss": 5.3605,
       "step": 151
     },
     {
-      "epoch": 0.5527272727272727,
-      "grad_norm": 4.960287570953369,
-      "learning_rate": 4.084848484848485e-05,
-      "loss": 5.5827,
+      "epoch": 0.5487364620938628,
+      "grad_norm": 2.709074020385742,
+      "learning_rate": 4.091456077015644e-05,
+      "loss": 5.433,
       "step": 152
     },
     {
-      "epoch": 0.5563636363636364,
-      "grad_norm": 3.0634284019470215,
-      "learning_rate": 4.078787878787879e-05,
-      "loss": 5.5706,
+      "epoch": 0.5523465703971119,
+      "grad_norm": 2.338242769241333,
+      "learning_rate": 4.0854392298435624e-05,
+      "loss": 5.3792,
       "step": 153
     },
     {
-      "epoch": 0.56,
-      "grad_norm": 2.428879499435425,
-      "learning_rate": 4.072727272727273e-05,
-      "loss": 5.4515,
+      "epoch": 0.555956678700361,
+      "grad_norm": 1.8731917142868042,
+      "learning_rate": 4.079422382671481e-05,
+      "loss": 5.3199,
       "step": 154
     },
     {
-      "epoch": 0.5636363636363636,
-      "grad_norm": 3.948068857192993,
-      "learning_rate": 4.066666666666667e-05,
-      "loss": 5.4016,
+      "epoch": 0.5595667870036101,
+      "grad_norm": 2.192985773086548,
+      "learning_rate": 4.0734055354993985e-05,
+      "loss": 5.2468,
       "step": 155
     },
     {
-      "epoch": 0.5672727272727273,
-      "grad_norm": 3.6819639205932617,
-      "learning_rate": 4.0606060606060606e-05,
-      "loss": 5.4725,
+      "epoch": 0.5631768953068592,
+      "grad_norm": 3.6917717456817627,
+      "learning_rate": 4.067388688327316e-05,
+      "loss": 5.3494,
       "step": 156
     },
     {
-      "epoch": 0.5709090909090909,
-      "grad_norm": 2.238288164138794,
-      "learning_rate": 4.0545454545454545e-05,
-      "loss": 5.3492,
+      "epoch": 0.5667870036101083,
+      "grad_norm": 4.642220497131348,
+      "learning_rate": 4.0613718411552346e-05,
+      "loss": 5.6593,
       "step": 157
     },
     {
-      "epoch": 0.5745454545454546,
-      "grad_norm": 2.360812187194824,
-      "learning_rate": 4.048484848484849e-05,
-      "loss": 5.379,
+      "epoch": 0.5703971119133574,
+      "grad_norm": 2.366976261138916,
+      "learning_rate": 4.055354993983153e-05,
+      "loss": 5.3336,
       "step": 158
     },
     {
-      "epoch": 0.5781818181818181,
-      "grad_norm": 2.7471134662628174,
-      "learning_rate": 4.042424242424243e-05,
-      "loss": 5.2598,
+      "epoch": 0.5740072202166066,
+      "grad_norm": 3.0058817863464355,
+      "learning_rate": 4.0493381468110714e-05,
+      "loss": 5.4665,
       "step": 159
     },
     {
-      "epoch": 0.5818181818181818,
-      "grad_norm": 2.7425880432128906,
-      "learning_rate": 4.0363636363636367e-05,
-      "loss": 5.2575,
+      "epoch": 0.5776173285198556,
+      "grad_norm": 2.297910690307617,
+      "learning_rate": 4.043321299638989e-05,
+      "loss": 5.232,
       "step": 160
     },
     {
-      "epoch": 0.5854545454545454,
-      "grad_norm": 2.0691795349121094,
-      "learning_rate": 4.0303030303030305e-05,
-      "loss": 5.4382,
+      "epoch": 0.5812274368231047,
+      "grad_norm": 1.971007227897644,
+      "learning_rate": 4.0373044524669075e-05,
+      "loss": 5.2868,
       "step": 161
     },
     {
-      "epoch": 0.5890909090909091,
-      "grad_norm": 2.2511022090911865,
-      "learning_rate": 4.024242424242424e-05,
-      "loss": 5.5356,
+      "epoch": 0.5848375451263538,
+      "grad_norm": 2.3164796829223633,
+      "learning_rate": 4.031287605294826e-05,
+      "loss": 5.42,
       "step": 162
     },
     {
-      "epoch": 0.5927272727272728,
-      "grad_norm": 2.0557987689971924,
-      "learning_rate": 4.018181818181818e-05,
-      "loss": 5.4373,
+      "epoch": 0.5884476534296029,
+      "grad_norm": 2.2059273719787598,
+      "learning_rate": 4.0252707581227436e-05,
+      "loss": 5.2749,
       "step": 163
     },
     {
-      "epoch": 0.5963636363636363,
-      "grad_norm": 2.4520606994628906,
-      "learning_rate": 4.012121212121212e-05,
-      "loss": 5.2067,
+      "epoch": 0.592057761732852,
+      "grad_norm": 2.0889828205108643,
+      "learning_rate": 4.019253910950662e-05,
+      "loss": 5.4344,
       "step": 164
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 1.9265565872192383,
-      "learning_rate": 4.0060606060606065e-05,
-      "loss": 5.3952,
+      "epoch": 0.5956678700361011,
+      "grad_norm": 2.5745677947998047,
+      "learning_rate": 4.0132370637785804e-05,
+      "loss": 5.5377,
       "step": 165
     },
     {
-      "epoch": 0.6036363636363636,
-      "grad_norm": 2.1164627075195312,
-      "learning_rate": 4e-05,
-      "loss": 5.134,
+      "epoch": 0.5992779783393501,
+      "grad_norm": 2.0397369861602783,
+      "learning_rate": 4.007220216606498e-05,
+      "loss": 5.2009,
       "step": 166
     },
     {
-      "epoch": 0.6072727272727273,
-      "grad_norm": 1.9591832160949707,
-      "learning_rate": 3.993939393939394e-05,
-      "loss": 5.2561,
+      "epoch": 0.6028880866425993,
+      "grad_norm": 2.9067304134368896,
+      "learning_rate": 4.0012033694344165e-05,
+      "loss": 5.2047,
       "step": 167
     },
     {
-      "epoch": 0.610909090909091,
-      "grad_norm": 1.9654712677001953,
-      "learning_rate": 3.987878787878788e-05,
-      "loss": 5.1538,
+      "epoch": 0.6064981949458483,
+      "grad_norm": 3.246255874633789,
+      "learning_rate": 3.995186522262335e-05,
+      "loss": 5.5057,
       "step": 168
     },
     {
-      "epoch": 0.6145454545454545,
-      "grad_norm": 1.8879733085632324,
-      "learning_rate": 3.981818181818182e-05,
-      "loss": 5.2923,
+      "epoch": 0.6101083032490975,
+      "grad_norm": 1.773790717124939,
+      "learning_rate": 3.989169675090253e-05,
+      "loss": 5.2973,
       "step": 169
     },
     {
-      "epoch": 0.6181818181818182,
-      "grad_norm": 2.7771313190460205,
-      "learning_rate": 3.975757575757576e-05,
-      "loss": 5.3083,
+      "epoch": 0.6137184115523465,
+      "grad_norm": 2.192772150039673,
+      "learning_rate": 3.983152827918171e-05,
+      "loss": 5.1046,
       "step": 170
     },
     {
-      "epoch": 0.6218181818181818,
-      "grad_norm": 2.2295308113098145,
-      "learning_rate": 3.96969696969697e-05,
-      "loss": 5.2658,
+      "epoch": 0.6173285198555957,
+      "grad_norm": 1.923920750617981,
+      "learning_rate": 3.9771359807460894e-05,
+      "loss": 5.2984,
       "step": 171
     },
     {
-      "epoch": 0.6254545454545455,
-      "grad_norm": 2.3644328117370605,
-      "learning_rate": 3.963636363636364e-05,
-      "loss": 5.1096,
+      "epoch": 0.6209386281588448,
+      "grad_norm": 2.1909148693084717,
+      "learning_rate": 3.971119133574007e-05,
+      "loss": 5.222,
       "step": 172
     },
     {
-      "epoch": 0.6290909090909091,
-      "grad_norm": 1.9340869188308716,
-      "learning_rate": 3.957575757575758e-05,
-      "loss": 5.1372,
+      "epoch": 0.6245487364620939,
+      "grad_norm": 1.6805565357208252,
+      "learning_rate": 3.9651022864019255e-05,
+      "loss": 5.1115,
       "step": 173
     },
     {
-      "epoch": 0.6327272727272727,
-      "grad_norm": 2.5413575172424316,
-      "learning_rate": 3.951515151515152e-05,
-      "loss": 5.085,
+      "epoch": 0.628158844765343,
+      "grad_norm": 1.7837809324264526,
+      "learning_rate": 3.959085439229844e-05,
+      "loss": 5.1649,
       "step": 174
     },
     {
-      "epoch": 0.6363636363636364,
-      "grad_norm": 2.119511842727661,
-      "learning_rate": 3.9454545454545455e-05,
-      "loss": 5.2124,
+      "epoch": 0.631768953068592,
+      "grad_norm": 1.7250087261199951,
+      "learning_rate": 3.953068592057762e-05,
+      "loss": 5.0921,
       "step": 175
     },
     {
-      "epoch": 0.64,
-      "grad_norm": 2.2437052726745605,
-      "learning_rate": 3.939393939393939e-05,
-      "loss": 5.1622,
+      "epoch": 0.6353790613718412,
+      "grad_norm": 1.9569655656814575,
+      "learning_rate": 3.94705174488568e-05,
+      "loss": 5.2318,
       "step": 176
     },
     {
-      "epoch": 0.6436363636363637,
-      "grad_norm": 4.928871154785156,
-      "learning_rate": 3.933333333333333e-05,
-      "loss": 5.1084,
+      "epoch": 0.6389891696750902,
+      "grad_norm": 1.8813940286636353,
+      "learning_rate": 3.941034897713598e-05,
+      "loss": 5.2228,
       "step": 177
     },
     {
-      "epoch": 0.6472727272727272,
-      "grad_norm": 4.648951053619385,
-      "learning_rate": 3.927272727272727e-05,
-      "loss": 5.36,
+      "epoch": 0.6425992779783394,
+      "grad_norm": 3.922642469406128,
+      "learning_rate": 3.935018050541516e-05,
+      "loss": 5.4068,
       "step": 178
     },
     {
-      "epoch": 0.6509090909090909,
-      "grad_norm": 2.124182939529419,
-      "learning_rate": 3.9212121212121215e-05,
-      "loss": 5.1618,
+      "epoch": 0.6462093862815884,
+      "grad_norm": 2.448709726333618,
+      "learning_rate": 3.9290012033694345e-05,
+      "loss": 5.2182,
       "step": 179
     },
     {
-      "epoch": 0.6545454545454545,
-      "grad_norm": 1.795525074005127,
-      "learning_rate": 3.9151515151515153e-05,
-      "loss": 5.1547,
+      "epoch": 0.6498194945848376,
+      "grad_norm": 2.0298540592193604,
+      "learning_rate": 3.922984356197353e-05,
+      "loss": 5.0192,
       "step": 180
     },
     {
-      "epoch": 0.6581818181818182,
-      "grad_norm": 7.342172145843506,
-      "learning_rate": 3.909090909090909e-05,
-      "loss": 5.1927,
+      "epoch": 0.6534296028880866,
+      "grad_norm": 2.1303651332855225,
+      "learning_rate": 3.916967509025271e-05,
+      "loss": 5.1934,
       "step": 181
     },
     {
-      "epoch": 0.6618181818181819,
-      "grad_norm": 7.812733173370361,
-      "learning_rate": 3.903030303030304e-05,
-      "loss": 5.0418,
+      "epoch": 0.6570397111913358,
+      "grad_norm": 1.6486132144927979,
+      "learning_rate": 3.91095066185319e-05,
+      "loss": 5.1027,
       "step": 182
     },
     {
-      "epoch": 0.6654545454545454,
-      "grad_norm": 7.93719482421875,
-      "learning_rate": 3.8969696969696975e-05,
-      "loss": 5.2858,
+      "epoch": 0.6606498194945848,
+      "grad_norm": 3.1995131969451904,
+      "learning_rate": 3.9049338146811074e-05,
+      "loss": 5.2469,
       "step": 183
     },
     {
-      "epoch": 0.6690909090909091,
-      "grad_norm": 2.118412971496582,
-      "learning_rate": 3.8909090909090914e-05,
-      "loss": 5.0045,
+      "epoch": 0.6642599277978339,
+      "grad_norm": 2.314117193222046,
+      "learning_rate": 3.898916967509025e-05,
+      "loss": 4.9994,
       "step": 184
     },
     {
-      "epoch": 0.6727272727272727,
-      "grad_norm": 2.4741883277893066,
-      "learning_rate": 3.884848484848485e-05,
-      "loss": 5.0554,
+      "epoch": 0.6678700361010831,
+      "grad_norm": 2.438479423522949,
+      "learning_rate": 3.8929001203369435e-05,
+      "loss": 5.1398,
       "step": 185
     },
     {
-      "epoch": 0.6763636363636364,
-      "grad_norm": 2.7295148372650146,
-      "learning_rate": 3.878787878787879e-05,
-      "loss": 5.0661,
+      "epoch": 0.6714801444043321,
+      "grad_norm": 2.5420024394989014,
+      "learning_rate": 3.886883273164862e-05,
+      "loss": 4.9167,
       "step": 186
     },
     {
-      "epoch": 0.68,
-      "grad_norm": 3.038841962814331,
-      "learning_rate": 3.872727272727273e-05,
-      "loss": 5.2243,
+      "epoch": 0.6750902527075813,
+      "grad_norm": 2.7453882694244385,
+      "learning_rate": 3.88086642599278e-05,
+      "loss": 5.1795,
       "step": 187
     },
     {
-      "epoch": 0.6836363636363636,
-      "grad_norm": 1.7481746673583984,
-      "learning_rate": 3.866666666666667e-05,
-      "loss": 5.1118,
+      "epoch": 0.6787003610108303,
+      "grad_norm": 2.1120471954345703,
+      "learning_rate": 3.874849578820698e-05,
+      "loss": 5.0016,
       "step": 188
     },
     {
-      "epoch": 0.6872727272727273,
-      "grad_norm": 1.9238699674606323,
-      "learning_rate": 3.8606060606060605e-05,
-      "loss": 5.0864,
+      "epoch": 0.6823104693140795,
+      "grad_norm": 1.9928067922592163,
+      "learning_rate": 3.8688327316486164e-05,
+      "loss": 5.0273,
       "step": 189
     },
     {
-      "epoch": 0.6909090909090909,
-      "grad_norm": 2.7689208984375,
-      "learning_rate": 3.8545454545454544e-05,
-      "loss": 5.165,
+      "epoch": 0.6859205776173285,
+      "grad_norm": 2.736645221710205,
+      "learning_rate": 3.862815884476535e-05,
+      "loss": 5.1239,
       "step": 190
     },
     {
-      "epoch": 0.6945454545454546,
-      "grad_norm": 1.9809805154800415,
-      "learning_rate": 3.848484848484848e-05,
-      "loss": 5.2885,
+      "epoch": 0.6895306859205776,
+      "grad_norm": 2.823556423187256,
+      "learning_rate": 3.8567990373044525e-05,
+      "loss": 5.1206,
       "step": 191
     },
     {
-      "epoch": 0.6981818181818182,
-      "grad_norm": 1.7771217823028564,
-      "learning_rate": 3.842424242424243e-05,
-      "loss": 5.0592,
+      "epoch": 0.6931407942238267,
+      "grad_norm": 2.5614190101623535,
+      "learning_rate": 3.850782190132371e-05,
+      "loss": 5.0496,
       "step": 192
     },
     {
-      "epoch": 0.7018181818181818,
-      "grad_norm": 2.0928571224212646,
-      "learning_rate": 3.8363636363636365e-05,
-      "loss": 5.0556,
+      "epoch": 0.6967509025270758,
+      "grad_norm": 3.6118035316467285,
+      "learning_rate": 3.844765342960289e-05,
+      "loss": 5.0111,
       "step": 193
     },
     {
-      "epoch": 0.7054545454545454,
-      "grad_norm": 1.9713197946548462,
-      "learning_rate": 3.830303030303031e-05,
-      "loss": 4.9863,
+      "epoch": 0.7003610108303249,
+      "grad_norm": 2.387235164642334,
+      "learning_rate": 3.838748495788207e-05,
+      "loss": 5.1377,
       "step": 194
     },
     {
-      "epoch": 0.7090909090909091,
-      "grad_norm": 1.7148576974868774,
-      "learning_rate": 3.824242424242425e-05,
-      "loss": 5.1573,
+      "epoch": 0.703971119133574,
+      "grad_norm": 2.7525899410247803,
+      "learning_rate": 3.8327316486161254e-05,
+      "loss": 4.9603,
       "step": 195
     },
     {
-      "epoch": 0.7127272727272728,
-      "grad_norm": 1.9550209045410156,
-      "learning_rate": 3.818181818181819e-05,
-      "loss": 5.1499,
+      "epoch": 0.7075812274368231,
+      "grad_norm": 2.7961931228637695,
+      "learning_rate": 3.826714801444044e-05,
+      "loss": 5.279,
       "step": 196
     },
     {
-      "epoch": 0.7163636363636363,
-      "grad_norm": 1.695175290107727,
-      "learning_rate": 3.8121212121212125e-05,
-      "loss": 5.0989,
+      "epoch": 0.7111913357400722,
+      "grad_norm": 2.0542795658111572,
+      "learning_rate": 3.8206979542719615e-05,
+      "loss": 4.8995,
       "step": 197
     },
     {
-      "epoch": 0.72,
-      "grad_norm": 3.0943408012390137,
-      "learning_rate": 3.8060606060606064e-05,
-      "loss": 5.0336,
+      "epoch": 0.7148014440433214,
+      "grad_norm": 3.907233953475952,
+      "learning_rate": 3.81468110709988e-05,
+      "loss": 4.9418,
       "step": 198
     },
     {
-      "epoch": 0.7236363636363636,
-      "grad_norm": 2.219883918762207,
-      "learning_rate": 3.8e-05,
-      "loss": 4.942,
+      "epoch": 0.7184115523465704,
+      "grad_norm": 4.625024795532227,
+      "learning_rate": 3.8086642599277976e-05,
+      "loss": 5.0614,
       "step": 199
     },
     {
-      "epoch": 0.7272727272727273,
-      "grad_norm": 3.4300808906555176,
-      "learning_rate": 3.793939393939394e-05,
-      "loss": 4.9841,
+      "epoch": 0.7220216606498195,
+      "grad_norm": 1.8027046918869019,
+      "learning_rate": 3.802647412755716e-05,
+      "loss": 4.8633,
       "step": 200
     },
     {
-      "epoch": 0.730909090909091,
-      "grad_norm": 2.2358174324035645,
-      "learning_rate": 3.787878787878788e-05,
-      "loss": 5.0942,
+      "epoch": 0.7256317689530686,
+      "grad_norm": 2.0051488876342773,
+      "learning_rate": 3.7966305655836343e-05,
+      "loss": 4.8928,
       "step": 201
     },
     {
-      "epoch": 0.7345454545454545,
-      "grad_norm": 2.1182949542999268,
-      "learning_rate": 3.781818181818182e-05,
-      "loss": 4.9778,
+      "epoch": 0.7292418772563177,
+      "grad_norm": 3.243194818496704,
+      "learning_rate": 3.790613718411553e-05,
+      "loss": 5.0681,
       "step": 202
     },
     {
-      "epoch": 0.7381818181818182,
-      "grad_norm": 2.5967414379119873,
-      "learning_rate": 3.7757575757575755e-05,
-      "loss": 4.8359,
+      "epoch": 0.7328519855595668,
+      "grad_norm": 2.727696418762207,
+      "learning_rate": 3.784596871239471e-05,
+      "loss": 5.1281,
       "step": 203
     },
     {
-      "epoch": 0.7418181818181818,
-      "grad_norm": 2.3327507972717285,
-      "learning_rate": 3.76969696969697e-05,
-      "loss": 5.0698,
+      "epoch": 0.7364620938628159,
+      "grad_norm": 1.5806025266647339,
+      "learning_rate": 3.778580024067389e-05,
+      "loss": 4.9068,
       "step": 204
     },
     {
-      "epoch": 0.7454545454545455,
-      "grad_norm": 2.0424389839172363,
-      "learning_rate": 3.763636363636364e-05,
-      "loss": 5.0554,
+      "epoch": 0.740072202166065,
+      "grad_norm": 4.9806318283081055,
+      "learning_rate": 3.7725631768953066e-05,
+      "loss": 5.0823,
       "step": 205
     },
     {
-      "epoch": 0.7490909090909091,
-      "grad_norm": 2.24621844291687,
-      "learning_rate": 3.757575757575758e-05,
-      "loss": 5.0766,
+      "epoch": 0.7436823104693141,
+      "grad_norm": 2.2012317180633545,
+      "learning_rate": 3.766546329723225e-05,
+      "loss": 4.8072,
       "step": 206
     },
     {
-      "epoch": 0.7527272727272727,
-      "grad_norm": 1.9795176982879639,
-      "learning_rate": 3.7515151515151516e-05,
-      "loss": 4.8173,
+      "epoch": 0.7472924187725631,
+      "grad_norm": 1.9060529470443726,
+      "learning_rate": 3.760529482551143e-05,
+      "loss": 4.9973,
       "step": 207
     },
     {
-      "epoch": 0.7563636363636363,
-      "grad_norm": 2.3668367862701416,
-      "learning_rate": 3.745454545454546e-05,
-      "loss": 4.8863,
+      "epoch": 0.7509025270758123,
+      "grad_norm": 1.9107259511947632,
+      "learning_rate": 3.754512635379062e-05,
+      "loss": 4.9668,
       "step": 208
     },
     {
-      "epoch": 0.76,
-      "grad_norm": 2.0429444313049316,
-      "learning_rate": 3.73939393939394e-05,
-      "loss": 4.8721,
+      "epoch": 0.7545126353790613,
+      "grad_norm": 1.9527552127838135,
+      "learning_rate": 3.74849578820698e-05,
+      "loss": 4.9307,
       "step": 209
     },
     {
-      "epoch": 0.7636363636363637,
-      "grad_norm": 2.5353472232818604,
-      "learning_rate": 3.733333333333334e-05,
-      "loss": 5.0184,
+      "epoch": 0.7581227436823105,
+      "grad_norm": 2.9125237464904785,
+      "learning_rate": 3.742478941034898e-05,
+      "loss": 4.8515,
       "step": 210
     },
     {
-      "epoch": 0.7672727272727272,
-      "grad_norm": 1.7212400436401367,
-      "learning_rate": 3.7272727272727276e-05,
-      "loss": 4.8016,
+      "epoch": 0.7617328519855595,
+      "grad_norm": 2.114863395690918,
+      "learning_rate": 3.7364620938628155e-05,
+      "loss": 4.9533,
       "step": 211
     },
     {
-      "epoch": 0.7709090909090909,
-      "grad_norm": 4.483168125152588,
-      "learning_rate": 3.7212121212121214e-05,
-      "loss": 4.9398,
+      "epoch": 0.7653429602888087,
+      "grad_norm": 3.5130767822265625,
+      "learning_rate": 3.730445246690734e-05,
+      "loss": 5.1674,
       "step": 212
     },
     {
-      "epoch": 0.7745454545454545,
-      "grad_norm": 2.3681108951568604,
-      "learning_rate": 3.715151515151515e-05,
-      "loss": 4.895,
+      "epoch": 0.7689530685920578,
+      "grad_norm": 3.569817304611206,
+      "learning_rate": 3.724428399518652e-05,
+      "loss": 4.78,
       "step": 213
     },
     {
-      "epoch": 0.7781818181818182,
-      "grad_norm": 3.49525785446167,
-      "learning_rate": 3.709090909090909e-05,
-      "loss": 4.8547,
+      "epoch": 0.7725631768953068,
+      "grad_norm": 3.456483840942383,
+      "learning_rate": 3.718411552346571e-05,
+      "loss": 4.6653,
       "step": 214
     },
     {
-      "epoch": 0.7818181818181819,
-      "grad_norm": 4.825103282928467,
-      "learning_rate": 3.703030303030303e-05,
-      "loss": 4.7677,
+      "epoch": 0.776173285198556,
+      "grad_norm": 2.1834640502929688,
+      "learning_rate": 3.712394705174489e-05,
+      "loss": 4.9685,
       "step": 215
     },
     {
-      "epoch": 0.7854545454545454,
-      "grad_norm": 3.3975582122802734,
-      "learning_rate": 3.6969696969696974e-05,
-      "loss": 5.0948,
+      "epoch": 0.779783393501805,
+      "grad_norm": 3.3436832427978516,
+      "learning_rate": 3.706377858002407e-05,
+      "loss": 4.9632,
       "step": 216
     },
     {
-      "epoch": 0.7890909090909091,
-      "grad_norm": 2.115818738937378,
-      "learning_rate": 3.690909090909091e-05,
-      "loss": 4.8962,
+      "epoch": 0.7833935018050542,
+      "grad_norm": 3.473731517791748,
+      "learning_rate": 3.700361010830325e-05,
+      "loss": 4.8625,
       "step": 217
     },
     {
-      "epoch": 0.7927272727272727,
-      "grad_norm": 3.3818588256835938,
-      "learning_rate": 3.684848484848485e-05,
-      "loss": 5.0175,
+      "epoch": 0.7870036101083032,
+      "grad_norm": 4.904114246368408,
+      "learning_rate": 3.694344163658243e-05,
+      "loss": 5.1025,
       "step": 218
     },
     {
-      "epoch": 0.7963636363636364,
-      "grad_norm": 2.1706228256225586,
-      "learning_rate": 3.678787878787879e-05,
-      "loss": 4.9794,
+      "epoch": 0.7906137184115524,
+      "grad_norm": 2.1981472969055176,
+      "learning_rate": 3.688327316486161e-05,
+      "loss": 4.7881,
       "step": 219
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 1.8316903114318848,
-      "learning_rate": 3.672727272727273e-05,
-      "loss": 5.0241,
+      "epoch": 0.7942238267148014,
+      "grad_norm": 2.3624002933502197,
+      "learning_rate": 3.68231046931408e-05,
+      "loss": 5.0446,
       "step": 220
     },
     {
-      "epoch": 0.8036363636363636,
-      "grad_norm": 2.90144944190979,
-      "learning_rate": 3.6666666666666666e-05,
-      "loss": 4.7593,
+      "epoch": 0.7978339350180506,
+      "grad_norm": 4.91404390335083,
+      "learning_rate": 3.676293622141998e-05,
+      "loss": 4.7575,
       "step": 221
     },
     {
-      "epoch": 0.8072727272727273,
-      "grad_norm": 3.2332253456115723,
-      "learning_rate": 3.660606060606061e-05,
-      "loss": 4.8894,
+      "epoch": 0.8014440433212996,
+      "grad_norm": 2.7553951740264893,
+      "learning_rate": 3.670276774969916e-05,
+      "loss": 5.0162,
       "step": 222
     },
     {
-      "epoch": 0.8109090909090909,
-      "grad_norm": 2.5061445236206055,
-      "learning_rate": 3.654545454545455e-05,
-      "loss": 4.7913,
+      "epoch": 0.8050541516245487,
+      "grad_norm": 2.2965950965881348,
+      "learning_rate": 3.664259927797834e-05,
+      "loss": 4.7732,
       "step": 223
     },
     {
-      "epoch": 0.8145454545454546,
-      "grad_norm": 1.689694881439209,
-      "learning_rate": 3.648484848484849e-05,
-      "loss": 4.8167,
+      "epoch": 0.8086642599277978,
+      "grad_norm": 4.340023040771484,
+      "learning_rate": 3.6582430806257526e-05,
+      "loss": 4.6869,
       "step": 224
     },
     {
-      "epoch": 0.8181818181818182,
-      "grad_norm": 2.99468994140625,
-      "learning_rate": 3.6424242424242426e-05,
-      "loss": 5.0467,
+      "epoch": 0.8122743682310469,
+      "grad_norm": 4.003706932067871,
+      "learning_rate": 3.65222623345367e-05,
+      "loss": 4.7567,
       "step": 225
     },
     {
-      "epoch": 0.8218181818181818,
-      "grad_norm": 1.5727870464324951,
-      "learning_rate": 3.6363636363636364e-05,
-      "loss": 4.8491,
+      "epoch": 0.8158844765342961,
+      "grad_norm": 1.9011428356170654,
+      "learning_rate": 3.646209386281589e-05,
+      "loss": 4.889,
       "step": 226
     },
     {
-      "epoch": 0.8254545454545454,
-      "grad_norm": 2.4910600185394287,
-      "learning_rate": 3.63030303030303e-05,
-      "loss": 4.7987,
+      "epoch": 0.8194945848375451,
+      "grad_norm": 4.650227069854736,
+      "learning_rate": 3.6401925391095064e-05,
+      "loss": 4.8348,
       "step": 227
     },
     {
-      "epoch": 0.8290909090909091,
-      "grad_norm": 2.5647313594818115,
-      "learning_rate": 3.624242424242425e-05,
-      "loss": 4.9892,
+      "epoch": 0.8231046931407943,
+      "grad_norm": 3.565837860107422,
+      "learning_rate": 3.634175691937425e-05,
+      "loss": 4.9042,
       "step": 228
     },
     {
-      "epoch": 0.8327272727272728,
-      "grad_norm": 2.091198205947876,
-      "learning_rate": 3.6181818181818186e-05,
-      "loss": 4.7671,
+      "epoch": 0.8267148014440433,
+      "grad_norm": 2.047588586807251,
+      "learning_rate": 3.628158844765343e-05,
+      "loss": 4.8082,
       "step": 229
     },
     {
-      "epoch": 0.8363636363636363,
-      "grad_norm": 2.494631290435791,
-      "learning_rate": 3.6121212121212124e-05,
-      "loss": 5.0782,
+      "epoch": 0.8303249097472925,
+      "grad_norm": 3.465941905975342,
+      "learning_rate": 3.6221419975932616e-05,
+      "loss": 4.5621,
       "step": 230
     },
     {
-      "epoch": 0.84,
-      "grad_norm": 6.3343000411987305,
-      "learning_rate": 3.606060606060606e-05,
-      "loss": 4.92,
+      "epoch": 0.8339350180505415,
+      "grad_norm": 2.2322192192077637,
+      "learning_rate": 3.616125150421179e-05,
+      "loss": 4.7779,
       "step": 231
     },
     {
-      "epoch": 0.8436363636363636,
-      "grad_norm": 2.6743531227111816,
-      "learning_rate": 3.6e-05,
-      "loss": 4.7505,
+      "epoch": 0.8375451263537906,
+      "grad_norm": 3.8442447185516357,
+      "learning_rate": 3.610108303249098e-05,
+      "loss": 4.7807,
       "step": 232
     },
     {
-      "epoch": 0.8472727272727273,
-      "grad_norm": 2.4769270420074463,
-      "learning_rate": 3.593939393939394e-05,
-      "loss": 4.6388,
+      "epoch": 0.8411552346570397,
+      "grad_norm": 3.0969555377960205,
+      "learning_rate": 3.6040914560770154e-05,
+      "loss": 4.7886,
       "step": 233
     },
     {
-      "epoch": 0.850909090909091,
-      "grad_norm": 1.9747099876403809,
-      "learning_rate": 3.587878787878788e-05,
-      "loss": 4.9499,
+      "epoch": 0.8447653429602888,
+      "grad_norm": 1.886405348777771,
+      "learning_rate": 3.598074608904934e-05,
+      "loss": 4.7991,
       "step": 234
     },
     {
-      "epoch": 0.8545454545454545,
-      "grad_norm": 3.063316822052002,
-      "learning_rate": 3.5818181818181816e-05,
-      "loss": 5.1008,
+      "epoch": 0.8483754512635379,
+      "grad_norm": 4.009521961212158,
+      "learning_rate": 3.592057761732852e-05,
+      "loss": 5.0554,
       "step": 235
     },
     {
-      "epoch": 0.8581818181818182,
-      "grad_norm": 1.9509135484695435,
-      "learning_rate": 3.575757575757576e-05,
-      "loss": 5.0562,
+      "epoch": 0.851985559566787,
+      "grad_norm": 3.61405086517334,
