diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,49034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.00915029091389184,
+  "eval_steps": 500,
+  "global_step": 7000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.307184416270263e-06,
+      "grad_norm": 6.268255710601807,
+      "learning_rate": 0.0,
+      "loss": 7.0486,
+      "step": 1
+    },
+    {
+      "epoch": 2.614368832540526e-06,
+      "grad_norm": 5.648464202880859,
+      "learning_rate": 5e-06,
+      "loss": 6.6548,
+      "step": 2
+    },
+    {
+      "epoch": 3.921553248810789e-06,
+      "grad_norm": 3.917975902557373,
+      "learning_rate": 1e-05,
+      "loss": 5.9005,
+      "step": 3
+    },
+    {
+      "epoch": 5.228737665081052e-06,
+      "grad_norm": 5.040616989135742,
+      "learning_rate": 1.5e-05,
+      "loss": 6.4287,
+      "step": 4
+    },
+    {
+      "epoch": 6.535922081351315e-06,
+      "grad_norm": 4.18554162979126,
+      "learning_rate": 2e-05,
+      "loss": 5.8283,
+      "step": 5
+    },
+    {
+      "epoch": 7.843106497621578e-06,
+      "grad_norm": 3.837583303451538,
+      "learning_rate": 2.5e-05,
+      "loss": 5.8711,
+      "step": 6
+    },
+    {
+      "epoch": 9.150290913891842e-06,
+      "grad_norm": 3.9937901496887207,
+      "learning_rate": 3e-05,
+      "loss": 6.2397,
+      "step": 7
+    },
+    {
+      "epoch": 1.0457475330162104e-05,
+      "grad_norm": 6.323297023773193,
+      "learning_rate": 3.5e-05,
+      "loss": 6.0184,
+      "step": 8
+    },
+    {
+      "epoch": 1.1764659746432368e-05,
+      "grad_norm": 4.541579246520996,
+      "learning_rate": 4e-05,
+      "loss": 4.9289,
+      "step": 9
+    },
+    {
+      "epoch": 1.307184416270263e-05,
+      "grad_norm": 4.3712358474731445,
+      "learning_rate": 4.5e-05,
+      "loss": 5.5431,
+      "step": 10
+    },
+    {
+      "epoch": 1.4379028578972894e-05,
+      "grad_norm": 5.535464763641357,
+      "learning_rate": 5e-05,
+      "loss": 5.4152,
+      "step": 11
+    },
+    {
+      "epoch": 1.5686212995243156e-05,
+      "grad_norm": 4.411280632019043,
+      "learning_rate": 4.9994994994994995e-05,
+      "loss": 5.0227,
+      "step": 12
+    },
+    {
+      "epoch": 1.6993397411513418e-05,
+      "grad_norm": 8.777021408081055,
+      "learning_rate": 4.9989989989989995e-05,
+      "loss": 4.628,
+      "step": 13
+    },
+    {
+      "epoch": 1.8300581827783683e-05,
+      "grad_norm": 11.268546104431152,
+      "learning_rate": 4.998498498498499e-05,
+      "loss": 4.4529,
+      "step": 14
+    },
+    {
+      "epoch": 1.9607766244053945e-05,
+      "grad_norm": 11.081679344177246,
+      "learning_rate": 4.997997997997999e-05,
+      "loss": 3.7927,
+      "step": 15
+    },
+    {
+      "epoch": 2.0914950660324208e-05,
+      "grad_norm": 17.39815330505371,
+      "learning_rate": 4.9974974974974974e-05,
+      "loss": 3.8354,
+      "step": 16
+    },
+    {
+      "epoch": 2.222213507659447e-05,
+      "grad_norm": 19.696117401123047,
+      "learning_rate": 4.9969969969969974e-05,
+      "loss": 2.6296,
+      "step": 17
+    },
+    {
+      "epoch": 2.3529319492864735e-05,
+      "grad_norm": 9.491320610046387,
+      "learning_rate": 4.996496496496497e-05,
+      "loss": 2.34,
+      "step": 18
+    },
+    {
+      "epoch": 2.4836503909134997e-05,
+      "grad_norm": 8.10347843170166,
+      "learning_rate": 4.995995995995996e-05,
+      "loss": 2.8097,
+      "step": 19
+    },
+    {
+      "epoch": 2.614368832540526e-05,
+      "grad_norm": 10.142881393432617,
+      "learning_rate": 4.995495495495496e-05,
+      "loss": 2.418,
+      "step": 20
+    },
+    {
+      "epoch": 2.745087274167552e-05,
+      "grad_norm": 5.666007041931152,
+      "learning_rate": 4.994994994994995e-05,
+      "loss": 2.4814,
+      "step": 21
+    },
+    {
+      "epoch": 2.8758057157945787e-05,
+      "grad_norm": 2.6253504753112793,
+      "learning_rate": 4.994494494494495e-05,
+      "loss": 3.0029,
+      "step": 22
+    },
+    {
+      "epoch": 3.006524157421605e-05,
+      "grad_norm": 1.9413844347000122,
+      "learning_rate": 4.9939939939939945e-05,
+      "loss": 2.7044,
+      "step": 23
+    },
+    {
+      "epoch": 3.137242599048631e-05,
+      "grad_norm": 2.3702781200408936,
+      "learning_rate": 4.993493493493494e-05,
+      "loss": 3.2683,
+      "step": 24
+    },
+    {
+      "epoch": 3.267961040675658e-05,
+      "grad_norm": 1.8796427249908447,
+      "learning_rate": 4.992992992992993e-05,
+      "loss": 2.6591,
+      "step": 25
+    },
+    {
+      "epoch": 3.3986794823026836e-05,
+      "grad_norm": 1.8859280347824097,
+      "learning_rate": 4.9924924924924924e-05,
+      "loss": 2.6567,
+      "step": 26
+    },
+    {
+      "epoch": 3.52939792392971e-05,
+      "grad_norm": 1.8021355867385864,
+      "learning_rate": 4.9919919919919924e-05,
+      "loss": 2.0643,
+      "step": 27
+    },
+    {
+      "epoch": 3.660116365556737e-05,
+      "grad_norm": 3.3398263454437256,
+      "learning_rate": 4.991491491491492e-05,
+      "loss": 1.9806,
+      "step": 28
+    },
+    {
+      "epoch": 3.7908348071837625e-05,
+      "grad_norm": 1.7556695938110352,
+      "learning_rate": 4.9909909909909917e-05,
+      "loss": 2.5847,
+      "step": 29
+    },
+    {
+      "epoch": 3.921553248810789e-05,
+      "grad_norm": 2.361534833908081,
+      "learning_rate": 4.990490490490491e-05,
+      "loss": 3.1982,
+      "step": 30
+    },
+    {
+      "epoch": 4.052271690437815e-05,
+      "grad_norm": 2.1116466522216797,
+      "learning_rate": 4.98998998998999e-05,
+      "loss": 2.1447,
+      "step": 31
+    },
+    {
+      "epoch": 4.1829901320648415e-05,
+      "grad_norm": 2.76759672164917,
+      "learning_rate": 4.9894894894894896e-05,
+      "loss": 2.2138,
+      "step": 32
+    },
+    {
+      "epoch": 4.313708573691868e-05,
+      "grad_norm": 1.744004726409912,
+      "learning_rate": 4.988988988988989e-05,
+      "loss": 2.6433,
+      "step": 33
+    },
+    {
+      "epoch": 4.444427015318894e-05,
+      "grad_norm": 1.946826457977295,
+      "learning_rate": 4.988488488488489e-05,
+      "loss": 2.814,
+      "step": 34
+    },
+    {
+      "epoch": 4.5751454569459205e-05,
+      "grad_norm": 1.7588024139404297,
+      "learning_rate": 4.987987987987988e-05,
+      "loss": 1.9231,
+      "step": 35
+    },
+    {
+      "epoch": 4.705863898572947e-05,
+      "grad_norm": 1.920486330986023,
+      "learning_rate": 4.9874874874874874e-05,
+      "loss": 2.3058,
+      "step": 36
+    },
+    {
+      "epoch": 4.836582340199973e-05,
+      "grad_norm": 2.172233819961548,
+      "learning_rate": 4.9869869869869874e-05,
+      "loss": 1.9368,
+      "step": 37
+    },
+    {
+      "epoch": 4.9673007818269995e-05,
+      "grad_norm": 1.9940109252929688,
+      "learning_rate": 4.986486486486487e-05,
+      "loss": 2.4671,
+      "step": 38
+    },
+    {
+      "epoch": 5.0980192234540254e-05,
+      "grad_norm": 2.1106722354888916,
+      "learning_rate": 4.985985985985987e-05,
+      "loss": 1.2266,
+      "step": 39
+    },
+    {
+      "epoch": 5.228737665081052e-05,
+      "grad_norm": 1.9987304210662842,
+      "learning_rate": 4.985485485485486e-05,
+      "loss": 2.5097,
+      "step": 40
+    },
+    {
+      "epoch": 5.3594561067080785e-05,
+      "grad_norm": 4.258189678192139,
+      "learning_rate": 4.984984984984985e-05,
+      "loss": 2.7884,
+      "step": 41
+    },
+    {
+      "epoch": 5.490174548335104e-05,
+      "grad_norm": 3.9918763637542725,
+      "learning_rate": 4.9844844844844846e-05,
+      "loss": 1.8792,
+      "step": 42
+    },
+    {
+      "epoch": 5.620892989962131e-05,
+      "grad_norm": 2.13405442237854,
+      "learning_rate": 4.983983983983984e-05,
+      "loss": 2.4141,
+      "step": 43
+    },
+    {
+      "epoch": 5.7516114315891574e-05,
+      "grad_norm": 2.369387626647949,
+      "learning_rate": 4.983483483483484e-05,
+      "loss": 2.4503,
+      "step": 44
+    },
+    {
+      "epoch": 5.882329873216183e-05,
+      "grad_norm": 2.4659979343414307,
+      "learning_rate": 4.982982982982983e-05,
+      "loss": 1.6704,
+      "step": 45
+    },
+    {
+      "epoch": 6.01304831484321e-05,
+      "grad_norm": 2.8049533367156982,
+      "learning_rate": 4.982482482482483e-05,
+      "loss": 2.2134,
+      "step": 46
+    },
+    {
+      "epoch": 6.143766756470236e-05,
+      "grad_norm": 3.652784824371338,
+      "learning_rate": 4.9819819819819824e-05,
+      "loss": 2.4586,
+      "step": 47
+    },
+    {
+      "epoch": 6.274485198097262e-05,
+      "grad_norm": 2.76751971244812,
+      "learning_rate": 4.981481481481482e-05,
+      "loss": 2.6559,
+      "step": 48
+    },
+    {
+      "epoch": 6.405203639724289e-05,
+      "grad_norm": 2.337362051010132,
+      "learning_rate": 4.980980980980981e-05,
+      "loss": 2.6549,
+      "step": 49
+    },
+    {
+      "epoch": 6.535922081351315e-05,
+      "grad_norm": 2.5444376468658447,
+      "learning_rate": 4.98048048048048e-05,
+      "loss": 2.041,
+      "step": 50
+    },
+    {
+      "epoch": 6.666640522978342e-05,
+      "grad_norm": 2.3144052028656006,
+      "learning_rate": 4.97997997997998e-05,
+      "loss": 2.4423,
+      "step": 51
+    },
+    {
+      "epoch": 6.797358964605367e-05,
+      "grad_norm": 2.647606372833252,
+      "learning_rate": 4.9794794794794796e-05,
+      "loss": 2.0042,
+      "step": 52
+    },
+    {
+      "epoch": 6.928077406232394e-05,
+      "grad_norm": 2.2047841548919678,
+      "learning_rate": 4.9789789789789795e-05,
+      "loss": 2.7494,
+      "step": 53
+    },
+    {
+      "epoch": 7.05879584785942e-05,
+      "grad_norm": 2.5686540603637695,
+      "learning_rate": 4.978478478478479e-05,
+      "loss": 2.6443,
+      "step": 54
+    },
+    {
+      "epoch": 7.189514289486447e-05,
+      "grad_norm": 2.634040594100952,
+      "learning_rate": 4.977977977977978e-05,
+      "loss": 2.3489,
+      "step": 55
+    },
+    {
+      "epoch": 7.320232731113473e-05,
+      "grad_norm": 3.5904624462127686,
+      "learning_rate": 4.977477477477478e-05,
+      "loss": 2.325,
+      "step": 56
+    },
+    {
+      "epoch": 7.450951172740499e-05,
+      "grad_norm": 2.7417197227478027,
+      "learning_rate": 4.976976976976977e-05,
+      "loss": 2.109,
+      "step": 57
+    },
+    {
+      "epoch": 7.581669614367525e-05,
+      "grad_norm": 3.90405011177063,
+      "learning_rate": 4.976476476476477e-05,
+      "loss": 1.9777,
+      "step": 58
+    },
+    {
+      "epoch": 7.712388055994552e-05,
+      "grad_norm": 3.0904104709625244,
+      "learning_rate": 4.975975975975976e-05,
+      "loss": 2.964,
+      "step": 59
+    },
+    {
+      "epoch": 7.843106497621578e-05,
+      "grad_norm": 6.283293724060059,
+      "learning_rate": 4.975475475475476e-05,
+      "loss": 2.7203,
+      "step": 60
+    },
+    {
+      "epoch": 7.973824939248605e-05,
+      "grad_norm": 15.482819557189941,
+      "learning_rate": 4.974974974974975e-05,
+      "loss": 2.0104,
+      "step": 61
+    },
+    {
+      "epoch": 8.10454338087563e-05,
+      "grad_norm": 5.350860118865967,
+      "learning_rate": 4.9744744744744746e-05,
+      "loss": 2.4738,
+      "step": 62
+    },
+    {
+      "epoch": 8.235261822502656e-05,
+      "grad_norm": 2.5188300609588623,
+      "learning_rate": 4.9739739739739745e-05,
+      "loss": 2.3628,
+      "step": 63
+    },
+    {
+      "epoch": 8.365980264129683e-05,
+      "grad_norm": 3.6089985370635986,
+      "learning_rate": 4.973473473473474e-05,
+      "loss": 2.0068,
+      "step": 64
+    },
+    {
+      "epoch": 8.49669870575671e-05,
+      "grad_norm": 4.064090251922607,
+      "learning_rate": 4.972972972972974e-05,
+      "loss": 1.9134,
+      "step": 65
+    },
+    {
+      "epoch": 8.627417147383736e-05,
+      "grad_norm": 4.058968544006348,
+      "learning_rate": 4.9724724724724724e-05,
+      "loss": 2.654,
+      "step": 66
+    },
+    {
+      "epoch": 8.758135589010763e-05,
+      "grad_norm": 4.4842023849487305,
+      "learning_rate": 4.971971971971972e-05,
+      "loss": 2.0954,
+      "step": 67
+    },
+    {
+      "epoch": 8.888854030637788e-05,
+      "grad_norm": 4.771529674530029,
+      "learning_rate": 4.971471471471472e-05,
+      "loss": 2.3041,
+      "step": 68
+    },
+    {
+      "epoch": 9.019572472264814e-05,
+      "grad_norm": 5.2841267585754395,
+      "learning_rate": 4.970970970970971e-05,
+      "loss": 1.6284,
+      "step": 69
+    },
+    {
+      "epoch": 9.150290913891841e-05,
+      "grad_norm": 4.454013824462891,
+      "learning_rate": 4.970470470470471e-05,
+      "loss": 2.4258,
+      "step": 70
+    },
+    {
+      "epoch": 9.281009355518868e-05,
+      "grad_norm": 4.686120986938477,
+      "learning_rate": 4.96996996996997e-05,
+      "loss": 1.7159,
+      "step": 71
+    },
+    {
+      "epoch": 9.411727797145894e-05,
+      "grad_norm": 4.039860248565674,
+      "learning_rate": 4.9694694694694696e-05,
+      "loss": 2.4571,
+      "step": 72
+    },
+    {
+      "epoch": 9.542446238772919e-05,
+      "grad_norm": 6.6892805099487305,
+      "learning_rate": 4.9689689689689696e-05,
+      "loss": 1.5466,
+      "step": 73
+    },
+    {
+      "epoch": 9.673164680399946e-05,
+      "grad_norm": 4.6148362159729,
+      "learning_rate": 4.968468468468468e-05,
+      "loss": 2.1115,
+      "step": 74
+    },
+    {
+      "epoch": 9.803883122026972e-05,
+      "grad_norm": 4.417403697967529,
+      "learning_rate": 4.967967967967968e-05,
+      "loss": 2.0975,
+      "step": 75
+    },
+    {
+      "epoch": 9.934601563653999e-05,
+      "grad_norm": 6.658927917480469,
+      "learning_rate": 4.9674674674674674e-05,
+      "loss": 1.4255,
+      "step": 76
+    },
+    {
+      "epoch": 0.00010065320005281026,
+      "grad_norm": 5.162082195281982,
+      "learning_rate": 4.9669669669669674e-05,
+      "loss": 2.1406,
+      "step": 77
+    },
+    {
+      "epoch": 0.00010196038446908051,
+      "grad_norm": 3.9566612243652344,
+      "learning_rate": 4.966466466466467e-05,
+      "loss": 2.6316,
+      "step": 78
+    },
+    {
+      "epoch": 0.00010326756888535077,
+      "grad_norm": 5.431234836578369,
+      "learning_rate": 4.965965965965966e-05,
+      "loss": 1.9831,
+      "step": 79
+    },
+    {
+      "epoch": 0.00010457475330162104,
+      "grad_norm": 7.695541858673096,
+      "learning_rate": 4.965465465465466e-05,
+      "loss": 1.3602,
+      "step": 80
+    },
+    {
+      "epoch": 0.0001058819377178913,
+      "grad_norm": 5.668361663818359,
+      "learning_rate": 4.964964964964965e-05,
+      "loss": 1.2717,
+      "step": 81
+    },
+    {
+      "epoch": 0.00010718912213416157,
+      "grad_norm": 6.2884202003479,
+      "learning_rate": 4.9644644644644646e-05,
+      "loss": 2.3994,
+      "step": 82
+    },
+    {
+      "epoch": 0.00010849630655043183,
+      "grad_norm": 5.497302055358887,
+      "learning_rate": 4.963963963963964e-05,
+      "loss": 2.0307,
+      "step": 83
+    },
+    {
+      "epoch": 0.00010980349096670209,
+      "grad_norm": 6.879200458526611,
+      "learning_rate": 4.963463463463464e-05,
+      "loss": 1.4382,
+      "step": 84
+    },
+    {
+      "epoch": 0.00011111067538297235,
+      "grad_norm": 5.275783538818359,
+      "learning_rate": 4.962962962962963e-05,
+      "loss": 1.8855,
+      "step": 85
+    },
+    {
+      "epoch": 0.00011241785979924262,
+      "grad_norm": 5.1828460693359375,
+      "learning_rate": 4.9624624624624625e-05,
+      "loss": 2.1653,
+      "step": 86
+    },
+    {
+      "epoch": 0.00011372504421551288,
+      "grad_norm": 4.819062232971191,
+      "learning_rate": 4.9619619619619624e-05,
+      "loss": 2.6426,
+      "step": 87
+    },
+    {
+      "epoch": 0.00011503222863178315,
+      "grad_norm": 4.319160461425781,
+      "learning_rate": 4.961461461461462e-05,
+      "loss": 2.6315,
+      "step": 88
+    },
+    {
+      "epoch": 0.0001163394130480534,
+      "grad_norm": 5.244660377502441,
+      "learning_rate": 4.960960960960962e-05,
+      "loss": 2.4654,
+      "step": 89
+    },
+    {
+      "epoch": 0.00011764659746432367,
+      "grad_norm": 4.433946132659912,
+      "learning_rate": 4.96046046046046e-05,
+      "loss": 2.1795,
+      "step": 90
+    },
+    {
+      "epoch": 0.00011895378188059393,
+      "grad_norm": 5.009888648986816,
+      "learning_rate": 4.95995995995996e-05,
+      "loss": 1.3696,
+      "step": 91
+    },
+    {
+      "epoch": 0.0001202609662968642,
+      "grad_norm": 4.033069133758545,
+      "learning_rate": 4.9594594594594596e-05,
+      "loss": 2.3803,
+      "step": 92
+    },
+    {
+      "epoch": 0.00012156815071313446,
+      "grad_norm": 4.739905834197998,
+      "learning_rate": 4.958958958958959e-05,
+      "loss": 1.7027,
+      "step": 93
+    },
+    {
+      "epoch": 0.00012287533512940471,
+      "grad_norm": 5.726681232452393,
+      "learning_rate": 4.958458458458459e-05,
+      "loss": 1.0941,
+      "step": 94
+    },
+    {
+      "epoch": 0.00012418251954567498,
+      "grad_norm": 4.398964881896973,
+      "learning_rate": 4.957957957957958e-05,
+      "loss": 1.8081,
+      "step": 95
+    },
+    {
+      "epoch": 0.00012548970396194525,
+      "grad_norm": 3.442424774169922,
+      "learning_rate": 4.957457457457458e-05,
+      "loss": 2.0943,
+      "step": 96
+    },
+    {
+      "epoch": 0.0001267968883782155,
+      "grad_norm": 3.42536997795105,
+      "learning_rate": 4.9569569569569574e-05,
+      "loss": 2.3138,
+      "step": 97
+    },
+    {
+      "epoch": 0.00012810407279448578,
+      "grad_norm": 3.0850987434387207,
+      "learning_rate": 4.956456456456457e-05,
+      "loss": 2.1422,
+      "step": 98
+    },
+    {
+      "epoch": 0.00012941125721075604,
+      "grad_norm": 3.327355146408081,
+      "learning_rate": 4.955955955955956e-05,
+      "loss": 1.607,
+      "step": 99
+    },
+    {
+      "epoch": 0.0001307184416270263,
+      "grad_norm": 3.0526223182678223,
+      "learning_rate": 4.955455455455455e-05,
+      "loss": 1.5995,
+      "step": 100
+    },
+    {
+      "epoch": 0.00013202562604329657,
+      "grad_norm": 3.1441359519958496,
+      "learning_rate": 4.954954954954955e-05,
+      "loss": 2.2129,
+      "step": 101
+    },
+    {
+      "epoch": 0.00013333281045956684,
+      "grad_norm": 2.209366798400879,
+      "learning_rate": 4.9544544544544546e-05,
+      "loss": 2.2904,
+      "step": 102
+    },
+    {
+      "epoch": 0.00013463999487583708,
+      "grad_norm": 2.4867329597473145,
+      "learning_rate": 4.953953953953954e-05,
+      "loss": 0.7869,
+      "step": 103
+    },
+    {
+      "epoch": 0.00013594717929210734,
+      "grad_norm": 2.2363412380218506,
+      "learning_rate": 4.953453453453454e-05,
+      "loss": 1.4709,
+      "step": 104
+    },
+    {
+      "epoch": 0.0001372543637083776,
+      "grad_norm": 2.6723251342773438,
+      "learning_rate": 4.952952952952953e-05,
+      "loss": 2.2871,
+      "step": 105
+    },
+    {
+      "epoch": 0.00013856154812464787,
+      "grad_norm": 1.9363102912902832,
+      "learning_rate": 4.952452452452453e-05,
+      "loss": 1.985,
+      "step": 106
+    },
+    {
+      "epoch": 0.00013986873254091814,
+      "grad_norm": 1.7824273109436035,
+      "learning_rate": 4.951951951951952e-05,
+      "loss": 1.189,
+      "step": 107
+    },
+    {
+      "epoch": 0.0001411759169571884,
+      "grad_norm": 1.8107385635375977,
+      "learning_rate": 4.951451451451452e-05,
+      "loss": 1.8943,
+      "step": 108
+    },
+    {
+      "epoch": 0.00014248310137345867,
+      "grad_norm": 2.283268451690674,
+      "learning_rate": 4.950950950950951e-05,
+      "loss": 1.7242,
+      "step": 109
+    },
+    {
+      "epoch": 0.00014379028578972894,
+      "grad_norm": 1.7572216987609863,
+      "learning_rate": 4.95045045045045e-05,
+      "loss": 1.8113,
+      "step": 110
+    },
+    {
+      "epoch": 0.0001450974702059992,
+      "grad_norm": 2.1798532009124756,
+      "learning_rate": 4.94994994994995e-05,
+      "loss": 1.9223,
+      "step": 111
+    },
+    {
+      "epoch": 0.00014640465462226947,
+      "grad_norm": 1.4888501167297363,
+      "learning_rate": 4.9494494494494496e-05,
+      "loss": 1.9747,
+      "step": 112
+    },
+    {
+      "epoch": 0.0001477118390385397,
+      "grad_norm": 1.8382623195648193,
+      "learning_rate": 4.9489489489489496e-05,
+      "loss": 2.199,
+      "step": 113
+    },
+    {
+      "epoch": 0.00014901902345480997,
+      "grad_norm": 2.766442060470581,
+      "learning_rate": 4.948448448448449e-05,
+      "loss": 1.433,
+      "step": 114
+    },
+    {
+      "epoch": 0.00015032620787108024,
+      "grad_norm": 1.3955047130584717,
+      "learning_rate": 4.947947947947948e-05,
+      "loss": 0.8611,
+      "step": 115
+    },
+    {
+      "epoch": 0.0001516333922873505,
+      "grad_norm": 1.3797332048416138,
+      "learning_rate": 4.9474474474474475e-05,
+      "loss": 1.8795,
+      "step": 116
+    },
+    {
+      "epoch": 0.00015294057670362077,
+      "grad_norm": 2.196755886077881,
+      "learning_rate": 4.946946946946947e-05,
+      "loss": 1.4086,
+      "step": 117
+    },
+    {
+      "epoch": 0.00015424776111989103,
+      "grad_norm": 1.6287676095962524,
+      "learning_rate": 4.946446446446447e-05,
+      "loss": 1.0982,
+      "step": 118
+    },
+    {
+      "epoch": 0.0001555549455361613,
+      "grad_norm": 2.1588571071624756,
+      "learning_rate": 4.945945945945946e-05,
+      "loss": 2.2011,
+      "step": 119
+    },
+    {
+      "epoch": 0.00015686212995243156,
+      "grad_norm": 1.66507887840271,
+      "learning_rate": 4.945445445445446e-05,
+      "loss": 1.6342,
+      "step": 120
+    },
+    {
+      "epoch": 0.00015816931436870183,
+      "grad_norm": 2.086681365966797,
+      "learning_rate": 4.944944944944945e-05,
+      "loss": 2.4012,
+      "step": 121
+    },
+    {
+      "epoch": 0.0001594764987849721,
+      "grad_norm": 2.2638509273529053,
+      "learning_rate": 4.9444444444444446e-05,
+      "loss": 0.2397,
+      "step": 122
+    },
+    {
+      "epoch": 0.00016078368320124236,
+      "grad_norm": 1.6955903768539429,
+      "learning_rate": 4.9439439439439446e-05,
+      "loss": 2.0733,
+      "step": 123
+    },
+    {
+      "epoch": 0.0001620908676175126,
+      "grad_norm": 1.7107908725738525,
+      "learning_rate": 4.943443443443443e-05,
+      "loss": 1.6862,
+      "step": 124
+    },
+    {
+      "epoch": 0.00016339805203378286,
+      "grad_norm": 2.117300033569336,
+      "learning_rate": 4.942942942942943e-05,
+      "loss": 1.6494,
+      "step": 125
+    },
+    {
+      "epoch": 0.00016470523645005313,
+      "grad_norm": 1.9111508131027222,
+      "learning_rate": 4.9424424424424425e-05,
+      "loss": 0.932,
+      "step": 126
+    },
+    {
+      "epoch": 0.0001660124208663234,
+      "grad_norm": 1.934354305267334,
+      "learning_rate": 4.9419419419419425e-05,
+      "loss": 1.6937,
+      "step": 127
+    },
+    {
+      "epoch": 0.00016731960528259366,
+      "grad_norm": 2.0459141731262207,
+      "learning_rate": 4.941441441441442e-05,
+      "loss": 2.1459,
+      "step": 128
+    },
+    {
+      "epoch": 0.00016862678969886393,
+      "grad_norm": 1.9724762439727783,
+      "learning_rate": 4.940940940940941e-05,
+      "loss": 2.2388,
+      "step": 129
+    },
+    {
+      "epoch": 0.0001699339741151342,
+      "grad_norm": 1.900592565536499,
+      "learning_rate": 4.940440440440441e-05,
+      "loss": 2.3512,
+      "step": 130
+    },
+    {
+      "epoch": 0.00017124115853140446,
+      "grad_norm": 1.978319764137268,
+      "learning_rate": 4.93993993993994e-05,
+      "loss": 1.7366,
+      "step": 131
+    },
+    {
+      "epoch": 0.00017254834294767472,
+      "grad_norm": 1.5513819456100464,
+      "learning_rate": 4.9394394394394396e-05,
+      "loss": 1.0159,
+      "step": 132
+    },
+    {
+      "epoch": 0.000173855527363945,
+      "grad_norm": 2.037893772125244,
+      "learning_rate": 4.938938938938939e-05,
+      "loss": 1.4329,
+      "step": 133
+    },
+    {
+      "epoch": 0.00017516271178021525,
+      "grad_norm": 1.7013237476348877,
+      "learning_rate": 4.938438438438439e-05,
+      "loss": 0.7663,
+      "step": 134
+    },
+    {
+      "epoch": 0.0001764698961964855,
+      "grad_norm": 1.8744912147521973,
+      "learning_rate": 4.937937937937938e-05,
+      "loss": 1.9305,
+      "step": 135
+    },
+    {
+      "epoch": 0.00017777708061275576,
+      "grad_norm": 1.606590747833252,
+      "learning_rate": 4.9374374374374375e-05,
+      "loss": 2.1672,
+      "step": 136
+    },
+    {
+      "epoch": 0.00017908426502902602,
+      "grad_norm": 2.2351033687591553,
+      "learning_rate": 4.9369369369369375e-05,
+      "loss": 1.0281,
+      "step": 137
+    },
+    {
+      "epoch": 0.0001803914494452963,
+      "grad_norm": 2.2086477279663086,
+      "learning_rate": 4.936436436436437e-05,
+      "loss": 1.6579,
+      "step": 138
+    },
+    {
+      "epoch": 0.00018169863386156655,
+      "grad_norm": 1.6347638368606567,
+      "learning_rate": 4.935935935935936e-05,
+      "loss": 1.7343,
+      "step": 139
+    },
+    {
+      "epoch": 0.00018300581827783682,
+      "grad_norm": 1.6986087560653687,
+      "learning_rate": 4.9354354354354354e-05,
+      "loss": 2.2753,
+      "step": 140
+    },
+    {
+      "epoch": 0.00018431300269410709,
+      "grad_norm": 1.8067278861999512,
+      "learning_rate": 4.9349349349349347e-05,
+      "loss": 0.8229,
+      "step": 141
+    },
+    {
+      "epoch": 0.00018562018711037735,
+      "grad_norm": 1.9648691415786743,
+      "learning_rate": 4.9344344344344346e-05,
+      "loss": 1.8146,
+      "step": 142
+    },
+    {
+      "epoch": 0.00018692737152664762,
+      "grad_norm": 1.3609802722930908,
+      "learning_rate": 4.933933933933934e-05,
+      "loss": 1.4422,
+      "step": 143
+    },
+    {
+      "epoch": 0.00018823455594291788,
+      "grad_norm": 2.002868890762329,
+      "learning_rate": 4.933433433433434e-05,
+      "loss": 0.0777,
+      "step": 144
+    },
+    {
+      "epoch": 0.00018954174035918812,
+      "grad_norm": 1.5004322528839111,
+      "learning_rate": 4.932932932932933e-05,
+      "loss": 1.5963,
+      "step": 145
+    },
+    {
+      "epoch": 0.00019084892477545839,
+      "grad_norm": 1.5936583280563354,
+      "learning_rate": 4.9324324324324325e-05,
+      "loss": 0.765,
+      "step": 146
+    },
+    {
+      "epoch": 0.00019215610919172865,
+      "grad_norm": 1.2652479410171509,
+      "learning_rate": 4.9319319319319325e-05,
+      "loss": 0.9597,
+      "step": 147
+    },
+    {
+      "epoch": 0.00019346329360799892,
+      "grad_norm": 1.7863197326660156,
+      "learning_rate": 4.931431431431432e-05,
+      "loss": 0.9391,
+      "step": 148
+    },
+    {
+      "epoch": 0.00019477047802426918,
+      "grad_norm": 1.9188827276229858,
+      "learning_rate": 4.930930930930931e-05,
+      "loss": 2.2355,
+      "step": 149
+    },
+    {
+      "epoch": 0.00019607766244053945,
+      "grad_norm": 1.5928704738616943,
+      "learning_rate": 4.9304304304304304e-05,
+      "loss": 1.1829,
+      "step": 150
+    },
+    {
+      "epoch": 0.0001973848468568097,
+      "grad_norm": 2.145599603652954,
+      "learning_rate": 4.92992992992993e-05,
+      "loss": 1.4609,
+      "step": 151
+    },
+    {
+      "epoch": 0.00019869203127307998,
+      "grad_norm": 1.4171509742736816,
+      "learning_rate": 4.9294294294294296e-05,
+      "loss": 1.9975,
+      "step": 152
+    },
+    {
+      "epoch": 0.00019999921568935024,
+      "grad_norm": 1.3210326433181763,
+      "learning_rate": 4.928928928928929e-05,
+      "loss": 1.0295,
+      "step": 153
+    },
+    {
+      "epoch": 0.0002013064001056205,
+      "grad_norm": 2.013183832168579,
+      "learning_rate": 4.928428428428429e-05,
+      "loss": 1.5366,
+      "step": 154
+    },
+    {
+      "epoch": 0.00020261358452189078,
+      "grad_norm": 1.594954490661621,
+      "learning_rate": 4.927927927927928e-05,
+      "loss": 1.0073,
+      "step": 155
+    },
+    {
+      "epoch": 0.00020392076893816101,
+      "grad_norm": 2.206082344055176,
+      "learning_rate": 4.927427427427428e-05,
+      "loss": 0.9262,
+      "step": 156
+    },
+    {
+      "epoch": 0.00020522795335443128,
+      "grad_norm": 2.40515398979187,
+      "learning_rate": 4.926926926926927e-05,
+      "loss": 1.4697,
+      "step": 157
+    },
+    {
+      "epoch": 0.00020653513777070155,
+      "grad_norm": 1.8254011869430542,
+      "learning_rate": 4.926426426426427e-05,
+      "loss": 1.8146,
+      "step": 158
+    },
+    {
+      "epoch": 0.0002078423221869718,
+      "grad_norm": 2.0213472843170166,
+      "learning_rate": 4.925925925925926e-05,
+      "loss": 3.1616,
+      "step": 159
+    },
+    {
+      "epoch": 0.00020914950660324208,
+      "grad_norm": 1.8707919120788574,
+      "learning_rate": 4.9254254254254254e-05,
+      "loss": 1.6477,
+      "step": 160
+    },
+    {
+      "epoch": 0.00021045669101951234,
+      "grad_norm": 1.9849202632904053,
+      "learning_rate": 4.9249249249249253e-05,
+      "loss": 1.5499,
+      "step": 161
+    },
+    {
+      "epoch": 0.0002117638754357826,
+      "grad_norm": 1.8460415601730347,
+      "learning_rate": 4.9244244244244246e-05,
+      "loss": 1.4391,
+      "step": 162
+    },
+    {
+      "epoch": 0.00021307105985205287,
+      "grad_norm": 1.8488810062408447,
+      "learning_rate": 4.9239239239239246e-05,
+      "loss": 1.8629,
+      "step": 163
+    },
+    {
+      "epoch": 0.00021437824426832314,
+      "grad_norm": 1.8770229816436768,
+      "learning_rate": 4.923423423423424e-05,
+      "loss": 2.3753,
+      "step": 164
+    },
+    {
+      "epoch": 0.0002156854286845934,
+      "grad_norm": 1.472601056098938,
+      "learning_rate": 4.922922922922923e-05,
+      "loss": 0.7564,
+      "step": 165
+    },
+    {
+      "epoch": 0.00021699261310086367,
+      "grad_norm": 2.860130548477173,
+      "learning_rate": 4.9224224224224225e-05,
+      "loss": 0.7373,
+      "step": 166
+    },
+    {
+      "epoch": 0.0002182997975171339,
+      "grad_norm": 1.886745572090149,
+      "learning_rate": 4.921921921921922e-05,
+      "loss": 1.0192,
+      "step": 167
+    },
+    {
+      "epoch": 0.00021960698193340417,
+      "grad_norm": 1.2749582529067993,
+      "learning_rate": 4.921421421421422e-05,
+      "loss": 0.4704,
+      "step": 168
+    },
+    {
+      "epoch": 0.00022091416634967444,
+      "grad_norm": 1.645389199256897,
+      "learning_rate": 4.920920920920921e-05,
+      "loss": 1.8075,
+      "step": 169
+    },
+    {
+      "epoch": 0.0002222213507659447,
+      "grad_norm": 1.7299386262893677,
+      "learning_rate": 4.920420420420421e-05,
+      "loss": 2.4038,
+      "step": 170
+    },
+    {
+      "epoch": 0.00022352853518221497,
+      "grad_norm": 1.6160026788711548,
+      "learning_rate": 4.9199199199199204e-05,
+      "loss": 1.1931,
+      "step": 171
+    },
+    {
+      "epoch": 0.00022483571959848524,
+      "grad_norm": 1.9258168935775757,
+      "learning_rate": 4.9194194194194196e-05,
+      "loss": 2.2323,
+      "step": 172
+    },
+    {
+      "epoch": 0.0002261429040147555,
+      "grad_norm": 2.079798936843872,
+      "learning_rate": 4.9189189189189196e-05,
+      "loss": 1.9468,
+      "step": 173
+    },
+    {
+      "epoch": 0.00022745008843102577,
+      "grad_norm": 2.0454630851745605,
+      "learning_rate": 4.918418418418418e-05,
+      "loss": 1.7755,
+      "step": 174
+    },
+    {
+      "epoch": 0.00022875727284729603,
+      "grad_norm": 1.9173661470413208,
+      "learning_rate": 4.917917917917918e-05,
+      "loss": 1.0856,
+      "step": 175
+    },
+    {
+      "epoch": 0.0002300644572635663,
+      "grad_norm": 1.6535844802856445,
+      "learning_rate": 4.9174174174174175e-05,
+      "loss": 1.6214,
+      "step": 176
+    },
+    {
+      "epoch": 0.00023137164167983656,
+      "grad_norm": 2.402019739151001,
+      "learning_rate": 4.916916916916917e-05,
+      "loss": 1.7331,
+      "step": 177
+    },
+    {
+      "epoch": 0.0002326788260961068,
+      "grad_norm": 1.7244210243225098,
+      "learning_rate": 4.916416416416417e-05,
+      "loss": 1.4885,
+      "step": 178
+    },
+    {
+      "epoch": 0.00023398601051237707,
+      "grad_norm": 2.702523946762085,
+      "learning_rate": 4.915915915915916e-05,
+      "loss": 2.0403,
+      "step": 179
+    },
+    {
+      "epoch": 0.00023529319492864733,
+      "grad_norm": 2.1688618659973145,
+      "learning_rate": 4.915415415415416e-05,
+      "loss": 0.986,
+      "step": 180
+    },
+    {
+      "epoch": 0.0002366003793449176,
+      "grad_norm": 2.3919677734375,
+      "learning_rate": 4.9149149149149154e-05,
+      "loss": 1.3291,
+      "step": 181
+    },
+    {
+      "epoch": 0.00023790756376118786,
+      "grad_norm": 1.8715122938156128,
+      "learning_rate": 4.9144144144144147e-05,
+      "loss": 1.7018,
+      "step": 182
+    },
+    {
+      "epoch": 0.00023921474817745813,
+      "grad_norm": 2.152589797973633,
+      "learning_rate": 4.913913913913914e-05,
+      "loss": 1.8486,
+      "step": 183
+    },
+    {
+      "epoch": 0.0002405219325937284,
+      "grad_norm": 1.5322504043579102,
+      "learning_rate": 4.913413413413413e-05,
+      "loss": 1.6896,
+      "step": 184
+    },
+    {
+      "epoch": 0.00024182911700999866,
+      "grad_norm": 1.6208884716033936,
+      "learning_rate": 4.912912912912913e-05,
+      "loss": 1.9508,
+      "step": 185
+    },
+    {
+      "epoch": 0.00024313630142626893,
+      "grad_norm": 1.5740910768508911,
+      "learning_rate": 4.9124124124124125e-05,
+      "loss": 1.1913,
+      "step": 186
+    },
+    {
+      "epoch": 0.00024444348584253916,
+      "grad_norm": 2.652970790863037,
+      "learning_rate": 4.9119119119119125e-05,
+      "loss": 1.5786,
+      "step": 187
+    },
+    {
+      "epoch": 0.00024575067025880943,
+      "grad_norm": 1.5115643739700317,
+      "learning_rate": 4.911411411411412e-05,
+      "loss": 2.3382,
+      "step": 188
+    },
+    {
+      "epoch": 0.0002470578546750797,
+      "grad_norm": 2.009474754333496,
+      "learning_rate": 4.910910910910911e-05,
+      "loss": 2.2713,
+      "step": 189
+    },
+    {
+      "epoch": 0.00024836503909134996,
+      "grad_norm": 3.25408673286438,
+      "learning_rate": 4.9104104104104104e-05,
+      "loss": 1.767,
+      "step": 190
+    },
+    {
+      "epoch": 0.0002496722235076202,
+      "grad_norm": 1.6666245460510254,
+      "learning_rate": 4.90990990990991e-05,
+      "loss": 1.1057,
+      "step": 191
+    },
+    {
+      "epoch": 0.0002509794079238905,
+      "grad_norm": 2.5899922847747803,
+      "learning_rate": 4.90940940940941e-05,
+      "loss": 1.7851,
+      "step": 192
+    },
+    {
+      "epoch": 0.00025228659234016076,
+      "grad_norm": 1.9656989574432373,
+      "learning_rate": 4.908908908908909e-05,
+      "loss": 2.5767,
+      "step": 193
+    },
+    {
+      "epoch": 0.000253593776756431,
+      "grad_norm": 1.7756541967391968,
+      "learning_rate": 4.908408408408409e-05,
+      "loss": 1.7306,
+      "step": 194
+    },
+    {
+      "epoch": 0.0002549009611727013,
+      "grad_norm": 1.9504581689834595,
+      "learning_rate": 4.907907907907908e-05,
+      "loss": 1.5729,
+      "step": 195
+    },
+    {
+      "epoch": 0.00025620814558897155,
+      "grad_norm": 0.957115888595581,
+      "learning_rate": 4.9074074074074075e-05,
+      "loss": 0.277,
+      "step": 196
+    },
+    {
+      "epoch": 0.0002575153300052418,
+      "grad_norm": 1.4767944812774658,
+      "learning_rate": 4.9069069069069075e-05,
+      "loss": 1.5084,
+      "step": 197
+    },
+    {
+      "epoch": 0.0002588225144215121,
+      "grad_norm": 1.2738802433013916,
+      "learning_rate": 4.906406406406407e-05,
+      "loss": 0.9465,
+      "step": 198
+    },
+    {
+      "epoch": 0.00026012969883778235,
+      "grad_norm": 1.6053591966629028,
+      "learning_rate": 4.905905905905906e-05,
+      "loss": 1.3578,
+      "step": 199
+    },
+    {
+      "epoch": 0.0002614368832540526,
+      "grad_norm": 1.664510726928711,
+      "learning_rate": 4.9054054054054054e-05,
+      "loss": 2.4543,
+      "step": 200
+    },
+    {
+      "epoch": 0.0002627440676703229,
+      "grad_norm": 1.6944023370742798,
+      "learning_rate": 4.9049049049049054e-05,
+      "loss": 1.3043,
+      "step": 201
+    },
+    {
+      "epoch": 0.00026405125208659315,
+      "grad_norm": 1.387467861175537,
+      "learning_rate": 4.904404404404405e-05,
+      "loss": 1.4816,
+      "step": 202
+    },
+    {
+      "epoch": 0.0002653584365028634,
+      "grad_norm": 1.925346851348877,
+      "learning_rate": 4.903903903903904e-05,
+      "loss": 1.6951,
+      "step": 203
+    },
+    {
+      "epoch": 0.0002666656209191337,
+      "grad_norm": 1.4878056049346924,
+      "learning_rate": 4.903403403403404e-05,
+      "loss": 1.7805,
+      "step": 204
+    },
+    {
+      "epoch": 0.0002679728053354039,
+      "grad_norm": 2.018648862838745,
+      "learning_rate": 4.902902902902903e-05,
+      "loss": 1.4652,
+      "step": 205
+    },
+    {
+      "epoch": 0.00026927998975167415,
+      "grad_norm": 1.8136177062988281,
+      "learning_rate": 4.902402402402403e-05,
+      "loss": 1.5613,
+      "step": 206
+    },
+    {
+      "epoch": 0.0002705871741679444,
+      "grad_norm": 2.323335886001587,
+      "learning_rate": 4.901901901901902e-05,
+      "loss": 1.6988,
+      "step": 207
+    },
+    {
+      "epoch": 0.0002718943585842147,
+      "grad_norm": 1.7906171083450317,
+      "learning_rate": 4.901401401401402e-05,
+      "loss": 1.5934,
+      "step": 208
+    },
+    {
+      "epoch": 0.00027320154300048495,
+      "grad_norm": 1.5515409708023071,
+      "learning_rate": 4.900900900900901e-05,
+      "loss": 1.5874,
+      "step": 209
+    },
+    {
+      "epoch": 0.0002745087274167552,
+      "grad_norm": 1.9337890148162842,
+      "learning_rate": 4.9004004004004004e-05,
+      "loss": 1.7147,
+      "step": 210
+    },
+    {
+      "epoch": 0.0002758159118330255,
+      "grad_norm": 2.102532148361206,
+      "learning_rate": 4.8998998998999004e-05,
+      "loss": 1.7885,
+      "step": 211
+    },
+    {
+      "epoch": 0.00027712309624929575,
+      "grad_norm": 1.6877130270004272,
+      "learning_rate": 4.8993993993994e-05,
+      "loss": 1.5048,
+      "step": 212
+    },
+    {
+      "epoch": 0.000278430280665566,
+      "grad_norm": 1.8921681642532349,
+      "learning_rate": 4.898898898898899e-05,
+      "loss": 1.8371,
+      "step": 213
+    },
+    {
+      "epoch": 0.0002797374650818363,
+      "grad_norm": 1.810315728187561,
+      "learning_rate": 4.898398398398399e-05,
+      "loss": 1.7009,
+      "step": 214
+    },
+    {
+      "epoch": 0.00028104464949810654,
+      "grad_norm": 2.759514331817627,
+      "learning_rate": 4.8978978978978976e-05,
+      "loss": 3.1437,
+      "step": 215
+    },
+    {
+      "epoch": 0.0002823518339143768,
+      "grad_norm": 2.4991655349731445,
+      "learning_rate": 4.8973973973973975e-05,
+      "loss": 2.3588,
+      "step": 216
+    },
+    {
+      "epoch": 0.0002836590183306471,
+      "grad_norm": 2.547462224960327,
+      "learning_rate": 4.896896896896897e-05,
+      "loss": 1.5821,
+      "step": 217
+    },
+    {
+      "epoch": 0.00028496620274691734,
+      "grad_norm": 1.3106446266174316,
+      "learning_rate": 4.896396396396397e-05,
+      "loss": 1.4013,
+      "step": 218
+    },
+    {
+      "epoch": 0.0002862733871631876,
+      "grad_norm": 1.6845613718032837,
+      "learning_rate": 4.895895895895896e-05,
+      "loss": 1.5136,
+      "step": 219
+    },
+    {
+      "epoch": 0.00028758057157945787,
+      "grad_norm": 1.659991979598999,
+      "learning_rate": 4.8953953953953954e-05,
+      "loss": 2.2554,
+      "step": 220
+    },
+    {
+      "epoch": 0.00028888775599572814,
+      "grad_norm": 1.3590683937072754,
+      "learning_rate": 4.8948948948948954e-05,
+      "loss": 0.7868,
+      "step": 221
+    },
+    {
+      "epoch": 0.0002901949404119984,
+      "grad_norm": 1.8435338735580444,
+      "learning_rate": 4.894394394394395e-05,
+      "loss": 1.8352,
+      "step": 222
+    },
+    {
+      "epoch": 0.00029150212482826867,
+      "grad_norm": 1.6166387796401978,
+      "learning_rate": 4.893893893893894e-05,
+      "loss": 1.3868,
+      "step": 223
+    },
+    {
+      "epoch": 0.00029280930924453893,
+      "grad_norm": 2.057384967803955,
+      "learning_rate": 4.893393393393393e-05,
+      "loss": 1.898,
+      "step": 224
+    },
+    {
+      "epoch": 0.0002941164936608092,
+      "grad_norm": 1.695014238357544,
+      "learning_rate": 4.892892892892893e-05,
+      "loss": 2.2495,
+      "step": 225
+    },
+    {
+      "epoch": 0.0002954236780770794,
+      "grad_norm": 2.4186930656433105,
+      "learning_rate": 4.8923923923923926e-05,
+      "loss": 2.1074,
+      "step": 226
+    },
+    {
+      "epoch": 0.0002967308624933497,
+      "grad_norm": 1.5932652950286865,
+      "learning_rate": 4.891891891891892e-05,
+      "loss": 1.1245,
+      "step": 227
+    },
+    {
+      "epoch": 0.00029803804690961994,
+      "grad_norm": 1.7992708683013916,
+      "learning_rate": 4.891391391391392e-05,
+      "loss": 1.6981,
+      "step": 228
+    },
+    {
+      "epoch": 0.0002993452313258902,
+      "grad_norm": 1.6720223426818848,
+      "learning_rate": 4.890890890890891e-05,
+      "loss": 1.499,
+      "step": 229
+    },
+    {
+      "epoch": 0.00030065241574216047,
+      "grad_norm": 1.7174367904663086,
+      "learning_rate": 4.890390390390391e-05,
+      "loss": 1.4546,
+      "step": 230
+    },
+    {
+      "epoch": 0.00030195960015843074,
+      "grad_norm": 2.5477240085601807,
+      "learning_rate": 4.8898898898898904e-05,
+      "loss": 1.0959,
+      "step": 231
+    },
+    {
+      "epoch": 0.000303266784574701,
+      "grad_norm": 1.551347017288208,
+      "learning_rate": 4.88938938938939e-05,
+      "loss": 1.2192,
+      "step": 232
+    },
+    {
+      "epoch": 0.00030457396899097127,
+      "grad_norm": 4.913365364074707,
+      "learning_rate": 4.888888888888889e-05,
+      "loss": 1.4216,
+      "step": 233
+    },
+    {
+      "epoch": 0.00030588115340724153,
+      "grad_norm": 2.8688876628875732,
+      "learning_rate": 4.888388388388388e-05,
+      "loss": 1.2954,
+      "step": 234
+    },
+    {
+      "epoch": 0.0003071883378235118,
+      "grad_norm": 1.8472936153411865,
+      "learning_rate": 4.887887887887888e-05,
+      "loss": 1.6743,
+      "step": 235
+    },
+    {
+      "epoch": 0.00030849552223978207,
+      "grad_norm": 2.6009321212768555,
+      "learning_rate": 4.8873873873873876e-05,
+      "loss": 0.8994,
+      "step": 236
+    },
+    {
+      "epoch": 0.00030980270665605233,
+      "grad_norm": 3.2686116695404053,
+      "learning_rate": 4.8868868868868875e-05,
+      "loss": 1.89,
+      "step": 237
+    },
+    {
+      "epoch": 0.0003111098910723226,
+      "grad_norm": 2.842278003692627,
+      "learning_rate": 4.886386386386387e-05,
+      "loss": 1.3352,
+      "step": 238
+    },
+    {
+      "epoch": 0.00031241707548859286,
+      "grad_norm": 1.6975959539413452,
+      "learning_rate": 4.885885885885886e-05,
+      "loss": 2.5637,
+      "step": 239
+    },
+    {
+      "epoch": 0.00031372425990486313,
+      "grad_norm": 1.6205472946166992,
+      "learning_rate": 4.8853853853853854e-05,
+      "loss": 1.7702,
+      "step": 240
+    },
+    {
+      "epoch": 0.0003150314443211334,
+      "grad_norm": 1.9137065410614014,
+      "learning_rate": 4.884884884884885e-05,
+      "loss": 2.128,
+      "step": 241
+    },
+    {
+      "epoch": 0.00031633862873740366,
+      "grad_norm": 1.6640243530273438,
+      "learning_rate": 4.884384384384385e-05,
+      "loss": 2.798,
+      "step": 242
+    },
+    {
+      "epoch": 0.0003176458131536739,
+      "grad_norm": 1.5545439720153809,
+      "learning_rate": 4.883883883883884e-05,
+      "loss": 1.9383,
+      "step": 243
+    },
+    {
+      "epoch": 0.0003189529975699442,
+      "grad_norm": 2.110470771789551,
+      "learning_rate": 4.883383383383384e-05,
+      "loss": 2.7099,
+      "step": 244
+    },
+    {
+      "epoch": 0.00032026018198621446,
+      "grad_norm": 1.3277205228805542,
+      "learning_rate": 4.882882882882883e-05,
+      "loss": 1.6492,
+      "step": 245
+    },
+    {
+      "epoch": 0.0003215673664024847,
+      "grad_norm": 1.317232608795166,
+      "learning_rate": 4.8823823823823826e-05,
+      "loss": 1.3825,
+      "step": 246
+    },
+    {
+      "epoch": 0.00032287455081875493,
+      "grad_norm": 1.5808682441711426,
+      "learning_rate": 4.8818818818818825e-05,
+      "loss": 1.7761,
+      "step": 247
+    },
+    {
+      "epoch": 0.0003241817352350252,
+      "grad_norm": 1.576471209526062,
+      "learning_rate": 4.881381381381381e-05,
+      "loss": 1.6476,
+      "step": 248
+    },
+    {
+      "epoch": 0.00032548891965129546,
+      "grad_norm": 1.5742518901824951,
+      "learning_rate": 4.880880880880881e-05,
+      "loss": 1.3871,
+      "step": 249
+    },
+    {
+      "epoch": 0.00032679610406756573,
+      "grad_norm": 1.22970449924469,
+      "learning_rate": 4.8803803803803804e-05,
+      "loss": 0.9665,
+      "step": 250
+    },
+    {
+      "epoch": 0.000328103288483836,
+      "grad_norm": 1.9720549583435059,
+      "learning_rate": 4.87987987987988e-05,
+      "loss": 1.5671,
+      "step": 251
+    },
+    {
+      "epoch": 0.00032941047290010626,
+      "grad_norm": 1.6531200408935547,
+      "learning_rate": 4.87937937937938e-05,
+      "loss": 2.0848,
+      "step": 252
+    },
+    {
+      "epoch": 0.0003307176573163765,
+      "grad_norm": 0.06442175805568695,
+      "learning_rate": 4.878878878878879e-05,
+      "loss": 0.0011,
+      "step": 253
+    },
+    {
+      "epoch": 0.0003320248417326468,
+      "grad_norm": 1.4050544500350952,
+      "learning_rate": 4.878378378378379e-05,
+      "loss": 0.779,
+      "step": 254
+    },
+    {
+      "epoch": 0.00033333202614891706,
+      "grad_norm": 1.4547470808029175,
+      "learning_rate": 4.877877877877878e-05,
+      "loss": 1.2172,
+      "step": 255
+    },
+    {
+      "epoch": 0.0003346392105651873,
+      "grad_norm": 1.7619500160217285,
+      "learning_rate": 4.8773773773773776e-05,
+      "loss": 2.2073,
+      "step": 256
+    },
+    {
+      "epoch": 0.0003359463949814576,
+      "grad_norm": 1.3657172918319702,
+      "learning_rate": 4.876876876876877e-05,
+      "loss": 1.0345,
+      "step": 257
+    },
+    {
+      "epoch": 0.00033725357939772785,
+      "grad_norm": 1.9810845851898193,
+      "learning_rate": 4.876376376376376e-05,
+      "loss": 2.0455,
+      "step": 258
+    },
+    {
+      "epoch": 0.0003385607638139981,
+      "grad_norm": 1.3832188844680786,
+      "learning_rate": 4.875875875875876e-05,
+      "loss": 1.9951,
+      "step": 259
+    },
+    {
+      "epoch": 0.0003398679482302684,
+      "grad_norm": 1.5447360277175903,
+      "learning_rate": 4.8753753753753754e-05,
+      "loss": 2.2018,
+      "step": 260
+    },
+    {
+      "epoch": 0.00034117513264653865,
+      "grad_norm": 1.6113301515579224,
+      "learning_rate": 4.8748748748748754e-05,
+      "loss": 1.9734,
+      "step": 261
+    },
+    {
+      "epoch": 0.0003424823170628089,
+      "grad_norm": 1.8055421113967896,
+      "learning_rate": 4.874374374374375e-05,
+      "loss": 2.1845,
+      "step": 262
+    },
+    {
+      "epoch": 0.0003437895014790792,
+      "grad_norm": 1.568477749824524,
+      "learning_rate": 4.873873873873874e-05,
+      "loss": 1.8695,
+      "step": 263
+    },
+    {
+      "epoch": 0.00034509668589534945,
+      "grad_norm": 1.6943508386611938,
+      "learning_rate": 4.873373373373374e-05,
+      "loss": 1.8148,
+      "step": 264
+    },
+    {
+      "epoch": 0.0003464038703116197,
+      "grad_norm": 1.0157723426818848,
+      "learning_rate": 4.8728728728728726e-05,
+      "loss": 0.6497,
+      "step": 265
+    },
+    {
+      "epoch": 0.00034771105472789,
+      "grad_norm": 1.506629228591919,
+      "learning_rate": 4.8723723723723726e-05,
+      "loss": 2.0373,
+      "step": 266
+    },
+    {
+      "epoch": 0.00034901823914416024,
+      "grad_norm": 1.9490463733673096,
+      "learning_rate": 4.871871871871872e-05,
+      "loss": 1.37,
+      "step": 267
+    },
+    {
+      "epoch": 0.0003503254235604305,
+      "grad_norm": 1.2232673168182373,
+      "learning_rate": 4.871371371371372e-05,
+      "loss": 0.8621,
+      "step": 268
+    },
+    {
+      "epoch": 0.0003516326079767007,
+      "grad_norm": 1.1460094451904297,
+      "learning_rate": 4.870870870870871e-05,
+      "loss": 0.6316,
+      "step": 269
+    },
+    {
+      "epoch": 0.000352939792392971,
+      "grad_norm": 1.7706618309020996,
+      "learning_rate": 4.8703703703703704e-05,
+      "loss": 1.8774,
+      "step": 270
+    },
+    {
+      "epoch": 0.00035424697680924125,
+      "grad_norm": 1.5327508449554443,
+      "learning_rate": 4.8698698698698704e-05,
+      "loss": 1.669,
+      "step": 271
+    },
+    {
+      "epoch": 0.0003555541612255115,
+      "grad_norm": 2.183750629425049,
+      "learning_rate": 4.86936936936937e-05,
+      "loss": 1.7982,
+      "step": 272
+    },
+    {
+      "epoch": 0.0003568613456417818,
+      "grad_norm": 1.3268592357635498,
+      "learning_rate": 4.868868868868869e-05,
+      "loss": 1.6348,
+      "step": 273
+    },
+    {
+      "epoch": 0.00035816853005805205,
+      "grad_norm": 1.3289145231246948,
+      "learning_rate": 4.868368368368368e-05,
+      "loss": 1.0636,
+      "step": 274
+    },
+    {
+      "epoch": 0.0003594757144743223,
+      "grad_norm": 1.543196201324463,
+      "learning_rate": 4.867867867867868e-05,
+      "loss": 1.6589,
+      "step": 275
+    },
+    {
+      "epoch": 0.0003607828988905926,
+      "grad_norm": 1.8786145448684692,
+      "learning_rate": 4.8673673673673676e-05,
+      "loss": 2.2636,
+      "step": 276
+    },
+    {
+      "epoch": 0.00036209008330686284,
+      "grad_norm": 1.9315189123153687,
+      "learning_rate": 4.866866866866867e-05,
+      "loss": 1.4786,
+      "step": 277
+    },
+    {
+      "epoch": 0.0003633972677231331,
+      "grad_norm": 3.0136470794677734,
+      "learning_rate": 4.866366366366367e-05,
+      "loss": 1.1689,
+      "step": 278
+    },
+    {
+      "epoch": 0.0003647044521394034,
+      "grad_norm": 1.5813599824905396,
+      "learning_rate": 4.865865865865866e-05,
+      "loss": 1.7059,
+      "step": 279
+    },
+    {
+      "epoch": 0.00036601163655567364,
+      "grad_norm": 1.8223422765731812,
+      "learning_rate": 4.865365365365366e-05,
+      "loss": 1.2049,
+      "step": 280
+    },
+    {
+      "epoch": 0.0003673188209719439,
+      "grad_norm": 1.1571829319000244,
+      "learning_rate": 4.8648648648648654e-05,
+      "loss": 0.9757,
+      "step": 281
+    },
+    {
+      "epoch": 0.00036862600538821417,
+      "grad_norm": 2.1226696968078613,
+      "learning_rate": 4.864364364364364e-05,
+      "loss": 2.0413,
+      "step": 282
+    },
+    {
+      "epoch": 0.00036993318980448444,
+      "grad_norm": 1.8020603656768799,
+      "learning_rate": 4.863863863863864e-05,
+      "loss": 1.5693,
+      "step": 283
+    },
+    {
+      "epoch": 0.0003712403742207547,
+      "grad_norm": 1.799559235572815,
+      "learning_rate": 4.863363363363363e-05,
+      "loss": 2.1869,
+      "step": 284
+    },
+    {
+      "epoch": 0.00037254755863702497,
+      "grad_norm": 2.161092519760132,
+      "learning_rate": 4.862862862862863e-05,
+      "loss": 1.903,
+      "step": 285
+    },
+    {
+      "epoch": 0.00037385474305329523,
+      "grad_norm": 1.8967934846878052,
+      "learning_rate": 4.8623623623623626e-05,
+      "loss": 1.9437,
+      "step": 286
+    },
+    {
+      "epoch": 0.0003751619274695655,
+      "grad_norm": 1.6356936693191528,
+      "learning_rate": 4.861861861861862e-05,
+      "loss": 1.8212,
+      "step": 287
+    },
+    {
+      "epoch": 0.00037646911188583576,
+      "grad_norm": 2.0186562538146973,
+      "learning_rate": 4.861361361361362e-05,
+      "loss": 1.9657,
+      "step": 288
+    },
+    {
+      "epoch": 0.00037777629630210603,
+      "grad_norm": 2.0139613151550293,
+      "learning_rate": 4.860860860860861e-05,
+      "loss": 1.3713,
+      "step": 289
+    },
+    {
+      "epoch": 0.00037908348071837624,
+      "grad_norm": 1.9778642654418945,
+      "learning_rate": 4.8603603603603605e-05,
+      "loss": 2.1371,
+      "step": 290
+    },
+    {
+      "epoch": 0.0003803906651346465,
+      "grad_norm": 2.013936996459961,
+      "learning_rate": 4.85985985985986e-05,
+      "loss": 2.1338,
+      "step": 291
+    },
+    {
+      "epoch": 0.00038169784955091677,
+      "grad_norm": 1.480167031288147,
+      "learning_rate": 4.85935935935936e-05,
+      "loss": 1.4518,
+      "step": 292
+    },
+    {
+      "epoch": 0.00038300503396718704,
+      "grad_norm": 1.0584605932235718,
+      "learning_rate": 4.858858858858859e-05,
+      "loss": 0.9891,
+      "step": 293
+    },
+    {
+      "epoch": 0.0003843122183834573,
+      "grad_norm": 1.5172793865203857,
+      "learning_rate": 4.858358358358358e-05,
+      "loss": 1.1713,
+      "step": 294
+    },
+    {
+      "epoch": 0.00038561940279972757,
+      "grad_norm": 1.6944522857666016,
+      "learning_rate": 4.857857857857858e-05,
+      "loss": 2.72,
+      "step": 295
+    },
+    {
+      "epoch": 0.00038692658721599783,
+      "grad_norm": 1.6618627309799194,
+      "learning_rate": 4.8573573573573576e-05,
+      "loss": 1.3664,
+      "step": 296
+    },
+    {
+      "epoch": 0.0003882337716322681,
+      "grad_norm": 1.5239827632904053,
+      "learning_rate": 4.8568568568568576e-05,
+      "loss": 1.8734,
+      "step": 297
+    },
+    {
+      "epoch": 0.00038954095604853837,
+      "grad_norm": 1.4096382856369019,
+      "learning_rate": 4.856356356356356e-05,
+      "loss": 1.4521,
+      "step": 298
+    },
+    {
+      "epoch": 0.00039084814046480863,
+      "grad_norm": 1.4097378253936768,
+      "learning_rate": 4.855855855855856e-05,
+      "loss": 1.4607,
+      "step": 299
+    },
+    {
+      "epoch": 0.0003921553248810789,
+      "grad_norm": 1.5461453199386597,
+      "learning_rate": 4.8553553553553555e-05,
+      "loss": 1.744,
+      "step": 300
+    },
+    {
+      "epoch": 0.00039346250929734916,
+      "grad_norm": 1.9085536003112793,
+      "learning_rate": 4.854854854854855e-05,
+      "loss": 2.6346,
+      "step": 301
+    },
+    {
+      "epoch": 0.0003947696937136194,
+      "grad_norm": 1.5974091291427612,
+      "learning_rate": 4.854354354354355e-05,
+      "loss": 2.779,
+      "step": 302
+    },
+    {
+      "epoch": 0.0003960768781298897,
+      "grad_norm": 1.4183374643325806,
+      "learning_rate": 4.853853853853854e-05,
+      "loss": 1.1477,
+      "step": 303
+    },
+    {
+      "epoch": 0.00039738406254615996,
+      "grad_norm": 1.7596776485443115,
+      "learning_rate": 4.853353353353354e-05,
+      "loss": 1.3722,
+      "step": 304
+    },
+    {
+      "epoch": 0.0003986912469624302,
+      "grad_norm": 2.0469233989715576,
+      "learning_rate": 4.852852852852853e-05,
+      "loss": 2.532,
+      "step": 305
+    },
+    {
+      "epoch": 0.0003999984313787005,
+      "grad_norm": 1.7058496475219727,
+      "learning_rate": 4.8523523523523526e-05,
+      "loss": 1.8808,
+      "step": 306
+    },
+    {
+      "epoch": 0.00040130561579497075,
+      "grad_norm": 2.928122043609619,
+      "learning_rate": 4.851851851851852e-05,
+      "loss": 1.4133,
+      "step": 307
+    },
+    {
+      "epoch": 0.000402612800211241,
+      "grad_norm": 2.3317997455596924,
+      "learning_rate": 4.851351351351351e-05,
+      "loss": 1.9976,
+      "step": 308
+    },
+    {
+      "epoch": 0.0004039199846275113,
+      "grad_norm": 1.520419716835022,
+      "learning_rate": 4.850850850850851e-05,
+      "loss": 1.7102,
+      "step": 309
+    },
+    {
+      "epoch": 0.00040522716904378155,
+      "grad_norm": 1.5943193435668945,
+      "learning_rate": 4.8503503503503505e-05,
+      "loss": 1.9234,
+      "step": 310
+    },
+    {
+      "epoch": 0.0004065343534600518,
+      "grad_norm": 0.8694736361503601,
+      "learning_rate": 4.8498498498498504e-05,
+      "loss": 0.4603,
+      "step": 311
+    },
+    {
+      "epoch": 0.00040784153787632203,
+      "grad_norm": 1.2508904933929443,
+      "learning_rate": 4.84934934934935e-05,
+      "loss": 0.7948,
+      "step": 312
+    },
+    {
+      "epoch": 0.0004091487222925923,
+      "grad_norm": 2.01047420501709,
+      "learning_rate": 4.848848848848849e-05,
+      "loss": 1.8304,
+      "step": 313
+    },
+    {
+      "epoch": 0.00041045590670886256,
+      "grad_norm": 2.1941375732421875,
+      "learning_rate": 4.848348348348349e-05,
+      "loss": 1.7869,
+      "step": 314
+    },
+    {
+      "epoch": 0.0004117630911251328,
+      "grad_norm": 1.6790601015090942,
+      "learning_rate": 4.8478478478478476e-05,
+      "loss": 1.2441,
+      "step": 315
+    },
+    {
+      "epoch": 0.0004130702755414031,
+      "grad_norm": 1.6825275421142578,
+      "learning_rate": 4.8473473473473476e-05,
+      "loss": 1.7431,
+      "step": 316
+    },
+    {
+      "epoch": 0.00041437745995767336,
+      "grad_norm": 1.328930377960205,
+      "learning_rate": 4.846846846846847e-05,
+      "loss": 1.3589,
+      "step": 317
+    },
+    {
+      "epoch": 0.0004156846443739436,
+      "grad_norm": 1.519547939300537,
+      "learning_rate": 4.846346346346346e-05,
+      "loss": 0.5922,
+      "step": 318
+    },
+    {
+      "epoch": 0.0004169918287902139,
+      "grad_norm": 1.8517762422561646,
+      "learning_rate": 4.845845845845846e-05,
+      "loss": 2.0995,
+      "step": 319
+    },
+    {
+      "epoch": 0.00041829901320648415,
+      "grad_norm": 1.561224341392517,
+      "learning_rate": 4.8453453453453455e-05,
+      "loss": 2.1158,
+      "step": 320
+    },
+    {
+      "epoch": 0.0004196061976227544,
+      "grad_norm": 1.9378067255020142,
+      "learning_rate": 4.8448448448448455e-05,
+      "loss": 2.2775,
+      "step": 321
+    },
+    {
+      "epoch": 0.0004209133820390247,
+      "grad_norm": 1.6161144971847534,
+      "learning_rate": 4.844344344344345e-05,
+      "loss": 2.2261,
+      "step": 322
+    },
+    {
+      "epoch": 0.00042222056645529495,
+      "grad_norm": 0.9984459280967712,
+      "learning_rate": 4.843843843843844e-05,
+      "loss": 0.8248,
+      "step": 323
+    },
+    {
+      "epoch": 0.0004235277508715652,
+      "grad_norm": 1.2390950918197632,
+      "learning_rate": 4.8433433433433433e-05,
+      "loss": 0.8056,
+      "step": 324
+    },
+    {
+      "epoch": 0.0004248349352878355,
+      "grad_norm": 1.8470146656036377,
+      "learning_rate": 4.8428428428428426e-05,
+      "loss": 1.8804,
+      "step": 325
+    },
+    {
+      "epoch": 0.00042614211970410575,
+      "grad_norm": 1.5264325141906738,
+      "learning_rate": 4.8423423423423426e-05,
+      "loss": 1.4915,
+      "step": 326
+    },
+    {
+      "epoch": 0.000427449304120376,
+      "grad_norm": 1.4650579690933228,
+      "learning_rate": 4.841841841841842e-05,
+      "loss": 1.9906,
+      "step": 327
+    },
+    {
+      "epoch": 0.0004287564885366463,
+      "grad_norm": 1.6912672519683838,
+      "learning_rate": 4.841341341341342e-05,
+      "loss": 2.5512,
+      "step": 328
+    },
+    {
+      "epoch": 0.00043006367295291654,
+      "grad_norm": 1.419115424156189,
+      "learning_rate": 4.840840840840841e-05,
+      "loss": 1.254,
+      "step": 329
+    },
+    {
+      "epoch": 0.0004313708573691868,
+      "grad_norm": 1.3889597654342651,
+      "learning_rate": 4.8403403403403405e-05,
+      "loss": 1.5753,
+      "step": 330
+    },
+    {
+      "epoch": 0.00043267804178545707,
+      "grad_norm": 1.9419002532958984,
+      "learning_rate": 4.83983983983984e-05,
+      "loss": 1.363,
+      "step": 331
+    },
+    {
+      "epoch": 0.00043398522620172734,
+      "grad_norm": 2.337916374206543,
+      "learning_rate": 4.839339339339339e-05,
+      "loss": 1.8217,
+      "step": 332
+    },
+    {
+      "epoch": 0.00043529241061799755,
+      "grad_norm": 1.259263277053833,
+      "learning_rate": 4.838838838838839e-05,
+      "loss": 1.4361,
+      "step": 333
+    },
+    {
+      "epoch": 0.0004365995950342678,
+      "grad_norm": 1.728857159614563,
+      "learning_rate": 4.8383383383383384e-05,
+      "loss": 1.8216,
+      "step": 334
+    },
+    {
+      "epoch": 0.0004379067794505381,
+      "grad_norm": 1.126977562904358,
+      "learning_rate": 4.837837837837838e-05,
+      "loss": 0.6098,
+      "step": 335
+    },
+    {
+      "epoch": 0.00043921396386680835,
+      "grad_norm": 1.5760388374328613,
+      "learning_rate": 4.8373373373373376e-05,
+      "loss": 1.9721,
+      "step": 336
+    },
+    {
+      "epoch": 0.0004405211482830786,
+      "grad_norm": 1.6654877662658691,
+      "learning_rate": 4.836836836836837e-05,
+      "loss": 2.0176,
+      "step": 337
+    },
+    {
+      "epoch": 0.0004418283326993489,
+      "grad_norm": 1.5803656578063965,