+      "learning_rate": 3.5860409145607706e-05,
+      "loss": 4.6868,
       "step": 236
     },
     {
-      "epoch": 0.8618181818181818,
-      "grad_norm": 2.897395372390747,
-      "learning_rate": 3.56969696969697e-05,
-      "loss": 4.9302,
+      "epoch": 0.855595667870036,
+      "grad_norm": 2.804062604904175,
+      "learning_rate": 3.580024067388689e-05,
+      "loss": 4.7725,
       "step": 237
     },
     {
-      "epoch": 0.8654545454545455,
-      "grad_norm": 2.996319055557251,
-      "learning_rate": 3.563636363636364e-05,
-      "loss": 4.8001,
+      "epoch": 0.8592057761732852,
+      "grad_norm": 2.4844532012939453,
+      "learning_rate": 3.574007220216607e-05,
+      "loss": 5.0849,
       "step": 238
     },
     {
-      "epoch": 0.8690909090909091,
-      "grad_norm": 2.526507616043091,
-      "learning_rate": 3.5575757575757576e-05,
-      "loss": 4.8748,
+      "epoch": 0.8628158844765343,
+      "grad_norm": 4.291991710662842,
+      "learning_rate": 3.5679903730445244e-05,
+      "loss": 4.954,
       "step": 239
     },
     {
-      "epoch": 0.8727272727272727,
-      "grad_norm": 3.4848058223724365,
-      "learning_rate": 3.551515151515152e-05,
-      "loss": 4.707,
+      "epoch": 0.8664259927797834,
+      "grad_norm": 7.508516788482666,
+      "learning_rate": 3.561973525872443e-05,
+      "loss": 4.6777,
       "step": 240
     },
     {
-      "epoch": 0.8763636363636363,
-      "grad_norm": 1.839698076248169,
-      "learning_rate": 3.545454545454546e-05,
-      "loss": 4.7075,
+      "epoch": 0.8700361010830325,
+      "grad_norm": 3.0717663764953613,
+      "learning_rate": 3.555956678700361e-05,
+      "loss": 4.9312,
       "step": 241
     },
     {
-      "epoch": 0.88,
-      "grad_norm": 2.332421064376831,
-      "learning_rate": 3.53939393939394e-05,
-      "loss": 4.8433,
+      "epoch": 0.8736462093862816,
+      "grad_norm": 2.553285837173462,
+      "learning_rate": 3.5499398315282796e-05,
+      "loss": 4.753,
       "step": 242
     },
     {
-      "epoch": 0.8836363636363637,
-      "grad_norm": 2.877429485321045,
-      "learning_rate": 3.5333333333333336e-05,
-      "loss": 4.6706,
+      "epoch": 0.8772563176895307,
+      "grad_norm": 3.743809461593628,
+      "learning_rate": 3.543922984356198e-05,
+      "loss": 4.8038,
       "step": 243
     },
     {
-      "epoch": 0.8872727272727273,
-      "grad_norm": 4.581256866455078,
-      "learning_rate": 3.5272727272727274e-05,
-      "loss": 4.7334,
+      "epoch": 0.8808664259927798,
+      "grad_norm": 2.7291336059570312,
+      "learning_rate": 3.537906137184116e-05,
+      "loss": 4.7678,
       "step": 244
     },
     {
-      "epoch": 0.8909090909090909,
-      "grad_norm": 2.1481051445007324,
-      "learning_rate": 3.521212121212121e-05,
-      "loss": 4.7936,
+      "epoch": 0.8844765342960289,
+      "grad_norm": 2.182262420654297,
+      "learning_rate": 3.531889290012034e-05,
+      "loss": 4.7459,
       "step": 245
     },
     {
-      "epoch": 0.8945454545454545,
-      "grad_norm": 3.5877878665924072,
-      "learning_rate": 3.515151515151515e-05,
-      "loss": 4.9524,
+      "epoch": 0.8880866425992779,
+      "grad_norm": 2.953906297683716,
+      "learning_rate": 3.525872442839952e-05,
+      "loss": 4.762,
       "step": 246
     },
     {
-      "epoch": 0.8981818181818182,
-      "grad_norm": 2.7592244148254395,
-      "learning_rate": 3.509090909090909e-05,
-      "loss": 4.6404,
+      "epoch": 0.8916967509025271,
+      "grad_norm": 5.394567012786865,
+      "learning_rate": 3.51985559566787e-05,
+      "loss": 4.7177,
       "step": 247
     },
     {
-      "epoch": 0.9018181818181819,
-      "grad_norm": 3.172548532485962,
-      "learning_rate": 3.503030303030303e-05,
-      "loss": 4.9291,
+      "epoch": 0.8953068592057761,
+      "grad_norm": 2.277156352996826,
+      "learning_rate": 3.5138387484957885e-05,
+      "loss": 4.6974,
       "step": 248
     },
     {
-      "epoch": 0.9054545454545454,
-      "grad_norm": 2.158982515335083,
-      "learning_rate": 3.4969696969696966e-05,
-      "loss": 4.6853,
+      "epoch": 0.8989169675090253,
+      "grad_norm": 2.9844322204589844,
+      "learning_rate": 3.507821901323707e-05,
+      "loss": 4.8035,
       "step": 249
     },
     {
-      "epoch": 0.9090909090909091,
-      "grad_norm": 2.6028811931610107,
-      "learning_rate": 3.490909090909091e-05,
-      "loss": 4.5646,
+      "epoch": 0.9025270758122743,
+      "grad_norm": 3.6745762825012207,
+      "learning_rate": 3.5018050541516247e-05,
+      "loss": 4.6441,
       "step": 250
     },
     {
-      "epoch": 0.9127272727272727,
-      "grad_norm": 5.963568210601807,
-      "learning_rate": 3.484848484848485e-05,
-      "loss": 4.5695,
+      "epoch": 0.9061371841155235,
+      "grad_norm": 2.259817600250244,
+      "learning_rate": 3.495788206979543e-05,
+      "loss": 4.8853,
       "step": 251
     },
     {
-      "epoch": 0.9163636363636364,
-      "grad_norm": 6.336101055145264,
-      "learning_rate": 3.4787878787878795e-05,
-      "loss": 5.0168,
+      "epoch": 0.9097472924187726,
+      "grad_norm": 2.207033634185791,
+      "learning_rate": 3.489771359807461e-05,
+      "loss": 4.7622,
       "step": 252
     },
     {
-      "epoch": 0.92,
-      "grad_norm": 3.318384885787964,
-      "learning_rate": 3.472727272727273e-05,
-      "loss": 4.7873,
+      "epoch": 0.9133574007220217,
+      "grad_norm": 1.8033169507980347,
+      "learning_rate": 3.483754512635379e-05,
+      "loss": 4.6108,
       "step": 253
     },
     {
-      "epoch": 0.9236363636363636,
-      "grad_norm": 3.579904079437256,
-      "learning_rate": 3.466666666666667e-05,
-      "loss": 4.7128,
+      "epoch": 0.9169675090252708,
+      "grad_norm": 2.7010810375213623,
+      "learning_rate": 3.4777376654632975e-05,
+      "loss": 4.8258,
       "step": 254
     },
     {
-      "epoch": 0.9272727272727272,
-      "grad_norm": 3.7936673164367676,
-      "learning_rate": 3.460606060606061e-05,
-      "loss": 4.7375,
+      "epoch": 0.9205776173285198,
+      "grad_norm": 2.7489020824432373,
+      "learning_rate": 3.471720818291215e-05,
+      "loss": 4.6758,
       "step": 255
     },
     {
-      "epoch": 0.9309090909090909,
-      "grad_norm": 3.784524440765381,
-      "learning_rate": 3.454545454545455e-05,
-      "loss": 4.7575,
+      "epoch": 0.924187725631769,
+      "grad_norm": 3.4038209915161133,
+      "learning_rate": 3.4657039711191336e-05,
+      "loss": 4.596,
       "step": 256
     },
     {
-      "epoch": 0.9345454545454546,
-      "grad_norm": 2.5060670375823975,
-      "learning_rate": 3.4484848484848486e-05,
-      "loss": 4.7153,
+      "epoch": 0.927797833935018,
+      "grad_norm": 3.105119228363037,
+      "learning_rate": 3.459687123947052e-05,
+      "loss": 4.883,
       "step": 257
     },
     {
-      "epoch": 0.9381818181818182,
-      "grad_norm": 2.5724377632141113,
-      "learning_rate": 3.4424242424242425e-05,
-      "loss": 4.8637,
+      "epoch": 0.9314079422382672,
+      "grad_norm": 6.055860996246338,
+      "learning_rate": 3.4536702767749704e-05,
+      "loss": 4.6512,
       "step": 258
     },
     {
-      "epoch": 0.9418181818181818,
-      "grad_norm": 3.9936795234680176,
-      "learning_rate": 3.436363636363636e-05,
-      "loss": 4.9075,
+      "epoch": 0.9350180505415162,
+      "grad_norm": 2.115447521209717,
+      "learning_rate": 3.447653429602888e-05,
+      "loss": 4.5428,
       "step": 259
     },
     {
-      "epoch": 0.9454545454545454,
-      "grad_norm": 1.9491581916809082,
-      "learning_rate": 3.43030303030303e-05,
-      "loss": 4.8538,
+      "epoch": 0.9386281588447654,
+      "grad_norm": 6.520136833190918,
+      "learning_rate": 3.4416365824308065e-05,
+      "loss": 4.9671,
       "step": 260
     },
     {
-      "epoch": 0.9490909090909091,
-      "grad_norm": 1.8183727264404297,
-      "learning_rate": 3.424242424242424e-05,
-      "loss": 4.7095,
+      "epoch": 0.9422382671480144,
+      "grad_norm": 5.740654945373535,
+      "learning_rate": 3.435619735258724e-05,
+      "loss": 4.9137,
       "step": 261
     },
     {
-      "epoch": 0.9527272727272728,
-      "grad_norm": 2.8100154399871826,
-      "learning_rate": 3.4181818181818185e-05,
-      "loss": 4.7456,
+      "epoch": 0.9458483754512635,
+      "grad_norm": 2.9576783180236816,
+      "learning_rate": 3.4296028880866426e-05,
+      "loss": 4.8369,
       "step": 262
     },
     {
-      "epoch": 0.9563636363636364,
-      "grad_norm": 2.25604248046875,
-      "learning_rate": 3.412121212121212e-05,
-      "loss": 4.5881,
+      "epoch": 0.9494584837545126,
+      "grad_norm": 2.0835630893707275,
+      "learning_rate": 3.423586040914561e-05,
+      "loss": 4.6395,
       "step": 263
     },
     {
-      "epoch": 0.96,
-      "grad_norm": 3.7915985584259033,
-      "learning_rate": 3.406060606060606e-05,
-      "loss": 4.6032,
+      "epoch": 0.9530685920577617,
+      "grad_norm": 3.2203140258789062,
+      "learning_rate": 3.4175691937424794e-05,
+      "loss": 4.7584,
       "step": 264
     },
     {
-      "epoch": 0.9636363636363636,
-      "grad_norm": 4.647451400756836,
-      "learning_rate": 3.4000000000000007e-05,
-      "loss": 4.9148,
+      "epoch": 0.9566787003610109,
+      "grad_norm": 3.108306646347046,
+      "learning_rate": 3.411552346570397e-05,
+      "loss": 4.8148,
       "step": 265
     },
     {
-      "epoch": 0.9672727272727273,
-      "grad_norm": 4.613765239715576,
-      "learning_rate": 3.3939393939393945e-05,
-      "loss": 4.6888,
+      "epoch": 0.9602888086642599,
+      "grad_norm": 4.311677932739258,
+      "learning_rate": 3.4055354993983155e-05,
+      "loss": 5.0121,
       "step": 266
     },
     {
-      "epoch": 0.9709090909090909,
-      "grad_norm": 2.291818857192993,
-      "learning_rate": 3.387878787878788e-05,
-      "loss": 4.8835,
+      "epoch": 0.9638989169675091,
+      "grad_norm": 2.4522058963775635,
+      "learning_rate": 3.399518652226233e-05,
+      "loss": 4.6426,
       "step": 267
     },
     {
-      "epoch": 0.9745454545454545,
-      "grad_norm": 2.959836483001709,
-      "learning_rate": 3.381818181818182e-05,
-      "loss": 4.781,
+      "epoch": 0.9675090252707581,
+      "grad_norm": 6.1543426513671875,
+      "learning_rate": 3.3935018050541516e-05,
+      "loss": 4.7737,
       "step": 268
     },
     {
-      "epoch": 0.9781818181818182,
-      "grad_norm": 2.353412389755249,
-      "learning_rate": 3.375757575757576e-05,
-      "loss": 4.6777,
+      "epoch": 0.9711191335740073,
+      "grad_norm": 6.448587417602539,
+      "learning_rate": 3.38748495788207e-05,
+      "loss": 4.9805,
       "step": 269
     },
     {
-      "epoch": 0.9818181818181818,
-      "grad_norm": 2.019970178604126,
-      "learning_rate": 3.36969696969697e-05,
-      "loss": 4.6844,
+      "epoch": 0.9747292418772563,
+      "grad_norm": 3.016731023788452,
+      "learning_rate": 3.3814681107099884e-05,
+      "loss": 4.701,
       "step": 270
     },
     {
-      "epoch": 0.9854545454545455,
-      "grad_norm": 3.597709894180298,
-      "learning_rate": 3.3636363636363636e-05,
-      "loss": 4.8324,
+      "epoch": 0.9783393501805054,
+      "grad_norm": 4.361660003662109,
+      "learning_rate": 3.375451263537907e-05,
+      "loss": 4.5368,
       "step": 271
     },
     {
-      "epoch": 0.9890909090909091,
-      "grad_norm": 1.8851228952407837,
-      "learning_rate": 3.3575757575757575e-05,
-      "loss": 4.7522,
+      "epoch": 0.9819494584837545,
+      "grad_norm": 17.387479782104492,
+      "learning_rate": 3.3694344163658245e-05,
+      "loss": 4.6907,
       "step": 272
     },
     {
-      "epoch": 0.9927272727272727,
-      "grad_norm": 2.674484968185425,
-      "learning_rate": 3.351515151515151e-05,
-      "loss": 4.6662,
+      "epoch": 0.9855595667870036,
+      "grad_norm": 3.6502737998962402,
+      "learning_rate": 3.363417569193742e-05,
+      "loss": 4.6492,
       "step": 273
     },
     {
-      "epoch": 0.9963636363636363,
-      "grad_norm": 4.3509840965271,
-      "learning_rate": 3.345454545454546e-05,
-      "loss": 4.7466,
+      "epoch": 0.9891696750902527,
+      "grad_norm": 3.673269510269165,
+      "learning_rate": 3.3574007220216606e-05,
+      "loss": 4.8245,
       "step": 274
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 3.465963363647461,
-      "learning_rate": 3.3393939393939397e-05,
-      "loss": 4.9775,
+      "epoch": 0.9927797833935018,
+      "grad_norm": 4.932320594787598,
+      "learning_rate": 3.351383874849579e-05,
+      "loss": 4.6652,
       "step": 275
     },
     {
-      "epoch": 1.0036363636363637,
-      "grad_norm": 3.604304313659668,
-      "learning_rate": 3.3333333333333335e-05,
-      "loss": 4.6624,
+      "epoch": 0.9963898916967509,
+      "grad_norm": 7.902830600738525,
+      "learning_rate": 3.3453670276774974e-05,
+      "loss": 4.6427,
       "step": 276
     },
     {
-      "epoch": 1.0072727272727273,
-      "grad_norm": 3.1077980995178223,
-      "learning_rate": 3.327272727272727e-05,
-      "loss": 4.5934,
+      "epoch": 1.0,
+      "grad_norm": 5.016904354095459,
+      "learning_rate": 3.339350180505416e-05,
+      "loss": 4.6971,
       "step": 277
     },
     {
-      "epoch": 1.010909090909091,
-      "grad_norm": 2.4643919467926025,
-      "learning_rate": 3.321212121212121e-05,
-      "loss": 4.7402,
+      "epoch": 1.0036101083032491,
+      "grad_norm": 4.204914569854736,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 4.6376,
       "step": 278
     },
     {
-      "epoch": 1.0145454545454546,
-      "grad_norm": 2.7680652141571045,
-      "learning_rate": 3.315151515151516e-05,
-      "loss": 4.8989,
+      "epoch": 1.0072202166064983,
+      "grad_norm": 2.7867376804351807,
+      "learning_rate": 3.327316486161252e-05,
+      "loss": 4.6176,
       "step": 279
     },
     {
-      "epoch": 1.018181818181818,
-      "grad_norm": 2.3318872451782227,
-      "learning_rate": 3.3090909090909095e-05,
-      "loss": 4.7498,
+      "epoch": 1.0108303249097472,
+      "grad_norm": 9.475534439086914,
+      "learning_rate": 3.3212996389891696e-05,
+      "loss": 4.8113,
       "step": 280
     },
     {
-      "epoch": 1.0218181818181817,
-      "grad_norm": 2.1118509769439697,
-      "learning_rate": 3.303030303030303e-05,
-      "loss": 4.5331,
+      "epoch": 1.0144404332129964,
+      "grad_norm": 9.185867309570312,
+      "learning_rate": 3.315282791817088e-05,
+      "loss": 4.8703,
       "step": 281
     },
     {
-      "epoch": 1.0254545454545454,
-      "grad_norm": 3.2147281169891357,
-      "learning_rate": 3.296969696969697e-05,
-      "loss": 4.8356,
+      "epoch": 1.0180505415162455,
+      "grad_norm": 3.52699875831604,
+      "learning_rate": 3.3092659446450064e-05,
+      "loss": 4.8145,
       "step": 282
     },
     {
-      "epoch": 1.029090909090909,
-      "grad_norm": 2.104762315750122,
-      "learning_rate": 3.290909090909091e-05,
-      "loss": 4.6182,
+      "epoch": 1.0216606498194947,
+      "grad_norm": 3.920712947845459,
+      "learning_rate": 3.303249097472924e-05,
+      "loss": 4.6849,
       "step": 283
     },
     {
-      "epoch": 1.0327272727272727,
-      "grad_norm": 2.818783760070801,
-      "learning_rate": 3.284848484848485e-05,
-      "loss": 4.8216,
+      "epoch": 1.0252707581227436,
+      "grad_norm": 4.423035144805908,
+      "learning_rate": 3.2972322503008425e-05,
+      "loss": 4.5551,
       "step": 284
     },
     {
-      "epoch": 1.0363636363636364,
-      "grad_norm": 2.765343427658081,
-      "learning_rate": 3.278787878787879e-05,
-      "loss": 4.8029,
+      "epoch": 1.0288808664259927,
+      "grad_norm": 3.3100874423980713,
+      "learning_rate": 3.291215403128761e-05,
+      "loss": 4.8097,
       "step": 285
     },
     {
-      "epoch": 1.04,
-      "grad_norm": 1.492432713508606,
-      "learning_rate": 3.272727272727273e-05,
-      "loss": 4.7011,
+      "epoch": 1.032490974729242,
+      "grad_norm": 6.985757350921631,
+      "learning_rate": 3.2851985559566786e-05,
+      "loss": 4.7497,
       "step": 286
     },
     {
-      "epoch": 1.0436363636363637,
-      "grad_norm": 2.7354354858398438,
-      "learning_rate": 3.266666666666667e-05,
-      "loss": 4.5074,
+      "epoch": 1.036101083032491,
+      "grad_norm": 5.6586222648620605,
+      "learning_rate": 3.279181708784597e-05,
+      "loss": 4.7276,
       "step": 287
     },
     {
-      "epoch": 1.0472727272727274,
-      "grad_norm": 3.9838192462921143,
-      "learning_rate": 3.260606060606061e-05,
-      "loss": 4.7455,
+      "epoch": 1.03971119133574,
+      "grad_norm": 3.2179510593414307,
+      "learning_rate": 3.2731648616125154e-05,
+      "loss": 4.6136,
       "step": 288
     },
     {
-      "epoch": 1.050909090909091,
-      "grad_norm": 3.0941343307495117,
-      "learning_rate": 3.254545454545455e-05,
-      "loss": 4.5761,
+      "epoch": 1.0433212996389891,
+      "grad_norm": 3.0860912799835205,
+      "learning_rate": 3.267148014440433e-05,
+      "loss": 4.7697,
       "step": 289
     },
     {
-      "epoch": 1.0545454545454545,
-      "grad_norm": 2.4687376022338867,
-      "learning_rate": 3.2484848484848485e-05,
-      "loss": 4.5997,
+      "epoch": 1.0469314079422383,
+      "grad_norm": 2.0248677730560303,
+      "learning_rate": 3.2611311672683515e-05,
+      "loss": 4.8089,
       "step": 290
     },
     {
-      "epoch": 1.0581818181818181,
-      "grad_norm": 3.4283993244171143,
-      "learning_rate": 3.2424242424242423e-05,
-      "loss": 4.6079,
+      "epoch": 1.0505415162454874,
+      "grad_norm": 3.7353522777557373,
+      "learning_rate": 3.25511432009627e-05,
+      "loss": 4.6567,
       "step": 291
     },
     {
-      "epoch": 1.0618181818181818,
-      "grad_norm": 2.7953782081604004,
-      "learning_rate": 3.236363636363636e-05,
-      "loss": 4.6912,
+      "epoch": 1.0541516245487366,
+      "grad_norm": 2.9157674312591553,
+      "learning_rate": 3.249097472924188e-05,
+      "loss": 4.5186,
       "step": 292
     },
     {
-      "epoch": 1.0654545454545454,
-      "grad_norm": 2.2899701595306396,
-      "learning_rate": 3.230303030303031e-05,
-      "loss": 4.6359,
+      "epoch": 1.0577617328519855,
+      "grad_norm": 3.010720729827881,
+      "learning_rate": 3.243080625752106e-05,
+      "loss": 4.9587,
       "step": 293
     },
     {
-      "epoch": 1.069090909090909,
-      "grad_norm": 3.88470721244812,
-      "learning_rate": 3.2242424242424245e-05,
-      "loss": 4.7514,
+      "epoch": 1.0613718411552346,
+      "grad_norm": 2.433742046356201,
+      "learning_rate": 3.237063778580024e-05,
+      "loss": 4.5569,
       "step": 294
     },
     {
-      "epoch": 1.0727272727272728,
-      "grad_norm": 2.5786900520324707,
-      "learning_rate": 3.2181818181818184e-05,
-      "loss": 4.5935,
+      "epoch": 1.0649819494584838,
+      "grad_norm": 2.5127952098846436,
+      "learning_rate": 3.231046931407942e-05,
+      "loss": 4.6183,
       "step": 295
     },
     {
-      "epoch": 1.0763636363636364,
-      "grad_norm": 2.5727601051330566,
-      "learning_rate": 3.212121212121212e-05,
-      "loss": 4.6504,
+      "epoch": 1.068592057761733,
+      "grad_norm": 2.83199143409729,
+      "learning_rate": 3.2250300842358605e-05,
+      "loss": 4.4713,
       "step": 296
     },
     {
-      "epoch": 1.08,
-      "grad_norm": 2.643254280090332,
-      "learning_rate": 3.206060606060606e-05,
-      "loss": 4.7769,
+      "epoch": 1.0722021660649819,
+      "grad_norm": 3.9529902935028076,
+      "learning_rate": 3.219013237063779e-05,
+      "loss": 4.7303,
       "step": 297
     },
     {
-      "epoch": 1.0836363636363637,
-      "grad_norm": 2.8821048736572266,
-      "learning_rate": 3.2000000000000005e-05,
-      "loss": 4.7027,
+      "epoch": 1.075812274368231,
+      "grad_norm": 3.1393637657165527,
+      "learning_rate": 3.212996389891697e-05,
+      "loss": 4.8081,
       "step": 298
     },
     {
-      "epoch": 1.0872727272727274,
-      "grad_norm": 4.047679424285889,
-      "learning_rate": 3.1939393939393944e-05,
-      "loss": 4.8341,
+      "epoch": 1.0794223826714802,
+      "grad_norm": 2.7979230880737305,
+      "learning_rate": 3.2069795427196156e-05,
+      "loss": 4.6623,
       "step": 299
     },
     {
-      "epoch": 1.0909090909090908,
-      "grad_norm": 3.6683619022369385,
-      "learning_rate": 3.187878787878788e-05,
-      "loss": 4.7663,
+      "epoch": 1.0830324909747293,
+      "grad_norm": 5.9944024085998535,
+      "learning_rate": 3.2009626955475333e-05,
+      "loss": 4.7301,
       "step": 300
     },
     {
-      "epoch": 1.0945454545454545,
-      "grad_norm": 5.787856578826904,
-      "learning_rate": 3.181818181818182e-05,
-      "loss": 4.697,
+      "epoch": 1.0866425992779782,
+      "grad_norm": 4.5418901443481445,
+      "learning_rate": 3.194945848375451e-05,
+      "loss": 4.5683,
       "step": 301
     },
     {
-      "epoch": 1.0981818181818181,
-      "grad_norm": 7.743192672729492,
-      "learning_rate": 3.175757575757576e-05,
-      "loss": 4.7046,
+      "epoch": 1.0902527075812274,
+      "grad_norm": 6.022171497344971,
+      "learning_rate": 3.1889290012033695e-05,
+      "loss": 4.8128,
       "step": 302
     },
     {
-      "epoch": 1.1018181818181818,
-      "grad_norm": 3.313638925552368,
-      "learning_rate": 3.16969696969697e-05,
-      "loss": 4.6653,
+      "epoch": 1.0938628158844765,
+      "grad_norm": 4.565722465515137,
+      "learning_rate": 3.182912154031288e-05,
+      "loss": 4.5517,
       "step": 303
     },
     {
-      "epoch": 1.1054545454545455,
-      "grad_norm": 4.846319198608398,
-      "learning_rate": 3.1636363636363635e-05,
-      "loss": 4.6066,
+      "epoch": 1.0974729241877257,
+      "grad_norm": 5.111222267150879,
+      "learning_rate": 3.176895306859206e-05,
+      "loss": 4.5883,
       "step": 304
     },
     {
-      "epoch": 1.1090909090909091,
-      "grad_norm": 4.69194221496582,
-      "learning_rate": 3.1575757575757574e-05,
-      "loss": 4.7325,
+      "epoch": 1.1010830324909748,
+      "grad_norm": 3.97930645942688,
+      "learning_rate": 3.170878459687124e-05,
+      "loss": 4.7255,
       "step": 305
     },
     {
-      "epoch": 1.1127272727272728,
-      "grad_norm": 3.353827476501465,
-      "learning_rate": 3.151515151515151e-05,
-      "loss": 4.6744,
+      "epoch": 1.1046931407942238,
+      "grad_norm": 3.147520065307617,
+      "learning_rate": 3.164861612515042e-05,
+      "loss": 4.5757,
       "step": 306
     },
     {
-      "epoch": 1.1163636363636364,
-      "grad_norm": 3.623960256576538,
-      "learning_rate": 3.145454545454546e-05,
-      "loss": 4.8799,
+      "epoch": 1.108303249097473,
+      "grad_norm": 5.289313793182373,
+      "learning_rate": 3.15884476534296e-05,
+      "loss": 4.7255,
       "step": 307
     },
     {
-      "epoch": 1.12,
-      "grad_norm": 4.136066436767578,
-      "learning_rate": 3.1393939393939395e-05,
-      "loss": 4.715,
+      "epoch": 1.111913357400722,
+      "grad_norm": 4.879770278930664,
+      "learning_rate": 3.1528279181708784e-05,
+      "loss": 4.9087,
       "step": 308
     },
     {
-      "epoch": 1.1236363636363635,
-      "grad_norm": 4.029481410980225,
-      "learning_rate": 3.1333333333333334e-05,
-      "loss": 4.6084,
+      "epoch": 1.1155234657039712,
+      "grad_norm": 3.533722162246704,
+      "learning_rate": 3.146811070998797e-05,
+      "loss": 4.4174,
       "step": 309
     },
     {
-      "epoch": 1.1272727272727272,
-      "grad_norm": 4.471688270568848,
-      "learning_rate": 3.127272727272728e-05,
-      "loss": 4.5595,
+      "epoch": 1.1191335740072201,
+      "grad_norm": 2.7764930725097656,
+      "learning_rate": 3.140794223826715e-05,
+      "loss": 4.634,
       "step": 310
     },
     {
-      "epoch": 1.1309090909090909,
-      "grad_norm": 2.3349573612213135,
-      "learning_rate": 3.121212121212122e-05,
-      "loss": 4.5042,
+      "epoch": 1.1227436823104693,
+      "grad_norm": 3.450363874435425,
+      "learning_rate": 3.134777376654633e-05,
+      "loss": 4.5882,
       "step": 311
     },
     {
-      "epoch": 1.1345454545454545,
-      "grad_norm": 3.2584612369537354,
-      "learning_rate": 3.1151515151515156e-05,
-      "loss": 4.5667,
+      "epoch": 1.1263537906137184,
+      "grad_norm": 2.2029852867126465,
+      "learning_rate": 3.128760529482551e-05,
+      "loss": 4.6923,
       "step": 312
     },
     {
-      "epoch": 1.1381818181818182,
-      "grad_norm": 4.187669277191162,
-      "learning_rate": 3.1090909090909094e-05,
-      "loss": 4.591,
+      "epoch": 1.1299638989169676,
+      "grad_norm": 13.167424201965332,
+      "learning_rate": 3.12274368231047e-05,
+      "loss": 4.6935,
       "step": 313
     },
     {
-      "epoch": 1.1418181818181818,
-      "grad_norm": 2.7688863277435303,
-      "learning_rate": 3.103030303030303e-05,
-      "loss": 4.6028,
+      "epoch": 1.1335740072202167,
+      "grad_norm": 7.503760814666748,
+      "learning_rate": 3.1167268351383874e-05,
+      "loss": 4.6496,
       "step": 314
     },
     {
-      "epoch": 1.1454545454545455,
-      "grad_norm": 2.8685548305511475,
-      "learning_rate": 3.096969696969697e-05,
-      "loss": 4.8637,
+      "epoch": 1.1371841155234657,
+      "grad_norm": 4.865889549255371,
+      "learning_rate": 3.110709987966306e-05,
+      "loss": 4.5744,
       "step": 315
     },
     {
-      "epoch": 1.1490909090909092,
-      "grad_norm": 3.441349506378174,
-      "learning_rate": 3.090909090909091e-05,
-      "loss": 4.3877,
+      "epoch": 1.1407942238267148,
+      "grad_norm": 3.753988742828369,
+      "learning_rate": 3.104693140794224e-05,
+      "loss": 4.534,
       "step": 316
     },
     {
-      "epoch": 1.1527272727272728,
-      "grad_norm": 2.61682391166687,
-      "learning_rate": 3.084848484848485e-05,
-      "loss": 4.5628,
+      "epoch": 1.144404332129964,
+      "grad_norm": 4.719662666320801,
+      "learning_rate": 3.098676293622142e-05,
+      "loss": 4.5527,
       "step": 317
     },
     {
-      "epoch": 1.1563636363636363,
-      "grad_norm": 2.719127655029297,
-      "learning_rate": 3.0787878787878786e-05,
-      "loss": 4.6051,
+      "epoch": 1.1480144404332129,
+      "grad_norm": 4.489358425140381,
+      "learning_rate": 3.09265944645006e-05,
+      "loss": 4.6453,
       "step": 318
     },
     {
-      "epoch": 1.16,
-      "grad_norm": 3.737633466720581,
-      "learning_rate": 3.0727272727272724e-05,
-      "loss": 4.4883,
+      "epoch": 1.151624548736462,
+      "grad_norm": 2.6681582927703857,
+      "learning_rate": 3.086642599277979e-05,
+      "loss": 4.5538,
       "step": 319
     },
     {
-      "epoch": 1.1636363636363636,
-      "grad_norm": 2.2933197021484375,
-      "learning_rate": 3.066666666666667e-05,
-      "loss": 4.5372,
+      "epoch": 1.1552346570397112,
+      "grad_norm": 2.0293869972229004,
+      "learning_rate": 3.080625752105897e-05,
+      "loss": 4.5606,
       "step": 320
     },
     {
-      "epoch": 1.1672727272727272,
-      "grad_norm": 2.3944780826568604,
-      "learning_rate": 3.060606060606061e-05,
-      "loss": 4.6886,