+      "learning_rate": 4.836336336336337e-05,
+      "loss": 1.6791,
+      "step": 338
+    },
+    {
+      "epoch": 0.00044313551711561914,
+      "grad_norm": 1.6455450057983398,
+      "learning_rate": 4.835835835835836e-05,
+      "loss": 2.1326,
+      "step": 339
+    },
+    {
+      "epoch": 0.0004444427015318894,
+      "grad_norm": 1.302128553390503,
+      "learning_rate": 4.8353353353353355e-05,
+      "loss": 1.3892,
+      "step": 340
+    },
+    {
+      "epoch": 0.0004457498859481597,
+      "grad_norm": 1.2957465648651123,
+      "learning_rate": 4.834834834834835e-05,
+      "loss": 1.2624,
+      "step": 341
+    },
+    {
+      "epoch": 0.00044705707036442994,
+      "grad_norm": 1.56611168384552,
+      "learning_rate": 4.834334334334335e-05,
+      "loss": 1.2925,
+      "step": 342
+    },
+    {
+      "epoch": 0.0004483642547807002,
+      "grad_norm": 1.5972720384597778,
+      "learning_rate": 4.833833833833834e-05,
+      "loss": 2.2345,
+      "step": 343
+    },
+    {
+      "epoch": 0.00044967143919697047,
+      "grad_norm": 1.6864628791809082,
+      "learning_rate": 4.8333333333333334e-05,
+      "loss": 2.1596,
+      "step": 344
+    },
+    {
+      "epoch": 0.00045097862361324074,
+      "grad_norm": 1.3631922006607056,
+      "learning_rate": 4.832832832832833e-05,
+      "loss": 0.8551,
+      "step": 345
+    },
+    {
+      "epoch": 0.000452285808029511,
+      "grad_norm": 1.2245639562606812,
+      "learning_rate": 4.8323323323323326e-05,
+      "loss": 1.613,
+      "step": 346
+    },
+    {
+      "epoch": 0.00045359299244578127,
+      "grad_norm": 1.5226577520370483,
+      "learning_rate": 4.8318318318318326e-05,
+      "loss": 1.3359,
+      "step": 347
+    },
+    {
+      "epoch": 0.00045490017686205153,
+      "grad_norm": 1.371559739112854,
+      "learning_rate": 4.831331331331331e-05,
+      "loss": 2.1276,
+      "step": 348
+    },
+    {
+      "epoch": 0.0004562073612783218,
+      "grad_norm": 1.8305604457855225,
+      "learning_rate": 4.830830830830831e-05,
+      "loss": 2.6008,
+      "step": 349
+    },
+    {
+      "epoch": 0.00045751454569459206,
+      "grad_norm": 1.6536656618118286,
+      "learning_rate": 4.8303303303303305e-05,
+      "loss": 1.9033,
+      "step": 350
+    },
+    {
+      "epoch": 0.00045882173011086233,
+      "grad_norm": 1.682886004447937,
+      "learning_rate": 4.82982982982983e-05,
+      "loss": 1.7702,
+      "step": 351
+    },
+    {
+      "epoch": 0.0004601289145271326,
+      "grad_norm": 1.8235169649124146,
+      "learning_rate": 4.82932932932933e-05,
+      "loss": 2.0686,
+      "step": 352
+    },
+    {
+      "epoch": 0.00046143609894340286,
+      "grad_norm": 1.3382108211517334,
+      "learning_rate": 4.828828828828829e-05,
+      "loss": 1.3496,
+      "step": 353
+    },
+    {
+      "epoch": 0.0004627432833596731,
+      "grad_norm": 1.3107976913452148,
+      "learning_rate": 4.828328328328329e-05,
+      "loss": 0.7386,
+      "step": 354
+    },
+    {
+      "epoch": 0.00046405046777594334,
+      "grad_norm": 1.5658841133117676,
+      "learning_rate": 4.8278278278278283e-05,
+      "loss": 1.1042,
+      "step": 355
+    },
+    {
+      "epoch": 0.0004653576521922136,
+      "grad_norm": 1.51932692527771,
+      "learning_rate": 4.827327327327327e-05,
+      "loss": 0.7015,
+      "step": 356
+    },
+    {
+      "epoch": 0.00046666483660848387,
+      "grad_norm": 1.8380029201507568,
+      "learning_rate": 4.826826826826827e-05,
+      "loss": 1.8174,
+      "step": 357
+    },
+    {
+      "epoch": 0.00046797202102475413,
+      "grad_norm": 1.3767679929733276,
+      "learning_rate": 4.826326326326326e-05,
+      "loss": 1.3915,
+      "step": 358
+    },
+    {
+      "epoch": 0.0004692792054410244,
+      "grad_norm": 1.7605303525924683,
+      "learning_rate": 4.825825825825826e-05,
+      "loss": 0.9956,
+      "step": 359
+    },
+    {
+      "epoch": 0.00047058638985729466,
+      "grad_norm": 1.5122482776641846,
+      "learning_rate": 4.8253253253253255e-05,
+      "loss": 1.2155,
+      "step": 360
+    },
+    {
+      "epoch": 0.00047189357427356493,
+      "grad_norm": 1.5101649761199951,
+      "learning_rate": 4.824824824824825e-05,
+      "loss": 1.1057,
+      "step": 361
+    },
+    {
+      "epoch": 0.0004732007586898352,
+      "grad_norm": 1.0014289617538452,
+      "learning_rate": 4.824324324324325e-05,
+      "loss": 0.6814,
+      "step": 362
+    },
+    {
+      "epoch": 0.00047450794310610546,
+      "grad_norm": 1.7926838397979736,
+      "learning_rate": 4.823823823823824e-05,
+      "loss": 2.0843,
+      "step": 363
+    },
+    {
+      "epoch": 0.0004758151275223757,
+      "grad_norm": 1.632832407951355,
+      "learning_rate": 4.823323323323324e-05,
+      "loss": 2.0709,
+      "step": 364
+    },
+    {
+      "epoch": 0.000477122311938646,
+      "grad_norm": 1.414014458656311,
+      "learning_rate": 4.822822822822823e-05,
+      "loss": 1.8019,
+      "step": 365
+    },
+    {
+      "epoch": 0.00047842949635491626,
+      "grad_norm": 1.4348331689834595,
+      "learning_rate": 4.8223223223223226e-05,
+      "loss": 1.7161,
+      "step": 366
+    },
+    {
+      "epoch": 0.0004797366807711865,
+      "grad_norm": 1.537306547164917,
+      "learning_rate": 4.821821821821822e-05,
+      "loss": 0.8713,
+      "step": 367
+    },
+    {
+      "epoch": 0.0004810438651874568,
+      "grad_norm": 1.8898135423660278,
+      "learning_rate": 4.821321321321321e-05,
+      "loss": 2.1803,
+      "step": 368
+    },
+    {
+      "epoch": 0.00048235104960372705,
+      "grad_norm": 1.8930268287658691,
+      "learning_rate": 4.820820820820821e-05,
+      "loss": 1.5469,
+      "step": 369
+    },
+    {
+      "epoch": 0.0004836582340199973,
+      "grad_norm": 1.4677132368087769,
+      "learning_rate": 4.8203203203203205e-05,
+      "loss": 0.9854,
+      "step": 370
+    },
+    {
+      "epoch": 0.0004849654184362676,
+      "grad_norm": 1.2428200244903564,
+      "learning_rate": 4.8198198198198205e-05,
+      "loss": 1.503,
+      "step": 371
+    },
+    {
+      "epoch": 0.00048627260285253785,
+      "grad_norm": 1.542529582977295,
+      "learning_rate": 4.81931931931932e-05,
+      "loss": 1.7597,
+      "step": 372
+    },
+    {
+      "epoch": 0.0004875797872688081,
+      "grad_norm": 1.6563135385513306,
+      "learning_rate": 4.818818818818819e-05,
+      "loss": 1.2799,
+      "step": 373
+    },
+    {
+      "epoch": 0.0004888869716850783,
+      "grad_norm": 1.8432083129882812,
+      "learning_rate": 4.8183183183183184e-05,
+      "loss": 2.2825,
+      "step": 374
+    },
+    {
+      "epoch": 0.0004901941561013486,
+      "grad_norm": 1.8847230672836304,
+      "learning_rate": 4.817817817817818e-05,
+      "loss": 1.7107,
+      "step": 375
+    },
+    {
+      "epoch": 0.0004915013405176189,
+      "grad_norm": 1.4317970275878906,
+      "learning_rate": 4.8173173173173177e-05,
+      "loss": 1.2853,
+      "step": 376
+    },
+    {
+      "epoch": 0.0004928085249338892,
+      "grad_norm": 1.9522016048431396,
+      "learning_rate": 4.816816816816817e-05,
+      "loss": 2.0247,
+      "step": 377
+    },
+    {
+      "epoch": 0.0004941157093501594,
+      "grad_norm": 1.979780912399292,
+      "learning_rate": 4.816316316316317e-05,
+      "loss": 2.8973,
+      "step": 378
+    },
+    {
+      "epoch": 0.0004954228937664297,
+      "grad_norm": 1.2153147459030151,
+      "learning_rate": 4.815815815815816e-05,
+      "loss": 0.918,
+      "step": 379
+    },
+    {
+      "epoch": 0.0004967300781826999,
+      "grad_norm": 1.7080423831939697,
+      "learning_rate": 4.8153153153153155e-05,
+      "loss": 2.3176,
+      "step": 380
+    },
+    {
+      "epoch": 0.0004980372625989702,
+      "grad_norm": 1.4782633781433105,
+      "learning_rate": 4.814814814814815e-05,
+      "loss": 1.4934,
+      "step": 381
+    },
+    {
+      "epoch": 0.0004993444470152405,
+      "grad_norm": 1.3005950450897217,
+      "learning_rate": 4.814314314314314e-05,
+      "loss": 0.9541,
+      "step": 382
+    },
+    {
+      "epoch": 0.0005006516314315108,
+      "grad_norm": 1.6197017431259155,
+      "learning_rate": 4.813813813813814e-05,
+      "loss": 1.5796,
+      "step": 383
+    },
+    {
+      "epoch": 0.000501958815847781,
+      "grad_norm": 1.5193967819213867,
+      "learning_rate": 4.8133133133133134e-05,
+      "loss": 1.5869,
+      "step": 384
+    },
+    {
+      "epoch": 0.0005032660002640513,
+      "grad_norm": 2.8697612285614014,
+      "learning_rate": 4.8128128128128134e-05,
+      "loss": 1.4886,
+      "step": 385
+    },
+    {
+      "epoch": 0.0005045731846803215,
+      "grad_norm": 1.835699200630188,
+      "learning_rate": 4.812312312312313e-05,
+      "loss": 1.6965,
+      "step": 386
+    },
+    {
+      "epoch": 0.0005058803690965917,
+      "grad_norm": 1.51179039478302,
+      "learning_rate": 4.811811811811812e-05,
+      "loss": 1.5248,
+      "step": 387
+    },
+    {
+      "epoch": 0.000507187553512862,
+      "grad_norm": 1.3816337585449219,
+      "learning_rate": 4.811311311311312e-05,
+      "loss": 1.8567,
+      "step": 388
+    },
+    {
+      "epoch": 0.0005084947379291323,
+      "grad_norm": 1.3670501708984375,
+      "learning_rate": 4.810810810810811e-05,
+      "loss": 0.9172,
+      "step": 389
+    },
+    {
+      "epoch": 0.0005098019223454026,
+      "grad_norm": 1.5191140174865723,
+      "learning_rate": 4.8103103103103105e-05,
+      "loss": 2.2351,
+      "step": 390
+    },
+    {
+      "epoch": 0.0005111091067616728,
+      "grad_norm": 1.9390945434570312,
+      "learning_rate": 4.80980980980981e-05,
+      "loss": 1.5947,
+      "step": 391
+    },
+    {
+      "epoch": 0.0005124162911779431,
+      "grad_norm": 1.0936216115951538,
+      "learning_rate": 4.809309309309309e-05,
+      "loss": 0.8947,
+      "step": 392
+    },
+    {
+      "epoch": 0.0005137234755942133,
+      "grad_norm": 3.829148530960083,
+      "learning_rate": 4.808808808808809e-05,
+      "loss": 2.7547,
+      "step": 393
+    },
+    {
+      "epoch": 0.0005150306600104836,
+      "grad_norm": 2.89119815826416,
+      "learning_rate": 4.8083083083083084e-05,
+      "loss": 1.7366,
+      "step": 394
+    },
+    {
+      "epoch": 0.0005163378444267538,
+      "grad_norm": 1.6022891998291016,
+      "learning_rate": 4.8078078078078084e-05,
+      "loss": 1.786,
+      "step": 395
+    },
+    {
+      "epoch": 0.0005176450288430242,
+      "grad_norm": 1.2842696905136108,
+      "learning_rate": 4.807307307307308e-05,
+      "loss": 0.2522,
+      "step": 396
+    },
+    {
+      "epoch": 0.0005189522132592944,
+      "grad_norm": 1.976442813873291,
+      "learning_rate": 4.806806806806807e-05,
+      "loss": 1.9734,
+      "step": 397
+    },
+    {
+      "epoch": 0.0005202593976755647,
+      "grad_norm": 1.5915790796279907,
+      "learning_rate": 4.806306306306306e-05,
+      "loss": 1.9909,
+      "step": 398
+    },
+    {
+      "epoch": 0.0005215665820918349,
+      "grad_norm": 1.658768892288208,
+      "learning_rate": 4.8058058058058056e-05,
+      "loss": 1.6076,
+      "step": 399
+    },
+    {
+      "epoch": 0.0005228737665081052,
+      "grad_norm": 1.4726157188415527,
+      "learning_rate": 4.8053053053053055e-05,
+      "loss": 2.0277,
+      "step": 400
+    },
+    {
+      "epoch": 0.0005241809509243754,
+      "grad_norm": 2.232145309448242,
+      "learning_rate": 4.804804804804805e-05,
+      "loss": 1.7958,
+      "step": 401
+    },
+    {
+      "epoch": 0.0005254881353406458,
+      "grad_norm": 1.6406078338623047,
+      "learning_rate": 4.804304304304305e-05,
+      "loss": 1.7463,
+      "step": 402
+    },
+    {
+      "epoch": 0.000526795319756916,
+      "grad_norm": 1.1940925121307373,
+      "learning_rate": 4.803803803803804e-05,
+      "loss": 0.7934,
+      "step": 403
+    },
+    {
+      "epoch": 0.0005281025041731863,
+      "grad_norm": 1.7259567975997925,
+      "learning_rate": 4.8033033033033034e-05,
+      "loss": 0.9874,
+      "step": 404
+    },
+    {
+      "epoch": 0.0005294096885894565,
+      "grad_norm": 1.845804214477539,
+      "learning_rate": 4.8028028028028034e-05,
+      "loss": 0.4897,
+      "step": 405
+    },
+    {
+      "epoch": 0.0005307168730057268,
+      "grad_norm": 1.6611266136169434,
+      "learning_rate": 4.802302302302302e-05,
+      "loss": 1.7425,
+      "step": 406
+    },
+    {
+      "epoch": 0.000532024057421997,
+      "grad_norm": 1.487190842628479,
+      "learning_rate": 4.801801801801802e-05,
+      "loss": 2.0357,
+      "step": 407
+    },
+    {
+      "epoch": 0.0005333312418382674,
+      "grad_norm": 0.8890923261642456,
+      "learning_rate": 4.801301301301301e-05,
+      "loss": 0.5583,
+      "step": 408
+    },
+    {
+      "epoch": 0.0005346384262545376,
+      "grad_norm": 1.4603744745254517,
+      "learning_rate": 4.800800800800801e-05,
+      "loss": 1.6225,
+      "step": 409
+    },
+    {
+      "epoch": 0.0005359456106708078,
+      "grad_norm": 1.9771097898483276,
+      "learning_rate": 4.8003003003003005e-05,
+      "loss": 1.7477,
+      "step": 410
+    },
+    {
+      "epoch": 0.0005372527950870781,
+      "grad_norm": 1.5973646640777588,
+      "learning_rate": 4.7997997997998e-05,
+      "loss": 1.5643,
+      "step": 411
+    },
+    {
+      "epoch": 0.0005385599795033483,
+      "grad_norm": 1.9672987461090088,
+      "learning_rate": 4.7992992992993e-05,
+      "loss": 2.1034,
+      "step": 412
+    },
+    {
+      "epoch": 0.0005398671639196186,
+      "grad_norm": 1.4598102569580078,
+      "learning_rate": 4.798798798798799e-05,
+      "loss": 1.9834,
+      "step": 413
+    },
+    {
+      "epoch": 0.0005411743483358888,
+      "grad_norm": 1.8543391227722168,
+      "learning_rate": 4.798298298298299e-05,
+      "loss": 1.8818,
+      "step": 414
+    },
+    {
+      "epoch": 0.0005424815327521592,
+      "grad_norm": 1.3450524806976318,
+      "learning_rate": 4.797797797797798e-05,
+      "loss": 1.0332,
+      "step": 415
+    },
+    {
+      "epoch": 0.0005437887171684294,
+      "grad_norm": 1.323103904724121,
+      "learning_rate": 4.797297297297298e-05,
+      "loss": 1.4493,
+      "step": 416
+    },
+    {
+      "epoch": 0.0005450959015846997,
+      "grad_norm": 3.834731101989746,
+      "learning_rate": 4.796796796796797e-05,
+      "loss": 1.7237,
+      "step": 417
+    },
+    {
+      "epoch": 0.0005464030860009699,
+      "grad_norm": 1.6012156009674072,
+      "learning_rate": 4.796296296296296e-05,
+      "loss": 1.7705,
+      "step": 418
+    },
+    {
+      "epoch": 0.0005477102704172402,
+      "grad_norm": 1.6005125045776367,
+      "learning_rate": 4.795795795795796e-05,
+      "loss": 1.2692,
+      "step": 419
+    },
+    {
+      "epoch": 0.0005490174548335104,
+      "grad_norm": 0.04926226660609245,
+      "learning_rate": 4.7952952952952956e-05,
+      "loss": 0.001,
+      "step": 420
+    },
+    {
+      "epoch": 0.0005503246392497808,
+      "grad_norm": 1.9261682033538818,
+      "learning_rate": 4.7947947947947955e-05,
+      "loss": 1.3089,
+      "step": 421
+    },
+    {
+      "epoch": 0.000551631823666051,
+      "grad_norm": 1.4734810590744019,
+      "learning_rate": 4.794294294294295e-05,
+      "loss": 1.9636,
+      "step": 422
+    },
+    {
+      "epoch": 0.0005529390080823213,
+      "grad_norm": 1.5746914148330688,
+      "learning_rate": 4.793793793793794e-05,
+      "loss": 1.481,
+      "step": 423
+    },
+    {
+      "epoch": 0.0005542461924985915,
+      "grad_norm": 1.6360576152801514,
+      "learning_rate": 4.7932932932932934e-05,
+      "loss": 2.0989,
+      "step": 424
+    },
+    {
+      "epoch": 0.0005555533769148618,
+      "grad_norm": 1.9103704690933228,
+      "learning_rate": 4.792792792792793e-05,
+      "loss": 2.5168,
+      "step": 425
+    },
+    {
+      "epoch": 0.000556860561331132,
+      "grad_norm": 1.6170457601547241,
+      "learning_rate": 4.792292292292293e-05,
+      "loss": 2.1027,
+      "step": 426
+    },
+    {
+      "epoch": 0.0005581677457474023,
+      "grad_norm": 1.6519997119903564,
+      "learning_rate": 4.791791791791792e-05,
+      "loss": 1.8065,
+      "step": 427
+    },
+    {
+      "epoch": 0.0005594749301636726,
+      "grad_norm": 2.1365749835968018,
+      "learning_rate": 4.791291291291291e-05,
+      "loss": 2.1404,
+      "step": 428
+    },
+    {
+      "epoch": 0.0005607821145799429,
+      "grad_norm": 1.484369158744812,
+      "learning_rate": 4.790790790790791e-05,
+      "loss": 1.6508,
+      "step": 429
+    },
+    {
+      "epoch": 0.0005620892989962131,
+      "grad_norm": 2.2153050899505615,
+      "learning_rate": 4.7902902902902906e-05,
+      "loss": 1.9617,
+      "step": 430
+    },
+    {
+      "epoch": 0.0005633964834124833,
+      "grad_norm": 1.4849168062210083,
+      "learning_rate": 4.78978978978979e-05,
+      "loss": 1.6652,
+      "step": 431
+    },
+    {
+      "epoch": 0.0005647036678287536,
+      "grad_norm": 1.4296141862869263,
+      "learning_rate": 4.789289289289289e-05,
+      "loss": 1.2995,
+      "step": 432
+    },
+    {
+      "epoch": 0.0005660108522450238,
+      "grad_norm": 1.6376841068267822,
+      "learning_rate": 4.788788788788789e-05,
+      "loss": 2.2176,
+      "step": 433
+    },
+    {
+      "epoch": 0.0005673180366612942,
+      "grad_norm": 1.422378420829773,
+      "learning_rate": 4.7882882882882884e-05,
+      "loss": 1.4486,
+      "step": 434
+    },
+    {
+      "epoch": 0.0005686252210775644,
+      "grad_norm": 1.5652416944503784,
+      "learning_rate": 4.787787787787788e-05,
+      "loss": 1.8622,
+      "step": 435
+    },
+    {
+      "epoch": 0.0005699324054938347,
+      "grad_norm": 1.2181988954544067,
+      "learning_rate": 4.787287287287288e-05,
+      "loss": 0.7264,
+      "step": 436
+    },
+    {
+      "epoch": 0.0005712395899101049,
+      "grad_norm": 1.4485244750976562,
+      "learning_rate": 4.786786786786787e-05,
+      "loss": 1.0068,
+      "step": 437
+    },
+    {
+      "epoch": 0.0005725467743263752,
+      "grad_norm": 1.6365692615509033,
+      "learning_rate": 4.786286286286287e-05,
+      "loss": 1.8623,
+      "step": 438
+    },
+    {
+      "epoch": 0.0005738539587426454,
+      "grad_norm": 1.362836241722107,
+      "learning_rate": 4.785785785785786e-05,
+      "loss": 1.2694,
+      "step": 439
+    },
+    {
+      "epoch": 0.0005751611431589157,
+      "grad_norm": 1.5164575576782227,
+      "learning_rate": 4.7852852852852856e-05,
+      "loss": 1.9222,
+      "step": 440
+    },
+    {
+      "epoch": 0.000576468327575186,
+      "grad_norm": 1.7084901332855225,
+      "learning_rate": 4.784784784784785e-05,
+      "loss": 1.8659,
+      "step": 441
+    },
+    {
+      "epoch": 0.0005777755119914563,
+      "grad_norm": 2.0331461429595947,
+      "learning_rate": 4.784284284284284e-05,
+      "loss": 1.4633,
+      "step": 442
+    },
+    {
+      "epoch": 0.0005790826964077265,
+      "grad_norm": 1.529123306274414,
+      "learning_rate": 4.783783783783784e-05,
+      "loss": 1.4277,
+      "step": 443
+    },
+    {
+      "epoch": 0.0005803898808239968,
+      "grad_norm": 1.4111024141311646,
+      "learning_rate": 4.7832832832832834e-05,
+      "loss": 1.7261,
+      "step": 444
+    },
+    {
+      "epoch": 0.000581697065240267,
+      "grad_norm": 0.98598313331604,
+      "learning_rate": 4.7827827827827834e-05,
+      "loss": 0.4229,
+      "step": 445
+    },
+    {
+      "epoch": 0.0005830042496565373,
+      "grad_norm": 2.0867719650268555,
+      "learning_rate": 4.782282282282283e-05,
+      "loss": 1.9944,
+      "step": 446
+    },
+    {
+      "epoch": 0.0005843114340728075,
+      "grad_norm": 1.6054155826568604,
+      "learning_rate": 4.781781781781782e-05,
+      "loss": 2.2063,
+      "step": 447
+    },
+    {
+      "epoch": 0.0005856186184890779,
+      "grad_norm": 1.374302864074707,
+      "learning_rate": 4.781281281281281e-05,
+      "loss": 1.4218,
+      "step": 448
+    },
+    {
+      "epoch": 0.0005869258029053481,
+      "grad_norm": 1.8737014532089233,
+      "learning_rate": 4.7807807807807806e-05,
+      "loss": 2.0067,
+      "step": 449
+    },
+    {
+      "epoch": 0.0005882329873216184,
+      "grad_norm": 1.221209168434143,
+      "learning_rate": 4.7802802802802806e-05,
+      "loss": 1.1187,
+      "step": 450
+    },
+    {
+      "epoch": 0.0005895401717378886,
+      "grad_norm": 2.31976580619812,
+      "learning_rate": 4.77977977977978e-05,
+      "loss": 2.244,
+      "step": 451
+    },
+    {
+      "epoch": 0.0005908473561541588,
+      "grad_norm": 2.0758728981018066,
+      "learning_rate": 4.77927927927928e-05,
+      "loss": 2.37,
+      "step": 452
+    },
+    {
+      "epoch": 0.0005921545405704291,
+      "grad_norm": 1.5117443799972534,
+      "learning_rate": 4.778778778778779e-05,
+      "loss": 1.4368,
+      "step": 453
+    },
+    {
+      "epoch": 0.0005934617249866994,
+      "grad_norm": 2.0739381313323975,
+      "learning_rate": 4.7782782782782784e-05,
+      "loss": 2.3445,
+      "step": 454
+    },
+    {
+      "epoch": 0.0005947689094029697,
+      "grad_norm": 1.6032236814498901,
+      "learning_rate": 4.7777777777777784e-05,
+      "loss": 1.3034,
+      "step": 455
+    },
+    {
+      "epoch": 0.0005960760938192399,
+      "grad_norm": 1.8953197002410889,
+      "learning_rate": 4.777277277277277e-05,
+      "loss": 2.4014,
+      "step": 456
+    },
+    {
+      "epoch": 0.0005973832782355102,
+      "grad_norm": 1.1892727613449097,
+      "learning_rate": 4.776776776776777e-05,
+      "loss": 1.0365,
+      "step": 457
+    },
+    {
+      "epoch": 0.0005986904626517804,
+      "grad_norm": 1.3845131397247314,
+      "learning_rate": 4.776276276276276e-05,
+      "loss": 1.4982,
+      "step": 458
+    },
+    {
+      "epoch": 0.0005999976470680507,
+      "grad_norm": 2.0147531032562256,
+      "learning_rate": 4.775775775775776e-05,
+      "loss": 1.0432,
+      "step": 459
+    },
+    {
+      "epoch": 0.0006013048314843209,
+      "grad_norm": 1.3901069164276123,
+      "learning_rate": 4.7752752752752756e-05,
+      "loss": 1.1874,
+      "step": 460
+    },
+    {
+      "epoch": 0.0006026120159005913,
+      "grad_norm": 1.7510122060775757,
+      "learning_rate": 4.774774774774775e-05,
+      "loss": 1.7532,
+      "step": 461
+    },
+    {
+      "epoch": 0.0006039192003168615,
+      "grad_norm": 1.4255083799362183,
+      "learning_rate": 4.774274274274275e-05,
+      "loss": 1.2638,
+      "step": 462
+    },
+    {
+      "epoch": 0.0006052263847331318,
+      "grad_norm": 1.3789587020874023,
+      "learning_rate": 4.773773773773774e-05,
+      "loss": 1.2885,
+      "step": 463
+    },
+    {
+      "epoch": 0.000606533569149402,
+      "grad_norm": 1.8223754167556763,
+      "learning_rate": 4.7732732732732734e-05,
+      "loss": 1.846,
+      "step": 464
+    },
+    {
+      "epoch": 0.0006078407535656723,
+      "grad_norm": 1.9569519758224487,
+      "learning_rate": 4.772772772772773e-05,
+      "loss": 2.0224,
+      "step": 465
+    },
+    {
+      "epoch": 0.0006091479379819425,
+      "grad_norm": 1.589490532875061,
+      "learning_rate": 4.772272272272272e-05,
+      "loss": 1.4883,
+      "step": 466
+    },
+    {
+      "epoch": 0.0006104551223982129,
+      "grad_norm": 1.4150370359420776,
+      "learning_rate": 4.771771771771772e-05,
+      "loss": 1.1537,
+      "step": 467
+    },
+    {
+      "epoch": 0.0006117623068144831,
+      "grad_norm": 1.3750197887420654,
+      "learning_rate": 4.771271271271271e-05,
+      "loss": 1.6263,
+      "step": 468
+    },
+    {
+      "epoch": 0.0006130694912307534,
+      "grad_norm": 1.5440406799316406,
+      "learning_rate": 4.770770770770771e-05,
+      "loss": 1.6786,
+      "step": 469
+    },
+    {
+      "epoch": 0.0006143766756470236,
+      "grad_norm": 1.7556005716323853,
+      "learning_rate": 4.7702702702702706e-05,
+      "loss": 2.0755,
+      "step": 470
+    },
+    {
+      "epoch": 0.0006156838600632939,
+      "grad_norm": 1.2925258874893188,
+      "learning_rate": 4.76976976976977e-05,
+      "loss": 0.8731,
+      "step": 471
+    },
+    {
+      "epoch": 0.0006169910444795641,
+      "grad_norm": 1.463951826095581,
+      "learning_rate": 4.76926926926927e-05,
+      "loss": 1.7757,
+      "step": 472
+    },
+    {
+      "epoch": 0.0006182982288958343,
+      "grad_norm": 1.1145371198654175,
+      "learning_rate": 4.7687687687687685e-05,
+      "loss": 0.3717,
+      "step": 473
+    },
+    {
+      "epoch": 0.0006196054133121047,
+      "grad_norm": 1.656201720237732,
+      "learning_rate": 4.7682682682682685e-05,
+      "loss": 1.3816,
+      "step": 474
+    },
+    {
+      "epoch": 0.0006209125977283749,
+      "grad_norm": 1.6448125839233398,
+      "learning_rate": 4.767767767767768e-05,
+      "loss": 1.3373,
+      "step": 475
+    },
+    {
+      "epoch": 0.0006222197821446452,
+      "grad_norm": 1.5466721057891846,
+      "learning_rate": 4.767267267267268e-05,
+      "loss": 1.757,
+      "step": 476
+    },
+    {
+      "epoch": 0.0006235269665609154,
+      "grad_norm": 1.4172669649124146,
+      "learning_rate": 4.766766766766767e-05,
+      "loss": 1.3798,
+      "step": 477
+    },
+    {
+      "epoch": 0.0006248341509771857,
+      "grad_norm": 1.4851760864257812,
+      "learning_rate": 4.766266266266266e-05,
+      "loss": 1.5471,
+      "step": 478
+    },
+    {
+      "epoch": 0.0006261413353934559,
+      "grad_norm": 1.2650772333145142,
+      "learning_rate": 4.765765765765766e-05,
+      "loss": 0.5324,
+      "step": 479
+    },
+    {
+      "epoch": 0.0006274485198097263,
+      "grad_norm": 1.645494818687439,
+      "learning_rate": 4.7652652652652656e-05,
+      "loss": 1.422,
+      "step": 480
+    },
+    {
+      "epoch": 0.0006287557042259965,
+      "grad_norm": 1.4462095499038696,
+      "learning_rate": 4.764764764764765e-05,
+      "loss": 0.7855,
+      "step": 481
+    },
+    {
+      "epoch": 0.0006300628886422668,
+      "grad_norm": 1.6467872858047485,
+      "learning_rate": 4.764264264264264e-05,
+      "loss": 1.7378,
+      "step": 482
+    },
+    {
+      "epoch": 0.000631370073058537,
+      "grad_norm": 1.4757755994796753,
+      "learning_rate": 4.763763763763764e-05,
+      "loss": 1.2541,
+      "step": 483
+    },
+    {
+      "epoch": 0.0006326772574748073,
+      "grad_norm": 1.649613618850708,
+      "learning_rate": 4.7632632632632635e-05,
+      "loss": 1.9402,
+      "step": 484
+    },
+    {
+      "epoch": 0.0006339844418910775,
+      "grad_norm": 1.5895005464553833,
+      "learning_rate": 4.762762762762763e-05,
+      "loss": 1.6001,
+      "step": 485
+    },
+    {
+      "epoch": 0.0006352916263073478,
+      "grad_norm": 1.9634181261062622,
+      "learning_rate": 4.762262262262263e-05,
+      "loss": 1.9354,
+      "step": 486
+    },
+    {
+      "epoch": 0.0006365988107236181,
+      "grad_norm": 1.5781902074813843,
+      "learning_rate": 4.761761761761762e-05,
+      "loss": 1.7391,
+      "step": 487
+    },
+    {
+      "epoch": 0.0006379059951398884,
+      "grad_norm": 1.7558811902999878,
+      "learning_rate": 4.761261261261262e-05,
+      "loss": 1.3157,
+      "step": 488
+    },
+    {
+      "epoch": 0.0006392131795561586,
+      "grad_norm": 1.4790571928024292,
+      "learning_rate": 4.7607607607607606e-05,
+      "loss": 1.2889,
+      "step": 489
+    },
+    {
+      "epoch": 0.0006405203639724289,
+      "grad_norm": 1.9054185152053833,
+      "learning_rate": 4.7602602602602606e-05,
+      "loss": 1.9227,
+      "step": 490
+    },
+    {
+      "epoch": 0.0006418275483886991,
+      "grad_norm": 1.3396105766296387,
+      "learning_rate": 4.75975975975976e-05,
+      "loss": 0.966,
+      "step": 491
+    },
+    {
+      "epoch": 0.0006431347328049694,
+      "grad_norm": 1.6935501098632812,
+      "learning_rate": 4.759259259259259e-05,
+      "loss": 1.3237,
+      "step": 492
+    },
+    {
+      "epoch": 0.0006444419172212397,
+      "grad_norm": 1.58639395236969,
+      "learning_rate": 4.758758758758759e-05,
+      "loss": 1.8338,
+      "step": 493
+    },
+    {
+      "epoch": 0.0006457491016375099,
+      "grad_norm": 1.8331503868103027,
+      "learning_rate": 4.7582582582582585e-05,
+      "loss": 1.5735,
+      "step": 494
+    },
+    {
+      "epoch": 0.0006470562860537802,
+      "grad_norm": 2.0812811851501465,
+      "learning_rate": 4.7577577577577584e-05,
+      "loss": 2.0272,
+      "step": 495
+    },
+    {
+      "epoch": 0.0006483634704700504,
+      "grad_norm": 1.8232816457748413,
+      "learning_rate": 4.757257257257258e-05,
+      "loss": 1.52,
+      "step": 496
+    },
+    {
+      "epoch": 0.0006496706548863207,
+      "grad_norm": 1.5954536199569702,
+      "learning_rate": 4.756756756756757e-05,
+      "loss": 1.8882,
+      "step": 497
+    },
+    {
+      "epoch": 0.0006509778393025909,
+      "grad_norm": 2.2882657051086426,
+      "learning_rate": 4.756256256256256e-05,
+      "loss": 2.3967,
+      "step": 498
+    },
+    {
+      "epoch": 0.0006522850237188612,
+      "grad_norm": 1.4601846933364868,
+      "learning_rate": 4.7557557557557556e-05,
+      "loss": 1.7912,
+      "step": 499
+    },
+    {
+      "epoch": 0.0006535922081351315,
+      "grad_norm": 1.9598947763442993,
+      "learning_rate": 4.7552552552552556e-05,
+      "loss": 2.0452,
+      "step": 500
+    },
+    {
+      "epoch": 0.0006548993925514018,
+      "grad_norm": 1.4029874801635742,
+      "learning_rate": 4.754754754754755e-05,
+      "loss": 2.2084,
+      "step": 501
+    },
+    {
+      "epoch": 0.000656206576967672,
+      "grad_norm": 1.3695791959762573,
+      "learning_rate": 4.754254254254254e-05,
+      "loss": 0.6875,
+      "step": 502
+    },
+    {
+      "epoch": 0.0006575137613839423,
+      "grad_norm": 1.6765706539154053,
+      "learning_rate": 4.753753753753754e-05,
+      "loss": 1.3212,
+      "step": 503
+    },
+    {
+      "epoch": 0.0006588209458002125,
+      "grad_norm": 1.437617301940918,
+      "learning_rate": 4.7532532532532535e-05,
+      "loss": 1.7261,
+      "step": 504
+    },
+    {
+      "epoch": 0.0006601281302164828,
+      "grad_norm": 1.1459113359451294,
+      "learning_rate": 4.7527527527527534e-05,
+      "loss": 1.1455,
+      "step": 505
+    },
+    {
+      "epoch": 0.000661435314632753,
+      "grad_norm": 2.0540614128112793,
+      "learning_rate": 4.752252252252252e-05,
+      "loss": 1.7546,
+      "step": 506
+    },
+    {
+      "epoch": 0.0006627424990490234,
+      "grad_norm": 1.8672137260437012,
+      "learning_rate": 4.751751751751752e-05,
+      "loss": 1.681,
+      "step": 507
+    },
+    {
+      "epoch": 0.0006640496834652936,
+      "grad_norm": 1.7728941440582275,
+      "learning_rate": 4.7512512512512513e-05,
+      "loss": 2.2605,
+      "step": 508
+    },
+    {
+      "epoch": 0.0006653568678815639,
+      "grad_norm": 1.5071007013320923,
+      "learning_rate": 4.7507507507507506e-05,
+      "loss": 1.6046,
+      "step": 509
+    },
+    {
+      "epoch": 0.0006666640522978341,
+      "grad_norm": 2.8894779682159424,
+      "learning_rate": 4.7502502502502506e-05,
+      "loss": 1.8453,
+      "step": 510
+    },
+    {
+      "epoch": 0.0006679712367141044,
+      "grad_norm": 1.3977031707763672,
+      "learning_rate": 4.74974974974975e-05,
+      "loss": 1.7809,
+      "step": 511
+    },
+    {
+      "epoch": 0.0006692784211303746,
+      "grad_norm": 1.560545802116394,
+      "learning_rate": 4.74924924924925e-05,
+      "loss": 1.3079,
+      "step": 512
+    },
+    {
+      "epoch": 0.000670585605546645,
+      "grad_norm": 2.0114705562591553,
+      "learning_rate": 4.748748748748749e-05,
+      "loss": 1.4561,
+      "step": 513
+    },
+    {
+      "epoch": 0.0006718927899629152,
+      "grad_norm": 1.2086541652679443,
+      "learning_rate": 4.7482482482482485e-05,
+      "loss": 1.1433,
+      "step": 514
+    },
+    {
+      "epoch": 0.0006731999743791855,
+      "grad_norm": 1.3815529346466064,
+      "learning_rate": 4.747747747747748e-05,
+      "loss": 1.0519,
+      "step": 515
+    },
+    {
+      "epoch": 0.0006745071587954557,
+      "grad_norm": 2.3520267009735107,
+      "learning_rate": 4.747247247247247e-05,
+      "loss": 1.7171,
+      "step": 516
+    },
+    {
+      "epoch": 0.0006758143432117259,
+      "grad_norm": 1.6693229675292969,
+      "learning_rate": 4.746746746746747e-05,
+      "loss": 2.1353,
+      "step": 517
+    },
+    {
+      "epoch": 0.0006771215276279962,
+      "grad_norm": 1.6879549026489258,
+      "learning_rate": 4.7462462462462463e-05,
+      "loss": 1.3734,
+      "step": 518
+    },
+    {
+      "epoch": 0.0006784287120442664,
+      "grad_norm": 0.8456169366836548,
+      "learning_rate": 4.745745745745746e-05,
+      "loss": 0.3061,
+      "step": 519
+    },
+    {
+      "epoch": 0.0006797358964605368,
+      "grad_norm": 1.6144449710845947,
+      "learning_rate": 4.7452452452452456e-05,
+      "loss": 1.8747,
+      "step": 520
+    },
+    {
+      "epoch": 0.000681043080876807,
+      "grad_norm": 1.71808922290802,
+      "learning_rate": 4.744744744744745e-05,
+      "loss": 1.8467,
+      "step": 521
+    },
+    {
+      "epoch": 0.0006823502652930773,
+      "grad_norm": 1.8412578105926514,
+      "learning_rate": 4.744244244244245e-05,
+      "loss": 2.0737,
+      "step": 522
+    },
+    {
+      "epoch": 0.0006836574497093475,
+      "grad_norm": 1.6421705484390259,
+      "learning_rate": 4.7437437437437435e-05,
+      "loss": 2.1484,
+      "step": 523
+    },
+    {
+      "epoch": 0.0006849646341256178,
+      "grad_norm": 2.2837939262390137,
+      "learning_rate": 4.7432432432432435e-05,
+      "loss": 2.1236,
+      "step": 524
+    },
+    {
+      "epoch": 0.000686271818541888,
+      "grad_norm": 1.500970721244812,
+      "learning_rate": 4.742742742742743e-05,
+      "loss": 1.2437,
+      "step": 525
+    },
+    {
+      "epoch": 0.0006875790029581584,
+      "grad_norm": 1.9431389570236206,
+      "learning_rate": 4.742242242242243e-05,
+      "loss": 1.8633,
+      "step": 526
+    },
+    {
+      "epoch": 0.0006888861873744286,
+      "grad_norm": 1.3766447305679321,
+      "learning_rate": 4.741741741741742e-05,
+      "loss": 1.5172,
+      "step": 527
+    },
+    {
+      "epoch": 0.0006901933717906989,
+      "grad_norm": 2.337942361831665,
+      "learning_rate": 4.7412412412412414e-05,
+      "loss": 1.0124,
+      "step": 528
+    },
+    {
+      "epoch": 0.0006915005562069691,
+      "grad_norm": 1.7455742359161377,
+      "learning_rate": 4.740740740740741e-05,
+      "loss": 1.3154,
+      "step": 529
+    },
+    {
+      "epoch": 0.0006928077406232394,
+      "grad_norm": 1.3493332862854004,
+      "learning_rate": 4.7402402402402406e-05,
+      "loss": 1.8971,
+      "step": 530
+    },
+    {
+      "epoch": 0.0006941149250395096,
+      "grad_norm": 1.5431979894638062,
+      "learning_rate": 4.73973973973974e-05,
+      "loss": 1.6901,
+      "step": 531
+    },
+    {
+      "epoch": 0.00069542210945578,
+      "grad_norm": 1.568732500076294,
+      "learning_rate": 4.739239239239239e-05,
+      "loss": 1.9977,
+      "step": 532
+    },
+    {
+      "epoch": 0.0006967292938720502,
+      "grad_norm": 1.2472697496414185,
+      "learning_rate": 4.738738738738739e-05,
+      "loss": 1.1996,
+      "step": 533
+    },
+    {
+      "epoch": 0.0006980364782883205,
+      "grad_norm": 1.3425178527832031,
+      "learning_rate": 4.7382382382382385e-05,
+      "loss": 1.1555,
+      "step": 534
+    },
+    {
+      "epoch": 0.0006993436627045907,
+      "grad_norm": 1.501718282699585,
+      "learning_rate": 4.737737737737738e-05,
+      "loss": 1.4577,
+      "step": 535
+    },
+    {
+      "epoch": 0.000700650847120861,
+      "grad_norm": 1.6779203414916992,
+      "learning_rate": 4.737237237237238e-05,
+      "loss": 1.8303,
+      "step": 536
+    },
+    {
+      "epoch": 0.0007019580315371312,
+      "grad_norm": 1.2691553831100464,
+      "learning_rate": 4.736736736736737e-05,
+      "loss": 1.7076,
+      "step": 537
+    },
+    {
+      "epoch": 0.0007032652159534014,
+      "grad_norm": 1.707943320274353,
+      "learning_rate": 4.7362362362362364e-05,
+      "loss": 1.1667,
+      "step": 538
+    },
+    {
+      "epoch": 0.0007045724003696718,
+      "grad_norm": 1.7803258895874023,
+      "learning_rate": 4.7357357357357357e-05,
+      "loss": 2.0336,
+      "step": 539
+    },
+    {
+      "epoch": 0.000705879584785942,
+      "grad_norm": 1.6252355575561523,
+      "learning_rate": 4.735235235235235e-05,
+      "loss": 2.6676,
+      "step": 540
+    },
+    {
+      "epoch": 0.0007071867692022123,
+      "grad_norm": 0.8783780932426453,
+      "learning_rate": 4.734734734734735e-05,
+      "loss": 0.4316,
+      "step": 541
+    },
+    {
+      "epoch": 0.0007084939536184825,
+      "grad_norm": 1.0812312364578247,
+      "learning_rate": 4.734234234234234e-05,
+      "loss": 0.8006,
+      "step": 542
+    },
+    {
+      "epoch": 0.0007098011380347528,
+      "grad_norm": 1.6457270383834839,
+      "learning_rate": 4.733733733733734e-05,
+      "loss": 1.2591,
+      "step": 543
+    },
+    {
+      "epoch": 0.000711108322451023,
+      "grad_norm": 1.405547857284546,
+      "learning_rate": 4.7332332332332335e-05,
+      "loss": 2.1359,
+      "step": 544
+    },
+    {
+      "epoch": 0.0007124155068672934,
+      "grad_norm": 1.1879029273986816,
+      "learning_rate": 4.732732732732733e-05,
+      "loss": 0.7761,
+      "step": 545
+    },
+    {
+      "epoch": 0.0007137226912835636,
+      "grad_norm": 1.9608098268508911,
+      "learning_rate": 4.732232232232233e-05,
+      "loss": 1.119,
+      "step": 546
+    },
+    {
+      "epoch": 0.0007150298756998339,
+      "grad_norm": 1.590875267982483,
+      "learning_rate": 4.731731731731732e-05,
+      "loss": 2.1024,
+      "step": 547
+    },
+    {
+      "epoch": 0.0007163370601161041,
+      "grad_norm": 1.3659666776657104,
+      "learning_rate": 4.7312312312312314e-05,
+      "loss": 1.8388,
+      "step": 548
+    },
+    {
+      "epoch": 0.0007176442445323744,
+      "grad_norm": 1.3668805360794067,
+      "learning_rate": 4.730730730730731e-05,
+      "loss": 1.7072,
+      "step": 549
+    },
+    {
+      "epoch": 0.0007189514289486446,
+      "grad_norm": 1.6230323314666748,
+      "learning_rate": 4.7302302302302306e-05,
+      "loss": 0.6986,
+      "step": 550
+    },
+    {
+      "epoch": 0.0007202586133649149,
+      "grad_norm": 1.7399420738220215,
+      "learning_rate": 4.72972972972973e-05,
+      "loss": 1.4688,
+      "step": 551
+    },
+    {
+      "epoch": 0.0007215657977811852,
+      "grad_norm": 1.507240653038025,
+      "learning_rate": 4.729229229229229e-05,
+      "loss": 1.8082,
+      "step": 552
+    },
+    {
+      "epoch": 0.0007228729821974555,
+      "grad_norm": 1.7650612592697144,
+      "learning_rate": 4.728728728728729e-05,
+      "loss": 2.0414,
+      "step": 553
+    },
+    {
+      "epoch": 0.0007241801666137257,
+      "grad_norm": 1.589437484741211,
+      "learning_rate": 4.7282282282282285e-05,
+      "loss": 1.574,
+      "step": 554
+    },
+    {
+      "epoch": 0.000725487351029996,
+      "grad_norm": 1.6265712976455688,
+      "learning_rate": 4.7277277277277285e-05,
+      "loss": 1.8274,
+      "step": 555
+    },
+    {
+      "epoch": 0.0007267945354462662,
+      "grad_norm": 2.1706082820892334,
+      "learning_rate": 4.727227227227227e-05,
+      "loss": 1.7992,
+      "step": 556
+    },
+    {
+      "epoch": 0.0007281017198625365,
+      "grad_norm": 0.9128278493881226,
+      "learning_rate": 4.726726726726727e-05,
+      "loss": 0.3874,
+      "step": 557
+    },
+    {
+      "epoch": 0.0007294089042788067,
+      "grad_norm": 1.6550322771072388,
+      "learning_rate": 4.7262262262262264e-05,
+      "loss": 1.9098,
+      "step": 558
+    },
+    {
+      "epoch": 0.000730716088695077,
+      "grad_norm": 1.6177189350128174,
+      "learning_rate": 4.725725725725726e-05,
+      "loss": 2.2427,
+      "step": 559
+    },
+    {
+      "epoch": 0.0007320232731113473,
+      "grad_norm": 2.232266902923584,
+      "learning_rate": 4.7252252252252256e-05,
+      "loss": 2.6294,
+      "step": 560
+    },
+    {
+      "epoch": 0.0007333304575276175,
+      "grad_norm": 1.5115314722061157,
+      "learning_rate": 4.724724724724725e-05,
+      "loss": 2.5073,
+      "step": 561
+    },
+    {
+      "epoch": 0.0007346376419438878,
+      "grad_norm": 1.4863507747650146,
+      "learning_rate": 4.724224224224225e-05,
+      "loss": 1.5661,
+      "step": 562
+    },
+    {
+      "epoch": 0.000735944826360158,
+      "grad_norm": 1.5657413005828857,
+      "learning_rate": 4.723723723723724e-05,
+      "loss": 1.6394,
+      "step": 563
+    },
+    {
+      "epoch": 0.0007372520107764283,
+      "grad_norm": 1.6362054347991943,
+      "learning_rate": 4.7232232232232235e-05,
+      "loss": 2.3414,
+      "step": 564
+    },
+    {
+      "epoch": 0.0007385591951926986,
+      "grad_norm": 2.6017391681671143,
+      "learning_rate": 4.722722722722723e-05,
+      "loss": 1.6355,
+      "step": 565
+    },
+    {
+      "epoch": 0.0007398663796089689,
+      "grad_norm": 1.5872087478637695,
+      "learning_rate": 4.722222222222222e-05,
+      "loss": 1.3947,
+      "step": 566
+    },
+    {
+      "epoch": 0.0007411735640252391,
+      "grad_norm": 1.5154848098754883,
+      "learning_rate": 4.721721721721722e-05,
+      "loss": 2.1958,
+      "step": 567
+    },
+    {
+      "epoch": 0.0007424807484415094,
+      "grad_norm": 1.678316593170166,
+      "learning_rate": 4.7212212212212214e-05,
+      "loss": 1.894,
+      "step": 568
+    },
+    {
+      "epoch": 0.0007437879328577796,
+      "grad_norm": 1.447416067123413,
+      "learning_rate": 4.7207207207207214e-05,
+      "loss": 1.857,
+      "step": 569
+    },
+    {
+      "epoch": 0.0007450951172740499,
+      "grad_norm": 1.4686932563781738,
+      "learning_rate": 4.7202202202202207e-05,
+      "loss": 2.1657,
+      "step": 570
+    },
+    {
+      "epoch": 0.0007464023016903201,
+      "grad_norm": 1.3969084024429321,
+      "learning_rate": 4.71971971971972e-05,
+      "loss": 1.2883,
+      "step": 571
+    },
+    {
+      "epoch": 0.0007477094861065905,
+      "grad_norm": 1.384464979171753,
+      "learning_rate": 4.71921921921922e-05,
+      "loss": 1.3122,
+      "step": 572
+    },
+    {
+      "epoch": 0.0007490166705228607,
+      "grad_norm": 1.4711626768112183,
+      "learning_rate": 4.7187187187187185e-05,
+      "loss": 2.0273,
+      "step": 573
+    },
+    {
+      "epoch": 0.000750323854939131,
+      "grad_norm": 1.5936601161956787,
+      "learning_rate": 4.7182182182182185e-05,
+      "loss": 1.6843,
+      "step": 574
+    },
+    {
+      "epoch": 0.0007516310393554012,
+      "grad_norm": 2.064631223678589,
+      "learning_rate": 4.717717717717718e-05,
+      "loss": 1.2199,
+      "step": 575
+    },
+    {
+      "epoch": 0.0007529382237716715,
+      "grad_norm": 1.2828699350357056,
+      "learning_rate": 4.717217217217217e-05,
+      "loss": 1.4969,
+      "step": 576
+    },
+    {
+      "epoch": 0.0007542454081879417,
+      "grad_norm": 1.5185223817825317,
+      "learning_rate": 4.716716716716717e-05,
+      "loss": 1.6038,
+      "step": 577
+    },
+    {
+      "epoch": 0.0007555525926042121,
+      "grad_norm": 1.2944966554641724,
+      "learning_rate": 4.7162162162162164e-05,
+      "loss": 1.6003,
+      "step": 578
+    },
+    {
+      "epoch": 0.0007568597770204823,
+      "grad_norm": 1.326327919960022,
+      "learning_rate": 4.7157157157157164e-05,
+      "loss": 0.8233,
+      "step": 579
+    },
+    {
+      "epoch": 0.0007581669614367525,
+      "grad_norm": 2.239767074584961,
+      "learning_rate": 4.715215215215216e-05,
+      "loss": 2.0641,
+      "step": 580
+    },
+    {
+      "epoch": 0.0007594741458530228,
+      "grad_norm": 1.5224183797836304,
+      "learning_rate": 4.714714714714715e-05,
+      "loss": 0.872,
+      "step": 581
+    },
+    {
+      "epoch": 0.000760781330269293,
+      "grad_norm": 1.5005768537521362,
+      "learning_rate": 4.714214214214214e-05,
+      "loss": 2.684,
+      "step": 582
+    },
+    {
+      "epoch": 0.0007620885146855633,
+      "grad_norm": 1.3663041591644287,
+      "learning_rate": 4.7137137137137136e-05,
+      "loss": 1.3607,
+      "step": 583
+    },
+    {
+      "epoch": 0.0007633956991018335,
+      "grad_norm": 1.1366426944732666,
+      "learning_rate": 4.7132132132132135e-05,
+      "loss": 0.7646,
+      "step": 584
+    },
+    {
+      "epoch": 0.0007647028835181039,
+      "grad_norm": 1.3268709182739258,
+      "learning_rate": 4.712712712712713e-05,
+      "loss": 1.6458,
+      "step": 585
+    },
+    {
+      "epoch": 0.0007660100679343741,
+      "grad_norm": 1.845065712928772,
+      "learning_rate": 4.712212212212213e-05,
+      "loss": 1.8178,
+      "step": 586
+    },
+    {
+      "epoch": 0.0007673172523506444,
+      "grad_norm": 1.4855918884277344,
+      "learning_rate": 4.711711711711712e-05,
+      "loss": 1.3352,
+      "step": 587
+    },
+    {
+      "epoch": 0.0007686244367669146,
+      "grad_norm": 1.6433022022247314,
+      "learning_rate": 4.7112112112112114e-05,
+      "loss": 1.9675,
+      "step": 588
+    },
+    {
+      "epoch": 0.0007699316211831849,
+      "grad_norm": 1.2438154220581055,
+      "learning_rate": 4.710710710710711e-05,
+      "loss": 1.1669,
+      "step": 589
+    },
+    {
+      "epoch": 0.0007712388055994551,
+      "grad_norm": 1.503528118133545,
+      "learning_rate": 4.71021021021021e-05,
+      "loss": 0.8251,
+      "step": 590
+    },
+    {
+      "epoch": 0.0007725459900157255,
+      "grad_norm": 2.111198902130127,
+      "learning_rate": 4.70970970970971e-05,
+      "loss": 1.6109,
+      "step": 591
+    },
+    {
+      "epoch": 0.0007738531744319957,
+      "grad_norm": 1.3497482538223267,
+      "learning_rate": 4.709209209209209e-05,
+      "loss": 1.3869,
+      "step": 592
+    },
+    {
+      "epoch": 0.000775160358848266,
+      "grad_norm": 1.7703373432159424,
+      "learning_rate": 4.708708708708709e-05,
+      "loss": 2.495,
+      "step": 593
+    },
+    {
+      "epoch": 0.0007764675432645362,
+      "grad_norm": 1.4268288612365723,
+      "learning_rate": 4.7082082082082085e-05,
+      "loss": 1.2705,
+      "step": 594
+    },
+    {
+      "epoch": 0.0007777747276808065,
+      "grad_norm": 2.1214709281921387,
+      "learning_rate": 4.707707707707708e-05,
+      "loss": 1.3228,
+      "step": 595
+    },
+    {
+      "epoch": 0.0007790819120970767,
+      "grad_norm": 1.483981966972351,
+      "learning_rate": 4.707207207207208e-05,
+      "loss": 2.1134,
+      "step": 596
+    },
+    {
+      "epoch": 0.000780389096513347,
+      "grad_norm": 1.4466556310653687,
+      "learning_rate": 4.7067067067067064e-05,
+      "loss": 0.9819,
+      "step": 597
+    },
+    {
+      "epoch": 0.0007816962809296173,
+      "grad_norm": 2.0418903827667236,
+      "learning_rate": 4.7062062062062064e-05,
+      "loss": 1.6331,
+      "step": 598
+    },
+    {
+      "epoch": 0.0007830034653458876,
+      "grad_norm": 1.142237901687622,
+      "learning_rate": 4.705705705705706e-05,
+      "loss": 0.4096,
+      "step": 599
+    },
+    {
+      "epoch": 0.0007843106497621578,
+      "grad_norm": 1.4778763055801392,
+      "learning_rate": 4.705205205205206e-05,
+      "loss": 1.5868,
+      "step": 600
+    },
+    {
+      "epoch": 0.0007856178341784281,
+      "grad_norm": 1.6162097454071045,
+      "learning_rate": 4.704704704704705e-05,
+      "loss": 1.2317,
+      "step": 601
+    },
+    {
+      "epoch": 0.0007869250185946983,
+      "grad_norm": 1.3901240825653076,
+      "learning_rate": 4.704204204204204e-05,
+      "loss": 1.94,
+      "step": 602
+    },
+    {
+      "epoch": 0.0007882322030109685,
+      "grad_norm": 1.5318193435668945,
+      "learning_rate": 4.703703703703704e-05,
+      "loss": 2.0653,
+      "step": 603
+    },
+    {
+      "epoch": 0.0007895393874272389,
+      "grad_norm": 1.466309905052185,
+      "learning_rate": 4.7032032032032035e-05,
+      "loss": 1.6146,
+      "step": 604
+    },
+    {
+      "epoch": 0.0007908465718435091,
+      "grad_norm": 1.4965687990188599,
+      "learning_rate": 4.7027027027027035e-05,
+      "loss": 1.3284,
+      "step": 605
+    },
+    {
+      "epoch": 0.0007921537562597794,
+      "grad_norm": 1.515586018562317,
+      "learning_rate": 4.702202202202202e-05,
+      "loss": 1.8626,
+      "step": 606
+    },
+    {
+      "epoch": 0.0007934609406760496,
+      "grad_norm": 1.9984151124954224,
+      "learning_rate": 4.7017017017017014e-05,
+      "loss": 1.9397,
+      "step": 607
+    },
+    {
+      "epoch": 0.0007947681250923199,
+      "grad_norm": 1.4615470170974731,
+      "learning_rate": 4.7012012012012014e-05,
+      "loss": 1.2842,
+      "step": 608
+    },
+    {
+      "epoch": 0.0007960753095085901,
+      "grad_norm": 1.4304986000061035,
+      "learning_rate": 4.700700700700701e-05,
+      "loss": 1.1272,
+      "step": 609
+    },
+    {
+      "epoch": 0.0007973824939248604,
+      "grad_norm": 0.4988657236099243,
+      "learning_rate": 4.700200200200201e-05,
+      "loss": 0.1799,
+      "step": 610
+    },
+    {
+      "epoch": 0.0007986896783411307,
+      "grad_norm": 1.984812617301941,
+      "learning_rate": 4.6996996996997e-05,
+      "loss": 2.4337,
+      "step": 611
+    },
+    {
+      "epoch": 0.000799996862757401,
+      "grad_norm": 1.6376155614852905,
+      "learning_rate": 4.699199199199199e-05,
+      "loss": 2.8601,
+      "step": 612
+    },
+    {
+      "epoch": 0.0008013040471736712,
+      "grad_norm": 1.282543659210205,
+      "learning_rate": 4.698698698698699e-05,
+      "loss": 1.2701,
+      "step": 613
+    },
+    {
+      "epoch": 0.0008026112315899415,
+      "grad_norm": 2.0118024349212646,
+      "learning_rate": 4.698198198198198e-05,
+      "loss": 0.913,
+      "step": 614
+    },
+    {
+      "epoch": 0.0008039184160062117,
+      "grad_norm": 1.4409403800964355,
+      "learning_rate": 4.697697697697698e-05,
+      "loss": 1.1211,
+      "step": 615
+    },
+    {
+      "epoch": 0.000805225600422482,
+      "grad_norm": 1.6272765398025513,
+      "learning_rate": 4.697197197197197e-05,
+      "loss": 1.5744,
+      "step": 616
+    },
+    {
+      "epoch": 0.0008065327848387523,
+      "grad_norm": 1.487402319908142,
+      "learning_rate": 4.696696696696697e-05,
+      "loss": 1.4778,
+      "step": 617
+    },
+    {
+      "epoch": 0.0008078399692550226,
+      "grad_norm": 1.682793140411377,
+      "learning_rate": 4.6961961961961964e-05,
+      "loss": 2.0187,
+      "step": 618
+    },
+    {
+      "epoch": 0.0008091471536712928,
+      "grad_norm": 1.9255287647247314,
+      "learning_rate": 4.695695695695696e-05,
+      "loss": 2.0679,
+      "step": 619
+    },
+    {
+      "epoch": 0.0008104543380875631,
+      "grad_norm": 1.6533180475234985,
+      "learning_rate": 4.695195195195196e-05,
+      "loss": 1.8739,
+      "step": 620
+    },
+    {
+      "epoch": 0.0008117615225038333,
+      "grad_norm": 1.622597336769104,
+      "learning_rate": 4.694694694694695e-05,
+      "loss": 1.5089,
+      "step": 621
+    },
+    {
+      "epoch": 0.0008130687069201036,
+      "grad_norm": 1.182200312614441,
+      "learning_rate": 4.694194194194194e-05,
+      "loss": 0.8188,
+      "step": 622
+    },
+    {
+      "epoch": 0.0008143758913363738,
+      "grad_norm": 1.1160887479782104,
+      "learning_rate": 4.6936936936936936e-05,
+      "loss": 0.7047,
+      "step": 623
+    },
+    {
+      "epoch": 0.0008156830757526441,
+      "grad_norm": 1.5866217613220215,
+      "learning_rate": 4.6931931931931936e-05,
+      "loss": 1.5625,
+      "step": 624
+    },
+    {
+      "epoch": 0.0008169902601689144,
+      "grad_norm": 1.6501386165618896,
+      "learning_rate": 4.692692692692693e-05,
+      "loss": 1.3065,
+      "step": 625
+    },
+    {
+      "epoch": 0.0008182974445851846,
+      "grad_norm": 1.5348718166351318,
+      "learning_rate": 4.692192192192192e-05,
+      "loss": 1.7483,
+      "step": 626
+    },
+    {
+      "epoch": 0.0008196046290014549,
+      "grad_norm": 1.7987422943115234,
+      "learning_rate": 4.691691691691692e-05,
+      "loss": 1.5134,
+      "step": 627
+    },
+    {
+      "epoch": 0.0008209118134177251,
+      "grad_norm": 2.0807056427001953,
+      "learning_rate": 4.6911911911911914e-05,
+      "loss": 1.7906,
+      "step": 628
+    },
+    {
+      "epoch": 0.0008222189978339954,
+      "grad_norm": 1.5999367237091064,
+      "learning_rate": 4.6906906906906914e-05,
+      "loss": 1.8923,
+      "step": 629
+    },
+    {
+      "epoch": 0.0008235261822502656,
+      "grad_norm": 1.6543577909469604,
+      "learning_rate": 4.690190190190191e-05,
+      "loss": 1.6049,
+      "step": 630
+    },
+    {
+      "epoch": 0.000824833366666536,
+      "grad_norm": 1.6403521299362183,
+      "learning_rate": 4.68968968968969e-05,
+      "loss": 2.2661,
+      "step": 631
+    },
+    {
+      "epoch": 0.0008261405510828062,
+      "grad_norm": 1.5473531484603882,
+      "learning_rate": 4.689189189189189e-05,
+      "loss": 1.4505,
+      "step": 632
+    },
+    {
+      "epoch": 0.0008274477354990765,
+      "grad_norm": 1.2692561149597168,
+      "learning_rate": 4.6886886886886886e-05,
+      "loss": 0.7423,
+      "step": 633
+    },
+    {
+      "epoch": 0.0008287549199153467,
+      "grad_norm": 1.5211836099624634,
+      "learning_rate": 4.6881881881881886e-05,
+      "loss": 1.8324,
+      "step": 634
+    },
+    {
+      "epoch": 0.000830062104331617,
+      "grad_norm": 1.5988188982009888,
+      "learning_rate": 4.687687687687688e-05,
+      "loss": 1.3321,
+      "step": 635
+    },
+    {
+      "epoch": 0.0008313692887478872,
+      "grad_norm": 1.7307522296905518,
+      "learning_rate": 4.687187187187188e-05,
+      "loss": 1.6279,
+      "step": 636
+    },
+    {
+      "epoch": 0.0008326764731641576,
+      "grad_norm": 2.006218194961548,
+      "learning_rate": 4.686686686686687e-05,
+      "loss": 1.6494,
+      "step": 637
+    },
+    {
+      "epoch": 0.0008339836575804278,
+      "grad_norm": 1.4637088775634766,
+      "learning_rate": 4.6861861861861864e-05,
+      "loss": 1.8269,
+      "step": 638
+    },
+    {
+      "epoch": 0.0008352908419966981,
+      "grad_norm": 1.680317759513855,
+      "learning_rate": 4.685685685685686e-05,
+      "loss": 2.0398,
+      "step": 639
+    },
+    {
+      "epoch": 0.0008365980264129683,
+      "grad_norm": 1.4945439100265503,
+      "learning_rate": 4.685185185185185e-05,
+      "loss": 2.064,
+      "step": 640
+    },
+    {
+      "epoch": 0.0008379052108292386,
+      "grad_norm": 1.4014527797698975,
+      "learning_rate": 4.684684684684685e-05,
+      "loss": 1.3195,
+      "step": 641
+    },
+    {
+      "epoch": 0.0008392123952455088,
+      "grad_norm": 1.392907977104187,
+      "learning_rate": 4.684184184184184e-05,
+      "loss": 1.3836,
+      "step": 642
+    },
+    {
+      "epoch": 0.0008405195796617792,
+      "grad_norm": 1.1703048944473267,
+      "learning_rate": 4.6836836836836836e-05,
+      "loss": 0.8156,
+      "step": 643
+    },
+    {
+      "epoch": 0.0008418267640780494,
+      "grad_norm": 1.5210392475128174,
+      "learning_rate": 4.6831831831831836e-05,
+      "loss": 1.7891,
+      "step": 644
+    },
+    {
+      "epoch": 0.0008431339484943196,
+      "grad_norm": 1.8993138074874878,
+      "learning_rate": 4.682682682682683e-05,
+      "loss": 1.7191,
+      "step": 645
+    },
+    {
+      "epoch": 0.0008444411329105899,
+      "grad_norm": 1.5546610355377197,
+      "learning_rate": 4.682182182182183e-05,
+      "loss": 1.1105,
+      "step": 646
+    },
+    {
+      "epoch": 0.0008457483173268601,
+      "grad_norm": 1.9155322313308716,
+      "learning_rate": 4.6816816816816815e-05,
+      "loss": 1.4803,
+      "step": 647
+    },
+    {
+      "epoch": 0.0008470555017431304,
+      "grad_norm": 1.6106840372085571,
+      "learning_rate": 4.6811811811811814e-05,
+      "loss": 2.0653,
+      "step": 648
+    },
+    {
+      "epoch": 0.0008483626861594006,
+      "grad_norm": 1.6042048931121826,
+      "learning_rate": 4.680680680680681e-05,
+      "loss": 2.255,
+      "step": 649
+    },
+    {
+      "epoch": 0.000849669870575671,
+      "grad_norm": 1.74154531955719,
+      "learning_rate": 4.68018018018018e-05,
+      "loss": 1.6876,
+      "step": 650
+    },
+    {
+      "epoch": 0.0008509770549919412,
+      "grad_norm": 1.889991044998169,
+      "learning_rate": 4.67967967967968e-05,
+      "loss": 2.7168,
+      "step": 651
+    },
+    {
+      "epoch": 0.0008522842394082115,
+      "grad_norm": 1.2411787509918213,
+      "learning_rate": 4.679179179179179e-05,
+      "loss": 1.236,
+      "step": 652
+    },
+    {
+      "epoch": 0.0008535914238244817,
+      "grad_norm": 1.6380681991577148,
+      "learning_rate": 4.678678678678679e-05,
+      "loss": 1.0434,
+      "step": 653
+    },
+    {
+      "epoch": 0.000854898608240752,
+      "grad_norm": 1.0483059883117676,
+      "learning_rate": 4.6781781781781786e-05,
+      "loss": 0.4068,
+      "step": 654
+    },
+    {
+      "epoch": 0.0008562057926570222,
+      "grad_norm": 1.4122055768966675,
+      "learning_rate": 4.677677677677678e-05,
+      "loss": 1.1278,
+      "step": 655
+    },
+    {
+      "epoch": 0.0008575129770732926,
+      "grad_norm": 1.9559366703033447,
+      "learning_rate": 4.677177177177177e-05,
+      "loss": 1.7594,
+      "step": 656
+    },
+    {
+      "epoch": 0.0008588201614895628,
+      "grad_norm": 1.5630970001220703,
+      "learning_rate": 4.6766766766766765e-05,
+      "loss": 1.9152,
+      "step": 657
+    },
+    {
+      "epoch": 0.0008601273459058331,
+      "grad_norm": 1.0776516199111938,
+      "learning_rate": 4.6761761761761764e-05,
+      "loss": 0.8263,
+      "step": 658
+    },
+    {
+      "epoch": 0.0008614345303221033,
+      "grad_norm": 1.4641155004501343,
+      "learning_rate": 4.675675675675676e-05,
+      "loss": 1.6466,
+      "step": 659
+    },
+    {
+      "epoch": 0.0008627417147383736,
+      "grad_norm": 1.797682762145996,
+      "learning_rate": 4.675175175175176e-05,
+      "loss": 1.2984,
+      "step": 660
+    },
+    {
+      "epoch": 0.0008640488991546438,
+      "grad_norm": 1.5468018054962158,
+      "learning_rate": 4.674674674674675e-05,
+      "loss": 2.0747,
+      "step": 661
+    },
+    {
+      "epoch": 0.0008653560835709141,
+      "grad_norm": 1.3826464414596558,
+      "learning_rate": 4.674174174174174e-05,
+      "loss": 1.7024,
+      "step": 662
+    },
+    {
+      "epoch": 0.0008666632679871844,
+      "grad_norm": 1.3884285688400269,
+      "learning_rate": 4.673673673673674e-05,
+      "loss": 1.7245,
+      "step": 663
+    },
+    {
+      "epoch": 0.0008679704524034547,
+      "grad_norm": 1.3933097124099731,
+      "learning_rate": 4.673173173173173e-05,
+      "loss": 2.0303,
+      "step": 664
+    },
+    {
+      "epoch": 0.0008692776368197249,
+      "grad_norm": 1.8334778547286987,
+      "learning_rate": 4.672672672672673e-05,
+      "loss": 2.0857,
+      "step": 665
+    },
+    {
+      "epoch": 0.0008705848212359951,
+      "grad_norm": 2.003793954849243,
+      "learning_rate": 4.672172172172172e-05,
+      "loss": 1.7374,
+      "step": 666
+    },
+    {
+      "epoch": 0.0008718920056522654,
+      "grad_norm": 1.4170873165130615,
+      "learning_rate": 4.671671671671672e-05,
+      "loss": 1.4093,
+      "step": 667
+    },
+    {
+      "epoch": 0.0008731991900685356,
+      "grad_norm": 1.9772340059280396,
+      "learning_rate": 4.6711711711711715e-05,
+      "loss": 2.6609,
+      "step": 668
+    },
+    {
+      "epoch": 0.000874506374484806,
+      "grad_norm": 1.5182138681411743,
+      "learning_rate": 4.670670670670671e-05,
+      "loss": 2.0109,
+      "step": 669
+    },
+    {
+      "epoch": 0.0008758135589010762,
+      "grad_norm": 1.2754665613174438,
+      "learning_rate": 4.670170170170171e-05,
+      "loss": 1.6225,
+      "step": 670
+    },
+    {
+      "epoch": 0.0008771207433173465,
+      "grad_norm": 1.5271803140640259,
+      "learning_rate": 4.66966966966967e-05,
+      "loss": 2.2835,
+      "step": 671
+    },
+    {
+      "epoch": 0.0008784279277336167,
+      "grad_norm": 1.3149797916412354,
+      "learning_rate": 4.669169169169169e-05,
+      "loss": 1.3654,
+      "step": 672
+    },
+    {
+      "epoch": 0.000879735112149887,
+      "grad_norm": 1.993091344833374,
+      "learning_rate": 4.6686686686686686e-05,
+      "loss": 2.0773,
+      "step": 673
+    },
+    {
+      "epoch": 0.0008810422965661572,
+      "grad_norm": 1.2878508567810059,
+      "learning_rate": 4.6681681681681686e-05,
+      "loss": 0.8774,
+      "step": 674
+    },
+    {
+      "epoch": 0.0008823494809824275,
+      "grad_norm": 1.3715171813964844,
+      "learning_rate": 4.667667667667668e-05,
+      "loss": 2.1742,
+      "step": 675
+    },
+    {
+      "epoch": 0.0008836566653986978,
+      "grad_norm": 2.5644302368164062,
+      "learning_rate": 4.667167167167167e-05,
+      "loss": 1.4282,
+      "step": 676
+    },
+    {
+      "epoch": 0.0008849638498149681,
+      "grad_norm": 1.311880111694336,
+      "learning_rate": 4.666666666666667e-05,
+      "loss": 1.7326,
+      "step": 677
+    },
+    {
+      "epoch": 0.0008862710342312383,
+      "grad_norm": 1.475303053855896,
+      "learning_rate": 4.6661661661661665e-05,
+      "loss": 1.3059,
+      "step": 678
+    },
+    {
+      "epoch": 0.0008875782186475086,
+      "grad_norm": 1.7172300815582275,
+      "learning_rate": 4.6656656656656664e-05,
+      "loss": 2.2081,
+      "step": 679
+    },
+    {
+      "epoch": 0.0008888854030637788,
+      "grad_norm": 1.8925265073776245,
+      "learning_rate": 4.665165165165166e-05,
+      "loss": 1.7082,
+      "step": 680
+    },
+    {
+      "epoch": 0.0008901925874800491,
+      "grad_norm": 1.5461525917053223,
+      "learning_rate": 4.6646646646646644e-05,
+      "loss": 2.4286,
+      "step": 681
+    },
+    {
+      "epoch": 0.0008914997718963193,
+      "grad_norm": 1.493786334991455,
+      "learning_rate": 4.664164164164164e-05,
+      "loss": 2.3482,
+      "step": 682
+    },
+    {
+      "epoch": 0.0008928069563125897,
+      "grad_norm": 1.0161991119384766,
+      "learning_rate": 4.6636636636636636e-05,
+      "loss": 0.6091,
+      "step": 683
+    },
+    {
+      "epoch": 0.0008941141407288599,
+      "grad_norm": 1.4707592725753784,
+      "learning_rate": 4.6631631631631636e-05,
+      "loss": 1.5476,
+      "step": 684
+    },
+    {
+      "epoch": 0.0008954213251451302,
+      "grad_norm": 1.3240089416503906,
+      "learning_rate": 4.662662662662663e-05,
+      "loss": 1.0464,
+      "step": 685
+    },
+    {
+      "epoch": 0.0008967285095614004,
+      "grad_norm": 1.9543989896774292,
+      "learning_rate": 4.662162162162162e-05,
+      "loss": 1.5915,
+      "step": 686
+    },
+    {
+      "epoch": 0.0008980356939776707,
+      "grad_norm": 1.274321436882019,
+      "learning_rate": 4.661661661661662e-05,
+      "loss": 1.8471,
+      "step": 687
+    },
+    {
+      "epoch": 0.0008993428783939409,
+      "grad_norm": 2.8131563663482666,
+      "learning_rate": 4.6611611611611615e-05,
+      "loss": 1.566,
+      "step": 688
+    },
+    {
+      "epoch": 0.0009006500628102112,
+      "grad_norm": 1.5721018314361572,
+      "learning_rate": 4.660660660660661e-05,
+      "loss": 1.1862,
+      "step": 689
+    },
+    {
+      "epoch": 0.0009019572472264815,
+      "grad_norm": 1.794669270515442,
+      "learning_rate": 4.66016016016016e-05,
+      "loss": 2.433,
+      "step": 690
+    },
+    {
+      "epoch": 0.0009032644316427517,
+      "grad_norm": 1.5412126779556274,
+      "learning_rate": 4.65965965965966e-05,
+      "loss": 1.5977,
+      "step": 691
+    },
+    {
+      "epoch": 0.000904571616059022,
+      "grad_norm": 1.315383791923523,
+      "learning_rate": 4.659159159159159e-05,
+      "loss": 0.9201,
+      "step": 692
+    },
+    {
+      "epoch": 0.0009058788004752922,
+      "grad_norm": 1.2493948936462402,
+      "learning_rate": 4.6586586586586586e-05,
+      "loss": 0.6013,
+      "step": 693
+    },
+    {
+      "epoch": 0.0009071859848915625,
+      "grad_norm": 1.4124208688735962,
+      "learning_rate": 4.6581581581581586e-05,
+      "loss": 2.0033,
+      "step": 694
+    },
+    {
+      "epoch": 0.0009084931693078327,
+      "grad_norm": 1.4190680980682373,
+      "learning_rate": 4.657657657657658e-05,
+      "loss": 1.7679,
+      "step": 695
+    },
+    {
+      "epoch": 0.0009098003537241031,
+      "grad_norm": 1.634192943572998,
+      "learning_rate": 4.657157157157158e-05,
+      "loss": 1.6089,
+      "step": 696
+    },
+    {
+      "epoch": 0.0009111075381403733,
+      "grad_norm": 1.8498886823654175,
+      "learning_rate": 4.6566566566566565e-05,
+      "loss": 1.8977,
+      "step": 697
+    },
+    {
+      "epoch": 0.0009124147225566436,
+      "grad_norm": 1.5007154941558838,
+      "learning_rate": 4.6561561561561565e-05,
+      "loss": 1.867,
+      "step": 698
+    },
+    {
+      "epoch": 0.0009137219069729138,
+      "grad_norm": 2.1808488368988037,
+      "learning_rate": 4.655655655655656e-05,
+      "loss": 2.2829,
+      "step": 699
+    },
+    {
+      "epoch": 0.0009150290913891841,
+      "grad_norm": 1.2694264650344849,
+      "learning_rate": 4.655155155155155e-05,
+      "loss": 0.9172,
+      "step": 700
+    },
+    {
+      "epoch": 0.0009163362758054543,
+      "grad_norm": 1.8196114301681519,
+      "learning_rate": 4.654654654654655e-05,
+      "loss": 1.4365,
+      "step": 701
+    },
+    {
+      "epoch": 0.0009176434602217247,
+      "grad_norm": 1.6108999252319336,
+      "learning_rate": 4.6541541541541543e-05,
+      "loss": 1.9053,
+      "step": 702
+    },
+    {
+      "epoch": 0.0009189506446379949,
+      "grad_norm": 1.2370092868804932,
+      "learning_rate": 4.653653653653654e-05,
+      "loss": 1.349,
+      "step": 703
+    },
+    {
+      "epoch": 0.0009202578290542652,
+      "grad_norm": 1.524497389793396,
+      "learning_rate": 4.6531531531531536e-05,
+      "loss": 1.7978,
+      "step": 704
+    },
+    {
+      "epoch": 0.0009215650134705354,
+      "grad_norm": 1.5805730819702148,
+      "learning_rate": 4.652652652652653e-05,
+      "loss": 1.6615,
+      "step": 705
+    },
+    {
+      "epoch": 0.0009228721978868057,
+      "grad_norm": 2.2942686080932617,
+      "learning_rate": 4.652152152152152e-05,
+      "loss": 2.1471,
+      "step": 706
+    },
+    {
+      "epoch": 0.0009241793823030759,
+      "grad_norm": 1.6951627731323242,
+      "learning_rate": 4.6516516516516515e-05,
+      "loss": 1.2362,
+      "step": 707
+    },
+    {
+      "epoch": 0.0009254865667193463,
+      "grad_norm": 1.9619264602661133,
+      "learning_rate": 4.6511511511511515e-05,
+      "loss": 1.5742,
+      "step": 708
+    },
+    {
+      "epoch": 0.0009267937511356165,
+      "grad_norm": 1.3662768602371216,
+      "learning_rate": 4.650650650650651e-05,
+      "loss": 1.1128,
+      "step": 709
+    },
+    {
+      "epoch": 0.0009281009355518867,
+      "grad_norm": 2.4068853855133057,
+      "learning_rate": 4.650150150150151e-05,
+      "loss": 2.2087,
+      "step": 710
+    },
+    {
+      "epoch": 0.000929408119968157,
+      "grad_norm": 2.019045114517212,
+      "learning_rate": 4.64964964964965e-05,
+      "loss": 1.7745,
+      "step": 711
+    },
+    {
+      "epoch": 0.0009307153043844272,
+      "grad_norm": 2.091775894165039,
+      "learning_rate": 4.6491491491491493e-05,
+      "loss": 1.7824,
+      "step": 712
+    },
+    {
+      "epoch": 0.0009320224888006975,
+      "grad_norm": 1.729496717453003,
+      "learning_rate": 4.648648648648649e-05,
+      "loss": 1.2614,
+      "step": 713
+    },
+    {
+      "epoch": 0.0009333296732169677,
+      "grad_norm": 1.2630985975265503,
+      "learning_rate": 4.648148148148148e-05,
+      "loss": 0.9162,
+      "step": 714
+    },
+    {
+      "epoch": 0.0009346368576332381,
+      "grad_norm": 1.1649032831192017,
+      "learning_rate": 4.647647647647648e-05,
+      "loss": 0.8089,
+      "step": 715
+    },
+    {
+      "epoch": 0.0009359440420495083,
+      "grad_norm": 1.670494556427002,
+      "learning_rate": 4.647147147147147e-05,
+      "loss": 1.6403,
+      "step": 716
+    },
+    {
+      "epoch": 0.0009372512264657786,
+      "grad_norm": 1.7459062337875366,
+      "learning_rate": 4.6466466466466465e-05,
+      "loss": 1.861,
+      "step": 717
+    },
+    {
+      "epoch": 0.0009385584108820488,
+      "grad_norm": 1.4365828037261963,
+      "learning_rate": 4.6461461461461465e-05,
+      "loss": 1.8736,
+      "step": 718
+    },
+    {
+      "epoch": 0.0009398655952983191,
+      "grad_norm": 1.5231200456619263,
+      "learning_rate": 4.645645645645646e-05,
+      "loss": 1.2991,
+      "step": 719
+    },
+    {
+      "epoch": 0.0009411727797145893,
+      "grad_norm": 1.3555145263671875,
+      "learning_rate": 4.645145145145146e-05,
+      "loss": 1.7721,
+      "step": 720
+    },
+    {
+      "epoch": 0.0009424799641308596,
+      "grad_norm": 1.4573489427566528,
+      "learning_rate": 4.644644644644645e-05,
+      "loss": 2.0622,
+      "step": 721
+    },
+    {
+      "epoch": 0.0009437871485471299,
+      "grad_norm": 1.3339567184448242,
+      "learning_rate": 4.6441441441441444e-05,
+      "loss": 1.8605,
+      "step": 722
+    },
+    {
+      "epoch": 0.0009450943329634002,
+      "grad_norm": 1.3285406827926636,
+      "learning_rate": 4.6436436436436437e-05,
+      "loss": 1.3524,
+      "step": 723
+    },
+    {
+      "epoch": 0.0009464015173796704,
+      "grad_norm": 1.576004147529602,
+      "learning_rate": 4.643143143143143e-05,
+      "loss": 1.5146,
+      "step": 724
+    },
+    {
+      "epoch": 0.0009477087017959407,
+      "grad_norm": 1.5992568731307983,
+      "learning_rate": 4.642642642642643e-05,
+      "loss": 2.0314,
+      "step": 725
+    },
+    {
+      "epoch": 0.0009490158862122109,
+      "grad_norm": 1.8096318244934082,
+      "learning_rate": 4.642142142142142e-05,
+      "loss": 1.0039,
+      "step": 726
+    },
+    {
+      "epoch": 0.0009503230706284812,
+      "grad_norm": 1.9867362976074219,
+      "learning_rate": 4.641641641641642e-05,
+      "loss": 2.0532,
+      "step": 727
+    },
+    {
+      "epoch": 0.0009516302550447515,
+      "grad_norm": 1.5593302249908447,
+      "learning_rate": 4.6411411411411415e-05,
+      "loss": 1.6883,
+      "step": 728
+    },
+    {
+      "epoch": 0.0009529374394610218,
+      "grad_norm": 2.027670383453369,
+      "learning_rate": 4.640640640640641e-05,
+      "loss": 1.2035,
+      "step": 729
+    },
+    {
+      "epoch": 0.000954244623877292,
+      "grad_norm": 1.6325725317001343,
+      "learning_rate": 4.64014014014014e-05,
+      "loss": 2.7577,
+      "step": 730
+    },
+    {
+      "epoch": 0.0009555518082935622,
+      "grad_norm": 1.5115816593170166,
+      "learning_rate": 4.6396396396396394e-05,
+      "loss": 2.1885,
+      "step": 731
+    },
+    {
+      "epoch": 0.0009568589927098325,
+      "grad_norm": 1.320870041847229,
+      "learning_rate": 4.6391391391391394e-05,
+      "loss": 1.2766,
+      "step": 732
+    },
+    {
+      "epoch": 0.0009581661771261027,
+      "grad_norm": 1.5859371423721313,
+      "learning_rate": 4.6386386386386387e-05,
+      "loss": 2.3353,
+      "step": 733
+    },
+    {
+      "epoch": 0.000959473361542373,
+      "grad_norm": 1.6014713048934937,
+      "learning_rate": 4.6381381381381386e-05,
+      "loss": 1.0897,
+      "step": 734
+    },
+    {
+      "epoch": 0.0009607805459586433,
+      "grad_norm": 1.3904573917388916,
+      "learning_rate": 4.637637637637638e-05,
+      "loss": 1.7004,
+      "step": 735
+    },
+    {
+      "epoch": 0.0009620877303749136,
+      "grad_norm": 1.5303046703338623,
+      "learning_rate": 4.637137137137137e-05,
+      "loss": 1.7498,
+      "step": 736
+    },
+    {
+      "epoch": 0.0009633949147911838,
+      "grad_norm": 1.551106572151184,
+      "learning_rate": 4.636636636636637e-05,
+      "loss": 2.5456,
+      "step": 737
+    },
+    {
+      "epoch": 0.0009647020992074541,
+      "grad_norm": 1.358823299407959,
+      "learning_rate": 4.6361361361361365e-05,
+      "loss": 1.1721,
+      "step": 738
+    },
+    {
+      "epoch": 0.0009660092836237243,
+      "grad_norm": 1.291875958442688,
+      "learning_rate": 4.635635635635636e-05,
+      "loss": 1.0625,
+      "step": 739
+    },
+    {
+      "epoch": 0.0009673164680399946,
+      "grad_norm": 1.935599684715271,
+      "learning_rate": 4.635135135135135e-05,
+      "loss": 2.526,
+      "step": 740
+    },
+    {
+      "epoch": 0.0009686236524562649,
+      "grad_norm": 1.629394292831421,
+      "learning_rate": 4.634634634634635e-05,
+      "loss": 1.7169,
+      "step": 741
+    },
+    {
+      "epoch": 0.0009699308368725352,
+      "grad_norm": 1.0493532419204712,
+      "learning_rate": 4.6341341341341344e-05,
+      "loss": 0.9184,
+      "step": 742
+    },
+    {
+      "epoch": 0.0009712380212888054,
+      "grad_norm": 1.8731212615966797,
+      "learning_rate": 4.633633633633634e-05,
+      "loss": 2.643,
+      "step": 743
+    },
+    {
+      "epoch": 0.0009725452057050757,
+      "grad_norm": 2.1282262802124023,
+      "learning_rate": 4.6331331331331336e-05,
+      "loss": 1.7461,
+      "step": 744
+    },
+    {
+      "epoch": 0.0009738523901213459,
+      "grad_norm": 1.4210848808288574,
+      "learning_rate": 4.632632632632633e-05,
+      "loss": 1.3042,
+      "step": 745
+    },
+    {
+      "epoch": 0.0009751595745376162,
+      "grad_norm": 1.7453669309616089,
+      "learning_rate": 4.632132132132133e-05,
+      "loss": 1.6472,
+      "step": 746
+    },
+    {
+      "epoch": 0.0009764667589538864,
+      "grad_norm": 1.2782055139541626,
+      "learning_rate": 4.6316316316316315e-05,
+      "loss": 1.3502,
+      "step": 747
+    },
+    {
+      "epoch": 0.0009777739433701567,
+      "grad_norm": 1.480959177017212,
+      "learning_rate": 4.6311311311311315e-05,
+      "loss": 1.5736,
+      "step": 748
+    },
+    {
+      "epoch": 0.000979081127786427,
+      "grad_norm": 1.492809772491455,
+      "learning_rate": 4.630630630630631e-05,
+      "loss": 2.0535,
+      "step": 749
+    },
+    {
+      "epoch": 0.0009803883122026973,
+      "grad_norm": 1.6125658750534058,
+      "learning_rate": 4.63013013013013e-05,
+      "loss": 1.6602,
+      "step": 750
+    },
+    {
+      "epoch": 0.0009816954966189676,
+      "grad_norm": 1.6419305801391602,
+      "learning_rate": 4.62962962962963e-05,
+      "loss": 1.3645,
+      "step": 751
+    },
+    {
+      "epoch": 0.0009830026810352377,
+      "grad_norm": 1.4012327194213867,
+      "learning_rate": 4.6291291291291294e-05,
+      "loss": 2.0916,
+      "step": 752
+    },
+    {
+      "epoch": 0.000984309865451508,
+      "grad_norm": 1.2971864938735962,
+      "learning_rate": 4.628628628628629e-05,
+      "loss": 1.4935,
+      "step": 753
+    },
+    {
+      "epoch": 0.0009856170498677784,
+      "grad_norm": 1.4981552362442017,
+      "learning_rate": 4.6281281281281286e-05,
+      "loss": 0.5875,
+      "step": 754
+    },
+    {
+      "epoch": 0.0009869242342840485,
+      "grad_norm": 2.1142072677612305,
+      "learning_rate": 4.627627627627627e-05,
+      "loss": 1.7751,
+      "step": 755
+    },
+    {
+      "epoch": 0.0009882314187003188,
+      "grad_norm": 0.7426319718360901,
+      "learning_rate": 4.627127127127127e-05,
+      "loss": 0.066,
+      "step": 756
+    },
+    {
+      "epoch": 0.000989538603116589,
+      "grad_norm": 1.6408412456512451,
+      "learning_rate": 4.6266266266266265e-05,
+      "loss": 2.5212,
+      "step": 757
+    },
+    {
+      "epoch": 0.0009908457875328594,
+      "grad_norm": 1.2834370136260986,
+      "learning_rate": 4.6261261261261265e-05,
+      "loss": 1.1589,
+      "step": 758
+    },
+    {
+      "epoch": 0.0009921529719491295,
+      "grad_norm": 1.4316962957382202,
+      "learning_rate": 4.625625625625626e-05,
+      "loss": 1.7252,
+      "step": 759
+    },
+    {
+      "epoch": 0.0009934601563653998,
+      "grad_norm": 1.2498915195465088,
+      "learning_rate": 4.625125125125125e-05,
+      "loss": 0.5978,
+      "step": 760
+    },
+    {
+      "epoch": 0.0009947673407816702,
+      "grad_norm": 2.1427066326141357,
+      "learning_rate": 4.624624624624625e-05,
+      "loss": 1.4767,
+      "step": 761
+    },
+    {
+      "epoch": 0.0009960745251979405,
+      "grad_norm": 1.8533881902694702,
+      "learning_rate": 4.6241241241241244e-05,
+      "loss": 0.9712,
+      "step": 762
+    },
+    {
+      "epoch": 0.0009973817096142106,
+      "grad_norm": 1.6610878705978394,
+      "learning_rate": 4.6236236236236244e-05,
+      "loss": 1.7298,
+      "step": 763
+    },
+    {
+      "epoch": 0.000998688894030481,
+      "grad_norm": 1.3472763299942017,
+      "learning_rate": 4.623123123123123e-05,
+      "loss": 2.3774,
+      "step": 764
+    },
+    {
+      "epoch": 0.0009999960784467512,
+      "grad_norm": 1.3953591585159302,
+      "learning_rate": 4.622622622622623e-05,
+      "loss": 1.8534,
+      "step": 765
+    },
+    {
+      "epoch": 0.0010013032628630215,
+      "grad_norm": 2.0223636627197266,
+      "learning_rate": 4.622122122122122e-05,
+      "loss": 2.241,
+      "step": 766
+    },
+    {
+      "epoch": 0.0010026104472792916,
+      "grad_norm": 1.3348544836044312,
+      "learning_rate": 4.6216216216216215e-05,
+      "loss": 0.9346,
+      "step": 767
+    },
+    {
+      "epoch": 0.001003917631695562,
+      "grad_norm": 1.5678869485855103,
+      "learning_rate": 4.6211211211211215e-05,
+      "loss": 1.8014,
+      "step": 768
+    },
+    {
+      "epoch": 0.0010052248161118323,
+      "grad_norm": 1.4671971797943115,
+      "learning_rate": 4.620620620620621e-05,
+      "loss": 1.6975,
+      "step": 769
+    },
+    {
+      "epoch": 0.0010065320005281026,
+      "grad_norm": 1.505881428718567,
+      "learning_rate": 4.620120120120121e-05,
+      "loss": 2.1595,
+      "step": 770
+    },
+    {
+      "epoch": 0.0010078391849443727,
+      "grad_norm": 1.7970147132873535,
+      "learning_rate": 4.61961961961962e-05,
+      "loss": 2.2557,
+      "step": 771
+    },
+    {
+      "epoch": 0.001009146369360643,
+      "grad_norm": 1.4663082361221313,
+      "learning_rate": 4.6191191191191194e-05,
+      "loss": 2.043,
+      "step": 772
+    },
+    {
+      "epoch": 0.0010104535537769133,
+      "grad_norm": 1.5121746063232422,
+      "learning_rate": 4.618618618618619e-05,
+      "loss": 1.6631,
+      "step": 773
+    },
+    {
+      "epoch": 0.0010117607381931834,
+      "grad_norm": 1.2969845533370972,
+      "learning_rate": 4.618118118118118e-05,
+      "loss": 1.5017,
+      "step": 774
+    },
+    {
+      "epoch": 0.0010130679226094538,
+      "grad_norm": 2.499478578567505,
+      "learning_rate": 4.617617617617618e-05,
+      "loss": 1.6606,
+      "step": 775
+    },
+    {
+      "epoch": 0.001014375107025724,
+      "grad_norm": 1.222113847732544,
+      "learning_rate": 4.617117117117117e-05,
+      "loss": 1.4345,
+      "step": 776
+    },
+    {
+      "epoch": 0.0010156822914419944,
+      "grad_norm": 1.8465626239776611,
+      "learning_rate": 4.616616616616617e-05,
+      "loss": 1.5905,
+      "step": 777
+    },
+    {
+      "epoch": 0.0010169894758582645,
+      "grad_norm": 1.0977576971054077,
+      "learning_rate": 4.6161161161161165e-05,
+      "loss": 1.4128,
+      "step": 778
+    },
+    {
+      "epoch": 0.0010182966602745348,
+      "grad_norm": 2.1433565616607666,
+      "learning_rate": 4.615615615615616e-05,
+      "loss": 1.9955,
+      "step": 779
+    },
+    {
+      "epoch": 0.0010196038446908052,
+      "grad_norm": 1.3921386003494263,
+      "learning_rate": 4.615115115115115e-05,
+      "loss": 1.5392,
+      "step": 780
+    },
+    {
+      "epoch": 0.0010209110291070755,
+      "grad_norm": 0.6536403298377991,
+      "learning_rate": 4.6146146146146144e-05,
+      "loss": 0.3118,
+      "step": 781
+    },
+    {
+      "epoch": 0.0010222182135233456,
+      "grad_norm": 1.5995150804519653,
+      "learning_rate": 4.6141141141141144e-05,
+      "loss": 1.4457,
+      "step": 782
+    },
+    {
+      "epoch": 0.001023525397939616,
+      "grad_norm": 1.3978142738342285,
+      "learning_rate": 4.613613613613614e-05,
+      "loss": 1.7488,
+      "step": 783
+    },
+    {
+      "epoch": 0.0010248325823558862,
+      "grad_norm": 1.577843189239502,
+      "learning_rate": 4.613113113113114e-05,
+      "loss": 1.0493,
+      "step": 784
+    },
+    {
+      "epoch": 0.0010261397667721565,
+      "grad_norm": 1.591992735862732,
+      "learning_rate": 4.612612612612613e-05,
+      "loss": 0.6235,
+      "step": 785
+    },
+    {
+      "epoch": 0.0010274469511884266,
+      "grad_norm": 1.9197914600372314,
+      "learning_rate": 4.612112112112112e-05,
+      "loss": 1.9289,
+      "step": 786
+    },
+    {
+      "epoch": 0.001028754135604697,
+      "grad_norm": 1.4725290536880493,
+      "learning_rate": 4.611611611611612e-05,
+      "loss": 1.054,
+      "step": 787
+    },
+    {
+      "epoch": 0.0010300613200209673,
+      "grad_norm": 1.9627829790115356,
+      "learning_rate": 4.6111111111111115e-05,
+      "loss": 1.7368,
+      "step": 788
+    },
+    {
+      "epoch": 0.0010313685044372376,
+      "grad_norm": 2.8953444957733154,
+      "learning_rate": 4.610610610610611e-05,
+      "loss": 2.1914,
+      "step": 789
+    },
+    {
+      "epoch": 0.0010326756888535077,
+      "grad_norm": 1.4363845586776733,
+      "learning_rate": 4.61011011011011e-05,
+      "loss": 1.5758,
+      "step": 790
+    },
+    {
+      "epoch": 0.001033982873269778,
+      "grad_norm": 1.8652585744857788,
+      "learning_rate": 4.6096096096096094e-05,
+      "loss": 1.9152,
+      "step": 791
+    },
+    {
+      "epoch": 0.0010352900576860483,
+      "grad_norm": 1.5264694690704346,
+      "learning_rate": 4.6091091091091094e-05,
+      "loss": 2.2221,
+      "step": 792
+    },
+    {
+      "epoch": 0.0010365972421023187,
+      "grad_norm": 1.1599723100662231,
+      "learning_rate": 4.608608608608609e-05,
+      "loss": 0.6811,
+      "step": 793
+    },
+    {
+      "epoch": 0.0010379044265185888,
+      "grad_norm": 1.674228310585022,
+      "learning_rate": 4.608108108108109e-05,
+      "loss": 2.2283,
+      "step": 794
+    },
+    {
+      "epoch": 0.001039211610934859,
+      "grad_norm": 2.127528429031372,
+      "learning_rate": 4.607607607607608e-05,
+      "loss": 1.9444,
+      "step": 795
+    },
+    {
+      "epoch": 0.0010405187953511294,
+      "grad_norm": 1.7764638662338257,
+      "learning_rate": 4.607107107107107e-05,
+      "loss": 2.1071,
+      "step": 796
+    },
+    {
+      "epoch": 0.0010418259797673995,
+      "grad_norm": 1.5508782863616943,
+      "learning_rate": 4.6066066066066066e-05,
+      "loss": 2.154,
+      "step": 797
+    },
+    {
+      "epoch": 0.0010431331641836698,
+      "grad_norm": 1.594293475151062,
+      "learning_rate": 4.606106106106106e-05,
+      "loss": 1.5922,
+      "step": 798
+    },
+    {
+      "epoch": 0.0010444403485999401,
+      "grad_norm": 1.8112678527832031,
+      "learning_rate": 4.605605605605606e-05,
+      "loss": 1.9487,
+      "step": 799
+    },
+    {
+      "epoch": 0.0010457475330162105,
+      "grad_norm": 1.5911704301834106,
+      "learning_rate": 4.605105105105105e-05,
+      "loss": 1.9454,
+      "step": 800
+    },
+    {
+      "epoch": 0.0010470547174324806,
+      "grad_norm": 2.074603796005249,
+      "learning_rate": 4.604604604604605e-05,
+      "loss": 2.069,
+      "step": 801
+    },
+    {
+      "epoch": 0.0010483619018487509,
+      "grad_norm": 1.3887349367141724,
+      "learning_rate": 4.6041041041041044e-05,
+      "loss": 1.0027,
+      "step": 802
+    },
+    {
+      "epoch": 0.0010496690862650212,
+      "grad_norm": 2.159118413925171,
+      "learning_rate": 4.603603603603604e-05,
+      "loss": 1.9919,
+      "step": 803
+    },
+    {
+      "epoch": 0.0010509762706812915,
+      "grad_norm": 0.8145937919616699,
+      "learning_rate": 4.603103103103104e-05,
+      "loss": 0.5479,
+      "step": 804
+    },
+    {
+      "epoch": 0.0010522834550975616,
+      "grad_norm": 1.0970402956008911,
+      "learning_rate": 4.602602602602602e-05,
+      "loss": 0.7958,
+      "step": 805
+    },
+    {
+      "epoch": 0.001053590639513832,
+      "grad_norm": 1.6436102390289307,
+      "learning_rate": 4.602102102102102e-05,
+      "loss": 2.2761,
+      "step": 806
+    },
+    {
+      "epoch": 0.0010548978239301023,
+      "grad_norm": 2.3100688457489014,
+      "learning_rate": 4.6016016016016016e-05,
+      "loss": 1.1421,
+      "step": 807
+    },
+    {
+      "epoch": 0.0010562050083463726,
+      "grad_norm": 1.1512593030929565,
+      "learning_rate": 4.6011011011011015e-05,
+      "loss": 0.9207,
+      "step": 808
+    },
+    {
+      "epoch": 0.0010575121927626427,
+      "grad_norm": 1.6368423700332642,
+      "learning_rate": 4.600600600600601e-05,
+      "loss": 2.2944,
+      "step": 809
+    },
+    {
+      "epoch": 0.001058819377178913,
+      "grad_norm": 1.241062879562378,
+      "learning_rate": 4.6001001001001e-05,
+      "loss": 1.0867,
+      "step": 810
+    },
+    {
+      "epoch": 0.0010601265615951833,
+      "grad_norm": 1.1022549867630005,
+      "learning_rate": 4.5995995995996e-05,
+      "loss": 0.8227,
+      "step": 811
+    },
+    {
+      "epoch": 0.0010614337460114536,
+      "grad_norm": 1.346975564956665,
+      "learning_rate": 4.5990990990990994e-05,
+      "loss": 1.571,
+      "step": 812
+    },
+    {
+      "epoch": 0.0010627409304277238,
+      "grad_norm": 1.4832590818405151,
+      "learning_rate": 4.5985985985985994e-05,
+      "loss": 1.5572,
+      "step": 813
+    },
+    {
+      "epoch": 0.001064048114843994,
+      "grad_norm": 1.3186007738113403,
+      "learning_rate": 4.598098098098098e-05,
+      "loss": 1.2784,
+      "step": 814
+    },
+    {
+      "epoch": 0.0010653552992602644,
+      "grad_norm": 1.409213900566101,
+      "learning_rate": 4.597597597597598e-05,
+      "loss": 1.7077,
+      "step": 815
+    },
+    {
+      "epoch": 0.0010666624836765347,
+      "grad_norm": 1.5419780015945435,
+      "learning_rate": 4.597097097097097e-05,
+      "loss": 1.3623,
+      "step": 816
+    },
+    {
+      "epoch": 0.0010679696680928048,
+      "grad_norm": 1.8298070430755615,
+      "learning_rate": 4.5965965965965966e-05,
+      "loss": 1.7875,
+      "step": 817
+    },
+    {
+      "epoch": 0.0010692768525090751,
+      "grad_norm": 1.176162838935852,
+      "learning_rate": 4.5960960960960966e-05,
+      "loss": 0.5726,
+      "step": 818
+    },
+    {
+      "epoch": 0.0010705840369253455,
+      "grad_norm": 1.8125083446502686,
+      "learning_rate": 4.595595595595596e-05,
+      "loss": 1.5066,
+      "step": 819
+    },
+    {
+      "epoch": 0.0010718912213416156,
+      "grad_norm": 1.5410789251327515,
+      "learning_rate": 4.595095095095096e-05,
+      "loss": 1.0769,
+      "step": 820
+    },
+    {
+      "epoch": 0.0010731984057578859,
+      "grad_norm": 1.330946683883667,
+      "learning_rate": 4.594594594594595e-05,
+      "loss": 0.9746,
+      "step": 821
+    },
+    {
+      "epoch": 0.0010745055901741562,
+      "grad_norm": 2.7522494792938232,
+      "learning_rate": 4.594094094094094e-05,
+      "loss": 1.4613,
+      "step": 822
+    },
+    {
+      "epoch": 0.0010758127745904265,
+      "grad_norm": 1.2615082263946533,
+      "learning_rate": 4.593593593593594e-05,
+      "loss": 1.0243,
+      "step": 823
+    },
+    {
+      "epoch": 0.0010771199590066966,
+      "grad_norm": 1.3613532781600952,
+      "learning_rate": 4.593093093093093e-05,
+      "loss": 1.3285,
+      "step": 824
+    },
+    {
+      "epoch": 0.001078427143422967,
+      "grad_norm": 1.9715328216552734,
+      "learning_rate": 4.592592592592593e-05,
+      "loss": 2.257,
+      "step": 825
+    },
+    {
+      "epoch": 0.0010797343278392373,
+      "grad_norm": 1.4096521139144897,
+      "learning_rate": 4.592092092092092e-05,
+      "loss": 1.1046,
+      "step": 826
+    },
+    {
+      "epoch": 0.0010810415122555076,
+      "grad_norm": 1.42073655128479,
+      "learning_rate": 4.5915915915915916e-05,
+      "loss": 1.6635,
+      "step": 827
+    },
+    {
+      "epoch": 0.0010823486966717777,
+      "grad_norm": 1.504815936088562,
+      "learning_rate": 4.5910910910910916e-05,
+      "loss": 1.4442,
+      "step": 828
+    },
+    {
+      "epoch": 0.001083655881088048,
+      "grad_norm": 1.7664979696273804,
+      "learning_rate": 4.590590590590591e-05,
+      "loss": 1.8573,
+      "step": 829
+    },
+    {
+      "epoch": 0.0010849630655043183,
+      "grad_norm": 1.8008695840835571,
+      "learning_rate": 4.59009009009009e-05,
+      "loss": 1.7079,
+      "step": 830
+    },
+    {
+      "epoch": 0.0010862702499205886,
+      "grad_norm": 1.1379114389419556,
+      "learning_rate": 4.5895895895895895e-05,
+      "loss": 0.9911,
+      "step": 831
+    },
+    {
+      "epoch": 0.0010875774343368587,
+      "grad_norm": 2.099579334259033,
+      "learning_rate": 4.5890890890890894e-05,
+      "loss": 1.4611,
+      "step": 832
+    },
+    {
+      "epoch": 0.001088884618753129,
+      "grad_norm": 0.23350459337234497,
+      "learning_rate": 4.588588588588589e-05,
+      "loss": 0.0053,
+      "step": 833
+    },
+    {
+      "epoch": 0.0010901918031693994,
+      "grad_norm": 1.6057443618774414,
+      "learning_rate": 4.588088088088088e-05,
+      "loss": 2.4532,
+      "step": 834
+    },
+    {
+      "epoch": 0.0010914989875856697,
+      "grad_norm": 1.6137644052505493,
+      "learning_rate": 4.587587587587588e-05,
+      "loss": 1.2097,
+      "step": 835
+    },
+    {
+      "epoch": 0.0010928061720019398,
+      "grad_norm": 1.387888789176941,
+      "learning_rate": 4.587087087087087e-05,
+      "loss": 1.6962,
+      "step": 836
+    },
+    {
+      "epoch": 0.0010941133564182101,
+      "grad_norm": 1.8844033479690552,
+      "learning_rate": 4.586586586586587e-05,
+      "loss": 2.1858,
+      "step": 837
+    },
+    {
+      "epoch": 0.0010954205408344804,
+      "grad_norm": 1.295705795288086,
+      "learning_rate": 4.586086086086086e-05,
+      "loss": 0.9426,
+      "step": 838
+    },
+    {
+      "epoch": 0.0010967277252507505,
+      "grad_norm": 1.432578206062317,
+      "learning_rate": 4.585585585585586e-05,
+      "loss": 1.9353,
+      "step": 839
+    },
+    {
+      "epoch": 0.0010980349096670209,
+      "grad_norm": 0.761978805065155,
+      "learning_rate": 4.585085085085085e-05,
+      "loss": 0.6169,
+      "step": 840
+    },
+    {
+      "epoch": 0.0010993420940832912,
+      "grad_norm": 1.270096778869629,
+      "learning_rate": 4.5845845845845845e-05,
+      "loss": 1.1971,
+      "step": 841
+    },
+    {
+      "epoch": 0.0011006492784995615,
+      "grad_norm": 1.4082361459732056,
+      "learning_rate": 4.5840840840840844e-05,
+      "loss": 2.1865,
+      "step": 842
+    },
+    {
+      "epoch": 0.0011019564629158316,
+      "grad_norm": 1.844780445098877,
+      "learning_rate": 4.583583583583584e-05,
+      "loss": 1.7291,
+      "step": 843
+    },
+    {
+      "epoch": 0.001103263647332102,
+      "grad_norm": 1.5229413509368896,
+      "learning_rate": 4.583083083083084e-05,
+      "loss": 1.7612,
+      "step": 844
+    },
+    {
+      "epoch": 0.0011045708317483722,
+      "grad_norm": 1.5335806608200073,
+      "learning_rate": 4.582582582582583e-05,
+      "loss": 0.9754,
+      "step": 845
+    },
+    {
+      "epoch": 0.0011058780161646426,
+      "grad_norm": 1.658276915550232,
+      "learning_rate": 4.582082082082082e-05,
+      "loss": 2.3253,
+      "step": 846
+    },
+    {
+      "epoch": 0.0011071852005809127,
+      "grad_norm": 2.3184142112731934,
+      "learning_rate": 4.5815815815815816e-05,
+      "loss": 1.7134,
+      "step": 847
+    },
+    {
+      "epoch": 0.001108492384997183,
+      "grad_norm": 1.7002702951431274,
+      "learning_rate": 4.581081081081081e-05,
+      "loss": 1.6376,
+      "step": 848
+    },
+    {
+      "epoch": 0.0011097995694134533,
+      "grad_norm": 1.4328159093856812,
+      "learning_rate": 4.580580580580581e-05,
+      "loss": 1.9185,
+      "step": 849
+    },
+    {
+      "epoch": 0.0011111067538297236,
+      "grad_norm": 1.6288138628005981,
+      "learning_rate": 4.58008008008008e-05,
+      "loss": 1.1558,
+      "step": 850
+    },
+    {
+      "epoch": 0.0011124139382459937,
+      "grad_norm": 3.06666898727417,
+      "learning_rate": 4.57957957957958e-05,
+      "loss": 1.8392,
+      "step": 851
+    },
+    {
+      "epoch": 0.001113721122662264,
+      "grad_norm": 1.3637505769729614,
+      "learning_rate": 4.5790790790790794e-05,
+      "loss": 1.224,
+      "step": 852
+    },
+    {
+      "epoch": 0.0011150283070785344,
+      "grad_norm": 1.414533257484436,
+      "learning_rate": 4.578578578578579e-05,
+      "loss": 1.4846,
+      "step": 853
+    },
+    {
+      "epoch": 0.0011163354914948047,
+      "grad_norm": 1.2439321279525757,
+      "learning_rate": 4.578078078078079e-05,
+      "loss": 1.4256,
+      "step": 854
+    },
+    {
+      "epoch": 0.0011176426759110748,
+      "grad_norm": 1.6864304542541504,
+      "learning_rate": 4.577577577577577e-05,
+      "loss": 1.9729,
+      "step": 855
+    },
+    {
+      "epoch": 0.0011189498603273451,
+      "grad_norm": 1.7443264722824097,
+      "learning_rate": 4.577077077077077e-05,
+      "loss": 1.8151,
+      "step": 856
+    },
+    {
+      "epoch": 0.0011202570447436154,
+      "grad_norm": 1.180742859840393,
+      "learning_rate": 4.5765765765765766e-05,
+      "loss": 1.3495,
+      "step": 857
+    },
+    {
+      "epoch": 0.0011215642291598858,
+      "grad_norm": 1.9463708400726318,
+      "learning_rate": 4.5760760760760766e-05,
+      "loss": 1.4778,
+      "step": 858
+    },
+    {
+      "epoch": 0.0011228714135761559,
+      "grad_norm": 1.5906401872634888,
+      "learning_rate": 4.575575575575576e-05,
+      "loss": 2.6368,
+      "step": 859
+    },
+    {
+      "epoch": 0.0011241785979924262,
+      "grad_norm": 1.2574598789215088,
+      "learning_rate": 4.575075075075075e-05,
+      "loss": 1.5985,
+      "step": 860
+    },
+    {
+      "epoch": 0.0011254857824086965,
+      "grad_norm": 2.0222814083099365,
+      "learning_rate": 4.574574574574575e-05,
+      "loss": 1.9934,
+      "step": 861
+    },
+    {
+      "epoch": 0.0011267929668249666,
+      "grad_norm": 1.55674147605896,
+      "learning_rate": 4.5740740740740745e-05,
+      "loss": 2.02,
+      "step": 862
+    },
+    {
+      "epoch": 0.001128100151241237,
+      "grad_norm": 1.769354224205017,
+      "learning_rate": 4.573573573573574e-05,
+      "loss": 2.18,
+      "step": 863
+    },
+    {
+      "epoch": 0.0011294073356575072,
+      "grad_norm": 1.098128080368042,
+      "learning_rate": 4.573073073073073e-05,
+      "loss": 0.9966,
+      "step": 864
+    },
+    {
+      "epoch": 0.0011307145200737776,
+      "grad_norm": 2.306091547012329,
+      "learning_rate": 4.5725725725725723e-05,
+      "loss": 1.622,
+      "step": 865
+    },
+    {
+      "epoch": 0.0011320217044900477,
+      "grad_norm": 1.7915018796920776,
+      "learning_rate": 4.572072072072072e-05,
+      "loss": 1.5752,
+      "step": 866
+    },
+    {
+      "epoch": 0.001133328888906318,
+      "grad_norm": 1.7172670364379883,
+      "learning_rate": 4.5715715715715716e-05,
+      "loss": 2.084,
+      "step": 867
+    },
+    {
+      "epoch": 0.0011346360733225883,
+      "grad_norm": 1.7435013055801392,
+      "learning_rate": 4.5710710710710716e-05,
+      "loss": 2.0701,
+      "step": 868
+    },
+    {
+      "epoch": 0.0011359432577388586,
+      "grad_norm": 1.3407816886901855,
+      "learning_rate": 4.570570570570571e-05,
+      "loss": 1.4594,
+      "step": 869
+    },
+    {
+      "epoch": 0.0011372504421551287,
+      "grad_norm": 1.6299376487731934,
+      "learning_rate": 4.57007007007007e-05,
+      "loss": 2.0859,
+      "step": 870
+    },
+    {
+      "epoch": 0.001138557626571399,
+      "grad_norm": 1.8910582065582275,
+      "learning_rate": 4.56956956956957e-05,
+      "loss": 1.6506,
+      "step": 871
+    },
+    {
+      "epoch": 0.0011398648109876694,
+      "grad_norm": 1.5794777870178223,
+      "learning_rate": 4.569069069069069e-05,
+      "loss": 1.7067,
+      "step": 872
+    },
+    {
+      "epoch": 0.0011411719954039397,
+      "grad_norm": 1.2254141569137573,
+      "learning_rate": 4.568568568568569e-05,
+      "loss": 0.9155,
+      "step": 873
+    },
+    {
+      "epoch": 0.0011424791798202098,
+      "grad_norm": 1.513360619544983,
+      "learning_rate": 4.568068068068068e-05,
+      "loss": 1.144,
+      "step": 874
+    },
+    {
+      "epoch": 0.00114378636423648,
+      "grad_norm": 1.1067694425582886,
+      "learning_rate": 4.567567567567568e-05,
+      "loss": 0.9639,
+      "step": 875
+    },
+    {
+      "epoch": 0.0011450935486527504,
+      "grad_norm": 1.3811447620391846,
+      "learning_rate": 4.567067067067067e-05,
+      "loss": 1.5599,
+      "step": 876
+    },
+    {
+      "epoch": 0.0011464007330690207,
+      "grad_norm": 1.2767783403396606,
+      "learning_rate": 4.5665665665665666e-05,
+      "loss": 0.6873,
+      "step": 877
+    },
+    {
+      "epoch": 0.0011477079174852908,
+      "grad_norm": 1.5702688694000244,
+      "learning_rate": 4.5660660660660666e-05,
+      "loss": 1.4588,
+      "step": 878
+    },
+    {
+      "epoch": 0.0011490151019015612,
+      "grad_norm": 1.3665337562561035,
+      "learning_rate": 4.565565565565566e-05,
+      "loss": 1.1774,
+      "step": 879
+    },
+    {
+      "epoch": 0.0011503222863178315,
+      "grad_norm": 1.4107836484909058,
+      "learning_rate": 4.565065065065065e-05,
+      "loss": 1.8697,
+      "step": 880
+    },
+    {
+      "epoch": 0.0011516294707341016,
+      "grad_norm": 1.9390654563903809,
+      "learning_rate": 4.5645645645645645e-05,
+      "loss": 1.4225,
+      "step": 881
+    },
+    {
+      "epoch": 0.001152936655150372,
+      "grad_norm": 1.0538872480392456,
+      "learning_rate": 4.5640640640640645e-05,
+      "loss": 0.3713,
+      "step": 882
+    },
+    {
+      "epoch": 0.0011542438395666422,
+      "grad_norm": 1.4992613792419434,
+      "learning_rate": 4.563563563563564e-05,
+      "loss": 1.206,
+      "step": 883
+    },
+    {
+      "epoch": 0.0011555510239829125,
+      "grad_norm": 1.3779773712158203,
+      "learning_rate": 4.563063063063063e-05,
+      "loss": 1.4261,
+      "step": 884
+    },
+    {
+      "epoch": 0.0011568582083991827,
+      "grad_norm": 1.8651567697525024,
+      "learning_rate": 4.562562562562563e-05,
+      "loss": 2.0045,
+      "step": 885
+    },
+    {
+      "epoch": 0.001158165392815453,
+      "grad_norm": 1.466630458831787,
+      "learning_rate": 4.562062062062062e-05,
+      "loss": 0.8037,
+      "step": 886
+    },
+    {
+      "epoch": 0.0011594725772317233,
+      "grad_norm": 1.6154310703277588,
+      "learning_rate": 4.561561561561562e-05,
+      "loss": 1.6574,
+      "step": 887
+    },
+    {
+      "epoch": 0.0011607797616479936,
+      "grad_norm": 1.209167718887329,
+      "learning_rate": 4.561061061061061e-05,
+      "loss": 0.8947,
+      "step": 888
+    },
+    {
+      "epoch": 0.0011620869460642637,
+      "grad_norm": 2.297548532485962,
+      "learning_rate": 4.560560560560561e-05,
+      "loss": 2.1663,
+      "step": 889
+    },
+    {
+      "epoch": 0.001163394130480534,
+      "grad_norm": 1.4887466430664062,
+      "learning_rate": 4.56006006006006e-05,
+      "loss": 1.4262,
+      "step": 890
+    },
+    {
+      "epoch": 0.0011647013148968044,
+      "grad_norm": 1.2562835216522217,
+      "learning_rate": 4.5595595595595595e-05,
+      "loss": 0.7872,
+      "step": 891
+    },
+    {
+      "epoch": 0.0011660084993130747,
+      "grad_norm": 1.5932337045669556,
+      "learning_rate": 4.5590590590590595e-05,
+      "loss": 1.8181,
+      "step": 892
+    },
+    {
+      "epoch": 0.0011673156837293448,
+      "grad_norm": 1.4483977556228638,
+      "learning_rate": 4.558558558558559e-05,
+      "loss": 0.9179,
+      "step": 893
+    },
+    {
+      "epoch": 0.001168622868145615,
+      "grad_norm": 1.6455111503601074,
+      "learning_rate": 4.558058058058059e-05,
+      "loss": 1.3642,
+      "step": 894
+    },
+    {
+      "epoch": 0.0011699300525618854,
+      "grad_norm": 1.7909139394760132,
+      "learning_rate": 4.557557557557558e-05,
+      "loss": 1.1367,
+      "step": 895
+    },
+    {
+      "epoch": 0.0011712372369781557,
+      "grad_norm": 1.8691556453704834,
+      "learning_rate": 4.5570570570570573e-05,
+      "loss": 1.9421,
+      "step": 896
+    },
+    {
+      "epoch": 0.0011725444213944258,
+      "grad_norm": 1.2419389486312866,
+      "learning_rate": 4.5565565565565566e-05,
+      "loss": 1.0851,
+      "step": 897
+    },
+    {
+      "epoch": 0.0011738516058106962,
+      "grad_norm": 1.4257506132125854,
+      "learning_rate": 4.556056056056056e-05,
+      "loss": 1.8685,
+      "step": 898
+    },
+    {
+      "epoch": 0.0011751587902269665,
+      "grad_norm": 1.3899787664413452,
+      "learning_rate": 4.555555555555556e-05,
+      "loss": 2.0175,
+      "step": 899
+    },
+    {
+      "epoch": 0.0011764659746432368,
+      "grad_norm": NaN,
+      "learning_rate": 4.555055055055055e-05,
+      "loss": 1.6648,
+      "step": 900
+    },
+    {
+      "epoch": 0.001177773159059507,
+      "grad_norm": 1.8874595165252686,
+      "learning_rate": 4.555055055055055e-05,
+      "loss": 2.4731,
+      "step": 901
+    },
+    {
+      "epoch": 0.0011790803434757772,
+      "grad_norm": 1.0181536674499512,
+      "learning_rate": 4.5545545545545545e-05,
+      "loss": 0.5193,
+      "step": 902
+    },
+    {
+      "epoch": 0.0011803875278920475,
+      "grad_norm": 1.5093921422958374,
+      "learning_rate": 4.5540540540540545e-05,
+      "loss": 1.1692,
+      "step": 903
+    },
+    {
+      "epoch": 0.0011816947123083176,
+      "grad_norm": 0.9943222999572754,
+      "learning_rate": 4.553553553553554e-05,
+      "loss": 0.4475,
+      "step": 904
+    },
+    {
+      "epoch": 0.001183001896724588,
+      "grad_norm": 1.2361242771148682,
+      "learning_rate": 4.553053053053054e-05,
+      "loss": 1.1539,
+      "step": 905
+    },
+    {
+      "epoch": 0.0011843090811408583,
+      "grad_norm": 1.9103820323944092,
+      "learning_rate": 4.5525525525525524e-05,
+      "loss": 2.2337,
+      "step": 906
+    },
+    {
+      "epoch": 0.0011856162655571286,
+      "grad_norm": 1.7128448486328125,
+      "learning_rate": 4.5520520520520523e-05,
+      "loss": 1.144,
+      "step": 907
+    },
+    {
+      "epoch": 0.0011869234499733987,
+      "grad_norm": 1.8571282625198364,
+      "learning_rate": 4.5515515515515516e-05,
+      "loss": 2.4864,
+      "step": 908
+    },
+    {
+      "epoch": 0.001188230634389669,
+      "grad_norm": 1.3939807415008545,
+      "learning_rate": 4.551051051051051e-05,
+      "loss": 1.7691,
+      "step": 909
+    },
+    {
+      "epoch": 0.0011895378188059393,
+      "grad_norm": 1.6099313497543335,
+      "learning_rate": 4.550550550550551e-05,
+      "loss": 1.4867,
+      "step": 910
+    },
+    {
+      "epoch": 0.0011908450032222097,
+      "grad_norm": 1.3069841861724854,
+      "learning_rate": 4.55005005005005e-05,
+      "loss": 1.5203,
+      "step": 911
+    },
+    {
+      "epoch": 0.0011921521876384798,
+      "grad_norm": 1.7743209600448608,
+      "learning_rate": 4.54954954954955e-05,
+      "loss": 2.202,
+      "step": 912
+    },
+    {
+      "epoch": 0.00119345937205475,
+      "grad_norm": 1.7642059326171875,
+      "learning_rate": 4.5490490490490495e-05,
+      "loss": 2.4379,
+      "step": 913
+    },
+    {
+      "epoch": 0.0011947665564710204,
+      "grad_norm": 2.95877742767334,
+      "learning_rate": 4.548548548548549e-05,
+      "loss": 1.8692,
+      "step": 914
+    },
+    {
+      "epoch": 0.0011960737408872907,
+      "grad_norm": 1.491555094718933,
+      "learning_rate": 4.548048048048048e-05,
+      "loss": 2.3894,
+      "step": 915
+    },
+    {
+      "epoch": 0.0011973809253035608,
+      "grad_norm": 1.8221832513809204,
+      "learning_rate": 4.5475475475475474e-05,
+      "loss": 2.4876,
+      "step": 916
+    },
+    {
+      "epoch": 0.0011986881097198311,
+      "grad_norm": 1.264717936515808,
+      "learning_rate": 4.5470470470470474e-05,
+      "loss": 1.5999,
+      "step": 917
+    },
+    {
+      "epoch": 0.0011999952941361015,
+      "grad_norm": 1.808779239654541,
+      "learning_rate": 4.5465465465465467e-05,
+      "loss": 1.9447,
+      "step": 918
+    },
+    {
+      "epoch": 0.0012013024785523718,
+      "grad_norm": 1.5189424753189087,
+      "learning_rate": 4.5460460460460466e-05,
+      "loss": 1.4584,
+      "step": 919
+    },
+    {
+      "epoch": 0.0012026096629686419,
+      "grad_norm": 1.3586452007293701,
+      "learning_rate": 4.545545545545546e-05,
+      "loss": 1.7887,
+      "step": 920
+    },
+    {
+      "epoch": 0.0012039168473849122,
+      "grad_norm": 1.398910641670227,
+      "learning_rate": 4.545045045045045e-05,
+      "loss": 1.1718,
+      "step": 921
+    },
+    {
+      "epoch": 0.0012052240318011825,
+      "grad_norm": 1.2478675842285156,
+      "learning_rate": 4.544544544544545e-05,
+      "loss": 1.2514,
+      "step": 922
+    },
+    {
+      "epoch": 0.0012065312162174528,
+      "grad_norm": 2.1135025024414062,
+      "learning_rate": 4.544044044044044e-05,
+      "loss": 1.8296,
+      "step": 923
+    },
+    {
+      "epoch": 0.001207838400633723,
+      "grad_norm": 1.5757348537445068,
+      "learning_rate": 4.543543543543544e-05,
+      "loss": 3.0286,
+      "step": 924
+    },
+    {
+      "epoch": 0.0012091455850499933,
+      "grad_norm": 1.4365657567977905,
+      "learning_rate": 4.543043043043043e-05,
+      "loss": 2.1078,
+      "step": 925
+    },
+    {
+      "epoch": 0.0012104527694662636,
+      "grad_norm": 1.9949811697006226,
+      "learning_rate": 4.542542542542543e-05,
+      "loss": 2.1051,
+      "step": 926
+    },
+    {
+      "epoch": 0.0012117599538825337,
+      "grad_norm": 1.3817561864852905,
+      "learning_rate": 4.5420420420420424e-05,
+      "loss": 1.1749,
+      "step": 927
+    },
+    {
+      "epoch": 0.001213067138298804,
+      "grad_norm": 1.6927233934402466,
+      "learning_rate": 4.5415415415415417e-05,
+      "loss": 1.9164,
+      "step": 928
+    },
+    {
+      "epoch": 0.0012143743227150743,
+      "grad_norm": 1.7584826946258545,
+      "learning_rate": 4.5410410410410416e-05,
+      "loss": 2.3891,
+      "step": 929
+    },
+    {
+      "epoch": 0.0012156815071313447,
+      "grad_norm": 1.7426416873931885,
+      "learning_rate": 4.540540540540541e-05,
+      "loss": 1.3118,
+      "step": 930
+    },
+    {
+      "epoch": 0.0012169886915476148,
+      "grad_norm": 1.632703423500061,
+      "learning_rate": 4.54004004004004e-05,
+      "loss": 1.6597,
+      "step": 931
+    },
+    {
+      "epoch": 0.001218295875963885,
+      "grad_norm": 1.4729312658309937,
+      "learning_rate": 4.5395395395395395e-05,
+      "loss": 1.4667,
+      "step": 932
+    },
+    {
+      "epoch": 0.0012196030603801554,
+      "grad_norm": 1.668464183807373,
+      "learning_rate": 4.539039039039039e-05,
+      "loss": 1.926,
+      "step": 933
+    },
+    {
+      "epoch": 0.0012209102447964257,
+      "grad_norm": 1.5949184894561768,
+      "learning_rate": 4.538538538538539e-05,
+      "loss": 1.6382,
+      "step": 934
+    },
+    {
+      "epoch": 0.0012222174292126958,
+      "grad_norm": 1.3693821430206299,
+      "learning_rate": 4.538038038038038e-05,
+      "loss": 1.576,
+      "step": 935
+    },
+    {
+      "epoch": 0.0012235246136289661,
+      "grad_norm": 1.9348069429397583,
+      "learning_rate": 4.537537537537538e-05,
+      "loss": 1.7365,
+      "step": 936
+    },
+    {
+      "epoch": 0.0012248317980452365,
+      "grad_norm": 1.4702868461608887,
+      "learning_rate": 4.5370370370370374e-05,
+      "loss": 2.0769,
+      "step": 937
+    },
+    {
+      "epoch": 0.0012261389824615068,
+      "grad_norm": 1.4736220836639404,
+      "learning_rate": 4.536536536536537e-05,
+      "loss": 1.2919,
+      "step": 938
+    },
+    {
+      "epoch": 0.0012274461668777769,
+      "grad_norm": 1.3041695356369019,
+      "learning_rate": 4.536036036036036e-05,
+      "loss": 2.0047,
+      "step": 939
+    },
+    {
+      "epoch": 0.0012287533512940472,
+      "grad_norm": 1.5080018043518066,
+      "learning_rate": 4.535535535535535e-05,
+      "loss": 0.8442,
+      "step": 940
+    },
+    {
+      "epoch": 0.0012300605357103175,
+      "grad_norm": 2.1177821159362793,
+      "learning_rate": 4.535035035035035e-05,
+      "loss": 1.4438,
+      "step": 941
+    },
+    {
+      "epoch": 0.0012313677201265878,
+      "grad_norm": 1.5005906820297241,
+      "learning_rate": 4.5345345345345345e-05,
+      "loss": 1.6536,
+      "step": 942
+    },
+    {
+      "epoch": 0.001232674904542858,
+      "grad_norm": 1.5183066129684448,
+      "learning_rate": 4.5340340340340345e-05,
+      "loss": 1.8566,
+      "step": 943
+    },
+    {
+      "epoch": 0.0012339820889591283,
+      "grad_norm": 1.8110268115997314,
+      "learning_rate": 4.533533533533534e-05,
+      "loss": 1.615,
+      "step": 944
+    },
+    {
+      "epoch": 0.0012352892733753986,
+      "grad_norm": 1.2820264101028442,
+      "learning_rate": 4.533033033033033e-05,
+      "loss": 1.2077,
+      "step": 945
+    },
+    {
+      "epoch": 0.0012365964577916687,
+      "grad_norm": 1.9272654056549072,
+      "learning_rate": 4.532532532532533e-05,
+      "loss": 2.1356,
+      "step": 946
+    },
+    {
+      "epoch": 0.001237903642207939,
+      "grad_norm": 1.486853837966919,
+      "learning_rate": 4.5320320320320324e-05,
+      "loss": 1.5343,
+      "step": 947
+    },
+    {
+      "epoch": 0.0012392108266242093,
+      "grad_norm": 1.2197450399398804,
+      "learning_rate": 4.531531531531532e-05,
+      "loss": 1.2302,
+      "step": 948
+    },
+    {
+      "epoch": 0.0012405180110404796,
+      "grad_norm": 1.7098912000656128,
+      "learning_rate": 4.531031031031031e-05,
+      "loss": 1.3453,
+      "step": 949
+    },
+    {
+      "epoch": 0.0012418251954567497,
+      "grad_norm": 1.7141612768173218,
+      "learning_rate": 4.530530530530531e-05,
+      "loss": 2.36,
+      "step": 950
+    },
+    {
+      "epoch": 0.00124313237987302,
+      "grad_norm": 1.648511290550232,
+      "learning_rate": 4.53003003003003e-05,
+      "loss": 1.6291,
+      "step": 951
+    },
+    {
+      "epoch": 0.0012444395642892904,
+      "grad_norm": 1.829443335533142,
+      "learning_rate": 4.5295295295295295e-05,
+      "loss": 2.2654,
+      "step": 952
+    },
+    {
+      "epoch": 0.0012457467487055607,
+      "grad_norm": 1.3765976428985596,
+      "learning_rate": 4.5290290290290295e-05,
+      "loss": 2.1441,
+      "step": 953
+    },
+    {
+      "epoch": 0.0012470539331218308,
+      "grad_norm": 1.6352201700210571,
+      "learning_rate": 4.528528528528529e-05,
+      "loss": 2.1715,
+      "step": 954
+    },
+    {
+      "epoch": 0.0012483611175381011,
+      "grad_norm": 1.5562580823898315,
+      "learning_rate": 4.528028028028029e-05,
+      "loss": 2.1396,
+      "step": 955
+    },
+    {
+      "epoch": 0.0012496683019543714,
+      "grad_norm": 1.5098286867141724,
+      "learning_rate": 4.5275275275275274e-05,
+      "loss": 1.7206,
+      "step": 956
+    },
+    {
+      "epoch": 0.0012509754863706418,
+      "grad_norm": 1.6444472074508667,
+      "learning_rate": 4.5270270270270274e-05,
+      "loss": 1.7656,
+      "step": 957
+    },
+    {
+      "epoch": 0.0012522826707869119,
+      "grad_norm": 2.4799416065216064,
+      "learning_rate": 4.526526526526527e-05,
+      "loss": 2.1838,
+      "step": 958
+    },
+    {
+      "epoch": 0.0012535898552031822,
+      "grad_norm": 1.590050220489502,
+      "learning_rate": 4.526026026026026e-05,
+      "loss": 1.4106,
+      "step": 959
+    },
+    {
+      "epoch": 0.0012548970396194525,
+      "grad_norm": 1.8405051231384277,
+      "learning_rate": 4.525525525525526e-05,
+      "loss": 2.2631,
+      "step": 960
+    },
+    {
+      "epoch": 0.0012562042240357228,
+      "grad_norm": 1.5832786560058594,
+      "learning_rate": 4.525025025025025e-05,
+      "loss": 1.3631,
+      "step": 961
+    },
+    {
+      "epoch": 0.001257511408451993,
+      "grad_norm": 1.4532434940338135,
+      "learning_rate": 4.524524524524525e-05,
+      "loss": 0.9313,
+      "step": 962
+    },
+    {
+      "epoch": 0.0012588185928682633,
+      "grad_norm": 1.4057062864303589,
+      "learning_rate": 4.5240240240240245e-05,
+      "loss": 1.4056,
+      "step": 963
+    },
+    {
+      "epoch": 0.0012601257772845336,
+      "grad_norm": 2.0283358097076416,
+      "learning_rate": 4.523523523523524e-05,
+      "loss": 1.8143,
+      "step": 964
+    },
+    {
+      "epoch": 0.001261432961700804,
+      "grad_norm": 1.1053264141082764,
+      "learning_rate": 4.523023023023023e-05,
+      "loss": 0.83,
+      "step": 965
+    },
+    {
+      "epoch": 0.001262740146117074,
+      "grad_norm": 1.7534080743789673,
+      "learning_rate": 4.5225225225225224e-05,
+      "loss": 1.7179,
+      "step": 966
+    },
+    {
+      "epoch": 0.0012640473305333443,
+      "grad_norm": 0.884555459022522,
+      "learning_rate": 4.5220220220220224e-05,
+      "loss": 0.3851,
+      "step": 967
+    },
+    {
+      "epoch": 0.0012653545149496146,
+      "grad_norm": 1.3067853450775146,
+      "learning_rate": 4.521521521521522e-05,
+      "loss": 1.5448,
+      "step": 968
+    },
+    {
+      "epoch": 0.0012666616993658847,
+      "grad_norm": 1.5509170293807983,
+      "learning_rate": 4.521021021021021e-05,
+      "loss": 1.9306,
+      "step": 969
+    },
+    {
+      "epoch": 0.001267968883782155,
+      "grad_norm": 1.7076313495635986,
+      "learning_rate": 4.520520520520521e-05,
+      "loss": 1.5645,
+      "step": 970
+    },
+    {
+      "epoch": 0.0012692760681984254,
+      "grad_norm": 1.488572359085083,
+      "learning_rate": 4.52002002002002e-05,
+      "loss": 1.8993,
+      "step": 971
+    },
+    {
+      "epoch": 0.0012705832526146957,
+      "grad_norm": 1.1463112831115723,
+      "learning_rate": 4.5195195195195196e-05,
+      "loss": 1.0441,
+      "step": 972
+    },
+    {
+      "epoch": 0.0012718904370309658,
+      "grad_norm": 1.3509596586227417,
+      "learning_rate": 4.519019019019019e-05,
+      "loss": 1.0284,
+      "step": 973
+    },
+    {
+      "epoch": 0.0012731976214472361,
+      "grad_norm": 1.3112398386001587,
+      "learning_rate": 4.518518518518519e-05,
+      "loss": 1.7263,
+      "step": 974
+    },
+    {
+      "epoch": 0.0012745048058635064,
+      "grad_norm": 1.512604832649231,
+      "learning_rate": 4.518018018018018e-05,
+      "loss": 1.8321,
+      "step": 975
+    },
+    {
+      "epoch": 0.0012758119902797768,
+      "grad_norm": 1.3571206331253052,
+      "learning_rate": 4.5175175175175174e-05,
+      "loss": 0.9064,
+      "step": 976
+    },
+    {
+      "epoch": 0.0012771191746960469,
+      "grad_norm": 1.6946697235107422,
+      "learning_rate": 4.5170170170170174e-05,
+      "loss": 1.3531,
+      "step": 977
+    },
+    {
+      "epoch": 0.0012784263591123172,
+      "grad_norm": 1.585479974746704,
+      "learning_rate": 4.516516516516517e-05,
+      "loss": 2.3439,
+      "step": 978
+    },
+    {
+      "epoch": 0.0012797335435285875,
+      "grad_norm": 1.1922357082366943,
+      "learning_rate": 4.516016016016017e-05,
+      "loss": 1.212,
+      "step": 979
+    },
+    {
+      "epoch": 0.0012810407279448578,
+      "grad_norm": 1.596510648727417,
+      "learning_rate": 4.515515515515516e-05,
+      "loss": 2.1316,
+      "step": 980
+    },
+    {
+      "epoch": 0.001282347912361128,
+      "grad_norm": 1.4152441024780273,
+      "learning_rate": 4.515015015015015e-05,
+      "loss": 2.1548,
+      "step": 981
+    },
+    {
+      "epoch": 0.0012836550967773982,
+      "grad_norm": 1.8281924724578857,
+      "learning_rate": 4.5145145145145146e-05,
+      "loss": 1.3201,
+      "step": 982
+    },
+    {
+      "epoch": 0.0012849622811936686,
+      "grad_norm": 1.3299903869628906,
+      "learning_rate": 4.514014014014014e-05,
+      "loss": 1.3265,
+      "step": 983
+    },
+    {
+      "epoch": 0.0012862694656099389,
+      "grad_norm": 1.4934943914413452,
+      "learning_rate": 4.513513513513514e-05,
+      "loss": 1.8198,
+      "step": 984
+    },
+    {
+      "epoch": 0.001287576650026209,
+      "grad_norm": 1.5048881769180298,
+      "learning_rate": 4.513013013013013e-05,
+      "loss": 0.6756,
+      "step": 985
+    },
+    {
+      "epoch": 0.0012888838344424793,
+      "grad_norm": 1.2431844472885132,
+      "learning_rate": 4.512512512512513e-05,
+      "loss": 1.072,
+      "step": 986
+    },
+    {
+      "epoch": 0.0012901910188587496,
+      "grad_norm": 1.5030936002731323,
+      "learning_rate": 4.5120120120120124e-05,
+      "loss": 1.803,
+      "step": 987
+    },
+    {
+      "epoch": 0.0012914982032750197,
+      "grad_norm": 1.7279001474380493,
+      "learning_rate": 4.511511511511512e-05,
+      "loss": 1.9023,
+      "step": 988
+    },
+    {
+      "epoch": 0.00129280538769129,
+      "grad_norm": 1.3523824214935303,
+      "learning_rate": 4.511011011011011e-05,
+      "loss": 2.3966,
+      "step": 989
+    },
+    {
+      "epoch": 0.0012941125721075604,
+      "grad_norm": 1.2615395784378052,
+      "learning_rate": 4.51051051051051e-05,
+      "loss": 0.8247,
+      "step": 990
+    },
+    {
+      "epoch": 0.0012954197565238307,
+      "grad_norm": 1.778281331062317,
+      "learning_rate": 4.51001001001001e-05,
+      "loss": 0.7731,
+      "step": 991
+    },
+    {
+      "epoch": 0.0012967269409401008,
+      "grad_norm": 2.3872108459472656,
+      "learning_rate": 4.5095095095095096e-05,
+      "loss": 0.952,
+      "step": 992
+    },
+    {
+      "epoch": 0.0012980341253563711,
+      "grad_norm": 1.3104918003082275,
+      "learning_rate": 4.5090090090090095e-05,
+      "loss": 1.2728,
+      "step": 993
+    },
+    {
+      "epoch": 0.0012993413097726414,
+      "grad_norm": 1.4704560041427612,
+      "learning_rate": 4.508508508508509e-05,
+      "loss": 2.7315,
+      "step": 994
+    },
+    {
+      "epoch": 0.0013006484941889117,
+      "grad_norm": 0.8565270304679871,
+      "learning_rate": 4.508008008008008e-05,
+      "loss": 0.4964,
+      "step": 995
+    },
+    {
+      "epoch": 0.0013019556786051819,
+      "grad_norm": 1.0428869724273682,
+      "learning_rate": 4.507507507507508e-05,
+      "loss": 0.6046,
+      "step": 996
+    },
+    {
+      "epoch": 0.0013032628630214522,
+      "grad_norm": 0.8731210231781006,
+      "learning_rate": 4.507007007007007e-05,
+      "loss": 0.3445,
+      "step": 997
+    },
+    {
+      "epoch": 0.0013045700474377225,
+      "grad_norm": 1.6904356479644775,
+      "learning_rate": 4.506506506506507e-05,
+      "loss": 1.8073,
+      "step": 998
+    },
+    {
+      "epoch": 0.0013058772318539928,
+      "grad_norm": 1.7629987001419067,
+      "learning_rate": 4.506006006006006e-05,
+      "loss": 1.1306,
+      "step": 999
+    },
+    {
+      "epoch": 0.001307184416270263,
+      "grad_norm": 2.73661732673645,
+      "learning_rate": 4.505505505505506e-05,
+      "loss": 1.7229,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0013084916006865332,
+      "grad_norm": 2.375129222869873,
+      "learning_rate": 4.505005005005005e-05,
+      "loss": 1.9766,
+      "step": 1001
+    },
+    {
+      "epoch": 0.0013097987851028036,
+      "grad_norm": 1.7902029752731323,
+      "learning_rate": 4.5045045045045046e-05,
+      "loss": 2.1272,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0013111059695190739,
+      "grad_norm": 1.4387460947036743,
+      "learning_rate": 4.5040040040040045e-05,
+      "loss": 1.3835,
+      "step": 1003
+    },
+    {
+      "epoch": 0.001312413153935344,
+      "grad_norm": 1.4268690347671509,
+      "learning_rate": 4.503503503503504e-05,
+      "loss": 1.5135,
+      "step": 1004
+    },
+    {
+      "epoch": 0.0013137203383516143,
+      "grad_norm": 1.4620217084884644,
+      "learning_rate": 4.503003003003004e-05,
+      "loss": 1.6779,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0013150275227678846,
+      "grad_norm": 1.7664227485656738,
+      "learning_rate": 4.5025025025025024e-05,
+      "loss": 1.621,
+      "step": 1006
+    },
+    {
+      "epoch": 0.001316334707184155,
+      "grad_norm": 1.4966270923614502,
+      "learning_rate": 4.502002002002002e-05,
+      "loss": 1.4523,
+      "step": 1007
+    },
+    {
+      "epoch": 0.001317641891600425,
+      "grad_norm": 1.4574472904205322,
+      "learning_rate": 4.501501501501502e-05,
+      "loss": 1.5881,
+      "step": 1008
+    },
+    {
+      "epoch": 0.0013189490760166954,
+      "grad_norm": 1.4857311248779297,
+      "learning_rate": 4.501001001001001e-05,
+      "loss": 1.1323,
+      "step": 1009
+    },
+    {
+      "epoch": 0.0013202562604329657,
+      "grad_norm": 1.4847455024719238,
+      "learning_rate": 4.500500500500501e-05,
+      "loss": 1.9606,
+      "step": 1010
+    },
+    {
+      "epoch": 0.0013215634448492358,
+      "grad_norm": 1.5814106464385986,
+      "learning_rate": 4.5e-05,
+      "loss": 1.0063,
+      "step": 1011
+    },
+    {
+      "epoch": 0.001322870629265506,
+      "grad_norm": 1.7457741498947144,
+      "learning_rate": 4.4994994994994996e-05,
+      "loss": 1.9615,
+      "step": 1012
+    },
+    {
+      "epoch": 0.0013241778136817764,
+      "grad_norm": 1.552299976348877,
+      "learning_rate": 4.4989989989989996e-05,
+      "loss": 1.6046,