+      "epoch": 1.1588447653429603,
+      "grad_norm": 15.385830879211426,
+      "learning_rate": 3.074608904933815e-05,
+      "loss": 4.8805,
       "step": 321
     },
     {
-      "epoch": 1.170909090909091,
-      "grad_norm": 2.4027483463287354,
-      "learning_rate": 3.054545454545455e-05,
-      "loss": 4.5962,
+      "epoch": 1.1624548736462095,
+      "grad_norm": 17.79530143737793,
+      "learning_rate": 3.0685920577617325e-05,
+      "loss": 4.5318,
       "step": 322
     },
     {
-      "epoch": 1.1745454545454546,
-      "grad_norm": 2.160446882247925,
-      "learning_rate": 3.0484848484848487e-05,
-      "loss": 4.7377,
+      "epoch": 1.1660649819494584,
+      "grad_norm": 9.239445686340332,
+      "learning_rate": 3.062575210589651e-05,
+      "loss": 4.5567,
       "step": 323
     },
     {
-      "epoch": 1.1781818181818182,
-      "grad_norm": 2.3969006538391113,
-      "learning_rate": 3.0424242424242426e-05,
-      "loss": 4.5241,
+      "epoch": 1.1696750902527075,
+      "grad_norm": 4.8387322425842285,
+      "learning_rate": 3.056558363417569e-05,
+      "loss": 4.6446,
       "step": 324
     },
     {
-      "epoch": 1.1818181818181819,
-      "grad_norm": 1.7160532474517822,
-      "learning_rate": 3.0363636363636367e-05,
-      "loss": 4.5891,
+      "epoch": 1.1732851985559567,
+      "grad_norm": 5.810051441192627,
+      "learning_rate": 3.0505415162454877e-05,
+      "loss": 4.5906,
       "step": 325
     },
     {
-      "epoch": 1.1854545454545455,
-      "grad_norm": 2.991136312484741,
-      "learning_rate": 3.0303030303030306e-05,
-      "loss": 4.5969,
+      "epoch": 1.1768953068592058,
+      "grad_norm": 6.55676794052124,
+      "learning_rate": 3.0445246690734057e-05,
+      "loss": 4.6755,
       "step": 326
     },
     {
-      "epoch": 1.189090909090909,
-      "grad_norm": 2.5131070613861084,
-      "learning_rate": 3.0242424242424244e-05,
-      "loss": 4.5446,
+      "epoch": 1.1805054151624548,
+      "grad_norm": 4.300464153289795,
+      "learning_rate": 3.038507821901324e-05,
+      "loss": 4.6347,
       "step": 327
     },
     {
-      "epoch": 1.1927272727272726,
-      "grad_norm": 2.6630470752716064,
-      "learning_rate": 3.0181818181818182e-05,
-      "loss": 4.7089,
+      "epoch": 1.184115523465704,
+      "grad_norm": 4.3174519538879395,
+      "learning_rate": 3.032490974729242e-05,
+      "loss": 4.7603,
       "step": 328
     },
     {
-      "epoch": 1.1963636363636363,
-      "grad_norm": 4.355704307556152,
-      "learning_rate": 3.012121212121212e-05,
-      "loss": 4.4316,
+      "epoch": 1.187725631768953,
+      "grad_norm": 5.036055564880371,
+      "learning_rate": 3.0264741275571602e-05,
+      "loss": 4.7223,
       "step": 329
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 2.5602145195007324,
-      "learning_rate": 3.0060606060606062e-05,
-      "loss": 4.5123,
+      "epoch": 1.1913357400722022,
+      "grad_norm": 4.322726726531982,
+      "learning_rate": 3.0204572803850783e-05,
+      "loss": 4.6241,
       "step": 330
     },
     {
-      "epoch": 1.2036363636363636,
-      "grad_norm": 2.802565574645996,
-      "learning_rate": 3e-05,
-      "loss": 4.4404,
+      "epoch": 1.1949458483754514,
+      "grad_norm": 3.5698063373565674,
+      "learning_rate": 3.0144404332129967e-05,
+      "loss": 4.6906,
       "step": 331
     },
     {
-      "epoch": 1.2072727272727273,
-      "grad_norm": 1.707558274269104,
-      "learning_rate": 2.993939393939394e-05,
-      "loss": 4.5435,
+      "epoch": 1.1985559566787003,
+      "grad_norm": 4.076971530914307,
+      "learning_rate": 3.0084235860409147e-05,
+      "loss": 4.5428,
       "step": 332
     },
     {
-      "epoch": 1.210909090909091,
-      "grad_norm": 3.6090481281280518,
-      "learning_rate": 2.9878787878787877e-05,
-      "loss": 4.5652,
+      "epoch": 1.2021660649819494,
+      "grad_norm": 4.875138759613037,
+      "learning_rate": 3.0024067388688324e-05,
+      "loss": 4.7874,
       "step": 333
     },
     {
-      "epoch": 1.2145454545454546,
-      "grad_norm": 2.4650418758392334,
-      "learning_rate": 2.9818181818181816e-05,
-      "loss": 4.5316,
+      "epoch": 1.2057761732851986,
+      "grad_norm": 3.3777639865875244,
+      "learning_rate": 2.996389891696751e-05,
+      "loss": 4.4553,
       "step": 334
     },
     {
-      "epoch": 1.2181818181818183,
-      "grad_norm": 4.170000076293945,
-      "learning_rate": 2.9757575757575757e-05,
-      "loss": 4.5409,
+      "epoch": 1.2093862815884477,
+      "grad_norm": 8.022859573364258,
+      "learning_rate": 2.9903730445246692e-05,
+      "loss": 4.5464,
       "step": 335
     },
     {
-      "epoch": 1.221818181818182,
-      "grad_norm": 3.2201592922210693,
-      "learning_rate": 2.96969696969697e-05,
-      "loss": 4.5562,
+      "epoch": 1.2129963898916967,
+      "grad_norm": 4.7307047843933105,
+      "learning_rate": 2.9843561973525873e-05,
+      "loss": 4.4503,
       "step": 336
     },
     {
-      "epoch": 1.2254545454545456,
-      "grad_norm": 2.6030731201171875,
-      "learning_rate": 2.963636363636364e-05,
-      "loss": 4.5289,
+      "epoch": 1.2166064981949458,
+      "grad_norm": 2.6915078163146973,
+      "learning_rate": 2.9783393501805057e-05,
+      "loss": 4.76,
       "step": 337
     },
     {
-      "epoch": 1.229090909090909,
-      "grad_norm": 5.5631513595581055,
-      "learning_rate": 2.957575757575758e-05,
-      "loss": 4.5547,
+      "epoch": 1.220216606498195,
+      "grad_norm": 4.561704158782959,
+      "learning_rate": 2.972322503008424e-05,
+      "loss": 4.8931,
       "step": 338
     },
     {
-      "epoch": 1.2327272727272727,
-      "grad_norm": 5.190867900848389,
-      "learning_rate": 2.9515151515151518e-05,
-      "loss": 4.4717,
+      "epoch": 1.2238267148014441,
+      "grad_norm": 3.5802175998687744,
+      "learning_rate": 2.9663056558363418e-05,
+      "loss": 4.7736,
       "step": 339
     },
     {
-      "epoch": 1.2363636363636363,
-      "grad_norm": 4.258591651916504,
-      "learning_rate": 2.9454545454545456e-05,
-      "loss": 4.5684,
+      "epoch": 1.2274368231046933,
+      "grad_norm": 2.2672626972198486,
+      "learning_rate": 2.9602888086642598e-05,
+      "loss": 4.6361,
       "step": 340
     },
     {
-      "epoch": 1.24,
-      "grad_norm": 6.548985004425049,
-      "learning_rate": 2.9393939393939394e-05,
-      "loss": 4.4463,
+      "epoch": 1.2310469314079422,
+      "grad_norm": 3.247804641723633,
+      "learning_rate": 2.9542719614921782e-05,
+      "loss": 4.5714,
       "step": 341
     },
     {
-      "epoch": 1.2436363636363637,
-      "grad_norm": 2.304765462875366,
-      "learning_rate": 2.9333333333333336e-05,
-      "loss": 4.514,
+      "epoch": 1.2346570397111913,
+      "grad_norm": 2.9589133262634277,
+      "learning_rate": 2.9482551143200966e-05,
+      "loss": 4.6753,
       "step": 342
     },
     {
-      "epoch": 1.2472727272727273,
-      "grad_norm": 4.920679569244385,
-      "learning_rate": 2.9272727272727274e-05,
-      "loss": 4.6245,
+      "epoch": 1.2382671480144405,
+      "grad_norm": 3.426271677017212,
+      "learning_rate": 2.9422382671480147e-05,
+      "loss": 4.5655,
       "step": 343
     },
     {
-      "epoch": 1.250909090909091,
-      "grad_norm": 4.364790916442871,
-      "learning_rate": 2.9212121212121213e-05,
-      "loss": 4.377,
+      "epoch": 1.2418772563176894,
+      "grad_norm": 3.625859498977661,
+      "learning_rate": 2.9362214199759324e-05,
+      "loss": 4.5216,
       "step": 344
     },
     {
-      "epoch": 1.2545454545454544,
-      "grad_norm": 2.6548376083374023,
-      "learning_rate": 2.915151515151515e-05,
-      "loss": 4.673,
+      "epoch": 1.2454873646209386,
+      "grad_norm": 2.466961145401001,
+      "learning_rate": 2.9302045728038508e-05,
+      "loss": 4.544,
       "step": 345
     },
     {
-      "epoch": 1.2581818181818183,
-      "grad_norm": 4.800558567047119,
-      "learning_rate": 2.909090909090909e-05,
-      "loss": 4.6142,
+      "epoch": 1.2490974729241877,
+      "grad_norm": 3.262009859085083,
+      "learning_rate": 2.924187725631769e-05,
+      "loss": 4.7037,
       "step": 346
     },
     {
-      "epoch": 1.2618181818181817,
-      "grad_norm": 2.561149835586548,
-      "learning_rate": 2.903030303030303e-05,
-      "loss": 4.6248,
+      "epoch": 1.2527075812274369,
+      "grad_norm": 2.6443886756896973,
+      "learning_rate": 2.9181708784596872e-05,
+      "loss": 4.5689,
       "step": 347
     },
     {
-      "epoch": 1.2654545454545454,
-      "grad_norm": 2.2318789958953857,
-      "learning_rate": 2.896969696969697e-05,
-      "loss": 4.6913,
+      "epoch": 1.256317689530686,
+      "grad_norm": 2.4855597019195557,
+      "learning_rate": 2.9121540312876056e-05,
+      "loss": 4.6886,
       "step": 348
     },
     {
-      "epoch": 1.269090909090909,
-      "grad_norm": 3.8998968601226807,
-      "learning_rate": 2.8909090909090908e-05,
-      "loss": 4.6961,
+      "epoch": 1.259927797833935,
+      "grad_norm": 2.964390754699707,
+      "learning_rate": 2.906137184115524e-05,
+      "loss": 4.5719,
       "step": 349
     },
     {
-      "epoch": 1.2727272727272727,
-      "grad_norm": 3.6434195041656494,
-      "learning_rate": 2.8848484848484853e-05,
-      "loss": 4.6966,
+      "epoch": 1.263537906137184,
+      "grad_norm": 2.4611830711364746,
+      "learning_rate": 2.9001203369434417e-05,
+      "loss": 4.6705,
       "step": 350
     },
     {
-      "epoch": 1.2763636363636364,
-      "grad_norm": 2.5094985961914062,
-      "learning_rate": 2.878787878787879e-05,
-      "loss": 4.6748,
+      "epoch": 1.2671480144404332,
+      "grad_norm": 3.128156900405884,
+      "learning_rate": 2.8941034897713598e-05,
+      "loss": 4.6805,
       "step": 351
     },
     {
-      "epoch": 1.28,
-      "grad_norm": 5.040907382965088,
-      "learning_rate": 2.872727272727273e-05,
-      "loss": 4.4934,
+      "epoch": 1.2707581227436824,
+      "grad_norm": 2.996750831604004,
+      "learning_rate": 2.888086642599278e-05,
+      "loss": 4.5068,
       "step": 352
     },
     {
-      "epoch": 1.2836363636363637,
-      "grad_norm": 4.69912576675415,
-      "learning_rate": 2.8666666666666668e-05,
-      "loss": 4.4348,
+      "epoch": 1.2743682310469313,
+      "grad_norm": 3.2886064052581787,
+      "learning_rate": 2.8820697954271962e-05,
+      "loss": 4.4811,
       "step": 353
     },
     {
-      "epoch": 1.2872727272727273,
-      "grad_norm": 2.9167287349700928,
-      "learning_rate": 2.860606060606061e-05,
-      "loss": 4.4813,
+      "epoch": 1.2779783393501805,
+      "grad_norm": 2.363309144973755,
+      "learning_rate": 2.8760529482551146e-05,
+      "loss": 4.4984,
       "step": 354
     },
     {
-      "epoch": 1.290909090909091,
-      "grad_norm": 2.7324609756469727,
-      "learning_rate": 2.8545454545454548e-05,
-      "loss": 4.5087,
+      "epoch": 1.2815884476534296,
+      "grad_norm": 5.747439861297607,
+      "learning_rate": 2.870036101083033e-05,
+      "loss": 4.5855,
       "step": 355
     },
     {
-      "epoch": 1.2945454545454544,
-      "grad_norm": 4.106588363647461,
-      "learning_rate": 2.8484848484848486e-05,
-      "loss": 4.8083,
+      "epoch": 1.2851985559566788,
+      "grad_norm": 2.802114486694336,
+      "learning_rate": 2.8640192539109507e-05,
+      "loss": 4.5466,
       "step": 356
     },
     {
-      "epoch": 1.298181818181818,
-      "grad_norm": 3.12602162361145,
-      "learning_rate": 2.8424242424242424e-05,
-      "loss": 4.5605,
+      "epoch": 1.288808664259928,
+      "grad_norm": 3.1506760120391846,
+      "learning_rate": 2.8580024067388687e-05,
+      "loss": 4.4259,
       "step": 357
     },
     {
-      "epoch": 1.3018181818181818,
-      "grad_norm": 3.1429829597473145,
-      "learning_rate": 2.8363636363636363e-05,
-      "loss": 4.6312,
+      "epoch": 1.2924187725631768,
+      "grad_norm": 6.545141220092773,
+      "learning_rate": 2.851985559566787e-05,
+      "loss": 4.4247,
       "step": 358
     },
     {
-      "epoch": 1.3054545454545454,
-      "grad_norm": 3.179574966430664,
-      "learning_rate": 2.8303030303030305e-05,
-      "loss": 4.6001,
+      "epoch": 1.296028880866426,
+      "grad_norm": 11.535502433776855,
+      "learning_rate": 2.8459687123947055e-05,
+      "loss": 4.5342,
       "step": 359
     },
     {
-      "epoch": 1.309090909090909,
-      "grad_norm": 3.6335721015930176,
-      "learning_rate": 2.8242424242424243e-05,
-      "loss": 4.4829,
+      "epoch": 1.2996389891696751,
+      "grad_norm": 6.203153133392334,
+      "learning_rate": 2.8399518652226236e-05,
+      "loss": 4.5254,
       "step": 360
     },
     {
-      "epoch": 1.3127272727272727,
-      "grad_norm": 3.5006937980651855,
-      "learning_rate": 2.818181818181818e-05,
-      "loss": 4.5427,
+      "epoch": 1.303249097472924,
+      "grad_norm": 3.3538057804107666,
+      "learning_rate": 2.8339350180505413e-05,
+      "loss": 4.5488,
       "step": 361
     },
     {
-      "epoch": 1.3163636363636364,
-      "grad_norm": 5.941033363342285,
-      "learning_rate": 2.812121212121212e-05,
-      "loss": 4.8228,
+      "epoch": 1.3068592057761732,
+      "grad_norm": 7.0459675788879395,
+      "learning_rate": 2.8279181708784597e-05,
+      "loss": 4.8362,
       "step": 362
     },
     {
-      "epoch": 1.32,
-      "grad_norm": 3.8232581615448,
-      "learning_rate": 2.8060606060606058e-05,
-      "loss": 4.6344,
+      "epoch": 1.3104693140794224,
+      "grad_norm": 6.513792037963867,
+      "learning_rate": 2.821901323706378e-05,
+      "loss": 4.5027,
       "step": 363
     },
     {
-      "epoch": 1.3236363636363637,
-      "grad_norm": 3.8346667289733887,
-      "learning_rate": 2.8000000000000003e-05,
-      "loss": 4.5342,
+      "epoch": 1.3140794223826715,
+      "grad_norm": 4.375041484832764,
+      "learning_rate": 2.815884476534296e-05,
+      "loss": 4.5529,
       "step": 364
     },
     {
-      "epoch": 1.3272727272727272,
-      "grad_norm": 3.590717315673828,
-      "learning_rate": 2.7939393939393945e-05,
-      "loss": 4.4932,
+      "epoch": 1.3176895306859207,
+      "grad_norm": 2.8317453861236572,
+      "learning_rate": 2.8098676293622145e-05,
+      "loss": 4.5939,
       "step": 365
     },
     {
-      "epoch": 1.330909090909091,
-      "grad_norm": 5.545670509338379,
-      "learning_rate": 2.7878787878787883e-05,
-      "loss": 4.6162,
+      "epoch": 1.3212996389891698,
+      "grad_norm": 5.5963640213012695,
+      "learning_rate": 2.803850782190133e-05,
+      "loss": 4.6253,
       "step": 366
     },
     {
-      "epoch": 1.3345454545454545,
-      "grad_norm": 2.9127771854400635,
-      "learning_rate": 2.781818181818182e-05,
-      "loss": 4.3389,
+      "epoch": 1.3249097472924187,
+      "grad_norm": 4.876716136932373,
+      "learning_rate": 2.7978339350180506e-05,
+      "loss": 4.5127,
       "step": 367
     },
     {
-      "epoch": 1.3381818181818181,
-      "grad_norm": 5.067335605621338,
-      "learning_rate": 2.775757575757576e-05,
-      "loss": 4.7217,
+      "epoch": 1.3285198555956679,
+      "grad_norm": 3.072681188583374,
+      "learning_rate": 2.7918170878459687e-05,
+      "loss": 4.756,
       "step": 368
     },
     {
-      "epoch": 1.3418181818181818,
-      "grad_norm": 6.718608856201172,
-      "learning_rate": 2.7696969696969698e-05,
-      "loss": 4.5497,
+      "epoch": 1.332129963898917,
+      "grad_norm": 2.941751480102539,
+      "learning_rate": 2.785800240673887e-05,
+      "loss": 4.694,
       "step": 369
     },
     {
-      "epoch": 1.3454545454545455,
-      "grad_norm": 3.0536184310913086,
-      "learning_rate": 2.7636363636363636e-05,
-      "loss": 4.48,
+      "epoch": 1.335740072202166,
+      "grad_norm": 5.169521331787109,
+      "learning_rate": 2.779783393501805e-05,
+      "loss": 4.5207,
       "step": 370
     },
     {
-      "epoch": 1.3490909090909091,
-      "grad_norm": 3.9842278957366943,
-      "learning_rate": 2.7575757575757578e-05,
-      "loss": 4.6911,
+      "epoch": 1.339350180505415,
+      "grad_norm": 4.792356491088867,
+      "learning_rate": 2.7737665463297235e-05,
+      "loss": 4.5512,
       "step": 371
     },
     {
-      "epoch": 1.3527272727272728,
-      "grad_norm": 4.863982677459717,
-      "learning_rate": 2.7515151515151516e-05,
-      "loss": 4.692,
+      "epoch": 1.3429602888086642,
+      "grad_norm": 6.515016555786133,
+      "learning_rate": 2.7677496991576412e-05,
+      "loss": 4.8404,
       "step": 372
     },
     {
-      "epoch": 1.3563636363636364,
-      "grad_norm": 3.385343551635742,
-      "learning_rate": 2.7454545454545455e-05,
-      "loss": 4.6677,
+      "epoch": 1.3465703971119134,
+      "grad_norm": 4.51361608505249,
+      "learning_rate": 2.7617328519855596e-05,
+      "loss": 4.3583,
       "step": 373
     },
     {
-      "epoch": 1.3599999999999999,
-      "grad_norm": 4.236108779907227,
-      "learning_rate": 2.7393939393939393e-05,
-      "loss": 4.5584,
+      "epoch": 1.3501805054151625,
+      "grad_norm": 3.141707181930542,
+      "learning_rate": 2.7557160048134777e-05,
+      "loss": 4.548,
       "step": 374
     },
     {
-      "epoch": 1.3636363636363638,
-      "grad_norm": 4.658186435699463,
-      "learning_rate": 2.733333333333333e-05,
-      "loss": 4.7851,
+      "epoch": 1.3537906137184115,
+      "grad_norm": 2.3701894283294678,
+      "learning_rate": 2.749699157641396e-05,
+      "loss": 4.5083,
       "step": 375
     },
     {
-      "epoch": 1.3672727272727272,
-      "grad_norm": 2.5289180278778076,
-      "learning_rate": 2.7272727272727273e-05,
-      "loss": 4.6179,
+      "epoch": 1.3574007220216606,
+      "grad_norm": 6.5997772216796875,
+      "learning_rate": 2.7436823104693144e-05,
+      "loss": 4.4702,
       "step": 376
     },
     {
-      "epoch": 1.3709090909090909,
-      "grad_norm": 3.179882049560547,
-      "learning_rate": 2.721212121212121e-05,
-      "loss": 4.5029,
+      "epoch": 1.3610108303249098,
+      "grad_norm": 6.290494441986084,
+      "learning_rate": 2.7376654632972325e-05,
+      "loss": 4.5668,
       "step": 377
     },
     {
-      "epoch": 1.3745454545454545,
-      "grad_norm": 3.1706180572509766,
-      "learning_rate": 2.7151515151515157e-05,
-      "loss": 4.4334,
+      "epoch": 1.364620938628159,
+      "grad_norm": 2.408001661300659,
+      "learning_rate": 2.7316486161251502e-05,
+      "loss": 4.7025,
       "step": 378
     },
     {
-      "epoch": 1.3781818181818182,
-      "grad_norm": 3.649819850921631,
-      "learning_rate": 2.7090909090909095e-05,
-      "loss": 4.3529,
+      "epoch": 1.3682310469314078,
+      "grad_norm": 4.783012390136719,
+      "learning_rate": 2.7256317689530686e-05,
+      "loss": 4.4451,
       "step": 379
     },
     {
-      "epoch": 1.3818181818181818,
-      "grad_norm": 4.352628231048584,
-      "learning_rate": 2.7030303030303033e-05,
-      "loss": 4.6518,
+      "epoch": 1.371841155234657,
+      "grad_norm": 5.366835594177246,
+      "learning_rate": 2.719614921780987e-05,
+      "loss": 4.6915,
       "step": 380
     },
     {
-      "epoch": 1.3854545454545455,
-      "grad_norm": 3.875267744064331,
-      "learning_rate": 2.696969696969697e-05,
-      "loss": 4.6,
+      "epoch": 1.3754512635379061,
+      "grad_norm": 2.9238317012786865,
+      "learning_rate": 2.713598074608905e-05,
+      "loss": 4.6786,
       "step": 381
     },
     {
-      "epoch": 1.3890909090909092,
-      "grad_norm": 7.508423328399658,
-      "learning_rate": 2.6909090909090913e-05,
-      "loss": 4.6121,
+      "epoch": 1.3790613718411553,
+      "grad_norm": 4.001113414764404,
+      "learning_rate": 2.7075812274368234e-05,
+      "loss": 4.5797,
       "step": 382
     },
     {
-      "epoch": 1.3927272727272726,
-      "grad_norm": 4.307608604431152,
-      "learning_rate": 2.684848484848485e-05,
-      "loss": 4.6758,
+      "epoch": 1.3826714801444044,
+      "grad_norm": 3.905136823654175,
+      "learning_rate": 2.7015643802647418e-05,
+      "loss": 4.5265,
       "step": 383
     },
     {
-      "epoch": 1.3963636363636365,
-      "grad_norm": 2.5840773582458496,
-      "learning_rate": 2.678787878787879e-05,
-      "loss": 4.5082,
+      "epoch": 1.3862815884476534,
+      "grad_norm": 3.3477468490600586,
+      "learning_rate": 2.6955475330926595e-05,
+      "loss": 4.5869,
       "step": 384
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 3.530118703842163,
-      "learning_rate": 2.6727272727272728e-05,
-      "loss": 4.3333,
+      "epoch": 1.3898916967509025,
+      "grad_norm": 2.695615768432617,
+      "learning_rate": 2.6895306859205776e-05,
+      "loss": 4.5366,
       "step": 385
     },
     {
-      "epoch": 1.4036363636363636,
-      "grad_norm": 6.304224491119385,
-      "learning_rate": 2.6666666666666667e-05,
-      "loss": 4.6398,
+      "epoch": 1.3935018050541517,
+      "grad_norm": 3.9516983032226562,
+      "learning_rate": 2.683513838748496e-05,
+      "loss": 4.5686,
       "step": 386
     },
     {
-      "epoch": 1.4072727272727272,
-      "grad_norm": 3.043519973754883,
-      "learning_rate": 2.6606060606060605e-05,
-      "loss": 4.4337,
+      "epoch": 1.3971119133574006,
+      "grad_norm": 5.092808723449707,
+      "learning_rate": 2.6774969915764144e-05,
+      "loss": 4.4119,
       "step": 387
     },
     {
-      "epoch": 1.410909090909091,
-      "grad_norm": 5.097027778625488,
-      "learning_rate": 2.6545454545454547e-05,
-      "loss": 4.6867,
+      "epoch": 1.4007220216606497,
+      "grad_norm": 4.993492126464844,
+      "learning_rate": 2.6714801444043324e-05,
+      "loss": 4.7174,
       "step": 388
     },
     {
-      "epoch": 1.4145454545454546,
-      "grad_norm": 5.522812366485596,
-      "learning_rate": 2.6484848484848485e-05,
-      "loss": 4.3685,
+      "epoch": 1.404332129963899,
+      "grad_norm": 3.519829750061035,
+      "learning_rate": 2.66546329723225e-05,
+      "loss": 4.7413,
       "step": 389
     },
     {
-      "epoch": 1.4181818181818182,
-      "grad_norm": 3.3053135871887207,
-      "learning_rate": 2.6424242424242423e-05,
-      "loss": 4.6564,
+      "epoch": 1.407942238267148,
+      "grad_norm": 3.6330387592315674,
+      "learning_rate": 2.6594464500601685e-05,
+      "loss": 4.5918,
       "step": 390
     },
     {
-      "epoch": 1.4218181818181819,
-      "grad_norm": 4.631997585296631,
-      "learning_rate": 2.636363636363636e-05,
-      "loss": 4.5055,
+      "epoch": 1.4115523465703972,
+      "grad_norm": 4.714821815490723,
+      "learning_rate": 2.6534296028880866e-05,
+      "loss": 4.4284,
       "step": 391
     },
     {
-      "epoch": 1.4254545454545455,
-      "grad_norm": 2.758570671081543,
-      "learning_rate": 2.63030303030303e-05,
-      "loss": 4.5271,
+      "epoch": 1.4151624548736463,
+      "grad_norm": 5.586686134338379,
+      "learning_rate": 2.647412755716005e-05,
+      "loss": 4.6257,
       "step": 392
     },
     {
-      "epoch": 1.4290909090909092,
-      "grad_norm": 2.726597785949707,
-      "learning_rate": 2.6242424242424245e-05,
-      "loss": 4.4968,
+      "epoch": 1.4187725631768953,
+      "grad_norm": 4.272150993347168,
+      "learning_rate": 2.6413959085439234e-05,
+      "loss": 4.7499,
       "step": 393
     },
     {
-      "epoch": 1.4327272727272726,
-      "grad_norm": 5.252767086029053,
-      "learning_rate": 2.6181818181818187e-05,
-      "loss": 4.8041,
+      "epoch": 1.4223826714801444,
+      "grad_norm": 3.671114683151245,
+      "learning_rate": 2.6353790613718414e-05,
+      "loss": 4.4255,
       "step": 394
     },
     {
-      "epoch": 1.4363636363636363,
-      "grad_norm": 3.7906062602996826,
-      "learning_rate": 2.6121212121212125e-05,
-      "loss": 4.3581,
+      "epoch": 1.4259927797833936,
+      "grad_norm": 3.9677507877349854,
+      "learning_rate": 2.629362214199759e-05,
+      "loss": 4.8504,
       "step": 395
     },
     {
-      "epoch": 1.44,
-      "grad_norm": 5.029653549194336,
-      "learning_rate": 2.6060606060606063e-05,
-      "loss": 4.4908,
+      "epoch": 1.4296028880866425,
+      "grad_norm": 3.9609200954437256,
+      "learning_rate": 2.6233453670276775e-05,
+      "loss": 4.6043,
       "step": 396
     },
     {
-      "epoch": 1.4436363636363636,
-      "grad_norm": 6.576402187347412,
-      "learning_rate": 2.6000000000000002e-05,
-      "loss": 4.4629,
+      "epoch": 1.4332129963898916,
+      "grad_norm": 3.998934030532837,
+      "learning_rate": 2.617328519855596e-05,
+      "loss": 4.4927,
       "step": 397
     },
     {
-      "epoch": 1.4472727272727273,
-      "grad_norm": 6.675891399383545,
-      "learning_rate": 2.593939393939394e-05,
-      "loss": 4.7604,
+      "epoch": 1.4368231046931408,
+      "grad_norm": 3.251746892929077,
+      "learning_rate": 2.611311672683514e-05,
+      "loss": 4.4543,
       "step": 398
     },
     {
-      "epoch": 1.450909090909091,
-      "grad_norm": 4.859799861907959,
-      "learning_rate": 2.587878787878788e-05,
-      "loss": 4.431,
+      "epoch": 1.44043321299639,
+      "grad_norm": 6.4932379722595215,
+      "learning_rate": 2.6052948255114323e-05,
+      "loss": 4.7533,
       "step": 399
     },
     {
-      "epoch": 1.4545454545454546,
-      "grad_norm": 4.277210712432861,
-      "learning_rate": 2.581818181818182e-05,
-      "loss": 4.4418,
+      "epoch": 1.444043321299639,
+      "grad_norm": 6.668980121612549,
+      "learning_rate": 2.59927797833935e-05,
+      "loss": 4.6553,
       "step": 400
     },
     {
-      "epoch": 1.4581818181818182,
-      "grad_norm": 3.4903981685638428,
-      "learning_rate": 2.575757575757576e-05,
-      "loss": 4.5139,
+      "epoch": 1.447653429602888,
+      "grad_norm": 3.2073581218719482,
+      "learning_rate": 2.5932611311672685e-05,
+      "loss": 4.711,
       "step": 401
     },
     {
-      "epoch": 1.461818181818182,
-      "grad_norm": 5.1570963859558105,
-      "learning_rate": 2.5696969696969697e-05,
-      "loss": 4.4146,
+      "epoch": 1.4512635379061372,
+      "grad_norm": 5.215062141418457,
+      "learning_rate": 2.5872442839951865e-05,
+      "loss": 4.519,
       "step": 402
     },
     {
-      "epoch": 1.4654545454545453,
-      "grad_norm": 7.002410411834717,
-      "learning_rate": 2.5636363636363635e-05,
-      "loss": 4.6652,
+      "epoch": 1.4548736462093863,
+      "grad_norm": 7.263779640197754,
+      "learning_rate": 2.581227436823105e-05,
+      "loss": 4.8241,
       "step": 403
     },
     {
-      "epoch": 1.4690909090909092,
-      "grad_norm": 3.277343273162842,
-      "learning_rate": 2.5575757575757573e-05,
-      "loss": 4.8023,
+      "epoch": 1.4584837545126355,
+      "grad_norm": 3.1585211753845215,
+      "learning_rate": 2.5752105896510233e-05,
+      "loss": 4.6324,
       "step": 404
     },
     {
-      "epoch": 1.4727272727272727,
-      "grad_norm": 4.8317341804504395,
-      "learning_rate": 2.5515151515151515e-05,
-      "loss": 4.7799,
+      "epoch": 1.4620938628158844,
+      "grad_norm": 8.619237899780273,
+      "learning_rate": 2.5691937424789413e-05,
+      "loss": 4.4658,
       "step": 405
     },
     {
-      "epoch": 1.4763636363636363,