+      "step": 1013
+    },
+    {
+      "epoch": 0.0013254849980980467,
+      "grad_norm": 1.4323080778121948,
+      "learning_rate": 4.498498498498498e-05,
+      "loss": 1.448,
+      "step": 1014
+    },
+    {
+      "epoch": 0.0013267921825143168,
+      "grad_norm": 1.2698408365249634,
+      "learning_rate": 4.497997997997998e-05,
+      "loss": 1.347,
+      "step": 1015
+    },
+    {
+      "epoch": 0.0013280993669305872,
+      "grad_norm": 1.4110513925552368,
+      "learning_rate": 4.4974974974974974e-05,
+      "loss": 1.4287,
+      "step": 1016
+    },
+    {
+      "epoch": 0.0013294065513468575,
+      "grad_norm": 1.4845863580703735,
+      "learning_rate": 4.4969969969969974e-05,
+      "loss": 1.4334,
+      "step": 1017
+    },
+    {
+      "epoch": 0.0013307137357631278,
+      "grad_norm": 1.2936795949935913,
+      "learning_rate": 4.496496496496497e-05,
+      "loss": 1.0426,
+      "step": 1018
+    },
+    {
+      "epoch": 0.001332020920179398,
+      "grad_norm": 1.7912980318069458,
+      "learning_rate": 4.495995995995996e-05,
+      "loss": 1.2248,
+      "step": 1019
+    },
+    {
+      "epoch": 0.0013333281045956682,
+      "grad_norm": 1.6292836666107178,
+      "learning_rate": 4.495495495495496e-05,
+      "loss": 2.5021,
+      "step": 1020
+    },
+    {
+      "epoch": 0.0013346352890119385,
+      "grad_norm": 1.6026805639266968,
+      "learning_rate": 4.494994994994995e-05,
+      "loss": 2.2754,
+      "step": 1021
+    },
+    {
+      "epoch": 0.0013359424734282089,
+      "grad_norm": 1.2606528997421265,
+      "learning_rate": 4.4944944944944946e-05,
+      "loss": 1.4579,
+      "step": 1022
+    },
+    {
+      "epoch": 0.001337249657844479,
+      "grad_norm": 1.5761170387268066,
+      "learning_rate": 4.493993993993994e-05,
+      "loss": 1.6152,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0013385568422607493,
+      "grad_norm": 1.8901695013046265,
+      "learning_rate": 4.493493493493494e-05,
+      "loss": 2.1698,
+      "step": 1024
+    },
+    {
+      "epoch": 0.0013398640266770196,
+      "grad_norm": 1.538041353225708,
+      "learning_rate": 4.492992992992993e-05,
+      "loss": 1.4151,
+      "step": 1025
+    },
+    {
+      "epoch": 0.00134117121109329,
+      "grad_norm": 1.6619324684143066,
+      "learning_rate": 4.4924924924924925e-05,
+      "loss": 1.3636,
+      "step": 1026
+    },
+    {
+      "epoch": 0.00134247839550956,
+      "grad_norm": 1.5119179487228394,
+      "learning_rate": 4.4919919919919924e-05,
+      "loss": 1.2279,
+      "step": 1027
+    },
+    {
+      "epoch": 0.0013437855799258303,
+      "grad_norm": 1.23798406124115,
+      "learning_rate": 4.491491491491492e-05,
+      "loss": 1.1753,
+      "step": 1028
+    },
+    {
+      "epoch": 0.0013450927643421007,
+      "grad_norm": 1.4199293851852417,
+      "learning_rate": 4.490990990990992e-05,
+      "loss": 1.1256,
+      "step": 1029
+    },
+    {
+      "epoch": 0.001346399948758371,
+      "grad_norm": 1.378620982170105,
+      "learning_rate": 4.490490490490491e-05,
+      "loss": 1.2436,
+      "step": 1030
+    },
+    {
+      "epoch": 0.001347707133174641,
+      "grad_norm": 1.5836337804794312,
+      "learning_rate": 4.48998998998999e-05,
+      "loss": 1.7565,
+      "step": 1031
+    },
+    {
+      "epoch": 0.0013490143175909114,
+      "grad_norm": 1.8844412565231323,
+      "learning_rate": 4.4894894894894896e-05,
+      "loss": 1.9317,
+      "step": 1032
+    },
+    {
+      "epoch": 0.0013503215020071817,
+      "grad_norm": 1.2800387144088745,
+      "learning_rate": 4.488988988988989e-05,
+      "loss": 1.767,
+      "step": 1033
+    },
+    {
+      "epoch": 0.0013516286864234518,
+      "grad_norm": 1.4740140438079834,
+      "learning_rate": 4.488488488488489e-05,
+      "loss": 1.5674,
+      "step": 1034
+    },
+    {
+      "epoch": 0.0013529358708397222,
+      "grad_norm": 1.7117799520492554,
+      "learning_rate": 4.487987987987988e-05,
+      "loss": 1.3394,
+      "step": 1035
+    },
+    {
+      "epoch": 0.0013542430552559925,
+      "grad_norm": 1.2293859720230103,
+      "learning_rate": 4.487487487487488e-05,
+      "loss": 0.5758,
+      "step": 1036
+    },
+    {
+      "epoch": 0.0013555502396722628,
+      "grad_norm": 1.6349135637283325,
+      "learning_rate": 4.4869869869869874e-05,
+      "loss": 2.1877,
+      "step": 1037
+    },
+    {
+      "epoch": 0.001356857424088533,
+      "grad_norm": 1.5475668907165527,
+      "learning_rate": 4.486486486486487e-05,
+      "loss": 1.9503,
+      "step": 1038
+    },
+    {
+      "epoch": 0.0013581646085048032,
+      "grad_norm": 1.531639575958252,
+      "learning_rate": 4.485985985985986e-05,
+      "loss": 2.7255,
+      "step": 1039
+    },
+    {
+      "epoch": 0.0013594717929210735,
+      "grad_norm": 1.5453858375549316,
+      "learning_rate": 4.485485485485485e-05,
+      "loss": 2.2243,
+      "step": 1040
+    },
+    {
+      "epoch": 0.0013607789773373439,
+      "grad_norm": 1.390790343284607,
+      "learning_rate": 4.484984984984985e-05,
+      "loss": 1.1866,
+      "step": 1041
+    },
+    {
+      "epoch": 0.001362086161753614,
+      "grad_norm": 1.5565235614776611,
+      "learning_rate": 4.4844844844844846e-05,
+      "loss": 2.1964,
+      "step": 1042
+    },
+    {
+      "epoch": 0.0013633933461698843,
+      "grad_norm": 1.446272611618042,
+      "learning_rate": 4.483983983983984e-05,
+      "loss": 1.5477,
+      "step": 1043
+    },
+    {
+      "epoch": 0.0013647005305861546,
+      "grad_norm": 1.6914466619491577,
+      "learning_rate": 4.483483483483484e-05,
+      "loss": 2.0997,
+      "step": 1044
+    },
+    {
+      "epoch": 0.001366007715002425,
+      "grad_norm": 1.078420877456665,
+      "learning_rate": 4.482982982982983e-05,
+      "loss": 0.826,
+      "step": 1045
+    },
+    {
+      "epoch": 0.001367314899418695,
+      "grad_norm": 1.3479429483413696,
+      "learning_rate": 4.482482482482483e-05,
+      "loss": 1.0279,
+      "step": 1046
+    },
+    {
+      "epoch": 0.0013686220838349653,
+      "grad_norm": 1.4340006113052368,
+      "learning_rate": 4.481981981981982e-05,
+      "loss": 1.4559,
+      "step": 1047
+    },
+    {
+      "epoch": 0.0013699292682512357,
+      "grad_norm": 1.1092137098312378,
+      "learning_rate": 4.481481481481482e-05,
+      "loss": 0.9114,
+      "step": 1048
+    },
+    {
+      "epoch": 0.001371236452667506,
+      "grad_norm": 1.5593024492263794,
+      "learning_rate": 4.480980980980981e-05,
+      "loss": 1.2339,
+      "step": 1049
+    },
+    {
+      "epoch": 0.001372543637083776,
+      "grad_norm": 1.5974435806274414,
+      "learning_rate": 4.48048048048048e-05,
+      "loss": 1.3428,
+      "step": 1050
+    },
+    {
+      "epoch": 0.0013738508215000464,
+      "grad_norm": 1.6131750345230103,
+      "learning_rate": 4.47997997997998e-05,
+      "loss": 1.6389,
+      "step": 1051
+    },
+    {
+      "epoch": 0.0013751580059163167,
+      "grad_norm": 1.2683908939361572,
+      "learning_rate": 4.4794794794794796e-05,
+      "loss": 2.207,
+      "step": 1052
+    },
+    {
+      "epoch": 0.0013764651903325868,
+      "grad_norm": 1.4106446504592896,
+      "learning_rate": 4.4789789789789796e-05,
+      "loss": 1.0792,
+      "step": 1053
+    },
+    {
+      "epoch": 0.0013777723747488571,
+      "grad_norm": 1.5027683973312378,
+      "learning_rate": 4.478478478478479e-05,
+      "loss": 1.0829,
+      "step": 1054
+    },
+    {
+      "epoch": 0.0013790795591651275,
+      "grad_norm": 1.6816227436065674,
+      "learning_rate": 4.477977977977978e-05,
+      "loss": 1.7634,
+      "step": 1055
+    },
+    {
+      "epoch": 0.0013803867435813978,
+      "grad_norm": 1.4186557531356812,
+      "learning_rate": 4.4774774774774775e-05,
+      "loss": 1.5171,
+      "step": 1056
+    },
+    {
+      "epoch": 0.0013816939279976679,
+      "grad_norm": 1.5925618410110474,
+      "learning_rate": 4.476976976976977e-05,
+      "loss": 1.5306,
+      "step": 1057
+    },
+    {
+      "epoch": 0.0013830011124139382,
+      "grad_norm": 1.6793149709701538,
+      "learning_rate": 4.476476476476477e-05,
+      "loss": 2.1562,
+      "step": 1058
+    },
+    {
+      "epoch": 0.0013843082968302085,
+      "grad_norm": 1.6331313848495483,
+      "learning_rate": 4.475975975975976e-05,
+      "loss": 1.0657,
+      "step": 1059
+    },
+    {
+      "epoch": 0.0013856154812464788,
+      "grad_norm": 1.4284422397613525,
+      "learning_rate": 4.475475475475476e-05,
+      "loss": 1.8906,
+      "step": 1060
+    },
+    {
+      "epoch": 0.001386922665662749,
+      "grad_norm": 1.2693400382995605,
+      "learning_rate": 4.474974974974975e-05,
+      "loss": 1.2615,
+      "step": 1061
+    },
+    {
+      "epoch": 0.0013882298500790193,
+      "grad_norm": 1.5865108966827393,
+      "learning_rate": 4.4744744744744746e-05,
+      "loss": 1.5963,
+      "step": 1062
+    },
+    {
+      "epoch": 0.0013895370344952896,
+      "grad_norm": 1.8226633071899414,
+      "learning_rate": 4.4739739739739746e-05,
+      "loss": 1.5368,
+      "step": 1063
+    },
+    {
+      "epoch": 0.00139084421891156,
+      "grad_norm": 1.5632004737854004,
+      "learning_rate": 4.473473473473473e-05,
+      "loss": 2.3385,
+      "step": 1064
+    },
+    {
+      "epoch": 0.00139215140332783,
+      "grad_norm": 1.6352211236953735,
+      "learning_rate": 4.472972972972973e-05,
+      "loss": 1.6503,
+      "step": 1065
+    },
+    {
+      "epoch": 0.0013934585877441003,
+      "grad_norm": 1.6075360774993896,
+      "learning_rate": 4.4724724724724725e-05,
+      "loss": 1.4972,
+      "step": 1066
+    },
+    {
+      "epoch": 0.0013947657721603706,
+      "grad_norm": 1.4677680730819702,
+      "learning_rate": 4.4719719719719725e-05,
+      "loss": 1.226,
+      "step": 1067
+    },
+    {
+      "epoch": 0.001396072956576641,
+      "grad_norm": 1.544594168663025,
+      "learning_rate": 4.471471471471472e-05,
+      "loss": 1.8329,
+      "step": 1068
+    },
+    {
+      "epoch": 0.001397380140992911,
+      "grad_norm": 1.6306288242340088,
+      "learning_rate": 4.470970970970971e-05,
+      "loss": 1.88,
+      "step": 1069
+    },
+    {
+      "epoch": 0.0013986873254091814,
+      "grad_norm": 1.0322602987289429,
+      "learning_rate": 4.470470470470471e-05,
+      "loss": 0.8276,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0013999945098254517,
+      "grad_norm": 1.4682077169418335,
+      "learning_rate": 4.46996996996997e-05,
+      "loss": 2.2163,
+      "step": 1071
+    },
+    {
+      "epoch": 0.001401301694241722,
+      "grad_norm": 1.5061390399932861,
+      "learning_rate": 4.4694694694694696e-05,
+      "loss": 1.6961,
+      "step": 1072
+    },
+    {
+      "epoch": 0.0014026088786579921,
+      "grad_norm": 1.7032755613327026,
+      "learning_rate": 4.468968968968969e-05,
+      "loss": 2.2922,
+      "step": 1073
+    },
+    {
+      "epoch": 0.0014039160630742625,
+      "grad_norm": 1.4059962034225464,
+      "learning_rate": 4.468468468468469e-05,
+      "loss": 1.0916,
+      "step": 1074
+    },
+    {
+      "epoch": 0.0014052232474905328,
+      "grad_norm": 1.0631825923919678,
+      "learning_rate": 4.467967967967968e-05,
+      "loss": 0.5614,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0014065304319068029,
+      "grad_norm": 1.2643707990646362,
+      "learning_rate": 4.4674674674674675e-05,
+      "loss": 2.9687,
+      "step": 1076
+    },
+    {
+      "epoch": 0.0014078376163230732,
+      "grad_norm": 1.7831271886825562,
+      "learning_rate": 4.4669669669669675e-05,
+      "loss": 1.5579,
+      "step": 1077
+    },
+    {
+      "epoch": 0.0014091448007393435,
+      "grad_norm": 1.3956787586212158,
+      "learning_rate": 4.466466466466467e-05,
+      "loss": 1.2812,
+      "step": 1078
+    },
+    {
+      "epoch": 0.0014104519851556138,
+      "grad_norm": 1.9840550422668457,
+      "learning_rate": 4.465965965965966e-05,
+      "loss": 1.7758,
+      "step": 1079
+    },
+    {
+      "epoch": 0.001411759169571884,
+      "grad_norm": 1.5351686477661133,
+      "learning_rate": 4.4654654654654654e-05,
+      "loss": 2.1252,
+      "step": 1080
+    },
+    {
+      "epoch": 0.0014130663539881543,
+      "grad_norm": 1.6291900873184204,
+      "learning_rate": 4.4649649649649647e-05,
+      "loss": 2.5752,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0014143735384044246,
+      "grad_norm": 1.5300661325454712,
+      "learning_rate": 4.4644644644644646e-05,
+      "loss": 1.1044,
+      "step": 1082
+    },
+    {
+      "epoch": 0.001415680722820695,
+      "grad_norm": 1.3166667222976685,
+      "learning_rate": 4.463963963963964e-05,
+      "loss": 1.2051,
+      "step": 1083
+    },
+    {
+      "epoch": 0.001416987907236965,
+      "grad_norm": 1.7525458335876465,
+      "learning_rate": 4.463463463463464e-05,
+      "loss": 1.4984,
+      "step": 1084
+    },
+    {
+      "epoch": 0.0014182950916532353,
+      "grad_norm": 1.681523323059082,
+      "learning_rate": 4.462962962962963e-05,
+      "loss": 1.981,
+      "step": 1085
+    },
+    {
+      "epoch": 0.0014196022760695056,
+      "grad_norm": 1.403407096862793,
+      "learning_rate": 4.4624624624624625e-05,
+      "loss": 1.6837,
+      "step": 1086
+    },
+    {
+      "epoch": 0.001420909460485776,
+      "grad_norm": 1.60698401927948,
+      "learning_rate": 4.4619619619619625e-05,
+      "loss": 1.2903,
+      "step": 1087
+    },
+    {
+      "epoch": 0.001422216644902046,
+      "grad_norm": 1.5527478456497192,
+      "learning_rate": 4.461461461461462e-05,
+      "loss": 2.3446,
+      "step": 1088
+    },
+    {
+      "epoch": 0.0014235238293183164,
+      "grad_norm": 1.3959760665893555,
+      "learning_rate": 4.460960960960961e-05,
+      "loss": 1.2579,
+      "step": 1089
+    },
+    {
+      "epoch": 0.0014248310137345867,
+      "grad_norm": 1.4291527271270752,
+      "learning_rate": 4.4604604604604604e-05,
+      "loss": 1.9366,
+      "step": 1090
+    },
+    {
+      "epoch": 0.001426138198150857,
+      "grad_norm": 1.8839575052261353,
+      "learning_rate": 4.4599599599599603e-05,
+      "loss": 1.9503,
+      "step": 1091
+    },
+    {
+      "epoch": 0.0014274453825671271,
+      "grad_norm": 1.7062782049179077,
+      "learning_rate": 4.4594594594594596e-05,
+      "loss": 2.0039,
+      "step": 1092
+    },
+    {
+      "epoch": 0.0014287525669833974,
+      "grad_norm": 1.2447699308395386,
+      "learning_rate": 4.458958958958959e-05,
+      "loss": 1.625,
+      "step": 1093
+    },
+    {
+      "epoch": 0.0014300597513996678,
+      "grad_norm": 1.73619544506073,
+      "learning_rate": 4.458458458458459e-05,
+      "loss": 1.732,
+      "step": 1094
+    },
+    {
+      "epoch": 0.001431366935815938,
+      "grad_norm": 1.52973210811615,
+      "learning_rate": 4.457957957957958e-05,
+      "loss": 1.8194,
+      "step": 1095
+    },
+    {
+      "epoch": 0.0014326741202322082,
+      "grad_norm": 1.4581701755523682,
+      "learning_rate": 4.457457457457458e-05,
+      "loss": 1.3531,
+      "step": 1096
+    },
+    {
+      "epoch": 0.0014339813046484785,
+      "grad_norm": 1.3048017024993896,
+      "learning_rate": 4.456956956956957e-05,
+      "loss": 1.0823,
+      "step": 1097
+    },
+    {
+      "epoch": 0.0014352884890647488,
+      "grad_norm": 2.210214376449585,
+      "learning_rate": 4.456456456456457e-05,
+      "loss": 1.984,
+      "step": 1098
+    },
+    {
+      "epoch": 0.001436595673481019,
+      "grad_norm": 1.324388027191162,
+      "learning_rate": 4.455955955955956e-05,
+      "loss": 1.9303,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0014379028578972892,
+      "grad_norm": 1.6504307985305786,
+      "learning_rate": 4.4554554554554554e-05,
+      "loss": 1.7924,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0014392100423135596,
+      "grad_norm": 1.3554632663726807,
+      "learning_rate": 4.4549549549549553e-05,
+      "loss": 1.4162,
+      "step": 1101
+    },
+    {
+      "epoch": 0.0014405172267298299,
+      "grad_norm": 1.3540107011795044,
+      "learning_rate": 4.4544544544544546e-05,
+      "loss": 1.7356,
+      "step": 1102
+    },
+    {
+      "epoch": 0.0014418244111461,
+      "grad_norm": 1.3198270797729492,
+      "learning_rate": 4.4539539539539546e-05,
+      "loss": 1.355,
+      "step": 1103
+    },
+    {
+      "epoch": 0.0014431315955623703,
+      "grad_norm": 1.3873863220214844,
+      "learning_rate": 4.453453453453454e-05,
+      "loss": 1.4008,
+      "step": 1104
+    },
+    {
+      "epoch": 0.0014444387799786406,
+      "grad_norm": 1.5417594909667969,
+      "learning_rate": 4.452952952952953e-05,
+      "loss": 2.0057,
+      "step": 1105
+    },
+    {
+      "epoch": 0.001445745964394911,
+      "grad_norm": 1.4490896463394165,
+      "learning_rate": 4.4524524524524525e-05,
+      "loss": 1.4947,
+      "step": 1106
+    },
+    {
+      "epoch": 0.001447053148811181,
+      "grad_norm": 1.7169790267944336,
+      "learning_rate": 4.451951951951952e-05,
+      "loss": 2.0466,
+      "step": 1107
+    },
+    {
+      "epoch": 0.0014483603332274514,
+      "grad_norm": 1.5183039903640747,
+      "learning_rate": 4.451451451451452e-05,
+      "loss": 1.8438,
+      "step": 1108
+    },
+    {
+      "epoch": 0.0014496675176437217,
+      "grad_norm": 1.5959415435791016,
+      "learning_rate": 4.450950950950951e-05,
+      "loss": 1.5984,
+      "step": 1109
+    },
+    {
+      "epoch": 0.001450974702059992,
+      "grad_norm": 1.9591600894927979,
+      "learning_rate": 4.450450450450451e-05,
+      "loss": 2.3898,
+      "step": 1110
+    },
+    {
+      "epoch": 0.0014522818864762621,
+      "grad_norm": 1.4900435209274292,
+      "learning_rate": 4.4499499499499504e-05,
+      "loss": 2.1395,
+      "step": 1111
+    },
+    {
+      "epoch": 0.0014535890708925324,
+      "grad_norm": 1.6837607622146606,
+      "learning_rate": 4.4494494494494497e-05,
+      "loss": 2.0868,
+      "step": 1112
+    },
+    {
+      "epoch": 0.0014548962553088028,
+      "grad_norm": 1.4912734031677246,
+      "learning_rate": 4.4489489489489496e-05,
+      "loss": 0.9278,
+      "step": 1113
+    },
+    {
+      "epoch": 0.001456203439725073,
+      "grad_norm": 1.4343236684799194,
+      "learning_rate": 4.448448448448448e-05,
+      "loss": 1.6078,
+      "step": 1114
+    },
+    {
+      "epoch": 0.0014575106241413432,
+      "grad_norm": 1.4999350309371948,
+      "learning_rate": 4.447947947947948e-05,
+      "loss": 1.341,
+      "step": 1115
+    },
+    {
+      "epoch": 0.0014588178085576135,
+      "grad_norm": 1.4038490056991577,
+      "learning_rate": 4.4474474474474475e-05,
+      "loss": 1.3936,
+      "step": 1116
+    },
+    {
+      "epoch": 0.0014601249929738838,
+      "grad_norm": 1.0398789644241333,
+      "learning_rate": 4.446946946946947e-05,
+      "loss": 0.578,
+      "step": 1117
+    },
+    {
+      "epoch": 0.001461432177390154,
+      "grad_norm": 1.4073609113693237,
+      "learning_rate": 4.446446446446447e-05,
+      "loss": 0.9457,
+      "step": 1118
+    },
+    {
+      "epoch": 0.0014627393618064242,
+      "grad_norm": 1.2467488050460815,
+      "learning_rate": 4.445945945945946e-05,
+      "loss": 0.9525,
+      "step": 1119
+    },
+    {
+      "epoch": 0.0014640465462226946,
+      "grad_norm": 1.9137269258499146,
+      "learning_rate": 4.445445445445446e-05,
+      "loss": 2.7107,
+      "step": 1120
+    },
+    {
+      "epoch": 0.0014653537306389649,
+      "grad_norm": 1.5996705293655396,
+      "learning_rate": 4.4449449449449454e-05,
+      "loss": 1.5259,
+      "step": 1121
+    },
+    {
+      "epoch": 0.001466660915055235,
+      "grad_norm": 1.280623197555542,
+      "learning_rate": 4.4444444444444447e-05,
+      "loss": 1.15,
+      "step": 1122
+    },
+    {
+      "epoch": 0.0014679680994715053,
+      "grad_norm": 1.50899338722229,
+      "learning_rate": 4.443943943943944e-05,
+      "loss": 2.2523,
+      "step": 1123
+    },
+    {
+      "epoch": 0.0014692752838877756,
+      "grad_norm": 1.5035184621810913,
+      "learning_rate": 4.443443443443443e-05,
+      "loss": 1.4263,
+      "step": 1124
+    },
+    {
+      "epoch": 0.001470582468304046,
+      "grad_norm": 1.273934006690979,
+      "learning_rate": 4.442942942942943e-05,
+      "loss": 1.065,
+      "step": 1125
+    },
+    {
+      "epoch": 0.001471889652720316,
+      "grad_norm": 1.5894581079483032,
+      "learning_rate": 4.4424424424424425e-05,
+      "loss": 2.3779,
+      "step": 1126
+    },
+    {
+      "epoch": 0.0014731968371365864,
+      "grad_norm": 1.4514482021331787,
+      "learning_rate": 4.4419419419419425e-05,
+      "loss": 1.1883,
+      "step": 1127
+    },
+    {
+      "epoch": 0.0014745040215528567,
+      "grad_norm": 1.4540423154830933,
+      "learning_rate": 4.441441441441442e-05,
+      "loss": 1.9158,
+      "step": 1128
+    },
+    {
+      "epoch": 0.001475811205969127,
+      "grad_norm": 1.4299407005310059,
+      "learning_rate": 4.440940940940941e-05,
+      "loss": 1.6812,
+      "step": 1129
+    },
+    {
+      "epoch": 0.001477118390385397,
+      "grad_norm": 1.3102803230285645,
+      "learning_rate": 4.4404404404404404e-05,
+      "loss": 1.0295,
+      "step": 1130
+    },
+    {
+      "epoch": 0.0014784255748016674,
+      "grad_norm": 1.6308422088623047,
+      "learning_rate": 4.43993993993994e-05,
+      "loss": 2.6251,
+      "step": 1131
+    },
+    {
+      "epoch": 0.0014797327592179377,
+      "grad_norm": 1.6849792003631592,
+      "learning_rate": 4.43943943943944e-05,
+      "loss": 1.5007,
+      "step": 1132
+    },
+    {
+      "epoch": 0.001481039943634208,
+      "grad_norm": 1.442339301109314,
+      "learning_rate": 4.438938938938939e-05,
+      "loss": 1.9181,
+      "step": 1133
+    },
+    {
+      "epoch": 0.0014823471280504782,
+      "grad_norm": 1.2693877220153809,
+      "learning_rate": 4.438438438438439e-05,
+      "loss": 1.4906,
+      "step": 1134
+    },
+    {
+      "epoch": 0.0014836543124667485,
+      "grad_norm": 1.720503807067871,
+      "learning_rate": 4.437937937937938e-05,
+      "loss": 1.8859,
+      "step": 1135
+    },
+    {
+      "epoch": 0.0014849614968830188,
+      "grad_norm": 1.665482521057129,
+      "learning_rate": 4.4374374374374375e-05,
+      "loss": 1.2516,
+      "step": 1136
+    },
+    {
+      "epoch": 0.0014862686812992891,
+      "grad_norm": 1.2259413003921509,
+      "learning_rate": 4.4369369369369375e-05,
+      "loss": 1.7504,
+      "step": 1137
+    },
+    {
+      "epoch": 0.0014875758657155592,
+      "grad_norm": 1.7044838666915894,
+      "learning_rate": 4.436436436436437e-05,
+      "loss": 1.4455,
+      "step": 1138
+    },
+    {
+      "epoch": 0.0014888830501318296,
+      "grad_norm": 1.4143483638763428,
+      "learning_rate": 4.435935935935936e-05,
+      "loss": 2.553,
+      "step": 1139
+    },
+    {
+      "epoch": 0.0014901902345480999,
+      "grad_norm": 1.480130672454834,
+      "learning_rate": 4.4354354354354354e-05,
+      "loss": 1.4149,
+      "step": 1140
+    },
+    {
+      "epoch": 0.00149149741896437,
+      "grad_norm": 1.1919716596603394,
+      "learning_rate": 4.4349349349349354e-05,
+      "loss": 1.0367,
+      "step": 1141
+    },
+    {
+      "epoch": 0.0014928046033806403,
+      "grad_norm": 1.563664436340332,
+      "learning_rate": 4.434434434434435e-05,
+      "loss": 1.5095,
+      "step": 1142
+    },
+    {
+      "epoch": 0.0014941117877969106,
+      "grad_norm": 1.324340581893921,
+      "learning_rate": 4.433933933933934e-05,
+      "loss": 1.2211,
+      "step": 1143
+    },
+    {
+      "epoch": 0.001495418972213181,
+      "grad_norm": 1.5284448862075806,
+      "learning_rate": 4.433433433433434e-05,
+      "loss": 2.5254,
+      "step": 1144
+    },
+    {
+      "epoch": 0.001496726156629451,
+      "grad_norm": 1.1125216484069824,
+      "learning_rate": 4.432932932932933e-05,
+      "loss": 1.259,
+      "step": 1145
+    },
+    {
+      "epoch": 0.0014980333410457214,
+      "grad_norm": 1.1863129138946533,
+      "learning_rate": 4.432432432432433e-05,
+      "loss": 1.1715,
+      "step": 1146
+    },
+    {
+      "epoch": 0.0014993405254619917,
+      "grad_norm": 1.4508670568466187,
+      "learning_rate": 4.431931931931932e-05,
+      "loss": 1.3734,
+      "step": 1147
+    },
+    {
+      "epoch": 0.001500647709878262,
+      "grad_norm": 1.281041145324707,
+      "learning_rate": 4.431431431431431e-05,
+      "loss": 1.8241,
+      "step": 1148
+    },
+    {
+      "epoch": 0.001501954894294532,
+      "grad_norm": 1.8391817808151245,
+      "learning_rate": 4.430930930930931e-05,
+      "loss": 2.3275,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0015032620787108024,
+      "grad_norm": 0.9320809841156006,
+      "learning_rate": 4.4304304304304304e-05,
+      "loss": 0.6891,
+      "step": 1150
+    },
+    {
+      "epoch": 0.0015045692631270727,
+      "grad_norm": 1.683285117149353,
+      "learning_rate": 4.4299299299299304e-05,
+      "loss": 2.2791,
+      "step": 1151
+    },
+    {
+      "epoch": 0.001505876447543343,
+      "grad_norm": 1.3434828519821167,
+      "learning_rate": 4.42942942942943e-05,
+      "loss": 1.8591,
+      "step": 1152
+    },
+    {
+      "epoch": 0.0015071836319596132,
+      "grad_norm": 2.2481801509857178,
+      "learning_rate": 4.428928928928929e-05,
+      "loss": 1.507,
+      "step": 1153
+    },
+    {
+      "epoch": 0.0015084908163758835,
+      "grad_norm": 1.3485732078552246,
+      "learning_rate": 4.428428428428429e-05,
+      "loss": 0.977,
+      "step": 1154
+    },
+    {
+      "epoch": 0.0015097980007921538,
+      "grad_norm": 2.09972882270813,
+      "learning_rate": 4.4279279279279276e-05,
+      "loss": 1.855,
+      "step": 1155
+    },
+    {
+      "epoch": 0.0015111051852084241,
+      "grad_norm": 1.4306321144104004,
+      "learning_rate": 4.4274274274274275e-05,
+      "loss": 1.4004,
+      "step": 1156
+    },
+    {
+      "epoch": 0.0015124123696246942,
+      "grad_norm": 1.4846643209457397,
+      "learning_rate": 4.426926926926927e-05,
+      "loss": 1.0552,
+      "step": 1157
+    },
+    {
+      "epoch": 0.0015137195540409645,
+      "grad_norm": 1.423444151878357,
+      "learning_rate": 4.426426426426427e-05,
+      "loss": 0.6854,
+      "step": 1158
+    },
+    {
+      "epoch": 0.0015150267384572349,
+      "grad_norm": 1.7255076169967651,
+      "learning_rate": 4.425925925925926e-05,
+      "loss": 1.6995,
+      "step": 1159
+    },
+    {
+      "epoch": 0.001516333922873505,
+      "grad_norm": 1.4593850374221802,
+      "learning_rate": 4.4254254254254254e-05,
+      "loss": 1.6136,
+      "step": 1160
+    },
+    {
+      "epoch": 0.0015176411072897753,
+      "grad_norm": 1.5913044214248657,
+      "learning_rate": 4.4249249249249254e-05,
+      "loss": 2.1368,
+      "step": 1161
+    },
+    {
+      "epoch": 0.0015189482917060456,
+      "grad_norm": 1.4174920320510864,
+      "learning_rate": 4.424424424424425e-05,
+      "loss": 1.5843,
+      "step": 1162
+    },
+    {
+      "epoch": 0.001520255476122316,
+      "grad_norm": 1.363776445388794,
+      "learning_rate": 4.4239239239239247e-05,
+      "loss": 1.4175,
+      "step": 1163
+    },
+    {
+      "epoch": 0.001521562660538586,
+      "grad_norm": 1.5941675901412964,
+      "learning_rate": 4.423423423423423e-05,
+      "loss": 1.7733,
+      "step": 1164
+    },
+    {
+      "epoch": 0.0015228698449548563,
+      "grad_norm": 1.4240998029708862,
+      "learning_rate": 4.422922922922923e-05,
+      "loss": 1.5153,
+      "step": 1165
+    },
+    {
+      "epoch": 0.0015241770293711267,
+      "grad_norm": 1.6118253469467163,
+      "learning_rate": 4.4224224224224226e-05,
+      "loss": 2.225,
+      "step": 1166
+    },
+    {
+      "epoch": 0.001525484213787397,
+      "grad_norm": 1.6440070867538452,
+      "learning_rate": 4.421921921921922e-05,
+      "loss": 1.1364,
+      "step": 1167
+    },
+    {
+      "epoch": 0.001526791398203667,
+      "grad_norm": 1.5479505062103271,
+      "learning_rate": 4.421421421421422e-05,
+      "loss": 1.8757,
+      "step": 1168
+    },
+    {
+      "epoch": 0.0015280985826199374,
+      "grad_norm": 1.4486463069915771,
+      "learning_rate": 4.420920920920921e-05,
+      "loss": 2.0286,
+      "step": 1169
+    },
+    {
+      "epoch": 0.0015294057670362077,
+      "grad_norm": 1.219124436378479,
+      "learning_rate": 4.420420420420421e-05,
+      "loss": 1.4304,
+      "step": 1170
+    },
+    {
+      "epoch": 0.001530712951452478,
+      "grad_norm": 1.0430004596710205,
+      "learning_rate": 4.4199199199199204e-05,
+      "loss": 0.2533,
+      "step": 1171
+    },
+    {
+      "epoch": 0.0015320201358687481,
+      "grad_norm": 2.2137339115142822,
+      "learning_rate": 4.41941941941942e-05,
+      "loss": 1.0365,
+      "step": 1172
+    },
+    {
+      "epoch": 0.0015333273202850185,
+      "grad_norm": 1.4467207193374634,
+      "learning_rate": 4.418918918918919e-05,
+      "loss": 2.2587,
+      "step": 1173
+    },
+    {
+      "epoch": 0.0015346345047012888,
+      "grad_norm": 1.0869256258010864,
+      "learning_rate": 4.418418418418418e-05,
+      "loss": 1.2913,
+      "step": 1174
+    },
+    {
+      "epoch": 0.0015359416891175591,
+      "grad_norm": 1.4638752937316895,
+      "learning_rate": 4.417917917917918e-05,
+      "loss": 1.6315,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0015372488735338292,
+      "grad_norm": 1.4086905717849731,
+      "learning_rate": 4.4174174174174176e-05,
+      "loss": 1.7919,
+      "step": 1176
+    },
+    {
+      "epoch": 0.0015385560579500995,
+      "grad_norm": 1.8316904306411743,
+      "learning_rate": 4.4169169169169175e-05,
+      "loss": 2.1954,
+      "step": 1177
+    },
+    {
+      "epoch": 0.0015398632423663699,
+      "grad_norm": 1.2402193546295166,
+      "learning_rate": 4.416416416416417e-05,
+      "loss": 1.2364,
+      "step": 1178
+    },
+    {
+      "epoch": 0.0015411704267826402,
+      "grad_norm": 1.315686821937561,
+      "learning_rate": 4.415915915915916e-05,
+      "loss": 1.2691,
+      "step": 1179
+    },
+    {
+      "epoch": 0.0015424776111989103,
+      "grad_norm": 1.3043485879898071,
+      "learning_rate": 4.4154154154154154e-05,
+      "loss": 1.3194,
+      "step": 1180
+    },
+    {
+      "epoch": 0.0015437847956151806,
+      "grad_norm": 1.8233788013458252,
+      "learning_rate": 4.414914914914915e-05,
+      "loss": 2.0747,
+      "step": 1181
+    },
+    {
+      "epoch": 0.001545091980031451,
+      "grad_norm": 1.4552195072174072,
+      "learning_rate": 4.414414414414415e-05,
+      "loss": 1.9802,
+      "step": 1182
+    },
+    {
+      "epoch": 0.001546399164447721,
+      "grad_norm": 1.519539713859558,
+      "learning_rate": 4.413913913913914e-05,
+      "loss": 1.2207,
+      "step": 1183
+    },
+    {
+      "epoch": 0.0015477063488639913,
+      "grad_norm": 1.5158014297485352,
+      "learning_rate": 4.413413413413414e-05,
+      "loss": 1.5761,
+      "step": 1184
+    },
+    {
+      "epoch": 0.0015490135332802617,
+      "grad_norm": 1.6284315586090088,
+      "learning_rate": 4.412912912912913e-05,
+      "loss": 2.2527,
+      "step": 1185
+    },
+    {
+      "epoch": 0.001550320717696532,
+      "grad_norm": 1.2745996713638306,
+      "learning_rate": 4.4124124124124126e-05,
+      "loss": 1.9546,
+      "step": 1186
+    },
+    {
+      "epoch": 0.001551627902112802,
+      "grad_norm": 1.6594645977020264,
+      "learning_rate": 4.4119119119119125e-05,
+      "loss": 1.3062,
+      "step": 1187
+    },
+    {
+      "epoch": 0.0015529350865290724,
+      "grad_norm": 1.3198025226593018,
+      "learning_rate": 4.411411411411412e-05,
+      "loss": 1.3369,
+      "step": 1188
+    },
+    {
+      "epoch": 0.0015542422709453427,
+      "grad_norm": 1.2512048482894897,
+      "learning_rate": 4.410910910910911e-05,
+      "loss": 1.2788,
+      "step": 1189
+    },
+    {
+      "epoch": 0.001555549455361613,
+      "grad_norm": 1.5593481063842773,
+      "learning_rate": 4.4104104104104104e-05,
+      "loss": 1.859,
+      "step": 1190
+    },
+    {
+      "epoch": 0.0015568566397778831,
+      "grad_norm": 1.7205810546875,
+      "learning_rate": 4.40990990990991e-05,
+      "loss": 1.4135,
+      "step": 1191
+    },
+    {
+      "epoch": 0.0015581638241941535,
+      "grad_norm": 0.6868228316307068,
+      "learning_rate": 4.40940940940941e-05,
+      "loss": 0.2401,
+      "step": 1192
+    },
+    {
+      "epoch": 0.0015594710086104238,
+      "grad_norm": 1.4841355085372925,
+      "learning_rate": 4.408908908908909e-05,
+      "loss": 1.6513,
+      "step": 1193
+    },
+    {
+      "epoch": 0.001560778193026694,
+      "grad_norm": 1.5980615615844727,
+      "learning_rate": 4.408408408408409e-05,
+      "loss": 1.7611,
+      "step": 1194
+    },
+    {
+      "epoch": 0.0015620853774429642,
+      "grad_norm": 1.1999614238739014,
+      "learning_rate": 4.407907907907908e-05,
+      "loss": 1.2208,
+      "step": 1195
+    },
+    {
+      "epoch": 0.0015633925618592345,
+      "grad_norm": 1.4964503049850464,
+      "learning_rate": 4.4074074074074076e-05,
+      "loss": 1.4227,
+      "step": 1196
+    },
+    {
+      "epoch": 0.0015646997462755048,
+      "grad_norm": 1.1320579051971436,
+      "learning_rate": 4.406906906906907e-05,
+      "loss": 1.0842,
+      "step": 1197
+    },
+    {
+      "epoch": 0.0015660069306917752,
+      "grad_norm": 1.6775199174880981,
+      "learning_rate": 4.406406406406406e-05,
+      "loss": 1.9969,
+      "step": 1198
+    },
+    {
+      "epoch": 0.0015673141151080453,
+      "grad_norm": 1.8392958641052246,
+      "learning_rate": 4.405905905905906e-05,
+      "loss": 1.9819,
+      "step": 1199
+    },
+    {
+      "epoch": 0.0015686212995243156,
+      "grad_norm": 1.163496494293213,
+      "learning_rate": 4.4054054054054054e-05,
+      "loss": 0.9046,
+      "step": 1200
+    },
+    {
+      "epoch": 0.001569928483940586,
+      "grad_norm": 1.4975576400756836,
+      "learning_rate": 4.4049049049049054e-05,
+      "loss": 1.9296,
+      "step": 1201
+    },
+    {
+      "epoch": 0.0015712356683568562,
+      "grad_norm": 1.1918988227844238,
+      "learning_rate": 4.404404404404405e-05,
+      "loss": 0.8539,
+      "step": 1202
+    },
+    {
+      "epoch": 0.0015725428527731263,
+      "grad_norm": 1.558987021446228,
+      "learning_rate": 4.403903903903904e-05,
+      "loss": 1.9624,
+      "step": 1203
+    },
+    {
+      "epoch": 0.0015738500371893966,
+      "grad_norm": 1.498814582824707,
+      "learning_rate": 4.403403403403404e-05,
+      "loss": 1.9889,
+      "step": 1204
+    },
+    {
+      "epoch": 0.001575157221605667,
+      "grad_norm": 1.4256953001022339,
+      "learning_rate": 4.4029029029029026e-05,
+      "loss": 1.5232,
+      "step": 1205
+    },
+    {
+      "epoch": 0.001576464406021937,
+      "grad_norm": 1.6388453245162964,
+      "learning_rate": 4.4024024024024026e-05,
+      "loss": 1.6846,
+      "step": 1206
+    },
+    {
+      "epoch": 0.0015777715904382074,
+      "grad_norm": 1.4944214820861816,
+      "learning_rate": 4.401901901901902e-05,
+      "loss": 1.7971,
+      "step": 1207
+    },
+    {
+      "epoch": 0.0015790787748544777,
+      "grad_norm": 1.376610279083252,
+      "learning_rate": 4.401401401401402e-05,
+      "loss": 1.986,
+      "step": 1208
+    },
+    {
+      "epoch": 0.001580385959270748,
+      "grad_norm": 1.7665876150131226,
+      "learning_rate": 4.400900900900901e-05,
+      "loss": 2.2618,
+      "step": 1209
+    },
+    {
+      "epoch": 0.0015816931436870181,
+      "grad_norm": 1.2744839191436768,
+      "learning_rate": 4.4004004004004004e-05,
+      "loss": 0.9995,
+      "step": 1210
+    },
+    {
+      "epoch": 0.0015830003281032885,
+      "grad_norm": 1.646520972251892,
+      "learning_rate": 4.3998998998999004e-05,
+      "loss": 2.1352,
+      "step": 1211
+    },
+    {
+      "epoch": 0.0015843075125195588,
+      "grad_norm": 1.1735550165176392,
+      "learning_rate": 4.3993993993994e-05,
+      "loss": 0.9999,
+      "step": 1212
+    },
+    {
+      "epoch": 0.001585614696935829,
+      "grad_norm": 1.1633192300796509,
+      "learning_rate": 4.398898898898899e-05,
+      "loss": 1.0543,
+      "step": 1213
+    },
+    {
+      "epoch": 0.0015869218813520992,
+      "grad_norm": 1.6410757303237915,
+      "learning_rate": 4.398398398398398e-05,
+      "loss": 1.5424,
+      "step": 1214
+    },
+    {
+      "epoch": 0.0015882290657683695,
+      "grad_norm": 1.5041000843048096,
+      "learning_rate": 4.397897897897898e-05,
+      "loss": 1.9636,
+      "step": 1215
+    },
+    {
+      "epoch": 0.0015895362501846398,
+      "grad_norm": 1.5833463668823242,
+      "learning_rate": 4.3973973973973976e-05,
+      "loss": 2.3137,
+      "step": 1216
+    },
+    {
+      "epoch": 0.0015908434346009102,
+      "grad_norm": 1.6321911811828613,
+      "learning_rate": 4.396896896896897e-05,
+      "loss": 2.1946,
+      "step": 1217
+    },
+    {
+      "epoch": 0.0015921506190171803,
+      "grad_norm": 1.7650401592254639,
+      "learning_rate": 4.396396396396397e-05,
+      "loss": 1.362,
+      "step": 1218
+    },
+    {
+      "epoch": 0.0015934578034334506,
+      "grad_norm": 1.3611323833465576,
+      "learning_rate": 4.395895895895896e-05,
+      "loss": 1.3303,
+      "step": 1219
+    },
+    {
+      "epoch": 0.001594764987849721,
+      "grad_norm": 1.2658058404922485,
+      "learning_rate": 4.395395395395396e-05,
+      "loss": 1.1709,
+      "step": 1220
+    },
+    {
+      "epoch": 0.0015960721722659912,
+      "grad_norm": 1.6971577405929565,
+      "learning_rate": 4.3948948948948954e-05,
+      "loss": 2.0272,
+      "step": 1221
+    },
+    {
+      "epoch": 0.0015973793566822613,
+      "grad_norm": 1.5502392053604126,
+      "learning_rate": 4.394394394394394e-05,
+      "loss": 1.8204,
+      "step": 1222
+    },
+    {
+      "epoch": 0.0015986865410985316,
+      "grad_norm": 1.6171573400497437,
+      "learning_rate": 4.393893893893894e-05,
+      "loss": 1.8864,
+      "step": 1223
+    },
+    {
+      "epoch": 0.001599993725514802,
+      "grad_norm": 1.5373413562774658,
+      "learning_rate": 4.393393393393393e-05,
+      "loss": 1.4704,
+      "step": 1224
+    },
+    {
+      "epoch": 0.001601300909931072,
+      "grad_norm": 1.4288482666015625,
+      "learning_rate": 4.392892892892893e-05,
+      "loss": 1.7857,
+      "step": 1225
+    },
+    {
+      "epoch": 0.0016026080943473424,
+      "grad_norm": 1.4504035711288452,
+      "learning_rate": 4.3923923923923926e-05,
+      "loss": 1.7675,
+      "step": 1226
+    },
+    {
+      "epoch": 0.0016039152787636127,
+      "grad_norm": 1.467299222946167,
+      "learning_rate": 4.391891891891892e-05,
+      "loss": 2.0677,
+      "step": 1227
+    },
+    {
+      "epoch": 0.001605222463179883,
+      "grad_norm": 1.6170086860656738,
+      "learning_rate": 4.391391391391392e-05,
+      "loss": 1.7516,
+      "step": 1228
+    },
+    {
+      "epoch": 0.0016065296475961531,
+      "grad_norm": 1.8216831684112549,
+      "learning_rate": 4.390890890890891e-05,
+      "loss": 1.2571,
+      "step": 1229
+    },
+    {
+      "epoch": 0.0016078368320124234,
+      "grad_norm": 1.494882583618164,
+      "learning_rate": 4.3903903903903905e-05,
+      "loss": 1.5727,
+      "step": 1230
+    },
+    {
+      "epoch": 0.0016091440164286938,
+      "grad_norm": 1.3008613586425781,
+      "learning_rate": 4.38988988988989e-05,
+      "loss": 0.9696,
+      "step": 1231
+    },
+    {
+      "epoch": 0.001610451200844964,
+      "grad_norm": 1.467699646949768,
+      "learning_rate": 4.38938938938939e-05,
+      "loss": 1.0714,
+      "step": 1232
+    },
+    {
+      "epoch": 0.0016117583852612342,
+      "grad_norm": 1.6584597826004028,
+      "learning_rate": 4.388888888888889e-05,
+      "loss": 2.1645,
+      "step": 1233
+    },
+    {
+      "epoch": 0.0016130655696775045,
+      "grad_norm": 1.5039458274841309,
+      "learning_rate": 4.388388388388388e-05,
+      "loss": 0.85,
+      "step": 1234
+    },
+    {
+      "epoch": 0.0016143727540937748,
+      "grad_norm": 1.4749056100845337,
+      "learning_rate": 4.387887887887888e-05,
+      "loss": 1.5049,
+      "step": 1235
+    },
+    {
+      "epoch": 0.0016156799385100451,
+      "grad_norm": 1.6271461248397827,
+      "learning_rate": 4.3873873873873876e-05,
+      "loss": 2.4346,
+      "step": 1236
+    },
+    {
+      "epoch": 0.0016169871229263152,
+      "grad_norm": 1.3466531038284302,
+      "learning_rate": 4.3868868868868876e-05,
+      "loss": 1.7227,
+      "step": 1237
+    },
+    {
+      "epoch": 0.0016182943073425856,
+      "grad_norm": 1.2843544483184814,
+      "learning_rate": 4.386386386386386e-05,
+      "loss": 1.2562,
+      "step": 1238
+    },
+    {
+      "epoch": 0.0016196014917588559,
+      "grad_norm": 1.4929187297821045,
+      "learning_rate": 4.385885885885886e-05,
+      "loss": 1.6347,
+      "step": 1239
+    },
+    {
+      "epoch": 0.0016209086761751262,
+      "grad_norm": 1.8974292278289795,
+      "learning_rate": 4.3853853853853855e-05,
+      "loss": 1.0762,
+      "step": 1240
+    },
+    {
+      "epoch": 0.0016222158605913963,
+      "grad_norm": 1.5485420227050781,
+      "learning_rate": 4.384884884884885e-05,
+      "loss": 1.4139,
+      "step": 1241
+    },
+    {
+      "epoch": 0.0016235230450076666,
+      "grad_norm": 1.703771710395813,
+      "learning_rate": 4.384384384384385e-05,
+      "loss": 1.5666,
+      "step": 1242
+    },
+    {
+      "epoch": 0.001624830229423937,
+      "grad_norm": 1.5430285930633545,
+      "learning_rate": 4.383883883883884e-05,
+      "loss": 1.7656,
+      "step": 1243
+    },
+    {
+      "epoch": 0.0016261374138402073,
+      "grad_norm": 1.935025930404663,
+      "learning_rate": 4.383383383383384e-05,
+      "loss": 1.8523,
+      "step": 1244
+    },
+    {
+      "epoch": 0.0016274445982564774,
+      "grad_norm": 1.3613533973693848,
+      "learning_rate": 4.382882882882883e-05,
+      "loss": 1.1636,
+      "step": 1245
+    },
+    {
+      "epoch": 0.0016287517826727477,
+      "grad_norm": 1.867212176322937,
+      "learning_rate": 4.3823823823823826e-05,
+      "loss": 2.3179,
+      "step": 1246
+    },
+    {
+      "epoch": 0.001630058967089018,
+      "grad_norm": 1.2737607955932617,
+      "learning_rate": 4.381881881881882e-05,
+      "loss": 1.3035,
+      "step": 1247
+    },
+    {
+      "epoch": 0.0016313661515052881,
+      "grad_norm": 1.649146318435669,
+      "learning_rate": 4.381381381381381e-05,
+      "loss": 1.6625,
+      "step": 1248
+    },
+    {
+      "epoch": 0.0016326733359215584,
+      "grad_norm": 1.5357623100280762,
+      "learning_rate": 4.380880880880881e-05,
+      "loss": 2.1161,
+      "step": 1249
+    },
+    {
+      "epoch": 0.0016339805203378288,
+      "grad_norm": 1.5750634670257568,
+      "learning_rate": 4.3803803803803805e-05,
+      "loss": 1.3742,
+      "step": 1250
+    },
+    {
+      "epoch": 0.001635287704754099,
+      "grad_norm": 1.2248564958572388,
+      "learning_rate": 4.3798798798798805e-05,
+      "loss": 1.2077,
+      "step": 1251
+    },
+    {
+      "epoch": 0.0016365948891703692,
+      "grad_norm": 1.7150086164474487,
+      "learning_rate": 4.37937937937938e-05,
+      "loss": 1.3169,
+      "step": 1252
+    },
+    {
+      "epoch": 0.0016379020735866395,
+      "grad_norm": 1.6553337574005127,
+      "learning_rate": 4.378878878878879e-05,
+      "loss": 1.4439,
+      "step": 1253
+    },
+    {
+      "epoch": 0.0016392092580029098,
+      "grad_norm": 1.6199713945388794,
+      "learning_rate": 4.378378378378379e-05,
+      "loss": 1.8177,
+      "step": 1254
+    },
+    {
+      "epoch": 0.0016405164424191801,
+      "grad_norm": 1.382757544517517,
+      "learning_rate": 4.3778778778778776e-05,
+      "loss": 2.0457,
+      "step": 1255
+    },
+    {
+      "epoch": 0.0016418236268354502,
+      "grad_norm": 1.4352965354919434,
+      "learning_rate": 4.3773773773773776e-05,
+      "loss": 1.5482,
+      "step": 1256
+    },
+    {
+      "epoch": 0.0016431308112517206,
+      "grad_norm": 1.1183345317840576,
+      "learning_rate": 4.376876876876877e-05,
+      "loss": 0.5715,
+      "step": 1257
+    },
+    {
+      "epoch": 0.0016444379956679909,
+      "grad_norm": 0.9702792167663574,
+      "learning_rate": 4.376376376376376e-05,
+      "loss": 0.614,
+      "step": 1258
+    },
+    {
+      "epoch": 0.0016457451800842612,
+      "grad_norm": 1.515434741973877,
+      "learning_rate": 4.375875875875876e-05,
+      "loss": 2.1585,
+      "step": 1259
+    },
+    {
+      "epoch": 0.0016470523645005313,
+      "grad_norm": 0.9589918255805969,
+      "learning_rate": 4.3753753753753755e-05,
+      "loss": 0.9146,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0016483595489168016,
+      "grad_norm": 1.6274853944778442,
+      "learning_rate": 4.3748748748748755e-05,
+      "loss": 2.0129,
+      "step": 1261
+    },
+    {
+      "epoch": 0.001649666733333072,
+      "grad_norm": 1.6729307174682617,
+      "learning_rate": 4.374374374374375e-05,
+      "loss": 1.0236,
+      "step": 1262
+    },
+    {
+      "epoch": 0.0016509739177493423,
+      "grad_norm": 1.5257456302642822,
+      "learning_rate": 4.373873873873874e-05,
+      "loss": 2.3678,
+      "step": 1263
+    },
+    {
+      "epoch": 0.0016522811021656124,
+      "grad_norm": 2.570962905883789,
+      "learning_rate": 4.3733733733733733e-05,
+      "loss": 1.3765,
+      "step": 1264
+    },
+    {
+      "epoch": 0.0016535882865818827,
+      "grad_norm": 2.3833565711975098,
+      "learning_rate": 4.3728728728728726e-05,
+      "loss": 1.9474,
+      "step": 1265
+    },
+    {
+      "epoch": 0.001654895470998153,
+      "grad_norm": 1.607430338859558,
+      "learning_rate": 4.3723723723723726e-05,
+      "loss": 1.4637,
+      "step": 1266
+    },
+    {
+      "epoch": 0.0016562026554144233,
+      "grad_norm": 1.2718642950057983,
+      "learning_rate": 4.371871871871872e-05,
+      "loss": 1.5711,
+      "step": 1267
+    },
+    {
+      "epoch": 0.0016575098398306934,
+      "grad_norm": 1.8313981294631958,
+      "learning_rate": 4.371371371371372e-05,
+      "loss": 1.9502,
+      "step": 1268
+    },
+    {
+      "epoch": 0.0016588170242469637,
+      "grad_norm": 0.4412385821342468,
+      "learning_rate": 4.370870870870871e-05,
+      "loss": 0.011,
+      "step": 1269
+    },
+    {
+      "epoch": 0.001660124208663234,
+      "grad_norm": 1.5772066116333008,
+      "learning_rate": 4.3703703703703705e-05,
+      "loss": 2.0022,
+      "step": 1270
+    },
+    {
+      "epoch": 0.0016614313930795042,
+      "grad_norm": 1.4418865442276,
+      "learning_rate": 4.3698698698698705e-05,
+      "loss": 1.8237,
+      "step": 1271
+    },
+    {
+      "epoch": 0.0016627385774957745,
+      "grad_norm": 1.0146071910858154,
+      "learning_rate": 4.369369369369369e-05,
+      "loss": 0.519,
+      "step": 1272
+    },
+    {
+      "epoch": 0.0016640457619120448,
+      "grad_norm": 1.2826814651489258,
+      "learning_rate": 4.368868868868869e-05,
+      "loss": 1.8981,
+      "step": 1273
+    },
+    {
+      "epoch": 0.0016653529463283151,
+      "grad_norm": 1.4657295942306519,
+      "learning_rate": 4.3683683683683684e-05,
+      "loss": 1.8027,
+      "step": 1274
+    },
+    {
+      "epoch": 0.0016666601307445852,
+      "grad_norm": 1.4937814474105835,
+      "learning_rate": 4.367867867867868e-05,
+      "loss": 1.501,
+      "step": 1275
+    },
+    {
+      "epoch": 0.0016679673151608555,
+      "grad_norm": 1.2491382360458374,
+      "learning_rate": 4.3673673673673676e-05,
+      "loss": 1.2338,
+      "step": 1276
+    },
+    {
+      "epoch": 0.0016692744995771259,
+      "grad_norm": 1.3882606029510498,
+      "learning_rate": 4.366866866866867e-05,
+      "loss": 1.2095,
+      "step": 1277
+    },
+    {
+      "epoch": 0.0016705816839933962,
+      "grad_norm": 1.4623945951461792,
+      "learning_rate": 4.366366366366367e-05,
+      "loss": 1.5785,
+      "step": 1278
+    },
+    {
+      "epoch": 0.0016718888684096663,
+      "grad_norm": 1.5898010730743408,
+      "learning_rate": 4.365865865865866e-05,
+      "loss": 1.2913,
+      "step": 1279
+    },
+    {
+      "epoch": 0.0016731960528259366,
+      "grad_norm": 0.7526304125785828,
+      "learning_rate": 4.3653653653653655e-05,
+      "loss": 0.1743,
+      "step": 1280
+    },
+    {
+      "epoch": 0.001674503237242207,
+      "grad_norm": 1.4021170139312744,
+      "learning_rate": 4.364864864864865e-05,
+      "loss": 2.2948,
+      "step": 1281
+    },
+    {
+      "epoch": 0.0016758104216584772,
+      "grad_norm": 1.0175584554672241,
+      "learning_rate": 4.364364364364365e-05,
+      "loss": 0.5163,
+      "step": 1282
+    },
+    {
+      "epoch": 0.0016771176060747474,
+      "grad_norm": 1.6596543788909912,
+      "learning_rate": 4.363863863863864e-05,
+      "loss": 2.3498,
+      "step": 1283
+    },
+    {
+      "epoch": 0.0016784247904910177,
+      "grad_norm": 1.184618353843689,
+      "learning_rate": 4.3633633633633634e-05,
+      "loss": 0.9975,
+      "step": 1284
+    },
+    {
+      "epoch": 0.001679731974907288,
+      "grad_norm": 1.7727652788162231,
+      "learning_rate": 4.362862862862863e-05,
+      "loss": 1.618,
+      "step": 1285
+    },
+    {
+      "epoch": 0.0016810391593235583,
+      "grad_norm": 1.3691402673721313,
+      "learning_rate": 4.3623623623623626e-05,
+      "loss": 1.1099,
+      "step": 1286
+    },
+    {
+      "epoch": 0.0016823463437398284,
+      "grad_norm": 1.4772765636444092,
+      "learning_rate": 4.3618618618618626e-05,
+      "loss": 1.6299,
+      "step": 1287
+    },
+    {
+      "epoch": 0.0016836535281560987,
+      "grad_norm": 1.486764669418335,
+      "learning_rate": 4.361361361361361e-05,
+      "loss": 2.0694,
+      "step": 1288
+    },
+    {
+      "epoch": 0.001684960712572369,
+      "grad_norm": 1.4340412616729736,
+      "learning_rate": 4.360860860860861e-05,
+      "loss": 1.9061,
+      "step": 1289
+    },
+    {
+      "epoch": 0.0016862678969886392,
+      "grad_norm": 1.3626625537872314,
+      "learning_rate": 4.3603603603603605e-05,
+      "loss": 1.4503,
+      "step": 1290
+    },
+    {
+      "epoch": 0.0016875750814049095,
+      "grad_norm": 1.7349869012832642,
+      "learning_rate": 4.35985985985986e-05,
+      "loss": 2.0538,
+      "step": 1291
+    },
+    {
+      "epoch": 0.0016888822658211798,
+      "grad_norm": 1.1891404390335083,
+      "learning_rate": 4.35935935935936e-05,
+      "loss": 1.0979,
+      "step": 1292
+    },
+    {
+      "epoch": 0.0016901894502374501,
+      "grad_norm": 1.5615184307098389,
+      "learning_rate": 4.358858858858859e-05,
+      "loss": 1.7685,
+      "step": 1293
+    },
+    {
+      "epoch": 0.0016914966346537202,
+      "grad_norm": 0.9962694048881531,
+      "learning_rate": 4.3583583583583584e-05,
+      "loss": 0.4108,
+      "step": 1294
+    },
+    {
+      "epoch": 0.0016928038190699905,
+      "grad_norm": 1.3363423347473145,
+      "learning_rate": 4.3578578578578583e-05,
+      "loss": 0.9138,
+      "step": 1295
+    },
+    {
+      "epoch": 0.0016941110034862609,
+      "grad_norm": 1.0401540994644165,
+      "learning_rate": 4.3573573573573576e-05,
+      "loss": 1.1761,
+      "step": 1296
+    },
+    {
+      "epoch": 0.0016954181879025312,
+      "grad_norm": 1.8245062828063965,
+      "learning_rate": 4.356856856856857e-05,
+      "loss": 2.1244,
+      "step": 1297
+    },
+    {
+      "epoch": 0.0016967253723188013,
+      "grad_norm": 1.7264286279678345,
+      "learning_rate": 4.356356356356356e-05,
+      "loss": 1.6421,
+      "step": 1298
+    },
+    {
+      "epoch": 0.0016980325567350716,
+      "grad_norm": 0.1397632360458374,
+      "learning_rate": 4.355855855855856e-05,
+      "loss": 0.0034,
+      "step": 1299
+    },
+    {
+      "epoch": 0.001699339741151342,
+      "grad_norm": 1.5455902814865112,
+      "learning_rate": 4.3553553553553555e-05,
+      "loss": 0.8684,
+      "step": 1300
+    },
+    {
+      "epoch": 0.0017006469255676122,
+      "grad_norm": 2.74151873588562,
+      "learning_rate": 4.354854854854855e-05,
+      "loss": 1.6336,
+      "step": 1301
+    },
+    {
+      "epoch": 0.0017019541099838823,
+      "grad_norm": 1.5082085132598877,
+      "learning_rate": 4.354354354354355e-05,
+      "loss": 2.1466,
+      "step": 1302
+    },
+    {
+      "epoch": 0.0017032612944001527,
+      "grad_norm": 1.5993309020996094,
+      "learning_rate": 4.353853853853854e-05,
+      "loss": 2.0628,
+      "step": 1303
+    },
+    {
+      "epoch": 0.001704568478816423,
+      "grad_norm": 1.1165951490402222,
+      "learning_rate": 4.353353353353354e-05,
+      "loss": 1.1247,
+      "step": 1304
+    },
+    {
+      "epoch": 0.0017058756632326933,
+      "grad_norm": 1.8766840696334839,
+      "learning_rate": 4.352852852852853e-05,
+      "loss": 1.1355,
+      "step": 1305
+    },
+    {
+      "epoch": 0.0017071828476489634,
+      "grad_norm": 1.5404255390167236,
+      "learning_rate": 4.3523523523523527e-05,
+      "loss": 0.3954,
+      "step": 1306
+    },
+    {
+      "epoch": 0.0017084900320652337,
+      "grad_norm": 1.755220651626587,
+      "learning_rate": 4.351851851851852e-05,
+      "loss": 1.859,
+      "step": 1307
+    },
+    {
+      "epoch": 0.001709797216481504,
+      "grad_norm": 1.5398303270339966,
+      "learning_rate": 4.351351351351351e-05,
+      "loss": 2.1824,
+      "step": 1308
+    },
+    {
+      "epoch": 0.0017111044008977744,
+      "grad_norm": 1.2593640089035034,
+      "learning_rate": 4.350850850850851e-05,
+      "loss": 1.4959,
+      "step": 1309
+    },
+    {
+      "epoch": 0.0017124115853140445,
+      "grad_norm": 1.7014564275741577,
+      "learning_rate": 4.3503503503503505e-05,
+      "loss": 2.5335,
+      "step": 1310
+    },
+    {
+      "epoch": 0.0017137187697303148,
+      "grad_norm": 1.4378031492233276,
+      "learning_rate": 4.3498498498498505e-05,
+      "loss": 1.151,
+      "step": 1311
+    },
+    {
+      "epoch": 0.001715025954146585,
+      "grad_norm": 1.9308477640151978,
+      "learning_rate": 4.34934934934935e-05,
+      "loss": 2.2516,
+      "step": 1312
+    },
+    {
+      "epoch": 0.0017163331385628552,
+      "grad_norm": 1.8909333944320679,
+      "learning_rate": 4.348848848848849e-05,
+      "loss": 2.1757,
+      "step": 1313
+    },
+    {
+      "epoch": 0.0017176403229791255,
+      "grad_norm": 1.9359357357025146,
+      "learning_rate": 4.3483483483483484e-05,
+      "loss": 2.0771,
+      "step": 1314
+    },
+    {
+      "epoch": 0.0017189475073953958,
+      "grad_norm": 1.7034690380096436,
+      "learning_rate": 4.347847847847848e-05,
+      "loss": 1.3766,
+      "step": 1315
+    },
+    {
+      "epoch": 0.0017202546918116662,
+      "grad_norm": 1.8465443849563599,
+      "learning_rate": 4.3473473473473477e-05,
+      "loss": 1.3459,
+      "step": 1316
+    },
+    {
+      "epoch": 0.0017215618762279363,
+      "grad_norm": 2.009443759918213,
+      "learning_rate": 4.346846846846847e-05,
+      "loss": 2.5599,
+      "step": 1317
+    },
+    {
+      "epoch": 0.0017228690606442066,
+      "grad_norm": 1.5030558109283447,
+      "learning_rate": 4.346346346346347e-05,
+      "loss": 2.1631,
+      "step": 1318
+    },
+    {
+      "epoch": 0.001724176245060477,
+      "grad_norm": 1.5249741077423096,
+      "learning_rate": 4.345845845845846e-05,
+      "loss": 1.637,
+      "step": 1319
+    },
+    {
+      "epoch": 0.0017254834294767472,
+      "grad_norm": 1.3070316314697266,
+      "learning_rate": 4.3453453453453455e-05,
+      "loss": 1.7166,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0017267906138930173,
+      "grad_norm": 1.939177393913269,
+      "learning_rate": 4.3448448448448455e-05,
+      "loss": 1.7653,
+      "step": 1321
+    },
+    {
+      "epoch": 0.0017280977983092877,
+      "grad_norm": 1.9239637851715088,
+      "learning_rate": 4.344344344344344e-05,
+      "loss": 1.3714,
+      "step": 1322
+    },
+    {
+      "epoch": 0.001729404982725558,
+      "grad_norm": 1.3498599529266357,
+      "learning_rate": 4.343843843843844e-05,
+      "loss": 1.2912,
+      "step": 1323
+    },
+    {
+      "epoch": 0.0017307121671418283,
+      "grad_norm": 1.576553225517273,
+      "learning_rate": 4.3433433433433434e-05,
+      "loss": 1.6707,
+      "step": 1324
+    },
+    {
+      "epoch": 0.0017320193515580984,
+      "grad_norm": 1.6087493896484375,
+      "learning_rate": 4.3428428428428434e-05,
+      "loss": 1.6374,
+      "step": 1325
+    },
+    {
+      "epoch": 0.0017333265359743687,
+      "grad_norm": 1.3529077768325806,
+      "learning_rate": 4.342342342342343e-05,
+      "loss": 1.9405,
+      "step": 1326
+    },
+    {
+      "epoch": 0.001734633720390639,
+      "grad_norm": 1.178276777267456,
+      "learning_rate": 4.341841841841842e-05,
+      "loss": 1.5034,
+      "step": 1327
+    },
+    {
+      "epoch": 0.0017359409048069094,
+      "grad_norm": 1.374110460281372,
+      "learning_rate": 4.341341341341342e-05,
+      "loss": 1.3932,
+      "step": 1328
+    },
+    {
+      "epoch": 0.0017372480892231795,
+      "grad_norm": 0.5865709781646729,
+      "learning_rate": 4.340840840840841e-05,
+      "loss": 0.3236,
+      "step": 1329
+    },
+    {
+      "epoch": 0.0017385552736394498,
+      "grad_norm": 1.4669052362442017,
+      "learning_rate": 4.3403403403403405e-05,
+      "loss": 1.3216,
+      "step": 1330
+    },
+    {
+      "epoch": 0.00173986245805572,
+      "grad_norm": 1.5241165161132812,
+      "learning_rate": 4.33983983983984e-05,
+      "loss": 1.5953,
+      "step": 1331
+    },
+    {
+      "epoch": 0.0017411696424719902,
+      "grad_norm": 1.5230462551116943,
+      "learning_rate": 4.339339339339339e-05,
+      "loss": 2.3941,
+      "step": 1332
+    },
+    {
+      "epoch": 0.0017424768268882605,
+      "grad_norm": 1.2815715074539185,
+      "learning_rate": 4.338838838838839e-05,
+      "loss": 0.9748,
+      "step": 1333
+    },
+    {
+      "epoch": 0.0017437840113045308,
+      "grad_norm": 1.426518440246582,
+      "learning_rate": 4.3383383383383384e-05,
+      "loss": 1.4167,
+      "step": 1334
+    },
+    {
+      "epoch": 0.0017450911957208012,
+      "grad_norm": 1.4060072898864746,
+      "learning_rate": 4.3378378378378384e-05,
+      "loss": 2.0181,
+      "step": 1335
+    },
+    {
+      "epoch": 0.0017463983801370713,
+      "grad_norm": 1.6321356296539307,
+      "learning_rate": 4.337337337337338e-05,
+      "loss": 1.2248,
+      "step": 1336
+    },
+    {
+      "epoch": 0.0017477055645533416,
+      "grad_norm": 1.8750109672546387,
+      "learning_rate": 4.336836836836837e-05,
+      "loss": 1.2176,
+      "step": 1337
+    },
+    {
+      "epoch": 0.001749012748969612,
+      "grad_norm": 1.7221183776855469,
+      "learning_rate": 4.336336336336336e-05,
+      "loss": 1.2798,
+      "step": 1338
+    },
+    {
+      "epoch": 0.0017503199333858822,
+      "grad_norm": 1.6384902000427246,
+      "learning_rate": 4.3358358358358356e-05,
+      "loss": 2.1856,
+      "step": 1339
+    },
+    {
+      "epoch": 0.0017516271178021523,
+      "grad_norm": 1.0511401891708374,
+      "learning_rate": 4.3353353353353355e-05,
+      "loss": 0.8308,
+      "step": 1340
+    },
+    {
+      "epoch": 0.0017529343022184226,
+      "grad_norm": 1.385850191116333,
+      "learning_rate": 4.334834834834835e-05,
+      "loss": 1.6717,
+      "step": 1341
+    },
+    {
+      "epoch": 0.001754241486634693,
+      "grad_norm": 1.1377594470977783,
+      "learning_rate": 4.334334334334335e-05,
+      "loss": 0.8367,
+      "step": 1342
+    },
+    {
+      "epoch": 0.0017555486710509633,
+      "grad_norm": 1.6207880973815918,
+      "learning_rate": 4.333833833833834e-05,
+      "loss": 2.3777,
+      "step": 1343
+    },
+    {
+      "epoch": 0.0017568558554672334,
+      "grad_norm": 0.6226940751075745,
+      "learning_rate": 4.3333333333333334e-05,
+      "loss": 0.2113,
+      "step": 1344
+    },
+    {
+      "epoch": 0.0017581630398835037,
+      "grad_norm": 1.6958853006362915,
+      "learning_rate": 4.3328328328328334e-05,
+      "loss": 2.2018,
+      "step": 1345
+    },
+    {
+      "epoch": 0.001759470224299774,
+      "grad_norm": 1.9238884449005127,
+      "learning_rate": 4.332332332332332e-05,
+      "loss": 1.7379,
+      "step": 1346
+    },
+    {
+      "epoch": 0.0017607774087160443,
+      "grad_norm": 1.5163981914520264,
+      "learning_rate": 4.331831831831832e-05,
+      "loss": 2.1165,
+      "step": 1347
+    },
+    {
+      "epoch": 0.0017620845931323144,
+      "grad_norm": 1.0818721055984497,
+      "learning_rate": 4.331331331331331e-05,
+      "loss": 0.9938,
+      "step": 1348
+    },
+    {
+      "epoch": 0.0017633917775485848,
+      "grad_norm": 1.5910513401031494,
+      "learning_rate": 4.330830830830831e-05,
+      "loss": 1.3771,
+      "step": 1349
+    },
+    {
+      "epoch": 0.001764698961964855,
+      "grad_norm": 1.4074985980987549,
+      "learning_rate": 4.3303303303303305e-05,
+      "loss": 1.0621,
+      "step": 1350
+    },
+    {
+      "epoch": 0.0017660061463811254,
+      "grad_norm": 1.3553016185760498,
+      "learning_rate": 4.32982982982983e-05,
+      "loss": 2.1296,
+      "step": 1351
+    },
+    {
+      "epoch": 0.0017673133307973955,
+      "grad_norm": 1.5584310293197632,
+      "learning_rate": 4.32932932932933e-05,
+      "loss": 2.5648,
+      "step": 1352
+    },
+    {
+      "epoch": 0.0017686205152136658,
+      "grad_norm": 0.7152749300003052,
+      "learning_rate": 4.328828828828829e-05,
+      "loss": 0.2797,
+      "step": 1353
+    },
+    {
+      "epoch": 0.0017699276996299361,
+      "grad_norm": 1.3317090272903442,
+      "learning_rate": 4.328328328328329e-05,
+      "loss": 1.986,
+      "step": 1354
+    },
+    {
+      "epoch": 0.0017712348840462063,
+      "grad_norm": 1.4769171476364136,
+      "learning_rate": 4.327827827827828e-05,
+      "loss": 1.8507,
+      "step": 1355
+    },
+    {
+      "epoch": 0.0017725420684624766,
+      "grad_norm": 1.673120379447937,
+      "learning_rate": 4.327327327327328e-05,
+      "loss": 2.4266,
+      "step": 1356
+    },
+    {
+      "epoch": 0.0017738492528787469,
+      "grad_norm": 1.5308825969696045,
+      "learning_rate": 4.326826826826827e-05,
+      "loss": 1.8016,
+      "step": 1357
+    },
+    {
+      "epoch": 0.0017751564372950172,
+      "grad_norm": 1.5092765092849731,
+      "learning_rate": 4.326326326326326e-05,
+      "loss": 2.2371,
+      "step": 1358
+    },
+    {
+      "epoch": 0.0017764636217112873,
+      "grad_norm": 0.7928272485733032,
+      "learning_rate": 4.325825825825826e-05,
+      "loss": 0.3143,
+      "step": 1359
+    },
+    {
+      "epoch": 0.0017777708061275576,
+      "grad_norm": 1.7217007875442505,
+      "learning_rate": 4.3253253253253256e-05,
+      "loss": 1.8566,
+      "step": 1360
+    },
+    {
+      "epoch": 0.001779077990543828,
+      "grad_norm": 1.5367738008499146,
+      "learning_rate": 4.3248248248248255e-05,
+      "loss": 2.1112,
+      "step": 1361
+    },
+    {
+      "epoch": 0.0017803851749600983,
+      "grad_norm": 1.9571820497512817,
+      "learning_rate": 4.324324324324325e-05,
+      "loss": 1.7161,
+      "step": 1362
+    },
+    {
+      "epoch": 0.0017816923593763684,
+      "grad_norm": 2.2658941745758057,
+      "learning_rate": 4.3238238238238234e-05,
+      "loss": 1.8393,
+      "step": 1363
+    },
+    {
+      "epoch": 0.0017829995437926387,
+      "grad_norm": 1.0749531984329224,
+      "learning_rate": 4.3233233233233234e-05,
+      "loss": 0.9335,
+      "step": 1364
+    },
+    {
+      "epoch": 0.001784306728208909,
+      "grad_norm": 1.6484084129333496,
+      "learning_rate": 4.322822822822823e-05,
+      "loss": 1.6231,
+      "step": 1365
+    },
+    {
+      "epoch": 0.0017856139126251793,
+      "grad_norm": 1.666406512260437,
+      "learning_rate": 4.322322322322323e-05,
+      "loss": 2.0575,
+      "step": 1366
+    },
+    {
+      "epoch": 0.0017869210970414494,
+      "grad_norm": 1.4402440786361694,
+      "learning_rate": 4.321821821821822e-05,
+      "loss": 1.5258,
+      "step": 1367
+    },
+    {
+      "epoch": 0.0017882282814577198,
+      "grad_norm": 1.3725416660308838,
+      "learning_rate": 4.321321321321321e-05,
+      "loss": 1.4985,
+      "step": 1368
+    },
+    {
+      "epoch": 0.00178953546587399,
+      "grad_norm": 1.1884535551071167,
+      "learning_rate": 4.320820820820821e-05,
+      "loss": 0.8457,
+      "step": 1369
+    },
+    {
+      "epoch": 0.0017908426502902604,
+      "grad_norm": 1.7568581104278564,
+      "learning_rate": 4.3203203203203206e-05,
+      "loss": 1.0493,
+      "step": 1370
+    },
+    {
+      "epoch": 0.0017921498347065305,
+      "grad_norm": 1.7199748754501343,
+      "learning_rate": 4.31981981981982e-05,
+      "loss": 1.9267,
+      "step": 1371
+    },
+    {
+      "epoch": 0.0017934570191228008,
+      "grad_norm": 1.5979365110397339,
+      "learning_rate": 4.319319319319319e-05,
+      "loss": 1.1105,
+      "step": 1372
+    },
+    {
+      "epoch": 0.0017947642035390711,
+      "grad_norm": 1.730568289756775,
+      "learning_rate": 4.318818818818819e-05,
+      "loss": 2.1069,
+      "step": 1373
+    },
+    {
+      "epoch": 0.0017960713879553415,
+      "grad_norm": 1.3659794330596924,
+      "learning_rate": 4.3183183183183184e-05,
+      "loss": 1.3182,
+      "step": 1374
+    },
+    {
+      "epoch": 0.0017973785723716116,
+      "grad_norm": 1.362149953842163,
+      "learning_rate": 4.317817817817818e-05,
+      "loss": 1.4367,
+      "step": 1375
+    },
+    {
+      "epoch": 0.0017986857567878819,
+      "grad_norm": 1.8486407995224,
+      "learning_rate": 4.317317317317318e-05,
+      "loss": 1.8463,
+      "step": 1376
+    },
+    {
+      "epoch": 0.0017999929412041522,
+      "grad_norm": 1.3113247156143188,
+      "learning_rate": 4.316816816816817e-05,
+      "loss": 1.0199,
+      "step": 1377
+    },
+    {
+      "epoch": 0.0018013001256204223,
+      "grad_norm": 1.1706315279006958,
+      "learning_rate": 4.316316316316317e-05,
+      "loss": 1.1681,
+      "step": 1378
+    },
+    {
+      "epoch": 0.0018026073100366926,
+      "grad_norm": 1.6893227100372314,
+      "learning_rate": 4.315815815815816e-05,
+      "loss": 1.9949,
+      "step": 1379
+    },
+    {
+      "epoch": 0.001803914494452963,
+      "grad_norm": 1.4375907182693481,
+      "learning_rate": 4.3153153153153156e-05,
+      "loss": 0.8079,
+      "step": 1380
+    },
+    {
+      "epoch": 0.0018052216788692333,
+      "grad_norm": 1.8611098527908325,
+      "learning_rate": 4.314814814814815e-05,
+      "loss": 2.5563,
+      "step": 1381
+    },
+    {
+      "epoch": 0.0018065288632855034,
+      "grad_norm": 1.32766854763031,
+      "learning_rate": 4.314314314314314e-05,
+      "loss": 0.7402,
+      "step": 1382
+    },
+    {
+      "epoch": 0.0018078360477017737,
+      "grad_norm": 1.5074973106384277,
+      "learning_rate": 4.313813813813814e-05,
+      "loss": 1.0876,
+      "step": 1383
+    },
+    {
+      "epoch": 0.001809143232118044,
+      "grad_norm": 1.2465490102767944,
+      "learning_rate": 4.3133133133133134e-05,
+      "loss": 1.6936,
+      "step": 1384
+    },
+    {
+      "epoch": 0.0018104504165343143,
+      "grad_norm": 1.4665617942810059,
+      "learning_rate": 4.3128128128128134e-05,
+      "loss": 1.9762,
+      "step": 1385
+    },
+    {
+      "epoch": 0.0018117576009505844,
+      "grad_norm": 1.4028939008712769,
+      "learning_rate": 4.312312312312313e-05,
+      "loss": 2.0752,
+      "step": 1386
+    },
+    {
+      "epoch": 0.0018130647853668547,
+      "grad_norm": 1.1653519868850708,
+      "learning_rate": 4.311811811811812e-05,
+      "loss": 0.7552,
+      "step": 1387
+    },
+    {
+      "epoch": 0.001814371969783125,
+      "grad_norm": 1.295170783996582,
+      "learning_rate": 4.311311311311311e-05,
+      "loss": 1.8702,
+      "step": 1388
+    },
+    {
+      "epoch": 0.0018156791541993954,
+      "grad_norm": 1.3392369747161865,
+      "learning_rate": 4.3108108108108106e-05,
+      "loss": 1.3585,
+      "step": 1389
+    },
+    {
+      "epoch": 0.0018169863386156655,
+      "grad_norm": 0.8655954003334045,
+      "learning_rate": 4.3103103103103106e-05,
+      "loss": 0.3256,
+      "step": 1390
+    },
+    {
+      "epoch": 0.0018182935230319358,
+      "grad_norm": 1.3352432250976562,
+      "learning_rate": 4.30980980980981e-05,
+      "loss": 2.382,
+      "step": 1391
+    },
+    {
+      "epoch": 0.0018196007074482061,
+      "grad_norm": 1.5272313356399536,
+      "learning_rate": 4.30930930930931e-05,
+      "loss": 1.7232,
+      "step": 1392
+    },
+    {
+      "epoch": 0.0018209078918644764,
+      "grad_norm": 1.8199968338012695,
+      "learning_rate": 4.308808808808809e-05,
+      "loss": 2.4559,
+      "step": 1393
+    },
+    {
+      "epoch": 0.0018222150762807466,
+      "grad_norm": 1.4391125440597534,
+      "learning_rate": 4.3083083083083084e-05,
+      "loss": 1.0684,
+      "step": 1394
+    },
+    {
+      "epoch": 0.0018235222606970169,
+      "grad_norm": 1.269695520401001,
+      "learning_rate": 4.3078078078078084e-05,
+      "loss": 1.7451,
+      "step": 1395
+    },
+    {
+      "epoch": 0.0018248294451132872,
+      "grad_norm": 1.411726713180542,
+      "learning_rate": 4.307307307307307e-05,
+      "loss": 1.4257,
+      "step": 1396
+    },
+    {
+      "epoch": 0.0018261366295295573,
+      "grad_norm": 1.3439668416976929,
+      "learning_rate": 4.306806806806807e-05,
+      "loss": 1.4274,
+      "step": 1397
+    },
+    {
+      "epoch": 0.0018274438139458276,
+      "grad_norm": 1.4007078409194946,
+      "learning_rate": 4.306306306306306e-05,
+      "loss": 1.2485,
+      "step": 1398
+    },
+    {
+      "epoch": 0.001828750998362098,
+      "grad_norm": 1.5390173196792603,
+      "learning_rate": 4.305805805805806e-05,
+      "loss": 1.8902,
+      "step": 1399
+    },
+    {
+      "epoch": 0.0018300581827783683,
+      "grad_norm": 1.3821252584457397,
+      "learning_rate": 4.3053053053053056e-05,
+      "loss": 1.8779,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0018313653671946384,
+      "grad_norm": 1.1391609907150269,
+      "learning_rate": 4.304804804804805e-05,
+      "loss": 0.9036,
+      "step": 1401
+    },
+    {
+      "epoch": 0.0018326725516109087,
+      "grad_norm": 1.6939865350723267,
+      "learning_rate": 4.304304304304305e-05,
+      "loss": 0.9605,
+      "step": 1402
+    },
+    {
+      "epoch": 0.001833979736027179,
+      "grad_norm": 1.2148393392562866,
+      "learning_rate": 4.303803803803804e-05,
+      "loss": 1.2693,
+      "step": 1403
+    },
+    {
+      "epoch": 0.0018352869204434493,
+      "grad_norm": 1.4309426546096802,
+      "learning_rate": 4.3033033033033034e-05,
+      "loss": 1.4887,
+      "step": 1404
+    },
+    {
+      "epoch": 0.0018365941048597194,
+      "grad_norm": 1.552206039428711,
+      "learning_rate": 4.302802802802803e-05,
+      "loss": 1.9924,
+      "step": 1405
+    },
+    {
+      "epoch": 0.0018379012892759897,
+      "grad_norm": 1.5694180727005005,
+      "learning_rate": 4.302302302302302e-05,
+      "loss": 1.7557,
+      "step": 1406
+    },
+    {
+      "epoch": 0.00183920847369226,
+      "grad_norm": 1.3013912439346313,
+      "learning_rate": 4.301801801801802e-05,
+      "loss": 1.487,
+      "step": 1407
+    },
+    {
+      "epoch": 0.0018405156581085304,
+      "grad_norm": 2.117809772491455,
+      "learning_rate": 4.301301301301301e-05,
+      "loss": 2.7287,
+      "step": 1408
+    },
+    {
+      "epoch": 0.0018418228425248005,
+      "grad_norm": 1.9013354778289795,
+      "learning_rate": 4.300800800800801e-05,
+      "loss": 2.6201,
+      "step": 1409
+    },
+    {
+      "epoch": 0.0018431300269410708,
+      "grad_norm": 1.445999026298523,
+      "learning_rate": 4.3003003003003006e-05,
+      "loss": 1.6531,
+      "step": 1410
+    },
+    {
+      "epoch": 0.0018444372113573411,
+      "grad_norm": 1.3536559343338013,
+      "learning_rate": 4.2997997997998e-05,
+      "loss": 1.4582,
+      "step": 1411
+    },
+    {
+      "epoch": 0.0018457443957736114,
+      "grad_norm": 1.5885568857192993,
+      "learning_rate": 4.2992992992993e-05,
+      "loss": 1.7563,
+      "step": 1412
+    },
+    {
+      "epoch": 0.0018470515801898815,
+      "grad_norm": 1.2025485038757324,
+      "learning_rate": 4.2987987987987985e-05,
+      "loss": 1.0646,
+      "step": 1413
+    },
+    {
+      "epoch": 0.0018483587646061519,
+      "grad_norm": 1.1941187381744385,
+      "learning_rate": 4.2982982982982985e-05,
+      "loss": 1.1471,
+      "step": 1414
+    },
+    {
+      "epoch": 0.0018496659490224222,
+      "grad_norm": 1.4847627878189087,
+      "learning_rate": 4.297797797797798e-05,
+      "loss": 2.4284,
+      "step": 1415
+    },
+    {
+      "epoch": 0.0018509731334386925,
+      "grad_norm": 1.4225730895996094,
+      "learning_rate": 4.297297297297298e-05,
+      "loss": 1.0716,
+      "step": 1416
+    },
+    {
+      "epoch": 0.0018522803178549626,
+      "grad_norm": 1.72896409034729,
+      "learning_rate": 4.296796796796797e-05,
+      "loss": 1.4967,
+      "step": 1417
+    },
+    {
+      "epoch": 0.001853587502271233,
+      "grad_norm": 0.8089171648025513,
+      "learning_rate": 4.296296296296296e-05,
+      "loss": 0.3262,
+      "step": 1418
+    },
+    {
+      "epoch": 0.0018548946866875032,
+      "grad_norm": 1.1190005540847778,
+      "learning_rate": 4.295795795795796e-05,
+      "loss": 1.8598,
+      "step": 1419
+    },
+    {
+      "epoch": 0.0018562018711037733,
+      "grad_norm": 1.3130826950073242,
+      "learning_rate": 4.2952952952952956e-05,
+      "loss": 1.6531,
+      "step": 1420
+    },
+    {
+      "epoch": 0.0018575090555200437,
+      "grad_norm": 1.4470422267913818,
+      "learning_rate": 4.294794794794795e-05,
+      "loss": 1.7035,
+      "step": 1421
+    },
+    {
+      "epoch": 0.001858816239936314,
+      "grad_norm": 1.2675118446350098,
+      "learning_rate": 4.294294294294294e-05,
+      "loss": 0.6884,
+      "step": 1422
+    },
+    {
+      "epoch": 0.0018601234243525843,
+      "grad_norm": 1.331605076789856,
+      "learning_rate": 4.293793793793794e-05,
+      "loss": 1.7179,
+      "step": 1423
+    },
+    {
+      "epoch": 0.0018614306087688544,
+      "grad_norm": 1.4828003644943237,
+      "learning_rate": 4.2932932932932935e-05,
+      "loss": 2.3641,
+      "step": 1424
+    },
+    {
+      "epoch": 0.0018627377931851247,
+      "grad_norm": 1.263620376586914,
+      "learning_rate": 4.292792792792793e-05,
+      "loss": 1.8103,
+      "step": 1425
+    },
+    {
+      "epoch": 0.001864044977601395,
+      "grad_norm": 1.2304747104644775,
+      "learning_rate": 4.292292292292293e-05,
+      "loss": 1.4039,
+      "step": 1426
+    },
+    {
+      "epoch": 0.0018653521620176654,
+      "grad_norm": 1.4697588682174683,
+      "learning_rate": 4.291791791791792e-05,
+      "loss": 2.4613,
+      "step": 1427
+    },
+    {
+      "epoch": 0.0018666593464339355,
+      "grad_norm": 1.4634438753128052,
+      "learning_rate": 4.291291291291292e-05,
+      "loss": 1.7761,
+      "step": 1428
+    },
+    {
+      "epoch": 0.0018679665308502058,
+      "grad_norm": 1.4904392957687378,
+      "learning_rate": 4.290790790790791e-05,
+      "loss": 1.4831,
+      "step": 1429
+    },
+    {
+      "epoch": 0.0018692737152664761,
+      "grad_norm": 1.358394980430603,
+      "learning_rate": 4.2902902902902906e-05,
+      "loss": 2.2373,
+      "step": 1430
+    },
+    {
+      "epoch": 0.0018705808996827464,
+      "grad_norm": 0.9131472110748291,
+      "learning_rate": 4.28978978978979e-05,
+      "loss": 0.6586,
+      "step": 1431
+    },
+    {
+      "epoch": 0.0018718880840990165,
+      "grad_norm": 1.2101187705993652,
+      "learning_rate": 4.289289289289289e-05,
+      "loss": 1.3421,
+      "step": 1432
+    },
+    {
+      "epoch": 0.0018731952685152869,
+      "grad_norm": 1.4582985639572144,
+      "learning_rate": 4.288788788788789e-05,
+      "loss": 2.089,
+      "step": 1433
+    },
+    {
+      "epoch": 0.0018745024529315572,
+      "grad_norm": 1.530697226524353,
+      "learning_rate": 4.2882882882882885e-05,
+      "loss": 2.0292,
+      "step": 1434
+    },
+    {
+      "epoch": 0.0018758096373478275,
+      "grad_norm": 1.5531808137893677,
+      "learning_rate": 4.2877877877877884e-05,
+      "loss": 1.7257,
+      "step": 1435
+    },
+    {
+      "epoch": 0.0018771168217640976,
+      "grad_norm": 2.0334112644195557,
+      "learning_rate": 4.287287287287288e-05,
+      "loss": 2.2395,
+      "step": 1436
+    },
+    {
+      "epoch": 0.001878424006180368,
+      "grad_norm": 1.5652258396148682,
+      "learning_rate": 4.286786786786787e-05,
+      "loss": 1.6322,
+      "step": 1437
+    },
+    {
+      "epoch": 0.0018797311905966382,
+      "grad_norm": 1.249422311782837,
+      "learning_rate": 4.286286286286286e-05,
+      "loss": 1.3683,
+      "step": 1438
+    },
+    {
+      "epoch": 0.0018810383750129083,
+      "grad_norm": 1.2398805618286133,
+      "learning_rate": 4.2857857857857856e-05,
+      "loss": 1.2385,
+      "step": 1439
+    },
+    {
+      "epoch": 0.0018823455594291787,
+      "grad_norm": 2.8785653114318848,
+      "learning_rate": 4.2852852852852856e-05,
+      "loss": 1.2533,
+      "step": 1440
+    },
+    {
+      "epoch": 0.001883652743845449,
+      "grad_norm": 1.5326817035675049,
+      "learning_rate": 4.284784784784785e-05,
+      "loss": 1.6618,
+      "step": 1441
+    },
+    {
+      "epoch": 0.0018849599282617193,
+      "grad_norm": 1.6503008604049683,
+      "learning_rate": 4.284284284284284e-05,
+      "loss": 2.5617,
+      "step": 1442
+    },
+    {
+      "epoch": 0.0018862671126779894,
+      "grad_norm": 1.3797667026519775,
+      "learning_rate": 4.283783783783784e-05,
+      "loss": 1.3771,
+      "step": 1443
+    },
+    {
+      "epoch": 0.0018875742970942597,
+      "grad_norm": 1.945457935333252,
+      "learning_rate": 4.2832832832832835e-05,
+      "loss": 0.4182,
+      "step": 1444
+    },
+    {
+      "epoch": 0.00188888148151053,
+      "grad_norm": 1.4646294116973877,
+      "learning_rate": 4.2827827827827835e-05,
+      "loss": 1.1078,
+      "step": 1445
+    },
+    {
+      "epoch": 0.0018901886659268004,
+      "grad_norm": 1.436588168144226,
+      "learning_rate": 4.282282282282282e-05,
+      "loss": 1.3253,
+      "step": 1446
+    },
+    {
+      "epoch": 0.0018914958503430705,
+      "grad_norm": 1.5396558046340942,
+      "learning_rate": 4.281781781781782e-05,
+      "loss": 1.2812,
+      "step": 1447
+    },
+    {
+      "epoch": 0.0018928030347593408,
+      "grad_norm": 1.5310207605361938,
+      "learning_rate": 4.2812812812812813e-05,
+      "loss": 1.3051,
+      "step": 1448
+    },
+    {
+      "epoch": 0.001894110219175611,
+      "grad_norm": 1.247341513633728,
+      "learning_rate": 4.2807807807807806e-05,
+      "loss": 1.5743,
+      "step": 1449
+    },
+    {
+      "epoch": 0.0018954174035918814,
+      "grad_norm": 1.5774229764938354,
+      "learning_rate": 4.2802802802802806e-05,
+      "loss": 1.4431,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0018967245880081515,
+      "grad_norm": 1.3610193729400635,
+      "learning_rate": 4.27977977977978e-05,
+      "loss": 1.2685,
+      "step": 1451
+    },
+    {
+      "epoch": 0.0018980317724244218,
+      "grad_norm": 1.5870070457458496,
+      "learning_rate": 4.27927927927928e-05,
+      "loss": 2.2065,
+      "step": 1452
+    },
+    {
+      "epoch": 0.0018993389568406922,
+      "grad_norm": 1.4996387958526611,
+      "learning_rate": 4.278778778778779e-05,
+      "loss": 2.0389,
+      "step": 1453
+    },
+    {
+      "epoch": 0.0019006461412569625,
+      "grad_norm": 1.431861400604248,
+      "learning_rate": 4.2782782782782785e-05,
+      "loss": 2.2997,
+      "step": 1454
+    },
+    {
+      "epoch": 0.0019019533256732326,
+      "grad_norm": 1.586350679397583,
+      "learning_rate": 4.277777777777778e-05,
+      "loss": 1.4156,
+      "step": 1455
+    },
+    {
+      "epoch": 0.001903260510089503,
+      "grad_norm": 0.7952925562858582,
+      "learning_rate": 4.277277277277277e-05,
+      "loss": 0.4465,
+      "step": 1456
+    },
+    {
+      "epoch": 0.0019045676945057732,
+      "grad_norm": 1.2405833005905151,
+      "learning_rate": 4.276776776776777e-05,
+      "loss": 1.3952,
+      "step": 1457
+    },
+    {
+      "epoch": 0.0019058748789220435,
+      "grad_norm": 0.6673822402954102,
+      "learning_rate": 4.2762762762762763e-05,
+      "loss": 0.2461,
+      "step": 1458
+    },
+    {
+      "epoch": 0.0019071820633383136,
+      "grad_norm": 1.347853660583496,
+      "learning_rate": 4.275775775775776e-05,
+      "loss": 2.4171,
+      "step": 1459
+    },
+    {
+      "epoch": 0.001908489247754584,
+      "grad_norm": 1.4121501445770264,
+      "learning_rate": 4.2752752752752756e-05,
+      "loss": 0.8398,
+      "step": 1460
+    },
+    {
+      "epoch": 0.0019097964321708543,
+      "grad_norm": 1.3706833124160767,
+      "learning_rate": 4.274774774774775e-05,
+      "loss": 1.7066,
+      "step": 1461
+    },
+    {
+      "epoch": 0.0019111036165871244,
+      "grad_norm": 1.2938553094863892,
+      "learning_rate": 4.274274274274275e-05,
+      "loss": 1.4554,
+      "step": 1462
+    },
+    {
+      "epoch": 0.0019124108010033947,
+      "grad_norm": 1.5075329542160034,
+      "learning_rate": 4.2737737737737735e-05,
+      "loss": 1.5185,
+      "step": 1463
+    },
+    {
+      "epoch": 0.001913717985419665,
+      "grad_norm": 1.165055274963379,
+      "learning_rate": 4.2732732732732735e-05,
+      "loss": 0.8594,
+      "step": 1464
+    },
+    {
+      "epoch": 0.0019150251698359353,
+      "grad_norm": 1.6534725427627563,
+      "learning_rate": 4.272772772772773e-05,
+      "loss": 2.1566,
+      "step": 1465
+    },
+    {
+      "epoch": 0.0019163323542522055,
+      "grad_norm": 1.9862817525863647,
+      "learning_rate": 4.272272272272273e-05,
+      "loss": 1.9241,
+      "step": 1466
+    },
+    {
+      "epoch": 0.0019176395386684758,
+      "grad_norm": 1.4614254236221313,
+      "learning_rate": 4.271771771771772e-05,
+      "loss": 1.6539,
+      "step": 1467
+    },
+    {
+      "epoch": 0.001918946723084746,
+      "grad_norm": 1.565486192703247,
+      "learning_rate": 4.2712712712712714e-05,
+      "loss": 2.4318,
+      "step": 1468
+    },
+    {
+      "epoch": 0.0019202539075010164,
+      "grad_norm": 1.9500453472137451,
+      "learning_rate": 4.270770770770771e-05,
+      "loss": 1.8941,
+      "step": 1469
+    },
+    {
+      "epoch": 0.0019215610919172865,
+      "grad_norm": 1.321319818496704,
+      "learning_rate": 4.2702702702702706e-05,
+      "loss": 1.0335,
+      "step": 1470
+    },
+    {
+      "epoch": 0.0019228682763335568,
+      "grad_norm": 1.7462786436080933,
+      "learning_rate": 4.26976976976977e-05,
+      "loss": 1.1837,
+      "step": 1471
+    },
+    {
+      "epoch": 0.0019241754607498272,
+      "grad_norm": 1.736907720565796,
+      "learning_rate": 4.269269269269269e-05,
+      "loss": 1.6386,
+      "step": 1472
+    },
+    {
+      "epoch": 0.0019254826451660975,
+      "grad_norm": 1.13442862033844,
+      "learning_rate": 4.2687687687687685e-05,
+      "loss": 0.9867,
+      "step": 1473
+    },
+    {
+      "epoch": 0.0019267898295823676,
+      "grad_norm": 1.7142646312713623,
+      "learning_rate": 4.2682682682682685e-05,
+      "loss": 1.9046,
+      "step": 1474
+    },
+    {
+      "epoch": 0.001928097013998638,
+      "grad_norm": 1.5673025846481323,
+      "learning_rate": 4.267767767767768e-05,
+      "loss": 1.3969,
+      "step": 1475
+    },
+    {
+      "epoch": 0.0019294041984149082,
+      "grad_norm": 1.5808483362197876,
+      "learning_rate": 4.267267267267268e-05,
+      "loss": 2.0955,
+      "step": 1476
+    },
+    {
+      "epoch": 0.0019307113828311785,
+      "grad_norm": 1.1170554161071777,
+      "learning_rate": 4.266766766766767e-05,
+      "loss": 1.4634,
+      "step": 1477
+    },
+    {
+      "epoch": 0.0019320185672474486,
+      "grad_norm": 1.804661512374878,
+      "learning_rate": 4.2662662662662664e-05,
+      "loss": 1.8844,
+      "step": 1478
+    },
+    {
+      "epoch": 0.001933325751663719,
+      "grad_norm": 1.494469404220581,
+      "learning_rate": 4.2657657657657657e-05,
+      "loss": 1.182,
+      "step": 1479
+    },
+    {
+      "epoch": 0.0019346329360799893,
+      "grad_norm": 1.882277250289917,
+      "learning_rate": 4.265265265265265e-05,
+      "loss": 1.9901,
+      "step": 1480
+    },
+    {
+      "epoch": 0.0019359401204962596,
+      "grad_norm": 1.6744542121887207,
+      "learning_rate": 4.264764764764765e-05,
+      "loss": 2.0032,
+      "step": 1481
+    },
+    {
+      "epoch": 0.0019372473049125297,
+      "grad_norm": 2.1882741451263428,
+      "learning_rate": 4.264264264264264e-05,
+      "loss": 1.0468,
+      "step": 1482
+    },
+    {
+      "epoch": 0.0019385544893288,
+      "grad_norm": 1.4654386043548584,
+      "learning_rate": 4.263763763763764e-05,
+      "loss": 1.737,
+      "step": 1483
+    },
+    {
+      "epoch": 0.0019398616737450703,
+      "grad_norm": 2.03692364692688,
+      "learning_rate": 4.2632632632632635e-05,
+      "loss": 2.329,
+      "step": 1484
+    },
+    {
+      "epoch": 0.0019411688581613404,
+      "grad_norm": 1.2058316469192505,
+      "learning_rate": 4.262762762762763e-05,
+      "loss": 0.9541,
+      "step": 1485
+    },
+    {
+      "epoch": 0.0019424760425776108,
+      "grad_norm": 1.7557787895202637,
+      "learning_rate": 4.262262262262263e-05,
+      "loss": 1.5912,
+      "step": 1486
+    },
+    {
+      "epoch": 0.001943783226993881,
+      "grad_norm": 1.5910978317260742,
+      "learning_rate": 4.261761761761762e-05,
+      "loss": 1.9267,
+      "step": 1487
+    },
+    {
+      "epoch": 0.0019450904114101514,
+      "grad_norm": 1.3717889785766602,
+      "learning_rate": 4.2612612612612614e-05,
+      "loss": 1.4328,
+      "step": 1488
+    },
+    {
+      "epoch": 0.0019463975958264215,
+      "grad_norm": 1.0328174829483032,
+      "learning_rate": 4.260760760760761e-05,
+      "loss": 0.7887,
+      "step": 1489
+    },
+    {
+      "epoch": 0.0019477047802426918,
+      "grad_norm": 1.3211932182312012,
+      "learning_rate": 4.2602602602602606e-05,
+      "loss": 1.0902,
+      "step": 1490
+    },
+    {
+      "epoch": 0.0019490119646589621,
+      "grad_norm": 1.3311086893081665,
+      "learning_rate": 4.25975975975976e-05,
+      "loss": 1.5207,
+      "step": 1491
+    },
+    {
+      "epoch": 0.0019503191490752325,
+      "grad_norm": 1.1878236532211304,
+      "learning_rate": 4.259259259259259e-05,
+      "loss": 1.934,
+      "step": 1492
+    },
+    {
+      "epoch": 0.0019516263334915026,
+      "grad_norm": 1.5886273384094238,
+      "learning_rate": 4.258758758758759e-05,
+      "loss": 1.7925,
+      "step": 1493
+    },
+    {
+      "epoch": 0.0019529335179077729,
+      "grad_norm": 1.430269718170166,
+      "learning_rate": 4.2582582582582585e-05,
+      "loss": 1.8998,
+      "step": 1494
+    },
+    {
+      "epoch": 0.001954240702324043,
+      "grad_norm": 1.321119785308838,
+      "learning_rate": 4.2577577577577585e-05,
+      "loss": 1.3459,
+      "step": 1495
+    },
+    {
+      "epoch": 0.0019555478867403133,
+      "grad_norm": 1.2092429399490356,
+      "learning_rate": 4.257257257257257e-05,
+      "loss": 1.4345,
+      "step": 1496
+    },
+    {
+      "epoch": 0.0019568550711565836,
+      "grad_norm": 1.5383827686309814,
+      "learning_rate": 4.256756756756757e-05,
+      "loss": 1.7497,
+      "step": 1497
+    },
+    {
+      "epoch": 0.001958162255572854,
+      "grad_norm": 1.990447759628296,
+      "learning_rate": 4.2562562562562564e-05,
+      "loss": 0.5758,
+      "step": 1498
+    },
+    {
+      "epoch": 0.0019594694399891243,
+      "grad_norm": 1.1384212970733643,
+      "learning_rate": 4.255755755755756e-05,
+      "loss": 0.9921,
+      "step": 1499
+    },
+    {
+      "epoch": 0.0019607766244053946,
+      "grad_norm": 1.3332409858703613,
+      "learning_rate": 4.2552552552552556e-05,
+      "loss": 1.5759,
+      "step": 1500
+    },
+    {
+      "epoch": 0.001962083808821665,
+      "grad_norm": 1.657772421836853,
+      "learning_rate": 4.254754754754755e-05,
+      "loss": 2.2981,
+      "step": 1501
+    },
+    {
+      "epoch": 0.0019633909932379352,
+      "grad_norm": 1.3510606288909912,
+      "learning_rate": 4.254254254254255e-05,
+      "loss": 1.7898,
+      "step": 1502
+    },
+    {
+      "epoch": 0.001964698177654205,
+      "grad_norm": 1.5047329664230347,
+      "learning_rate": 4.253753753753754e-05,
+      "loss": 2.3882,
+      "step": 1503
+    },
+    {
+      "epoch": 0.0019660053620704754,
+      "grad_norm": 0.4834231734275818,
+      "learning_rate": 4.2532532532532535e-05,
+      "loss": 0.1744,
+      "step": 1504
+    },
+    {
+      "epoch": 0.0019673125464867458,
+      "grad_norm": 1.4605001211166382,
+      "learning_rate": 4.252752752752753e-05,
+      "loss": 1.4694,
+      "step": 1505
+    },
+    {
+      "epoch": 0.001968619730903016,
+      "grad_norm": 1.6646137237548828,
+      "learning_rate": 4.252252252252252e-05,
+      "loss": 1.4732,
+      "step": 1506
+    },
+    {
+      "epoch": 0.0019699269153192864,
+      "grad_norm": 1.8789254426956177,
+      "learning_rate": 4.251751751751752e-05,
+      "loss": 1.511,
+      "step": 1507
+    },
+    {
+      "epoch": 0.0019712340997355567,
+      "grad_norm": 1.6691675186157227,
+      "learning_rate": 4.2512512512512514e-05,
+      "loss": 1.9486,
+      "step": 1508
+    },
+    {
+      "epoch": 0.001972541284151827,
+      "grad_norm": 1.3629869222640991,
+      "learning_rate": 4.2507507507507514e-05,
+      "loss": 1.1078,
+      "step": 1509
+    },
+    {
+      "epoch": 0.001973848468568097,
+      "grad_norm": 2.598057508468628,
+      "learning_rate": 4.2502502502502507e-05,
+      "loss": 1.3169,
+      "step": 1510
+    },
+    {
+      "epoch": 0.0019751556529843672,
+      "grad_norm": 1.5059863328933716,
+      "learning_rate": 4.24974974974975e-05,
+      "loss": 1.7909,
+      "step": 1511
+    },
+    {
+      "epoch": 0.0019764628374006376,
+      "grad_norm": 1.3698683977127075,
+      "learning_rate": 4.24924924924925e-05,
+      "loss": 1.7412,
+      "step": 1512
+    },
+    {
+      "epoch": 0.001977770021816908,
+      "grad_norm": 1.7377406358718872,
+      "learning_rate": 4.2487487487487485e-05,
+      "loss": 1.8956,
+      "step": 1513
+    },
+    {
+      "epoch": 0.001979077206233178,
+      "grad_norm": 1.6791915893554688,
+      "learning_rate": 4.2482482482482485e-05,
+      "loss": 1.9273,
+      "step": 1514
+    },
+    {
+      "epoch": 0.0019803843906494485,
+      "grad_norm": 1.6627063751220703,
+      "learning_rate": 4.247747747747748e-05,
+      "loss": 0.764,
+      "step": 1515
+    },
+    {
+      "epoch": 0.001981691575065719,
+      "grad_norm": 1.6537545919418335,
+      "learning_rate": 4.247247247247247e-05,
+      "loss": 1.9555,
+      "step": 1516
+    },
+    {
+      "epoch": 0.001982998759481989,
+      "grad_norm": 1.8646193742752075,
+      "learning_rate": 4.246746746746747e-05,
+      "loss": 1.7097,
+      "step": 1517
+    },
+    {
+      "epoch": 0.001984305943898259,
+      "grad_norm": 1.2620574235916138,
+      "learning_rate": 4.2462462462462464e-05,
+      "loss": 1.1305,
+      "step": 1518
+    },
+    {
+      "epoch": 0.0019856131283145294,
+      "grad_norm": 1.2493481636047363,
+      "learning_rate": 4.2457457457457464e-05,
+      "loss": 1.5161,
+      "step": 1519
+    },
+    {
+      "epoch": 0.0019869203127307997,
+      "grad_norm": 1.4736047983169556,
+      "learning_rate": 4.245245245245246e-05,
+      "loss": 2.1012,
+      "step": 1520
+    },
+    {
+      "epoch": 0.00198822749714707,
+      "grad_norm": 1.6359034776687622,
+      "learning_rate": 4.244744744744745e-05,
+      "loss": 1.8516,
+      "step": 1521
+    },
+    {
+      "epoch": 0.0019895346815633403,
+      "grad_norm": 1.6401679515838623,
+      "learning_rate": 4.244244244244244e-05,
+      "loss": 1.257,
+      "step": 1522
+    },
+    {
+      "epoch": 0.0019908418659796106,
+      "grad_norm": 1.313605785369873,
+      "learning_rate": 4.2437437437437436e-05,
+      "loss": 1.5808,
+      "step": 1523
+    },
+    {
+      "epoch": 0.001992149050395881,
+      "grad_norm": 1.5139002799987793,
+      "learning_rate": 4.2432432432432435e-05,
+      "loss": 1.3616,
+      "step": 1524
+    },
+    {
+      "epoch": 0.001993456234812151,
+      "grad_norm": 3.3174641132354736,
+      "learning_rate": 4.242742742742743e-05,
+      "loss": 1.1115,
+      "step": 1525
+    },
+    {
+      "epoch": 0.001994763419228421,
+      "grad_norm": 1.9150266647338867,
+      "learning_rate": 4.242242242242243e-05,
+      "loss": 2.1945,
+      "step": 1526
+    },
+    {
+      "epoch": 0.0019960706036446915,
+      "grad_norm": 1.8043403625488281,
+      "learning_rate": 4.241741741741742e-05,
+      "loss": 1.3139,
+      "step": 1527
+    },
+    {
+      "epoch": 0.001997377788060962,
+      "grad_norm": 1.2553147077560425,
+      "learning_rate": 4.2412412412412414e-05,
+      "loss": 1.376,
+      "step": 1528
+    },
+    {
+      "epoch": 0.001998684972477232,
+      "grad_norm": 1.8379456996917725,
+      "learning_rate": 4.240740740740741e-05,
+      "loss": 2.2553,
+      "step": 1529
+    },
+    {
+      "epoch": 0.0019999921568935024,
+      "grad_norm": 1.485284686088562,
+      "learning_rate": 4.24024024024024e-05,
+      "loss": 1.477,
+      "step": 1530
+    },
+    {
+      "epoch": 0.0020012993413097728,
+      "grad_norm": 1.2125955820083618,
+      "learning_rate": 4.23973973973974e-05,
+      "loss": 1.1023,
+      "step": 1531
+    },
+    {
+      "epoch": 0.002002606525726043,
+      "grad_norm": 1.6815204620361328,
+      "learning_rate": 4.239239239239239e-05,
+      "loss": 2.1045,
+      "step": 1532
+    },
+    {
+      "epoch": 0.002003913710142313,
+      "grad_norm": 1.2421667575836182,
+      "learning_rate": 4.238738738738739e-05,
+      "loss": 1.3327,
+      "step": 1533
+    },
+    {
+      "epoch": 0.0020052208945585833,
+      "grad_norm": 0.9187124967575073,
+      "learning_rate": 4.2382382382382385e-05,
+      "loss": 0.8634,
+      "step": 1534
+    },
+    {
+      "epoch": 0.0020065280789748536,
+      "grad_norm": 1.307012677192688,
+      "learning_rate": 4.237737737737738e-05,
+      "loss": 1.4249,
+      "step": 1535
+    },
+    {
+      "epoch": 0.002007835263391124,
+      "grad_norm": 1.9232230186462402,
+      "learning_rate": 4.237237237237238e-05,
+      "loss": 1.7143,
+      "step": 1536
+    },
+    {
+      "epoch": 0.0020091424478073943,
+      "grad_norm": 1.4440085887908936,
+      "learning_rate": 4.236736736736737e-05,
+      "loss": 0.8506,
+      "step": 1537
+    },
+    {
+      "epoch": 0.0020104496322236646,
+      "grad_norm": 1.3741660118103027,
+      "learning_rate": 4.2362362362362364e-05,
+      "loss": 1.8211,
+      "step": 1538
+    },
+    {
+      "epoch": 0.002011756816639935,
+      "grad_norm": 1.9500904083251953,
+      "learning_rate": 4.235735735735736e-05,
+      "loss": 2.1474,
+      "step": 1539
+    },
+    {
+      "epoch": 0.002013064001056205,
+      "grad_norm": 1.7212913036346436,
+      "learning_rate": 4.235235235235236e-05,
+      "loss": 1.2252,
+      "step": 1540
+    },
+    {
+      "epoch": 0.002014371185472475,
+      "grad_norm": 1.3449397087097168,
+      "learning_rate": 4.234734734734735e-05,
+      "loss": 1.3493,
+      "step": 1541
+    },
+    {
+      "epoch": 0.0020156783698887454,
+      "grad_norm": 1.8118581771850586,
+      "learning_rate": 4.234234234234234e-05,
+      "loss": 2.0876,
+      "step": 1542
+    },
+    {
+      "epoch": 0.0020169855543050157,
+      "grad_norm": 1.5977129936218262,
+      "learning_rate": 4.233733733733734e-05,
+      "loss": 1.5148,
+      "step": 1543
+    },
+    {
+      "epoch": 0.002018292738721286,
+      "grad_norm": 1.248093843460083,
+      "learning_rate": 4.2332332332332335e-05,
+      "loss": 1.2295,
+      "step": 1544
+    },
+    {
+      "epoch": 0.0020195999231375564,
+      "grad_norm": 1.273603916168213,
+      "learning_rate": 4.2327327327327335e-05,
+      "loss": 1.076,
+      "step": 1545
+    },
+    {
+      "epoch": 0.0020209071075538267,
+      "grad_norm": 1.5577911138534546,
+      "learning_rate": 4.232232232232232e-05,
+      "loss": 1.8095,
+      "step": 1546
+    },
+    {
+      "epoch": 0.002022214291970097,
+      "grad_norm": 1.2679510116577148,
+      "learning_rate": 4.2317317317317314e-05,
+      "loss": 1.6844,
+      "step": 1547
+    },
+    {
+      "epoch": 0.002023521476386367,
+      "grad_norm": 0.7545862197875977,
+      "learning_rate": 4.2312312312312314e-05,
+      "loss": 0.3103,
+      "step": 1548
+    },
+    {
+      "epoch": 0.0020248286608026372,
+      "grad_norm": 1.546135425567627,
+      "learning_rate": 4.230730730730731e-05,
+      "loss": 1.6428,
+      "step": 1549
+    },
+    {
+      "epoch": 0.0020261358452189075,
+      "grad_norm": 1.3285560607910156,
+      "learning_rate": 4.230230230230231e-05,
+      "loss": 1.6233,
+      "step": 1550
+    },
+    {
+      "epoch": 0.002027443029635178,
+      "grad_norm": 1.5606275796890259,
+      "learning_rate": 4.22972972972973e-05,
+      "loss": 1.6699,
+      "step": 1551
+    },
+    {
+      "epoch": 0.002028750214051448,
+      "grad_norm": 1.850281834602356,
+      "learning_rate": 4.229229229229229e-05,
+      "loss": 1.4488,
+      "step": 1552
+    },
+    {
+      "epoch": 0.0020300573984677185,
+      "grad_norm": 1.7698073387145996,
+      "learning_rate": 4.228728728728729e-05,
+      "loss": 2.6492,
+      "step": 1553
+    },
+    {
+      "epoch": 0.002031364582883989,
+      "grad_norm": 1.8112493753433228,
+      "learning_rate": 4.228228228228228e-05,
+      "loss": 1.605,
+      "step": 1554
+    },
+    {
+      "epoch": 0.002032671767300259,
+      "grad_norm": 1.41746187210083,
+      "learning_rate": 4.227727727727728e-05,
+      "loss": 1.3595,
+      "step": 1555
+    },
+    {
+      "epoch": 0.002033978951716529,
+      "grad_norm": 1.5460761785507202,
+      "learning_rate": 4.227227227227227e-05,
+      "loss": 2.1071,
+      "step": 1556
+    },
+    {
+      "epoch": 0.0020352861361327993,
+      "grad_norm": 1.6772769689559937,
+      "learning_rate": 4.226726726726727e-05,
+      "loss": 1.9404,
+      "step": 1557
+    },
+    {
+      "epoch": 0.0020365933205490697,
+      "grad_norm": 1.3576667308807373,
+      "learning_rate": 4.2262262262262264e-05,
+      "loss": 1.0275,
+      "step": 1558
+    },
+    {
+      "epoch": 0.00203790050496534,
+      "grad_norm": 1.4979108572006226,
+      "learning_rate": 4.225725725725726e-05,
+      "loss": 2.3442,
+      "step": 1559
+    },
+    {
+      "epoch": 0.0020392076893816103,
+      "grad_norm": 1.4685587882995605,
+      "learning_rate": 4.225225225225226e-05,
+      "loss": 1.8375,
+      "step": 1560
+    },
+    {
+      "epoch": 0.0020405148737978806,
+      "grad_norm": 0.7958675026893616,
+      "learning_rate": 4.224724724724725e-05,
+      "loss": 0.4731,
+      "step": 1561
+    },
+    {
+      "epoch": 0.002041822058214151,
+      "grad_norm": 1.4936842918395996,
+      "learning_rate": 4.224224224224225e-05,
+      "loss": 1.6163,
+      "step": 1562
+    },
+    {
+      "epoch": 0.0020431292426304213,
+      "grad_norm": 1.1718482971191406,
+      "learning_rate": 4.2237237237237236e-05,
+      "loss": 1.6056,
+      "step": 1563
+    },
+    {
+      "epoch": 0.002044436427046691,
+      "grad_norm": 2.012826681137085,
+      "learning_rate": 4.2232232232232236e-05,
+      "loss": 1.2933,
+      "step": 1564
+    },
+    {
+      "epoch": 0.0020457436114629615,
+      "grad_norm": 1.8661749362945557,
+      "learning_rate": 4.222722722722723e-05,
+      "loss": 2.0133,
+      "step": 1565
+    },
+    {
+      "epoch": 0.002047050795879232,
+      "grad_norm": 1.1587417125701904,
+      "learning_rate": 4.222222222222222e-05,
+      "loss": 1.4065,
+      "step": 1566
+    },
+    {
+      "epoch": 0.002048357980295502,
+      "grad_norm": 1.3768486976623535,
+      "learning_rate": 4.221721721721722e-05,
+      "loss": 1.4668,
+      "step": 1567
+    },
+    {
+      "epoch": 0.0020496651647117724,
+      "grad_norm": 1.2806345224380493,
+      "learning_rate": 4.2212212212212214e-05,
+      "loss": 1.0595,
+      "step": 1568
+    },
+    {
+      "epoch": 0.0020509723491280427,
+      "grad_norm": 1.6019591093063354,
+      "learning_rate": 4.2207207207207214e-05,
+      "loss": 2.4652,
+      "step": 1569
+    },
+    {
+      "epoch": 0.002052279533544313,
+      "grad_norm": 1.3496835231781006,
+      "learning_rate": 4.220220220220221e-05,
+      "loss": 1.4763,
+      "step": 1570
+    },
+    {
+      "epoch": 0.002053586717960583,
+      "grad_norm": 2.1939737796783447,
+      "learning_rate": 4.21971971971972e-05,
+      "loss": 1.8059,
+      "step": 1571
+    },
+    {
+      "epoch": 0.0020548939023768533,
+      "grad_norm": 0.9771382212638855,
+      "learning_rate": 4.219219219219219e-05,
+      "loss": 0.852,
+      "step": 1572
+    },
+    {
+      "epoch": 0.0020562010867931236,
+      "grad_norm": 1.3086856603622437,
+      "learning_rate": 4.2187187187187186e-05,
+      "loss": 1.1098,
+      "step": 1573
+    },
+    {
+      "epoch": 0.002057508271209394,
+      "grad_norm": 1.036381721496582,
+      "learning_rate": 4.2182182182182186e-05,
+      "loss": 0.783,
+      "step": 1574
+    },
+    {
+      "epoch": 0.0020588154556256642,
+      "grad_norm": 1.2067959308624268,
+      "learning_rate": 4.217717717717718e-05,
+      "loss": 1.3124,
+      "step": 1575
+    },
+    {
+      "epoch": 0.0020601226400419346,
+      "grad_norm": 1.142862319946289,
+      "learning_rate": 4.217217217217218e-05,
+      "loss": 0.4757,
+      "step": 1576
+    },
+    {
+      "epoch": 0.002061429824458205,
+      "grad_norm": 1.682342767715454,
+      "learning_rate": 4.216716716716717e-05,
+      "loss": 1.1961,
+      "step": 1577
+    },
+    {
+      "epoch": 0.002062737008874475,
+      "grad_norm": 1.0279693603515625,
+      "learning_rate": 4.2162162162162164e-05,
+      "loss": 0.7677,
+      "step": 1578
+    },
+    {
+      "epoch": 0.002064044193290745,
+      "grad_norm": 1.4873472452163696,
+      "learning_rate": 4.215715715715716e-05,
+      "loss": 1.0744,
+      "step": 1579
+    },
+    {
+      "epoch": 0.0020653513777070154,
+      "grad_norm": 1.9146010875701904,
+      "learning_rate": 4.215215215215215e-05,
+      "loss": 1.9371,
+      "step": 1580
+    },
+    {
+      "epoch": 0.0020666585621232857,
+      "grad_norm": 1.4062408208847046,
+      "learning_rate": 4.214714714714715e-05,
+      "loss": 1.5336,
+      "step": 1581
+    },
+    {
+      "epoch": 0.002067965746539556,
+      "grad_norm": 1.6379565000534058,
+      "learning_rate": 4.214214214214214e-05,
+      "loss": 1.9558,
+      "step": 1582
+    },
+    {
+      "epoch": 0.0020692729309558264,
+      "grad_norm": 1.4497932195663452,
+      "learning_rate": 4.2137137137137136e-05,
+      "loss": 1.2788,
+      "step": 1583
+    },
+    {
+      "epoch": 0.0020705801153720967,
+      "grad_norm": 1.4147257804870605,
+      "learning_rate": 4.2132132132132136e-05,
+      "loss": 1.9033,
+      "step": 1584
+    },
+    {
+      "epoch": 0.002071887299788367,
+      "grad_norm": 1.3680346012115479,
+      "learning_rate": 4.212712712712713e-05,
+      "loss": 1.029,
+      "step": 1585
+    },
+    {
+      "epoch": 0.0020731944842046373,
+      "grad_norm": 1.5454055070877075,
+      "learning_rate": 4.212212212212213e-05,
+      "loss": 1.3537,
+      "step": 1586
+    },
+    {
+      "epoch": 0.002074501668620907,
+      "grad_norm": 1.7991704940795898,
+      "learning_rate": 4.2117117117117115e-05,
+      "loss": 1.7358,
+      "step": 1587
+    },
+    {
+      "epoch": 0.0020758088530371775,
+      "grad_norm": 1.5681918859481812,
+      "learning_rate": 4.2112112112112114e-05,
+      "loss": 1.943,
+      "step": 1588
+    },
+    {
+      "epoch": 0.002077116037453448,
+      "grad_norm": 1.6188732385635376,
+      "learning_rate": 4.210710710710711e-05,
+      "loss": 2.5814,
+      "step": 1589
+    },
+    {
+      "epoch": 0.002078423221869718,
+      "grad_norm": 1.2546133995056152,
+      "learning_rate": 4.21021021021021e-05,
+      "loss": 1.0761,
+      "step": 1590
+    },
+    {
+      "epoch": 0.0020797304062859885,
+      "grad_norm": 1.4694972038269043,
+      "learning_rate": 4.20970970970971e-05,
+      "loss": 1.6272,
+      "step": 1591
+    },
+    {
+      "epoch": 0.002081037590702259,
+      "grad_norm": 1.6937707662582397,
+      "learning_rate": 4.209209209209209e-05,
+      "loss": 2.2769,
+      "step": 1592
+    },
+    {
+      "epoch": 0.002082344775118529,
+      "grad_norm": 1.551703929901123,
+      "learning_rate": 4.208708708708709e-05,
+      "loss": 1.4276,
+      "step": 1593
+    },
+    {
+      "epoch": 0.002083651959534799,
+      "grad_norm": 0.9847801327705383,
+      "learning_rate": 4.2082082082082086e-05,
+      "loss": 0.2801,
+      "step": 1594
+    },
+    {
+      "epoch": 0.0020849591439510693,
+      "grad_norm": 1.7075239419937134,
+      "learning_rate": 4.207707707707708e-05,
+      "loss": 2.1934,
+      "step": 1595
+    },
+    {
+      "epoch": 0.0020862663283673396,
+      "grad_norm": 2.2408063411712646,
+      "learning_rate": 4.207207207207207e-05,
+      "loss": 2.6122,
+      "step": 1596
+    },
+    {
+      "epoch": 0.00208757351278361,
+      "grad_norm": 1.3065295219421387,
+      "learning_rate": 4.2067067067067065e-05,
+      "loss": 1.5063,
+      "step": 1597
+    },
+    {
+      "epoch": 0.0020888806971998803,
+      "grad_norm": 1.5663440227508545,
+      "learning_rate": 4.2062062062062064e-05,
+      "loss": 2.5296,
+      "step": 1598
+    },
+    {
+      "epoch": 0.0020901878816161506,
+      "grad_norm": 0.7946434020996094,
+      "learning_rate": 4.205705705705706e-05,
+      "loss": 0.2782,
+      "step": 1599
+    },
+    {
+      "epoch": 0.002091495066032421,
+      "grad_norm": 1.102759838104248,
+      "learning_rate": 4.205205205205206e-05,
+      "loss": 0.4698,
+      "step": 1600
+    },
+    {
+      "epoch": 0.0020928022504486912,
+      "grad_norm": 1.840031623840332,
+      "learning_rate": 4.204704704704705e-05,
+      "loss": 1.914,
+      "step": 1601
+    },
+    {
+      "epoch": 0.002094109434864961,
+      "grad_norm": 1.1779990196228027,
+      "learning_rate": 4.204204204204204e-05,
+      "loss": 1.4124,
+      "step": 1602
+    },
+    {
+      "epoch": 0.0020954166192812314,
+      "grad_norm": 1.5375138521194458,
+      "learning_rate": 4.203703703703704e-05,
+      "loss": 1.7214,
+      "step": 1603
+    },
+    {
+      "epoch": 0.0020967238036975018,
+      "grad_norm": 1.3905709981918335,
+      "learning_rate": 4.203203203203203e-05,
+      "loss": 1.4159,
+      "step": 1604
+    },
+    {
+      "epoch": 0.002098030988113772,
+      "grad_norm": 1.6796258687973022,
+      "learning_rate": 4.202702702702703e-05,
+      "loss": 1.2918,
+      "step": 1605
+    },
+    {
+      "epoch": 0.0020993381725300424,
+      "grad_norm": 2.1706652641296387,
+      "learning_rate": 4.202202202202202e-05,
+      "loss": 1.9185,
+      "step": 1606
+    },
+    {
+      "epoch": 0.0021006453569463127,
+      "grad_norm": 1.3634697198867798,
+      "learning_rate": 4.201701701701702e-05,
+      "loss": 1.0133,
+      "step": 1607
+    },
+    {
+      "epoch": 0.002101952541362583,
+      "grad_norm": 1.629964828491211,
+      "learning_rate": 4.2012012012012015e-05,
+      "loss": 0.5669,
+      "step": 1608
+    },
+    {
+      "epoch": 0.0021032597257788534,
+      "grad_norm": 1.3685063123703003,
+      "learning_rate": 4.200700700700701e-05,
+      "loss": 2.0132,
+      "step": 1609
+    },
+    {
+      "epoch": 0.0021045669101951233,
+      "grad_norm": 1.620974063873291,
+      "learning_rate": 4.200200200200201e-05,
+      "loss": 1.7746,
+      "step": 1610
+    },
+    {
+      "epoch": 0.0021058740946113936,
+      "grad_norm": 1.383302092552185,
+      "learning_rate": 4.1996996996997e-05,
+      "loss": 1.5981,
+      "step": 1611
+    },
+    {
+      "epoch": 0.002107181279027664,
+      "grad_norm": 1.7801709175109863,
+      "learning_rate": 4.199199199199199e-05,
+      "loss": 2.1991,
+      "step": 1612
+    },
+    {
+      "epoch": 0.002108488463443934,
+      "grad_norm": 1.602144479751587,
+      "learning_rate": 4.1986986986986986e-05,
+      "loss": 1.2544,
+      "step": 1613
+    },
+    {
+      "epoch": 0.0021097956478602045,
+      "grad_norm": 1.5597031116485596,
+      "learning_rate": 4.1981981981981986e-05,
+      "loss": 2.3005,
+      "step": 1614
+    },
+    {
+      "epoch": 0.002111102832276475,
+      "grad_norm": 1.4823471307754517,
+      "learning_rate": 4.197697697697698e-05,
+      "loss": 1.8555,
+      "step": 1615
+    },
+    {
+      "epoch": 0.002112410016692745,
+      "grad_norm": 1.3974865674972534,
+      "learning_rate": 4.197197197197197e-05,
+      "loss": 1.3607,
+      "step": 1616
+    },
+    {
+      "epoch": 0.002113717201109015,
+      "grad_norm": 2.2095515727996826,
+      "learning_rate": 4.196696696696697e-05,
+      "loss": 2.5273,
+      "step": 1617
+    },
+    {
+      "epoch": 0.0021150243855252854,
+      "grad_norm": 1.433372974395752,
+      "learning_rate": 4.1961961961961965e-05,
+      "loss": 1.4491,
+      "step": 1618
+    },
+    {
+      "epoch": 0.0021163315699415557,
+      "grad_norm": 1.0379494428634644,
+      "learning_rate": 4.195695695695696e-05,
+      "loss": 0.8464,
+      "step": 1619
+    },
+    {
+      "epoch": 0.002117638754357826,
+      "grad_norm": 0.589991569519043,
+      "learning_rate": 4.195195195195196e-05,
+      "loss": 0.1329,
+      "step": 1620
+    },
+    {
+      "epoch": 0.0021189459387740963,
+      "grad_norm": 1.6323260068893433,
+      "learning_rate": 4.1946946946946944e-05,
+      "loss": 1.9577,
+      "step": 1621
+    },
+    {
+      "epoch": 0.0021202531231903667,
+      "grad_norm": 1.1113680601119995,
+      "learning_rate": 4.194194194194194e-05,
+      "loss": 0.7364,
+      "step": 1622
+    },
+    {
+      "epoch": 0.002121560307606637,
+      "grad_norm": 1.4257352352142334,
+      "learning_rate": 4.1936936936936936e-05,
+      "loss": 1.5237,
+      "step": 1623
+    },
+    {
+      "epoch": 0.0021228674920229073,
+      "grad_norm": 1.2944560050964355,
+      "learning_rate": 4.1931931931931936e-05,
+      "loss": 1.4138,
+      "step": 1624
+    },
+    {
+      "epoch": 0.002124174676439177,
+      "grad_norm": 1.3862648010253906,
+      "learning_rate": 4.192692692692693e-05,
+      "loss": 2.1726,
+      "step": 1625
+    },
+    {
+      "epoch": 0.0021254818608554475,
+      "grad_norm": 1.3448632955551147,
+      "learning_rate": 4.192192192192192e-05,
+      "loss": 1.7745,
+      "step": 1626
+    },
+    {
+      "epoch": 0.002126789045271718,
+      "grad_norm": 1.390791416168213,
+      "learning_rate": 4.191691691691692e-05,
+      "loss": 1.4272,
+      "step": 1627
+    },
+    {
+      "epoch": 0.002128096229687988,
+      "grad_norm": 1.081365704536438,
+      "learning_rate": 4.1911911911911915e-05,
+      "loss": 0.6313,
+      "step": 1628
+    },
+    {
+      "epoch": 0.0021294034141042585,
+      "grad_norm": 1.2875622510910034,
+      "learning_rate": 4.190690690690691e-05,
+      "loss": 1.0294,
+      "step": 1629
+    },
+    {
+      "epoch": 0.0021307105985205288,
+      "grad_norm": 1.556660532951355,
+      "learning_rate": 4.19019019019019e-05,
+      "loss": 1.5291,
+      "step": 1630
+    },
+    {
+      "epoch": 0.002132017782936799,
+      "grad_norm": 1.832720160484314,
+      "learning_rate": 4.18968968968969e-05,
+      "loss": 1.2926,
+      "step": 1631
+    },
+    {
+      "epoch": 0.0021333249673530694,
+      "grad_norm": 2.2108466625213623,
+      "learning_rate": 4.189189189189189e-05,
+      "loss": 1.9928,
+      "step": 1632
+    },
+    {
+      "epoch": 0.0021346321517693393,
+      "grad_norm": 1.7715742588043213,
+      "learning_rate": 4.1886886886886886e-05,
+      "loss": 1.0429,
+      "step": 1633
+    },
+    {
+      "epoch": 0.0021359393361856096,
+      "grad_norm": 0.8824874758720398,
+      "learning_rate": 4.1881881881881886e-05,
+      "loss": 0.353,
+      "step": 1634
+    },
+    {
+      "epoch": 0.00213724652060188,
+      "grad_norm": 1.4567463397979736,
+      "learning_rate": 4.187687687687688e-05,
+      "loss": 2.7918,
+      "step": 1635
+    },
+    {
+      "epoch": 0.0021385537050181503,
+      "grad_norm": 1.4053136110305786,
+      "learning_rate": 4.187187187187188e-05,
+      "loss": 2.0669,
+      "step": 1636
+    },
+    {
+      "epoch": 0.0021398608894344206,
+      "grad_norm": 1.1038392782211304,
+      "learning_rate": 4.1866866866866865e-05,
+      "loss": 1.1557,
+      "step": 1637
+    },
+    {
+      "epoch": 0.002141168073850691,
+      "grad_norm": 1.3500014543533325,
+      "learning_rate": 4.1861861861861865e-05,
+      "loss": 1.44,
+      "step": 1638
+    },
+    {
+      "epoch": 0.0021424752582669612,
+      "grad_norm": 1.1549291610717773,
+      "learning_rate": 4.185685685685686e-05,
+      "loss": 1.5516,
+      "step": 1639
+    },
+    {
+      "epoch": 0.002143782442683231,
+      "grad_norm": 1.349645972251892,
+      "learning_rate": 4.185185185185185e-05,
+      "loss": 1.0438,
+      "step": 1640
+    },
+    {
+      "epoch": 0.0021450896270995014,
+      "grad_norm": 1.4057704210281372,
+      "learning_rate": 4.184684684684685e-05,
+      "loss": 1.2328,
+      "step": 1641
+    },
+    {
+      "epoch": 0.0021463968115157717,
+      "grad_norm": 1.7850967645645142,
+      "learning_rate": 4.1841841841841843e-05,
+      "loss": 1.6124,
+      "step": 1642
+    },
+    {
+      "epoch": 0.002147703995932042,
+      "grad_norm": 1.4541049003601074,
+      "learning_rate": 4.183683683683684e-05,
+      "loss": 1.87,
+      "step": 1643
+    },
+    {
+      "epoch": 0.0021490111803483124,
+      "grad_norm": 1.5384831428527832,
+      "learning_rate": 4.1831831831831836e-05,
+      "loss": 2.2349,
+      "step": 1644
+    },
+    {
+      "epoch": 0.0021503183647645827,
+      "grad_norm": 1.6927258968353271,
+      "learning_rate": 4.182682682682683e-05,
+      "loss": 1.0959,
+      "step": 1645
+    },
+    {
+      "epoch": 0.002151625549180853,
+      "grad_norm": 1.430592656135559,
+      "learning_rate": 4.182182182182182e-05,
+      "loss": 1.0466,
+      "step": 1646
+    },
+    {
+      "epoch": 0.0021529327335971233,
+      "grad_norm": 1.3500418663024902,
+      "learning_rate": 4.1816816816816815e-05,
+      "loss": 1.4112,
+      "step": 1647
+    },
+    {
+      "epoch": 0.0021542399180133932,
+      "grad_norm": 1.6099777221679688,
+      "learning_rate": 4.1811811811811815e-05,
+      "loss": 2.0719,
+      "step": 1648
+    },
+    {
+      "epoch": 0.0021555471024296636,
+      "grad_norm": 1.6342360973358154,
+      "learning_rate": 4.180680680680681e-05,
+      "loss": 1.936,
+      "step": 1649
+    },
+    {
+      "epoch": 0.002156854286845934,
+      "grad_norm": 1.5729727745056152,
+      "learning_rate": 4.180180180180181e-05,
+      "loss": 1.0408,
+      "step": 1650
+    },
+    {
+      "epoch": 0.002158161471262204,
+      "grad_norm": 1.5231847763061523,
+      "learning_rate": 4.17967967967968e-05,
+      "loss": 1.7468,
+      "step": 1651
+    },
+    {
+      "epoch": 0.0021594686556784745,
+      "grad_norm": 1.1770910024642944,
+      "learning_rate": 4.1791791791791793e-05,
+      "loss": 1.1546,
+      "step": 1652
+    },
+    {
+      "epoch": 0.002160775840094745,
+      "grad_norm": 1.3753366470336914,
+      "learning_rate": 4.178678678678679e-05,
+      "loss": 1.1226,
+      "step": 1653
+    },
+    {
+      "epoch": 0.002162083024511015,
+      "grad_norm": 1.8559156656265259,
+      "learning_rate": 4.178178178178178e-05,
+      "loss": 1.4849,
+      "step": 1654
+    },
+    {
+      "epoch": 0.002163390208927285,
+      "grad_norm": 1.669425129890442,
+      "learning_rate": 4.177677677677678e-05,
+      "loss": 1.7637,
+      "step": 1655
+    },
+    {
+      "epoch": 0.0021646973933435554,
+      "grad_norm": 1.5557461977005005,
+      "learning_rate": 4.177177177177177e-05,
+      "loss": 1.6608,
+      "step": 1656
+    },
+    {
+      "epoch": 0.0021660045777598257,
+      "grad_norm": 1.1244462728500366,
+      "learning_rate": 4.1766766766766765e-05,
+      "loss": 0.7272,
+      "step": 1657
+    },
+    {
+      "epoch": 0.002167311762176096,
+      "grad_norm": 1.4152379035949707,
+      "learning_rate": 4.1761761761761765e-05,
+      "loss": 1.1406,
+      "step": 1658
+    },
+    {
+      "epoch": 0.0021686189465923663,
+      "grad_norm": 1.4741320610046387,
+      "learning_rate": 4.175675675675676e-05,
+      "loss": 1.5492,
+      "step": 1659
+    },
+    {
+      "epoch": 0.0021699261310086366,
+      "grad_norm": 1.617621898651123,
+      "learning_rate": 4.175175175175176e-05,
+      "loss": 1.1048,
+      "step": 1660
+    },
+    {
+      "epoch": 0.002171233315424907,
+      "grad_norm": 1.7727829217910767,
+      "learning_rate": 4.174674674674675e-05,
+      "loss": 1.6242,
+      "step": 1661
+    },
+    {
+      "epoch": 0.0021725404998411773,
+      "grad_norm": 3.710369825363159,
+      "learning_rate": 4.1741741741741744e-05,
+      "loss": 1.7355,
+      "step": 1662
+    },
+    {
+      "epoch": 0.002173847684257447,
+      "grad_norm": 1.6971349716186523,
+      "learning_rate": 4.1736736736736737e-05,
+      "loss": 1.968,
+      "step": 1663
+    },
+    {
+      "epoch": 0.0021751548686737175,
+      "grad_norm": 1.1078214645385742,
+      "learning_rate": 4.173173173173173e-05,
+      "loss": 0.9752,
+      "step": 1664
+    },
+    {
+      "epoch": 0.002176462053089988,
+      "grad_norm": 1.6114851236343384,
+      "learning_rate": 4.172672672672673e-05,
+      "loss": 1.4643,
+      "step": 1665
+    },
+    {
+      "epoch": 0.002177769237506258,
+      "grad_norm": 1.2387334108352661,
+      "learning_rate": 4.172172172172172e-05,
+      "loss": 1.3744,
+      "step": 1666
+    },
+    {
+      "epoch": 0.0021790764219225284,
+      "grad_norm": 1.488434910774231,
+      "learning_rate": 4.171671671671672e-05,
+      "loss": 1.225,
+      "step": 1667
+    },
+    {
+      "epoch": 0.0021803836063387988,
+      "grad_norm": 1.2892612218856812,
+      "learning_rate": 4.1711711711711715e-05,
+      "loss": 2.0891,
+      "step": 1668
+    },
+    {
+      "epoch": 0.002181690790755069,
+      "grad_norm": 1.8534702062606812,
+      "learning_rate": 4.170670670670671e-05,
+      "loss": 1.6232,
+      "step": 1669
+    },
+    {
+      "epoch": 0.0021829979751713394,
+      "grad_norm": 1.5404982566833496,
+      "learning_rate": 4.170170170170171e-05,
+      "loss": 1.1884,
+      "step": 1670
+    },
+    {
+      "epoch": 0.0021843051595876093,
+      "grad_norm": 1.1658841371536255,
+      "learning_rate": 4.1696696696696694e-05,
+      "loss": 1.2897,
+      "step": 1671
+    },
+    {
+      "epoch": 0.0021856123440038796,
+      "grad_norm": 1.5434014797210693,
+      "learning_rate": 4.1691691691691694e-05,
+      "loss": 1.4142,
+      "step": 1672
+    },
+    {
+      "epoch": 0.00218691952842015,
+      "grad_norm": 1.6983811855316162,
+      "learning_rate": 4.1686686686686687e-05,
+      "loss": 1.4291,
+      "step": 1673
+    },
+    {
+      "epoch": 0.0021882267128364202,
+      "grad_norm": 2.0139994621276855,
+      "learning_rate": 4.1681681681681686e-05,
+      "loss": 1.2576,
+      "step": 1674
+    },
+    {
+      "epoch": 0.0021895338972526906,
+      "grad_norm": 1.5874844789505005,
+      "learning_rate": 4.167667667667668e-05,
+      "loss": 1.9113,
+      "step": 1675
+    },
+    {
+      "epoch": 0.002190841081668961,
+      "grad_norm": 1.1104267835617065,
+      "learning_rate": 4.167167167167167e-05,
+      "loss": 1.408,
+      "step": 1676
+    },
+    {
+      "epoch": 0.002192148266085231,
+      "grad_norm": 1.3420699834823608,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.5189,
+      "step": 1677
+    },
+    {
+      "epoch": 0.002193455450501501,
+      "grad_norm": 4.363579750061035,
+      "learning_rate": 4.1661661661661665e-05,
+      "loss": 1.868,
+      "step": 1678
+    },
+    {
+      "epoch": 0.0021947626349177714,
+      "grad_norm": 1.7833904027938843,
+      "learning_rate": 4.165665665665666e-05,
+      "loss": 1.9687,
+      "step": 1679
+    },
+    {
+      "epoch": 0.0021960698193340417,
+      "grad_norm": 1.716072916984558,
+      "learning_rate": 4.165165165165165e-05,
+      "loss": 0.8284,
+      "step": 1680
+    },
+    {
+      "epoch": 0.002197377003750312,
+      "grad_norm": 1.3779525756835938,
+      "learning_rate": 4.164664664664665e-05,
+      "loss": 1.5269,
+      "step": 1681
+    },
+    {
+      "epoch": 0.0021986841881665824,
+      "grad_norm": 1.417640209197998,
+      "learning_rate": 4.1641641641641644e-05,
+      "loss": 1.7857,
+      "step": 1682
+    },
+    {
+      "epoch": 0.0021999913725828527,
+      "grad_norm": 1.205691933631897,
+      "learning_rate": 4.163663663663664e-05,
+      "loss": 0.7655,
+      "step": 1683
+    },
+    {
+      "epoch": 0.002201298556999123,
+      "grad_norm": 1.6471670866012573,
+      "learning_rate": 4.1631631631631636e-05,
+      "loss": 1.714,
+      "step": 1684
+    },
+    {
+      "epoch": 0.0022026057414153933,
+      "grad_norm": 2.206336736679077,
+      "learning_rate": 4.162662662662663e-05,
+      "loss": 2.3123,
+      "step": 1685
+    },
+    {
+      "epoch": 0.002203912925831663,