-      "grad_norm": 3.615807056427002,
-      "learning_rate": 2.5454545454545454e-05,
-      "loss": 4.4125,
+      "epoch": 1.4657039711191335,
+      "grad_norm": 9.702713012695312,
+      "learning_rate": 2.563176895306859e-05,
+      "loss": 4.7969,
       "step": 406
     },
     {
-      "epoch": 1.48,
-      "grad_norm": 2.5169756412506104,
-      "learning_rate": 2.53939393939394e-05,
-      "loss": 4.6356,
+      "epoch": 1.4693140794223827,
+      "grad_norm": 6.16801118850708,
+      "learning_rate": 2.5571600481347774e-05,
+      "loss": 4.5924,
       "step": 407
     },
     {
-      "epoch": 1.4836363636363636,
-      "grad_norm": 5.736434459686279,
-      "learning_rate": 2.5333333333333337e-05,
-      "loss": 4.3922,
+      "epoch": 1.4729241877256318,
+      "grad_norm": 3.287752151489258,
+      "learning_rate": 2.5511432009626955e-05,
+      "loss": 4.8517,
       "step": 408
     },
     {
-      "epoch": 1.4872727272727273,
-      "grad_norm": 5.380589008331299,
-      "learning_rate": 2.5272727272727275e-05,
-      "loss": 4.4222,
+      "epoch": 1.476534296028881,
+      "grad_norm": 6.46322774887085,
+      "learning_rate": 2.545126353790614e-05,
+      "loss": 4.588,
       "step": 409
     },
     {
-      "epoch": 1.490909090909091,
-      "grad_norm": 4.923427581787109,
-      "learning_rate": 2.5212121212121214e-05,
-      "loss": 4.5611,
+      "epoch": 1.48014440433213,
+      "grad_norm": 5.652312755584717,
+      "learning_rate": 2.5391095066185323e-05,
+      "loss": 4.649,
       "step": 410
     },
     {
-      "epoch": 1.4945454545454546,
-      "grad_norm": 6.039291858673096,
-      "learning_rate": 2.5151515151515155e-05,
-      "loss": 4.4644,
+      "epoch": 1.483754512635379,
+      "grad_norm": 4.650730609893799,
+      "learning_rate": 2.53309265944645e-05,
+      "loss": 4.5696,
       "step": 411
     },
     {
-      "epoch": 1.498181818181818,
-      "grad_norm": 4.344960689544678,
-      "learning_rate": 2.5090909090909094e-05,
-      "loss": 4.6914,
+      "epoch": 1.4873646209386282,
+      "grad_norm": 3.9885752201080322,
+      "learning_rate": 2.527075812274368e-05,
+      "loss": 4.3658,
       "step": 412
     },
     {
-      "epoch": 1.501818181818182,
-      "grad_norm": 2.4840219020843506,
-      "learning_rate": 2.5030303030303032e-05,
-      "loss": 4.4634,
+      "epoch": 1.4909747292418771,
+      "grad_norm": 7.778469085693359,
+      "learning_rate": 2.5210589651022864e-05,
+      "loss": 4.5531,
       "step": 413
     },
     {
-      "epoch": 1.5054545454545454,
-      "grad_norm": 4.167297840118408,
-      "learning_rate": 2.496969696969697e-05,
-      "loss": 4.5234,
+      "epoch": 1.4945848375451263,
+      "grad_norm": 4.358502388000488,
+      "learning_rate": 2.5150421179302048e-05,
+      "loss": 4.8113,
       "step": 414
     },
     {
-      "epoch": 1.509090909090909,
-      "grad_norm": 3.043473482131958,
-      "learning_rate": 2.490909090909091e-05,
-      "loss": 4.5519,
+      "epoch": 1.4981949458483754,
+      "grad_norm": 3.1924219131469727,
+      "learning_rate": 2.509025270758123e-05,
+      "loss": 4.597,
       "step": 415
     },
     {
-      "epoch": 1.5127272727272727,
-      "grad_norm": 2.029343366622925,
-      "learning_rate": 2.4848484848484847e-05,
-      "loss": 4.4137,
+      "epoch": 1.5018050541516246,
+      "grad_norm": 2.0474438667297363,
+      "learning_rate": 2.5030084235860413e-05,
+      "loss": 4.6747,
       "step": 416
     },
     {
-      "epoch": 1.5163636363636364,
-      "grad_norm": 3.7380053997039795,
-      "learning_rate": 2.478787878787879e-05,
-      "loss": 4.7946,
+      "epoch": 1.5054151624548737,
+      "grad_norm": 3.988816261291504,
+      "learning_rate": 2.4969915764139593e-05,
+      "loss": 4.4763,
       "step": 417
     },
     {
-      "epoch": 1.52,
-      "grad_norm": 3.8255624771118164,
-      "learning_rate": 2.472727272727273e-05,
-      "loss": 4.6077,
+      "epoch": 1.5090252707581229,
+      "grad_norm": 4.126523971557617,
+      "learning_rate": 2.4909747292418774e-05,
+      "loss": 4.5102,
       "step": 418
     },
     {
-      "epoch": 1.5236363636363637,
-      "grad_norm": 3.6973726749420166,
-      "learning_rate": 2.466666666666667e-05,
-      "loss": 4.4438,
+      "epoch": 1.5126353790613718,
+      "grad_norm": 3.799755334854126,
+      "learning_rate": 2.4849578820697954e-05,
+      "loss": 4.3798,
       "step": 419
     },
     {
-      "epoch": 1.5272727272727273,
-      "grad_norm": 2.9415085315704346,
-      "learning_rate": 2.4606060606060607e-05,
-      "loss": 4.5242,
+      "epoch": 1.516245487364621,
+      "grad_norm": 5.314898490905762,
+      "learning_rate": 2.4789410348977138e-05,
+      "loss": 4.4691,
       "step": 420
     },
     {
-      "epoch": 1.5309090909090908,
-      "grad_norm": 2.4222817420959473,
-      "learning_rate": 2.4545454545454545e-05,
-      "loss": 4.4857,
+      "epoch": 1.5198555956678699,
+      "grad_norm": 4.028231620788574,
+      "learning_rate": 2.472924187725632e-05,
+      "loss": 4.5874,
       "step": 421
     },
     {
-      "epoch": 1.5345454545454547,
-      "grad_norm": 3.7326831817626953,
-      "learning_rate": 2.4484848484848484e-05,
-      "loss": 4.3882,
+      "epoch": 1.523465703971119,
+      "grad_norm": 3.6319522857666016,
+      "learning_rate": 2.46690734055355e-05,
+      "loss": 4.5248,
       "step": 422
     },
     {
-      "epoch": 1.538181818181818,
-      "grad_norm": 5.729616165161133,
-      "learning_rate": 2.4424242424242426e-05,
-      "loss": 4.6133,
+      "epoch": 1.5270758122743682,
+      "grad_norm": 6.449628829956055,
+      "learning_rate": 2.4608904933814683e-05,
+      "loss": 4.4882,
       "step": 423
     },
     {
-      "epoch": 1.541818181818182,
-      "grad_norm": 3.818107843399048,
-      "learning_rate": 2.4363636363636364e-05,
-      "loss": 4.5399,
+      "epoch": 1.5306859205776173,
+      "grad_norm": 7.905242919921875,
+      "learning_rate": 2.4548736462093864e-05,
+      "loss": 4.7121,
       "step": 424
     },
     {
-      "epoch": 1.5454545454545454,
-      "grad_norm": 6.078042984008789,
-      "learning_rate": 2.4303030303030306e-05,
-      "loss": 4.4185,
+      "epoch": 1.5342960288808665,
+      "grad_norm": 4.377833843231201,
+      "learning_rate": 2.4488567990373047e-05,
+      "loss": 4.5071,
       "step": 425
     },
     {
-      "epoch": 1.549090909090909,
-      "grad_norm": 3.017300844192505,
-      "learning_rate": 2.4242424242424244e-05,
-      "loss": 4.3699,
+      "epoch": 1.5379061371841156,
+      "grad_norm": 3.6568801403045654,
+      "learning_rate": 2.4428399518652228e-05,
+      "loss": 4.4466,
       "step": 426
     },
     {
-      "epoch": 1.5527272727272727,
-      "grad_norm": 2.781771421432495,
-      "learning_rate": 2.4181818181818182e-05,
-      "loss": 4.4404,
+      "epoch": 1.5415162454873648,
+      "grad_norm": 11.142749786376953,
+      "learning_rate": 2.436823104693141e-05,
+      "loss": 4.515,
       "step": 427
     },
     {
-      "epoch": 1.5563636363636364,
-      "grad_norm": 5.910660743713379,
-      "learning_rate": 2.4121212121212124e-05,
-      "loss": 4.2755,
+      "epoch": 1.5451263537906137,
+      "grad_norm": 5.1178083419799805,
+      "learning_rate": 2.4308062575210592e-05,
+      "loss": 4.4591,
       "step": 428
     },
     {
-      "epoch": 1.56,
-      "grad_norm": 6.957926273345947,
-      "learning_rate": 2.4060606060606062e-05,
-      "loss": 4.6575,
+      "epoch": 1.5487364620938628,
+      "grad_norm": 2.430281639099121,
+      "learning_rate": 2.4247894103489773e-05,
+      "loss": 4.5393,
       "step": 429
     },
     {
-      "epoch": 1.5636363636363635,
-      "grad_norm": 5.411463737487793,
-      "learning_rate": 2.4e-05,
-      "loss": 4.5792,
+      "epoch": 1.5523465703971118,
+      "grad_norm": 3.443852424621582,
+      "learning_rate": 2.4187725631768953e-05,
+      "loss": 4.5422,
       "step": 430
     },
     {
-      "epoch": 1.5672727272727274,
-      "grad_norm": 4.2124924659729,
-      "learning_rate": 2.393939393939394e-05,
-      "loss": 4.4813,
+      "epoch": 1.555956678700361,
+      "grad_norm": 4.571306228637695,
+      "learning_rate": 2.4127557160048137e-05,
+      "loss": 4.673,
       "step": 431
     },
     {
-      "epoch": 1.5709090909090908,
-      "grad_norm": 5.113527297973633,
-      "learning_rate": 2.387878787878788e-05,
-      "loss": 4.6248,
+      "epoch": 1.55956678700361,
+      "grad_norm": 3.1203410625457764,
+      "learning_rate": 2.4067388688327318e-05,
+      "loss": 4.6205,
       "step": 432
     },
     {
-      "epoch": 1.5745454545454547,
-      "grad_norm": 5.174093246459961,
-      "learning_rate": 2.381818181818182e-05,
-      "loss": 4.4863,
+      "epoch": 1.5631768953068592,
+      "grad_norm": 5.54405403137207,
+      "learning_rate": 2.40072202166065e-05,
+      "loss": 4.6493,
       "step": 433
     },
     {
-      "epoch": 1.5781818181818181,
-      "grad_norm": 4.3809027671813965,
-      "learning_rate": 2.375757575757576e-05,
-      "loss": 4.4467,
+      "epoch": 1.5667870036101084,
+      "grad_norm": 4.841291904449463,
+      "learning_rate": 2.3947051744885682e-05,
+      "loss": 4.4992,
       "step": 434
     },
     {
-      "epoch": 1.5818181818181818,
-      "grad_norm": 3.806034564971924,
-      "learning_rate": 2.36969696969697e-05,
-      "loss": 4.458,
+      "epoch": 1.5703971119133575,
+      "grad_norm": 5.931495189666748,
+      "learning_rate": 2.3886883273164863e-05,
+      "loss": 4.773,
       "step": 435
     },
     {
-      "epoch": 1.5854545454545454,
-      "grad_norm": 6.237285614013672,
-      "learning_rate": 2.3636363636363637e-05,
-      "loss": 4.584,
+      "epoch": 1.5740072202166067,
+      "grad_norm": 3.611367702484131,
+      "learning_rate": 2.3826714801444043e-05,
+      "loss": 4.5129,
       "step": 436
     },
     {
-      "epoch": 1.589090909090909,
-      "grad_norm": 8.469368934631348,
-      "learning_rate": 2.3575757575757576e-05,
-      "loss": 4.5677,
+      "epoch": 1.5776173285198556,
+      "grad_norm": 7.546535015106201,
+      "learning_rate": 2.3766546329723227e-05,
+      "loss": 4.4994,
       "step": 437
     },
     {
-      "epoch": 1.5927272727272728,
-      "grad_norm": 6.997206211090088,
-      "learning_rate": 2.3515151515151514e-05,
-      "loss": 4.4851,
+      "epoch": 1.5812274368231047,
+      "grad_norm": 5.833862781524658,
+      "learning_rate": 2.3706377858002408e-05,
+      "loss": 4.8386,
       "step": 438
     },
     {
-      "epoch": 1.5963636363636362,
-      "grad_norm": 3.9165611267089844,
-      "learning_rate": 2.3454545454545456e-05,
-      "loss": 4.6739,
+      "epoch": 1.5848375451263537,
+      "grad_norm": 3.0301411151885986,
+      "learning_rate": 2.3646209386281588e-05,
+      "loss": 4.3775,
       "step": 439
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 4.296258449554443,
-      "learning_rate": 2.3393939393939397e-05,
-      "loss": 4.3954,
+      "epoch": 1.5884476534296028,
+      "grad_norm": 5.783769607543945,
+      "learning_rate": 2.3586040914560772e-05,
+      "loss": 4.6718,
       "step": 440
     },
     {
-      "epoch": 1.6036363636363635,
-      "grad_norm": 4.895616054534912,
-      "learning_rate": 2.3333333333333336e-05,
-      "loss": 4.6904,
+      "epoch": 1.592057761732852,
+      "grad_norm": 3.89277982711792,
+      "learning_rate": 2.3525872442839953e-05,
+      "loss": 4.4917,
       "step": 441
     },
     {
-      "epoch": 1.6072727272727274,
-      "grad_norm": 3.0008482933044434,
-      "learning_rate": 2.3272727272727274e-05,
-      "loss": 4.4606,
+      "epoch": 1.595667870036101,
+      "grad_norm": 2.7134242057800293,
+      "learning_rate": 2.3465703971119137e-05,
+      "loss": 4.5437,
       "step": 442
     },
     {
-      "epoch": 1.6109090909090908,
-      "grad_norm": 3.4518730640411377,
-      "learning_rate": 2.3212121212121212e-05,
-      "loss": 4.3835,
+      "epoch": 1.5992779783393503,
+      "grad_norm": 4.376542568206787,
+      "learning_rate": 2.3405535499398314e-05,
+      "loss": 4.5702,
       "step": 443
     },
     {
-      "epoch": 1.6145454545454545,
-      "grad_norm": 7.562982559204102,
-      "learning_rate": 2.315151515151515e-05,
-      "loss": 4.7344,
+      "epoch": 1.6028880866425994,
+      "grad_norm": 5.11259651184082,
+      "learning_rate": 2.3345367027677498e-05,
+      "loss": 4.3343,
       "step": 444
     },
     {
-      "epoch": 1.6181818181818182,
-      "grad_norm": 3.1272881031036377,
-      "learning_rate": 2.309090909090909e-05,
-      "loss": 4.6143,
+      "epoch": 1.6064981949458483,
+      "grad_norm": 3.1635055541992188,
+      "learning_rate": 2.328519855595668e-05,
+      "loss": 4.3862,
       "step": 445
     },
     {
-      "epoch": 1.6218181818181818,
-      "grad_norm": 2.709015369415283,
-      "learning_rate": 2.3030303030303034e-05,
-      "loss": 4.4113,
+      "epoch": 1.6101083032490975,
+      "grad_norm": 5.924020767211914,
+      "learning_rate": 2.322503008423586e-05,
+      "loss": 4.3922,
       "step": 446
     },
     {
-      "epoch": 1.6254545454545455,
-      "grad_norm": 3.9435224533081055,
-      "learning_rate": 2.2969696969696973e-05,
-      "loss": 4.4467,
+      "epoch": 1.6137184115523464,
+      "grad_norm": 6.799149513244629,
+      "learning_rate": 2.3164861612515043e-05,
+      "loss": 4.5604,
       "step": 447
     },
     {
-      "epoch": 1.6290909090909091,
-      "grad_norm": 6.414844989776611,
-      "learning_rate": 2.290909090909091e-05,
-      "loss": 4.8062,
+      "epoch": 1.6173285198555956,
+      "grad_norm": 3.7257330417633057,
+      "learning_rate": 2.3104693140794227e-05,
+      "loss": 4.468,
       "step": 448
     },
     {
-      "epoch": 1.6327272727272728,
-      "grad_norm": 3.1530921459198,
-      "learning_rate": 2.284848484848485e-05,
-      "loss": 4.4613,
+      "epoch": 1.6209386281588447,
+      "grad_norm": 5.105785369873047,
+      "learning_rate": 2.3044524669073407e-05,
+      "loss": 4.5101,
       "step": 449
     },
     {
-      "epoch": 1.6363636363636362,
-      "grad_norm": 3.0431370735168457,
-      "learning_rate": 2.2787878787878788e-05,
-      "loss": 4.3202,
+      "epoch": 1.6245487364620939,
+      "grad_norm": 7.325453758239746,
+      "learning_rate": 2.2984356197352588e-05,
+      "loss": 4.4697,
       "step": 450
     },
     {
-      "epoch": 1.6400000000000001,
-      "grad_norm": 4.826960563659668,
-      "learning_rate": 2.272727272727273e-05,
-      "loss": 4.5825,
+      "epoch": 1.628158844765343,
+      "grad_norm": 3.203767776489258,
+      "learning_rate": 2.292418772563177e-05,
+      "loss": 4.5219,
       "step": 451
     },
     {
-      "epoch": 1.6436363636363636,
-      "grad_norm": 8.548043251037598,
-      "learning_rate": 2.2666666666666668e-05,
-      "loss": 4.3152,
+      "epoch": 1.6317689530685922,
+      "grad_norm": 3.2828943729400635,
+      "learning_rate": 2.2864019253910952e-05,
+      "loss": 4.4196,
       "step": 452
     },
     {
-      "epoch": 1.6472727272727272,
-      "grad_norm": 4.511161804199219,
-      "learning_rate": 2.260606060606061e-05,
-      "loss": 4.4899,
+      "epoch": 1.6353790613718413,
+      "grad_norm": 4.563648700714111,
+      "learning_rate": 2.2803850782190133e-05,
+      "loss": 4.6709,
       "step": 453
     },
     {
-      "epoch": 1.6509090909090909,
-      "grad_norm": 3.710296154022217,
-      "learning_rate": 2.2545454545454548e-05,
-      "loss": 4.6325,
+      "epoch": 1.6389891696750902,
+      "grad_norm": 4.947784900665283,
+      "learning_rate": 2.2743682310469316e-05,
+      "loss": 4.7708,
       "step": 454
     },
     {
-      "epoch": 1.6545454545454545,
-      "grad_norm": 1.9863725900650024,
-      "learning_rate": 2.2484848484848486e-05,
-      "loss": 4.4654,
+      "epoch": 1.6425992779783394,
+      "grad_norm": 3.3646271228790283,
+      "learning_rate": 2.2683513838748497e-05,
+      "loss": 4.5654,
       "step": 455
     },
     {
-      "epoch": 1.6581818181818182,
-      "grad_norm": 4.258086204528809,
-      "learning_rate": 2.2424242424242424e-05,
-      "loss": 4.561,
+      "epoch": 1.6462093862815883,
+      "grad_norm": 3.2443368434906006,
+      "learning_rate": 2.2623345367027677e-05,
+      "loss": 4.5126,
       "step": 456
     },
     {
-      "epoch": 1.6618181818181819,
-      "grad_norm": 5.20453405380249,
-      "learning_rate": 2.2363636363636366e-05,
-      "loss": 4.4563,
+      "epoch": 1.6498194945848375,
+      "grad_norm": 2.8289527893066406,
+      "learning_rate": 2.2563176895306858e-05,
+      "loss": 4.5584,
       "step": 457
     },
     {
-      "epoch": 1.6654545454545455,
-      "grad_norm": 5.107168674468994,
-      "learning_rate": 2.2303030303030304e-05,
-      "loss": 4.6385,
+      "epoch": 1.6534296028880866,
+      "grad_norm": 4.497228622436523,
+      "learning_rate": 2.2503008423586042e-05,
+      "loss": 4.6282,
       "step": 458
     },
     {
-      "epoch": 1.669090909090909,
-      "grad_norm": 6.3191304206848145,
-      "learning_rate": 2.2242424242424243e-05,
-      "loss": 4.5206,
+      "epoch": 1.6570397111913358,
+      "grad_norm": 2.863171100616455,
+      "learning_rate": 2.2442839951865226e-05,
+      "loss": 4.4665,
       "step": 459
     },
     {
-      "epoch": 1.6727272727272728,
-      "grad_norm": 3.3349220752716064,
-      "learning_rate": 2.2181818181818184e-05,
-      "loss": 4.3053,
+      "epoch": 1.660649819494585,
+      "grad_norm": 2.3778388500213623,
+      "learning_rate": 2.2382671480144403e-05,
+      "loss": 4.6985,
       "step": 460
     },
     {
-      "epoch": 1.6763636363636363,
-      "grad_norm": 4.218853950500488,
-      "learning_rate": 2.2121212121212123e-05,
-      "loss": 4.5629,
+      "epoch": 1.664259927797834,
+      "grad_norm": 4.169567108154297,
+      "learning_rate": 2.2322503008423587e-05,
+      "loss": 4.8346,
       "step": 461
     },
     {
-      "epoch": 1.6800000000000002,
-      "grad_norm": 5.860531806945801,
-      "learning_rate": 2.206060606060606e-05,
-      "loss": 4.4272,
+      "epoch": 1.6678700361010832,
+      "grad_norm": 5.816603183746338,
+      "learning_rate": 2.226233453670277e-05,
+      "loss": 4.4637,
       "step": 462
     },
     {
-      "epoch": 1.6836363636363636,
-      "grad_norm": 4.101426601409912,
-      "learning_rate": 2.2000000000000003e-05,
-      "loss": 4.573,
+      "epoch": 1.6714801444043321,
+      "grad_norm": 4.912841796875,
+      "learning_rate": 2.220216606498195e-05,
+      "loss": 4.5214,
       "step": 463
     },
     {
-      "epoch": 1.6872727272727273,
-      "grad_norm": 3.375948429107666,
-      "learning_rate": 2.193939393939394e-05,
-      "loss": 4.3781,
+      "epoch": 1.6750902527075813,
+      "grad_norm": 3.094038724899292,
+      "learning_rate": 2.2141997593261132e-05,
+      "loss": 4.5753,
       "step": 464
     },
     {
-      "epoch": 1.690909090909091,
-      "grad_norm": 3.112006902694702,
-      "learning_rate": 2.187878787878788e-05,
-      "loss": 4.4511,
+      "epoch": 1.6787003610108302,
+      "grad_norm": 6.050639629364014,
+      "learning_rate": 2.2081829121540316e-05,
+      "loss": 4.5278,
       "step": 465
     },
     {
-      "epoch": 1.6945454545454546,
-      "grad_norm": 3.9260976314544678,
-      "learning_rate": 2.1818181818181818e-05,
-      "loss": 4.4448,
+      "epoch": 1.6823104693140793,
+      "grad_norm": 5.237624168395996,
+      "learning_rate": 2.2021660649819496e-05,
+      "loss": 4.4212,
       "step": 466
     },
     {
-      "epoch": 1.6981818181818182,
-      "grad_norm": 2.8007559776306152,
-      "learning_rate": 2.175757575757576e-05,
-      "loss": 4.4512,
+      "epoch": 1.6859205776173285,
+      "grad_norm": 2.529154062271118,
+      "learning_rate": 2.1961492178098677e-05,
+      "loss": 4.5208,
       "step": 467
     },
     {
-      "epoch": 1.7018181818181817,
-      "grad_norm": 3.95316743850708,
-      "learning_rate": 2.1696969696969698e-05,
-      "loss": 4.6404,
+      "epoch": 1.6895306859205776,
+      "grad_norm": 2.7793679237365723,
+      "learning_rate": 2.190132370637786e-05,
+      "loss": 4.458,
       "step": 468
     },
     {
-      "epoch": 1.7054545454545456,
-      "grad_norm": 3.3578267097473145,
-      "learning_rate": 2.163636363636364e-05,
-      "loss": 4.4527,
+      "epoch": 1.6931407942238268,
+      "grad_norm": 3.213160991668701,
+      "learning_rate": 2.184115523465704e-05,
+      "loss": 4.4169,
       "step": 469
     },
     {
-      "epoch": 1.709090909090909,
-      "grad_norm": 4.38985013961792,
-      "learning_rate": 2.1575757575757578e-05,
-      "loss": 4.531,
+      "epoch": 1.696750902527076,
+      "grad_norm": 6.508391380310059,
+      "learning_rate": 2.178098676293622e-05,
+      "loss": 4.55,
       "step": 470
     },
     {
-      "epoch": 1.7127272727272729,
-      "grad_norm": 3.9264490604400635,
-      "learning_rate": 2.1515151515151516e-05,
-      "loss": 4.495,
+      "epoch": 1.7003610108303249,
+      "grad_norm": 8.193519592285156,
+      "learning_rate": 2.1720818291215402e-05,
+      "loss": 4.6847,
       "step": 471
     },
     {
-      "epoch": 1.7163636363636363,
-      "grad_norm": 4.830046653747559,
-      "learning_rate": 2.1454545454545455e-05,
-      "loss": 4.4321,
+      "epoch": 1.703971119133574,
+      "grad_norm": 3.6836369037628174,
+      "learning_rate": 2.1660649819494586e-05,
+      "loss": 4.607,
       "step": 472
     },
     {
-      "epoch": 1.72,
-      "grad_norm": 4.556273937225342,
-      "learning_rate": 2.1393939393939393e-05,
-      "loss": 4.4907,
+      "epoch": 1.707581227436823,
+      "grad_norm": 5.093707084655762,
+      "learning_rate": 2.1600481347773767e-05,
+      "loss": 4.3498,
       "step": 473
     },
     {
-      "epoch": 1.7236363636363636,
-      "grad_norm": 5.617473125457764,
-      "learning_rate": 2.1333333333333335e-05,
-      "loss": 4.7541,
+      "epoch": 1.711191335740072,
+      "grad_norm": 5.675652980804443,
+      "learning_rate": 2.1540312876052947e-05,
+      "loss": 4.5075,
       "step": 474
     },
     {
-      "epoch": 1.7272727272727273,
-      "grad_norm": 3.0827252864837646,
-      "learning_rate": 2.1272727272727276e-05,
-      "loss": 4.3495,
+      "epoch": 1.7148014440433212,
+      "grad_norm": 4.682228088378906,
+      "learning_rate": 2.148014440433213e-05,
+      "loss": 4.7128,
       "step": 475
     },
     {
-      "epoch": 1.730909090909091,
-      "grad_norm": 3.5010452270507812,
-      "learning_rate": 2.1212121212121215e-05,
-      "loss": 4.6724,
+      "epoch": 1.7184115523465704,
+      "grad_norm": 4.707435607910156,
+      "learning_rate": 2.1419975932611315e-05,
+      "loss": 4.4863,
       "step": 476
     },
     {
-      "epoch": 1.7345454545454544,
-      "grad_norm": 3.626228094100952,
-      "learning_rate": 2.1151515151515153e-05,
-      "loss": 4.7117,
+      "epoch": 1.7220216606498195,
+      "grad_norm": 7.476378440856934,
+      "learning_rate": 2.1359807460890492e-05,
+      "loss": 4.7813,
       "step": 477
     },
     {
-      "epoch": 1.7381818181818183,
-      "grad_norm": 3.944805383682251,
-      "learning_rate": 2.109090909090909e-05,
-      "loss": 4.4103,
+      "epoch": 1.7256317689530687,
+      "grad_norm": 3.5324857234954834,
+      "learning_rate": 2.1299638989169676e-05,
+      "loss": 4.4661,
       "step": 478
     },
     {
-      "epoch": 1.7418181818181817,
-      "grad_norm": 2.4511728286743164,
-      "learning_rate": 2.103030303030303e-05,
-      "loss": 4.5316,
+      "epoch": 1.7292418772563178,
+      "grad_norm": 2.770537853240967,
+      "learning_rate": 2.123947051744886e-05,
+      "loss": 4.5247,
       "step": 479
     },
     {
-      "epoch": 1.7454545454545456,
-      "grad_norm": 6.276540756225586,
-      "learning_rate": 2.096969696969697e-05,
-      "loss": 4.4635,
+      "epoch": 1.7328519855595668,
+      "grad_norm": 2.7567989826202393,
+      "learning_rate": 2.117930204572804e-05,
+      "loss": 4.6836,
       "step": 480
     },
     {
-      "epoch": 1.749090909090909,
-      "grad_norm": 5.2754082679748535,
-      "learning_rate": 2.090909090909091e-05,
-      "loss": 4.5604,
+      "epoch": 1.736462093862816,
+      "grad_norm": 3.8367581367492676,
+      "learning_rate": 2.111913357400722e-05,
+      "loss": 4.7117,
       "step": 481
     },
     {
-      "epoch": 1.7527272727272727,
-      "grad_norm": 4.733824253082275,
-      "learning_rate": 2.084848484848485e-05,
-      "loss": 4.3724,
+      "epoch": 1.7400722021660648,
+      "grad_norm": 2.3137428760528564,
+      "learning_rate": 2.1058965102286405e-05,
+      "loss": 4.5121,
       "step": 482
     },
     {
-      "epoch": 1.7563636363636363,
-      "grad_norm": 4.327937126159668,
-      "learning_rate": 2.078787878787879e-05,
-      "loss": 4.6649,
+      "epoch": 1.743682310469314,
+      "grad_norm": 3.6655476093292236,
+      "learning_rate": 2.0998796630565585e-05,
+      "loss": 4.6839,
       "step": 483
     },
     {
-      "epoch": 1.76,
-      "grad_norm": 3.6761820316314697,
-      "learning_rate": 2.0727272727272728e-05,
-      "loss": 4.466,
+      "epoch": 1.7472924187725631,
+      "grad_norm": 2.493283987045288,
+      "learning_rate": 2.0938628158844766e-05,
+      "loss": 4.4769,
       "step": 484
     },
     {
-      "epoch": 1.7636363636363637,
-      "grad_norm": 3.9023444652557373,
-      "learning_rate": 2.0666666666666666e-05,
-      "loss": 4.5215,
+      "epoch": 1.7509025270758123,
+      "grad_norm": 6.7742919921875,
+      "learning_rate": 2.0878459687123946e-05,
+      "loss": 4.7192,
       "step": 485
     },
     {
-      "epoch": 1.767272727272727,
-      "grad_norm": 9.362757682800293,
-      "learning_rate": 2.0606060606060608e-05,
-      "loss": 4.31,
+      "epoch": 1.7545126353790614,
+      "grad_norm": 5.782177925109863,
+      "learning_rate": 2.081829121540313e-05,
+      "loss": 4.6141,
       "step": 486
     },
     {
-      "epoch": 1.770909090909091,
-      "grad_norm": 6.796205043792725,
-      "learning_rate": 2.0545454545454546e-05,
-      "loss": 4.4703,
+      "epoch": 1.7581227436823106,
+      "grad_norm": 3.9292287826538086,
+      "learning_rate": 2.075812274368231e-05,
+      "loss": 4.4728,
       "step": 487
     },
     {
-      "epoch": 1.7745454545454544,
-      "grad_norm": 5.368844509124756,
-      "learning_rate": 2.0484848484848485e-05,
-      "loss": 4.5281,
+      "epoch": 1.7617328519855595,
+      "grad_norm": 5.2916717529296875,
+      "learning_rate": 2.069795427196149e-05,
+      "loss": 4.5358,
       "step": 488
     },
     {
-      "epoch": 1.7781818181818183,
-      "grad_norm": 4.443175315856934,
-      "learning_rate": 2.0424242424242427e-05,
-      "loss": 4.5208,
+      "epoch": 1.7653429602888087,
+      "grad_norm": 9.929496765136719,
+      "learning_rate": 2.0637785800240675e-05,
+      "loss": 4.8103,
       "step": 489
     },
     {
-      "epoch": 1.7818181818181817,
-      "grad_norm": 4.688342094421387,
-      "learning_rate": 2.0363636363636365e-05,
-      "loss": 4.5241,
+      "epoch": 1.7689530685920578,
+      "grad_norm": 4.576100826263428,
+      "learning_rate": 2.057761732851986e-05,