+      "grad_norm": 1.4241451025009155,
+      "learning_rate": 4.162162162162163e-05,
+      "loss": 2.0536,
+      "step": 1686
+    },
+    {
+      "epoch": 0.0022052201102479335,
+      "grad_norm": 1.4296327829360962,
+      "learning_rate": 4.1616616616616615e-05,
+      "loss": 2.3104,
+      "step": 1687
+    },
+    {
+      "epoch": 0.002206527294664204,
+      "grad_norm": 2.1866962909698486,
+      "learning_rate": 4.161161161161161e-05,
+      "loss": 1.3986,
+      "step": 1688
+    },
+    {
+      "epoch": 0.002207834479080474,
+      "grad_norm": 1.4900760650634766,
+      "learning_rate": 4.160660660660661e-05,
+      "loss": 1.5956,
+      "step": 1689
+    },
+    {
+      "epoch": 0.0022091416634967445,
+      "grad_norm": 1.5100595951080322,
+      "learning_rate": 4.16016016016016e-05,
+      "loss": 1.6542,
+      "step": 1690
+    },
+    {
+      "epoch": 0.002210448847913015,
+      "grad_norm": 1.4321930408477783,
+      "learning_rate": 4.15965965965966e-05,
+      "loss": 2.0535,
+      "step": 1691
+    },
+    {
+      "epoch": 0.002211756032329285,
+      "grad_norm": 1.3654512166976929,
+      "learning_rate": 4.1591591591591594e-05,
+      "loss": 1.5635,
+      "step": 1692
+    },
+    {
+      "epoch": 0.0022130632167455555,
+      "grad_norm": 1.0412864685058594,
+      "learning_rate": 4.158658658658659e-05,
+      "loss": 0.8502,
+      "step": 1693
+    },
+    {
+      "epoch": 0.0022143704011618253,
+      "grad_norm": 2.0201199054718018,
+      "learning_rate": 4.1581581581581586e-05,
+      "loss": 1.7143,
+      "step": 1694
+    },
+    {
+      "epoch": 0.0022156775855780957,
+      "grad_norm": 1.277818202972412,
+      "learning_rate": 4.157657657657658e-05,
+      "loss": 1.3427,
+      "step": 1695
+    },
+    {
+      "epoch": 0.002216984769994366,
+      "grad_norm": 1.3864494562149048,
+      "learning_rate": 4.157157157157157e-05,
+      "loss": 1.6466,
+      "step": 1696
+    },
+    {
+      "epoch": 0.0022182919544106363,
+      "grad_norm": 1.5996184349060059,
+      "learning_rate": 4.1566566566566565e-05,
+      "loss": 1.934,
+      "step": 1697
+    },
+    {
+      "epoch": 0.0022195991388269066,
+      "grad_norm": 1.4885029792785645,
+      "learning_rate": 4.1561561561561565e-05,
+      "loss": 1.4581,
+      "step": 1698
+    },
+    {
+      "epoch": 0.002220906323243177,
+      "grad_norm": 1.244058609008789,
+      "learning_rate": 4.155655655655656e-05,
+      "loss": 1.84,
+      "step": 1699
+    },
+    {
+      "epoch": 0.0022222135076594473,
+      "grad_norm": 1.43318772315979,
+      "learning_rate": 4.155155155155155e-05,
+      "loss": 1.5146,
+      "step": 1700
+    },
+    {
+      "epoch": 0.002223520692075717,
+      "grad_norm": 2.0152251720428467,
+      "learning_rate": 4.154654654654655e-05,
+      "loss": 2.0638,
+      "step": 1701
+    },
+    {
+      "epoch": 0.0022248278764919875,
+      "grad_norm": 1.3837087154388428,
+      "learning_rate": 4.1541541541541544e-05,
+      "loss": 1.3273,
+      "step": 1702
+    },
+    {
+      "epoch": 0.002226135060908258,
+      "grad_norm": 1.1935254335403442,
+      "learning_rate": 4.1536536536536544e-05,
+      "loss": 1.384,
+      "step": 1703
+    },
+    {
+      "epoch": 0.002227442245324528,
+      "grad_norm": 1.0881271362304688,
+      "learning_rate": 4.153153153153153e-05,
+      "loss": 1.2483,
+      "step": 1704
+    },
+    {
+      "epoch": 0.0022287494297407984,
+      "grad_norm": 1.438117265701294,
+      "learning_rate": 4.152652652652653e-05,
+      "loss": 1.4326,
+      "step": 1705
+    },
+    {
+      "epoch": 0.0022300566141570687,
+      "grad_norm": 1.3385870456695557,
+      "learning_rate": 4.152152152152152e-05,
+      "loss": 1.3195,
+      "step": 1706
+    },
+    {
+      "epoch": 0.002231363798573339,
+      "grad_norm": 1.416751742362976,
+      "learning_rate": 4.1516516516516515e-05,
+      "loss": 2.0428,
+      "step": 1707
+    },
+    {
+      "epoch": 0.0022326709829896094,
+      "grad_norm": 1.2597955465316772,
+      "learning_rate": 4.1511511511511515e-05,
+      "loss": 1.6776,
+      "step": 1708
+    },
+    {
+      "epoch": 0.0022339781674058793,
+      "grad_norm": 1.1180474758148193,
+      "learning_rate": 4.150650650650651e-05,
+      "loss": 1.1397,
+      "step": 1709
+    },
+    {
+      "epoch": 0.0022352853518221496,
+      "grad_norm": 1.326080083847046,
+      "learning_rate": 4.150150150150151e-05,
+      "loss": 1.9762,
+      "step": 1710
+    },
+    {
+      "epoch": 0.00223659253623842,
+      "grad_norm": 1.2600030899047852,
+      "learning_rate": 4.14964964964965e-05,
+      "loss": 1.3425,
+      "step": 1711
+    },
+    {
+      "epoch": 0.0022378997206546902,
+      "grad_norm": 1.6297385692596436,
+      "learning_rate": 4.1491491491491494e-05,
+      "loss": 1.7848,
+      "step": 1712
+    },
+    {
+      "epoch": 0.0022392069050709605,
+      "grad_norm": 1.193619966506958,
+      "learning_rate": 4.148648648648649e-05,
+      "loss": 1.1092,
+      "step": 1713
+    },
+    {
+      "epoch": 0.002240514089487231,
+      "grad_norm": 1.6174732446670532,
+      "learning_rate": 4.148148148148148e-05,
+      "loss": 1.8455,
+      "step": 1714
+    },
+    {
+      "epoch": 0.002241821273903501,
+      "grad_norm": 1.5751314163208008,
+      "learning_rate": 4.147647647647648e-05,
+      "loss": 1.3083,
+      "step": 1715
+    },
+    {
+      "epoch": 0.0022431284583197715,
+      "grad_norm": 0.8099761605262756,
+      "learning_rate": 4.147147147147147e-05,
+      "loss": 0.2987,
+      "step": 1716
+    },
+    {
+      "epoch": 0.0022444356427360414,
+      "grad_norm": 1.5695183277130127,
+      "learning_rate": 4.146646646646647e-05,
+      "loss": 1.8786,
+      "step": 1717
+    },
+    {
+      "epoch": 0.0022457428271523117,
+      "grad_norm": 1.4828535318374634,
+      "learning_rate": 4.1461461461461465e-05,
+      "loss": 1.495,
+      "step": 1718
+    },
+    {
+      "epoch": 0.002247050011568582,
+      "grad_norm": 1.6301383972167969,
+      "learning_rate": 4.145645645645646e-05,
+      "loss": 1.9807,
+      "step": 1719
+    },
+    {
+      "epoch": 0.0022483571959848524,
+      "grad_norm": 1.2270265817642212,
+      "learning_rate": 4.145145145145145e-05,
+      "loss": 1.1866,
+      "step": 1720
+    },
+    {
+      "epoch": 0.0022496643804011227,
+      "grad_norm": 1.500815749168396,
+      "learning_rate": 4.1446446446446444e-05,
+      "loss": 2.1331,
+      "step": 1721
+    },
+    {
+      "epoch": 0.002250971564817393,
+      "grad_norm": 1.529885172843933,
+      "learning_rate": 4.1441441441441444e-05,
+      "loss": 1.7827,
+      "step": 1722
+    },
+    {
+      "epoch": 0.0022522787492336633,
+      "grad_norm": 1.4575401544570923,
+      "learning_rate": 4.143643643643644e-05,
+      "loss": 1.9545,
+      "step": 1723
+    },
+    {
+      "epoch": 0.002253585933649933,
+      "grad_norm": 2.4542195796966553,
+      "learning_rate": 4.143143143143144e-05,
+      "loss": 1.6215,
+      "step": 1724
+    },
+    {
+      "epoch": 0.0022548931180662035,
+      "grad_norm": 1.4880352020263672,
+      "learning_rate": 4.142642642642643e-05,
+      "loss": 2.0558,
+      "step": 1725
+    },
+    {
+      "epoch": 0.002256200302482474,
+      "grad_norm": 1.7764296531677246,
+      "learning_rate": 4.142142142142142e-05,
+      "loss": 1.5755,
+      "step": 1726
+    },
+    {
+      "epoch": 0.002257507486898744,
+      "grad_norm": 1.4144059419631958,
+      "learning_rate": 4.141641641641642e-05,
+      "loss": 2.2151,
+      "step": 1727
+    },
+    {
+      "epoch": 0.0022588146713150145,
+      "grad_norm": 1.3584964275360107,
+      "learning_rate": 4.1411411411411415e-05,
+      "loss": 1.5215,
+      "step": 1728
+    },
+    {
+      "epoch": 0.002260121855731285,
+      "grad_norm": 1.5217913389205933,
+      "learning_rate": 4.140640640640641e-05,
+      "loss": 1.9313,
+      "step": 1729
+    },
+    {
+      "epoch": 0.002261429040147555,
+      "grad_norm": 1.3413269519805908,
+      "learning_rate": 4.14014014014014e-05,
+      "loss": 1.763,
+      "step": 1730
+    },
+    {
+      "epoch": 0.0022627362245638254,
+      "grad_norm": 1.1921231746673584,
+      "learning_rate": 4.1396396396396394e-05,
+      "loss": 1.1547,
+      "step": 1731
+    },
+    {
+      "epoch": 0.0022640434089800953,
+      "grad_norm": 1.4440979957580566,
+      "learning_rate": 4.1391391391391394e-05,
+      "loss": 1.6559,
+      "step": 1732
+    },
+    {
+      "epoch": 0.0022653505933963656,
+      "grad_norm": 2.1122000217437744,
+      "learning_rate": 4.138638638638639e-05,
+      "loss": 1.6125,
+      "step": 1733
+    },
+    {
+      "epoch": 0.002266657777812636,
+      "grad_norm": 1.5255876779556274,
+      "learning_rate": 4.138138138138139e-05,
+      "loss": 1.684,
+      "step": 1734
+    },
+    {
+      "epoch": 0.0022679649622289063,
+      "grad_norm": 2.026237726211548,
+      "learning_rate": 4.137637637637638e-05,
+      "loss": 1.6675,
+      "step": 1735
+    },
+    {
+      "epoch": 0.0022692721466451766,
+      "grad_norm": 1.1640340089797974,
+      "learning_rate": 4.137137137137137e-05,
+      "loss": 2.804,
+      "step": 1736
+    },
+    {
+      "epoch": 0.002270579331061447,
+      "grad_norm": 1.4460965394973755,
+      "learning_rate": 4.1366366366366366e-05,
+      "loss": 1.705,
+      "step": 1737
+    },
+    {
+      "epoch": 0.0022718865154777172,
+      "grad_norm": 1.4600083827972412,
+      "learning_rate": 4.136136136136136e-05,
+      "loss": 1.2554,
+      "step": 1738
+    },
+    {
+      "epoch": 0.0022731936998939876,
+      "grad_norm": 2.2812397480010986,
+      "learning_rate": 4.135635635635636e-05,
+      "loss": 1.9847,
+      "step": 1739
+    },
+    {
+      "epoch": 0.0022745008843102574,
+      "grad_norm": 1.4344135522842407,
+      "learning_rate": 4.135135135135135e-05,
+      "loss": 1.689,
+      "step": 1740
+    },
+    {
+      "epoch": 0.0022758080687265278,
+      "grad_norm": 1.0502251386642456,
+      "learning_rate": 4.134634634634635e-05,
+      "loss": 1.112,
+      "step": 1741
+    },
+    {
+      "epoch": 0.002277115253142798,
+      "grad_norm": 1.7020087242126465,
+      "learning_rate": 4.1341341341341344e-05,
+      "loss": 1.846,
+      "step": 1742
+    },
+    {
+      "epoch": 0.0022784224375590684,
+      "grad_norm": 1.3225630521774292,
+      "learning_rate": 4.133633633633634e-05,
+      "loss": 2.2314,
+      "step": 1743
+    },
+    {
+      "epoch": 0.0022797296219753387,
+      "grad_norm": 0.8521056771278381,
+      "learning_rate": 4.133133133133134e-05,
+      "loss": 0.6071,
+      "step": 1744
+    },
+    {
+      "epoch": 0.002281036806391609,
+      "grad_norm": 1.6148325204849243,
+      "learning_rate": 4.132632632632632e-05,
+      "loss": 1.3506,
+      "step": 1745
+    },
+    {
+      "epoch": 0.0022823439908078794,
+      "grad_norm": 1.4945685863494873,
+      "learning_rate": 4.132132132132132e-05,
+      "loss": 1.6745,
+      "step": 1746
+    },
+    {
+      "epoch": 0.0022836511752241492,
+      "grad_norm": 1.4772586822509766,
+      "learning_rate": 4.1316316316316316e-05,
+      "loss": 1.5633,
+      "step": 1747
+    },
+    {
+      "epoch": 0.0022849583596404196,
+      "grad_norm": 1.5608842372894287,
+      "learning_rate": 4.1311311311311316e-05,
+      "loss": 1.9242,
+      "step": 1748
+    },
+    {
+      "epoch": 0.00228626554405669,
+      "grad_norm": 1.6354891061782837,
+      "learning_rate": 4.130630630630631e-05,
+      "loss": 1.9201,
+      "step": 1749
+    },
+    {
+      "epoch": 0.00228757272847296,
+      "grad_norm": 1.755149245262146,
+      "learning_rate": 4.13013013013013e-05,
+      "loss": 1.5432,
+      "step": 1750
+    },
+    {
+      "epoch": 0.0022888799128892305,
+      "grad_norm": 1.4602986574172974,
+      "learning_rate": 4.12962962962963e-05,
+      "loss": 1.304,
+      "step": 1751
+    },
+    {
+      "epoch": 0.002290187097305501,
+      "grad_norm": 1.1095424890518188,
+      "learning_rate": 4.1291291291291294e-05,
+      "loss": 0.6218,
+      "step": 1752
+    },
+    {
+      "epoch": 0.002291494281721771,
+      "grad_norm": 1.2594953775405884,
+      "learning_rate": 4.1286286286286294e-05,
+      "loss": 0.8238,
+      "step": 1753
+    },
+    {
+      "epoch": 0.0022928014661380415,
+      "grad_norm": 2.0432305335998535,
+      "learning_rate": 4.128128128128128e-05,
+      "loss": 1.9433,
+      "step": 1754
+    },
+    {
+      "epoch": 0.0022941086505543114,
+      "grad_norm": 1.2049952745437622,
+      "learning_rate": 4.127627627627628e-05,
+      "loss": 1.1553,
+      "step": 1755
+    },
+    {
+      "epoch": 0.0022954158349705817,
+      "grad_norm": 1.4440003633499146,
+      "learning_rate": 4.127127127127127e-05,
+      "loss": 1.5383,
+      "step": 1756
+    },
+    {
+      "epoch": 0.002296723019386852,
+      "grad_norm": 1.5628281831741333,
+      "learning_rate": 4.1266266266266266e-05,
+      "loss": 1.8146,
+      "step": 1757
+    },
+    {
+      "epoch": 0.0022980302038031223,
+      "grad_norm": 1.5044841766357422,
+      "learning_rate": 4.1261261261261266e-05,
+      "loss": 2.0446,
+      "step": 1758
+    },
+    {
+      "epoch": 0.0022993373882193927,
+      "grad_norm": 1.7014210224151611,
+      "learning_rate": 4.125625625625626e-05,
+      "loss": 2.0255,
+      "step": 1759
+    },
+    {
+      "epoch": 0.002300644572635663,
+      "grad_norm": 1.363365888595581,
+      "learning_rate": 4.125125125125126e-05,
+      "loss": 1.8351,
+      "step": 1760
+    },
+    {
+      "epoch": 0.0023019517570519333,
+      "grad_norm": 1.6271450519561768,
+      "learning_rate": 4.124624624624625e-05,
+      "loss": 1.4666,
+      "step": 1761
+    },
+    {
+      "epoch": 0.002303258941468203,
+      "grad_norm": 1.4153316020965576,
+      "learning_rate": 4.124124124124124e-05,
+      "loss": 2.1277,
+      "step": 1762
+    },
+    {
+      "epoch": 0.0023045661258844735,
+      "grad_norm": 1.3964701890945435,
+      "learning_rate": 4.123623623623624e-05,
+      "loss": 1.6534,
+      "step": 1763
+    },
+    {
+      "epoch": 0.002305873310300744,
+      "grad_norm": 1.9422136545181274,
+      "learning_rate": 4.123123123123123e-05,
+      "loss": 1.7777,
+      "step": 1764
+    },
+    {
+      "epoch": 0.002307180494717014,
+      "grad_norm": 1.6836040019989014,
+      "learning_rate": 4.122622622622623e-05,
+      "loss": 2.6265,
+      "step": 1765
+    },
+    {
+      "epoch": 0.0023084876791332845,
+      "grad_norm": 1.56594717502594,
+      "learning_rate": 4.122122122122122e-05,
+      "loss": 2.0814,
+      "step": 1766
+    },
+    {
+      "epoch": 0.0023097948635495548,
+      "grad_norm": 1.9876917600631714,
+      "learning_rate": 4.1216216216216216e-05,
+      "loss": 1.3663,
+      "step": 1767
+    },
+    {
+      "epoch": 0.002311102047965825,
+      "grad_norm": 1.649053931236267,
+      "learning_rate": 4.1211211211211216e-05,
+      "loss": 0.8615,
+      "step": 1768
+    },
+    {
+      "epoch": 0.0023124092323820954,
+      "grad_norm": 1.8021605014801025,
+      "learning_rate": 4.120620620620621e-05,
+      "loss": 1.6938,
+      "step": 1769
+    },
+    {
+      "epoch": 0.0023137164167983653,
+      "grad_norm": 0.9655439853668213,
+      "learning_rate": 4.12012012012012e-05,
+      "loss": 0.3956,
+      "step": 1770
+    },
+    {
+      "epoch": 0.0023150236012146356,
+      "grad_norm": 1.3140373229980469,
+      "learning_rate": 4.1196196196196195e-05,
+      "loss": 1.205,
+      "step": 1771
+    },
+    {
+      "epoch": 0.002316330785630906,
+      "grad_norm": 1.230553388595581,
+      "learning_rate": 4.1191191191191194e-05,
+      "loss": 1.399,
+      "step": 1772
+    },
+    {
+      "epoch": 0.0023176379700471763,
+      "grad_norm": 1.3285608291625977,
+      "learning_rate": 4.118618618618619e-05,
+      "loss": 1.6984,
+      "step": 1773
+    },
+    {
+      "epoch": 0.0023189451544634466,
+      "grad_norm": 1.6323069334030151,
+      "learning_rate": 4.118118118118118e-05,
+      "loss": 2.0829,
+      "step": 1774
+    },
+    {
+      "epoch": 0.002320252338879717,
+      "grad_norm": 1.3711373805999756,
+      "learning_rate": 4.117617617617618e-05,
+      "loss": 1.9417,
+      "step": 1775
+    },
+    {
+      "epoch": 0.0023215595232959872,
+      "grad_norm": 1.34134840965271,
+      "learning_rate": 4.117117117117117e-05,
+      "loss": 1.0252,
+      "step": 1776
+    },
+    {
+      "epoch": 0.0023228667077122575,
+      "grad_norm": 1.3405780792236328,
+      "learning_rate": 4.116616616616617e-05,
+      "loss": 1.5023,
+      "step": 1777
+    },
+    {
+      "epoch": 0.0023241738921285274,
+      "grad_norm": 1.1379036903381348,
+      "learning_rate": 4.1161161161161166e-05,
+      "loss": 1.3012,
+      "step": 1778
+    },
+    {
+      "epoch": 0.0023254810765447977,
+      "grad_norm": 1.2922269105911255,
+      "learning_rate": 4.115615615615616e-05,
+      "loss": 1.8476,
+      "step": 1779
+    },
+    {
+      "epoch": 0.002326788260961068,
+      "grad_norm": 1.4011834859848022,
+      "learning_rate": 4.115115115115115e-05,
+      "loss": 2.4286,
+      "step": 1780
+    },
+    {
+      "epoch": 0.0023280954453773384,
+      "grad_norm": 1.2978572845458984,
+      "learning_rate": 4.1146146146146145e-05,
+      "loss": 1.7427,
+      "step": 1781
+    },
+    {
+      "epoch": 0.0023294026297936087,
+      "grad_norm": 1.1687116622924805,
+      "learning_rate": 4.1141141141141144e-05,
+      "loss": 0.9572,
+      "step": 1782
+    },
+    {
+      "epoch": 0.002330709814209879,
+      "grad_norm": 1.1250890493392944,
+      "learning_rate": 4.113613613613614e-05,
+      "loss": 0.996,
+      "step": 1783
+    },
+    {
+      "epoch": 0.0023320169986261493,
+      "grad_norm": 1.327339768409729,
+      "learning_rate": 4.113113113113114e-05,
+      "loss": 1.2198,
+      "step": 1784
+    },
+    {
+      "epoch": 0.0023333241830424192,
+      "grad_norm": 1.1620149612426758,
+      "learning_rate": 4.112612612612613e-05,
+      "loss": 1.1023,
+      "step": 1785
+    },
+    {
+      "epoch": 0.0023346313674586896,
+      "grad_norm": 1.1286476850509644,
+      "learning_rate": 4.112112112112112e-05,
+      "loss": 0.9342,
+      "step": 1786
+    },
+    {
+      "epoch": 0.00233593855187496,
+      "grad_norm": 1.5960549116134644,
+      "learning_rate": 4.1116116116116116e-05,
+      "loss": 2.2209,
+      "step": 1787
+    },
+    {
+      "epoch": 0.00233724573629123,
+      "grad_norm": 1.4619271755218506,
+      "learning_rate": 4.111111111111111e-05,
+      "loss": 1.4747,
+      "step": 1788
+    },
+    {
+      "epoch": 0.0023385529207075005,
+      "grad_norm": 1.3447003364562988,
+      "learning_rate": 4.110610610610611e-05,
+      "loss": 1.3043,
+      "step": 1789
+    },
+    {
+      "epoch": 0.002339860105123771,
+      "grad_norm": 1.162828803062439,
+      "learning_rate": 4.11011011011011e-05,
+      "loss": 0.5281,
+      "step": 1790
+    },
+    {
+      "epoch": 0.002341167289540041,
+      "grad_norm": 1.2304607629776,
+      "learning_rate": 4.10960960960961e-05,
+      "loss": 1.4179,
+      "step": 1791
+    },
+    {
+      "epoch": 0.0023424744739563115,
+      "grad_norm": 3.0006136894226074,
+      "learning_rate": 4.1091091091091094e-05,
+      "loss": 1.8818,
+      "step": 1792
+    },
+    {
+      "epoch": 0.0023437816583725814,
+      "grad_norm": 1.3793811798095703,
+      "learning_rate": 4.108608608608609e-05,
+      "loss": 1.8891,
+      "step": 1793
+    },
+    {
+      "epoch": 0.0023450888427888517,
+      "grad_norm": 1.3870460987091064,
+      "learning_rate": 4.108108108108109e-05,
+      "loss": 1.9547,
+      "step": 1794
+    },
+    {
+      "epoch": 0.002346396027205122,
+      "grad_norm": 1.5216132402420044,
+      "learning_rate": 4.107607607607607e-05,
+      "loss": 0.892,
+      "step": 1795
+    },
+    {
+      "epoch": 0.0023477032116213923,
+      "grad_norm": 1.814724087715149,
+      "learning_rate": 4.107107107107107e-05,
+      "loss": 1.7994,
+      "step": 1796
+    },
+    {
+      "epoch": 0.0023490103960376626,
+      "grad_norm": 1.4178881645202637,
+      "learning_rate": 4.1066066066066066e-05,
+      "loss": 1.0964,
+      "step": 1797
+    },
+    {
+      "epoch": 0.002350317580453933,
+      "grad_norm": 1.1005735397338867,
+      "learning_rate": 4.106106106106106e-05,
+      "loss": 1.3296,
+      "step": 1798
+    },
+    {
+      "epoch": 0.0023516247648702033,
+      "grad_norm": 1.7554911375045776,
+      "learning_rate": 4.105605605605606e-05,
+      "loss": 1.1702,
+      "step": 1799
+    },
+    {
+      "epoch": 0.0023529319492864736,
+      "grad_norm": 1.382185935974121,
+      "learning_rate": 4.105105105105105e-05,
+      "loss": 1.8294,
+      "step": 1800
+    },
+    {
+      "epoch": 0.0023542391337027435,
+      "grad_norm": 1.3976444005966187,
+      "learning_rate": 4.104604604604605e-05,
+      "loss": 1.3881,
+      "step": 1801
+    },
+    {
+      "epoch": 0.002355546318119014,
+      "grad_norm": 1.7478222846984863,
+      "learning_rate": 4.1041041041041045e-05,
+      "loss": 1.7133,
+      "step": 1802
+    },
+    {
+      "epoch": 0.002356853502535284,
+      "grad_norm": 1.5657448768615723,
+      "learning_rate": 4.103603603603604e-05,
+      "loss": 2.0619,
+      "step": 1803
+    },
+    {
+      "epoch": 0.0023581606869515544,
+      "grad_norm": 1.176893949508667,
+      "learning_rate": 4.103103103103103e-05,
+      "loss": 0.8228,
+      "step": 1804
+    },
+    {
+      "epoch": 0.0023594678713678248,
+      "grad_norm": 1.5230913162231445,
+      "learning_rate": 4.1026026026026023e-05,
+      "loss": 2.5582,
+      "step": 1805
+    },
+    {
+      "epoch": 0.002360775055784095,
+      "grad_norm": 1.6636601686477661,
+      "learning_rate": 4.102102102102102e-05,
+      "loss": 2.0737,
+      "step": 1806
+    },
+    {
+      "epoch": 0.0023620822402003654,
+      "grad_norm": 1.8782610893249512,
+      "learning_rate": 4.1016016016016016e-05,
+      "loss": 1.9774,
+      "step": 1807
+    },
+    {
+      "epoch": 0.0023633894246166353,
+      "grad_norm": 1.6633241176605225,
+      "learning_rate": 4.1011011011011016e-05,
+      "loss": 2.1803,
+      "step": 1808
+    },
+    {
+      "epoch": 0.0023646966090329056,
+      "grad_norm": 1.5373061895370483,
+      "learning_rate": 4.100600600600601e-05,
+      "loss": 1.7308,
+      "step": 1809
+    },
+    {
+      "epoch": 0.002366003793449176,
+      "grad_norm": 1.0809903144836426,
+      "learning_rate": 4.1001001001001e-05,
+      "loss": 0.6819,
+      "step": 1810
+    },
+    {
+      "epoch": 0.0023673109778654462,
+      "grad_norm": 1.524808645248413,
+      "learning_rate": 4.0995995995996e-05,
+      "loss": 1.9502,
+      "step": 1811
+    },
+    {
+      "epoch": 0.0023686181622817166,
+      "grad_norm": 1.3418346643447876,
+      "learning_rate": 4.099099099099099e-05,
+      "loss": 0.9002,
+      "step": 1812
+    },
+    {
+      "epoch": 0.002369925346697987,
+      "grad_norm": 1.2214679718017578,
+      "learning_rate": 4.098598598598599e-05,
+      "loss": 1.1164,
+      "step": 1813
+    },
+    {
+      "epoch": 0.002371232531114257,
+      "grad_norm": 1.233942985534668,
+      "learning_rate": 4.098098098098098e-05,
+      "loss": 0.7907,
+      "step": 1814
+    },
+    {
+      "epoch": 0.0023725397155305275,
+      "grad_norm": 1.2380965948104858,
+      "learning_rate": 4.097597597597598e-05,
+      "loss": 1.2465,
+      "step": 1815
+    },
+    {
+      "epoch": 0.0023738468999467974,
+      "grad_norm": 1.9149025678634644,
+      "learning_rate": 4.097097097097097e-05,
+      "loss": 1.6182,
+      "step": 1816
+    },
+    {
+      "epoch": 0.0023751540843630677,
+      "grad_norm": 1.27927827835083,
+      "learning_rate": 4.0965965965965966e-05,
+      "loss": 1.7556,
+      "step": 1817
+    },
+    {
+      "epoch": 0.002376461268779338,
+      "grad_norm": 1.1898010969161987,
+      "learning_rate": 4.0960960960960966e-05,
+      "loss": 1.2351,
+      "step": 1818
+    },
+    {
+      "epoch": 0.0023777684531956084,
+      "grad_norm": 1.8980669975280762,
+      "learning_rate": 4.095595595595596e-05,
+      "loss": 1.1549,
+      "step": 1819
+    },
+    {
+      "epoch": 0.0023790756376118787,
+      "grad_norm": 1.3719185590744019,
+      "learning_rate": 4.095095095095095e-05,
+      "loss": 1.4162,
+      "step": 1820
+    },
+    {
+      "epoch": 0.002380382822028149,
+      "grad_norm": 1.7730140686035156,
+      "learning_rate": 4.0945945945945945e-05,
+      "loss": 1.5128,
+      "step": 1821
+    },
+    {
+      "epoch": 0.0023816900064444193,
+      "grad_norm": 0.7194238305091858,
+      "learning_rate": 4.0940940940940945e-05,
+      "loss": 0.4416,
+      "step": 1822
+    },
+    {
+      "epoch": 0.0023829971908606896,
+      "grad_norm": 1.5401561260223389,
+      "learning_rate": 4.093593593593594e-05,
+      "loss": 1.9482,
+      "step": 1823
+    },
+    {
+      "epoch": 0.0023843043752769595,
+      "grad_norm": 1.78561532497406,
+      "learning_rate": 4.093093093093093e-05,
+      "loss": 1.8882,
+      "step": 1824
+    },
+    {
+      "epoch": 0.00238561155969323,
+      "grad_norm": 1.2883561849594116,
+      "learning_rate": 4.092592592592593e-05,
+      "loss": 1.3162,
+      "step": 1825
+    },
+    {
+      "epoch": 0.0023869187441095,
+      "grad_norm": 1.4590256214141846,
+      "learning_rate": 4.092092092092092e-05,
+      "loss": 1.8209,
+      "step": 1826
+    },
+    {
+      "epoch": 0.0023882259285257705,
+      "grad_norm": 1.4966717958450317,
+      "learning_rate": 4.091591591591592e-05,
+      "loss": 2.4019,
+      "step": 1827
+    },
+    {
+      "epoch": 0.002389533112942041,
+      "grad_norm": 1.2926746606826782,
+      "learning_rate": 4.091091091091091e-05,
+      "loss": 1.4351,
+      "step": 1828
+    },
+    {
+      "epoch": 0.002390840297358311,
+      "grad_norm": 2.1648569107055664,
+      "learning_rate": 4.090590590590591e-05,
+      "loss": 1.7974,
+      "step": 1829
+    },
+    {
+      "epoch": 0.0023921474817745815,
+      "grad_norm": 2.367114305496216,
+      "learning_rate": 4.09009009009009e-05,
+      "loss": 1.4981,
+      "step": 1830
+    },
+    {
+      "epoch": 0.0023934546661908513,
+      "grad_norm": 1.542820692062378,
+      "learning_rate": 4.0895895895895895e-05,
+      "loss": 1.8831,
+      "step": 1831
+    },
+    {
+      "epoch": 0.0023947618506071217,
+      "grad_norm": 1.3952254056930542,
+      "learning_rate": 4.0890890890890895e-05,
+      "loss": 1.3602,
+      "step": 1832
+    },
+    {
+      "epoch": 0.002396069035023392,
+      "grad_norm": 1.5103827714920044,
+      "learning_rate": 4.088588588588589e-05,
+      "loss": 2.11,
+      "step": 1833
+    },
+    {
+      "epoch": 0.0023973762194396623,
+      "grad_norm": 1.9455229043960571,
+      "learning_rate": 4.088088088088089e-05,
+      "loss": 2.1666,
+      "step": 1834
+    },
+    {
+      "epoch": 0.0023986834038559326,
+      "grad_norm": 1.4509729146957397,
+      "learning_rate": 4.087587587587588e-05,
+      "loss": 1.0151,
+      "step": 1835
+    },
+    {
+      "epoch": 0.002399990588272203,
+      "grad_norm": 1.115557074546814,
+      "learning_rate": 4.0870870870870873e-05,
+      "loss": 0.875,
+      "step": 1836
+    },
+    {
+      "epoch": 0.0024012977726884733,
+      "grad_norm": 1.4523847103118896,
+      "learning_rate": 4.0865865865865866e-05,
+      "loss": 1.6126,
+      "step": 1837
+    },
+    {
+      "epoch": 0.0024026049571047436,
+      "grad_norm": 1.2691198587417603,
+      "learning_rate": 4.086086086086086e-05,
+      "loss": 1.5322,
+      "step": 1838
+    },
+    {
+      "epoch": 0.0024039121415210135,
+      "grad_norm": 2.0400454998016357,
+      "learning_rate": 4.085585585585586e-05,
+      "loss": 1.9487,
+      "step": 1839
+    },
+    {
+      "epoch": 0.0024052193259372838,
+      "grad_norm": 2.055110454559326,
+      "learning_rate": 4.085085085085085e-05,
+      "loss": 1.1972,
+      "step": 1840
+    },
+    {
+      "epoch": 0.002406526510353554,
+      "grad_norm": 1.1029541492462158,
+      "learning_rate": 4.0845845845845845e-05,
+      "loss": 1.085,
+      "step": 1841
+    },
+    {
+      "epoch": 0.0024078336947698244,
+      "grad_norm": 1.6044203042984009,
+      "learning_rate": 4.0840840840840845e-05,
+      "loss": 1.8803,
+      "step": 1842
+    },
+    {
+      "epoch": 0.0024091408791860947,
+      "grad_norm": 1.359581708908081,
+      "learning_rate": 4.083583583583584e-05,
+      "loss": 1.7361,
+      "step": 1843
+    },
+    {
+      "epoch": 0.002410448063602365,
+      "grad_norm": 1.980474591255188,
+      "learning_rate": 4.083083083083084e-05,
+      "loss": 1.7598,
+      "step": 1844
+    },
+    {
+      "epoch": 0.0024117552480186354,
+      "grad_norm": 1.3238515853881836,
+      "learning_rate": 4.0825825825825824e-05,
+      "loss": 1.2756,
+      "step": 1845
+    },
+    {
+      "epoch": 0.0024130624324349057,
+      "grad_norm": 1.2607429027557373,
+      "learning_rate": 4.0820820820820823e-05,
+      "loss": 0.993,
+      "step": 1846
+    },
+    {
+      "epoch": 0.0024143696168511756,
+      "grad_norm": 1.588477611541748,
+      "learning_rate": 4.0815815815815816e-05,
+      "loss": 1.5152,
+      "step": 1847
+    },
+    {
+      "epoch": 0.002415676801267446,
+      "grad_norm": 1.5605615377426147,
+      "learning_rate": 4.081081081081081e-05,
+      "loss": 1.7639,
+      "step": 1848
+    },
+    {
+      "epoch": 0.0024169839856837162,
+      "grad_norm": 1.6800919771194458,
+      "learning_rate": 4.080580580580581e-05,
+      "loss": 1.9297,
+      "step": 1849
+    },
+    {
+      "epoch": 0.0024182911700999865,
+      "grad_norm": 2.24501895904541,
+      "learning_rate": 4.08008008008008e-05,
+      "loss": 1.5989,
+      "step": 1850
+    },
+    {
+      "epoch": 0.002419598354516257,
+      "grad_norm": 1.3869489431381226,
+      "learning_rate": 4.07957957957958e-05,
+      "loss": 1.5776,
+      "step": 1851
+    },
+    {
+      "epoch": 0.002420905538932527,
+      "grad_norm": 1.3336656093597412,
+      "learning_rate": 4.0790790790790795e-05,
+      "loss": 1.4882,
+      "step": 1852
+    },
+    {
+      "epoch": 0.0024222127233487975,
+      "grad_norm": 1.4191880226135254,
+      "learning_rate": 4.078578578578579e-05,
+      "loss": 1.8135,
+      "step": 1853
+    },
+    {
+      "epoch": 0.0024235199077650674,
+      "grad_norm": 1.401842474937439,
+      "learning_rate": 4.078078078078078e-05,
+      "loss": 1.0761,
+      "step": 1854
+    },
+    {
+      "epoch": 0.0024248270921813377,
+      "grad_norm": 1.2266510725021362,
+      "learning_rate": 4.0775775775775774e-05,
+      "loss": 1.3525,
+      "step": 1855
+    },
+    {
+      "epoch": 0.002426134276597608,
+      "grad_norm": 1.4883300065994263,
+      "learning_rate": 4.0770770770770774e-05,
+      "loss": 1.8111,
+      "step": 1856
+    },
+    {
+      "epoch": 0.0024274414610138783,
+      "grad_norm": 1.2926830053329468,
+      "learning_rate": 4.0765765765765767e-05,
+      "loss": 0.8082,
+      "step": 1857
+    },
+    {
+      "epoch": 0.0024287486454301487,
+      "grad_norm": 1.3773339986801147,
+      "learning_rate": 4.0760760760760766e-05,
+      "loss": 1.1261,
+      "step": 1858
+    },
+    {
+      "epoch": 0.002430055829846419,
+      "grad_norm": 1.3188133239746094,
+      "learning_rate": 4.075575575575576e-05,
+      "loss": 1.4432,
+      "step": 1859
+    },
+    {
+      "epoch": 0.0024313630142626893,
+      "grad_norm": 1.6541621685028076,
+      "learning_rate": 4.075075075075075e-05,
+      "loss": 1.9469,
+      "step": 1860
+    },
+    {
+      "epoch": 0.0024326701986789596,
+      "grad_norm": 1.2363802194595337,
+      "learning_rate": 4.074574574574575e-05,
+      "loss": 1.0811,
+      "step": 1861
+    },
+    {
+      "epoch": 0.0024339773830952295,
+      "grad_norm": 1.7349998950958252,
+      "learning_rate": 4.074074074074074e-05,
+      "loss": 1.9156,
+      "step": 1862
+    },
+    {
+      "epoch": 0.0024352845675115,
+      "grad_norm": 1.9161877632141113,
+      "learning_rate": 4.073573573573574e-05,
+      "loss": 1.7063,
+      "step": 1863
+    },
+    {
+      "epoch": 0.00243659175192777,
+      "grad_norm": 1.2534571886062622,
+      "learning_rate": 4.073073073073073e-05,
+      "loss": 1.8379,
+      "step": 1864
+    },
+    {
+      "epoch": 0.0024378989363440405,
+      "grad_norm": 1.8141995668411255,
+      "learning_rate": 4.072572572572573e-05,
+      "loss": 2.2152,
+      "step": 1865
+    },
+    {
+      "epoch": 0.002439206120760311,
+      "grad_norm": 1.1039230823516846,
+      "learning_rate": 4.0720720720720724e-05,
+      "loss": 0.9145,
+      "step": 1866
+    },
+    {
+      "epoch": 0.002440513305176581,
+      "grad_norm": 1.3672388792037964,
+      "learning_rate": 4.0715715715715717e-05,
+      "loss": 1.63,
+      "step": 1867
+    },
+    {
+      "epoch": 0.0024418204895928514,
+      "grad_norm": 1.5401558876037598,
+      "learning_rate": 4.0710710710710716e-05,
+      "loss": 1.8055,
+      "step": 1868
+    },
+    {
+      "epoch": 0.0024431276740091213,
+      "grad_norm": 1.5458837747573853,
+      "learning_rate": 4.070570570570571e-05,
+      "loss": 1.5483,
+      "step": 1869
+    },
+    {
+      "epoch": 0.0024444348584253916,
+      "grad_norm": 1.4851856231689453,
+      "learning_rate": 4.07007007007007e-05,
+      "loss": 1.1762,
+      "step": 1870
+    },
+    {
+      "epoch": 0.002445742042841662,
+      "grad_norm": 1.5440478324890137,
+      "learning_rate": 4.0695695695695695e-05,
+      "loss": 1.7384,
+      "step": 1871
+    },
+    {
+      "epoch": 0.0024470492272579323,
+      "grad_norm": 1.5548888444900513,
+      "learning_rate": 4.069069069069069e-05,
+      "loss": 1.1735,
+      "step": 1872
+    },
+    {
+      "epoch": 0.0024483564116742026,
+      "grad_norm": 1.3652890920639038,
+      "learning_rate": 4.068568568568569e-05,
+      "loss": 1.3469,
+      "step": 1873
+    },
+    {
+      "epoch": 0.002449663596090473,
+      "grad_norm": 1.7792798280715942,
+      "learning_rate": 4.068068068068068e-05,
+      "loss": 2.4703,
+      "step": 1874
+    },
+    {
+      "epoch": 0.0024509707805067432,
+      "grad_norm": 2.163743495941162,
+      "learning_rate": 4.067567567567568e-05,
+      "loss": 1.7874,
+      "step": 1875
+    },
+    {
+      "epoch": 0.0024522779649230136,
+      "grad_norm": 1.1725695133209229,
+      "learning_rate": 4.0670670670670674e-05,
+      "loss": 1.0448,
+      "step": 1876
+    },
+    {
+      "epoch": 0.0024535851493392834,
+      "grad_norm": 1.5567781925201416,
+      "learning_rate": 4.066566566566567e-05,
+      "loss": 1.7695,
+      "step": 1877
+    },
+    {
+      "epoch": 0.0024548923337555538,
+      "grad_norm": 1.2569833993911743,
+      "learning_rate": 4.066066066066066e-05,
+      "loss": 1.1311,
+      "step": 1878
+    },
+    {
+      "epoch": 0.002456199518171824,
+      "grad_norm": 1.9324285984039307,
+      "learning_rate": 4.065565565565565e-05,
+      "loss": 1.9852,
+      "step": 1879
+    },
+    {
+      "epoch": 0.0024575067025880944,
+      "grad_norm": 1.5613092184066772,
+      "learning_rate": 4.065065065065065e-05,
+      "loss": 1.6148,
+      "step": 1880
+    },
+    {
+      "epoch": 0.0024588138870043647,
+      "grad_norm": 1.5827393531799316,
+      "learning_rate": 4.0645645645645645e-05,
+      "loss": 2.1625,
+      "step": 1881
+    },
+    {
+      "epoch": 0.002460121071420635,
+      "grad_norm": 1.3100626468658447,
+      "learning_rate": 4.0640640640640645e-05,
+      "loss": 1.2584,
+      "step": 1882
+    },
+    {
+      "epoch": 0.0024614282558369054,
+      "grad_norm": 1.599975347518921,
+      "learning_rate": 4.063563563563564e-05,
+      "loss": 1.9911,
+      "step": 1883
+    },
+    {
+      "epoch": 0.0024627354402531757,
+      "grad_norm": 1.5361524820327759,
+      "learning_rate": 4.063063063063063e-05,
+      "loss": 1.8284,
+      "step": 1884
+    },
+    {
+      "epoch": 0.0024640426246694456,
+      "grad_norm": 1.5422202348709106,
+      "learning_rate": 4.062562562562563e-05,
+      "loss": 1.4201,
+      "step": 1885
+    },
+    {
+      "epoch": 0.002465349809085716,
+      "grad_norm": 1.3174513578414917,
+      "learning_rate": 4.0620620620620624e-05,
+      "loss": 1.984,
+      "step": 1886
+    },
+    {
+      "epoch": 0.002466656993501986,
+      "grad_norm": 1.3149579763412476,
+      "learning_rate": 4.061561561561562e-05,
+      "loss": 1.0747,
+      "step": 1887
+    },
+    {
+      "epoch": 0.0024679641779182565,
+      "grad_norm": 1.3696186542510986,
+      "learning_rate": 4.061061061061061e-05,
+      "loss": 1.3481,
+      "step": 1888
+    },
+    {
+      "epoch": 0.002469271362334527,
+      "grad_norm": 1.2815920114517212,
+      "learning_rate": 4.060560560560561e-05,
+      "loss": 1.0309,
+      "step": 1889
+    },
+    {
+      "epoch": 0.002470578546750797,
+      "grad_norm": 1.7338346242904663,
+      "learning_rate": 4.06006006006006e-05,
+      "loss": 1.7711,
+      "step": 1890
+    },
+    {
+      "epoch": 0.0024718857311670675,
+      "grad_norm": 1.3611040115356445,
+      "learning_rate": 4.0595595595595595e-05,
+      "loss": 1.563,
+      "step": 1891
+    },
+    {
+      "epoch": 0.0024731929155833374,
+      "grad_norm": 0.8128279447555542,
+      "learning_rate": 4.0590590590590595e-05,
+      "loss": 0.6275,
+      "step": 1892
+    },
+    {
+      "epoch": 0.0024745000999996077,
+      "grad_norm": 1.2525824308395386,
+      "learning_rate": 4.058558558558559e-05,
+      "loss": 1.5625,
+      "step": 1893
+    },
+    {
+      "epoch": 0.002475807284415878,
+      "grad_norm": 1.875914454460144,
+      "learning_rate": 4.058058058058059e-05,
+      "loss": 1.7791,
+      "step": 1894
+    },
+    {
+      "epoch": 0.0024771144688321483,
+      "grad_norm": 1.9506193399429321,
+      "learning_rate": 4.0575575575575574e-05,
+      "loss": 1.9472,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0024784216532484186,
+      "grad_norm": 1.4316461086273193,
+      "learning_rate": 4.0570570570570574e-05,
+      "loss": 1.0125,
+      "step": 1896
+    },
+    {
+      "epoch": 0.002479728837664689,
+      "grad_norm": 1.3549818992614746,
+      "learning_rate": 4.056556556556557e-05,
+      "loss": 1.5268,
+      "step": 1897
+    },
+    {
+      "epoch": 0.0024810360220809593,
+      "grad_norm": 1.5343955755233765,
+      "learning_rate": 4.056056056056056e-05,
+      "loss": 2.3454,
+      "step": 1898
+    },
+    {
+      "epoch": 0.0024823432064972296,
+      "grad_norm": 1.478969931602478,
+      "learning_rate": 4.055555555555556e-05,
+      "loss": 1.942,
+      "step": 1899
+    },
+    {
+      "epoch": 0.0024836503909134995,
+      "grad_norm": 1.3843867778778076,
+      "learning_rate": 4.055055055055055e-05,
+      "loss": 2.5211,
+      "step": 1900
+    },
+    {
+      "epoch": 0.00248495757532977,
+      "grad_norm": 1.547825813293457,
+      "learning_rate": 4.054554554554555e-05,
+      "loss": 2.1782,
+      "step": 1901
+    },
+    {
+      "epoch": 0.00248626475974604,
+      "grad_norm": 1.6056628227233887,
+      "learning_rate": 4.0540540540540545e-05,
+      "loss": 1.7122,
+      "step": 1902
+    },
+    {
+      "epoch": 0.0024875719441623105,
+      "grad_norm": 1.370471715927124,
+      "learning_rate": 4.053553553553554e-05,
+      "loss": 1.3515,
+      "step": 1903
+    },
+    {
+      "epoch": 0.0024888791285785808,
+      "grad_norm": 1.641126036643982,
+      "learning_rate": 4.053053053053053e-05,
+      "loss": 1.9688,
+      "step": 1904
+    },
+    {
+      "epoch": 0.002490186312994851,
+      "grad_norm": 1.337172508239746,
+      "learning_rate": 4.0525525525525524e-05,
+      "loss": 1.9834,
+      "step": 1905
+    },
+    {
+      "epoch": 0.0024914934974111214,
+      "grad_norm": 1.321199893951416,
+      "learning_rate": 4.0520520520520524e-05,
+      "loss": 2.0242,
+      "step": 1906
+    },
+    {
+      "epoch": 0.0024928006818273917,
+      "grad_norm": 1.5153032541275024,
+      "learning_rate": 4.051551551551552e-05,
+      "loss": 2.0397,
+      "step": 1907
+    },
+    {
+      "epoch": 0.0024941078662436616,
+      "grad_norm": 1.1876680850982666,
+      "learning_rate": 4.051051051051051e-05,
+      "loss": 0.9109,
+      "step": 1908
+    },
+    {
+      "epoch": 0.002495415050659932,
+      "grad_norm": 1.0358036756515503,
+      "learning_rate": 4.050550550550551e-05,
+      "loss": 0.6898,
+      "step": 1909
+    },
+    {
+      "epoch": 0.0024967222350762023,
+      "grad_norm": 1.5187417268753052,
+      "learning_rate": 4.05005005005005e-05,
+      "loss": 1.6835,
+      "step": 1910
+    },
+    {
+      "epoch": 0.0024980294194924726,
+      "grad_norm": 1.3309277296066284,
+      "learning_rate": 4.04954954954955e-05,
+      "loss": 1.7915,
+      "step": 1911
+    },
+    {
+      "epoch": 0.002499336603908743,
+      "grad_norm": 1.4643735885620117,
+      "learning_rate": 4.049049049049049e-05,
+      "loss": 2.1732,
+      "step": 1912
+    },
+    {
+      "epoch": 0.0025006437883250132,
+      "grad_norm": 1.4383355379104614,
+      "learning_rate": 4.048548548548549e-05,
+      "loss": 1.4313,
+      "step": 1913
+    },
+    {
+      "epoch": 0.0025019509727412835,
+      "grad_norm": 1.5094817876815796,
+      "learning_rate": 4.048048048048048e-05,
+      "loss": 1.8269,
+      "step": 1914
+    },
+    {
+      "epoch": 0.0025032581571575534,
+      "grad_norm": 1.3402355909347534,
+      "learning_rate": 4.0475475475475474e-05,
+      "loss": 1.8552,
+      "step": 1915
+    },
+    {
+      "epoch": 0.0025045653415738237,
+      "grad_norm": 1.8498746156692505,
+      "learning_rate": 4.0470470470470474e-05,
+      "loss": 1.8181,
+      "step": 1916
+    },
+    {
+      "epoch": 0.002505872525990094,
+      "grad_norm": 1.3429100513458252,
+      "learning_rate": 4.046546546546547e-05,
+      "loss": 1.0496,
+      "step": 1917
+    },
+    {
+      "epoch": 0.0025071797104063644,
+      "grad_norm": 1.4095878601074219,
+      "learning_rate": 4.046046046046047e-05,
+      "loss": 1.5425,
+      "step": 1918
+    },
+    {
+      "epoch": 0.0025084868948226347,
+      "grad_norm": 1.2426705360412598,
+      "learning_rate": 4.045545545545546e-05,
+      "loss": 1.4998,
+      "step": 1919
+    },
+    {
+      "epoch": 0.002509794079238905,
+      "grad_norm": 1.6159253120422363,
+      "learning_rate": 4.045045045045045e-05,
+      "loss": 1.0457,
+      "step": 1920
+    },
+    {
+      "epoch": 0.0025111012636551753,
+      "grad_norm": 2.152106761932373,
+      "learning_rate": 4.0445445445445446e-05,
+      "loss": 2.252,
+      "step": 1921
+    },
+    {
+      "epoch": 0.0025124084480714457,
+      "grad_norm": 1.119509220123291,
+      "learning_rate": 4.044044044044044e-05,
+      "loss": 1.1803,
+      "step": 1922
+    },
+    {
+      "epoch": 0.0025137156324877155,
+      "grad_norm": 1.5565276145935059,
+      "learning_rate": 4.043543543543544e-05,
+      "loss": 1.261,
+      "step": 1923
+    },
+    {
+      "epoch": 0.002515022816903986,
+      "grad_norm": 1.8533421754837036,
+      "learning_rate": 4.043043043043043e-05,
+      "loss": 1.809,
+      "step": 1924
+    },
+    {
+      "epoch": 0.002516330001320256,
+      "grad_norm": 1.7663089036941528,
+      "learning_rate": 4.042542542542543e-05,
+      "loss": 1.6891,
+      "step": 1925
+    },
+    {
+      "epoch": 0.0025176371857365265,
+      "grad_norm": 1.69199800491333,
+      "learning_rate": 4.0420420420420424e-05,
+      "loss": 2.1076,
+      "step": 1926
+    },
+    {
+      "epoch": 0.002518944370152797,
+      "grad_norm": 1.298394799232483,
+      "learning_rate": 4.041541541541542e-05,
+      "loss": 1.1322,
+      "step": 1927
+    },
+    {
+      "epoch": 0.002520251554569067,
+      "grad_norm": 1.31342613697052,
+      "learning_rate": 4.041041041041041e-05,
+      "loss": 1.5266,
+      "step": 1928
+    },
+    {
+      "epoch": 0.0025215587389853375,
+      "grad_norm": 1.822007417678833,
+      "learning_rate": 4.04054054054054e-05,
+      "loss": 2.1545,
+      "step": 1929
+    },
+    {
+      "epoch": 0.002522865923401608,
+      "grad_norm": 1.5367997884750366,
+      "learning_rate": 4.04004004004004e-05,
+      "loss": 2.3631,
+      "step": 1930
+    },
+    {
+      "epoch": 0.0025241731078178777,
+      "grad_norm": 1.2048085927963257,
+      "learning_rate": 4.0395395395395396e-05,
+      "loss": 1.5117,
+      "step": 1931
+    },
+    {
+      "epoch": 0.002525480292234148,
+      "grad_norm": 1.5233961343765259,
+      "learning_rate": 4.0390390390390395e-05,
+      "loss": 1.388,
+      "step": 1932
+    },
+    {
+      "epoch": 0.0025267874766504183,
+      "grad_norm": 1.1803560256958008,
+      "learning_rate": 4.038538538538539e-05,
+      "loss": 1.2915,
+      "step": 1933
+    },
+    {
+      "epoch": 0.0025280946610666886,
+      "grad_norm": 1.5423097610473633,
+      "learning_rate": 4.038038038038038e-05,
+      "loss": 1.4207,
+      "step": 1934
+    },
+    {
+      "epoch": 0.002529401845482959,
+      "grad_norm": 1.8734697103500366,
+      "learning_rate": 4.037537537537538e-05,
+      "loss": 0.9798,
+      "step": 1935
+    },
+    {
+      "epoch": 0.0025307090298992293,
+      "grad_norm": 1.6678602695465088,
+      "learning_rate": 4.0370370370370374e-05,
+      "loss": 1.3318,
+      "step": 1936
+    },
+    {
+      "epoch": 0.0025320162143154996,
+      "grad_norm": 1.0457793474197388,
+      "learning_rate": 4.036536536536537e-05,
+      "loss": 0.9126,
+      "step": 1937
+    },
+    {
+      "epoch": 0.0025333233987317695,
+      "grad_norm": 1.256769061088562,
+      "learning_rate": 4.036036036036036e-05,
+      "loss": 1.0549,
+      "step": 1938
+    },
+    {
+      "epoch": 0.00253463058314804,
+      "grad_norm": 0.8580578565597534,
+      "learning_rate": 4.035535535535536e-05,
+      "loss": 0.6175,
+      "step": 1939
+    },
+    {
+      "epoch": 0.00253593776756431,
+      "grad_norm": 1.4798988103866577,
+      "learning_rate": 4.035035035035035e-05,
+      "loss": 1.3529,
+      "step": 1940
+    },
+    {
+      "epoch": 0.0025372449519805804,
+      "grad_norm": 1.7928217649459839,
+      "learning_rate": 4.0345345345345346e-05,
+      "loss": 2.6954,
+      "step": 1941
+    },
+    {
+      "epoch": 0.0025385521363968508,
+      "grad_norm": 1.3230857849121094,
+      "learning_rate": 4.0340340340340346e-05,
+      "loss": 1.1748,
+      "step": 1942
+    },
+    {
+      "epoch": 0.002539859320813121,
+      "grad_norm": 1.2928690910339355,
+      "learning_rate": 4.033533533533534e-05,
+      "loss": 1.2122,
+      "step": 1943
+    },
+    {
+      "epoch": 0.0025411665052293914,
+      "grad_norm": 2.0625734329223633,
+      "learning_rate": 4.033033033033033e-05,
+      "loss": 1.9035,
+      "step": 1944
+    },
+    {
+      "epoch": 0.0025424736896456617,
+      "grad_norm": 1.2294127941131592,
+      "learning_rate": 4.0325325325325324e-05,
+      "loss": 1.8425,
+      "step": 1945
+    },
+    {
+      "epoch": 0.0025437808740619316,
+      "grad_norm": 0.802227795124054,
+      "learning_rate": 4.032032032032032e-05,
+      "loss": 0.4178,
+      "step": 1946
+    },
+    {
+      "epoch": 0.002545088058478202,
+      "grad_norm": 1.7256674766540527,
+      "learning_rate": 4.031531531531532e-05,
+      "loss": 1.9568,
+      "step": 1947
+    },
+    {
+      "epoch": 0.0025463952428944722,
+      "grad_norm": 1.5038264989852905,
+      "learning_rate": 4.031031031031031e-05,
+      "loss": 1.4808,
+      "step": 1948
+    },
+    {
+      "epoch": 0.0025477024273107426,
+      "grad_norm": 0.9010105133056641,
+      "learning_rate": 4.030530530530531e-05,
+      "loss": 0.5353,
+      "step": 1949
+    },
+    {
+      "epoch": 0.002549009611727013,
+      "grad_norm": 3.577693223953247,
+      "learning_rate": 4.03003003003003e-05,
+      "loss": 1.5282,
+      "step": 1950
+    },
+    {
+      "epoch": 0.002550316796143283,
+      "grad_norm": 1.4891083240509033,
+      "learning_rate": 4.0295295295295296e-05,
+      "loss": 1.7715,
+      "step": 1951
+    },
+    {
+      "epoch": 0.0025516239805595535,
+      "grad_norm": 1.3520947694778442,
+      "learning_rate": 4.0290290290290296e-05,
+      "loss": 1.4332,
+      "step": 1952
+    },
+    {
+      "epoch": 0.002552931164975824,
+      "grad_norm": 1.3019458055496216,
+      "learning_rate": 4.028528528528528e-05,
+      "loss": 0.7576,
+      "step": 1953
+    },
+    {
+      "epoch": 0.0025542383493920937,
+      "grad_norm": 1.379707932472229,
+      "learning_rate": 4.028028028028028e-05,
+      "loss": 1.0357,
+      "step": 1954
+    },
+    {
+      "epoch": 0.002555545533808364,
+      "grad_norm": 1.4250497817993164,
+      "learning_rate": 4.0275275275275275e-05,
+      "loss": 1.1909,
+      "step": 1955
+    },
+    {
+      "epoch": 0.0025568527182246344,
+      "grad_norm": 1.26385498046875,
+      "learning_rate": 4.0270270270270274e-05,
+      "loss": 0.9951,
+      "step": 1956
+    },
+    {
+      "epoch": 0.0025581599026409047,
+      "grad_norm": 1.3548882007598877,
+      "learning_rate": 4.026526526526527e-05,
+      "loss": 1.9509,
+      "step": 1957
+    },
+    {
+      "epoch": 0.002559467087057175,
+      "grad_norm": 1.2190290689468384,
+      "learning_rate": 4.026026026026026e-05,
+      "loss": 0.722,
+      "step": 1958
+    },
+    {
+      "epoch": 0.0025607742714734453,
+      "grad_norm": 1.4614123106002808,
+      "learning_rate": 4.025525525525526e-05,
+      "loss": 2.0323,
+      "step": 1959
+    },
+    {
+      "epoch": 0.0025620814558897156,
+      "grad_norm": 1.51620352268219,
+      "learning_rate": 4.025025025025025e-05,
+      "loss": 1.7395,
+      "step": 1960
+    },
+    {
+      "epoch": 0.0025633886403059855,
+      "grad_norm": 1.1835839748382568,
+      "learning_rate": 4.0245245245245246e-05,
+      "loss": 1.2997,
+      "step": 1961
+    },
+    {
+      "epoch": 0.002564695824722256,
+      "grad_norm": 1.3911057710647583,
+      "learning_rate": 4.024024024024024e-05,
+      "loss": 1.6148,
+      "step": 1962
+    },
+    {
+      "epoch": 0.002566003009138526,
+      "grad_norm": 2.3794503211975098,
+      "learning_rate": 4.023523523523524e-05,
+      "loss": 1.079,
+      "step": 1963
+    },
+    {
+      "epoch": 0.0025673101935547965,
+      "grad_norm": 1.6634477376937866,
+      "learning_rate": 4.023023023023023e-05,
+      "loss": 2.1782,
+      "step": 1964
+    },
+    {
+      "epoch": 0.002568617377971067,
+      "grad_norm": 1.5287396907806396,
+      "learning_rate": 4.0225225225225225e-05,
+      "loss": 1.4655,
+      "step": 1965
+    },
+    {
+      "epoch": 0.002569924562387337,
+      "grad_norm": 1.8702428340911865,
+      "learning_rate": 4.0220220220220224e-05,
+      "loss": 1.2162,
+      "step": 1966
+    },
+    {
+      "epoch": 0.0025712317468036074,
+      "grad_norm": 1.9908095598220825,
+      "learning_rate": 4.021521521521522e-05,
+      "loss": 2.7337,
+      "step": 1967
+    },
+    {
+      "epoch": 0.0025725389312198778,
+      "grad_norm": 1.1542783975601196,
+      "learning_rate": 4.021021021021022e-05,
+      "loss": 1.2294,
+      "step": 1968
+    },
+    {
+      "epoch": 0.0025738461156361477,
+      "grad_norm": 1.4157741069793701,
+      "learning_rate": 4.020520520520521e-05,
+      "loss": 1.3498,
+      "step": 1969
+    },
+    {
+      "epoch": 0.002575153300052418,
+      "grad_norm": 1.5515310764312744,
+      "learning_rate": 4.02002002002002e-05,
+      "loss": 1.8184,
+      "step": 1970
+    },
+    {
+      "epoch": 0.0025764604844686883,
+      "grad_norm": 1.3329187631607056,
+      "learning_rate": 4.0195195195195196e-05,
+      "loss": 1.1111,
+      "step": 1971
+    },
+    {
+      "epoch": 0.0025777676688849586,
+      "grad_norm": 2.2171847820281982,
+      "learning_rate": 4.019019019019019e-05,
+      "loss": 1.5755,
+      "step": 1972
+    },
+    {
+      "epoch": 0.002579074853301229,
+      "grad_norm": 1.5465325117111206,
+      "learning_rate": 4.018518518518519e-05,
+      "loss": 1.8367,
+      "step": 1973
+    },
+    {
+      "epoch": 0.0025803820377174993,
+      "grad_norm": 1.8722420930862427,
+      "learning_rate": 4.018018018018018e-05,
+      "loss": 2.1407,
+      "step": 1974
+    },
+    {
+      "epoch": 0.0025816892221337696,
+      "grad_norm": 1.4971617460250854,
+      "learning_rate": 4.017517517517518e-05,
+      "loss": 1.193,
+      "step": 1975
+    },
+    {
+      "epoch": 0.0025829964065500395,
+      "grad_norm": 1.862038016319275,
+      "learning_rate": 4.0170170170170174e-05,
+      "loss": 1.8063,
+      "step": 1976
+    },
+    {
+      "epoch": 0.0025843035909663098,
+      "grad_norm": 1.9676222801208496,
+      "learning_rate": 4.016516516516517e-05,
+      "loss": 1.899,
+      "step": 1977
+    },
+    {
+      "epoch": 0.00258561077538258,
+      "grad_norm": 0.8395018577575684,
+      "learning_rate": 4.016016016016016e-05,
+      "loss": 0.3616,
+      "step": 1978
+    },
+    {
+      "epoch": 0.0025869179597988504,
+      "grad_norm": 1.3533786535263062,
+      "learning_rate": 4.015515515515515e-05,
+      "loss": 1.9378,
+      "step": 1979
+    },
+    {
+      "epoch": 0.0025882251442151207,
+      "grad_norm": 1.3205751180648804,
+      "learning_rate": 4.015015015015015e-05,
+      "loss": 1.5552,
+      "step": 1980
+    },
+    {
+      "epoch": 0.002589532328631391,
+      "grad_norm": 1.4081898927688599,
+      "learning_rate": 4.0145145145145146e-05,
+      "loss": 2.2908,
+      "step": 1981
+    },
+    {
+      "epoch": 0.0025908395130476614,
+      "grad_norm": 1.3975245952606201,
+      "learning_rate": 4.014014014014014e-05,
+      "loss": 1.0861,
+      "step": 1982
+    },
+    {
+      "epoch": 0.0025921466974639317,
+      "grad_norm": 1.5519052743911743,
+      "learning_rate": 4.013513513513514e-05,
+      "loss": 2.4197,
+      "step": 1983
+    },
+    {
+      "epoch": 0.0025934538818802016,
+      "grad_norm": 1.4379628896713257,
+      "learning_rate": 4.013013013013013e-05,
+      "loss": 1.2807,
+      "step": 1984
+    },
+    {
+      "epoch": 0.002594761066296472,
+      "grad_norm": 1.399878978729248,
+      "learning_rate": 4.012512512512513e-05,
+      "loss": 1.6349,
+      "step": 1985
+    },
+    {
+      "epoch": 0.0025960682507127422,
+      "grad_norm": 1.855602502822876,
+      "learning_rate": 4.012012012012012e-05,
+      "loss": 1.8093,
+      "step": 1986
+    },
+    {
+      "epoch": 0.0025973754351290125,
+      "grad_norm": 1.6053335666656494,
+      "learning_rate": 4.011511511511512e-05,
+      "loss": 2.0968,
+      "step": 1987
+    },
+    {
+      "epoch": 0.002598682619545283,
+      "grad_norm": 1.2241007089614868,
+      "learning_rate": 4.011011011011011e-05,
+      "loss": 1.3309,
+      "step": 1988
+    },
+    {
+      "epoch": 0.002599989803961553,
+      "grad_norm": 1.6596977710723877,
+      "learning_rate": 4.01051051051051e-05,
+      "loss": 1.9482,
+      "step": 1989
+    },
+    {
+      "epoch": 0.0026012969883778235,
+      "grad_norm": 1.5003933906555176,
+      "learning_rate": 4.01001001001001e-05,
+      "loss": 1.4497,
+      "step": 1990
+    },
+    {
+      "epoch": 0.002602604172794094,
+      "grad_norm": 1.357008934020996,
+      "learning_rate": 4.0095095095095096e-05,
+      "loss": 0.8863,
+      "step": 1991
+    },
+    {
+      "epoch": 0.0026039113572103637,
+      "grad_norm": 1.210019588470459,
+      "learning_rate": 4.0090090090090096e-05,
+      "loss": 0.9477,
+      "step": 1992
+    },
+    {
+      "epoch": 0.002605218541626634,
+      "grad_norm": 1.0514651536941528,
+      "learning_rate": 4.008508508508509e-05,
+      "loss": 0.6505,
+      "step": 1993
+    },
+    {
+      "epoch": 0.0026065257260429043,
+      "grad_norm": 1.6571133136749268,
+      "learning_rate": 4.008008008008008e-05,
+      "loss": 2.1003,
+      "step": 1994
+    },
+    {
+      "epoch": 0.0026078329104591747,
+      "grad_norm": 0.6768779158592224,
+      "learning_rate": 4.0075075075075075e-05,
+      "loss": 0.0896,
+      "step": 1995
+    },
+    {
+      "epoch": 0.002609140094875445,
+      "grad_norm": 1.8878520727157593,
+      "learning_rate": 4.007007007007007e-05,
+      "loss": 1.2328,
+      "step": 1996
+    },
+    {
+      "epoch": 0.0026104472792917153,
+      "grad_norm": 1.4446895122528076,
+      "learning_rate": 4.006506506506507e-05,
+      "loss": 1.6896,
+      "step": 1997
+    },
+    {
+      "epoch": 0.0026117544637079856,
+      "grad_norm": 1.5758342742919922,
+      "learning_rate": 4.006006006006006e-05,
+      "loss": 1.9928,
+      "step": 1998
+    },
+    {
+      "epoch": 0.0026130616481242555,
+      "grad_norm": 1.4952720403671265,
+      "learning_rate": 4.005505505505506e-05,
+      "loss": 1.2204,
+      "step": 1999
+    },
+    {
+      "epoch": 0.002614368832540526,
+      "grad_norm": 2.0945303440093994,
+      "learning_rate": 4.005005005005005e-05,
+      "loss": 1.3566,
+      "step": 2000
+    },
+    {
+      "epoch": 0.002615676016956796,
+      "grad_norm": 1.2122732400894165,
+      "learning_rate": 4.0045045045045046e-05,
+      "loss": 1.4865,
+      "step": 2001
+    },
+    {
+      "epoch": 0.0026169832013730665,
+      "grad_norm": 1.2634263038635254,
+      "learning_rate": 4.0040040040040046e-05,
+      "loss": 0.9814,
+      "step": 2002
+    },
+    {
+      "epoch": 0.002618290385789337,
+      "grad_norm": 1.1941736936569214,
+      "learning_rate": 4.003503503503503e-05,
+      "loss": 1.0253,
+      "step": 2003
+    },
+    {
+      "epoch": 0.002619597570205607,
+      "grad_norm": 1.5035587549209595,
+      "learning_rate": 4.003003003003003e-05,
+      "loss": 1.5252,
+      "step": 2004
+    },
+    {
+      "epoch": 0.0026209047546218774,
+      "grad_norm": 0.754768431186676,
+      "learning_rate": 4.0025025025025025e-05,
+      "loss": 0.5366,
+      "step": 2005
+    },
+    {
+      "epoch": 0.0026222119390381477,
+      "grad_norm": 1.6248060464859009,
+      "learning_rate": 4.0020020020020025e-05,
+      "loss": 2.1113,
+      "step": 2006
+    },
+    {
+      "epoch": 0.0026235191234544176,
+      "grad_norm": 1.7528574466705322,
+      "learning_rate": 4.001501501501502e-05,
+      "loss": 1.6915,
+      "step": 2007
+    },
+    {
+      "epoch": 0.002624826307870688,
+      "grad_norm": 0.9582585692405701,
+      "learning_rate": 4.001001001001001e-05,
+      "loss": 0.6811,
+      "step": 2008
+    },
+    {
+      "epoch": 0.0026261334922869583,
+      "grad_norm": 1.5348907709121704,
+      "learning_rate": 4.000500500500501e-05,
+      "loss": 2.089,
+      "step": 2009
+    },
+    {
+      "epoch": 0.0026274406767032286,
+      "grad_norm": 1.8672436475753784,
+      "learning_rate": 4e-05,
+      "loss": 1.9044,
+      "step": 2010
+    },
+    {
+      "epoch": 0.002628747861119499,
+      "grad_norm": 1.296150803565979,
+      "learning_rate": 3.9994994994994996e-05,
+      "loss": 0.7564,
+      "step": 2011
+    },
+    {
+      "epoch": 0.0026300550455357692,
+      "grad_norm": 0.9761872887611389,
+      "learning_rate": 3.998998998998999e-05,
+      "loss": 1.1528,
+      "step": 2012
+    },
+    {
+      "epoch": 0.0026313622299520396,
+      "grad_norm": 1.28115975856781,
+      "learning_rate": 3.998498498498498e-05,
+      "loss": 1.3665,
+      "step": 2013
+    },
+    {
+      "epoch": 0.00263266941436831,
+      "grad_norm": 0.9534834623336792,
+      "learning_rate": 3.997997997997998e-05,
+      "loss": 0.7384,
+      "step": 2014
+    },
+    {
+      "epoch": 0.0026339765987845798,
+      "grad_norm": 1.3732578754425049,
+      "learning_rate": 3.9974974974974975e-05,
+      "loss": 1.2486,
+      "step": 2015
+    },
+    {
+      "epoch": 0.00263528378320085,
+      "grad_norm": 0.9614124894142151,
+      "learning_rate": 3.9969969969969975e-05,
+      "loss": 0.6426,
+      "step": 2016
+    },
+    {
+      "epoch": 0.0026365909676171204,
+      "grad_norm": 1.8391082286834717,
+      "learning_rate": 3.996496496496497e-05,
+      "loss": 1.4386,
+      "step": 2017
+    },
+    {
+      "epoch": 0.0026378981520333907,
+      "grad_norm": 1.3796110153198242,
+      "learning_rate": 3.995995995995996e-05,
+      "loss": 1.4904,
+      "step": 2018
+    },
+    {
+      "epoch": 0.002639205336449661,
+      "grad_norm": 1.57155442237854,
+      "learning_rate": 3.995495495495496e-05,
+      "loss": 1.977,
+      "step": 2019
+    },
+    {
+      "epoch": 0.0026405125208659314,
+      "grad_norm": 1.2409968376159668,
+      "learning_rate": 3.9949949949949947e-05,
+      "loss": 1.4712,
+      "step": 2020
+    },
+    {
+      "epoch": 0.0026418197052822017,
+      "grad_norm": 1.2684842348098755,
+      "learning_rate": 3.9944944944944946e-05,
+      "loss": 1.6219,
+      "step": 2021
+    },
+    {
+      "epoch": 0.0026431268896984716,
+      "grad_norm": 1.610879898071289,
+      "learning_rate": 3.993993993993994e-05,
+      "loss": 2.3733,