+      "loss": 4.4048,
       "step": 490
     },
     {
-      "epoch": 1.7854545454545454,
-      "grad_norm": 5.826910972595215,
-      "learning_rate": 2.0303030303030303e-05,
-      "loss": 4.5866,
+      "epoch": 1.7725631768953067,
+      "grad_norm": 3.692178964614868,
+      "learning_rate": 2.0517448856799036e-05,
+      "loss": 4.6001,
       "step": 491
     },
     {
-      "epoch": 1.789090909090909,
-      "grad_norm": 4.945174217224121,
-      "learning_rate": 2.0242424242424245e-05,
-      "loss": 4.5894,
+      "epoch": 1.7761732851985559,
+      "grad_norm": 2.9772250652313232,
+      "learning_rate": 2.045728038507822e-05,
+      "loss": 4.345,
       "step": 492
     },
     {
-      "epoch": 1.7927272727272727,
-      "grad_norm": 3.059390068054199,
-      "learning_rate": 2.0181818181818183e-05,
-      "loss": 4.5118,
+      "epoch": 1.779783393501805,
+      "grad_norm": 4.7923407554626465,
+      "learning_rate": 2.0397111913357404e-05,
+      "loss": 4.5239,
       "step": 493
     },
     {
-      "epoch": 1.7963636363636364,
-      "grad_norm": 3.689739942550659,
-      "learning_rate": 2.012121212121212e-05,
-      "loss": 4.4408,
+      "epoch": 1.7833935018050542,
+      "grad_norm": 3.105140209197998,
+      "learning_rate": 2.033694344163658e-05,
+      "loss": 4.4268,
       "step": 494
     },
     {
-      "epoch": 1.8,
-      "grad_norm": 4.748984336853027,
-      "learning_rate": 2.006060606060606e-05,
-      "loss": 4.4197,
+      "epoch": 1.7870036101083033,
+      "grad_norm": 2.8429927825927734,
+      "learning_rate": 2.0276774969915765e-05,
+      "loss": 4.4264,
       "step": 495
     },
     {
-      "epoch": 1.8036363636363637,
-      "grad_norm": 2.963027000427246,
-      "learning_rate": 2e-05,
-      "loss": 4.4785,
+      "epoch": 1.7906137184115525,
+      "grad_norm": 3.325678825378418,
+      "learning_rate": 2.0216606498194946e-05,
+      "loss": 4.423,
       "step": 496
     },
     {
-      "epoch": 1.8072727272727271,
-      "grad_norm": 2.5443899631500244,
-      "learning_rate": 1.993939393939394e-05,
-      "loss": 4.5684,
+      "epoch": 1.7942238267148014,
+      "grad_norm": 4.431451320648193,
+      "learning_rate": 2.015643802647413e-05,
+      "loss": 4.6256,
       "step": 497
     },
     {
-      "epoch": 1.810909090909091,
-      "grad_norm": 4.136449813842773,
-      "learning_rate": 1.987878787878788e-05,
-      "loss": 4.3071,
+      "epoch": 1.7978339350180506,
+      "grad_norm": 6.648149490356445,
+      "learning_rate": 2.009626955475331e-05,
+      "loss": 4.4858,
       "step": 498
     },
     {
-      "epoch": 1.8145454545454545,
-      "grad_norm": 4.769286155700684,
-      "learning_rate": 1.981818181818182e-05,
-      "loss": 4.3815,
+      "epoch": 1.8014440433212995,
+      "grad_norm": 3.306225538253784,
+      "learning_rate": 2.003610108303249e-05,
+      "loss": 4.6129,
       "step": 499
     },
     {
-      "epoch": 1.8181818181818183,
-      "grad_norm": 5.660130500793457,
-      "learning_rate": 1.975757575757576e-05,
-      "loss": 4.4983,
+      "epoch": 1.8050541516245486,
+      "grad_norm": 3.573742389678955,
+      "learning_rate": 1.9975932611311675e-05,
+      "loss": 4.4521,
       "step": 500
     },
     {
-      "epoch": 1.8218181818181818,
-      "grad_norm": 3.919520616531372,
-      "learning_rate": 1.9696969696969697e-05,
-      "loss": 4.4176,
+      "epoch": 1.8086642599277978,
+      "grad_norm": 2.5690858364105225,
+      "learning_rate": 1.9915764139590855e-05,
+      "loss": 4.339,
       "step": 501
     },
     {
-      "epoch": 1.8254545454545454,
-      "grad_norm": 4.681278228759766,
-      "learning_rate": 1.9636363636363635e-05,
-      "loss": 4.4025,
+      "epoch": 1.812274368231047,
+      "grad_norm": 5.37463903427124,
+      "learning_rate": 1.9855595667870036e-05,
+      "loss": 4.5435,
       "step": 502
     },
     {
-      "epoch": 1.829090909090909,
-      "grad_norm": 3.8179829120635986,
-      "learning_rate": 1.9575757575757577e-05,
-      "loss": 4.4669,
+      "epoch": 1.815884476534296,
+      "grad_norm": 2.394864082336426,
+      "learning_rate": 1.979542719614922e-05,
+      "loss": 4.3385,
       "step": 503
     },
     {
-      "epoch": 1.8327272727272728,
-      "grad_norm": 3.794281005859375,
-      "learning_rate": 1.951515151515152e-05,
-      "loss": 4.4807,
+      "epoch": 1.8194945848375452,
+      "grad_norm": 3.3190903663635254,
+      "learning_rate": 1.97352587244284e-05,
+      "loss": 4.4239,
       "step": 504
     },
     {
-      "epoch": 1.8363636363636364,
-      "grad_norm": 2.3946359157562256,
-      "learning_rate": 1.9454545454545457e-05,
-      "loss": 4.4092,
+      "epoch": 1.8231046931407944,
+      "grad_norm": 4.373250484466553,
+      "learning_rate": 1.967509025270758e-05,
+      "loss": 4.4304,
       "step": 505
     },
     {
-      "epoch": 1.8399999999999999,
-      "grad_norm": 4.364572525024414,
-      "learning_rate": 1.9393939393939395e-05,
-      "loss": 4.3869,
+      "epoch": 1.8267148014440433,
+      "grad_norm": 2.63519287109375,
+      "learning_rate": 1.9614921780986764e-05,
+      "loss": 4.5574,
       "step": 506
     },
     {
-      "epoch": 1.8436363636363637,
-      "grad_norm": 2.095834493637085,
-      "learning_rate": 1.9333333333333333e-05,
-      "loss": 4.5834,
+      "epoch": 1.8303249097472925,
+      "grad_norm": 5.139864444732666,
+      "learning_rate": 1.955475330926595e-05,
+      "loss": 4.4422,
       "step": 507
     },
     {
-      "epoch": 1.8472727272727272,
-      "grad_norm": 5.8669114112854,
-      "learning_rate": 1.9272727272727272e-05,
-      "loss": 4.6442,
+      "epoch": 1.8339350180505414,
+      "grad_norm": 4.551270961761475,
+      "learning_rate": 1.9494584837545125e-05,
+      "loss": 4.6433,
       "step": 508
     },
     {
-      "epoch": 1.850909090909091,
-      "grad_norm": 3.7711021900177,
-      "learning_rate": 1.9212121212121213e-05,
-      "loss": 4.3915,
+      "epoch": 1.8375451263537905,
+      "grad_norm": 3.421534299850464,
+      "learning_rate": 1.943441636582431e-05,
+      "loss": 4.4937,
       "step": 509
     },
     {
-      "epoch": 1.8545454545454545,
-      "grad_norm": 3.1917080879211426,
-      "learning_rate": 1.9151515151515155e-05,
-      "loss": 4.4269,
+      "epoch": 1.8411552346570397,
+      "grad_norm": 3.374833822250366,
+      "learning_rate": 1.937424789410349e-05,
+      "loss": 4.544,
       "step": 510
     },
     {
-      "epoch": 1.8581818181818182,
-      "grad_norm": 3.4579732418060303,
-      "learning_rate": 1.9090909090909094e-05,
-      "loss": 4.248,
+      "epoch": 1.8447653429602888,
+      "grad_norm": 3.4238624572753906,
+      "learning_rate": 1.9314079422382674e-05,
+      "loss": 4.4343,
       "step": 511
     },
     {
-      "epoch": 1.8618181818181818,
-      "grad_norm": 2.2174642086029053,
-      "learning_rate": 1.9030303030303032e-05,
-      "loss": 4.2461,
+      "epoch": 1.848375451263538,
+      "grad_norm": 2.8447346687316895,
+      "learning_rate": 1.9253910950661854e-05,
+      "loss": 4.3405,
       "step": 512
     },
     {
-      "epoch": 1.8654545454545455,
-      "grad_norm": 9.098743438720703,
-      "learning_rate": 1.896969696969697e-05,
-      "loss": 4.3802,
+      "epoch": 1.8519855595667871,
+      "grad_norm": 2.3235862255096436,
+      "learning_rate": 1.9193742478941035e-05,
+      "loss": 4.3878,
       "step": 513
     },
     {
-      "epoch": 1.8690909090909091,
-      "grad_norm": 4.291934967041016,
-      "learning_rate": 1.890909090909091e-05,
-      "loss": 4.5362,
+      "epoch": 1.855595667870036,
+      "grad_norm": 2.4331719875335693,
+      "learning_rate": 1.913357400722022e-05,
+      "loss": 4.3489,
       "step": 514
     },
     {
-      "epoch": 1.8727272727272726,
-      "grad_norm": 4.486403942108154,
-      "learning_rate": 1.884848484848485e-05,
-      "loss": 4.4608,
+      "epoch": 1.8592057761732852,
+      "grad_norm": 3.433101177215576,
+      "learning_rate": 1.90734055354994e-05,
+      "loss": 4.5183,
       "step": 515
     },
     {
-      "epoch": 1.8763636363636365,
-      "grad_norm": 3.7184762954711914,
-      "learning_rate": 1.878787878787879e-05,
-      "loss": 4.4732,
+      "epoch": 1.8628158844765343,
+      "grad_norm": 4.398335933685303,
+      "learning_rate": 1.901323706377858e-05,
+      "loss": 4.6618,
       "step": 516
     },
     {
-      "epoch": 1.88,
-      "grad_norm": 3.1857259273529053,
-      "learning_rate": 1.872727272727273e-05,
-      "loss": 4.3346,
+      "epoch": 1.8664259927797833,
+      "grad_norm": 3.5990779399871826,
+      "learning_rate": 1.8953068592057764e-05,
+      "loss": 4.4411,
       "step": 517
     },
     {
-      "epoch": 1.8836363636363638,
-      "grad_norm": 5.259026050567627,
-      "learning_rate": 1.866666666666667e-05,
-      "loss": 4.5613,
+      "epoch": 1.8700361010830324,
+      "grad_norm": 2.9849636554718018,
+      "learning_rate": 1.8892900120336944e-05,
+      "loss": 4.5298,
       "step": 518
     },
     {
-      "epoch": 1.8872727272727272,
-      "grad_norm": 4.3462233543396,
-      "learning_rate": 1.8606060606060607e-05,
-      "loss": 4.5914,
+      "epoch": 1.8736462093862816,
+      "grad_norm": 3.6262741088867188,
+      "learning_rate": 1.8832731648616125e-05,
+      "loss": 4.514,
       "step": 519
     },
     {
-      "epoch": 1.8909090909090909,
-      "grad_norm": 2.839916944503784,
-      "learning_rate": 1.8545454545454545e-05,
-      "loss": 4.4223,
+      "epoch": 1.8772563176895307,
+      "grad_norm": 4.894564151763916,
+      "learning_rate": 1.877256317689531e-05,
+      "loss": 4.4753,
       "step": 520
     },
     {
-      "epoch": 1.8945454545454545,
-      "grad_norm": 3.4585793018341064,
-      "learning_rate": 1.8484848484848487e-05,
-      "loss": 4.4892,
+      "epoch": 1.8808664259927799,
+      "grad_norm": 4.75285530090332,
+      "learning_rate": 1.871239470517449e-05,
+      "loss": 4.4204,
       "step": 521
     },
     {
-      "epoch": 1.8981818181818182,
-      "grad_norm": 4.047860145568848,
-      "learning_rate": 1.8424242424242425e-05,
-      "loss": 4.4238,
+      "epoch": 1.884476534296029,
+      "grad_norm": 4.635976791381836,
+      "learning_rate": 1.865222623345367e-05,
+      "loss": 4.4292,
       "step": 522
     },
     {
-      "epoch": 1.9018181818181819,
-      "grad_norm": 3.6828198432922363,
-      "learning_rate": 1.8363636363636364e-05,
-      "loss": 4.4123,
+      "epoch": 1.888086642599278,
+      "grad_norm": 3.461372137069702,
+      "learning_rate": 1.8592057761732854e-05,
+      "loss": 4.2734,
       "step": 523
     },
     {
-      "epoch": 1.9054545454545453,
-      "grad_norm": 3.2874972820281982,
-      "learning_rate": 1.8303030303030305e-05,
-      "loss": 4.4527,
+      "epoch": 1.891696750902527,
+      "grad_norm": 7.233281135559082,
+      "learning_rate": 1.8531889290012034e-05,
+      "loss": 4.6022,
       "step": 524
     },
     {
-      "epoch": 1.9090909090909092,
-      "grad_norm": 2.8080523014068604,
-      "learning_rate": 1.8242424242424244e-05,
-      "loss": 4.3722,
+      "epoch": 1.895306859205776,
+      "grad_norm": 6.590332508087158,
+      "learning_rate": 1.8471720818291215e-05,
+      "loss": 4.5391,
       "step": 525
     },
     {
-      "epoch": 1.9127272727272726,
-      "grad_norm": 3.2030367851257324,
-      "learning_rate": 1.8181818181818182e-05,
-      "loss": 4.5094,
+      "epoch": 1.8989169675090252,
+      "grad_norm": 6.628108024597168,
+      "learning_rate": 1.84115523465704e-05,
+      "loss": 4.6294,
       "step": 526
     },
     {
-      "epoch": 1.9163636363636365,
-      "grad_norm": 4.297696590423584,
-      "learning_rate": 1.8121212121212124e-05,
-      "loss": 4.392,
+      "epoch": 1.9025270758122743,
+      "grad_norm": 6.2744460105896,
+      "learning_rate": 1.835138387484958e-05,
+      "loss": 4.5166,
       "step": 527
     },
     {
-      "epoch": 1.92,
-      "grad_norm": 6.101680755615234,
-      "learning_rate": 1.8060606060606062e-05,
-      "loss": 4.502,
+      "epoch": 1.9061371841155235,
+      "grad_norm": 8.246423721313477,
+      "learning_rate": 1.8291215403128763e-05,
+      "loss": 4.4344,
       "step": 528
     },
     {
-      "epoch": 1.9236363636363636,
-      "grad_norm": 4.8301215171813965,
-      "learning_rate": 1.8e-05,
-      "loss": 4.4693,
+      "epoch": 1.9097472924187726,
+      "grad_norm": 5.218391418457031,
+      "learning_rate": 1.8231046931407943e-05,
+      "loss": 4.4208,
       "step": 529
     },
     {
-      "epoch": 1.9272727272727272,
-      "grad_norm": 4.521356582641602,
-      "learning_rate": 1.793939393939394e-05,
-      "loss": 4.478,
+      "epoch": 1.9133574007220218,
+      "grad_norm": 2.6028380393981934,
+      "learning_rate": 1.8170878459687124e-05,
+      "loss": 4.4997,
       "step": 530
     },
     {
-      "epoch": 1.930909090909091,
-      "grad_norm": 3.493539571762085,
-      "learning_rate": 1.787878787878788e-05,
-      "loss": 4.4816,
+      "epoch": 1.916967509025271,
+      "grad_norm": 8.268327713012695,
+      "learning_rate": 1.8110709987966308e-05,
+      "loss": 4.3713,
       "step": 531
     },
     {
-      "epoch": 1.9345454545454546,
-      "grad_norm": 3.7430574893951416,
-      "learning_rate": 1.781818181818182e-05,
-      "loss": 4.5463,
+      "epoch": 1.9205776173285198,
+      "grad_norm": 4.178406238555908,
+      "learning_rate": 1.805054151624549e-05,
+      "loss": 4.4769,
       "step": 532
     },
     {
-      "epoch": 1.9381818181818182,
-      "grad_norm": 3.7508163452148438,
-      "learning_rate": 1.775757575757576e-05,
-      "loss": 4.293,
+      "epoch": 1.924187725631769,
+      "grad_norm": 3.8857920169830322,
+      "learning_rate": 1.799037304452467e-05,
+      "loss": 4.6029,
       "step": 533
     },
     {
-      "epoch": 1.9418181818181819,
-      "grad_norm": 6.832218170166016,
-      "learning_rate": 1.76969696969697e-05,
-      "loss": 4.5343,
+      "epoch": 1.927797833935018,
+      "grad_norm": 3.7289071083068848,
+      "learning_rate": 1.7930204572803853e-05,
+      "loss": 4.3475,
       "step": 534
     },
     {
-      "epoch": 1.9454545454545453,
-      "grad_norm": 3.9975361824035645,
-      "learning_rate": 1.7636363636363637e-05,
-      "loss": 4.3917,
+      "epoch": 1.931407942238267,
+      "grad_norm": 3.547478199005127,
+      "learning_rate": 1.7870036101083033e-05,
+      "loss": 4.4758,
       "step": 535
     },
     {
-      "epoch": 1.9490909090909092,
-      "grad_norm": 3.4691479206085205,
-      "learning_rate": 1.7575757575757576e-05,
-      "loss": 4.4656,
+      "epoch": 1.9350180505415162,
+      "grad_norm": 4.167591571807861,
+      "learning_rate": 1.7809867629362214e-05,
+      "loss": 4.4301,
       "step": 536
     },
     {
-      "epoch": 1.9527272727272726,
-      "grad_norm": 4.352736949920654,
-      "learning_rate": 1.7515151515151514e-05,
-      "loss": 4.3993,
+      "epoch": 1.9386281588447654,
+      "grad_norm": 3.6818253993988037,
+      "learning_rate": 1.7749699157641398e-05,
+      "loss": 4.3256,
       "step": 537
     },
     {
-      "epoch": 1.9563636363636365,
-      "grad_norm": 4.212103843688965,
-      "learning_rate": 1.7454545454545456e-05,
-      "loss": 4.3395,
+      "epoch": 1.9422382671480145,
+      "grad_norm": 6.802430152893066,
+      "learning_rate": 1.768953068592058e-05,
+      "loss": 4.4358,
       "step": 538
     },
     {
-      "epoch": 1.96,
-      "grad_norm": 4.381357192993164,
-      "learning_rate": 1.7393939393939397e-05,
-      "loss": 4.3113,
+      "epoch": 1.9458483754512637,
+      "grad_norm": 7.140280723571777,
+      "learning_rate": 1.762936221419976e-05,
+      "loss": 4.4219,
       "step": 539
     },
     {
-      "epoch": 1.9636363636363636,
-      "grad_norm": 4.0517706871032715,
-      "learning_rate": 1.7333333333333336e-05,
-      "loss": 4.4248,
+      "epoch": 1.9494584837545126,
+      "grad_norm": 7.254362106323242,
+      "learning_rate": 1.7569193742478943e-05,
+      "loss": 4.4482,
       "step": 540
     },
     {
-      "epoch": 1.9672727272727273,
-      "grad_norm": 3.827972650527954,
-      "learning_rate": 1.7272727272727274e-05,
-      "loss": 4.2242,
+      "epoch": 1.9530685920577617,
+      "grad_norm": 4.236722469329834,
+      "learning_rate": 1.7509025270758123e-05,
+      "loss": 4.624,
       "step": 541
     },
     {
-      "epoch": 1.970909090909091,
-      "grad_norm": 3.4165265560150146,
-      "learning_rate": 1.7212121212121212e-05,
-      "loss": 4.4312,
+      "epoch": 1.9566787003610109,
+      "grad_norm": 5.225027561187744,
+      "learning_rate": 1.7448856799037304e-05,
+      "loss": 4.3267,
       "step": 542
     },
     {
-      "epoch": 1.9745454545454546,
-      "grad_norm": 15.285648345947266,
-      "learning_rate": 1.715151515151515e-05,
-      "loss": 4.5811,
+      "epoch": 1.9602888086642598,
+      "grad_norm": 6.063048362731934,
+      "learning_rate": 1.7388688327316488e-05,
+      "loss": 4.2722,
       "step": 543
     },
     {
-      "epoch": 1.978181818181818,
-      "grad_norm": 19.892024993896484,
-      "learning_rate": 1.7090909090909092e-05,
-      "loss": 4.467,
+      "epoch": 1.963898916967509,
+      "grad_norm": 3.3228302001953125,
+      "learning_rate": 1.7328519855595668e-05,
+      "loss": 4.4031,
       "step": 544
     },
     {
-      "epoch": 1.981818181818182,
-      "grad_norm": 7.771790981292725,
-      "learning_rate": 1.703030303030303e-05,
-      "loss": 4.4484,
+      "epoch": 1.967509025270758,
+      "grad_norm": 5.654480934143066,
+      "learning_rate": 1.7268351383874852e-05,
+      "loss": 4.5082,
       "step": 545
     },
     {
-      "epoch": 1.9854545454545454,
-      "grad_norm": 4.017660140991211,
-      "learning_rate": 1.6969696969696972e-05,
-      "loss": 4.3429,
+      "epoch": 1.9711191335740073,
+      "grad_norm": 3.153670072555542,
+      "learning_rate": 1.7208182912154033e-05,
+      "loss": 4.547,
       "step": 546
     },
     {
-      "epoch": 1.9890909090909092,
-      "grad_norm": 5.259058475494385,
-      "learning_rate": 1.690909090909091e-05,
-      "loss": 4.4623,
+      "epoch": 1.9747292418772564,
+      "grad_norm": 3.0730340480804443,
+      "learning_rate": 1.7148014440433213e-05,
+      "loss": 4.3911,
       "step": 547
     },
     {
-      "epoch": 1.9927272727272727,
-      "grad_norm": 10.823943138122559,
-      "learning_rate": 1.684848484848485e-05,
-      "loss": 4.5293,
+      "epoch": 1.9783393501805056,
+      "grad_norm": 6.021193981170654,
+      "learning_rate": 1.7087845968712397e-05,
+      "loss": 4.7361,
       "step": 548
     },
     {
-      "epoch": 1.9963636363636363,
-      "grad_norm": 6.573860168457031,
-      "learning_rate": 1.6787878787878787e-05,
-      "loss": 4.6044,
+      "epoch": 1.9819494584837545,
+      "grad_norm": 4.3618550300598145,
+      "learning_rate": 1.7027677496991578e-05,
+      "loss": 4.4272,
       "step": 549
     },
     {
-      "epoch": 2.0,
-      "grad_norm": 5.922011375427246,
-      "learning_rate": 1.672727272727273e-05,
-      "loss": 4.3309,
+      "epoch": 1.9855595667870036,
+      "grad_norm": 6.582357883453369,
+      "learning_rate": 1.6967509025270758e-05,
+      "loss": 4.4994,
       "step": 550
     },
     {
-      "epoch": 2.0036363636363634,
-      "grad_norm": 3.63747501373291,
-      "learning_rate": 1.6666666666666667e-05,
-      "loss": 4.4179,
+      "epoch": 1.9891696750902526,
+      "grad_norm": 3.4479753971099854,
+      "learning_rate": 1.6907340553549942e-05,
+      "loss": 4.6324,
       "step": 551
     },
     {
-      "epoch": 2.0072727272727273,
-      "grad_norm": 6.124152183532715,
-      "learning_rate": 1.6606060606060606e-05,
-      "loss": 4.612,
+      "epoch": 1.9927797833935017,
+      "grad_norm": 6.284206867218018,
+      "learning_rate": 1.6847172081829123e-05,
+      "loss": 4.5219,
       "step": 552
     },
     {
-      "epoch": 2.0109090909090908,
-      "grad_norm": 4.838201522827148,
-      "learning_rate": 1.6545454545454548e-05,
-      "loss": 4.4896,
+      "epoch": 1.9963898916967509,
+      "grad_norm": 8.223616600036621,
+      "learning_rate": 1.6787003610108303e-05,
+      "loss": 4.4365,
       "step": 553
     },
     {
-      "epoch": 2.0145454545454546,
-      "grad_norm": 5.851705074310303,
-      "learning_rate": 1.6484848484848486e-05,
-      "loss": 4.3424,
+      "epoch": 2.0,
+      "grad_norm": 4.31699275970459,
+      "learning_rate": 1.6726835138387487e-05,
+      "loss": 4.2858,
       "step": 554
     },
     {
-      "epoch": 2.018181818181818,
-      "grad_norm": 4.73307991027832,
-      "learning_rate": 1.6424242424242424e-05,
-      "loss": 4.4876,
+      "epoch": 2.003610108303249,
+      "grad_norm": 4.425889015197754,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 4.4514,
       "step": 555
     },
     {
-      "epoch": 2.021818181818182,
-      "grad_norm": 3.129345417022705,
-      "learning_rate": 1.6363636363636366e-05,
-      "loss": 4.4337,
+      "epoch": 2.0072202166064983,
+      "grad_norm": 4.568549633026123,
+      "learning_rate": 1.6606498194945848e-05,
+      "loss": 4.4904,
       "step": 556
     },
     {
-      "epoch": 2.0254545454545454,
-      "grad_norm": 4.917591094970703,
-      "learning_rate": 1.6303030303030304e-05,
-      "loss": 4.6596,
+      "epoch": 2.0108303249097474,
+      "grad_norm": 4.659919261932373,
+      "learning_rate": 1.6546329723225032e-05,
+      "loss": 4.3659,
       "step": 557
     },
     {
-      "epoch": 2.0290909090909093,
-      "grad_norm": 6.358683109283447,
-      "learning_rate": 1.6242424242424243e-05,
-      "loss": 4.5853,
+      "epoch": 2.0144404332129966,
+      "grad_norm": 3.3128418922424316,
+      "learning_rate": 1.6486161251504212e-05,
+      "loss": 4.5661,
       "step": 558
     },
     {
-      "epoch": 2.0327272727272727,
-      "grad_norm": 4.329516887664795,
-      "learning_rate": 1.618181818181818e-05,
-      "loss": 4.3207,
+      "epoch": 2.0180505415162453,
+      "grad_norm": 5.247345924377441,
+      "learning_rate": 1.6425992779783393e-05,
+      "loss": 4.5473,
       "step": 559
     },
     {
-      "epoch": 2.036363636363636,
-      "grad_norm": 2.4272570610046387,
-      "learning_rate": 1.6121212121212123e-05,
-      "loss": 4.3553,
+      "epoch": 2.0216606498194944,
+      "grad_norm": 2.952227830886841,
+      "learning_rate": 1.6365824308062577e-05,
+      "loss": 4.4017,
       "step": 560
     },
     {
-      "epoch": 2.04,
-      "grad_norm": 3.0999836921691895,
-      "learning_rate": 1.606060606060606e-05,
-      "loss": 4.5796,
+      "epoch": 2.0252707581227436,
+      "grad_norm": 4.103919506072998,
+      "learning_rate": 1.6305655836341757e-05,
+      "loss": 4.5626,
       "step": 561
     },
     {
-      "epoch": 2.0436363636363635,
-      "grad_norm": 4.398182392120361,
-      "learning_rate": 1.6000000000000003e-05,
-      "loss": 4.4053,
+      "epoch": 2.0288808664259927,
+      "grad_norm": 3.0998098850250244,
+      "learning_rate": 1.624548736462094e-05,
+      "loss": 4.6575,
       "step": 562
     },
     {
-      "epoch": 2.0472727272727274,
-      "grad_norm": 4.124515533447266,
-      "learning_rate": 1.593939393939394e-05,
-      "loss": 4.5866,
+      "epoch": 2.032490974729242,
+      "grad_norm": 3.9067933559417725,
+      "learning_rate": 1.618531889290012e-05,
+      "loss": 4.5003,
       "step": 563
     },
     {
-      "epoch": 2.050909090909091,
-      "grad_norm": 10.85811710357666,
-      "learning_rate": 1.587878787878788e-05,
-      "loss": 4.5517,
+      "epoch": 2.036101083032491,
+      "grad_norm": 5.084564208984375,
+      "learning_rate": 1.6125150421179302e-05,
+      "loss": 4.593,
       "step": 564
     },
     {
-      "epoch": 2.0545454545454547,
-      "grad_norm": 3.445768117904663,
-      "learning_rate": 1.5818181818181818e-05,
-      "loss": 4.3176,
+      "epoch": 2.03971119133574,
+      "grad_norm": 2.939908027648926,
+      "learning_rate": 1.6064981949458486e-05,
+      "loss": 4.4549,
       "step": 565
     },
     {
-      "epoch": 2.058181818181818,
-      "grad_norm": 3.5251712799072266,
-      "learning_rate": 1.5757575757575756e-05,
-      "loss": 4.3987,
+      "epoch": 2.0433212996389893,
+      "grad_norm": 4.826956272125244,
+      "learning_rate": 1.6004813477737667e-05,
+      "loss": 4.5792,
       "step": 566
     },
     {
-      "epoch": 2.061818181818182,
-      "grad_norm": 4.528774738311768,
-      "learning_rate": 1.5696969696969698e-05,
-      "loss": 4.5358,
+      "epoch": 2.046931407942238,
+      "grad_norm": 4.092348575592041,
+      "learning_rate": 1.5944645006016847e-05,
+      "loss": 4.5905,
       "step": 567
     },
     {
-      "epoch": 2.0654545454545454,
-      "grad_norm": 8.901700019836426,
-      "learning_rate": 1.563636363636364e-05,
-      "loss": 4.6986,
+      "epoch": 2.050541516245487,
+      "grad_norm": 5.71950626373291,
+      "learning_rate": 1.588447653429603e-05,
+      "loss": 4.429,
       "step": 568
     },
     {
-      "epoch": 2.0690909090909093,
-      "grad_norm": 4.355597019195557,
-      "learning_rate": 1.5575757575757578e-05,
-      "loss": 4.3603,
+      "epoch": 2.0541516245487363,
+      "grad_norm": 3.808638572692871,
+      "learning_rate": 1.582430806257521e-05,
+      "loss": 4.4797,
       "step": 569
     },
     {
-      "epoch": 2.0727272727272728,
-      "grad_norm": 3.189401149749756,
-      "learning_rate": 1.5515151515151516e-05,
-      "loss": 4.3137,
+      "epoch": 2.0577617328519855,
+      "grad_norm": 4.150345325469971,
+      "learning_rate": 1.5764139590854392e-05,
+      "loss": 4.5639,
       "step": 570
     },
     {
-      "epoch": 2.076363636363636,
-      "grad_norm": 5.457395076751709,
-      "learning_rate": 1.5454545454545454e-05,
-      "loss": 4.5963,
+      "epoch": 2.0613718411552346,
+      "grad_norm": 4.915743350982666,
+      "learning_rate": 1.5703971119133576e-05,
+      "loss": 4.4055,
       "step": 571
     },
     {
-      "epoch": 2.08,
-      "grad_norm": 5.1099934577941895,
-      "learning_rate": 1.5393939393939393e-05,
-      "loss": 4.3938,
+      "epoch": 2.064981949458484,
+      "grad_norm": 5.808039665222168,
+      "learning_rate": 1.5643802647412757e-05,
+      "loss": 4.521,
       "step": 572
     },
     {
-      "epoch": 2.0836363636363635,
-      "grad_norm": 4.458206653594971,
-      "learning_rate": 1.5333333333333334e-05,
-      "loss": 4.3399,
+      "epoch": 2.068592057761733,
+      "grad_norm": 5.758335113525391,
+      "learning_rate": 1.5583634175691937e-05,
+      "loss": 4.5548,
       "step": 573
     },
     {
-      "epoch": 2.0872727272727274,
-      "grad_norm": 3.7500548362731934,
-      "learning_rate": 1.5272727272727276e-05,
-      "loss": 4.2681,
+      "epoch": 2.072202166064982,
+      "grad_norm": 5.1010422706604,
+      "learning_rate": 1.552346570397112e-05,
+      "loss": 4.3622,
       "step": 574
     },
     {
-      "epoch": 2.090909090909091,
-      "grad_norm": 3.342282295227051,
-      "learning_rate": 1.5212121212121213e-05,
-      "loss": 4.4396,
+      "epoch": 2.0758122743682312,
+      "grad_norm": 4.795560836791992,
+      "learning_rate": 1.54632972322503e-05,
+      "loss": 4.5221,
       "step": 575
     },
     {
-      "epoch": 2.0945454545454547,
-      "grad_norm": 5.0651092529296875,
-      "learning_rate": 1.5151515151515153e-05,
-      "loss": 4.5387,
+      "epoch": 2.07942238267148,
+      "grad_norm": 3.5267210006713867,
+      "learning_rate": 1.5403128760529485e-05,
+      "loss": 4.4669,
       "step": 576
     },
     {
-      "epoch": 2.098181818181818,
-      "grad_norm": 4.204915523529053,
-      "learning_rate": 1.5090909090909091e-05,
-      "loss": 4.4134,
+      "epoch": 2.083032490974729,
+      "grad_norm": 4.934731960296631,
+      "learning_rate": 1.5342960288808663e-05,
+      "loss": 4.4068,
       "step": 577
     },
     {
-      "epoch": 2.101818181818182,
-      "grad_norm": 3.690837860107422,
-      "learning_rate": 1.5030303030303031e-05,
-      "loss": 4.3563,
+      "epoch": 2.0866425992779782,
+      "grad_norm": 2.6810762882232666,
+      "learning_rate": 1.5282791817087847e-05,
+      "loss": 4.6031,
       "step": 578
     },
     {
-      "epoch": 2.1054545454545455,
-      "grad_norm": 3.8858180046081543,
-      "learning_rate": 1.496969696969697e-05,
-      "loss": 4.5194,
+      "epoch": 2.0902527075812274,
+      "grad_norm": 3.5045320987701416,
+      "learning_rate": 1.5222623345367029e-05,
+      "loss": 4.4259,
       "step": 579
     },
     {
-      "epoch": 2.109090909090909,
-      "grad_norm": 4.591732501983643,
-      "learning_rate": 1.4909090909090908e-05,
-      "loss": 4.5678,
+      "epoch": 2.0938628158844765,
+      "grad_norm": 4.098846435546875,
+      "learning_rate": 1.516245487364621e-05,
+      "loss": 4.3841,
       "step": 580
     },
     {
-      "epoch": 2.112727272727273,
-      "grad_norm": 2.2091257572174072,
-      "learning_rate": 1.484848484848485e-05,
-      "loss": 4.4385,
+      "epoch": 2.0974729241877257,
+      "grad_norm": 2.795208215713501,
+      "learning_rate": 1.5102286401925391e-05,
+      "loss": 4.4416,
       "step": 581
     },
     {
-      "epoch": 2.1163636363636362,
-      "grad_norm": 4.992697715759277,
-      "learning_rate": 1.478787878787879e-05,
-      "loss": 4.3667,
+      "epoch": 2.101083032490975,
+      "grad_norm": 5.189493179321289,
+      "learning_rate": 1.5042117930204574e-05,
+      "loss": 4.454,
       "step": 582
     },
     {
-      "epoch": 2.12,
-      "grad_norm": 4.349731922149658,
-      "learning_rate": 1.4727272727272728e-05,
-      "loss": 4.4421,
+      "epoch": 2.104693140794224,
+      "grad_norm": 3.3796346187591553,
+      "learning_rate": 1.4981949458483754e-05,
+      "loss": 4.581,
       "step": 583
     },
     {
-      "epoch": 2.1236363636363635,
-      "grad_norm": 3.431900978088379,
-      "learning_rate": 1.4666666666666668e-05,
-      "loss": 4.6203,
+      "epoch": 2.108303249097473,
+      "grad_norm": 3.021270513534546,
+      "learning_rate": 1.4921780986762936e-05,
+      "loss": 4.5721,
       "step": 584
     },
     {
-      "epoch": 2.1272727272727274,
-      "grad_norm": 7.2620062828063965,
-      "learning_rate": 1.4606060606060606e-05,
-      "loss": 4.4006,
+      "epoch": 2.111913357400722,
+      "grad_norm": 4.579837799072266,
+      "learning_rate": 1.486161251504212e-05,
+      "loss": 4.3886,
       "step": 585
     },
     {
-      "epoch": 2.130909090909091,
-      "grad_norm": 3.902266263961792,
-      "learning_rate": 1.4545454545454545e-05,
-      "loss": 4.6113,
+      "epoch": 2.115523465703971,
+      "grad_norm": 5.8261566162109375,
+      "learning_rate": 1.4801444043321299e-05,
+      "loss": 4.4218,
       "step": 586
     },
     {
-      "epoch": 2.1345454545454547,
-      "grad_norm": 3.3681981563568115,
-      "learning_rate": 1.4484848484848485e-05,
-      "loss": 4.212,
+      "epoch": 2.11913357400722,
+      "grad_norm": 3.2095069885253906,
+      "learning_rate": 1.4741275571600483e-05,
+      "loss": 4.3596,
       "step": 587
     },
     {
-      "epoch": 2.138181818181818,
-      "grad_norm": 3.5023534297943115,
-      "learning_rate": 1.4424242424242426e-05,
-      "loss": 4.23,
+      "epoch": 2.1227436823104693,
+      "grad_norm": 3.292442560195923,
+      "learning_rate": 1.4681107099879662e-05,
+      "loss": 4.4458,
       "step": 588
     },
     {
-      "epoch": 2.1418181818181816,
-      "grad_norm": 4.29148530960083,
-      "learning_rate": 1.4363636363636365e-05,
-      "loss": 4.5829,
+      "epoch": 2.1263537906137184,
+      "grad_norm": 5.677742958068848,
+      "learning_rate": 1.4620938628158846e-05,
+      "loss": 4.6095,
       "step": 589
     },
     {
-      "epoch": 2.1454545454545455,
-      "grad_norm": 2.9856648445129395,
-      "learning_rate": 1.4303030303030305e-05,
-      "loss": 4.4755,
+      "epoch": 2.1299638989169676,
+      "grad_norm": 3.2105185985565186,
+      "learning_rate": 1.4560770156438028e-05,
+      "loss": 4.333,
       "step": 590
     },
     {
-      "epoch": 2.149090909090909,
-      "grad_norm": 2.5396695137023926,
-      "learning_rate": 1.4242424242424243e-05,
-      "loss": 4.3903,
+      "epoch": 2.1335740072202167,
+      "grad_norm": 3.4787943363189697,
+      "learning_rate": 1.4500601684717209e-05,
+      "loss": 4.6703,
       "step": 591
     },
     {
-      "epoch": 2.152727272727273,
-      "grad_norm": 2.6252663135528564,
-      "learning_rate": 1.4181818181818181e-05,
-      "loss": 4.2966,
+      "epoch": 2.137184115523466,
+      "grad_norm": 2.8096866607666016,
+      "learning_rate": 1.444043321299639e-05,
+      "loss": 4.312,
       "step": 592
     },
     {
-      "epoch": 2.1563636363636363,
-      "grad_norm": 2.7963287830352783,
-      "learning_rate": 1.4121212121212121e-05,
-      "loss": 4.4912,
+      "epoch": 2.140794223826715,
+      "grad_norm": 3.680739164352417,
+      "learning_rate": 1.4380264741275573e-05,
+      "loss": 4.438,
       "step": 593
     },
     {
-      "epoch": 2.16,
-      "grad_norm": 4.679149150848389,
-      "learning_rate": 1.406060606060606e-05,
-      "loss": 4.6129,
+      "epoch": 2.1444043321299637,
+      "grad_norm": 3.65415620803833,
+      "learning_rate": 1.4320096269554753e-05,
+      "loss": 4.4032,
       "step": 594
     },
     {
-      "epoch": 2.1636363636363636,
-      "grad_norm": 4.0482869148254395,
-      "learning_rate": 1.4000000000000001e-05,
-      "loss": 4.5545,
+      "epoch": 2.148014440433213,
+      "grad_norm": 3.834942102432251,
+      "learning_rate": 1.4259927797833936e-05,
+      "loss": 4.499,
       "step": 595
     },
     {
-      "epoch": 2.1672727272727275,
-      "grad_norm": 2.992244005203247,
-      "learning_rate": 1.3939393939393942e-05,
-      "loss": 4.3336,
+      "epoch": 2.151624548736462,
+      "grad_norm": 2.716878890991211,
+      "learning_rate": 1.4199759326113118e-05,
+      "loss": 4.4475,
       "step": 596
     },
     {
-      "epoch": 2.170909090909091,
-      "grad_norm": 3.422591209411621,
-      "learning_rate": 1.387878787878788e-05,
-      "loss": 4.3752,
+      "epoch": 2.155234657039711,
+      "grad_norm": 3.9994101524353027,
+      "learning_rate": 1.4139590854392298e-05,
+      "loss": 4.4935,
       "step": 597
     },
     {
-      "epoch": 2.174545454545455,
-      "grad_norm": 3.826042652130127,
-      "learning_rate": 1.3818181818181818e-05,
-      "loss": 4.2918,
+      "epoch": 2.1588447653429603,
+      "grad_norm": 7.912623882293701,
+      "learning_rate": 1.407942238267148e-05,
+      "loss": 4.5646,
       "step": 598
     },
     {
-      "epoch": 2.178181818181818,
-      "grad_norm": 3.569441795349121,
-      "learning_rate": 1.3757575757575758e-05,
-      "loss": 4.4492,
+      "epoch": 2.1624548736462095,
+      "grad_norm": 5.011529445648193,
+      "learning_rate": 1.4019253910950665e-05,
+      "loss": 4.4788,
       "step": 599
     },
     {
-      "epoch": 2.1818181818181817,
-      "grad_norm": 3.8144829273223877,
-      "learning_rate": 1.3696969696969697e-05,
-      "loss": 4.3355,
+      "epoch": 2.1660649819494586,
+      "grad_norm": 4.918196201324463,
+      "learning_rate": 1.3959085439229843e-05,
+      "loss": 4.5599,
       "step": 600
     },
     {
-      "epoch": 2.1854545454545455,
-      "grad_norm": 3.672569990158081,
-      "learning_rate": 1.3636363636363637e-05,
-      "loss": 4.4492,
+      "epoch": 2.1696750902527078,
+      "grad_norm": 3.1637089252471924,
+      "learning_rate": 1.3898916967509026e-05,
+      "loss": 4.6395,
       "step": 601
     },
     {
-      "epoch": 2.189090909090909,
-      "grad_norm": 3.1186609268188477,
-      "learning_rate": 1.3575757575757578e-05,
-      "loss": 4.4051,
+      "epoch": 2.1732851985559565,
+      "grad_norm": 4.921415328979492,
+      "learning_rate": 1.3838748495788206e-05,
+      "loss": 4.3406,
       "step": 602
     },
     {
-      "epoch": 2.192727272727273,
-      "grad_norm": 2.243263006210327,
-      "learning_rate": 1.3515151515151517e-05,
-      "loss": 4.3317,
+      "epoch": 2.1768953068592056,
+      "grad_norm": 3.5986857414245605,
+      "learning_rate": 1.3778580024067388e-05,
+      "loss": 4.6468,
       "step": 603
     },
     {
-      "epoch": 2.1963636363636363,
-      "grad_norm": 3.677445650100708,
-      "learning_rate": 1.3454545454545457e-05,
-      "loss": 4.3382,
+      "epoch": 2.1805054151624548,
+      "grad_norm": 4.023871898651123,
+      "learning_rate": 1.3718411552346572e-05,
+      "loss": 4.4085,
       "step": 604
     },
     {
-      "epoch": 2.2,
-      "grad_norm": 2.7815980911254883,
-      "learning_rate": 1.3393939393939395e-05,
-      "loss": 4.4145,
+      "epoch": 2.184115523465704,
+      "grad_norm": 3.7364799976348877,
+      "learning_rate": 1.3658243080625751e-05,
+      "loss": 4.3773,
       "step": 605
     },
     {
-      "epoch": 2.2036363636363636,
-      "grad_norm": 4.602207183837891,
-      "learning_rate": 1.3333333333333333e-05,
-      "loss": 4.6324,
+      "epoch": 2.187725631768953,
+      "grad_norm": 3.5880355834960938,
+      "learning_rate": 1.3598074608904935e-05,
+      "loss": 4.3806,
       "step": 606
     },
     {
-      "epoch": 2.207272727272727,
-      "grad_norm": 4.399771690368652,
-      "learning_rate": 1.3272727272727273e-05,
-      "loss": 4.5507,
+      "epoch": 2.191335740072202,
+      "grad_norm": 3.8788704872131348,
+      "learning_rate": 1.3537906137184117e-05,
+      "loss": 4.3904,
       "step": 607
     },
     {
-      "epoch": 2.210909090909091,
-      "grad_norm": 4.964302062988281,
-      "learning_rate": 1.3212121212121212e-05,
-      "loss": 4.5691,
+      "epoch": 2.1949458483754514,
+      "grad_norm": 4.633315086364746,
+      "learning_rate": 1.3477737665463298e-05,
+      "loss": 4.7456,
       "step": 608
     },
     {
-      "epoch": 2.2145454545454544,
-      "grad_norm": 6.282228946685791,
-      "learning_rate": 1.315151515151515e-05,
-      "loss": 4.5208,
+      "epoch": 2.1985559566787005,
+      "grad_norm": 5.435410499572754,
+      "learning_rate": 1.341756919374248e-05,
+      "loss": 4.5496,
       "step": 609
     },
     {
-      "epoch": 2.2181818181818183,
-      "grad_norm": 4.12927770614624,
-      "learning_rate": 1.3090909090909093e-05,
-      "loss": 4.4183,
+      "epoch": 2.2021660649819497,
+      "grad_norm": 5.3415985107421875,
+      "learning_rate": 1.3357400722021662e-05,
+      "loss": 4.6066,
       "step": 610
     },
     {
-      "epoch": 2.2218181818181817,
-      "grad_norm": 3.556312322616577,
-      "learning_rate": 1.3030303030303032e-05,
-      "loss": 4.3785,
+      "epoch": 2.2057761732851984,
+      "grad_norm": 3.8118133544921875,
+      "learning_rate": 1.3297232250300843e-05,
+      "loss": 4.4713,
       "step": 611
     },
     {
-      "epoch": 2.2254545454545456,
-      "grad_norm": 3.0980546474456787,
-      "learning_rate": 1.296969696969697e-05,
-      "loss": 4.4019,
+      "epoch": 2.2093862815884475,
+      "grad_norm": 4.059695720672607,
+      "learning_rate": 1.3237063778580025e-05,
+      "loss": 4.4897,
       "step": 612
     },
     {
-      "epoch": 2.229090909090909,
-      "grad_norm": 3.8354101181030273,
-      "learning_rate": 1.290909090909091e-05,
-      "loss": 4.424,
+      "epoch": 2.2129963898916967,
+      "grad_norm": 2.7088205814361572,
+      "learning_rate": 1.3176895306859207e-05,
+      "loss": 4.5617,
       "step": 613
     },
     {
-      "epoch": 2.232727272727273,
-      "grad_norm": 5.496599197387695,
-      "learning_rate": 1.2848484848484848e-05,
-      "loss": 4.5742,
+      "epoch": 2.216606498194946,
+      "grad_norm": 4.490012168884277,
+      "learning_rate": 1.3116726835138388e-05,
+      "loss": 4.4686,
       "step": 614
     },
     {
-      "epoch": 2.2363636363636363,
-      "grad_norm": 3.4373464584350586,
-      "learning_rate": 1.2787878787878787e-05,
-      "loss": 4.5655,
+      "epoch": 2.220216606498195,
+      "grad_norm": 3.641652822494507,
+      "learning_rate": 1.305655836341757e-05,
+      "loss": 4.3731,
       "step": 615
     },
     {
-      "epoch": 2.24,
-      "grad_norm": 2.642533779144287,
-      "learning_rate": 1.2727272727272727e-05,
-      "loss": 4.3463,
+      "epoch": 2.223826714801444,
+      "grad_norm": 3.919235944747925,
+      "learning_rate": 1.299638989169675e-05,
+      "loss": 4.3683,
       "step": 616
     },
     {
-      "epoch": 2.2436363636363637,
-      "grad_norm": 7.2525224685668945,
-      "learning_rate": 1.2666666666666668e-05,
-      "loss": 4.3718,
+      "epoch": 2.2274368231046933,
+      "grad_norm": 3.3261141777038574,
+      "learning_rate": 1.2936221419975933e-05,
+      "loss": 4.3962,
       "step": 617
     },
     {
-      "epoch": 2.247272727272727,
-      "grad_norm": 5.643246173858643,
-      "learning_rate": 1.2606060606060607e-05,
-      "loss": 4.5915,
+      "epoch": 2.2310469314079424,
+      "grad_norm": 5.451700687408447,
+      "learning_rate": 1.2876052948255116e-05,
+      "loss": 4.5434,
       "step": 618
     },
     {
-      "epoch": 2.250909090909091,
-      "grad_norm": 6.5598835945129395,
-      "learning_rate": 1.2545454545454547e-05,
-      "loss": 4.5327,
+      "epoch": 2.234657039711191,
+      "grad_norm": 4.056484699249268,
+      "learning_rate": 1.2815884476534295e-05,
+      "loss": 4.4165,
       "step": 619
     },
     {
-      "epoch": 2.2545454545454544,
-      "grad_norm": 3.6604361534118652,
-      "learning_rate": 1.2484848484848485e-05,
-      "loss": 4.3982,
+      "epoch": 2.2382671480144403,
+      "grad_norm": 2.935309648513794,
+      "learning_rate": 1.2755716004813477e-05,
+      "loss": 4.3765,
       "step": 620
     },
     {
-      "epoch": 2.2581818181818183,
-      "grad_norm": 4.660340309143066,
-      "learning_rate": 1.2424242424242424e-05,
-      "loss": 4.3316,
+      "epoch": 2.2418772563176894,
+      "grad_norm": 2.4981465339660645,
+      "learning_rate": 1.2695547533092661e-05,
+      "loss": 4.4208,
       "step": 621
     },
     {
-      "epoch": 2.2618181818181817,
-      "grad_norm": 2.6217048168182373,
-      "learning_rate": 1.2363636363636365e-05,
-      "loss": 4.4005,
+      "epoch": 2.2454873646209386,
+      "grad_norm": 4.763179302215576,
+      "learning_rate": 1.263537906137184e-05,
+      "loss": 4.3193,
       "step": 622
     },
     {
-      "epoch": 2.2654545454545456,
-      "grad_norm": 5.1133952140808105,
-      "learning_rate": 1.2303030303030304e-05,
-      "loss": 4.6143,
+      "epoch": 2.2490974729241877,
+      "grad_norm": 6.148734092712402,
+      "learning_rate": 1.2575210589651024e-05,
+      "loss": 4.4633,
       "step": 623
     },
     {
-      "epoch": 2.269090909090909,
-      "grad_norm": 4.391026020050049,
-      "learning_rate": 1.2242424242424242e-05,
-      "loss": 4.5866,
+      "epoch": 2.252707581227437,
+      "grad_norm": 4.5394744873046875,
+      "learning_rate": 1.2515042117930206e-05,
+      "loss": 4.376,
       "step": 624
     },
     {
-      "epoch": 2.2727272727272725,
-      "grad_norm": 3.473971366882324,
-      "learning_rate": 1.2181818181818182e-05,
-      "loss": 4.5554,
+      "epoch": 2.256317689530686,
+      "grad_norm": 4.363958358764648,
+      "learning_rate": 1.2454873646209387e-05,
+      "loss": 4.5333,
       "step": 625
     },
     {
-      "epoch": 2.2763636363636364,
-      "grad_norm": 3.6338303089141846,
-      "learning_rate": 1.2121212121212122e-05,
-      "loss": 4.5832,
+      "epoch": 2.259927797833935,
+      "grad_norm": 3.6951801776885986,
+      "learning_rate": 1.2394705174488569e-05,
+      "loss": 4.4411,
       "step": 626
     },
     {
-      "epoch": 2.2800000000000002,
-      "grad_norm": 3.404435157775879,
-      "learning_rate": 1.2060606060606062e-05,
-      "loss": 4.4887,
+      "epoch": 2.2635379061371843,
+      "grad_norm": 4.908221244812012,
+      "learning_rate": 1.233453670276775e-05,
+      "loss": 4.4036,
       "step": 627
     },
     {
-      "epoch": 2.2836363636363637,
-      "grad_norm": 5.865675449371338,
-      "learning_rate": 1.2e-05,
-      "loss": 4.5747,
+      "epoch": 2.2671480144404335,
+      "grad_norm": 6.928462505340576,
+      "learning_rate": 1.2274368231046932e-05,
+      "loss": 4.5572,
       "step": 628
     },
     {
-      "epoch": 2.287272727272727,
-      "grad_norm": 3.2075819969177246,
-      "learning_rate": 1.193939393939394e-05,
-      "loss": 4.3799,
+      "epoch": 2.270758122743682,
+      "grad_norm": 4.029715061187744,
+      "learning_rate": 1.2214199759326114e-05,
+      "loss": 4.5028,
       "step": 629
     },
     {
-      "epoch": 2.290909090909091,
-      "grad_norm": 4.172214031219482,
-      "learning_rate": 1.187878787878788e-05,
-      "loss": 4.5288,
+      "epoch": 2.2743682310469313,
+      "grad_norm": 4.783502101898193,
+      "learning_rate": 1.2154031287605296e-05,
+      "loss": 4.5988,
       "step": 630
     },
     {
-      "epoch": 2.2945454545454544,
-      "grad_norm": 5.905372142791748,
-      "learning_rate": 1.1818181818181819e-05,
-      "loss": 4.3566,
+      "epoch": 2.2779783393501805,
+      "grad_norm": 5.311758041381836,
+      "learning_rate": 1.2093862815884477e-05,
+      "loss": 4.3549,
       "step": 631
     },
     {
-      "epoch": 2.2981818181818183,
-      "grad_norm": 3.021374225616455,
-      "learning_rate": 1.1757575757575757e-05,
-      "loss": 4.2457,
+      "epoch": 2.2815884476534296,
+      "grad_norm": 5.8430094718933105,
+      "learning_rate": 1.2033694344163659e-05,
+      "loss": 4.4828,
       "step": 632
     },
     {
-      "epoch": 2.3018181818181818,
-      "grad_norm": 6.168385982513428,
-      "learning_rate": 1.1696969696969699e-05,
-      "loss": 4.5722,
+      "epoch": 2.2851985559566788,
+      "grad_norm": 2.834203004837036,
+      "learning_rate": 1.1973525872442841e-05,
+      "loss": 4.4867,
       "step": 633
     },
     {
-      "epoch": 2.3054545454545456,
-      "grad_norm": 4.84739351272583,
-      "learning_rate": 1.1636363636363637e-05,
-      "loss": 4.5149,
+      "epoch": 2.288808664259928,
+      "grad_norm": 3.2368834018707275,
+      "learning_rate": 1.1913357400722022e-05,
+      "loss": 4.5147,
       "step": 634
     },
     {
-      "epoch": 2.309090909090909,
-      "grad_norm": 5.882564067840576,
-      "learning_rate": 1.1575757575757575e-05,
-      "loss": 4.6114,
+      "epoch": 2.292418772563177,
+      "grad_norm": 4.3523359298706055,
+      "learning_rate": 1.1853188929001204e-05,
+      "loss": 4.3972,
       "step": 635
     },
     {
-      "epoch": 2.3127272727272725,
-      "grad_norm": 3.714647054672241,
-      "learning_rate": 1.1515151515151517e-05,
-      "loss": 4.5177,
+      "epoch": 2.2960288808664258,
+      "grad_norm": 4.379057884216309,
+      "learning_rate": 1.1793020457280386e-05,
+      "loss": 4.3681,
       "step": 636
     },
     {
-      "epoch": 2.3163636363636364,
-      "grad_norm": 3.449841260910034,
-      "learning_rate": 1.1454545454545455e-05,
-      "loss": 4.5713,
+      "epoch": 2.299638989169675,
+      "grad_norm": 6.070941925048828,
+      "learning_rate": 1.1732851985559568e-05,
+      "loss": 4.4543,
       "step": 637
     },
     {
-      "epoch": 2.32,
-      "grad_norm": 6.447396755218506,
-      "learning_rate": 1.1393939393939394e-05,
-      "loss": 4.501,
+      "epoch": 2.303249097472924,
+      "grad_norm": 4.886351585388184,
+      "learning_rate": 1.1672683513838749e-05,
+      "loss": 4.4359,
       "step": 638
     },
     {
-      "epoch": 2.3236363636363637,
-      "grad_norm": 3.258431911468506,
-      "learning_rate": 1.1333333333333334e-05,
-      "loss": 4.5533,
+      "epoch": 2.306859205776173,
+      "grad_norm": 3.127772808074951,
+      "learning_rate": 1.161251504211793e-05,
+      "loss": 4.463,
       "step": 639
     },
     {
-      "epoch": 2.327272727272727,
-      "grad_norm": 4.90877103805542,
-      "learning_rate": 1.1272727272727274e-05,
-      "loss": 4.3269,
+      "epoch": 2.3104693140794224,
+      "grad_norm": 3.3918819427490234,
+      "learning_rate": 1.1552346570397113e-05,
+      "loss": 4.5662,
       "step": 640
     },
     {
-      "epoch": 2.330909090909091,
-      "grad_norm": 4.926637172698975,
-      "learning_rate": 1.1212121212121212e-05,
-      "loss": 4.395,
+      "epoch": 2.3140794223826715,
+      "grad_norm": 3.3989920616149902,
+      "learning_rate": 1.1492178098676294e-05,
+      "loss": 4.4219,
       "step": 641
     },
     {
-      "epoch": 2.3345454545454545,
-      "grad_norm": 4.344498634338379,
-      "learning_rate": 1.1151515151515152e-05,
-      "loss": 4.2733,
+      "epoch": 2.3176895306859207,
+      "grad_norm": 3.8738067150115967,
+      "learning_rate": 1.1432009626955476e-05,
+      "loss": 4.5089,
       "step": 642
     },
     {
-      "epoch": 2.3381818181818184,
-      "grad_norm": 2.261861562728882,
-      "learning_rate": 1.1090909090909092e-05,
-      "loss": 4.3516,
+      "epoch": 2.32129963898917,
+      "grad_norm": 4.018595218658447,
+      "learning_rate": 1.1371841155234658e-05,
+      "loss": 4.6628,
       "step": 643
     },
     {
-      "epoch": 2.341818181818182,
-      "grad_norm": 3.712153434753418,
-      "learning_rate": 1.103030303030303e-05,
-      "loss": 4.2867,
+      "epoch": 2.324909747292419,
+      "grad_norm": 4.790584087371826,
+      "learning_rate": 1.1311672683513839e-05,
+      "loss": 4.5279,
       "step": 644
     },
     {
-      "epoch": 2.3454545454545457,
-      "grad_norm": 2.1240344047546387,
-      "learning_rate": 1.096969696969697e-05,
-      "loss": 4.4041,
+      "epoch": 2.328519855595668,
+      "grad_norm": 3.1995787620544434,
+      "learning_rate": 1.1251504211793021e-05,
+      "loss": 4.6064,
       "step": 645
     },
     {
-      "epoch": 2.349090909090909,
-      "grad_norm": 5.3182053565979,
-      "learning_rate": 1.0909090909090909e-05,
-      "loss": 4.5118,
+      "epoch": 2.332129963898917,
+      "grad_norm": 2.8555033206939697,
+      "learning_rate": 1.1191335740072201e-05,
+      "loss": 4.4816,
       "step": 646
     },
     {
-      "epoch": 2.3527272727272726,
-      "grad_norm": 3.5492019653320312,
-      "learning_rate": 1.0848484848484849e-05,
-      "loss": 4.5899,
+      "epoch": 2.335740072202166,
+      "grad_norm": 3.3136844635009766,
+      "learning_rate": 1.1131167268351385e-05,
+      "loss": 4.5662,
       "step": 647
     },
     {
-      "epoch": 2.3563636363636364,
-      "grad_norm": 3.8747949600219727,
-      "learning_rate": 1.0787878787878789e-05,
-      "loss": 4.3194,
+      "epoch": 2.339350180505415,
+      "grad_norm": 3.995872735977173,
+      "learning_rate": 1.1070998796630566e-05,
+      "loss": 4.3727,
       "step": 648
     },
     {
-      "epoch": 2.36,
-      "grad_norm": 2.2515571117401123,
-      "learning_rate": 1.0727272727272727e-05,
-      "loss": 4.4239,
+      "epoch": 2.3429602888086642,
+      "grad_norm": 4.499544143676758,
+      "learning_rate": 1.1010830324909748e-05,
+      "loss": 4.478,
       "step": 649
     },
     {
-      "epoch": 2.3636363636363638,
-      "grad_norm": 3.838587760925293,
-      "learning_rate": 1.0666666666666667e-05,
-      "loss": 4.5496,
+      "epoch": 2.3465703971119134,
+      "grad_norm": 3.3210017681121826,
+      "learning_rate": 1.095066185318893e-05,
+      "loss": 4.396,
       "step": 650
     },
     {
-      "epoch": 2.367272727272727,
-      "grad_norm": 2.633479595184326,
-      "learning_rate": 1.0606060606060607e-05,
-      "loss": 4.2623,
+      "epoch": 2.3501805054151625,
+      "grad_norm": 2.9872148036956787,
+      "learning_rate": 1.089049338146811e-05,
+      "loss": 4.4108,
       "step": 651
     },
     {
-      "epoch": 2.370909090909091,
-      "grad_norm": 3.465851306915283,
-      "learning_rate": 1.0545454545454546e-05,
-      "loss": 4.3402,
+      "epoch": 2.3537906137184117,
+      "grad_norm": 5.219914436340332,
+      "learning_rate": 1.0830324909747293e-05,
+      "loss": 4.4073,
       "step": 652
     },
     {
-      "epoch": 2.3745454545454545,
-      "grad_norm": 4.0814595222473145,
-      "learning_rate": 1.0484848484848486e-05,
-      "loss": 4.2627,
+      "epoch": 2.357400722021661,
+      "grad_norm": 3.3450937271118164,
+      "learning_rate": 1.0770156438026474e-05,
+      "loss": 4.2831,
       "step": 653
     },
     {
-      "epoch": 2.378181818181818,
-      "grad_norm": 1.8538293838500977,
-      "learning_rate": 1.0424242424242426e-05,
-      "loss": 4.3685,
+      "epoch": 2.3610108303249095,
+      "grad_norm": 4.793975353240967,
+      "learning_rate": 1.0709987966305657e-05,
+      "loss": 4.6626,
       "step": 654
     },
     {
-      "epoch": 2.381818181818182,
-      "grad_norm": 3.5009477138519287,
-      "learning_rate": 1.0363636363636364e-05,
-      "loss": 4.4817,
+      "epoch": 2.3646209386281587,
+      "grad_norm": 4.04519510269165,
+      "learning_rate": 1.0649819494584838e-05,
+      "loss": 4.3976,
       "step": 655
     },
     {
-      "epoch": 2.3854545454545453,
-      "grad_norm": 5.759241580963135,
-      "learning_rate": 1.0303030303030304e-05,
-      "loss": 4.229,
+      "epoch": 2.368231046931408,
+      "grad_norm": 3.0251882076263428,
+      "learning_rate": 1.058965102286402e-05,
+      "loss": 4.4383,
       "step": 656
     },
     {
-      "epoch": 2.389090909090909,
-      "grad_norm": 8.428985595703125,
-      "learning_rate": 1.0242424242424242e-05,
-      "loss": 4.7398,
+      "epoch": 2.371841155234657,
+      "grad_norm": 3.5139055252075195,
+      "learning_rate": 1.0529482551143202e-05,
+      "loss": 4.5331,
       "step": 657
     },
     {
-      "epoch": 2.3927272727272726,
-      "grad_norm": 4.199151515960693,
-      "learning_rate": 1.0181818181818182e-05,
-      "loss": 4.2961,
+      "epoch": 2.375451263537906,
+      "grad_norm": 3.8762271404266357,
+      "learning_rate": 1.0469314079422383e-05,
+      "loss": 4.4937,
       "step": 658
     },
     {
-      "epoch": 2.3963636363636365,
-      "grad_norm": 4.035223484039307,
-      "learning_rate": 1.0121212121212122e-05,
-      "loss": 4.491,
+      "epoch": 2.3790613718411553,
+      "grad_norm": 4.291984558105469,
+      "learning_rate": 1.0409145607701565e-05,
+      "loss": 4.366,
       "step": 659
     },
     {
-      "epoch": 2.4,
-      "grad_norm": 3.7876856327056885,