+      "step": 2022
+    },
+    {
+      "epoch": 0.002644434074114742,
+      "grad_norm": 1.48215651512146,
+      "learning_rate": 3.993493493493494e-05,
+      "loss": 1.1412,
+      "step": 2023
+    },
+    {
+      "epoch": 0.002645741258531012,
+      "grad_norm": 1.7541826963424683,
+      "learning_rate": 3.992992992992993e-05,
+      "loss": 1.427,
+      "step": 2024
+    },
+    {
+      "epoch": 0.0026470484429472825,
+      "grad_norm": 1.2269361019134521,
+      "learning_rate": 3.9924924924924925e-05,
+      "loss": 1.16,
+      "step": 2025
+    },
+    {
+      "epoch": 0.002648355627363553,
+      "grad_norm": 1.3882238864898682,
+      "learning_rate": 3.9919919919919925e-05,
+      "loss": 1.2286,
+      "step": 2026
+    },
+    {
+      "epoch": 0.002649662811779823,
+      "grad_norm": 1.5924344062805176,
+      "learning_rate": 3.991491491491492e-05,
+      "loss": 2.1051,
+      "step": 2027
+    },
+    {
+      "epoch": 0.0026509699961960935,
+      "grad_norm": 1.5570005178451538,
+      "learning_rate": 3.990990990990991e-05,
+      "loss": 1.8338,
+      "step": 2028
+    },
+    {
+      "epoch": 0.002652277180612364,
+      "grad_norm": 1.8023189306259155,
+      "learning_rate": 3.9904904904904904e-05,
+      "loss": 1.5709,
+      "step": 2029
+    },
+    {
+      "epoch": 0.0026535843650286337,
+      "grad_norm": 2.0389351844787598,
+      "learning_rate": 3.9899899899899903e-05,
+      "loss": 1.7629,
+      "step": 2030
+    },
+    {
+      "epoch": 0.002654891549444904,
+      "grad_norm": 1.1661126613616943,
+      "learning_rate": 3.9894894894894896e-05,
+      "loss": 1.2144,
+      "step": 2031
+    },
+    {
+      "epoch": 0.0026561987338611743,
+      "grad_norm": 1.9049928188323975,
+      "learning_rate": 3.988988988988989e-05,
+      "loss": 2.3964,
+      "step": 2032
+    },
+    {
+      "epoch": 0.0026575059182774446,
+      "grad_norm": 1.5809181928634644,
+      "learning_rate": 3.988488488488489e-05,
+      "loss": 2.2846,
+      "step": 2033
+    },
+    {
+      "epoch": 0.002658813102693715,
+      "grad_norm": 1.7905234098434448,
+      "learning_rate": 3.987987987987988e-05,
+      "loss": 2.039,
+      "step": 2034
+    },
+    {
+      "epoch": 0.0026601202871099853,
+      "grad_norm": 1.642026662826538,
+      "learning_rate": 3.987487487487488e-05,
+      "loss": 0.831,
+      "step": 2035
+    },
+    {
+      "epoch": 0.0026614274715262556,
+      "grad_norm": 1.367571234703064,
+      "learning_rate": 3.986986986986987e-05,
+      "loss": 1.8172,
+      "step": 2036
+    },
+    {
+      "epoch": 0.002662734655942526,
+      "grad_norm": 0.6757227182388306,
+      "learning_rate": 3.986486486486487e-05,
+      "loss": 0.2978,
+      "step": 2037
+    },
+    {
+      "epoch": 0.002664041840358796,
+      "grad_norm": 2.139070510864258,
+      "learning_rate": 3.985985985985986e-05,
+      "loss": 1.9697,
+      "step": 2038
+    },
+    {
+      "epoch": 0.002665349024775066,
+      "grad_norm": 1.1820554733276367,
+      "learning_rate": 3.9854854854854854e-05,
+      "loss": 1.0919,
+      "step": 2039
+    },
+    {
+      "epoch": 0.0026666562091913364,
+      "grad_norm": 1.531373143196106,
+      "learning_rate": 3.9849849849849853e-05,
+      "loss": 1.8173,
+      "step": 2040
+    },
+    {
+      "epoch": 0.0026679633936076068,
+      "grad_norm": 1.1624926328659058,
+      "learning_rate": 3.9844844844844846e-05,
+      "loss": 1.3679,
+      "step": 2041
+    },
+    {
+      "epoch": 0.002669270578023877,
+      "grad_norm": 1.6291719675064087,
+      "learning_rate": 3.9839839839839846e-05,
+      "loss": 1.4781,
+      "step": 2042
+    },
+    {
+      "epoch": 0.0026705777624401474,
+      "grad_norm": 1.6225045919418335,
+      "learning_rate": 3.983483483483484e-05,
+      "loss": 2.1463,
+      "step": 2043
+    },
+    {
+      "epoch": 0.0026718849468564177,
+      "grad_norm": 1.9937926530838013,
+      "learning_rate": 3.982982982982983e-05,
+      "loss": 1.7029,
+      "step": 2044
+    },
+    {
+      "epoch": 0.0026731921312726876,
+      "grad_norm": 1.173494815826416,
+      "learning_rate": 3.9824824824824825e-05,
+      "loss": 1.0099,
+      "step": 2045
+    },
+    {
+      "epoch": 0.002674499315688958,
+      "grad_norm": 1.7637724876403809,
+      "learning_rate": 3.981981981981982e-05,
+      "loss": 1.3889,
+      "step": 2046
+    },
+    {
+      "epoch": 0.0026758065001052283,
+      "grad_norm": 1.5185253620147705,
+      "learning_rate": 3.981481481481482e-05,
+      "loss": 1.2063,
+      "step": 2047
+    },
+    {
+      "epoch": 0.0026771136845214986,
+      "grad_norm": 1.339429497718811,
+      "learning_rate": 3.980980980980981e-05,
+      "loss": 1.1688,
+      "step": 2048
+    },
+    {
+      "epoch": 0.002678420868937769,
+      "grad_norm": 1.5028117895126343,
+      "learning_rate": 3.980480480480481e-05,
+      "loss": 2.4854,
+      "step": 2049
+    },
+    {
+      "epoch": 0.002679728053354039,
+      "grad_norm": 1.3770437240600586,
+      "learning_rate": 3.9799799799799804e-05,
+      "loss": 1.693,
+      "step": 2050
+    },
+    {
+      "epoch": 0.0026810352377703095,
+      "grad_norm": 1.437799096107483,
+      "learning_rate": 3.9794794794794797e-05,
+      "loss": 1.0761,
+      "step": 2051
+    },
+    {
+      "epoch": 0.00268234242218658,
+      "grad_norm": 1.3654391765594482,
+      "learning_rate": 3.9789789789789796e-05,
+      "loss": 1.6745,
+      "step": 2052
+    },
+    {
+      "epoch": 0.0026836496066028497,
+      "grad_norm": 1.440726637840271,
+      "learning_rate": 3.978478478478478e-05,
+      "loss": 1.4871,
+      "step": 2053
+    },
+    {
+      "epoch": 0.00268495679101912,
+      "grad_norm": 1.3313815593719482,
+      "learning_rate": 3.977977977977978e-05,
+      "loss": 1.5516,
+      "step": 2054
+    },
+    {
+      "epoch": 0.0026862639754353904,
+      "grad_norm": 1.403893232345581,
+      "learning_rate": 3.9774774774774775e-05,
+      "loss": 1.2817,
+      "step": 2055
+    },
+    {
+      "epoch": 0.0026875711598516607,
+      "grad_norm": 1.3653924465179443,
+      "learning_rate": 3.976976976976977e-05,
+      "loss": 1.9225,
+      "step": 2056
+    },
+    {
+      "epoch": 0.002688878344267931,
+      "grad_norm": 1.4287811517715454,
+      "learning_rate": 3.976476476476477e-05,
+      "loss": 1.3453,
+      "step": 2057
+    },
+    {
+      "epoch": 0.0026901855286842013,
+      "grad_norm": 1.4424251317977905,
+      "learning_rate": 3.975975975975976e-05,
+      "loss": 1.8064,
+      "step": 2058
+    },
+    {
+      "epoch": 0.0026914927131004717,
+      "grad_norm": 1.3656929731369019,
+      "learning_rate": 3.975475475475476e-05,
+      "loss": 1.253,
+      "step": 2059
+    },
+    {
+      "epoch": 0.002692799897516742,
+      "grad_norm": 1.2642220258712769,
+      "learning_rate": 3.9749749749749754e-05,
+      "loss": 1.3694,
+      "step": 2060
+    },
+    {
+      "epoch": 0.002694107081933012,
+      "grad_norm": 1.655188798904419,
+      "learning_rate": 3.9744744744744747e-05,
+      "loss": 1.8618,
+      "step": 2061
+    },
+    {
+      "epoch": 0.002695414266349282,
+      "grad_norm": 0.5160999894142151,
+      "learning_rate": 3.973973973973974e-05,
+      "loss": 0.185,
+      "step": 2062
+    },
+    {
+      "epoch": 0.0026967214507655525,
+      "grad_norm": 1.4837048053741455,
+      "learning_rate": 3.973473473473473e-05,
+      "loss": 2.0436,
+      "step": 2063
+    },
+    {
+      "epoch": 0.002698028635181823,
+      "grad_norm": 2.9401626586914062,
+      "learning_rate": 3.972972972972973e-05,
+      "loss": 1.5095,
+      "step": 2064
+    },
+    {
+      "epoch": 0.002699335819598093,
+      "grad_norm": 1.6877011060714722,
+      "learning_rate": 3.9724724724724725e-05,
+      "loss": 1.8706,
+      "step": 2065
+    },
+    {
+      "epoch": 0.0027006430040143635,
+      "grad_norm": 1.6647981405258179,
+      "learning_rate": 3.9719719719719725e-05,
+      "loss": 2.3054,
+      "step": 2066
+    },
+    {
+      "epoch": 0.0027019501884306338,
+      "grad_norm": 1.5611270666122437,
+      "learning_rate": 3.971471471471472e-05,
+      "loss": 1.3985,
+      "step": 2067
+    },
+    {
+      "epoch": 0.0027032573728469037,
+      "grad_norm": 1.2352690696716309,
+      "learning_rate": 3.970970970970971e-05,
+      "loss": 1.6765,
+      "step": 2068
+    },
+    {
+      "epoch": 0.002704564557263174,
+      "grad_norm": 1.3239823579788208,
+      "learning_rate": 3.970470470470471e-05,
+      "loss": 1.5506,
+      "step": 2069
+    },
+    {
+      "epoch": 0.0027058717416794443,
+      "grad_norm": 1.5649746656417847,
+      "learning_rate": 3.96996996996997e-05,
+      "loss": 1.6246,
+      "step": 2070
+    },
+    {
+      "epoch": 0.0027071789260957146,
+      "grad_norm": 1.377916932106018,
+      "learning_rate": 3.96946946946947e-05,
+      "loss": 1.3012,
+      "step": 2071
+    },
+    {
+      "epoch": 0.002708486110511985,
+      "grad_norm": 1.4477307796478271,
+      "learning_rate": 3.968968968968969e-05,
+      "loss": 1.2674,
+      "step": 2072
+    },
+    {
+      "epoch": 0.0027097932949282553,
+      "grad_norm": 0.9115507006645203,
+      "learning_rate": 3.968468468468469e-05,
+      "loss": 0.6036,
+      "step": 2073
+    },
+    {
+      "epoch": 0.0027111004793445256,
+      "grad_norm": 1.3488483428955078,
+      "learning_rate": 3.967967967967968e-05,
+      "loss": 1.8283,
+      "step": 2074
+    },
+    {
+      "epoch": 0.002712407663760796,
+      "grad_norm": 1.2957956790924072,
+      "learning_rate": 3.9674674674674675e-05,
+      "loss": 1.5714,
+      "step": 2075
+    },
+    {
+      "epoch": 0.002713714848177066,
+      "grad_norm": 1.2624030113220215,
+      "learning_rate": 3.9669669669669675e-05,
+      "loss": 1.1737,
+      "step": 2076
+    },
+    {
+      "epoch": 0.002715022032593336,
+      "grad_norm": 1.1368193626403809,
+      "learning_rate": 3.966466466466467e-05,
+      "loss": 1.038,
+      "step": 2077
+    },
+    {
+      "epoch": 0.0027163292170096064,
+      "grad_norm": 1.3290246725082397,
+      "learning_rate": 3.965965965965966e-05,
+      "loss": 0.955,
+      "step": 2078
+    },
+    {
+      "epoch": 0.0027176364014258768,
+      "grad_norm": 1.326411247253418,
+      "learning_rate": 3.9654654654654654e-05,
+      "loss": 1.3841,
+      "step": 2079
+    },
+    {
+      "epoch": 0.002718943585842147,
+      "grad_norm": 1.5473278760910034,
+      "learning_rate": 3.9649649649649654e-05,
+      "loss": 1.5988,
+      "step": 2080
+    },
+    {
+      "epoch": 0.0027202507702584174,
+      "grad_norm": 1.520667314529419,
+      "learning_rate": 3.964464464464465e-05,
+      "loss": 2.2677,
+      "step": 2081
+    },
+    {
+      "epoch": 0.0027215579546746877,
+      "grad_norm": 3.00425386428833,
+      "learning_rate": 3.963963963963964e-05,
+      "loss": 1.5589,
+      "step": 2082
+    },
+    {
+      "epoch": 0.002722865139090958,
+      "grad_norm": 1.8552840948104858,
+      "learning_rate": 3.963463463463464e-05,
+      "loss": 1.5665,
+      "step": 2083
+    },
+    {
+      "epoch": 0.002724172323507228,
+      "grad_norm": 1.1176494359970093,
+      "learning_rate": 3.962962962962963e-05,
+      "loss": 1.0438,
+      "step": 2084
+    },
+    {
+      "epoch": 0.0027254795079234982,
+      "grad_norm": 1.0063292980194092,
+      "learning_rate": 3.962462462462463e-05,
+      "loss": 0.4489,
+      "step": 2085
+    },
+    {
+      "epoch": 0.0027267866923397686,
+      "grad_norm": 1.409346580505371,
+      "learning_rate": 3.961961961961962e-05,
+      "loss": 1.8401,
+      "step": 2086
+    },
+    {
+      "epoch": 0.002728093876756039,
+      "grad_norm": 1.6015211343765259,
+      "learning_rate": 3.961461461461461e-05,
+      "loss": 2.4007,
+      "step": 2087
+    },
+    {
+      "epoch": 0.002729401061172309,
+      "grad_norm": 1.79840886592865,
+      "learning_rate": 3.960960960960961e-05,
+      "loss": 1.6485,
+      "step": 2088
+    },
+    {
+      "epoch": 0.0027307082455885795,
+      "grad_norm": 1.3854961395263672,
+      "learning_rate": 3.9604604604604604e-05,
+      "loss": 1.6016,
+      "step": 2089
+    },
+    {
+      "epoch": 0.00273201543000485,
+      "grad_norm": 1.5748035907745361,
+      "learning_rate": 3.9599599599599604e-05,
+      "loss": 1.4871,
+      "step": 2090
+    },
+    {
+      "epoch": 0.0027333226144211197,
+      "grad_norm": 1.7774291038513184,
+      "learning_rate": 3.95945945945946e-05,
+      "loss": 1.2524,
+      "step": 2091
+    },
+    {
+      "epoch": 0.00273462979883739,
+      "grad_norm": 1.506029725074768,
+      "learning_rate": 3.958958958958959e-05,
+      "loss": 1.7196,
+      "step": 2092
+    },
+    {
+      "epoch": 0.0027359369832536604,
+      "grad_norm": 1.5348867177963257,
+      "learning_rate": 3.958458458458459e-05,
+      "loss": 1.5803,
+      "step": 2093
+    },
+    {
+      "epoch": 0.0027372441676699307,
+      "grad_norm": 1.5212199687957764,
+      "learning_rate": 3.9579579579579576e-05,
+      "loss": 1.7199,
+      "step": 2094
+    },
+    {
+      "epoch": 0.002738551352086201,
+      "grad_norm": 1.3874843120574951,
+      "learning_rate": 3.9574574574574575e-05,
+      "loss": 2.0295,
+      "step": 2095
+    },
+    {
+      "epoch": 0.0027398585365024713,
+      "grad_norm": 1.4956467151641846,
+      "learning_rate": 3.956956956956957e-05,
+      "loss": 1.5223,
+      "step": 2096
+    },
+    {
+      "epoch": 0.0027411657209187416,
+      "grad_norm": 1.3175606727600098,
+      "learning_rate": 3.956456456456457e-05,
+      "loss": 1.3091,
+      "step": 2097
+    },
+    {
+      "epoch": 0.002742472905335012,
+      "grad_norm": 1.4790319204330444,
+      "learning_rate": 3.955955955955956e-05,
+      "loss": 1.6883,
+      "step": 2098
+    },
+    {
+      "epoch": 0.002743780089751282,
+      "grad_norm": 1.3629701137542725,
+      "learning_rate": 3.9554554554554554e-05,
+      "loss": 1.5378,
+      "step": 2099
+    },
+    {
+      "epoch": 0.002745087274167552,
+      "grad_norm": 1.339185357093811,
+      "learning_rate": 3.9549549549549554e-05,
+      "loss": 2.4322,
+      "step": 2100
+    },
+    {
+      "epoch": 0.0027463944585838225,
+      "grad_norm": 1.505027413368225,
+      "learning_rate": 3.954454454454455e-05,
+      "loss": 1.6292,
+      "step": 2101
+    },
+    {
+      "epoch": 0.002747701643000093,
+      "grad_norm": 1.6119734048843384,
+      "learning_rate": 3.953953953953955e-05,
+      "loss": 2.4844,
+      "step": 2102
+    },
+    {
+      "epoch": 0.002749008827416363,
+      "grad_norm": 1.2258923053741455,
+      "learning_rate": 3.953453453453453e-05,
+      "loss": 1.2969,
+      "step": 2103
+    },
+    {
+      "epoch": 0.0027503160118326334,
+      "grad_norm": 1.59468412399292,
+      "learning_rate": 3.952952952952953e-05,
+      "loss": 1.9291,
+      "step": 2104
+    },
+    {
+      "epoch": 0.0027516231962489038,
+      "grad_norm": 1.5653488636016846,
+      "learning_rate": 3.9524524524524526e-05,
+      "loss": 1.9199,
+      "step": 2105
+    },
+    {
+      "epoch": 0.0027529303806651736,
+      "grad_norm": 1.4996395111083984,
+      "learning_rate": 3.951951951951952e-05,
+      "loss": 2.5863,
+      "step": 2106
+    },
+    {
+      "epoch": 0.002754237565081444,
+      "grad_norm": 1.3964852094650269,
+      "learning_rate": 3.951451451451452e-05,
+      "loss": 1.0641,
+      "step": 2107
+    },
+    {
+      "epoch": 0.0027555447494977143,
+      "grad_norm": 1.1878902912139893,
+      "learning_rate": 3.950950950950951e-05,
+      "loss": 1.1114,
+      "step": 2108
+    },
+    {
+      "epoch": 0.0027568519339139846,
+      "grad_norm": 1.1396305561065674,
+      "learning_rate": 3.950450450450451e-05,
+      "loss": 1.07,
+      "step": 2109
+    },
+    {
+      "epoch": 0.002758159118330255,
+      "grad_norm": 1.4439995288848877,
+      "learning_rate": 3.9499499499499504e-05,
+      "loss": 1.4961,
+      "step": 2110
+    },
+    {
+      "epoch": 0.0027594663027465252,
+      "grad_norm": 1.9901820421218872,
+      "learning_rate": 3.94944944944945e-05,
+      "loss": 2.1474,
+      "step": 2111
+    },
+    {
+      "epoch": 0.0027607734871627956,
+      "grad_norm": 1.5734364986419678,
+      "learning_rate": 3.948948948948949e-05,
+      "loss": 1.6936,
+      "step": 2112
+    },
+    {
+      "epoch": 0.002762080671579066,
+      "grad_norm": 1.4398524761199951,
+      "learning_rate": 3.948448448448448e-05,
+      "loss": 1.7767,
+      "step": 2113
+    },
+    {
+      "epoch": 0.0027633878559953358,
+      "grad_norm": 1.5927658081054688,
+      "learning_rate": 3.947947947947948e-05,
+      "loss": 1.5136,
+      "step": 2114
+    },
+    {
+      "epoch": 0.002764695040411606,
+      "grad_norm": 0.9344616532325745,
+      "learning_rate": 3.9474474474474476e-05,
+      "loss": 0.7111,
+      "step": 2115
+    },
+    {
+      "epoch": 0.0027660022248278764,
+      "grad_norm": 1.7013565301895142,
+      "learning_rate": 3.9469469469469475e-05,
+      "loss": 2.4446,
+      "step": 2116
+    },
+    {
+      "epoch": 0.0027673094092441467,
+      "grad_norm": 1.1245629787445068,
+      "learning_rate": 3.946446446446447e-05,
+      "loss": 0.7081,
+      "step": 2117
+    },
+    {
+      "epoch": 0.002768616593660417,
+      "grad_norm": 1.5265649557113647,
+      "learning_rate": 3.945945945945946e-05,
+      "loss": 1.5882,
+      "step": 2118
+    },
+    {
+      "epoch": 0.0027699237780766874,
+      "grad_norm": 2.080834150314331,
+      "learning_rate": 3.9454454454454454e-05,
+      "loss": 1.4114,
+      "step": 2119
+    },
+    {
+      "epoch": 0.0027712309624929577,
+      "grad_norm": 1.7935351133346558,
+      "learning_rate": 3.944944944944945e-05,
+      "loss": 1.7366,
+      "step": 2120
+    },
+    {
+      "epoch": 0.002772538146909228,
+      "grad_norm": 2.073263168334961,
+      "learning_rate": 3.944444444444445e-05,
+      "loss": 1.7597,
+      "step": 2121
+    },
+    {
+      "epoch": 0.002773845331325498,
+      "grad_norm": 1.472176194190979,
+      "learning_rate": 3.943943943943944e-05,
+      "loss": 2.3025,
+      "step": 2122
+    },
+    {
+      "epoch": 0.002775152515741768,
+      "grad_norm": 1.6525713205337524,
+      "learning_rate": 3.943443443443443e-05,
+      "loss": 2.1587,
+      "step": 2123
+    },
+    {
+      "epoch": 0.0027764597001580385,
+      "grad_norm": 1.493241310119629,
+      "learning_rate": 3.942942942942943e-05,
+      "loss": 1.6936,
+      "step": 2124
+    },
+    {
+      "epoch": 0.002777766884574309,
+      "grad_norm": 1.3473711013793945,
+      "learning_rate": 3.9424424424424426e-05,
+      "loss": 1.4847,
+      "step": 2125
+    },
+    {
+      "epoch": 0.002779074068990579,
+      "grad_norm": 1.1274248361587524,
+      "learning_rate": 3.9419419419419425e-05,
+      "loss": 1.0142,
+      "step": 2126
+    },
+    {
+      "epoch": 0.0027803812534068495,
+      "grad_norm": 1.5922311544418335,
+      "learning_rate": 3.941441441441442e-05,
+      "loss": 1.2503,
+      "step": 2127
+    },
+    {
+      "epoch": 0.00278168843782312,
+      "grad_norm": 0.7048544883728027,
+      "learning_rate": 3.940940940940941e-05,
+      "loss": 0.3283,
+      "step": 2128
+    },
+    {
+      "epoch": 0.0027829956222393897,
+      "grad_norm": 1.491128921508789,
+      "learning_rate": 3.9404404404404404e-05,
+      "loss": 2.0905,
+      "step": 2129
+    },
+    {
+      "epoch": 0.00278430280665566,
+      "grad_norm": 1.5215622186660767,
+      "learning_rate": 3.93993993993994e-05,
+      "loss": 1.3431,
+      "step": 2130
+    },
+    {
+      "epoch": 0.0027856099910719303,
+      "grad_norm": 1.4988982677459717,
+      "learning_rate": 3.93943943943944e-05,
+      "loss": 2.4214,
+      "step": 2131
+    },
+    {
+      "epoch": 0.0027869171754882007,
+      "grad_norm": 1.0583879947662354,
+      "learning_rate": 3.938938938938939e-05,
+      "loss": 1.2983,
+      "step": 2132
+    },
+    {
+      "epoch": 0.002788224359904471,
+      "grad_norm": 2.0345451831817627,
+      "learning_rate": 3.938438438438439e-05,
+      "loss": 2.0542,
+      "step": 2133
+    },
+    {
+      "epoch": 0.0027895315443207413,
+      "grad_norm": 1.2454980611801147,
+      "learning_rate": 3.937937937937938e-05,
+      "loss": 1.5754,
+      "step": 2134
+    },
+    {
+      "epoch": 0.0027908387287370116,
+      "grad_norm": 1.5364166498184204,
+      "learning_rate": 3.9374374374374376e-05,
+      "loss": 1.4142,
+      "step": 2135
+    },
+    {
+      "epoch": 0.002792145913153282,
+      "grad_norm": 1.4162671566009521,
+      "learning_rate": 3.936936936936937e-05,
+      "loss": 1.395,
+      "step": 2136
+    },
+    {
+      "epoch": 0.002793453097569552,
+      "grad_norm": 1.391393780708313,
+      "learning_rate": 3.936436436436436e-05,
+      "loss": 1.6288,
+      "step": 2137
+    },
+    {
+      "epoch": 0.002794760281985822,
+      "grad_norm": 1.166745662689209,
+      "learning_rate": 3.935935935935936e-05,
+      "loss": 1.5029,
+      "step": 2138
+    },
+    {
+      "epoch": 0.0027960674664020925,
+      "grad_norm": 1.2580697536468506,
+      "learning_rate": 3.9354354354354354e-05,
+      "loss": 1.1606,
+      "step": 2139
+    },
+    {
+      "epoch": 0.002797374650818363,
+      "grad_norm": 1.1910089254379272,
+      "learning_rate": 3.9349349349349354e-05,
+      "loss": 1.0285,
+      "step": 2140
+    },
+    {
+      "epoch": 0.002798681835234633,
+      "grad_norm": 1.6236528158187866,
+      "learning_rate": 3.934434434434435e-05,
+      "loss": 1.4381,
+      "step": 2141
+    },
+    {
+      "epoch": 0.0027999890196509034,
+      "grad_norm": 1.2432808876037598,
+      "learning_rate": 3.933933933933934e-05,
+      "loss": 1.0405,
+      "step": 2142
+    },
+    {
+      "epoch": 0.0028012962040671737,
+      "grad_norm": 1.3887481689453125,
+      "learning_rate": 3.933433433433434e-05,
+      "loss": 2.0166,
+      "step": 2143
+    },
+    {
+      "epoch": 0.002802603388483444,
+      "grad_norm": 1.300026535987854,
+      "learning_rate": 3.9329329329329326e-05,
+      "loss": 1.0104,
+      "step": 2144
+    },
+    {
+      "epoch": 0.002803910572899714,
+      "grad_norm": 1.3419066667556763,
+      "learning_rate": 3.9324324324324326e-05,
+      "loss": 1.3359,
+      "step": 2145
+    },
+    {
+      "epoch": 0.0028052177573159843,
+      "grad_norm": 1.5531896352767944,
+      "learning_rate": 3.931931931931932e-05,
+      "loss": 1.7291,
+      "step": 2146
+    },
+    {
+      "epoch": 0.0028065249417322546,
+      "grad_norm": 1.5282682180404663,
+      "learning_rate": 3.931431431431432e-05,
+      "loss": 1.4634,
+      "step": 2147
+    },
+    {
+      "epoch": 0.002807832126148525,
+      "grad_norm": 1.3880059719085693,
+      "learning_rate": 3.930930930930931e-05,
+      "loss": 1.9265,
+      "step": 2148
+    },
+    {
+      "epoch": 0.0028091393105647952,
+      "grad_norm": 1.8446824550628662,
+      "learning_rate": 3.9304304304304304e-05,
+      "loss": 1.6512,
+      "step": 2149
+    },
+    {
+      "epoch": 0.0028104464949810655,
+      "grad_norm": 0.3314052224159241,
+      "learning_rate": 3.9299299299299304e-05,
+      "loss": 0.1176,
+      "step": 2150
+    },
+    {
+      "epoch": 0.002811753679397336,
+      "grad_norm": 1.5311682224273682,
+      "learning_rate": 3.92942942942943e-05,
+      "loss": 1.5512,
+      "step": 2151
+    },
+    {
+      "epoch": 0.0028130608638136058,
+      "grad_norm": 1.2343051433563232,
+      "learning_rate": 3.92892892892893e-05,
+      "loss": 1.1211,
+      "step": 2152
+    },
+    {
+      "epoch": 0.002814368048229876,
+      "grad_norm": 1.4304108619689941,
+      "learning_rate": 3.928428428428428e-05,
+      "loss": 1.6917,
+      "step": 2153
+    },
+    {
+      "epoch": 0.0028156752326461464,
+      "grad_norm": 1.0727342367172241,
+      "learning_rate": 3.927927927927928e-05,
+      "loss": 0.8996,
+      "step": 2154
+    },
+    {
+      "epoch": 0.0028169824170624167,
+      "grad_norm": 1.126792073249817,
+      "learning_rate": 3.9274274274274276e-05,
+      "loss": 1.2973,
+      "step": 2155
+    },
+    {
+      "epoch": 0.002818289601478687,
+      "grad_norm": 1.6876169443130493,
+      "learning_rate": 3.926926926926927e-05,
+      "loss": 1.2195,
+      "step": 2156
+    },
+    {
+      "epoch": 0.0028195967858949574,
+      "grad_norm": 1.1861119270324707,
+      "learning_rate": 3.926426426426427e-05,
+      "loss": 0.9566,
+      "step": 2157
+    },
+    {
+      "epoch": 0.0028209039703112277,
+      "grad_norm": 1.1142774820327759,
+      "learning_rate": 3.925925925925926e-05,
+      "loss": 0.8374,
+      "step": 2158
+    },
+    {
+      "epoch": 0.002822211154727498,
+      "grad_norm": 1.7273575067520142,
+      "learning_rate": 3.925425425425426e-05,
+      "loss": 2.1927,
+      "step": 2159
+    },
+    {
+      "epoch": 0.002823518339143768,
+      "grad_norm": 2.968243360519409,
+      "learning_rate": 3.9249249249249254e-05,
+      "loss": 1.3633,
+      "step": 2160
+    },
+    {
+      "epoch": 0.002824825523560038,
+      "grad_norm": 1.1221617460250854,
+      "learning_rate": 3.924424424424424e-05,
+      "loss": 1.1931,
+      "step": 2161
+    },
+    {
+      "epoch": 0.0028261327079763085,
+      "grad_norm": 1.2373301982879639,
+      "learning_rate": 3.923923923923924e-05,
+      "loss": 1.321,
+      "step": 2162
+    },
+    {
+      "epoch": 0.002827439892392579,
+      "grad_norm": 1.3540993928909302,
+      "learning_rate": 3.923423423423423e-05,
+      "loss": 1.5032,
+      "step": 2163
+    },
+    {
+      "epoch": 0.002828747076808849,
+      "grad_norm": 2.1110341548919678,
+      "learning_rate": 3.922922922922923e-05,
+      "loss": 1.6869,
+      "step": 2164
+    },
+    {
+      "epoch": 0.0028300542612251195,
+      "grad_norm": 3.4193215370178223,
+      "learning_rate": 3.9224224224224226e-05,
+      "loss": 1.9028,
+      "step": 2165
+    },
+    {
+      "epoch": 0.00283136144564139,
+      "grad_norm": 1.1574500799179077,
+      "learning_rate": 3.921921921921922e-05,
+      "loss": 1.8632,
+      "step": 2166
+    },
+    {
+      "epoch": 0.00283266863005766,
+      "grad_norm": 0.9121442437171936,
+      "learning_rate": 3.921421421421422e-05,
+      "loss": 0.3304,
+      "step": 2167
+    },
+    {
+      "epoch": 0.00283397581447393,
+      "grad_norm": 1.4046697616577148,
+      "learning_rate": 3.920920920920921e-05,
+      "loss": 0.9095,
+      "step": 2168
+    },
+    {
+      "epoch": 0.0028352829988902003,
+      "grad_norm": 1.408126950263977,
+      "learning_rate": 3.9204204204204205e-05,
+      "loss": 2.2093,
+      "step": 2169
+    },
+    {
+      "epoch": 0.0028365901833064706,
+      "grad_norm": 1.8400304317474365,
+      "learning_rate": 3.91991991991992e-05,
+      "loss": 2.9623,
+      "step": 2170
+    },
+    {
+      "epoch": 0.002837897367722741,
+      "grad_norm": 1.1644660234451294,
+      "learning_rate": 3.91941941941942e-05,
+      "loss": 0.8549,
+      "step": 2171
+    },
+    {
+      "epoch": 0.0028392045521390113,
+      "grad_norm": 1.324384331703186,
+      "learning_rate": 3.918918918918919e-05,
+      "loss": 1.4598,
+      "step": 2172
+    },
+    {
+      "epoch": 0.0028405117365552816,
+      "grad_norm": 1.6433312892913818,
+      "learning_rate": 3.918418418418418e-05,
+      "loss": 1.793,
+      "step": 2173
+    },
+    {
+      "epoch": 0.002841818920971552,
+      "grad_norm": 1.2081270217895508,
+      "learning_rate": 3.917917917917918e-05,
+      "loss": 1.4221,
+      "step": 2174
+    },
+    {
+      "epoch": 0.002843126105387822,
+      "grad_norm": 1.2499949932098389,
+      "learning_rate": 3.9174174174174176e-05,
+      "loss": 1.4104,
+      "step": 2175
+    },
+    {
+      "epoch": 0.002844433289804092,
+      "grad_norm": 1.7963576316833496,
+      "learning_rate": 3.9169169169169176e-05,
+      "loss": 1.4704,
+      "step": 2176
+    },
+    {
+      "epoch": 0.0028457404742203624,
+      "grad_norm": 1.283759593963623,
+      "learning_rate": 3.916416416416417e-05,
+      "loss": 1.2633,
+      "step": 2177
+    },
+    {
+      "epoch": 0.0028470476586366328,
+      "grad_norm": 1.4032162427902222,
+      "learning_rate": 3.915915915915916e-05,
+      "loss": 1.9242,
+      "step": 2178
+    },
+    {
+      "epoch": 0.002848354843052903,
+      "grad_norm": 1.305837631225586,
+      "learning_rate": 3.9154154154154155e-05,
+      "loss": 1.0144,
+      "step": 2179
+    },
+    {
+      "epoch": 0.0028496620274691734,
+      "grad_norm": 1.4017144441604614,
+      "learning_rate": 3.914914914914915e-05,
+      "loss": 1.2406,
+      "step": 2180
+    },
+    {
+      "epoch": 0.0028509692118854437,
+      "grad_norm": 1.0714243650436401,
+      "learning_rate": 3.914414414414415e-05,
+      "loss": 1.2344,
+      "step": 2181
+    },
+    {
+      "epoch": 0.002852276396301714,
+      "grad_norm": 1.263034462928772,
+      "learning_rate": 3.913913913913914e-05,
+      "loss": 1.1296,
+      "step": 2182
+    },
+    {
+      "epoch": 0.002853583580717984,
+      "grad_norm": 0.8641487956047058,
+      "learning_rate": 3.913413413413414e-05,
+      "loss": 0.7715,
+      "step": 2183
+    },
+    {
+      "epoch": 0.0028548907651342543,
+      "grad_norm": 1.2568964958190918,
+      "learning_rate": 3.912912912912913e-05,
+      "loss": 1.0284,
+      "step": 2184
+    },
+    {
+      "epoch": 0.0028561979495505246,
+      "grad_norm": 1.445867657661438,
+      "learning_rate": 3.9124124124124126e-05,
+      "loss": 1.2917,
+      "step": 2185
+    },
+    {
+      "epoch": 0.002857505133966795,
+      "grad_norm": 1.3942453861236572,
+      "learning_rate": 3.911911911911912e-05,
+      "loss": 1.7379,
+      "step": 2186
+    },
+    {
+      "epoch": 0.002858812318383065,
+      "grad_norm": 1.395135760307312,
+      "learning_rate": 3.911411411411411e-05,
+      "loss": 2.0884,
+      "step": 2187
+    },
+    {
+      "epoch": 0.0028601195027993355,
+      "grad_norm": 1.5817264318466187,
+      "learning_rate": 3.910910910910911e-05,
+      "loss": 0.9085,
+      "step": 2188
+    },
+    {
+      "epoch": 0.002861426687215606,
+      "grad_norm": 1.2675480842590332,
+      "learning_rate": 3.9104104104104105e-05,
+      "loss": 1.8649,
+      "step": 2189
+    },
+    {
+      "epoch": 0.002862733871631876,
+      "grad_norm": 1.3925424814224243,
+      "learning_rate": 3.9099099099099105e-05,
+      "loss": 1.4546,
+      "step": 2190
+    },
+    {
+      "epoch": 0.002864041056048146,
+      "grad_norm": 1.3031902313232422,
+      "learning_rate": 3.90940940940941e-05,
+      "loss": 1.7122,
+      "step": 2191
+    },
+    {
+      "epoch": 0.0028653482404644164,
+      "grad_norm": 1.2801841497421265,
+      "learning_rate": 3.908908908908909e-05,
+      "loss": 1.29,
+      "step": 2192
+    },
+    {
+      "epoch": 0.0028666554248806867,
+      "grad_norm": 1.0257128477096558,
+      "learning_rate": 3.908408408408409e-05,
+      "loss": 0.9683,
+      "step": 2193
+    },
+    {
+      "epoch": 0.002867962609296957,
+      "grad_norm": 1.3909313678741455,
+      "learning_rate": 3.9079079079079076e-05,
+      "loss": 2.0009,
+      "step": 2194
+    },
+    {
+      "epoch": 0.0028692697937132273,
+      "grad_norm": 1.3106316328048706,
+      "learning_rate": 3.9074074074074076e-05,
+      "loss": 1.4901,
+      "step": 2195
+    },
+    {
+      "epoch": 0.0028705769781294977,
+      "grad_norm": 1.658611536026001,
+      "learning_rate": 3.906906906906907e-05,
+      "loss": 1.771,
+      "step": 2196
+    },
+    {
+      "epoch": 0.002871884162545768,
+      "grad_norm": 1.3780211210250854,
+      "learning_rate": 3.906406406406406e-05,
+      "loss": 1.3118,
+      "step": 2197
+    },
+    {
+      "epoch": 0.002873191346962038,
+      "grad_norm": 1.1381900310516357,
+      "learning_rate": 3.905905905905906e-05,
+      "loss": 1.877,
+      "step": 2198
+    },
+    {
+      "epoch": 0.002874498531378308,
+      "grad_norm": 1.5120329856872559,
+      "learning_rate": 3.9054054054054055e-05,
+      "loss": 2.2779,
+      "step": 2199
+    },
+    {
+      "epoch": 0.0028758057157945785,
+      "grad_norm": 1.9007819890975952,
+      "learning_rate": 3.9049049049049055e-05,
+      "loss": 1.8522,
+      "step": 2200
+    },
+    {
+      "epoch": 0.002877112900210849,
+      "grad_norm": 1.4512531757354736,
+      "learning_rate": 3.904404404404405e-05,
+      "loss": 1.331,
+      "step": 2201
+    },
+    {
+      "epoch": 0.002878420084627119,
+      "grad_norm": 1.5178358554840088,
+      "learning_rate": 3.903903903903904e-05,
+      "loss": 1.7808,
+      "step": 2202
+    },
+    {
+      "epoch": 0.0028797272690433895,
+      "grad_norm": 1.539099097251892,
+      "learning_rate": 3.9034034034034034e-05,
+      "loss": 1.3691,
+      "step": 2203
+    },
+    {
+      "epoch": 0.0028810344534596598,
+      "grad_norm": 1.3301935195922852,
+      "learning_rate": 3.9029029029029026e-05,
+      "loss": 1.1288,
+      "step": 2204
+    },
+    {
+      "epoch": 0.00288234163787593,
+      "grad_norm": 1.2755948305130005,
+      "learning_rate": 3.9024024024024026e-05,
+      "loss": 0.979,
+      "step": 2205
+    },
+    {
+      "epoch": 0.0028836488222922,
+      "grad_norm": 1.5008995532989502,
+      "learning_rate": 3.901901901901902e-05,
+      "loss": 1.8601,
+      "step": 2206
+    },
+    {
+      "epoch": 0.0028849560067084703,
+      "grad_norm": 1.4420750141143799,
+      "learning_rate": 3.901401401401402e-05,
+      "loss": 1.7191,
+      "step": 2207
+    },
+    {
+      "epoch": 0.0028862631911247406,
+      "grad_norm": 1.577452540397644,
+      "learning_rate": 3.900900900900901e-05,
+      "loss": 1.0463,
+      "step": 2208
+    },
+    {
+      "epoch": 0.002887570375541011,
+      "grad_norm": 1.262462854385376,
+      "learning_rate": 3.9004004004004005e-05,
+      "loss": 1.325,
+      "step": 2209
+    },
+    {
+      "epoch": 0.0028888775599572813,
+      "grad_norm": 1.1211475133895874,
+      "learning_rate": 3.8998998998999005e-05,
+      "loss": 0.6703,
+      "step": 2210
+    },
+    {
+      "epoch": 0.0028901847443735516,
+      "grad_norm": 2.6080334186553955,
+      "learning_rate": 3.899399399399399e-05,
+      "loss": 2.5353,
+      "step": 2211
+    },
+    {
+      "epoch": 0.002891491928789822,
+      "grad_norm": 1.357974648475647,
+      "learning_rate": 3.898898898898899e-05,
+      "loss": 2.0341,
+      "step": 2212
+    },
+    {
+      "epoch": 0.002892799113206092,
+      "grad_norm": 1.2836050987243652,
+      "learning_rate": 3.8983983983983984e-05,
+      "loss": 1.8832,
+      "step": 2213
+    },
+    {
+      "epoch": 0.002894106297622362,
+      "grad_norm": 1.2315454483032227,
+      "learning_rate": 3.897897897897898e-05,
+      "loss": 1.345,
+      "step": 2214
+    },
+    {
+      "epoch": 0.0028954134820386324,
+      "grad_norm": 1.9007593393325806,
+      "learning_rate": 3.8973973973973976e-05,
+      "loss": 1.9913,
+      "step": 2215
+    },
+    {
+      "epoch": 0.0028967206664549027,
+      "grad_norm": 1.6237214803695679,
+      "learning_rate": 3.896896896896897e-05,
+      "loss": 1.8533,
+      "step": 2216
+    },
+    {
+      "epoch": 0.002898027850871173,
+      "grad_norm": 1.1567765474319458,
+      "learning_rate": 3.896396396396397e-05,
+      "loss": 1.3661,
+      "step": 2217
+    },
+    {
+      "epoch": 0.0028993350352874434,
+      "grad_norm": 1.8322334289550781,
+      "learning_rate": 3.895895895895896e-05,
+      "loss": 1.9348,
+      "step": 2218
+    },
+    {
+      "epoch": 0.0029006422197037137,
+      "grad_norm": 1.3179900646209717,
+      "learning_rate": 3.8953953953953955e-05,
+      "loss": 1.3918,
+      "step": 2219
+    },
+    {
+      "epoch": 0.002901949404119984,
+      "grad_norm": 1.5900468826293945,
+      "learning_rate": 3.894894894894895e-05,
+      "loss": 1.4963,
+      "step": 2220
+    },
+    {
+      "epoch": 0.002903256588536254,
+      "grad_norm": 1.6042488813400269,
+      "learning_rate": 3.894394394394395e-05,
+      "loss": 2.605,
+      "step": 2221
+    },
+    {
+      "epoch": 0.0029045637729525242,
+      "grad_norm": 1.6375967264175415,
+      "learning_rate": 3.893893893893894e-05,
+      "loss": 2.1692,
+      "step": 2222
+    },
+    {
+      "epoch": 0.0029058709573687946,
+      "grad_norm": 1.8711655139923096,
+      "learning_rate": 3.8933933933933934e-05,
+      "loss": 2.0317,
+      "step": 2223
+    },
+    {
+      "epoch": 0.002907178141785065,
+      "grad_norm": 1.4329065084457397,
+      "learning_rate": 3.8928928928928933e-05,
+      "loss": 1.6011,
+      "step": 2224
+    },
+    {
+      "epoch": 0.002908485326201335,
+      "grad_norm": 1.6335432529449463,
+      "learning_rate": 3.8923923923923926e-05,
+      "loss": 1.6219,
+      "step": 2225
+    },
+    {
+      "epoch": 0.0029097925106176055,
+      "grad_norm": 1.0949170589447021,
+      "learning_rate": 3.8918918918918926e-05,
+      "loss": 0.9806,
+      "step": 2226
+    },
+    {
+      "epoch": 0.002911099695033876,
+      "grad_norm": 1.288183331489563,
+      "learning_rate": 3.891391391391391e-05,
+      "loss": 2.1031,
+      "step": 2227
+    },
+    {
+      "epoch": 0.002912406879450146,
+      "grad_norm": 1.5958274602890015,
+      "learning_rate": 3.890890890890891e-05,
+      "loss": 1.8244,
+      "step": 2228
+    },
+    {
+      "epoch": 0.002913714063866416,
+      "grad_norm": 1.1349316835403442,
+      "learning_rate": 3.8903903903903905e-05,
+      "loss": 1.3589,
+      "step": 2229
+    },
+    {
+      "epoch": 0.0029150212482826864,
+      "grad_norm": 1.8477003574371338,
+      "learning_rate": 3.88988988988989e-05,
+      "loss": 1.3192,
+      "step": 2230
+    },
+    {
+      "epoch": 0.0029163284326989567,
+      "grad_norm": 1.476585865020752,
+      "learning_rate": 3.88938938938939e-05,
+      "loss": 1.4037,
+      "step": 2231
+    },
+    {
+      "epoch": 0.002917635617115227,
+      "grad_norm": 1.2419575452804565,
+      "learning_rate": 3.888888888888889e-05,
+      "loss": 1.6337,
+      "step": 2232
+    },
+    {
+      "epoch": 0.0029189428015314973,
+      "grad_norm": 1.4536539316177368,
+      "learning_rate": 3.8883883883883884e-05,
+      "loss": 0.9798,
+      "step": 2233
+    },
+    {
+      "epoch": 0.0029202499859477676,
+      "grad_norm": 1.684038758277893,
+      "learning_rate": 3.8878878878878883e-05,
+      "loss": 2.0592,
+      "step": 2234
+    },
+    {
+      "epoch": 0.002921557170364038,
+      "grad_norm": 1.1105613708496094,
+      "learning_rate": 3.8873873873873876e-05,
+      "loss": 0.7923,
+      "step": 2235
+    },
+    {
+      "epoch": 0.002922864354780308,
+      "grad_norm": 1.5902795791625977,
+      "learning_rate": 3.886886886886887e-05,
+      "loss": 2.1564,
+      "step": 2236
+    },
+    {
+      "epoch": 0.002924171539196578,
+      "grad_norm": 1.145974040031433,
+      "learning_rate": 3.886386386386386e-05,
+      "loss": 1.0234,
+      "step": 2237
+    },
+    {
+      "epoch": 0.0029254787236128485,
+      "grad_norm": 1.591015100479126,
+      "learning_rate": 3.885885885885886e-05,
+      "loss": 1.6718,
+      "step": 2238
+    },
+    {
+      "epoch": 0.002926785908029119,
+      "grad_norm": 1.8055939674377441,
+      "learning_rate": 3.8853853853853855e-05,
+      "loss": 1.3664,
+      "step": 2239
+    },
+    {
+      "epoch": 0.002928093092445389,
+      "grad_norm": 1.3488624095916748,
+      "learning_rate": 3.884884884884885e-05,
+      "loss": 0.9705,
+      "step": 2240
+    },
+    {
+      "epoch": 0.0029294002768616594,
+      "grad_norm": 2.829200267791748,
+      "learning_rate": 3.884384384384385e-05,
+      "loss": 1.4442,
+      "step": 2241
+    },
+    {
+      "epoch": 0.0029307074612779298,
+      "grad_norm": 1.1419140100479126,
+      "learning_rate": 3.883883883883884e-05,
+      "loss": 1.1223,
+      "step": 2242
+    },
+    {
+      "epoch": 0.0029320146456942,
+      "grad_norm": 2.2440803050994873,
+      "learning_rate": 3.883383383383384e-05,
+      "loss": 2.1641,
+      "step": 2243
+    },
+    {
+      "epoch": 0.00293332183011047,
+      "grad_norm": 1.8819011449813843,
+      "learning_rate": 3.882882882882883e-05,
+      "loss": 1.4028,
+      "step": 2244
+    },
+    {
+      "epoch": 0.0029346290145267403,
+      "grad_norm": 1.4086968898773193,
+      "learning_rate": 3.8823823823823827e-05,
+      "loss": 1.6778,
+      "step": 2245
+    },
+    {
+      "epoch": 0.0029359361989430106,
+      "grad_norm": 1.5371630191802979,
+      "learning_rate": 3.881881881881882e-05,
+      "loss": 1.3815,
+      "step": 2246
+    },
+    {
+      "epoch": 0.002937243383359281,
+      "grad_norm": 1.3543468713760376,
+      "learning_rate": 3.881381381381381e-05,
+      "loss": 1.8969,
+      "step": 2247
+    },
+    {
+      "epoch": 0.0029385505677755512,
+      "grad_norm": 1.968708872795105,
+      "learning_rate": 3.880880880880881e-05,
+      "loss": 1.9086,
+      "step": 2248
+    },
+    {
+      "epoch": 0.0029398577521918216,
+      "grad_norm": 1.2748903036117554,
+      "learning_rate": 3.8803803803803805e-05,
+      "loss": 1.4398,
+      "step": 2249
+    },
+    {
+      "epoch": 0.002941164936608092,
+      "grad_norm": 1.6713937520980835,
+      "learning_rate": 3.8798798798798805e-05,
+      "loss": 2.1194,
+      "step": 2250
+    },
+    {
+      "epoch": 0.002942472121024362,
+      "grad_norm": 1.390107274055481,
+      "learning_rate": 3.87937937937938e-05,
+      "loss": 1.9787,
+      "step": 2251
+    },
+    {
+      "epoch": 0.002943779305440632,
+      "grad_norm": 1.2540277242660522,
+      "learning_rate": 3.878878878878879e-05,
+      "loss": 1.5649,
+      "step": 2252
+    },
+    {
+      "epoch": 0.0029450864898569024,
+      "grad_norm": 1.0322459936141968,
+      "learning_rate": 3.8783783783783784e-05,
+      "loss": 1.3197,
+      "step": 2253
+    },
+    {
+      "epoch": 0.0029463936742731727,
+      "grad_norm": 1.2263007164001465,
+      "learning_rate": 3.877877877877878e-05,
+      "loss": 1.0607,
+      "step": 2254
+    },
+    {
+      "epoch": 0.002947700858689443,
+      "grad_norm": 1.3015462160110474,
+      "learning_rate": 3.8773773773773777e-05,
+      "loss": 0.6064,
+      "step": 2255
+    },
+    {
+      "epoch": 0.0029490080431057134,
+      "grad_norm": 1.4123313426971436,
+      "learning_rate": 3.876876876876877e-05,
+      "loss": 2.0811,
+      "step": 2256
+    },
+    {
+      "epoch": 0.0029503152275219837,
+      "grad_norm": 1.5940463542938232,
+      "learning_rate": 3.876376376376377e-05,
+      "loss": 1.8167,
+      "step": 2257
+    },
+    {
+      "epoch": 0.002951622411938254,
+      "grad_norm": 1.7560492753982544,
+      "learning_rate": 3.875875875875876e-05,
+      "loss": 1.5764,
+      "step": 2258
+    },
+    {
+      "epoch": 0.002952929596354524,
+      "grad_norm": 1.4442381858825684,
+      "learning_rate": 3.8753753753753755e-05,
+      "loss": 1.7969,
+      "step": 2259
+    },
+    {
+      "epoch": 0.002954236780770794,
+      "grad_norm": 1.4383872747421265,
+      "learning_rate": 3.8748748748748755e-05,
+      "loss": 2.0515,
+      "step": 2260
+    },
+    {
+      "epoch": 0.0029555439651870645,
+      "grad_norm": 1.1254631280899048,
+      "learning_rate": 3.874374374374374e-05,
+      "loss": 1.3704,
+      "step": 2261
+    },
+    {
+      "epoch": 0.002956851149603335,
+      "grad_norm": 1.3781318664550781,
+      "learning_rate": 3.873873873873874e-05,
+      "loss": 1.8193,
+      "step": 2262
+    },
+    {
+      "epoch": 0.002958158334019605,
+      "grad_norm": 1.485600471496582,
+      "learning_rate": 3.8733733733733734e-05,
+      "loss": 2.2992,
+      "step": 2263
+    },
+    {
+      "epoch": 0.0029594655184358755,
+      "grad_norm": 1.2209676504135132,
+      "learning_rate": 3.8728728728728734e-05,
+      "loss": 1.4156,
+      "step": 2264
+    },
+    {
+      "epoch": 0.002960772702852146,
+      "grad_norm": 1.5215741395950317,
+      "learning_rate": 3.872372372372373e-05,
+      "loss": 2.4092,
+      "step": 2265
+    },
+    {
+      "epoch": 0.002962079887268416,
+      "grad_norm": 1.5584136247634888,
+      "learning_rate": 3.871871871871872e-05,
+      "loss": 2.3831,
+      "step": 2266
+    },
+    {
+      "epoch": 0.002963387071684686,
+      "grad_norm": 1.4257521629333496,
+      "learning_rate": 3.871371371371372e-05,
+      "loss": 1.1079,
+      "step": 2267
+    },
+    {
+      "epoch": 0.0029646942561009563,
+      "grad_norm": 1.547614574432373,
+      "learning_rate": 3.870870870870871e-05,
+      "loss": 2.0314,
+      "step": 2268
+    },
+    {
+      "epoch": 0.0029660014405172267,
+      "grad_norm": 1.739943504333496,
+      "learning_rate": 3.8703703703703705e-05,
+      "loss": 2.0542,
+      "step": 2269
+    },
+    {
+      "epoch": 0.002967308624933497,
+      "grad_norm": 1.2751171588897705,
+      "learning_rate": 3.86986986986987e-05,
+      "loss": 1.3616,
+      "step": 2270
+    },
+    {
+      "epoch": 0.0029686158093497673,
+      "grad_norm": 1.6274546384811401,
+      "learning_rate": 3.869369369369369e-05,
+      "loss": 1.4672,
+      "step": 2271
+    },
+    {
+      "epoch": 0.0029699229937660376,
+      "grad_norm": 1.175938367843628,
+      "learning_rate": 3.868868868868869e-05,
+      "loss": 1.0324,
+      "step": 2272
+    },
+    {
+      "epoch": 0.002971230178182308,
+      "grad_norm": 1.3523229360580444,
+      "learning_rate": 3.8683683683683684e-05,
+      "loss": 1.3527,
+      "step": 2273
+    },
+    {
+      "epoch": 0.0029725373625985783,
+      "grad_norm": 1.3575667142868042,
+      "learning_rate": 3.8678678678678684e-05,
+      "loss": 1.9407,
+      "step": 2274
+    },
+    {
+      "epoch": 0.002973844547014848,
+      "grad_norm": 1.0458022356033325,
+      "learning_rate": 3.867367367367368e-05,
+      "loss": 0.548,
+      "step": 2275
+    },
+    {
+      "epoch": 0.0029751517314311185,
+      "grad_norm": 1.2714483737945557,
+      "learning_rate": 3.866866866866867e-05,
+      "loss": 1.5105,
+      "step": 2276
+    },
+    {
+      "epoch": 0.0029764589158473888,
+      "grad_norm": 1.620856761932373,
+      "learning_rate": 3.866366366366366e-05,
+      "loss": 2.2875,
+      "step": 2277
+    },
+    {
+      "epoch": 0.002977766100263659,
+      "grad_norm": 1.180593729019165,
+      "learning_rate": 3.8658658658658656e-05,
+      "loss": 1.2648,
+      "step": 2278
+    },
+    {
+      "epoch": 0.0029790732846799294,
+      "grad_norm": 1.298000454902649,
+      "learning_rate": 3.8653653653653655e-05,
+      "loss": 1.5332,
+      "step": 2279
+    },
+    {
+      "epoch": 0.0029803804690961997,
+      "grad_norm": 1.1109254360198975,
+      "learning_rate": 3.864864864864865e-05,
+      "loss": 0.6654,
+      "step": 2280
+    },
+    {
+      "epoch": 0.00298168765351247,
+      "grad_norm": 1.3182607889175415,
+      "learning_rate": 3.864364364364365e-05,
+      "loss": 1.6461,
+      "step": 2281
+    },
+    {
+      "epoch": 0.00298299483792874,
+      "grad_norm": 1.3815674781799316,
+      "learning_rate": 3.863863863863864e-05,
+      "loss": 1.3511,
+      "step": 2282
+    },
+    {
+      "epoch": 0.0029843020223450103,
+      "grad_norm": 1.5189648866653442,
+      "learning_rate": 3.8633633633633634e-05,
+      "loss": 1.8139,
+      "step": 2283
+    },
+    {
+      "epoch": 0.0029856092067612806,
+      "grad_norm": 1.2007641792297363,
+      "learning_rate": 3.8628628628628634e-05,
+      "loss": 0.9596,
+      "step": 2284
+    },
+    {
+      "epoch": 0.002986916391177551,
+      "grad_norm": 1.4110480546951294,
+      "learning_rate": 3.862362362362363e-05,
+      "loss": 1.9044,
+      "step": 2285
+    },
+    {
+      "epoch": 0.0029882235755938212,
+      "grad_norm": 1.304974913597107,
+      "learning_rate": 3.861861861861862e-05,
+      "loss": 2.2202,
+      "step": 2286
+    },
+    {
+      "epoch": 0.0029895307600100915,
+      "grad_norm": 0.7858055233955383,
+      "learning_rate": 3.861361361361361e-05,
+      "loss": 0.6063,
+      "step": 2287
+    },
+    {
+      "epoch": 0.002990837944426362,
+      "grad_norm": 1.0783709287643433,
+      "learning_rate": 3.860860860860861e-05,
+      "loss": 1.2962,
+      "step": 2288
+    },
+    {
+      "epoch": 0.002992145128842632,
+      "grad_norm": 1.5946030616760254,
+      "learning_rate": 3.8603603603603605e-05,
+      "loss": 2.6741,
+      "step": 2289
+    },
+    {
+      "epoch": 0.002993452313258902,
+      "grad_norm": 1.89840829372406,
+      "learning_rate": 3.85985985985986e-05,
+      "loss": 1.7973,
+      "step": 2290
+    },
+    {
+      "epoch": 0.0029947594976751724,
+      "grad_norm": 1.0373282432556152,
+      "learning_rate": 3.85935935935936e-05,
+      "loss": 0.5151,
+      "step": 2291
+    },
+    {
+      "epoch": 0.0029960666820914427,
+      "grad_norm": 1.4962284564971924,
+      "learning_rate": 3.858858858858859e-05,
+      "loss": 1.7401,
+      "step": 2292
+    },
+    {
+      "epoch": 0.002997373866507713,
+      "grad_norm": 1.4326640367507935,
+      "learning_rate": 3.858358358358359e-05,
+      "loss": 1.2665,
+      "step": 2293
+    },
+    {
+      "epoch": 0.0029986810509239833,
+      "grad_norm": 1.0460174083709717,
+      "learning_rate": 3.857857857857858e-05,
+      "loss": 1.1553,
+      "step": 2294
+    },
+    {
+      "epoch": 0.0029999882353402537,
+      "grad_norm": 1.3629928827285767,
+      "learning_rate": 3.857357357357358e-05,
+      "loss": 1.2976,
+      "step": 2295
+    },
+    {
+      "epoch": 0.003001295419756524,
+      "grad_norm": 1.2213658094406128,
+      "learning_rate": 3.856856856856857e-05,
+      "loss": 1.2349,
+      "step": 2296
+    },
+    {
+      "epoch": 0.0030026026041727943,
+      "grad_norm": 1.7497223615646362,
+      "learning_rate": 3.856356356356356e-05,
+      "loss": 1.9588,
+      "step": 2297
+    },
+    {
+      "epoch": 0.003003909788589064,
+      "grad_norm": 4.239988327026367,
+      "learning_rate": 3.855855855855856e-05,
+      "loss": 1.1426,
+      "step": 2298
+    },
+    {
+      "epoch": 0.0030052169730053345,
+      "grad_norm": 1.3595997095108032,
+      "learning_rate": 3.8553553553553556e-05,
+      "loss": 1.6484,
+      "step": 2299
+    },
+    {
+      "epoch": 0.003006524157421605,
+      "grad_norm": 1.3246182203292847,
+      "learning_rate": 3.8548548548548555e-05,
+      "loss": 1.1512,
+      "step": 2300
+    },
+    {
+      "epoch": 0.003007831341837875,
+      "grad_norm": 1.4714034795761108,
+      "learning_rate": 3.854354354354355e-05,
+      "loss": 1.1349,
+      "step": 2301
+    },
+    {
+      "epoch": 0.0030091385262541455,
+      "grad_norm": 1.2542957067489624,
+      "learning_rate": 3.8538538538538534e-05,
+      "loss": 1.343,
+      "step": 2302
+    },
+    {
+      "epoch": 0.003010445710670416,
+      "grad_norm": 1.2829868793487549,
+      "learning_rate": 3.8533533533533534e-05,
+      "loss": 1.3385,
+      "step": 2303
+    },
+    {
+      "epoch": 0.003011752895086686,
+      "grad_norm": 1.705217719078064,
+      "learning_rate": 3.852852852852853e-05,
+      "loss": 2.4716,
+      "step": 2304
+    },
+    {
+      "epoch": 0.003013060079502956,
+      "grad_norm": 2.4408164024353027,
+      "learning_rate": 3.852352352352353e-05,
+      "loss": 2.2166,
+      "step": 2305
+    },
+    {
+      "epoch": 0.0030143672639192263,
+      "grad_norm": 0.9980645775794983,
+      "learning_rate": 3.851851851851852e-05,
+      "loss": 0.8752,
+      "step": 2306
+    },
+    {
+      "epoch": 0.0030156744483354966,
+      "grad_norm": 1.7413454055786133,
+      "learning_rate": 3.851351351351351e-05,
+      "loss": 1.9609,
+      "step": 2307
+    },
+    {
+      "epoch": 0.003016981632751767,
+      "grad_norm": 1.515007495880127,
+      "learning_rate": 3.850850850850851e-05,
+      "loss": 1.4392,
+      "step": 2308
+    },
+    {
+      "epoch": 0.0030182888171680373,
+      "grad_norm": 1.2530882358551025,
+      "learning_rate": 3.8503503503503506e-05,
+      "loss": 1.6739,
+      "step": 2309
+    },
+    {
+      "epoch": 0.0030195960015843076,
+      "grad_norm": 1.3557156324386597,
+      "learning_rate": 3.8498498498498505e-05,
+      "loss": 1.7048,
+      "step": 2310
+    },
+    {
+      "epoch": 0.003020903186000578,
+      "grad_norm": 1.4481651782989502,
+      "learning_rate": 3.849349349349349e-05,
+      "loss": 1.8929,
+      "step": 2311
+    },
+    {
+      "epoch": 0.0030222103704168482,
+      "grad_norm": 1.2424492835998535,
+      "learning_rate": 3.848848848848849e-05,
+      "loss": 1.6868,
+      "step": 2312
+    },
+    {
+      "epoch": 0.003023517554833118,
+      "grad_norm": 1.2703713178634644,
+      "learning_rate": 3.8483483483483484e-05,
+      "loss": 1.4057,
+      "step": 2313
+    },
+    {
+      "epoch": 0.0030248247392493884,
+      "grad_norm": 1.158390998840332,
+      "learning_rate": 3.847847847847848e-05,
+      "loss": 1.0573,
+      "step": 2314
+    },
+    {
+      "epoch": 0.0030261319236656588,
+      "grad_norm": 1.540954828262329,
+      "learning_rate": 3.847347347347348e-05,
+      "loss": 1.5629,
+      "step": 2315
+    },
+    {
+      "epoch": 0.003027439108081929,
+      "grad_norm": 1.433392882347107,
+      "learning_rate": 3.846846846846847e-05,
+      "loss": 1.8786,
+      "step": 2316
+    },
+    {
+      "epoch": 0.0030287462924981994,
+      "grad_norm": 1.6428567171096802,
+      "learning_rate": 3.846346346346347e-05,
+      "loss": 2.0449,
+      "step": 2317
+    },
+    {
+      "epoch": 0.0030300534769144697,
+      "grad_norm": 1.546736240386963,
+      "learning_rate": 3.845845845845846e-05,
+      "loss": 1.6832,
+      "step": 2318
+    },
+    {
+      "epoch": 0.00303136066133074,
+      "grad_norm": 1.1138025522232056,
+      "learning_rate": 3.8453453453453456e-05,
+      "loss": 1.0244,
+      "step": 2319
+    },
+    {
+      "epoch": 0.00303266784574701,
+      "grad_norm": 1.5228224992752075,
+      "learning_rate": 3.844844844844845e-05,
+      "loss": 2.0575,
+      "step": 2320
+    },
+    {
+      "epoch": 0.0030339750301632802,
+      "grad_norm": 1.3450860977172852,
+      "learning_rate": 3.844344344344344e-05,
+      "loss": 1.8457,
+      "step": 2321
+    },
+    {
+      "epoch": 0.0030352822145795506,
+      "grad_norm": 0.6514826416969299,
+      "learning_rate": 3.843843843843844e-05,
+      "loss": 0.4795,
+      "step": 2322
+    },
+    {
+      "epoch": 0.003036589398995821,
+      "grad_norm": 1.129919409751892,
+      "learning_rate": 3.8433433433433434e-05,
+      "loss": 0.8703,
+      "step": 2323
+    },
+    {
+      "epoch": 0.003037896583412091,
+      "grad_norm": 1.5874124765396118,
+      "learning_rate": 3.8428428428428434e-05,
+      "loss": 1.1049,
+      "step": 2324
+    },
+    {
+      "epoch": 0.0030392037678283615,
+      "grad_norm": 2.051623582839966,
+      "learning_rate": 3.842342342342343e-05,
+      "loss": 2.1051,
+      "step": 2325
+    },
+    {
+      "epoch": 0.003040510952244632,
+      "grad_norm": 1.4082225561141968,
+      "learning_rate": 3.841841841841842e-05,
+      "loss": 2.1579,
+      "step": 2326
+    },
+    {
+      "epoch": 0.003041818136660902,
+      "grad_norm": 1.3686426877975464,
+      "learning_rate": 3.841341341341341e-05,
+      "loss": 1.9073,
+      "step": 2327
+    },
+    {
+      "epoch": 0.003043125321077172,
+      "grad_norm": 1.306125521659851,
+      "learning_rate": 3.8408408408408406e-05,
+      "loss": 1.5586,
+      "step": 2328
+    },
+    {
+      "epoch": 0.0030444325054934424,
+      "grad_norm": 1.4132188558578491,
+      "learning_rate": 3.8403403403403406e-05,
+      "loss": 2.4167,
+      "step": 2329
+    },
+    {
+      "epoch": 0.0030457396899097127,
+      "grad_norm": 1.2818210124969482,
+      "learning_rate": 3.83983983983984e-05,
+      "loss": 1.2619,
+      "step": 2330
+    },
+    {
+      "epoch": 0.003047046874325983,
+      "grad_norm": 1.3048535585403442,
+      "learning_rate": 3.83933933933934e-05,
+      "loss": 1.5503,
+      "step": 2331
+    },
+    {
+      "epoch": 0.0030483540587422533,
+      "grad_norm": 1.3159483671188354,
+      "learning_rate": 3.838838838838839e-05,
+      "loss": 1.8206,
+      "step": 2332
+    },
+    {
+      "epoch": 0.0030496612431585236,
+      "grad_norm": 1.3755145072937012,
+      "learning_rate": 3.8383383383383384e-05,
+      "loss": 2.184,
+      "step": 2333
+    },
+    {
+      "epoch": 0.003050968427574794,
+      "grad_norm": 1.5807467699050903,
+      "learning_rate": 3.8378378378378384e-05,
+      "loss": 1.0137,
+      "step": 2334
+    },
+    {
+      "epoch": 0.0030522756119910643,
+      "grad_norm": 1.158003807067871,
+      "learning_rate": 3.837337337337337e-05,
+      "loss": 1.9711,
+      "step": 2335
+    },
+    {
+      "epoch": 0.003053582796407334,
+      "grad_norm": 1.3143372535705566,
+      "learning_rate": 3.836836836836837e-05,
+      "loss": 1.7441,
+      "step": 2336
+    },
+    {
+      "epoch": 0.0030548899808236045,
+      "grad_norm": 1.4322724342346191,
+      "learning_rate": 3.836336336336336e-05,
+      "loss": 1.4078,
+      "step": 2337
+    },
+    {
+      "epoch": 0.003056197165239875,
+      "grad_norm": 1.5119435787200928,
+      "learning_rate": 3.8358358358358356e-05,
+      "loss": 1.8212,
+      "step": 2338
+    },
+    {
+      "epoch": 0.003057504349656145,
+      "grad_norm": 1.3865551948547363,
+      "learning_rate": 3.8353353353353356e-05,
+      "loss": 1.4005,
+      "step": 2339
+    },
+    {
+      "epoch": 0.0030588115340724155,
+      "grad_norm": 1.2224832773208618,
+      "learning_rate": 3.834834834834835e-05,
+      "loss": 2.3001,
+      "step": 2340
+    },
+    {
+      "epoch": 0.0030601187184886858,
+      "grad_norm": 1.322066307067871,
+      "learning_rate": 3.834334334334335e-05,
+      "loss": 2.2216,
+      "step": 2341
+    },
+    {
+      "epoch": 0.003061425902904956,
+      "grad_norm": 1.562170386314392,
+      "learning_rate": 3.833833833833834e-05,
+      "loss": 1.2638,
+      "step": 2342
+    },
+    {
+      "epoch": 0.003062733087321226,
+      "grad_norm": 1.9327040910720825,
+      "learning_rate": 3.8333333333333334e-05,
+      "loss": 2.4613,
+      "step": 2343
+    },
+    {
+      "epoch": 0.0030640402717374963,
+      "grad_norm": 1.3933007717132568,
+      "learning_rate": 3.832832832832833e-05,
+      "loss": 1.6335,
+      "step": 2344
+    },
+    {
+      "epoch": 0.0030653474561537666,
+      "grad_norm": 1.1436634063720703,
+      "learning_rate": 3.832332332332332e-05,
+      "loss": 1.3338,
+      "step": 2345
+    },
+    {
+      "epoch": 0.003066654640570037,
+      "grad_norm": 0.1630355417728424,
+      "learning_rate": 3.831831831831832e-05,
+      "loss": 0.0046,
+      "step": 2346
+    },
+    {
+      "epoch": 0.0030679618249863073,
+      "grad_norm": 1.4216572046279907,
+      "learning_rate": 3.831331331331331e-05,
+      "loss": 2.4403,
+      "step": 2347
+    },
+    {
+      "epoch": 0.0030692690094025776,
+      "grad_norm": 1.2502360343933105,
+      "learning_rate": 3.830830830830831e-05,
+      "loss": 1.4582,
+      "step": 2348
+    },
+    {
+      "epoch": 0.003070576193818848,
+      "grad_norm": 1.5646591186523438,
+      "learning_rate": 3.8303303303303306e-05,
+      "loss": 1.8622,
+      "step": 2349
+    },
+    {
+      "epoch": 0.0030718833782351182,
+      "grad_norm": 1.8474674224853516,
+      "learning_rate": 3.82982982982983e-05,
+      "loss": 1.6097,
+      "step": 2350
+    },
+    {
+      "epoch": 0.003073190562651388,
+      "grad_norm": 1.6026530265808105,
+      "learning_rate": 3.82932932932933e-05,
+      "loss": 2.4436,
+      "step": 2351
+    },
+    {
+      "epoch": 0.0030744977470676584,
+      "grad_norm": 1.337257742881775,
+      "learning_rate": 3.8288288288288285e-05,
+      "loss": 1.7445,
+      "step": 2352
+    },
+    {
+      "epoch": 0.0030758049314839287,
+      "grad_norm": 0.9350708723068237,
+      "learning_rate": 3.8283283283283285e-05,
+      "loss": 0.8731,
+      "step": 2353
+    },
+    {
+      "epoch": 0.003077112115900199,
+      "grad_norm": 1.1121307611465454,
+      "learning_rate": 3.827827827827828e-05,
+      "loss": 1.3636,
+      "step": 2354
+    },
+    {
+      "epoch": 0.0030784193003164694,
+      "grad_norm": 1.5718612670898438,
+      "learning_rate": 3.827327327327328e-05,
+      "loss": 1.5452,
+      "step": 2355
+    },
+    {
+      "epoch": 0.0030797264847327397,
+      "grad_norm": 1.517712950706482,
+      "learning_rate": 3.826826826826827e-05,
+      "loss": 2.2907,
+      "step": 2356
+    },
+    {
+      "epoch": 0.00308103366914901,
+      "grad_norm": 1.458537220954895,
+      "learning_rate": 3.826326326326326e-05,
+      "loss": 1.6181,
+      "step": 2357
+    },
+    {
+      "epoch": 0.0030823408535652803,
+      "grad_norm": 1.2954556941986084,
+      "learning_rate": 3.825825825825826e-05,
+      "loss": 2.2099,
+      "step": 2358
+    },
+    {
+      "epoch": 0.0030836480379815502,