-      "learning_rate": 1.006060606060606e-05,
-      "loss": 4.4866,
+      "epoch": 2.3826714801444044,
+      "grad_norm": 3.6737313270568848,
+      "learning_rate": 1.0348977135980746e-05,
+      "loss": 4.3613,
       "step": 660
     },
     {
-      "epoch": 2.403636363636364,
-      "grad_norm": 4.28821325302124,
-      "learning_rate": 1e-05,
-      "loss": 4.3898,
+      "epoch": 2.3862815884476536,
+      "grad_norm": 4.335361003875732,
+      "learning_rate": 1.028880866425993e-05,
+      "loss": 4.3535,
       "step": 661
     },
     {
-      "epoch": 2.4072727272727272,
-      "grad_norm": 5.3424391746521,
-      "learning_rate": 9.93939393939394e-06,
-      "loss": 4.3163,
+      "epoch": 2.3898916967509027,
+      "grad_norm": 2.939689874649048,
+      "learning_rate": 1.022864019253911e-05,
+      "loss": 4.6187,
       "step": 662
     },
     {
-      "epoch": 2.410909090909091,
-      "grad_norm": 3.205174207687378,
-      "learning_rate": 9.87878787878788e-06,
-      "loss": 4.4189,
+      "epoch": 2.3935018050541514,
+      "grad_norm": 2.5564160346984863,
+      "learning_rate": 1.016847172081829e-05,
+      "loss": 4.5736,
       "step": 663
     },
     {
-      "epoch": 2.4145454545454546,
-      "grad_norm": 2.7838244438171387,
-      "learning_rate": 9.818181818181818e-06,
-      "loss": 4.3889,
+      "epoch": 2.3971119133574006,
+      "grad_norm": 7.256919860839844,
+      "learning_rate": 1.0108303249097473e-05,
+      "loss": 4.6371,
       "step": 664
     },
     {
-      "epoch": 2.418181818181818,
-      "grad_norm": 4.061773777008057,
-      "learning_rate": 9.75757575757576e-06,
-      "loss": 4.5237,
+      "epoch": 2.4007220216606497,
+      "grad_norm": 5.051013946533203,
+      "learning_rate": 1.0048134777376655e-05,
+      "loss": 4.5431,
       "step": 665
     },
     {
-      "epoch": 2.421818181818182,
-      "grad_norm": 5.466277122497559,
-      "learning_rate": 9.696969696969698e-06,
-      "loss": 4.5694,
+      "epoch": 2.404332129963899,
+      "grad_norm": 4.401142597198486,
+      "learning_rate": 9.987966305655837e-06,
+      "loss": 4.419,
       "step": 666
     },
     {
-      "epoch": 2.4254545454545453,
-      "grad_norm": 3.7190425395965576,
-      "learning_rate": 9.636363636363636e-06,
-      "loss": 4.4085,
+      "epoch": 2.407942238267148,
+      "grad_norm": 2.8643991947174072,
+      "learning_rate": 9.927797833935018e-06,
+      "loss": 4.3324,
       "step": 667
     },
     {
-      "epoch": 2.429090909090909,
-      "grad_norm": 5.519341468811035,
-      "learning_rate": 9.575757575757578e-06,
-      "loss": 4.5653,
+      "epoch": 2.411552346570397,
+      "grad_norm": 7.718422889709473,
+      "learning_rate": 9.8676293622142e-06,
+      "loss": 4.3999,
       "step": 668
     },
     {
-      "epoch": 2.4327272727272726,
-      "grad_norm": 3.7906625270843506,
-      "learning_rate": 9.515151515151516e-06,
-      "loss": 4.3678,
+      "epoch": 2.4151624548736463,
+      "grad_norm": 6.474762916564941,
+      "learning_rate": 9.807460890493382e-06,
+      "loss": 4.3473,
       "step": 669
     },
     {
-      "epoch": 2.4363636363636365,
-      "grad_norm": 4.553290843963623,
-      "learning_rate": 9.454545454545454e-06,
-      "loss": 4.6579,
+      "epoch": 2.4187725631768955,
+      "grad_norm": 5.406153678894043,
+      "learning_rate": 9.747292418772563e-06,
+      "loss": 4.4933,
       "step": 670
     },
     {
-      "epoch": 2.44,
-      "grad_norm": 4.562989711761475,
-      "learning_rate": 9.393939393939394e-06,
-      "loss": 4.3674,
+      "epoch": 2.422382671480144,
+      "grad_norm": 4.858792781829834,
+      "learning_rate": 9.687123947051745e-06,
+      "loss": 4.658,
       "step": 671
     },
     {
-      "epoch": 2.443636363636364,
-      "grad_norm": 4.022787094116211,
-      "learning_rate": 9.333333333333334e-06,
-      "loss": 4.3675,
+      "epoch": 2.4259927797833933,
+      "grad_norm": 5.371998310089111,
+      "learning_rate": 9.626955475330927e-06,
+      "loss": 4.478,
       "step": 672
     },
     {
-      "epoch": 2.4472727272727273,
-      "grad_norm": 2.7684528827667236,
-      "learning_rate": 9.272727272727273e-06,
-      "loss": 4.3793,
+      "epoch": 2.4296028880866425,
+      "grad_norm": 6.951625347137451,
+      "learning_rate": 9.56678700361011e-06,
+      "loss": 4.5687,
       "step": 673
     },
     {
-      "epoch": 2.450909090909091,
-      "grad_norm": 3.1357200145721436,
-      "learning_rate": 9.212121212121213e-06,
-      "loss": 4.2807,
+      "epoch": 2.4332129963898916,
+      "grad_norm": 2.8884341716766357,
+      "learning_rate": 9.50661853188929e-06,
+      "loss": 4.3574,
       "step": 674
     },
     {
-      "epoch": 2.4545454545454546,
-      "grad_norm": 2.534762144088745,
-      "learning_rate": 9.151515151515153e-06,
-      "loss": 4.3097,
+      "epoch": 2.436823104693141,
+      "grad_norm": 4.6059699058532715,
+      "learning_rate": 9.446450060168472e-06,
+      "loss": 4.5776,
       "step": 675
     },
     {
-      "epoch": 2.458181818181818,
-      "grad_norm": 3.3176114559173584,
-      "learning_rate": 9.090909090909091e-06,
-      "loss": 4.6301,
+      "epoch": 2.44043321299639,
+      "grad_norm": 4.730526924133301,
+      "learning_rate": 9.386281588447654e-06,
+      "loss": 4.592,
       "step": 676
     },
     {
-      "epoch": 2.461818181818182,
-      "grad_norm": 3.287240505218506,
-      "learning_rate": 9.030303030303031e-06,
-      "loss": 4.5127,
+      "epoch": 2.444043321299639,
+      "grad_norm": 3.3078360557556152,
+      "learning_rate": 9.326113116726835e-06,
+      "loss": 4.5253,
       "step": 677
     },
     {
-      "epoch": 2.4654545454545453,
-      "grad_norm": 3.7666358947753906,
-      "learning_rate": 8.96969696969697e-06,
-      "loss": 4.2394,
+      "epoch": 2.4476534296028882,
+      "grad_norm": 3.702000856399536,
+      "learning_rate": 9.265944645006017e-06,
+      "loss": 4.7267,
       "step": 678
     },
     {
-      "epoch": 2.4690909090909092,
-      "grad_norm": 2.5276758670806885,
-      "learning_rate": 8.90909090909091e-06,
-      "loss": 4.3502,
+      "epoch": 2.4512635379061374,
+      "grad_norm": 3.359959363937378,
+      "learning_rate": 9.2057761732852e-06,
+      "loss": 4.4877,
       "step": 679
     },
     {
-      "epoch": 2.4727272727272727,
-      "grad_norm": 3.2294423580169678,
-      "learning_rate": 8.84848484848485e-06,
-      "loss": 4.5801,
+      "epoch": 2.4548736462093865,
+      "grad_norm": 5.103507995605469,
+      "learning_rate": 9.145607701564381e-06,
+      "loss": 4.6424,
       "step": 680
     },
     {
-      "epoch": 2.4763636363636365,
-      "grad_norm": 5.495777130126953,
-      "learning_rate": 8.787878787878788e-06,
-      "loss": 4.7307,
+      "epoch": 2.4584837545126352,
+      "grad_norm": 5.196680068969727,
+      "learning_rate": 9.085439229843562e-06,
+      "loss": 4.4706,
       "step": 681
     },
     {
-      "epoch": 2.48,
-      "grad_norm": 7.053243637084961,
-      "learning_rate": 8.727272727272728e-06,
-      "loss": 4.3913,
+      "epoch": 2.4620938628158844,
+      "grad_norm": 3.826127052307129,
+      "learning_rate": 9.025270758122744e-06,
+      "loss": 4.381,
       "step": 682
     },
     {
-      "epoch": 2.4836363636363634,
-      "grad_norm": 8.957902908325195,
-      "learning_rate": 8.666666666666668e-06,
-      "loss": 4.4562,
+      "epoch": 2.4657039711191335,
+      "grad_norm": 2.5819101333618164,
+      "learning_rate": 8.965102286401926e-06,
+      "loss": 4.3937,
       "step": 683
     },
     {
-      "epoch": 2.4872727272727273,
-      "grad_norm": 4.16687536239624,
-      "learning_rate": 8.606060606060606e-06,
-      "loss": 4.3475,
+      "epoch": 2.4693140794223827,
+      "grad_norm": 3.1093242168426514,
+      "learning_rate": 8.904933814681107e-06,
+      "loss": 4.3896,
       "step": 684
     },
     {
-      "epoch": 2.4909090909090907,
-      "grad_norm": 9.852190971374512,
-      "learning_rate": 8.545454545454546e-06,
-      "loss": 4.5089,
+      "epoch": 2.472924187725632,
+      "grad_norm": 4.864704132080078,
+      "learning_rate": 8.84476534296029e-06,
+      "loss": 4.5112,
       "step": 685
     },
     {
-      "epoch": 2.4945454545454546,
-      "grad_norm": 12.333488464355469,
-      "learning_rate": 8.484848484848486e-06,
-      "loss": 4.4475,
+      "epoch": 2.476534296028881,
+      "grad_norm": 3.471543073654175,
+      "learning_rate": 8.784596871239471e-06,
+      "loss": 4.5498,
       "step": 686
     },
     {
-      "epoch": 2.498181818181818,
-      "grad_norm": 7.947030544281006,
-      "learning_rate": 8.424242424242425e-06,
-      "loss": 4.2607,
+      "epoch": 2.48014440433213,
+      "grad_norm": 4.402616500854492,
+      "learning_rate": 8.724428399518652e-06,
+      "loss": 4.3387,
       "step": 687
     },
     {
-      "epoch": 2.501818181818182,
-      "grad_norm": 4.91619348526001,
-      "learning_rate": 8.363636363636365e-06,
-      "loss": 4.3802,
+      "epoch": 2.483754512635379,
+      "grad_norm": 6.379435062408447,
+      "learning_rate": 8.664259927797834e-06,
+      "loss": 4.5544,
       "step": 688
     },
     {
-      "epoch": 2.5054545454545454,
-      "grad_norm": 4.144915580749512,
-      "learning_rate": 8.303030303030303e-06,
-      "loss": 4.6197,
+      "epoch": 2.487364620938628,
+      "grad_norm": 6.62040901184082,
+      "learning_rate": 8.604091456077016e-06,
+      "loss": 4.4874,
       "step": 689
     },
     {
-      "epoch": 2.509090909090909,
-      "grad_norm": 4.914970874786377,
-      "learning_rate": 8.242424242424243e-06,
-      "loss": 4.4078,
+      "epoch": 2.490974729241877,
+      "grad_norm": 5.348603248596191,
+      "learning_rate": 8.543922984356199e-06,
+      "loss": 4.3741,
       "step": 690
     },
     {
-      "epoch": 2.5127272727272727,
-      "grad_norm": 5.826870918273926,
-      "learning_rate": 8.181818181818183e-06,
-      "loss": 4.3835,
+      "epoch": 2.4945848375451263,
+      "grad_norm": 2.902223825454712,
+      "learning_rate": 8.483754512635379e-06,
+      "loss": 4.3933,
       "step": 691
     },
     {
-      "epoch": 2.5163636363636366,
-      "grad_norm": 4.456042766571045,
-      "learning_rate": 8.121212121212121e-06,
-      "loss": 4.4389,
+      "epoch": 2.4981949458483754,
+      "grad_norm": 3.2909133434295654,
+      "learning_rate": 8.423586040914561e-06,
+      "loss": 4.5229,
       "step": 692
     },
     {
-      "epoch": 2.52,
-      "grad_norm": 7.257143497467041,
-      "learning_rate": 8.060606060606061e-06,
-      "loss": 4.4145,
+      "epoch": 2.5018050541516246,
+      "grad_norm": 3.990521192550659,
+      "learning_rate": 8.363417569193743e-06,
+      "loss": 4.5907,
       "step": 693
     },
     {
-      "epoch": 2.5236363636363635,
-      "grad_norm": 6.106967926025391,
-      "learning_rate": 8.000000000000001e-06,
-      "loss": 4.3769,
+      "epoch": 2.5054151624548737,
+      "grad_norm": 5.001603126525879,
+      "learning_rate": 8.303249097472924e-06,
+      "loss": 4.4247,
       "step": 694
     },
     {
-      "epoch": 2.5272727272727273,
-      "grad_norm": 3.0552563667297363,
-      "learning_rate": 7.93939393939394e-06,
-      "loss": 4.4448,
+      "epoch": 2.509025270758123,
+      "grad_norm": 4.88585090637207,
+      "learning_rate": 8.243080625752106e-06,
+      "loss": 4.4019,
       "step": 695
     },
     {
-      "epoch": 2.5309090909090908,
-      "grad_norm": 4.240468502044678,
-      "learning_rate": 7.878787878787878e-06,
-      "loss": 4.6053,
+      "epoch": 2.512635379061372,
+      "grad_norm": 3.634761333465576,
+      "learning_rate": 8.182912154031288e-06,
+      "loss": 4.3335,
       "step": 696
     },
     {
-      "epoch": 2.5345454545454547,
-      "grad_norm": 4.00745153427124,
-      "learning_rate": 7.81818181818182e-06,
-      "loss": 4.715,
+      "epoch": 2.516245487364621,
+      "grad_norm": 4.154645919799805,
+      "learning_rate": 8.12274368231047e-06,
+      "loss": 4.4616,
       "step": 697
     },
     {
-      "epoch": 2.538181818181818,
-      "grad_norm": 7.360689640045166,
-      "learning_rate": 7.757575757575758e-06,
-      "loss": 4.2768,
+      "epoch": 2.51985559566787,
+      "grad_norm": 2.943950891494751,
+      "learning_rate": 8.062575210589651e-06,
+      "loss": 4.4305,
       "step": 698
     },
     {
-      "epoch": 2.541818181818182,
-      "grad_norm": 10.462032318115234,
-      "learning_rate": 7.696969696969696e-06,
-      "loss": 4.4312,
+      "epoch": 2.523465703971119,
+      "grad_norm": 5.359757900238037,
+      "learning_rate": 8.002406738868833e-06,
+      "loss": 4.4419,
       "step": 699
     },
     {
-      "epoch": 2.5454545454545454,
-      "grad_norm": 7.483114719390869,
-      "learning_rate": 7.636363636363638e-06,
-      "loss": 4.386,
+      "epoch": 2.527075812274368,
+      "grad_norm": 3.205974817276001,
+      "learning_rate": 7.942238267148016e-06,
+      "loss": 4.4804,
       "step": 700
     },
     {
-      "epoch": 2.549090909090909,
-      "grad_norm": 2.8339669704437256,
-      "learning_rate": 7.5757575757575764e-06,
-      "loss": 4.495,
+      "epoch": 2.5306859205776173,
+      "grad_norm": 3.1769614219665527,
+      "learning_rate": 7.882069795427196e-06,
+      "loss": 4.5115,
       "step": 701
     },
     {
-      "epoch": 2.5527272727272727,
-      "grad_norm": 4.9380340576171875,
-      "learning_rate": 7.515151515151516e-06,
-      "loss": 4.3816,
+      "epoch": 2.5342960288808665,
+      "grad_norm": 3.652250051498413,
+      "learning_rate": 7.821901323706378e-06,
+      "loss": 4.2911,
       "step": 702
     },
     {
-      "epoch": 2.5563636363636366,
-      "grad_norm": 5.354694366455078,
-      "learning_rate": 7.454545454545454e-06,
-      "loss": 4.3048,
+      "epoch": 2.5379061371841156,
+      "grad_norm": 6.6311235427856445,
+      "learning_rate": 7.76173285198556e-06,
+      "loss": 4.4701,
       "step": 703
     },
     {
-      "epoch": 2.56,
-      "grad_norm": 8.832294464111328,
-      "learning_rate": 7.393939393939395e-06,
-      "loss": 4.3648,
+      "epoch": 2.5415162454873648,
+      "grad_norm": 4.466462135314941,
+      "learning_rate": 7.701564380264743e-06,
+      "loss": 4.2806,
       "step": 704
     },
     {
-      "epoch": 2.5636363636363635,
-      "grad_norm": 5.9402971267700195,
-      "learning_rate": 7.333333333333334e-06,
-      "loss": 4.2456,
+      "epoch": 2.5451263537906135,
+      "grad_norm": 2.840956687927246,
+      "learning_rate": 7.641395908543923e-06,
+      "loss": 4.3817,
       "step": 705
     },
     {
-      "epoch": 2.5672727272727274,
-      "grad_norm": 4.423623561859131,
-      "learning_rate": 7.272727272727272e-06,
-      "loss": 4.629,
+      "epoch": 2.5487364620938626,
+      "grad_norm": 3.926969289779663,
+      "learning_rate": 7.581227436823105e-06,
+      "loss": 4.4675,
       "step": 706
     },
     {
-      "epoch": 2.570909090909091,
-      "grad_norm": 3.3913474082946777,
-      "learning_rate": 7.212121212121213e-06,
-      "loss": 4.3109,
+      "epoch": 2.5523465703971118,
+      "grad_norm": 7.378762245178223,
+      "learning_rate": 7.521058965102287e-06,
+      "loss": 4.2906,
       "step": 707
     },
     {
-      "epoch": 2.5745454545454547,
-      "grad_norm": 4.365673542022705,
-      "learning_rate": 7.151515151515152e-06,
-      "loss": 4.4317,
+      "epoch": 2.555956678700361,
+      "grad_norm": 6.083559036254883,
+      "learning_rate": 7.460890493381468e-06,
+      "loss": 4.5136,
       "step": 708
     },
     {
-      "epoch": 2.578181818181818,
-      "grad_norm": 5.201317310333252,
-      "learning_rate": 7.090909090909091e-06,
-      "loss": 4.361,
+      "epoch": 2.55956678700361,
+      "grad_norm": 3.145780324935913,
+      "learning_rate": 7.4007220216606496e-06,
+      "loss": 4.3858,
       "step": 709
     },
     {
-      "epoch": 2.581818181818182,
-      "grad_norm": 8.283641815185547,
-      "learning_rate": 7.03030303030303e-06,
-      "loss": 4.4209,
+      "epoch": 2.563176895306859,
+      "grad_norm": 3.72505784034729,
+      "learning_rate": 7.340553549939831e-06,
+      "loss": 4.5418,
       "step": 710
     },
     {
-      "epoch": 2.5854545454545454,
-      "grad_norm": 6.324493885040283,
-      "learning_rate": 6.969696969696971e-06,
-      "loss": 4.572,
+      "epoch": 2.5667870036101084,
+      "grad_norm": 4.158870697021484,
+      "learning_rate": 7.280385078219014e-06,
+      "loss": 4.5686,
       "step": 711
     },
     {
-      "epoch": 2.589090909090909,
-      "grad_norm": 4.416196823120117,
-      "learning_rate": 6.909090909090909e-06,
-      "loss": 4.4043,
+      "epoch": 2.5703971119133575,
+      "grad_norm": 6.947221279144287,
+      "learning_rate": 7.220216606498195e-06,
+      "loss": 4.2351,
       "step": 712
     },
     {
-      "epoch": 2.5927272727272728,
-      "grad_norm": 3.493542194366455,
-      "learning_rate": 6.848484848484848e-06,
-      "loss": 4.3647,
+      "epoch": 2.5740072202166067,
+      "grad_norm": 4.236288070678711,
+      "learning_rate": 7.160048134777377e-06,
+      "loss": 4.451,
       "step": 713
     },
     {
-      "epoch": 2.596363636363636,
-      "grad_norm": 3.4340717792510986,
-      "learning_rate": 6.787878787878789e-06,
-      "loss": 4.354,
+      "epoch": 2.577617328519856,
+      "grad_norm": 4.43533182144165,
+      "learning_rate": 7.099879663056559e-06,
+      "loss": 4.3672,
       "step": 714
     },
     {
-      "epoch": 2.6,
-      "grad_norm": 3.536820650100708,
-      "learning_rate": 6.727272727272728e-06,
-      "loss": 4.5004,
+      "epoch": 2.581227436823105,
+      "grad_norm": 2.0811402797698975,
+      "learning_rate": 7.03971119133574e-06,
+      "loss": 4.4889,
       "step": 715
     },
     {
-      "epoch": 2.6036363636363635,
-      "grad_norm": 4.423953533172607,
-      "learning_rate": 6.666666666666667e-06,
-      "loss": 4.2383,
+      "epoch": 2.5848375451263537,
+      "grad_norm": 3.2417030334472656,
+      "learning_rate": 6.979542719614922e-06,
+      "loss": 4.3176,
       "step": 716
     },
     {
-      "epoch": 2.6072727272727274,
-      "grad_norm": 3.8500802516937256,
-      "learning_rate": 6.606060606060606e-06,
-      "loss": 4.3031,
+      "epoch": 2.588447653429603,
+      "grad_norm": 4.629251956939697,
+      "learning_rate": 6.919374247894103e-06,
+      "loss": 4.6087,
       "step": 717
     },
     {
-      "epoch": 2.610909090909091,
-      "grad_norm": 5.0357747077941895,
-      "learning_rate": 6.545454545454547e-06,
-      "loss": 4.3652,
+      "epoch": 2.592057761732852,
+      "grad_norm": 3.3105053901672363,
+      "learning_rate": 6.859205776173286e-06,
+      "loss": 4.6441,
       "step": 718
     },
     {
-      "epoch": 2.6145454545454543,
-      "grad_norm": 4.3186492919921875,
-      "learning_rate": 6.484848484848485e-06,
-      "loss": 4.5205,
+      "epoch": 2.595667870036101,
+      "grad_norm": 3.484877586364746,
+      "learning_rate": 6.7990373044524675e-06,
+      "loss": 4.5328,
       "step": 719
     },
     {
-      "epoch": 2.618181818181818,
-      "grad_norm": 4.643374443054199,
-      "learning_rate": 6.424242424242424e-06,
-      "loss": 4.3247,
+      "epoch": 2.5992779783393503,
+      "grad_norm": 3.027569532394409,
+      "learning_rate": 6.738868832731649e-06,
+      "loss": 4.5395,
       "step": 720
     },
     {
-      "epoch": 2.621818181818182,
-      "grad_norm": 7.076257228851318,
-      "learning_rate": 6.363636363636363e-06,
-      "loss": 4.4034,
+      "epoch": 2.6028880866425994,
+      "grad_norm": 3.3761277198791504,
+      "learning_rate": 6.678700361010831e-06,
+      "loss": 4.4973,
       "step": 721
     },
     {
-      "epoch": 2.6254545454545455,
-      "grad_norm": 3.3600053787231445,
-      "learning_rate": 6.303030303030303e-06,
-      "loss": 4.3745,
+      "epoch": 2.606498194945848,
+      "grad_norm": 4.028012275695801,
+      "learning_rate": 6.618531889290012e-06,
+      "loss": 4.3345,
       "step": 722
     },
     {
-      "epoch": 2.629090909090909,
-      "grad_norm": 6.4890642166137695,
-      "learning_rate": 6.242424242424243e-06,
-      "loss": 4.3592,
+      "epoch": 2.6101083032490973,
+      "grad_norm": 3.6526448726654053,
+      "learning_rate": 6.558363417569194e-06,
+      "loss": 4.5195,
       "step": 723
     },
     {
-      "epoch": 2.632727272727273,
-      "grad_norm": 8.538745880126953,
-      "learning_rate": 6.181818181818183e-06,
-      "loss": 4.3828,
+      "epoch": 2.6137184115523464,
+      "grad_norm": 3.5621955394744873,
+      "learning_rate": 6.498194945848375e-06,
+      "loss": 4.3931,
       "step": 724
     },
     {
-      "epoch": 2.6363636363636362,
-      "grad_norm": 6.684054374694824,
-      "learning_rate": 6.121212121212121e-06,
-      "loss": 4.3507,
+      "epoch": 2.6173285198555956,
+      "grad_norm": 4.076904773712158,
+      "learning_rate": 6.438026474127558e-06,
+      "loss": 4.3022,
       "step": 725
     },
     {
-      "epoch": 2.64,
-      "grad_norm": 9.749906539916992,
-      "learning_rate": 6.060606060606061e-06,
-      "loss": 4.6677,
+      "epoch": 2.6209386281588447,
+      "grad_norm": 3.3727328777313232,
+      "learning_rate": 6.377858002406739e-06,
+      "loss": 4.5659,
       "step": 726
     },
     {
-      "epoch": 2.6436363636363636,
-      "grad_norm": 2.9998209476470947,
-      "learning_rate": 6e-06,
-      "loss": 4.4207,
+      "epoch": 2.624548736462094,
+      "grad_norm": 3.9651689529418945,
+      "learning_rate": 6.31768953068592e-06,
+      "loss": 4.4302,
       "step": 727
     },
     {
-      "epoch": 2.6472727272727274,
-      "grad_norm": 4.262009143829346,
-      "learning_rate": 5.93939393939394e-06,
-      "loss": 4.5837,
+      "epoch": 2.628158844765343,
+      "grad_norm": 5.752009391784668,
+      "learning_rate": 6.257521058965103e-06,
+      "loss": 4.6044,
       "step": 728
     },
     {
-      "epoch": 2.650909090909091,
-      "grad_norm": 5.290808200836182,
-      "learning_rate": 5.8787878787878785e-06,
-      "loss": 4.5531,
+      "epoch": 2.631768953068592,
+      "grad_norm": 4.023155689239502,
+      "learning_rate": 6.1973525872442845e-06,
+      "loss": 4.562,
       "step": 729
     },
     {
-      "epoch": 2.6545454545454543,
-      "grad_norm": 6.753263473510742,
-      "learning_rate": 5.8181818181818185e-06,
-      "loss": 4.4299,
+      "epoch": 2.6353790613718413,
+      "grad_norm": 5.812756538391113,
+      "learning_rate": 6.137184115523466e-06,
+      "loss": 4.4802,
       "step": 730
     },
     {
-      "epoch": 2.658181818181818,
-      "grad_norm": 4.626419544219971,
-      "learning_rate": 5.7575757575757586e-06,
-      "loss": 4.2893,
+      "epoch": 2.6389891696750905,
+      "grad_norm": 4.468845367431641,
+      "learning_rate": 6.077015643802648e-06,
+      "loss": 4.3375,
       "step": 731
     },
     {
-      "epoch": 2.661818181818182,
-      "grad_norm": 4.211031436920166,
-      "learning_rate": 5.696969696969697e-06,
-      "loss": 4.3628,
+      "epoch": 2.6425992779783396,
+      "grad_norm": 2.6663763523101807,
+      "learning_rate": 6.0168471720818295e-06,
+      "loss": 4.455,
       "step": 732
     },
     {
-      "epoch": 2.6654545454545455,
-      "grad_norm": 4.562239170074463,
-      "learning_rate": 5.636363636363637e-06,
-      "loss": 4.5267,
+      "epoch": 2.6462093862815883,
+      "grad_norm": 3.7850840091705322,
+      "learning_rate": 5.956678700361011e-06,
+      "loss": 4.5436,
       "step": 733
     },
     {
-      "epoch": 2.669090909090909,
-      "grad_norm": 3.194110155105591,
-      "learning_rate": 5.575757575757576e-06,
-      "loss": 4.2716,
+      "epoch": 2.6498194945848375,
+      "grad_norm": 3.1074018478393555,
+      "learning_rate": 5.896510228640193e-06,
+      "loss": 4.4719,
       "step": 734
     },
     {
-      "epoch": 2.672727272727273,
-      "grad_norm": 4.2862749099731445,
-      "learning_rate": 5.515151515151515e-06,
-      "loss": 4.4384,
+      "epoch": 2.6534296028880866,
+      "grad_norm": 2.576409339904785,
+      "learning_rate": 5.836341756919374e-06,
+      "loss": 4.5036,
       "step": 735
     },
     {
-      "epoch": 2.6763636363636363,
-      "grad_norm": 3.666160821914673,
-      "learning_rate": 5.4545454545454545e-06,
-      "loss": 4.4241,
+      "epoch": 2.6570397111913358,
+      "grad_norm": 5.051802158355713,
+      "learning_rate": 5.776173285198557e-06,
+      "loss": 4.3509,
       "step": 736
     },
     {
-      "epoch": 2.68,
-      "grad_norm": 6.476161003112793,
-      "learning_rate": 5.3939393939393945e-06,
-      "loss": 4.402,
+      "epoch": 2.660649819494585,
+      "grad_norm": 4.346753120422363,
+      "learning_rate": 5.716004813477738e-06,
+      "loss": 4.3802,
       "step": 737
     },
     {
-      "epoch": 2.6836363636363636,
-      "grad_norm": 6.069520950317383,
-      "learning_rate": 5.333333333333334e-06,
-      "loss": 4.5149,
+      "epoch": 2.664259927797834,
+      "grad_norm": 3.5581212043762207,
+      "learning_rate": 5.655836341756919e-06,
+      "loss": 4.6375,
       "step": 738
     },
     {
-      "epoch": 2.6872727272727275,
-      "grad_norm": 3.9702067375183105,
-      "learning_rate": 5.272727272727273e-06,
-      "loss": 4.2622,
+      "epoch": 2.667870036101083,
+      "grad_norm": 3.9020423889160156,
+      "learning_rate": 5.595667870036101e-06,
+      "loss": 4.4485,
       "step": 739
     },
     {
-      "epoch": 2.690909090909091,
-      "grad_norm": 4.099255561828613,
-      "learning_rate": 5.212121212121213e-06,
-      "loss": 4.5067,
+      "epoch": 2.671480144404332,
+      "grad_norm": 3.6418099403381348,
+      "learning_rate": 5.535499398315283e-06,
+      "loss": 4.4073,
       "step": 740
     },
     {
-      "epoch": 2.6945454545454544,
-      "grad_norm": 4.346926212310791,
-      "learning_rate": 5.151515151515152e-06,
-      "loss": 4.585,
+      "epoch": 2.675090252707581,
+      "grad_norm": 5.49989128112793,
+      "learning_rate": 5.475330926594465e-06,
+      "loss": 4.3374,
       "step": 741
     },
     {
-      "epoch": 2.6981818181818182,
-      "grad_norm": 4.047079086303711,
-      "learning_rate": 5.090909090909091e-06,
-      "loss": 4.4456,
+      "epoch": 2.67870036101083,
+      "grad_norm": 3.2897582054138184,
+      "learning_rate": 5.4151624548736465e-06,
+      "loss": 4.3568,
       "step": 742
     },
     {
-      "epoch": 2.7018181818181817,
-      "grad_norm": 3.6299378871917725,
-      "learning_rate": 5.03030303030303e-06,
-      "loss": 4.2703,
+      "epoch": 2.6823104693140793,
+      "grad_norm": 5.068946838378906,
+      "learning_rate": 5.354993983152829e-06,
+      "loss": 4.5797,
       "step": 743
     },
     {
-      "epoch": 2.7054545454545456,
-      "grad_norm": 6.5678534507751465,
-      "learning_rate": 4.96969696969697e-06,
-      "loss": 4.6659,
+      "epoch": 2.6859205776173285,
+      "grad_norm": 3.4122793674468994,
+      "learning_rate": 5.29482551143201e-06,
+      "loss": 4.4701,
       "step": 744
     },
     {
-      "epoch": 2.709090909090909,
-      "grad_norm": 5.289844989776611,
-      "learning_rate": 4.909090909090909e-06,
-      "loss": 4.4673,
+      "epoch": 2.6895306859205776,
+      "grad_norm": 4.280747890472412,
+      "learning_rate": 5.2346570397111915e-06,
+      "loss": 4.5973,
       "step": 745
     },
     {
-      "epoch": 2.712727272727273,
-      "grad_norm": 6.299747467041016,
-      "learning_rate": 4.848484848484849e-06,
-      "loss": 4.6312,
+      "epoch": 2.693140794223827,
+      "grad_norm": 5.901773452758789,
+      "learning_rate": 5.174488567990373e-06,
+      "loss": 4.4262,
       "step": 746
     },
     {
-      "epoch": 2.7163636363636363,
-      "grad_norm": 3.0065627098083496,
-      "learning_rate": 4.787878787878789e-06,
-      "loss": 4.4757,
+      "epoch": 2.696750902527076,
+      "grad_norm": 10.026396751403809,
+      "learning_rate": 5.114320096269555e-06,
+      "loss": 4.3797,
       "step": 747
     },
     {
-      "epoch": 2.7199999999999998,
-      "grad_norm": 3.4834516048431396,
-      "learning_rate": 4.727272727272727e-06,
-      "loss": 4.3239,
+      "epoch": 2.700361010830325,
+      "grad_norm": 8.616659164428711,
+      "learning_rate": 5.054151624548736e-06,
+      "loss": 4.3961,
       "step": 748
     },
     {
-      "epoch": 2.7236363636363636,
-      "grad_norm": 4.76310396194458,
-      "learning_rate": 4.666666666666667e-06,
-      "loss": 4.3299,
+      "epoch": 2.7039711191335742,
+      "grad_norm": 6.453322887420654,
+      "learning_rate": 4.993983152827919e-06,
+      "loss": 4.515,
       "step": 749
     },
     {
-      "epoch": 2.7272727272727275,
-      "grad_norm": 2.904684543609619,
-      "learning_rate": 4.606060606060606e-06,
-      "loss": 4.4176,
+      "epoch": 2.707581227436823,
+      "grad_norm": 4.593882083892822,
+      "learning_rate": 4.9338146811071e-06,
+      "loss": 4.1834,
       "step": 750
     },
     {
-      "epoch": 2.730909090909091,
-      "grad_norm": 4.35284423828125,
-      "learning_rate": 4.5454545454545455e-06,
-      "loss": 4.5483,
+      "epoch": 2.711191335740072,
+      "grad_norm": 2.6859915256500244,
+      "learning_rate": 4.873646209386281e-06,
+      "loss": 4.2954,
       "step": 751
     },
     {
-      "epoch": 2.7345454545454544,
-      "grad_norm": 4.889273166656494,
-      "learning_rate": 4.484848484848485e-06,
-      "loss": 4.4221,
+      "epoch": 2.7148014440433212,
+      "grad_norm": 2.83162522315979,
+      "learning_rate": 4.813477737665464e-06,
+      "loss": 4.4438,
       "step": 752
     },
     {
-      "epoch": 2.7381818181818183,
-      "grad_norm": 3.620983600616455,
-      "learning_rate": 4.424242424242425e-06,
-      "loss": 4.4576,
+      "epoch": 2.7184115523465704,
+      "grad_norm": 4.4429402351379395,
+      "learning_rate": 4.753309265944645e-06,
+      "loss": 4.4136,
       "step": 753
     },
     {
-      "epoch": 2.7418181818181817,
-      "grad_norm": 2.826190233230591,
-      "learning_rate": 4.363636363636364e-06,
-      "loss": 4.5104,
+      "epoch": 2.7220216606498195,
+      "grad_norm": 6.28361177444458,
+      "learning_rate": 4.693140794223827e-06,
+      "loss": 4.3655,
       "step": 754
     },
     {
-      "epoch": 2.7454545454545456,
-      "grad_norm": 2.733161687850952,
-      "learning_rate": 4.303030303030303e-06,
-      "loss": 4.5294,
+      "epoch": 2.7256317689530687,
+      "grad_norm": 5.667455673217773,
+      "learning_rate": 4.6329723225030085e-06,
+      "loss": 4.3888,
       "step": 755
     },
     {
-      "epoch": 2.749090909090909,
-      "grad_norm": 2.110649347305298,
-      "learning_rate": 4.242424242424243e-06,
-      "loss": 4.3034,
+      "epoch": 2.729241877256318,
+      "grad_norm": 3.5756163597106934,
+      "learning_rate": 4.572803850782191e-06,
+      "loss": 4.3934,
       "step": 756
     },
     {
-      "epoch": 2.752727272727273,
-      "grad_norm": 3.5162851810455322,
-      "learning_rate": 4.181818181818182e-06,
-      "loss": 4.3989,
+      "epoch": 2.7328519855595665,
+      "grad_norm": 4.108674049377441,
+      "learning_rate": 4.512635379061372e-06,
+      "loss": 4.3017,
       "step": 757
     },
     {
-      "epoch": 2.7563636363636363,
-      "grad_norm": 4.206791400909424,
-      "learning_rate": 4.1212121212121215e-06,
-      "loss": 4.5553,
+      "epoch": 2.7364620938628157,
+      "grad_norm": 2.353041648864746,
+      "learning_rate": 4.4524669073405535e-06,
+      "loss": 4.3216,
       "step": 758
     },
     {
-      "epoch": 2.76,
-      "grad_norm": 4.430637836456299,
-      "learning_rate": 4.060606060606061e-06,
-      "loss": 4.3779,
+      "epoch": 2.740072202166065,
+      "grad_norm": 5.459471225738525,
+      "learning_rate": 4.392298435619736e-06,
+      "loss": 4.4489,
       "step": 759
     },
     {
-      "epoch": 2.7636363636363637,
-      "grad_norm": 2.928250551223755,
-      "learning_rate": 4.000000000000001e-06,
-      "loss": 4.5767,
+      "epoch": 2.743682310469314,
+      "grad_norm": 3.5886504650115967,
+      "learning_rate": 4.332129963898917e-06,
+      "loss": 4.4121,
       "step": 760
     },
     {
-      "epoch": 2.767272727272727,
-      "grad_norm": 5.950421333312988,
-      "learning_rate": 3.939393939393939e-06,
-      "loss": 4.5452,
+      "epoch": 2.747292418772563,
+      "grad_norm": 6.772552967071533,
+      "learning_rate": 4.271961492178099e-06,
+      "loss": 4.4531,
       "step": 761
     },
     {
-      "epoch": 2.770909090909091,
-      "grad_norm": 4.050290107727051,
-      "learning_rate": 3.878787878787879e-06,
-      "loss": 4.3672,
+      "epoch": 2.7509025270758123,
+      "grad_norm": 3.167595148086548,
+      "learning_rate": 4.211793020457281e-06,
+      "loss": 4.6096,
       "step": 762
     },
     {
-      "epoch": 2.7745454545454544,
-      "grad_norm": 4.645779132843018,
-      "learning_rate": 3.818181818181819e-06,
-      "loss": 4.5808,
+      "epoch": 2.7545126353790614,
+      "grad_norm": 4.510582447052002,
+      "learning_rate": 4.151624548736462e-06,
+      "loss": 4.4543,
       "step": 763
     },
     {
-      "epoch": 2.7781818181818183,
-      "grad_norm": 2.2464754581451416,
-      "learning_rate": 3.757575757575758e-06,
-      "loss": 4.3538,
+      "epoch": 2.7581227436823106,
+      "grad_norm": 3.3659489154815674,
+      "learning_rate": 4.091456077015644e-06,
+      "loss": 4.2664,
       "step": 764
     },
     {
-      "epoch": 2.7818181818181817,
-      "grad_norm": 2.465804100036621,
-      "learning_rate": 3.6969696969696974e-06,
-      "loss": 4.5622,
+      "epoch": 2.7617328519855597,
+      "grad_norm": 2.412745237350464,
+      "learning_rate": 4.0312876052948256e-06,
+      "loss": 4.2983,
       "step": 765
     },
     {
-      "epoch": 2.785454545454545,
-      "grad_norm": 3.098994731903076,
-      "learning_rate": 3.636363636363636e-06,
-      "loss": 4.379,
+      "epoch": 2.765342960288809,
+      "grad_norm": 4.937220573425293,
+      "learning_rate": 3.971119133574008e-06,
+      "loss": 4.5444,
       "step": 766
     },
     {
-      "epoch": 2.789090909090909,
-      "grad_norm": 4.835888385772705,
-      "learning_rate": 3.575757575757576e-06,
-      "loss": 4.535,
+      "epoch": 2.768953068592058,
+      "grad_norm": 5.845764636993408,
+      "learning_rate": 3.910950661853189e-06,
+      "loss": 4.7001,
       "step": 767
     },
     {
-      "epoch": 2.792727272727273,
-      "grad_norm": 4.923304557800293,
-      "learning_rate": 3.515151515151515e-06,
-      "loss": 4.4813,
+      "epoch": 2.7725631768953067,
+      "grad_norm": 3.4525749683380127,
+      "learning_rate": 3.850782190132371e-06,
+      "loss": 4.2573,
       "step": 768
     },
     {
-      "epoch": 2.7963636363636364,
-      "grad_norm": 5.952007293701172,
-      "learning_rate": 3.4545454545454545e-06,
-      "loss": 4.7257,
+      "epoch": 2.776173285198556,
+      "grad_norm": 4.180619716644287,
+      "learning_rate": 3.7906137184115523e-06,
+      "loss": 4.6078,
       "step": 769
     },
     {
-      "epoch": 2.8,
-      "grad_norm": 2.3326032161712646,
-      "learning_rate": 3.3939393939393946e-06,
-      "loss": 4.4968,
+      "epoch": 2.779783393501805,
+      "grad_norm": 2.957162857055664,
+      "learning_rate": 3.730445246690734e-06,
+      "loss": 4.576,
       "step": 770
     },
     {
-      "epoch": 2.8036363636363637,
-      "grad_norm": 5.354976177215576,
-      "learning_rate": 3.3333333333333333e-06,
-      "loss": 4.1594,
+      "epoch": 2.783393501805054,
+      "grad_norm": 3.526676654815674,
+      "learning_rate": 3.6702767749699155e-06,
+      "loss": 4.3551,
       "step": 771
     },
     {
-      "epoch": 2.807272727272727,
-      "grad_norm": 3.175776481628418,
-      "learning_rate": 3.2727272727272733e-06,
-      "loss": 4.406,
+      "epoch": 2.7870036101083033,
+      "grad_norm": 3.771137237548828,
+      "learning_rate": 3.6101083032490977e-06,
+      "loss": 4.5584,
       "step": 772
     },
     {
-      "epoch": 2.810909090909091,
-      "grad_norm": 6.175566673278809,
-      "learning_rate": 3.212121212121212e-06,
-      "loss": 4.3835,
+      "epoch": 2.7906137184115525,
+      "grad_norm": 5.286788463592529,
+      "learning_rate": 3.5499398315282795e-06,
+      "loss": 4.5008,
       "step": 773
     },
     {
-      "epoch": 2.8145454545454545,
-      "grad_norm": 4.814353942871094,
-      "learning_rate": 3.1515151515151517e-06,
-      "loss": 4.3803,
+      "epoch": 2.794223826714801,
+      "grad_norm": 3.8384170532226562,
+      "learning_rate": 3.489771359807461e-06,
+      "loss": 4.6235,
       "step": 774
     },
     {
-      "epoch": 2.8181818181818183,
-      "grad_norm": 2.532165288925171,
-      "learning_rate": 3.0909090909090913e-06,
-      "loss": 4.6085,
+      "epoch": 2.7978339350180503,
+      "grad_norm": 5.631938457489014,
+      "learning_rate": 3.429602888086643e-06,
+      "loss": 4.7163,
       "step": 775
     },
     {
-      "epoch": 2.821818181818182,
-      "grad_norm": 3.019963264465332,
-      "learning_rate": 3.0303030303030305e-06,
-      "loss": 4.4147,
+      "epoch": 2.8014440433212995,
+      "grad_norm": 3.3230323791503906,
+      "learning_rate": 3.3694344163658244e-06,
+      "loss": 4.3598,
       "step": 776
     },
     {
-      "epoch": 2.825454545454545,
-      "grad_norm": 5.928140640258789,
-      "learning_rate": 2.96969696969697e-06,
-      "loss": 4.5954,
+      "epoch": 2.8050541516245486,
+      "grad_norm": 3.211805582046509,
+      "learning_rate": 3.309265944645006e-06,
+      "loss": 4.5694,
       "step": 777
     },
     {
-      "epoch": 2.829090909090909,
-      "grad_norm": 4.487292289733887,
-      "learning_rate": 2.9090909090909093e-06,
-      "loss": 4.5333,
+      "epoch": 2.808664259927798,
+      "grad_norm": 4.486380577087402,
+      "learning_rate": 3.2490974729241876e-06,
+      "loss": 4.5402,
       "step": 778
     },
     {
-      "epoch": 2.832727272727273,
-      "grad_norm": 7.2252349853515625,
-      "learning_rate": 2.8484848484848484e-06,
-      "loss": 4.3552,
+      "epoch": 2.812274368231047,
+      "grad_norm": 3.37566876411438,
+      "learning_rate": 3.1889290012033694e-06,
+      "loss": 4.3392,
       "step": 779
     },
     {
-      "epoch": 2.8363636363636364,
-      "grad_norm": 2.668553113937378,
-      "learning_rate": 2.787878787878788e-06,
-      "loss": 4.5654,
+      "epoch": 2.815884476534296,
+      "grad_norm": 3.2982022762298584,
+      "learning_rate": 3.1287605294825516e-06,
+      "loss": 4.2883,
       "step": 780
     },
     {
-      "epoch": 2.84,
-      "grad_norm": 2.857837438583374,
-      "learning_rate": 2.7272727272727272e-06,
-      "loss": 4.5263,
+      "epoch": 2.8194945848375452,
+      "grad_norm": 3.964442253112793,
+      "learning_rate": 3.068592057761733e-06,
+      "loss": 4.4602,
       "step": 781
     },
     {
-      "epoch": 2.8436363636363637,
-      "grad_norm": 2.993135929107666,
-      "learning_rate": 2.666666666666667e-06,
-      "loss": 4.6839,
+      "epoch": 2.8231046931407944,
+      "grad_norm": 2.6601409912109375,
+      "learning_rate": 3.0084235860409147e-06,
+      "loss": 4.5541,
       "step": 782
     },
     {
-      "epoch": 2.847272727272727,
-      "grad_norm": 5.227684020996094,
-      "learning_rate": 2.6060606060606064e-06,
-      "loss": 4.5512,
+      "epoch": 2.8267148014440435,
+      "grad_norm": 2.461411237716675,
+      "learning_rate": 2.9482551143200965e-06,
+      "loss": 4.2491,
       "step": 783
     },
     {
-      "epoch": 2.850909090909091,
-      "grad_norm": 3.6668307781219482,
-      "learning_rate": 2.5454545454545456e-06,
-      "loss": 4.4039,
+      "epoch": 2.8303249097472927,
+      "grad_norm": 3.935789108276367,
+      "learning_rate": 2.8880866425992783e-06,
+      "loss": 4.7376,
       "step": 784
     },
     {
-      "epoch": 2.8545454545454545,
-      "grad_norm": 2.9002740383148193,
-      "learning_rate": 2.484848484848485e-06,
-      "loss": 4.4902,
+      "epoch": 2.8339350180505414,
+      "grad_norm": 3.178452491760254,
+      "learning_rate": 2.8279181708784597e-06,
+      "loss": 4.2914,
       "step": 785
     },
     {
-      "epoch": 2.8581818181818184,
-      "grad_norm": 2.6871190071105957,
-      "learning_rate": 2.4242424242424244e-06,
-      "loss": 4.4637,
+      "epoch": 2.8375451263537905,
+      "grad_norm": 2.401991128921509,
+      "learning_rate": 2.7677496991576415e-06,
+      "loss": 4.3846,
       "step": 786
     },
     {
-      "epoch": 2.861818181818182,
-      "grad_norm": 2.605130672454834,
-      "learning_rate": 2.3636363636363636e-06,
-      "loss": 4.3348,
+      "epoch": 2.8411552346570397,
+      "grad_norm": 2.8064956665039062,
+      "learning_rate": 2.7075812274368233e-06,
+      "loss": 4.5398,
       "step": 787
     },
     {
-      "epoch": 2.8654545454545453,
-      "grad_norm": 2.137845754623413,
-      "learning_rate": 2.303030303030303e-06,
-      "loss": 4.3367,
+      "epoch": 2.844765342960289,
+      "grad_norm": 3.338486433029175,
+      "learning_rate": 2.647412755716005e-06,
+      "loss": 4.5519,
       "step": 788
     },
     {
-      "epoch": 2.869090909090909,
-      "grad_norm": 4.47348165512085,
-      "learning_rate": 2.2424242424242423e-06,
-      "loss": 4.3623,
+      "epoch": 2.848375451263538,
+      "grad_norm": 1.800193190574646,
+      "learning_rate": 2.5872442839951864e-06,
+      "loss": 4.4321,
       "step": 789
     },
     {
-      "epoch": 2.8727272727272726,
-      "grad_norm": 2.943335771560669,
-      "learning_rate": 2.181818181818182e-06,
-      "loss": 4.3887,
+      "epoch": 2.851985559566787,
+      "grad_norm": 2.2621030807495117,
+      "learning_rate": 2.527075812274368e-06,
+      "loss": 4.5175,
       "step": 790
     },
     {
-      "epoch": 2.8763636363636365,
-      "grad_norm": 3.716531753540039,
-      "learning_rate": 2.1212121212121216e-06,
-      "loss": 4.3412,
+      "epoch": 2.855595667870036,
+      "grad_norm": 3.4302749633789062,
+      "learning_rate": 2.46690734055355e-06,
+      "loss": 4.4695,
       "step": 791
     },
     {
-      "epoch": 2.88,
-      "grad_norm": 2.894035577774048,
-      "learning_rate": 2.0606060606060607e-06,
-      "loss": 4.3741,
+      "epoch": 2.859205776173285,
+      "grad_norm": 5.186544418334961,
+      "learning_rate": 2.406738868832732e-06,
+      "loss": 4.3755,
       "step": 792
     },
     {
-      "epoch": 2.8836363636363638,
-      "grad_norm": 3.7420289516448975,
-      "learning_rate": 2.0000000000000003e-06,
-      "loss": 4.5828,
+      "epoch": 2.862815884476534,
+      "grad_norm": 2.793041706085205,
+      "learning_rate": 2.3465703971119136e-06,
+      "loss": 4.2424,
       "step": 793
     },
     {
-      "epoch": 2.887272727272727,
-      "grad_norm": 2.9738407135009766,
-      "learning_rate": 1.9393939393939395e-06,
-      "loss": 4.4787,
+      "epoch": 2.8664259927797833,
+      "grad_norm": 2.6105570793151855,
+      "learning_rate": 2.2864019253910954e-06,
+      "loss": 4.441,
       "step": 794
     },
     {
-      "epoch": 2.8909090909090907,
-      "grad_norm": 2.441652297973633,
-      "learning_rate": 1.878787878787879e-06,
-      "loss": 4.51,
+      "epoch": 2.8700361010830324,
+      "grad_norm": 2.990581750869751,
+      "learning_rate": 2.2262334536702767e-06,
+      "loss": 4.4562,
       "step": 795
     },
     {
-      "epoch": 2.8945454545454545,
-      "grad_norm": 4.284375190734863,
-      "learning_rate": 1.818181818181818e-06,
-      "loss": 4.2759,
+      "epoch": 2.8736462093862816,
+      "grad_norm": 2.425126791000366,
+      "learning_rate": 2.1660649819494585e-06,
+      "loss": 4.3576,
       "step": 796
     },
     {
-      "epoch": 2.8981818181818184,
-      "grad_norm": 5.655579090118408,
-      "learning_rate": 1.7575757575757575e-06,
-      "loss": 4.4516,
+      "epoch": 2.8772563176895307,
+      "grad_norm": 3.0731112957000732,
+      "learning_rate": 2.1058965102286403e-06,
+      "loss": 4.4477,
       "step": 797
     },
     {
-      "epoch": 2.901818181818182,
-      "grad_norm": 3.4644057750701904,
-      "learning_rate": 1.6969696969696973e-06,
-      "loss": 4.3542,
+      "epoch": 2.88086642599278,
+      "grad_norm": 2.6812827587127686,
+      "learning_rate": 2.045728038507822e-06,
+      "loss": 4.536,
       "step": 798
     },
     {
-      "epoch": 2.9054545454545453,
-      "grad_norm": 2.9349281787872314,
-      "learning_rate": 1.6363636363636367e-06,
-      "loss": 4.4623,
+      "epoch": 2.884476534296029,
+      "grad_norm": 2.6943960189819336,
+      "learning_rate": 1.985559566787004e-06,
+      "loss": 4.4028,
       "step": 799
     },
     {
-      "epoch": 2.909090909090909,
-      "grad_norm": 2.0899038314819336,
-      "learning_rate": 1.5757575757575759e-06,
-      "loss": 4.3693,
+      "epoch": 2.888086642599278,
+      "grad_norm": 2.157318592071533,
+      "learning_rate": 1.9253910950661857e-06,
+      "loss": 4.4981,
       "step": 800
     },
     {
-      "epoch": 2.9127272727272726,
-      "grad_norm": 3.156378746032715,
-      "learning_rate": 1.5151515151515152e-06,
-      "loss": 4.4346,
+      "epoch": 2.8916967509025273,
+      "grad_norm": 3.369413375854492,
+      "learning_rate": 1.865222623345367e-06,
+      "loss": 4.3188,
       "step": 801
     },
     {
-      "epoch": 2.9163636363636365,
-      "grad_norm": 1.8763753175735474,
-      "learning_rate": 1.4545454545454546e-06,
-      "loss": 4.3416,
+      "epoch": 2.895306859205776,
+      "grad_norm": 2.947744607925415,
+      "learning_rate": 1.8050541516245488e-06,
+      "loss": 4.3764,
       "step": 802
     },
     {
-      "epoch": 2.92,
-      "grad_norm": 2.1865854263305664,
-      "learning_rate": 1.393939393939394e-06,
-      "loss": 4.4388,
+      "epoch": 2.898916967509025,
+      "grad_norm": 2.9001619815826416,
+      "learning_rate": 1.7448856799037304e-06,
+      "loss": 4.3952,
       "step": 803
     },
     {
-      "epoch": 2.923636363636364,
-      "grad_norm": 2.035835027694702,
-      "learning_rate": 1.3333333333333334e-06,
-      "loss": 4.3105,
+      "epoch": 2.9025270758122743,
+      "grad_norm": 3.887981653213501,
+      "learning_rate": 1.6847172081829122e-06,
+      "loss": 4.4897,
       "step": 804
     },
     {
-      "epoch": 2.9272727272727272,
-      "grad_norm": 2.976594924926758,
-      "learning_rate": 1.2727272727272728e-06,
-      "loss": 4.4205,
+      "epoch": 2.9061371841155235,
+      "grad_norm": 2.611881732940674,
+      "learning_rate": 1.6245487364620938e-06,
+      "loss": 4.4093,
       "step": 805
     },
     {
-      "epoch": 2.9309090909090907,
-      "grad_norm": 4.820901393890381,
-      "learning_rate": 1.2121212121212122e-06,
-      "loss": 4.2923,
+      "epoch": 2.9097472924187726,
+      "grad_norm": 3.8872454166412354,
+      "learning_rate": 1.5643802647412758e-06,
+      "loss": 4.3923,
       "step": 806
     },
     {
-      "epoch": 2.9345454545454546,
-      "grad_norm": 3.00866436958313,
-      "learning_rate": 1.1515151515151516e-06,
-      "loss": 4.5881,
+      "epoch": 2.9133574007220218,
+      "grad_norm": 2.8728034496307373,
+      "learning_rate": 1.5042117930204574e-06,
+      "loss": 4.4359,
       "step": 807
     },
     {
-      "epoch": 2.9381818181818184,
-      "grad_norm": 2.354827880859375,
-      "learning_rate": 1.090909090909091e-06,
-      "loss": 4.2351,
+      "epoch": 2.916967509025271,
+      "grad_norm": 1.9554171562194824,
+      "learning_rate": 1.4440433212996392e-06,
+      "loss": 4.2807,
       "step": 808
     },
     {
-      "epoch": 2.941818181818182,
-      "grad_norm": 2.7474520206451416,
-      "learning_rate": 1.0303030303030304e-06,
-      "loss": 4.4496,
+      "epoch": 2.9205776173285196,
+      "grad_norm": 2.1981494426727295,
+      "learning_rate": 1.3838748495788207e-06,
+      "loss": 4.4444,
       "step": 809
     },
     {
-      "epoch": 2.9454545454545453,
-      "grad_norm": 3.4006338119506836,
-      "learning_rate": 9.696969696969698e-07,
-      "loss": 4.5303,
+      "epoch": 2.9241877256317688,
+      "grad_norm": 2.9973559379577637,
+      "learning_rate": 1.3237063778580025e-06,
+      "loss": 4.4673,
       "step": 810
     },
     {
-      "epoch": 2.949090909090909,
-      "grad_norm": 2.1111907958984375,
-      "learning_rate": 9.09090909090909e-07,
-      "loss": 4.3686,
+      "epoch": 2.927797833935018,
+      "grad_norm": 2.8037450313568115,
+      "learning_rate": 1.263537906137184e-06,
+      "loss": 4.5387,
       "step": 811
     },
     {
-      "epoch": 2.9527272727272726,
-      "grad_norm": 2.6187009811401367,
-      "learning_rate": 8.484848484848486e-07,
-      "loss": 4.236,
+      "epoch": 2.931407942238267,
+      "grad_norm": 1.915396809577942,
+      "learning_rate": 1.203369434416366e-06,
+      "loss": 4.3888,
       "step": 812
     },
     {
-      "epoch": 2.9563636363636365,
-      "grad_norm": 7.847300052642822,
-      "learning_rate": 7.878787878787879e-07,
-      "loss": 4.3803,
+      "epoch": 2.935018050541516,
+      "grad_norm": 4.414287090301514,
+      "learning_rate": 1.1432009626955477e-06,
+      "loss": 4.3828,
       "step": 813
     },
     {
-      "epoch": 2.96,
-      "grad_norm": 2.505014419555664,
-      "learning_rate": 7.272727272727273e-07,
-      "loss": 4.4579,
+      "epoch": 2.9386281588447654,
+      "grad_norm": 2.6727025508880615,
+      "learning_rate": 1.0830324909747293e-06,
+      "loss": 4.4088,
       "step": 814
     },
     {
-      "epoch": 2.963636363636364,
-      "grad_norm": 2.3602616786956787,
-      "learning_rate": 6.666666666666667e-07,
-      "loss": 4.3183,
+      "epoch": 2.9422382671480145,
+      "grad_norm": 4.590742588043213,
+      "learning_rate": 1.022864019253911e-06,
+      "loss": 4.3833,
       "step": 815
     },
     {
-      "epoch": 2.9672727272727273,
-      "grad_norm": 2.405813694000244,
-      "learning_rate": 6.060606060606061e-07,
-      "loss": 4.3485,
+      "epoch": 2.9458483754512637,
+      "grad_norm": 4.188723087310791,
+      "learning_rate": 9.626955475330928e-07,
+      "loss": 4.2543,
       "step": 816
     },
     {
-      "epoch": 2.9709090909090907,
-      "grad_norm": 3.003232717514038,
-      "learning_rate": 5.454545454545455e-07,
-      "loss": 4.5807,
+      "epoch": 2.949458483754513,
+      "grad_norm": 1.739635944366455,
+      "learning_rate": 9.025270758122744e-07,
+      "loss": 4.2869,
       "step": 817
     },
     {
-      "epoch": 2.9745454545454546,
-      "grad_norm": 2.4426870346069336,
-      "learning_rate": 4.848484848484849e-07,
-      "loss": 4.3509,
+      "epoch": 2.953068592057762,
+      "grad_norm": 2.8158509731292725,
+      "learning_rate": 8.423586040914561e-07,
+      "loss": 4.5182,
       "step": 818
     },
     {
-      "epoch": 2.978181818181818,
-      "grad_norm": 3.0719637870788574,
-      "learning_rate": 4.242424242424243e-07,
-      "loss": 4.3542,
+      "epoch": 2.956678700361011,
+      "grad_norm": 3.629409074783325,
+      "learning_rate": 7.821901323706379e-07,
+      "loss": 4.6912,
       "step": 819
     },
     {
-      "epoch": 2.981818181818182,
-      "grad_norm": 2.4720072746276855,
-      "learning_rate": 3.6363636363636366e-07,
-      "loss": 4.4289,
+      "epoch": 2.96028880866426,
+      "grad_norm": 2.2690486907958984,
+      "learning_rate": 7.220216606498196e-07,
+      "loss": 4.55,
       "step": 820
     },
     {
-      "epoch": 2.9854545454545454,
-      "grad_norm": 3.5286593437194824,
-      "learning_rate": 3.0303030303030305e-07,
-      "loss": 4.6304,
+      "epoch": 2.963898916967509,
+      "grad_norm": 3.6004555225372314,
+      "learning_rate": 6.618531889290013e-07,
+      "loss": 4.6055,
       "step": 821
     },
     {
-      "epoch": 2.9890909090909092,
-      "grad_norm": 2.771474838256836,
-      "learning_rate": 2.4242424242424244e-07,
-      "loss": 4.4305,
+      "epoch": 2.967509025270758,
+      "grad_norm": 3.75754451751709,
+      "learning_rate": 6.01684717208183e-07,
+      "loss": 4.4194,
       "step": 822
     },
     {
-      "epoch": 2.9927272727272727,
-      "grad_norm": 1.9957259893417358,
-      "learning_rate": 1.8181818181818183e-07,
-      "loss": 4.3848,
+      "epoch": 2.9711191335740073,
+      "grad_norm": 5.5864691734313965,
+      "learning_rate": 5.415162454873646e-07,
+      "loss": 4.3831,
       "step": 823
     },
     {
-      "epoch": 2.996363636363636,
-      "grad_norm": 3.018747568130493,
-      "learning_rate": 1.2121212121212122e-07,
-      "loss": 4.3957,
+      "epoch": 2.9747292418772564,
+      "grad_norm": 2.2573602199554443,
+      "learning_rate": 4.813477737665464e-07,
+      "loss": 4.2861,
       "step": 824
     },
     {
-      "epoch": 3.0,
-      "grad_norm": 5.8460917472839355,
-      "learning_rate": 6.060606060606061e-08,
-      "loss": 4.7331,
+      "epoch": 2.9783393501805056,
+      "grad_norm": 2.9080207347869873,
+      "learning_rate": 4.2117930204572805e-07,
+      "loss": 4.6265,
       "step": 825
+    },
+    {
+      "epoch": 2.9819494584837543,
+      "grad_norm": 1.611305594444275,
+      "learning_rate": 3.610108303249098e-07,
+      "loss": 4.3987,
+      "step": 826
+    },
+    {
+      "epoch": 2.9855595667870034,
+      "grad_norm": 4.546231269836426,
+      "learning_rate": 3.008423586040915e-07,
+      "loss": 4.3495,
+      "step": 827
+    },
+    {
+      "epoch": 2.9891696750902526,
+      "grad_norm": 1.9480468034744263,
+      "learning_rate": 2.406738868832732e-07,
+      "loss": 4.3835,
+      "step": 828
+    },
+    {
+      "epoch": 2.9927797833935017,
+      "grad_norm": 2.429931879043579,
+      "learning_rate": 1.805054151624549e-07,
+      "loss": 4.3572,
+      "step": 829
+    },
+    {
+      "epoch": 2.996389891696751,
+      "grad_norm": 3.491809844970703,
+      "learning_rate": 1.203369434416366e-07,
+      "loss": 4.3179,
+      "step": 830
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 2.282447576522827,
+      "learning_rate": 6.01684717208183e-08,
+      "loss": 4.4063,
+      "step": 831
     }
   ],
   "logging_steps": 1,
-  "max_steps": 825,
+  "max_steps": 831,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
   "save_steps": 10,
@@ -5802,7 +5844,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.0032173593657344e+16,
+  "total_flos": 1.0123541294874624e+16,
   "train_batch_size": 2,
   "trial_name": null,
   "trial_params": null