+      "grad_norm": 1.016371250152588,
+      "learning_rate": 3.8253253253253256e-05,
+      "loss": 0.5338,
+      "step": 2359
+    },
+    {
+      "epoch": 0.0030849552223978205,
+      "grad_norm": 1.8459898233413696,
+      "learning_rate": 3.824824824824825e-05,
+      "loss": 1.9441,
+      "step": 2360
+    },
+    {
+      "epoch": 0.003086262406814091,
+      "grad_norm": 1.407071828842163,
+      "learning_rate": 3.824324324324324e-05,
+      "loss": 1.5063,
+      "step": 2361
+    },
+    {
+      "epoch": 0.003087569591230361,
+      "grad_norm": 1.4593595266342163,
+      "learning_rate": 3.823823823823824e-05,
+      "loss": 2.0347,
+      "step": 2362
+    },
+    {
+      "epoch": 0.0030888767756466315,
+      "grad_norm": 1.5268009901046753,
+      "learning_rate": 3.8233233233233235e-05,
+      "loss": 1.6052,
+      "step": 2363
+    },
+    {
+      "epoch": 0.003090183960062902,
+      "grad_norm": 1.2676761150360107,
+      "learning_rate": 3.822822822822823e-05,
+      "loss": 1.5751,
+      "step": 2364
+    },
+    {
+      "epoch": 0.003091491144479172,
+      "grad_norm": 1.2443203926086426,
+      "learning_rate": 3.822322322322323e-05,
+      "loss": 1.1462,
+      "step": 2365
+    },
+    {
+      "epoch": 0.003092798328895442,
+      "grad_norm": 1.4173095226287842,
+      "learning_rate": 3.821821821821822e-05,
+      "loss": 1.8226,
+      "step": 2366
+    },
+    {
+      "epoch": 0.0030941055133117124,
+      "grad_norm": 1.5984820127487183,
+      "learning_rate": 3.821321321321322e-05,
+      "loss": 2.3388,
+      "step": 2367
+    },
+    {
+      "epoch": 0.0030954126977279827,
+      "grad_norm": 1.204606294631958,
+      "learning_rate": 3.820820820820821e-05,
+      "loss": 2.3189,
+      "step": 2368
+    },
+    {
+      "epoch": 0.003096719882144253,
+      "grad_norm": 1.324674367904663,
+      "learning_rate": 3.8203203203203206e-05,
+      "loss": 1.1288,
+      "step": 2369
+    },
+    {
+      "epoch": 0.0030980270665605233,
+      "grad_norm": 1.57764732837677,
+      "learning_rate": 3.81981981981982e-05,
+      "loss": 1.999,
+      "step": 2370
+    },
+    {
+      "epoch": 0.0030993342509767936,
+      "grad_norm": 1.1876490116119385,
+      "learning_rate": 3.819319319319319e-05,
+      "loss": 0.9063,
+      "step": 2371
+    },
+    {
+      "epoch": 0.003100641435393064,
+      "grad_norm": 1.239383578300476,
+      "learning_rate": 3.818818818818819e-05,
+      "loss": 0.9412,
+      "step": 2372
+    },
+    {
+      "epoch": 0.0031019486198093343,
+      "grad_norm": 1.3771158456802368,
+      "learning_rate": 3.8183183183183185e-05,
+      "loss": 1.567,
+      "step": 2373
+    },
+    {
+      "epoch": 0.003103255804225604,
+      "grad_norm": 1.5067546367645264,
+      "learning_rate": 3.8178178178178184e-05,
+      "loss": 1.1937,
+      "step": 2374
+    },
+    {
+      "epoch": 0.0031045629886418745,
+      "grad_norm": 0.9543949365615845,
+      "learning_rate": 3.817317317317318e-05,
+      "loss": 0.7236,
+      "step": 2375
+    },
+    {
+      "epoch": 0.003105870173058145,
+      "grad_norm": 1.7176799774169922,
+      "learning_rate": 3.816816816816817e-05,
+      "loss": 1.427,
+      "step": 2376
+    },
+    {
+      "epoch": 0.003107177357474415,
+      "grad_norm": 1.2930257320404053,
+      "learning_rate": 3.816316316316316e-05,
+      "loss": 1.541,
+      "step": 2377
+    },
+    {
+      "epoch": 0.0031084845418906854,
+      "grad_norm": 2.007073402404785,
+      "learning_rate": 3.8158158158158156e-05,
+      "loss": 2.1588,
+      "step": 2378
+    },
+    {
+      "epoch": 0.0031097917263069558,
+      "grad_norm": 1.1775492429733276,
+      "learning_rate": 3.8153153153153156e-05,
+      "loss": 1.0083,
+      "step": 2379
+    },
+    {
+      "epoch": 0.003111098910723226,
+      "grad_norm": 1.3931069374084473,
+      "learning_rate": 3.814814814814815e-05,
+      "loss": 1.275,
+      "step": 2380
+    },
+    {
+      "epoch": 0.0031124060951394964,
+      "grad_norm": 1.1491038799285889,
+      "learning_rate": 3.814314314314314e-05,
+      "loss": 1.348,
+      "step": 2381
+    },
+    {
+      "epoch": 0.0031137132795557663,
+      "grad_norm": 1.3869198560714722,
+      "learning_rate": 3.813813813813814e-05,
+      "loss": 1.4164,
+      "step": 2382
+    },
+    {
+      "epoch": 0.0031150204639720366,
+      "grad_norm": 1.6015539169311523,
+      "learning_rate": 3.8133133133133135e-05,
+      "loss": 1.4229,
+      "step": 2383
+    },
+    {
+      "epoch": 0.003116327648388307,
+      "grad_norm": 1.2124028205871582,
+      "learning_rate": 3.8128128128128135e-05,
+      "loss": 1.5294,
+      "step": 2384
+    },
+    {
+      "epoch": 0.0031176348328045772,
+      "grad_norm": 1.58836030960083,
+      "learning_rate": 3.812312312312312e-05,
+      "loss": 1.3201,
+      "step": 2385
+    },
+    {
+      "epoch": 0.0031189420172208476,
+      "grad_norm": 1.98471200466156,
+      "learning_rate": 3.811811811811812e-05,
+      "loss": 1.779,
+      "step": 2386
+    },
+    {
+      "epoch": 0.003120249201637118,
+      "grad_norm": 1.1605335474014282,
+      "learning_rate": 3.8113113113113113e-05,
+      "loss": 1.1444,
+      "step": 2387
+    },
+    {
+      "epoch": 0.003121556386053388,
+      "grad_norm": 1.7739923000335693,
+      "learning_rate": 3.8108108108108106e-05,
+      "loss": 1.8621,
+      "step": 2388
+    },
+    {
+      "epoch": 0.003122863570469658,
+      "grad_norm": 1.191175937652588,
+      "learning_rate": 3.8103103103103106e-05,
+      "loss": 0.9456,
+      "step": 2389
+    },
+    {
+      "epoch": 0.0031241707548859284,
+      "grad_norm": 1.5076278448104858,
+      "learning_rate": 3.80980980980981e-05,
+      "loss": 2.0756,
+      "step": 2390
+    },
+    {
+      "epoch": 0.0031254779393021987,
+      "grad_norm": 1.6646863222122192,
+      "learning_rate": 3.80930930930931e-05,
+      "loss": 1.6496,
+      "step": 2391
+    },
+    {
+      "epoch": 0.003126785123718469,
+      "grad_norm": 1.3822599649429321,
+      "learning_rate": 3.808808808808809e-05,
+      "loss": 1.4816,
+      "step": 2392
+    },
+    {
+      "epoch": 0.0031280923081347394,
+      "grad_norm": 1.3920314311981201,
+      "learning_rate": 3.8083083083083085e-05,
+      "loss": 1.6005,
+      "step": 2393
+    },
+    {
+      "epoch": 0.0031293994925510097,
+      "grad_norm": 1.2975728511810303,
+      "learning_rate": 3.807807807807808e-05,
+      "loss": 1.9922,
+      "step": 2394
+    },
+    {
+      "epoch": 0.00313070667696728,
+      "grad_norm": 1.3219879865646362,
+      "learning_rate": 3.807307307307307e-05,
+      "loss": 1.4267,
+      "step": 2395
+    },
+    {
+      "epoch": 0.0031320138613835503,
+      "grad_norm": 1.5182386636734009,
+      "learning_rate": 3.806806806806807e-05,
+      "loss": 1.0976,
+      "step": 2396
+    },
+    {
+      "epoch": 0.00313332104579982,
+      "grad_norm": 1.069297194480896,
+      "learning_rate": 3.8063063063063064e-05,
+      "loss": 0.8538,
+      "step": 2397
+    },
+    {
+      "epoch": 0.0031346282302160905,
+      "grad_norm": 1.2279318571090698,
+      "learning_rate": 3.805805805805806e-05,
+      "loss": 1.728,
+      "step": 2398
+    },
+    {
+      "epoch": 0.003135935414632361,
+      "grad_norm": 1.2709479331970215,
+      "learning_rate": 3.8053053053053056e-05,
+      "loss": 1.5271,
+      "step": 2399
+    },
+    {
+      "epoch": 0.003137242599048631,
+      "grad_norm": 1.3891257047653198,
+      "learning_rate": 3.804804804804805e-05,
+      "loss": 1.8637,
+      "step": 2400
+    },
+    {
+      "epoch": 0.0031385497834649015,
+      "grad_norm": 1.4494059085845947,
+      "learning_rate": 3.804304304304305e-05,
+      "loss": 1.3501,
+      "step": 2401
+    },
+    {
+      "epoch": 0.003139856967881172,
+      "grad_norm": 1.71126127243042,
+      "learning_rate": 3.8038038038038035e-05,
+      "loss": 1.3453,
+      "step": 2402
+    },
+    {
+      "epoch": 0.003141164152297442,
+      "grad_norm": 1.2366821765899658,
+      "learning_rate": 3.8033033033033035e-05,
+      "loss": 1.2837,
+      "step": 2403
+    },
+    {
+      "epoch": 0.0031424713367137124,
+      "grad_norm": 1.8568717241287231,
+      "learning_rate": 3.802802802802803e-05,
+      "loss": 1.8979,
+      "step": 2404
+    },
+    {
+      "epoch": 0.0031437785211299823,
+      "grad_norm": 1.1883028745651245,
+      "learning_rate": 3.802302302302303e-05,
+      "loss": 1.7925,
+      "step": 2405
+    },
+    {
+      "epoch": 0.0031450857055462527,
+      "grad_norm": 1.389450192451477,
+      "learning_rate": 3.801801801801802e-05,
+      "loss": 1.5351,
+      "step": 2406
+    },
+    {
+      "epoch": 0.003146392889962523,
+      "grad_norm": 1.4794111251831055,
+      "learning_rate": 3.8013013013013014e-05,
+      "loss": 1.8135,
+      "step": 2407
+    },
+    {
+      "epoch": 0.0031477000743787933,
+      "grad_norm": 1.4817980527877808,
+      "learning_rate": 3.800800800800801e-05,
+      "loss": 2.1043,
+      "step": 2408
+    },
+    {
+      "epoch": 0.0031490072587950636,
+      "grad_norm": 1.4443860054016113,
+      "learning_rate": 3.8003003003003006e-05,
+      "loss": 1.5531,
+      "step": 2409
+    },
+    {
+      "epoch": 0.003150314443211334,
+      "grad_norm": 1.643570065498352,
+      "learning_rate": 3.7997997997998e-05,
+      "loss": 1.1345,
+      "step": 2410
+    },
+    {
+      "epoch": 0.0031516216276276043,
+      "grad_norm": 1.2511066198349,
+      "learning_rate": 3.799299299299299e-05,
+      "loss": 1.1368,
+      "step": 2411
+    },
+    {
+      "epoch": 0.003152928812043874,
+      "grad_norm": 1.5390712022781372,
+      "learning_rate": 3.7987987987987985e-05,
+      "loss": 1.9769,
+      "step": 2412
+    },
+    {
+      "epoch": 0.0031542359964601445,
+      "grad_norm": 1.255158543586731,
+      "learning_rate": 3.7982982982982985e-05,
+      "loss": 0.9355,
+      "step": 2413
+    },
+    {
+      "epoch": 0.0031555431808764148,
+      "grad_norm": 1.523741602897644,
+      "learning_rate": 3.797797797797798e-05,
+      "loss": 1.8523,
+      "step": 2414
+    },
+    {
+      "epoch": 0.003156850365292685,
+      "grad_norm": 1.315379023551941,
+      "learning_rate": 3.797297297297298e-05,
+      "loss": 1.9278,
+      "step": 2415
+    },
+    {
+      "epoch": 0.0031581575497089554,
+      "grad_norm": 1.2756330966949463,
+      "learning_rate": 3.796796796796797e-05,
+      "loss": 1.8843,
+      "step": 2416
+    },
+    {
+      "epoch": 0.0031594647341252257,
+      "grad_norm": 1.210466980934143,
+      "learning_rate": 3.7962962962962964e-05,
+      "loss": 1.7963,
+      "step": 2417
+    },
+    {
+      "epoch": 0.003160771918541496,
+      "grad_norm": 1.2296085357666016,
+      "learning_rate": 3.7957957957957963e-05,
+      "loss": 1.123,
+      "step": 2418
+    },
+    {
+      "epoch": 0.0031620791029577664,
+      "grad_norm": 1.673952579498291,
+      "learning_rate": 3.795295295295295e-05,
+      "loss": 1.9576,
+      "step": 2419
+    },
+    {
+      "epoch": 0.0031633862873740363,
+      "grad_norm": 2.145232915878296,
+      "learning_rate": 3.794794794794795e-05,
+      "loss": 1.9309,
+      "step": 2420
+    },
+    {
+      "epoch": 0.0031646934717903066,
+      "grad_norm": 1.6996355056762695,
+      "learning_rate": 3.794294294294294e-05,
+      "loss": 1.8433,
+      "step": 2421
+    },
+    {
+      "epoch": 0.003166000656206577,
+      "grad_norm": 1.3869062662124634,
+      "learning_rate": 3.793793793793794e-05,
+      "loss": 1.1144,
+      "step": 2422
+    },
+    {
+      "epoch": 0.0031673078406228472,
+      "grad_norm": 1.3878297805786133,
+      "learning_rate": 3.7932932932932935e-05,
+      "loss": 2.1702,
+      "step": 2423
+    },
+    {
+      "epoch": 0.0031686150250391175,
+      "grad_norm": 1.2861216068267822,
+      "learning_rate": 3.792792792792793e-05,
+      "loss": 1.6801,
+      "step": 2424
+    },
+    {
+      "epoch": 0.003169922209455388,
+      "grad_norm": 1.3688194751739502,
+      "learning_rate": 3.792292292292293e-05,
+      "loss": 1.9133,
+      "step": 2425
+    },
+    {
+      "epoch": 0.003171229393871658,
+      "grad_norm": 1.355769395828247,
+      "learning_rate": 3.791791791791792e-05,
+      "loss": 1.6653,
+      "step": 2426
+    },
+    {
+      "epoch": 0.003172536578287928,
+      "grad_norm": 1.6889708042144775,
+      "learning_rate": 3.7912912912912914e-05,
+      "loss": 1.4349,
+      "step": 2427
+    },
+    {
+      "epoch": 0.0031738437627041984,
+      "grad_norm": 1.9899072647094727,
+      "learning_rate": 3.790790790790791e-05,
+      "loss": 1.3739,
+      "step": 2428
+    },
+    {
+      "epoch": 0.0031751509471204687,
+      "grad_norm": 1.5568135976791382,
+      "learning_rate": 3.7902902902902906e-05,
+      "loss": 1.9218,
+      "step": 2429
+    },
+    {
+      "epoch": 0.003176458131536739,
+      "grad_norm": 1.3795119524002075,
+      "learning_rate": 3.78978978978979e-05,
+      "loss": 1.8197,
+      "step": 2430
+    },
+    {
+      "epoch": 0.0031777653159530093,
+      "grad_norm": 1.4101924896240234,
+      "learning_rate": 3.789289289289289e-05,
+      "loss": 1.6277,
+      "step": 2431
+    },
+    {
+      "epoch": 0.0031790725003692797,
+      "grad_norm": 1.2726935148239136,
+      "learning_rate": 3.788788788788789e-05,
+      "loss": 1.2851,
+      "step": 2432
+    },
+    {
+      "epoch": 0.00318037968478555,
+      "grad_norm": 0.9670968055725098,
+      "learning_rate": 3.7882882882882885e-05,
+      "loss": 0.2231,
+      "step": 2433
+    },
+    {
+      "epoch": 0.0031816868692018203,
+      "grad_norm": 1.888349175453186,
+      "learning_rate": 3.7877877877877885e-05,
+      "loss": 1.9776,
+      "step": 2434
+    },
+    {
+      "epoch": 0.00318299405361809,
+      "grad_norm": 1.4718064069747925,
+      "learning_rate": 3.787287287287287e-05,
+      "loss": 1.9079,
+      "step": 2435
+    },
+    {
+      "epoch": 0.0031843012380343605,
+      "grad_norm": 1.3652464151382446,
+      "learning_rate": 3.786786786786787e-05,
+      "loss": 2.4355,
+      "step": 2436
+    },
+    {
+      "epoch": 0.003185608422450631,
+      "grad_norm": 1.380580186843872,
+      "learning_rate": 3.7862862862862864e-05,
+      "loss": 1.3303,
+      "step": 2437
+    },
+    {
+      "epoch": 0.003186915606866901,
+      "grad_norm": 1.4008508920669556,
+      "learning_rate": 3.785785785785786e-05,
+      "loss": 1.8553,
+      "step": 2438
+    },
+    {
+      "epoch": 0.0031882227912831715,
+      "grad_norm": 1.4369511604309082,
+      "learning_rate": 3.7852852852852857e-05,
+      "loss": 1.7748,
+      "step": 2439
+    },
+    {
+      "epoch": 0.003189529975699442,
+      "grad_norm": 1.6873114109039307,
+      "learning_rate": 3.784784784784785e-05,
+      "loss": 1.097,
+      "step": 2440
+    },
+    {
+      "epoch": 0.003190837160115712,
+      "grad_norm": 1.56759512424469,
+      "learning_rate": 3.784284284284285e-05,
+      "loss": 1.7484,
+      "step": 2441
+    },
+    {
+      "epoch": 0.0031921443445319824,
+      "grad_norm": 1.442975401878357,
+      "learning_rate": 3.783783783783784e-05,
+      "loss": 1.6264,
+      "step": 2442
+    },
+    {
+      "epoch": 0.0031934515289482523,
+      "grad_norm": 1.6204643249511719,
+      "learning_rate": 3.7832832832832835e-05,
+      "loss": 1.5728,
+      "step": 2443
+    },
+    {
+      "epoch": 0.0031947587133645226,
+      "grad_norm": 1.1744952201843262,
+      "learning_rate": 3.782782782782783e-05,
+      "loss": 1.011,
+      "step": 2444
+    },
+    {
+      "epoch": 0.003196065897780793,
+      "grad_norm": 1.2430851459503174,
+      "learning_rate": 3.782282282282282e-05,
+      "loss": 1.3694,
+      "step": 2445
+    },
+    {
+      "epoch": 0.0031973730821970633,
+      "grad_norm": 1.3541752099990845,
+      "learning_rate": 3.781781781781782e-05,
+      "loss": 0.9227,
+      "step": 2446
+    },
+    {
+      "epoch": 0.0031986802666133336,
+      "grad_norm": 1.1479287147521973,
+      "learning_rate": 3.7812812812812814e-05,
+      "loss": 1.5477,
+      "step": 2447
+    },
+    {
+      "epoch": 0.003199987451029604,
+      "grad_norm": 1.1613905429840088,
+      "learning_rate": 3.780780780780781e-05,
+      "loss": 1.2767,
+      "step": 2448
+    },
+    {
+      "epoch": 0.0032012946354458742,
+      "grad_norm": 1.526404857635498,
+      "learning_rate": 3.7802802802802807e-05,
+      "loss": 1.849,
+      "step": 2449
+    },
+    {
+      "epoch": 0.003202601819862144,
+      "grad_norm": 1.1873849630355835,
+      "learning_rate": 3.77977977977978e-05,
+      "loss": 1.1686,
+      "step": 2450
+    },
+    {
+      "epoch": 0.0032039090042784144,
+      "grad_norm": 1.2901942729949951,
+      "learning_rate": 3.77927927927928e-05,
+      "loss": 1.7549,
+      "step": 2451
+    },
+    {
+      "epoch": 0.0032052161886946848,
+      "grad_norm": 1.1057497262954712,
+      "learning_rate": 3.7787787787787786e-05,
+      "loss": 1.2608,
+      "step": 2452
+    },
+    {
+      "epoch": 0.003206523373110955,
+      "grad_norm": 1.2475495338439941,
+      "learning_rate": 3.7782782782782785e-05,
+      "loss": 1.4811,
+      "step": 2453
+    },
+    {
+      "epoch": 0.0032078305575272254,
+      "grad_norm": 1.6205404996871948,
+      "learning_rate": 3.777777777777778e-05,
+      "loss": 2.1142,
+      "step": 2454
+    },
+    {
+      "epoch": 0.0032091377419434957,
+      "grad_norm": 2.0467660427093506,
+      "learning_rate": 3.777277277277277e-05,
+      "loss": 1.8853,
+      "step": 2455
+    },
+    {
+      "epoch": 0.003210444926359766,
+      "grad_norm": 1.635558843612671,
+      "learning_rate": 3.776776776776777e-05,
+      "loss": 2.2032,
+      "step": 2456
+    },
+    {
+      "epoch": 0.0032117521107760364,
+      "grad_norm": 1.429296851158142,
+      "learning_rate": 3.7762762762762764e-05,
+      "loss": 2.5942,
+      "step": 2457
+    },
+    {
+      "epoch": 0.0032130592951923062,
+      "grad_norm": 1.3980841636657715,
+      "learning_rate": 3.7757757757757764e-05,
+      "loss": 1.5398,
+      "step": 2458
+    },
+    {
+      "epoch": 0.0032143664796085766,
+      "grad_norm": 1.4883211851119995,
+      "learning_rate": 3.775275275275276e-05,
+      "loss": 1.6078,
+      "step": 2459
+    },
+    {
+      "epoch": 0.003215673664024847,
+      "grad_norm": 1.0815743207931519,
+      "learning_rate": 3.774774774774775e-05,
+      "loss": 0.9861,
+      "step": 2460
+    },
+    {
+      "epoch": 0.003216980848441117,
+      "grad_norm": 1.3284704685211182,
+      "learning_rate": 3.774274274274274e-05,
+      "loss": 1.4686,
+      "step": 2461
+    },
+    {
+      "epoch": 0.0032182880328573875,
+      "grad_norm": 1.7349952459335327,
+      "learning_rate": 3.7737737737737736e-05,
+      "loss": 1.6738,
+      "step": 2462
+    },
+    {
+      "epoch": 0.003219595217273658,
+      "grad_norm": 1.3510106801986694,
+      "learning_rate": 3.7732732732732735e-05,
+      "loss": 1.1229,
+      "step": 2463
+    },
+    {
+      "epoch": 0.003220902401689928,
+      "grad_norm": 1.8990131616592407,
+      "learning_rate": 3.772772772772773e-05,
+      "loss": 2.0683,
+      "step": 2464
+    },
+    {
+      "epoch": 0.0032222095861061985,
+      "grad_norm": 1.4208893775939941,
+      "learning_rate": 3.772272272272273e-05,
+      "loss": 2.2948,
+      "step": 2465
+    },
+    {
+      "epoch": 0.0032235167705224684,
+      "grad_norm": 1.2444639205932617,
+      "learning_rate": 3.771771771771772e-05,
+      "loss": 1.3323,
+      "step": 2466
+    },
+    {
+      "epoch": 0.0032248239549387387,
+      "grad_norm": 1.6971265077590942,
+      "learning_rate": 3.7712712712712714e-05,
+      "loss": 2.3399,
+      "step": 2467
+    },
+    {
+      "epoch": 0.003226131139355009,
+      "grad_norm": 1.3800829648971558,
+      "learning_rate": 3.770770770770771e-05,
+      "loss": 1.4378,
+      "step": 2468
+    },
+    {
+      "epoch": 0.0032274383237712793,
+      "grad_norm": 0.9919152855873108,
+      "learning_rate": 3.77027027027027e-05,
+      "loss": 1.0721,
+      "step": 2469
+    },
+    {
+      "epoch": 0.0032287455081875496,
+      "grad_norm": 1.2109402418136597,
+      "learning_rate": 3.76976976976977e-05,
+      "loss": 0.8518,
+      "step": 2470
+    },
+    {
+      "epoch": 0.00323005269260382,
+      "grad_norm": 2.381208658218384,
+      "learning_rate": 3.769269269269269e-05,
+      "loss": 2.6183,
+      "step": 2471
+    },
+    {
+      "epoch": 0.0032313598770200903,
+      "grad_norm": 1.3683842420578003,
+      "learning_rate": 3.768768768768769e-05,
+      "loss": 1.618,
+      "step": 2472
+    },
+    {
+      "epoch": 0.00323266706143636,
+      "grad_norm": 1.5521780252456665,
+      "learning_rate": 3.7682682682682685e-05,
+      "loss": 1.8936,
+      "step": 2473
+    },
+    {
+      "epoch": 0.0032339742458526305,
+      "grad_norm": 1.155938982963562,
+      "learning_rate": 3.767767767767768e-05,
+      "loss": 1.2625,
+      "step": 2474
+    },
+    {
+      "epoch": 0.003235281430268901,
+      "grad_norm": 1.5065397024154663,
+      "learning_rate": 3.767267267267268e-05,
+      "loss": 1.3006,
+      "step": 2475
+    },
+    {
+      "epoch": 0.003236588614685171,
+      "grad_norm": 1.2928768396377563,
+      "learning_rate": 3.766766766766767e-05,
+      "loss": 0.3742,
+      "step": 2476
+    },
+    {
+      "epoch": 0.0032378957991014415,
+      "grad_norm": 1.2482107877731323,
+      "learning_rate": 3.7662662662662664e-05,
+      "loss": 1.7933,
+      "step": 2477
+    },
+    {
+      "epoch": 0.0032392029835177118,
+      "grad_norm": 1.7992655038833618,
+      "learning_rate": 3.765765765765766e-05,
+      "loss": 2.138,
+      "step": 2478
+    },
+    {
+      "epoch": 0.003240510167933982,
+      "grad_norm": 1.5207356214523315,
+      "learning_rate": 3.765265265265266e-05,
+      "loss": 1.5466,
+      "step": 2479
+    },
+    {
+      "epoch": 0.0032418173523502524,
+      "grad_norm": 1.6592912673950195,
+      "learning_rate": 3.764764764764765e-05,
+      "loss": 2.0527,
+      "step": 2480
+    },
+    {
+      "epoch": 0.0032431245367665223,
+      "grad_norm": 1.3103289604187012,
+      "learning_rate": 3.764264264264264e-05,
+      "loss": 1.8252,
+      "step": 2481
+    },
+    {
+      "epoch": 0.0032444317211827926,
+      "grad_norm": 1.458674669265747,
+      "learning_rate": 3.763763763763764e-05,
+      "loss": 2.0628,
+      "step": 2482
+    },
+    {
+      "epoch": 0.003245738905599063,
+      "grad_norm": 1.570713758468628,
+      "learning_rate": 3.7632632632632635e-05,
+      "loss": 1.5234,
+      "step": 2483
+    },
+    {
+      "epoch": 0.0032470460900153333,
+      "grad_norm": 1.4717050790786743,
+      "learning_rate": 3.762762762762763e-05,
+      "loss": 1.678,
+      "step": 2484
+    },
+    {
+      "epoch": 0.0032483532744316036,
+      "grad_norm": 1.5631067752838135,
+      "learning_rate": 3.762262262262262e-05,
+      "loss": 1.872,
+      "step": 2485
+    },
+    {
+      "epoch": 0.003249660458847874,
+      "grad_norm": 1.3663777112960815,
+      "learning_rate": 3.7617617617617614e-05,
+      "loss": 2.263,
+      "step": 2486
+    },
+    {
+      "epoch": 0.003250967643264144,
+      "grad_norm": 1.408795952796936,
+      "learning_rate": 3.7612612612612614e-05,
+      "loss": 1.6472,
+      "step": 2487
+    },
+    {
+      "epoch": 0.0032522748276804145,
+      "grad_norm": 1.5756348371505737,
+      "learning_rate": 3.760760760760761e-05,
+      "loss": 2.4986,
+      "step": 2488
+    },
+    {
+      "epoch": 0.0032535820120966844,
+      "grad_norm": 1.2400751113891602,
+      "learning_rate": 3.760260260260261e-05,
+      "loss": 1.1463,
+      "step": 2489
+    },
+    {
+      "epoch": 0.0032548891965129547,
+      "grad_norm": 1.253077745437622,
+      "learning_rate": 3.75975975975976e-05,
+      "loss": 0.7989,
+      "step": 2490
+    },
+    {
+      "epoch": 0.003256196380929225,
+      "grad_norm": 1.044288992881775,
+      "learning_rate": 3.759259259259259e-05,
+      "loss": 1.0988,
+      "step": 2491
+    },
+    {
+      "epoch": 0.0032575035653454954,
+      "grad_norm": 1.7320916652679443,
+      "learning_rate": 3.758758758758759e-05,
+      "loss": 2.0028,
+      "step": 2492
+    },
+    {
+      "epoch": 0.0032588107497617657,
+      "grad_norm": 2.357729196548462,
+      "learning_rate": 3.758258258258258e-05,
+      "loss": 1.9613,
+      "step": 2493
+    },
+    {
+      "epoch": 0.003260117934178036,
+      "grad_norm": 1.6686489582061768,
+      "learning_rate": 3.757757757757758e-05,
+      "loss": 1.8527,
+      "step": 2494
+    },
+    {
+      "epoch": 0.0032614251185943063,
+      "grad_norm": 1.3781086206436157,
+      "learning_rate": 3.757257257257257e-05,
+      "loss": 1.3669,
+      "step": 2495
+    },
+    {
+      "epoch": 0.0032627323030105762,
+      "grad_norm": 1.394272804260254,
+      "learning_rate": 3.756756756756757e-05,
+      "loss": 2.0849,
+      "step": 2496
+    },
+    {
+      "epoch": 0.0032640394874268465,
+      "grad_norm": 1.386030912399292,
+      "learning_rate": 3.7562562562562564e-05,
+      "loss": 1.1837,
+      "step": 2497
+    },
+    {
+      "epoch": 0.003265346671843117,
+      "grad_norm": 1.4294358491897583,
+      "learning_rate": 3.755755755755756e-05,
+      "loss": 1.5788,
+      "step": 2498
+    },
+    {
+      "epoch": 0.003266653856259387,
+      "grad_norm": 1.3983025550842285,
+      "learning_rate": 3.755255255255256e-05,
+      "loss": 2.0227,
+      "step": 2499
+    },
+    {
+      "epoch": 0.0032679610406756575,
+      "grad_norm": 1.149295687675476,
+      "learning_rate": 3.754754754754755e-05,
+      "loss": 0.8955,
+      "step": 2500
+    },
+    {
+      "epoch": 0.003269268225091928,
+      "grad_norm": 1.6406313180923462,
+      "learning_rate": 3.754254254254255e-05,
+      "loss": 2.45,
+      "step": 2501
+    },
+    {
+      "epoch": 0.003270575409508198,
+      "grad_norm": 1.1378432512283325,
+      "learning_rate": 3.7537537537537536e-05,
+      "loss": 1.5229,
+      "step": 2502
+    },
+    {
+      "epoch": 0.0032718825939244685,
+      "grad_norm": 1.4857556819915771,
+      "learning_rate": 3.7532532532532536e-05,
+      "loss": 2.4252,
+      "step": 2503
+    },
+    {
+      "epoch": 0.0032731897783407383,
+      "grad_norm": 1.4522247314453125,
+      "learning_rate": 3.752752752752753e-05,
+      "loss": 2.0181,
+      "step": 2504
+    },
+    {
+      "epoch": 0.0032744969627570087,
+      "grad_norm": 1.4195541143417358,
+      "learning_rate": 3.752252252252252e-05,
+      "loss": 1.5216,
+      "step": 2505
+    },
+    {
+      "epoch": 0.003275804147173279,
+      "grad_norm": 1.2635961771011353,
+      "learning_rate": 3.751751751751752e-05,
+      "loss": 1.5098,
+      "step": 2506
+    },
+    {
+      "epoch": 0.0032771113315895493,
+      "grad_norm": 1.6600862741470337,
+      "learning_rate": 3.7512512512512514e-05,
+      "loss": 1.8719,
+      "step": 2507
+    },
+    {
+      "epoch": 0.0032784185160058196,
+      "grad_norm": 1.4099420309066772,
+      "learning_rate": 3.7507507507507514e-05,
+      "loss": 1.5088,
+      "step": 2508
+    },
+    {
+      "epoch": 0.00327972570042209,
+      "grad_norm": 1.4241465330123901,
+      "learning_rate": 3.750250250250251e-05,
+      "loss": 2.4648,
+      "step": 2509
+    },
+    {
+      "epoch": 0.0032810328848383603,
+      "grad_norm": 1.4309364557266235,
+      "learning_rate": 3.74974974974975e-05,
+      "loss": 1.9792,
+      "step": 2510
+    },
+    {
+      "epoch": 0.0032823400692546306,
+      "grad_norm": 1.1072784662246704,
+      "learning_rate": 3.749249249249249e-05,
+      "loss": 1.0823,
+      "step": 2511
+    },
+    {
+      "epoch": 0.0032836472536709005,
+      "grad_norm": 0.9064189195632935,
+      "learning_rate": 3.7487487487487486e-05,
+      "loss": 0.9415,
+      "step": 2512
+    },
+    {
+      "epoch": 0.003284954438087171,
+      "grad_norm": 1.3355475664138794,
+      "learning_rate": 3.7482482482482486e-05,
+      "loss": 1.4494,
+      "step": 2513
+    },
+    {
+      "epoch": 0.003286261622503441,
+      "grad_norm": 1.4068266153335571,
+      "learning_rate": 3.747747747747748e-05,
+      "loss": 2.0119,
+      "step": 2514
+    },
+    {
+      "epoch": 0.0032875688069197114,
+      "grad_norm": 1.383216142654419,
+      "learning_rate": 3.747247247247248e-05,
+      "loss": 0.9547,
+      "step": 2515
+    },
+    {
+      "epoch": 0.0032888759913359818,
+      "grad_norm": 1.4354058504104614,
+      "learning_rate": 3.746746746746747e-05,
+      "loss": 1.7676,
+      "step": 2516
+    },
+    {
+      "epoch": 0.003290183175752252,
+      "grad_norm": 0.12145306169986725,
+      "learning_rate": 3.7462462462462464e-05,
+      "loss": 0.0035,
+      "step": 2517
+    },
+    {
+      "epoch": 0.0032914903601685224,
+      "grad_norm": 1.6261683702468872,
+      "learning_rate": 3.745745745745746e-05,
+      "loss": 1.4693,
+      "step": 2518
+    },
+    {
+      "epoch": 0.0032927975445847923,
+      "grad_norm": 1.241979718208313,
+      "learning_rate": 3.745245245245245e-05,
+      "loss": 1.4207,
+      "step": 2519
+    },
+    {
+      "epoch": 0.0032941047290010626,
+      "grad_norm": 1.7058180570602417,
+      "learning_rate": 3.744744744744745e-05,
+      "loss": 2.2038,
+      "step": 2520
+    },
+    {
+      "epoch": 0.003295411913417333,
+      "grad_norm": 1.3528460264205933,
+      "learning_rate": 3.744244244244244e-05,
+      "loss": 1.6228,
+      "step": 2521
+    },
+    {
+      "epoch": 0.0032967190978336032,
+      "grad_norm": 1.1599621772766113,
+      "learning_rate": 3.7437437437437436e-05,
+      "loss": 0.8089,
+      "step": 2522
+    },
+    {
+      "epoch": 0.0032980262822498736,
+      "grad_norm": 1.7642662525177002,
+      "learning_rate": 3.7432432432432436e-05,
+      "loss": 2.86,
+      "step": 2523
+    },
+    {
+      "epoch": 0.003299333466666144,
+      "grad_norm": 1.5030518770217896,
+      "learning_rate": 3.742742742742743e-05,
+      "loss": 2.1198,
+      "step": 2524
+    },
+    {
+      "epoch": 0.003300640651082414,
+      "grad_norm": 1.6342583894729614,
+      "learning_rate": 3.742242242242243e-05,
+      "loss": 1.8132,
+      "step": 2525
+    },
+    {
+      "epoch": 0.0033019478354986845,
+      "grad_norm": 1.2257287502288818,
+      "learning_rate": 3.741741741741742e-05,
+      "loss": 1.0882,
+      "step": 2526
+    },
+    {
+      "epoch": 0.0033032550199149544,
+      "grad_norm": 1.740208625793457,
+      "learning_rate": 3.7412412412412414e-05,
+      "loss": 1.8535,
+      "step": 2527
+    },
+    {
+      "epoch": 0.0033045622043312247,
+      "grad_norm": 1.3340537548065186,
+      "learning_rate": 3.740740740740741e-05,
+      "loss": 2.1866,
+      "step": 2528
+    },
+    {
+      "epoch": 0.003305869388747495,
+      "grad_norm": 1.806502342224121,
+      "learning_rate": 3.74024024024024e-05,
+      "loss": 1.8411,
+      "step": 2529
+    },
+    {
+      "epoch": 0.0033071765731637654,
+      "grad_norm": 1.4850990772247314,
+      "learning_rate": 3.73973973973974e-05,
+      "loss": 2.177,
+      "step": 2530
+    },
+    {
+      "epoch": 0.0033084837575800357,
+      "grad_norm": 1.441721796989441,
+      "learning_rate": 3.739239239239239e-05,
+      "loss": 0.6232,
+      "step": 2531
+    },
+    {
+      "epoch": 0.003309790941996306,
+      "grad_norm": 1.1141750812530518,
+      "learning_rate": 3.738738738738739e-05,
+      "loss": 1.252,
+      "step": 2532
+    },
+    {
+      "epoch": 0.0033110981264125763,
+      "grad_norm": 1.3428772687911987,
+      "learning_rate": 3.7382382382382386e-05,
+      "loss": 1.3888,
+      "step": 2533
+    },
+    {
+      "epoch": 0.0033124053108288466,
+      "grad_norm": 1.4811742305755615,
+      "learning_rate": 3.737737737737738e-05,
+      "loss": 1.3163,
+      "step": 2534
+    },
+    {
+      "epoch": 0.0033137124952451165,
+      "grad_norm": 0.9579393863677979,
+      "learning_rate": 3.737237237237237e-05,
+      "loss": 1.0307,
+      "step": 2535
+    },
+    {
+      "epoch": 0.003315019679661387,
+      "grad_norm": 1.0461969375610352,
+      "learning_rate": 3.7367367367367365e-05,
+      "loss": 0.6798,
+      "step": 2536
+    },
+    {
+      "epoch": 0.003316326864077657,
+      "grad_norm": 1.275619626045227,
+      "learning_rate": 3.7362362362362364e-05,
+      "loss": 1.8447,
+      "step": 2537
+    },
+    {
+      "epoch": 0.0033176340484939275,
+      "grad_norm": 0.8445043563842773,
+      "learning_rate": 3.735735735735736e-05,
+      "loss": 0.4665,
+      "step": 2538
+    },
+    {
+      "epoch": 0.003318941232910198,
+      "grad_norm": 1.431227445602417,
+      "learning_rate": 3.735235235235236e-05,
+      "loss": 1.3295,
+      "step": 2539
+    },
+    {
+      "epoch": 0.003320248417326468,
+      "grad_norm": 1.4216489791870117,
+      "learning_rate": 3.734734734734735e-05,
+      "loss": 1.8289,
+      "step": 2540
+    },
+    {
+      "epoch": 0.0033215556017427384,
+      "grad_norm": 1.4469658136367798,
+      "learning_rate": 3.734234234234234e-05,
+      "loss": 1.6159,
+      "step": 2541
+    },
+    {
+      "epoch": 0.0033228627861590083,
+      "grad_norm": 1.6238259077072144,
+      "learning_rate": 3.733733733733734e-05,
+      "loss": 1.4307,
+      "step": 2542
+    },
+    {
+      "epoch": 0.0033241699705752786,
+      "grad_norm": 1.8158258199691772,
+      "learning_rate": 3.733233233233233e-05,
+      "loss": 0.9965,
+      "step": 2543
+    },
+    {
+      "epoch": 0.003325477154991549,
+      "grad_norm": 1.9678561687469482,
+      "learning_rate": 3.732732732732733e-05,
+      "loss": 1.7413,
+      "step": 2544
+    },
+    {
+      "epoch": 0.0033267843394078193,
+      "grad_norm": 1.184171199798584,
+      "learning_rate": 3.732232232232232e-05,
+      "loss": 0.6777,
+      "step": 2545
+    },
+    {
+      "epoch": 0.0033280915238240896,
+      "grad_norm": 2.1912789344787598,
+      "learning_rate": 3.731731731731732e-05,
+      "loss": 2.4439,
+      "step": 2546
+    },
+    {
+      "epoch": 0.00332939870824036,
+      "grad_norm": 1.0247867107391357,
+      "learning_rate": 3.7312312312312315e-05,
+      "loss": 1.314,
+      "step": 2547
+    },
+    {
+      "epoch": 0.0033307058926566302,
+      "grad_norm": 1.0994993448257446,
+      "learning_rate": 3.730730730730731e-05,
+      "loss": 0.7828,
+      "step": 2548
+    },
+    {
+      "epoch": 0.0033320130770729006,
+      "grad_norm": 1.3058141469955444,
+      "learning_rate": 3.730230230230231e-05,
+      "loss": 1.2037,
+      "step": 2549
+    },
+    {
+      "epoch": 0.0033333202614891705,
+      "grad_norm": 1.4790083169937134,
+      "learning_rate": 3.72972972972973e-05,
+      "loss": 1.7682,
+      "step": 2550
+    },
+    {
+      "epoch": 0.0033346274459054408,
+      "grad_norm": 1.4255295991897583,
+      "learning_rate": 3.72922922922923e-05,
+      "loss": 1.2565,
+      "step": 2551
+    },
+    {
+      "epoch": 0.003335934630321711,
+      "grad_norm": 1.2772984504699707,
+      "learning_rate": 3.7287287287287286e-05,
+      "loss": 0.8638,
+      "step": 2552
+    },
+    {
+      "epoch": 0.0033372418147379814,
+      "grad_norm": 4.120028972625732,
+      "learning_rate": 3.7282282282282286e-05,
+      "loss": 1.4935,
+      "step": 2553
+    },
+    {
+      "epoch": 0.0033385489991542517,
+      "grad_norm": 1.8560603857040405,
+      "learning_rate": 3.727727727727728e-05,
+      "loss": 1.7107,
+      "step": 2554
+    },
+    {
+      "epoch": 0.003339856183570522,
+      "grad_norm": 1.3424088954925537,
+      "learning_rate": 3.727227227227227e-05,
+      "loss": 2.1343,
+      "step": 2555
+    },
+    {
+      "epoch": 0.0033411633679867924,
+      "grad_norm": 1.4994310140609741,
+      "learning_rate": 3.726726726726727e-05,
+      "loss": 1.3606,
+      "step": 2556
+    },
+    {
+      "epoch": 0.0033424705524030623,
+      "grad_norm": 1.5294674634933472,
+      "learning_rate": 3.7262262262262265e-05,
+      "loss": 1.6676,
+      "step": 2557
+    },
+    {
+      "epoch": 0.0033437777368193326,
+      "grad_norm": 1.3413041830062866,
+      "learning_rate": 3.725725725725726e-05,
+      "loss": 1.8258,
+      "step": 2558
+    },
+    {
+      "epoch": 0.003345084921235603,
+      "grad_norm": 1.3173037767410278,
+      "learning_rate": 3.725225225225226e-05,
+      "loss": 1.1467,
+      "step": 2559
+    },
+    {
+      "epoch": 0.0033463921056518732,
+      "grad_norm": 1.3749955892562866,
+      "learning_rate": 3.7247247247247244e-05,
+      "loss": 1.406,
+      "step": 2560
+    },
+    {
+      "epoch": 0.0033476992900681435,
+      "grad_norm": 1.1830005645751953,
+      "learning_rate": 3.724224224224224e-05,
+      "loss": 1.6144,
+      "step": 2561
+    },
+    {
+      "epoch": 0.003349006474484414,
+      "grad_norm": 1.6450592279434204,
+      "learning_rate": 3.7237237237237236e-05,
+      "loss": 2.2723,
+      "step": 2562
+    },
+    {
+      "epoch": 0.003350313658900684,
+      "grad_norm": 1.0381273031234741,
+      "learning_rate": 3.7232232232232236e-05,
+      "loss": 0.618,
+      "step": 2563
+    },
+    {
+      "epoch": 0.0033516208433169545,
+      "grad_norm": 1.7593207359313965,
+      "learning_rate": 3.722722722722723e-05,
+      "loss": 1.8748,
+      "step": 2564
+    },
+    {
+      "epoch": 0.0033529280277332244,
+      "grad_norm": 0.8685852885246277,
+      "learning_rate": 3.722222222222222e-05,
+      "loss": 0.6301,
+      "step": 2565
+    },
+    {
+      "epoch": 0.0033542352121494947,
+      "grad_norm": 1.2465909719467163,
+      "learning_rate": 3.721721721721722e-05,
+      "loss": 1.413,
+      "step": 2566
+    },
+    {
+      "epoch": 0.003355542396565765,
+      "grad_norm": 1.0940313339233398,
+      "learning_rate": 3.7212212212212215e-05,
+      "loss": 1.3433,
+      "step": 2567
+    },
+    {
+      "epoch": 0.0033568495809820353,
+      "grad_norm": 1.554767370223999,
+      "learning_rate": 3.720720720720721e-05,
+      "loss": 1.9255,
+      "step": 2568
+    },
+    {
+      "epoch": 0.0033581567653983057,
+      "grad_norm": 1.2741312980651855,
+      "learning_rate": 3.72022022022022e-05,
+      "loss": 1.3144,
+      "step": 2569
+    },
+    {
+      "epoch": 0.003359463949814576,
+      "grad_norm": 1.7261019945144653,
+      "learning_rate": 3.71971971971972e-05,
+      "loss": 1.7591,
+      "step": 2570
+    },
+    {
+      "epoch": 0.0033607711342308463,
+      "grad_norm": 1.1016452312469482,
+      "learning_rate": 3.719219219219219e-05,
+      "loss": 0.876,
+      "step": 2571
+    },
+    {
+      "epoch": 0.0033620783186471166,
+      "grad_norm": 1.3392622470855713,
+      "learning_rate": 3.7187187187187186e-05,
+      "loss": 2.4038,
+      "step": 2572
+    },
+    {
+      "epoch": 0.0033633855030633865,
+      "grad_norm": 1.591395616531372,
+      "learning_rate": 3.7182182182182186e-05,
+      "loss": 2.2325,
+      "step": 2573
+    },
+    {
+      "epoch": 0.003364692687479657,
+      "grad_norm": 1.1574420928955078,
+      "learning_rate": 3.717717717717718e-05,
+      "loss": 1.694,
+      "step": 2574
+    },
+    {
+      "epoch": 0.003365999871895927,
+      "grad_norm": 1.4675060510635376,
+      "learning_rate": 3.717217217217218e-05,
+      "loss": 1.37,
+      "step": 2575
+    },
+    {
+      "epoch": 0.0033673070563121975,
+      "grad_norm": 1.2404601573944092,
+      "learning_rate": 3.7167167167167165e-05,
+      "loss": 1.2372,
+      "step": 2576
+    },
+    {
+      "epoch": 0.003368614240728468,
+      "grad_norm": 1.384019136428833,
+      "learning_rate": 3.7162162162162165e-05,
+      "loss": 1.1857,
+      "step": 2577
+    },
+    {
+      "epoch": 0.003369921425144738,
+      "grad_norm": 1.1118812561035156,
+      "learning_rate": 3.715715715715716e-05,
+      "loss": 0.4314,
+      "step": 2578
+    },
+    {
+      "epoch": 0.0033712286095610084,
+      "grad_norm": 1.2168631553649902,
+      "learning_rate": 3.715215215215215e-05,
+      "loss": 1.939,
+      "step": 2579
+    },
+    {
+      "epoch": 0.0033725357939772783,
+      "grad_norm": 1.7780351638793945,
+      "learning_rate": 3.714714714714715e-05,
+      "loss": 2.0179,
+      "step": 2580
+    },
+    {
+      "epoch": 0.0033738429783935486,
+      "grad_norm": 1.5881954431533813,
+      "learning_rate": 3.7142142142142143e-05,
+      "loss": 1.614,
+      "step": 2581
+    },
+    {
+      "epoch": 0.003375150162809819,
+      "grad_norm": 1.0476269721984863,
+      "learning_rate": 3.713713713713714e-05,
+      "loss": 1.6065,
+      "step": 2582
+    },
+    {
+      "epoch": 0.0033764573472260893,
+      "grad_norm": 1.156821608543396,
+      "learning_rate": 3.7132132132132136e-05,
+      "loss": 1.559,
+      "step": 2583
+    },
+    {
+      "epoch": 0.0033777645316423596,
+      "grad_norm": 1.3882472515106201,
+      "learning_rate": 3.712712712712713e-05,
+      "loss": 1.9964,
+      "step": 2584
+    },
+    {
+      "epoch": 0.00337907171605863,
+      "grad_norm": 1.3765472173690796,
+      "learning_rate": 3.712212212212212e-05,
+      "loss": 1.5688,
+      "step": 2585
+    },
+    {
+      "epoch": 0.0033803789004749002,
+      "grad_norm": 1.234465479850769,
+      "learning_rate": 3.7117117117117115e-05,
+      "loss": 1.1902,
+      "step": 2586
+    },
+    {
+      "epoch": 0.0033816860848911705,
+      "grad_norm": 1.4619053602218628,
+      "learning_rate": 3.7112112112112115e-05,
+      "loss": 1.2633,
+      "step": 2587
+    },
+    {
+      "epoch": 0.0033829932693074404,
+      "grad_norm": 1.4315576553344727,
+      "learning_rate": 3.710710710710711e-05,
+      "loss": 1.8368,
+      "step": 2588
+    },
+    {
+      "epoch": 0.0033843004537237108,
+      "grad_norm": 1.488231897354126,
+      "learning_rate": 3.710210210210211e-05,
+      "loss": 2.2069,
+      "step": 2589
+    },
+    {
+      "epoch": 0.003385607638139981,
+      "grad_norm": 1.0521318912506104,
+      "learning_rate": 3.70970970970971e-05,
+      "loss": 0.8635,
+      "step": 2590
+    },
+    {
+      "epoch": 0.0033869148225562514,
+      "grad_norm": 1.3446044921875,
+      "learning_rate": 3.7092092092092094e-05,
+      "loss": 1.3441,
+      "step": 2591
+    },
+    {
+      "epoch": 0.0033882220069725217,
+      "grad_norm": 1.30499267578125,
+      "learning_rate": 3.708708708708709e-05,
+      "loss": 1.3296,
+      "step": 2592
+    },
+    {
+      "epoch": 0.003389529191388792,
+      "grad_norm": 1.1514889001846313,
+      "learning_rate": 3.708208208208208e-05,
+      "loss": 1.4173,
+      "step": 2593
+    },
+    {
+      "epoch": 0.0033908363758050624,
+      "grad_norm": 0.853351891040802,
+      "learning_rate": 3.707707707707708e-05,
+      "loss": 0.6029,
+      "step": 2594
+    },
+    {
+      "epoch": 0.0033921435602213327,
+      "grad_norm": 1.3075495958328247,
+      "learning_rate": 3.707207207207207e-05,
+      "loss": 1.5047,
+      "step": 2595
+    },
+    {
+      "epoch": 0.0033934507446376026,
+      "grad_norm": 1.3809174299240112,
+      "learning_rate": 3.7067067067067065e-05,
+      "loss": 1.2783,
+      "step": 2596
+    },
+    {
+      "epoch": 0.003394757929053873,
+      "grad_norm": 3.566250801086426,
+      "learning_rate": 3.7062062062062065e-05,
+      "loss": 1.0969,
+      "step": 2597
+    },
+    {
+      "epoch": 0.003396065113470143,
+      "grad_norm": 1.419678807258606,
+      "learning_rate": 3.705705705705706e-05,
+      "loss": 1.7541,
+      "step": 2598
+    },
+    {
+      "epoch": 0.0033973722978864135,
+      "grad_norm": 1.0927537679672241,
+      "learning_rate": 3.705205205205206e-05,
+      "loss": 0.6953,
+      "step": 2599
+    },
+    {
+      "epoch": 0.003398679482302684,
+      "grad_norm": 1.4772865772247314,
+      "learning_rate": 3.704704704704705e-05,
+      "loss": 1.7545,
+      "step": 2600
+    },
+    {
+      "epoch": 0.003399986666718954,
+      "grad_norm": 1.1703388690948486,
+      "learning_rate": 3.7042042042042044e-05,
+      "loss": 1.0485,
+      "step": 2601
+    },
+    {
+      "epoch": 0.0034012938511352245,
+      "grad_norm": 0.9048022627830505,
+      "learning_rate": 3.7037037037037037e-05,
+      "loss": 0.6695,
+      "step": 2602
+    },
+    {
+      "epoch": 0.0034026010355514944,
+      "grad_norm": 1.5391570329666138,
+      "learning_rate": 3.703203203203203e-05,
+      "loss": 2.0867,
+      "step": 2603
+    },
+    {
+      "epoch": 0.0034039082199677647,
+      "grad_norm": 1.7999333143234253,
+      "learning_rate": 3.702702702702703e-05,
+      "loss": 1.9692,
+      "step": 2604
+    },
+    {
+      "epoch": 0.003405215404384035,
+      "grad_norm": 1.4057328701019287,
+      "learning_rate": 3.702202202202202e-05,
+      "loss": 1.5347,
+      "step": 2605
+    },
+    {
+      "epoch": 0.0034065225888003053,
+      "grad_norm": 1.5669541358947754,
+      "learning_rate": 3.701701701701702e-05,
+      "loss": 1.2676,
+      "step": 2606
+    },
+    {
+      "epoch": 0.0034078297732165756,
+      "grad_norm": 1.4838526248931885,
+      "learning_rate": 3.7012012012012015e-05,
+      "loss": 1.1219,
+      "step": 2607
+    },
+    {
+      "epoch": 0.003409136957632846,
+      "grad_norm": 1.279561161994934,
+      "learning_rate": 3.700700700700701e-05,
+      "loss": 2.0665,
+      "step": 2608
+    },
+    {
+      "epoch": 0.0034104441420491163,
+      "grad_norm": 1.0365511178970337,
+      "learning_rate": 3.700200200200201e-05,
+      "loss": 1.1248,
+      "step": 2609
+    },
+    {
+      "epoch": 0.0034117513264653866,
+      "grad_norm": 1.375585913658142,
+      "learning_rate": 3.6996996996996994e-05,
+      "loss": 1.9316,
+      "step": 2610
+    },
+    {
+      "epoch": 0.0034130585108816565,
+      "grad_norm": 1.3522917032241821,
+      "learning_rate": 3.6991991991991994e-05,
+      "loss": 1.3612,
+      "step": 2611
+    },
+    {
+      "epoch": 0.003414365695297927,
+      "grad_norm": 1.5002617835998535,
+      "learning_rate": 3.698698698698699e-05,
+      "loss": 1.487,
+      "step": 2612
+    },
+    {
+      "epoch": 0.003415672879714197,
+      "grad_norm": 1.509361743927002,
+      "learning_rate": 3.6981981981981986e-05,
+      "loss": 1.6028,
+      "step": 2613
+    },
+    {
+      "epoch": 0.0034169800641304674,
+      "grad_norm": 1.6538374423980713,
+      "learning_rate": 3.697697697697698e-05,
+      "loss": 1.5026,
+      "step": 2614
+    },
+    {
+      "epoch": 0.0034182872485467378,
+      "grad_norm": 1.4952061176300049,
+      "learning_rate": 3.697197197197197e-05,
+      "loss": 1.1088,
+      "step": 2615
+    },
+    {
+      "epoch": 0.003419594432963008,
+      "grad_norm": 1.0907396078109741,
+      "learning_rate": 3.696696696696697e-05,
+      "loss": 1.0042,
+      "step": 2616
+    },
+    {
+      "epoch": 0.0034209016173792784,
+      "grad_norm": 1.0086358785629272,
+      "learning_rate": 3.6961961961961965e-05,
+      "loss": 0.9976,
+      "step": 2617
+    },
+    {
+      "epoch": 0.0034222088017955487,
+      "grad_norm": 1.3923410177230835,
+      "learning_rate": 3.695695695695696e-05,
+      "loss": 1.6113,
+      "step": 2618
+    },
+    {
+      "epoch": 0.0034235159862118186,
+      "grad_norm": 1.5339523553848267,
+      "learning_rate": 3.695195195195195e-05,
+      "loss": 1.5962,
+      "step": 2619
+    },
+    {
+      "epoch": 0.003424823170628089,
+      "grad_norm": 1.1996715068817139,
+      "learning_rate": 3.694694694694695e-05,
+      "loss": 1.7018,
+      "step": 2620
+    },
+    {
+      "epoch": 0.0034261303550443593,
+      "grad_norm": 1.3185124397277832,
+      "learning_rate": 3.6941941941941944e-05,
+      "loss": 1.0407,
+      "step": 2621
+    },
+    {
+      "epoch": 0.0034274375394606296,
+      "grad_norm": 1.3789622783660889,
+      "learning_rate": 3.693693693693694e-05,
+      "loss": 1.691,
+      "step": 2622
+    },
+    {
+      "epoch": 0.0034287447238769,
+      "grad_norm": 1.4358243942260742,
+      "learning_rate": 3.6931931931931936e-05,
+      "loss": 1.1309,
+      "step": 2623
+    },
+    {
+      "epoch": 0.00343005190829317,
+      "grad_norm": 1.6063718795776367,
+      "learning_rate": 3.692692692692693e-05,
+      "loss": 1.8654,
+      "step": 2624
+    },
+    {
+      "epoch": 0.0034313590927094405,
+      "grad_norm": 1.8692991733551025,
+      "learning_rate": 3.692192192192193e-05,
+      "loss": 1.8019,
+      "step": 2625
+    },
+    {
+      "epoch": 0.0034326662771257104,
+      "grad_norm": 1.570290207862854,
+      "learning_rate": 3.6916916916916915e-05,
+      "loss": 1.8528,
+      "step": 2626
+    },
+    {
+      "epoch": 0.0034339734615419807,
+      "grad_norm": 1.3903871774673462,
+      "learning_rate": 3.691191191191191e-05,
+      "loss": 2.4141,
+      "step": 2627
+    },
+    {
+      "epoch": 0.003435280645958251,
+      "grad_norm": 1.2677329778671265,
+      "learning_rate": 3.690690690690691e-05,
+      "loss": 1.3082,
+      "step": 2628
+    },
+    {
+      "epoch": 0.0034365878303745214,
+      "grad_norm": 1.3496392965316772,
+      "learning_rate": 3.69019019019019e-05,
+      "loss": 1.5881,
+      "step": 2629
+    },
+    {
+      "epoch": 0.0034378950147907917,
+      "grad_norm": 1.5704869031906128,
+      "learning_rate": 3.68968968968969e-05,
+      "loss": 0.9427,
+      "step": 2630
+    },
+    {
+      "epoch": 0.003439202199207062,
+      "grad_norm": 0.8594569563865662,
+      "learning_rate": 3.6891891891891894e-05,
+      "loss": 0.4712,
+      "step": 2631
+    },
+    {
+      "epoch": 0.0034405093836233323,
+      "grad_norm": 1.6882586479187012,
+      "learning_rate": 3.688688688688689e-05,
+      "loss": 0.9979,
+      "step": 2632
+    },
+    {
+      "epoch": 0.0034418165680396027,
+      "grad_norm": 1.4572855234146118,
+      "learning_rate": 3.6881881881881887e-05,
+      "loss": 1.4093,
+      "step": 2633
+    },
+    {
+      "epoch": 0.0034431237524558725,
+      "grad_norm": 1.7281708717346191,
+      "learning_rate": 3.687687687687688e-05,
+      "loss": 1.1234,
+      "step": 2634
+    },
+    {
+      "epoch": 0.003444430936872143,
+      "grad_norm": 1.3713233470916748,
+      "learning_rate": 3.687187187187187e-05,
+      "loss": 1.3957,
+      "step": 2635
+    },
+    {
+      "epoch": 0.003445738121288413,
+      "grad_norm": 1.2914820909500122,
+      "learning_rate": 3.6866866866866865e-05,
+      "loss": 0.9736,
+      "step": 2636
+    },
+    {
+      "epoch": 0.0034470453057046835,
+      "grad_norm": 1.169939398765564,
+      "learning_rate": 3.6861861861861865e-05,
+      "loss": 1.3383,
+      "step": 2637
+    },
+    {
+      "epoch": 0.003448352490120954,
+      "grad_norm": 2.135481119155884,
+      "learning_rate": 3.685685685685686e-05,
+      "loss": 1.7115,
+      "step": 2638
+    },
+    {
+      "epoch": 0.003449659674537224,
+      "grad_norm": 1.7390172481536865,
+      "learning_rate": 3.685185185185185e-05,
+      "loss": 1.3168,
+      "step": 2639
+    },
+    {
+      "epoch": 0.0034509668589534945,
+      "grad_norm": 1.8733125925064087,
+      "learning_rate": 3.684684684684685e-05,
+      "loss": 1.5531,
+      "step": 2640
+    },
+    {
+      "epoch": 0.0034522740433697648,
+      "grad_norm": 1.510364294052124,
+      "learning_rate": 3.6841841841841844e-05,
+      "loss": 1.5805,
+      "step": 2641
+    },
+    {
+      "epoch": 0.0034535812277860347,
+      "grad_norm": 1.436781644821167,
+      "learning_rate": 3.6836836836836844e-05,
+      "loss": 1.824,
+      "step": 2642
+    },
+    {
+      "epoch": 0.003454888412202305,
+      "grad_norm": 1.4862667322158813,
+      "learning_rate": 3.683183183183183e-05,
+      "loss": 1.6155,
+      "step": 2643
+    },
+    {
+      "epoch": 0.0034561955966185753,
+      "grad_norm": 1.3869256973266602,
+      "learning_rate": 3.682682682682683e-05,
+      "loss": 2.1824,
+      "step": 2644
+    },
+    {
+      "epoch": 0.0034575027810348456,
+      "grad_norm": 0.7538400292396545,
+      "learning_rate": 3.682182182182182e-05,
+      "loss": 0.4827,
+      "step": 2645
+    },
+    {
+      "epoch": 0.003458809965451116,
+      "grad_norm": 1.8622817993164062,
+      "learning_rate": 3.6816816816816816e-05,
+      "loss": 1.9395,
+      "step": 2646
+    },
+    {
+      "epoch": 0.0034601171498673863,
+      "grad_norm": 1.4432042837142944,
+      "learning_rate": 3.6811811811811815e-05,
+      "loss": 1.5478,
+      "step": 2647
+    },
+    {
+      "epoch": 0.0034614243342836566,
+      "grad_norm": 1.3654472827911377,
+      "learning_rate": 3.680680680680681e-05,
+      "loss": 1.3516,
+      "step": 2648
+    },
+    {
+      "epoch": 0.0034627315186999265,
+      "grad_norm": 1.2091203927993774,
+      "learning_rate": 3.680180180180181e-05,
+      "loss": 1.3675,
+      "step": 2649
+    },
+    {
+      "epoch": 0.003464038703116197,
+      "grad_norm": 1.437963604927063,
+      "learning_rate": 3.67967967967968e-05,
+      "loss": 1.8549,
+      "step": 2650
+    },
+    {
+      "epoch": 0.003465345887532467,
+      "grad_norm": 1.4133902788162231,
+      "learning_rate": 3.6791791791791794e-05,
+      "loss": 1.7605,
+      "step": 2651
+    },
+    {
+      "epoch": 0.0034666530719487374,
+      "grad_norm": 1.4858729839324951,
+      "learning_rate": 3.678678678678679e-05,
+      "loss": 1.5194,
+      "step": 2652
+    },
+    {
+      "epoch": 0.0034679602563650077,
+      "grad_norm": 1.954944372177124,
+      "learning_rate": 3.678178178178178e-05,
+      "loss": 2.347,
+      "step": 2653
+    },
+    {
+      "epoch": 0.003469267440781278,
+      "grad_norm": 1.2268645763397217,
+      "learning_rate": 3.677677677677678e-05,
+      "loss": 1.2957,
+      "step": 2654
+    },
+    {
+      "epoch": 0.0034705746251975484,
+      "grad_norm": 1.6797181367874146,
+      "learning_rate": 3.677177177177177e-05,
+      "loss": 2.0235,
+      "step": 2655
+    },
+    {
+      "epoch": 0.0034718818096138187,
+      "grad_norm": 1.3829026222229004,
+      "learning_rate": 3.676676676676677e-05,
+      "loss": 2.2154,
+      "step": 2656
+    },
+    {
+      "epoch": 0.0034731889940300886,
+      "grad_norm": 1.412377119064331,
+      "learning_rate": 3.6761761761761765e-05,
+      "loss": 1.7845,
+      "step": 2657
+    },
+    {
+      "epoch": 0.003474496178446359,
+      "grad_norm": 1.576741099357605,
+      "learning_rate": 3.675675675675676e-05,
+      "loss": 1.8147,
+      "step": 2658
+    },
+    {
+      "epoch": 0.0034758033628626292,
+      "grad_norm": 1.5428481101989746,
+      "learning_rate": 3.675175175175176e-05,
+      "loss": 1.9701,
+      "step": 2659
+    },
+    {
+      "epoch": 0.0034771105472788996,
+      "grad_norm": 1.3539575338363647,
+      "learning_rate": 3.6746746746746744e-05,
+      "loss": 1.3606,
+      "step": 2660
+    },
+    {
+      "epoch": 0.00347841773169517,
+      "grad_norm": 1.4454861879348755,
+      "learning_rate": 3.6741741741741744e-05,
+      "loss": 1.2304,
+      "step": 2661
+    },
+    {
+      "epoch": 0.00347972491611144,
+      "grad_norm": 1.478310227394104,
+      "learning_rate": 3.673673673673674e-05,
+      "loss": 0.6951,
+      "step": 2662
+    },
+    {
+      "epoch": 0.0034810321005277105,
+      "grad_norm": 1.3475639820098877,
+      "learning_rate": 3.673173173173173e-05,
+      "loss": 1.4297,
+      "step": 2663
+    },
+    {
+      "epoch": 0.0034823392849439804,
+      "grad_norm": 1.413616418838501,
+      "learning_rate": 3.672672672672673e-05,
+      "loss": 1.6689,
+      "step": 2664
+    },
+    {
+      "epoch": 0.0034836464693602507,
+      "grad_norm": 1.2733666896820068,
+      "learning_rate": 3.672172172172172e-05,
+      "loss": 1.8637,
+      "step": 2665
+    },
+    {
+      "epoch": 0.003484953653776521,
+      "grad_norm": 1.2959364652633667,
+      "learning_rate": 3.671671671671672e-05,
+      "loss": 1.613,
+      "step": 2666
+    },
+    {
+      "epoch": 0.0034862608381927914,
+      "grad_norm": 1.3828407526016235,
+      "learning_rate": 3.6711711711711715e-05,
+      "loss": 1.6525,
+      "step": 2667
+    },
+    {
+      "epoch": 0.0034875680226090617,
+      "grad_norm": 1.0925780534744263,
+      "learning_rate": 3.670670670670671e-05,
+      "loss": 0.7482,
+      "step": 2668
+    },
+    {
+      "epoch": 0.003488875207025332,
+      "grad_norm": 1.2582857608795166,
+      "learning_rate": 3.67017017017017e-05,
+      "loss": 1.2766,
+      "step": 2669
+    },
+    {
+      "epoch": 0.0034901823914416023,
+      "grad_norm": 1.250939130783081,
+      "learning_rate": 3.6696696696696694e-05,
+      "loss": 0.6919,
+      "step": 2670
+    },
+    {
+      "epoch": 0.0034914895758578726,
+      "grad_norm": 1.37144935131073,
+      "learning_rate": 3.6691691691691694e-05,
+      "loss": 1.281,
+      "step": 2671
+    },
+    {
+      "epoch": 0.0034927967602741425,
+      "grad_norm": 1.5297462940216064,
+      "learning_rate": 3.668668668668669e-05,
+      "loss": 0.913,
+      "step": 2672
+    },
+    {
+      "epoch": 0.003494103944690413,
+      "grad_norm": 2.2115800380706787,
+      "learning_rate": 3.668168168168169e-05,
+      "loss": 2.1554,
+      "step": 2673
+    },
+    {
+      "epoch": 0.003495411129106683,
+      "grad_norm": 1.4597219228744507,
+      "learning_rate": 3.667667667667668e-05,
+      "loss": 1.5685,
+      "step": 2674
+    },
+    {
+      "epoch": 0.0034967183135229535,
+      "grad_norm": 1.1744719743728638,
+      "learning_rate": 3.667167167167167e-05,
+      "loss": 1.3026,
+      "step": 2675
+    },
+    {
+      "epoch": 0.003498025497939224,
+      "grad_norm": 1.484411358833313,
+      "learning_rate": 3.6666666666666666e-05,
+      "loss": 1.8545,
+      "step": 2676
+    },
+    {
+      "epoch": 0.003499332682355494,
+      "grad_norm": 1.54093337059021,
+      "learning_rate": 3.666166166166166e-05,
+      "loss": 1.5979,
+      "step": 2677
+    },
+    {
+      "epoch": 0.0035006398667717644,
+      "grad_norm": 1.6028257608413696,
+      "learning_rate": 3.665665665665666e-05,
+      "loss": 1.2319,
+      "step": 2678
+    },
+    {
+      "epoch": 0.0035019470511880348,
+      "grad_norm": 1.3732366561889648,
+      "learning_rate": 3.665165165165165e-05,
+      "loss": 1.6199,
+      "step": 2679
+    },
+    {
+      "epoch": 0.0035032542356043046,
+      "grad_norm": 1.4588109254837036,
+      "learning_rate": 3.664664664664665e-05,
+      "loss": 1.7777,
+      "step": 2680
+    },
+    {
+      "epoch": 0.003504561420020575,
+      "grad_norm": 1.7026045322418213,
+      "learning_rate": 3.6641641641641644e-05,
+      "loss": 2.4748,
+      "step": 2681
+    },
+    {
+      "epoch": 0.0035058686044368453,
+      "grad_norm": 1.0623811483383179,
+      "learning_rate": 3.663663663663664e-05,
+      "loss": 1.4156,
+      "step": 2682
+    },
+    {
+      "epoch": 0.0035071757888531156,
+      "grad_norm": 1.321517825126648,
+      "learning_rate": 3.663163163163164e-05,
+      "loss": 1.2143,
+      "step": 2683
+    },
+    {
+      "epoch": 0.003508482973269386,
+      "grad_norm": 1.692028522491455,
+      "learning_rate": 3.662662662662663e-05,
+      "loss": 1.3013,
+      "step": 2684
+    },
+    {
+      "epoch": 0.0035097901576856562,
+      "grad_norm": 1.2708861827850342,
+      "learning_rate": 3.662162162162162e-05,
+      "loss": 1.8202,
+      "step": 2685
+    },
+    {
+      "epoch": 0.0035110973421019266,
+      "grad_norm": 1.907790184020996,
+      "learning_rate": 3.6616616616616616e-05,
+      "loss": 2.3492,
+      "step": 2686
+    },
+    {
+      "epoch": 0.0035124045265181965,
+      "grad_norm": 1.4270598888397217,
+      "learning_rate": 3.6611611611611616e-05,
+      "loss": 1.7876,
+      "step": 2687
+    },
+    {
+      "epoch": 0.0035137117109344668,
+      "grad_norm": 1.4862713813781738,
+      "learning_rate": 3.660660660660661e-05,
+      "loss": 1.6357,
+      "step": 2688
+    },
+    {
+      "epoch": 0.003515018895350737,
+      "grad_norm": 1.1964284181594849,
+      "learning_rate": 3.66016016016016e-05,
+      "loss": 1.699,
+      "step": 2689
+    },
+    {
+      "epoch": 0.0035163260797670074,
+      "grad_norm": 1.6031752824783325,
+      "learning_rate": 3.65965965965966e-05,
+      "loss": 1.6452,
+      "step": 2690
+    },
+    {
+      "epoch": 0.0035176332641832777,
+      "grad_norm": 1.1613372564315796,
+      "learning_rate": 3.6591591591591594e-05,
+      "loss": 1.3814,
+      "step": 2691
+    },
+    {
+      "epoch": 0.003518940448599548,
+      "grad_norm": 1.4869496822357178,
+      "learning_rate": 3.6586586586586594e-05,
+      "loss": 2.1448,
+      "step": 2692
+    },
+    {
+      "epoch": 0.0035202476330158184,
+      "grad_norm": 1.744122862815857,
+      "learning_rate": 3.658158158158158e-05,
+      "loss": 2.3145,
+      "step": 2693
+    },
+    {
+      "epoch": 0.0035215548174320887,
+      "grad_norm": 1.8008646965026855,
+      "learning_rate": 3.657657657657658e-05,
+      "loss": 1.8546,
+      "step": 2694
+    },
+    {
+      "epoch": 0.0035228620018483586,
+      "grad_norm": 1.1975582838058472,
+      "learning_rate": 3.657157157157157e-05,
+      "loss": 0.9121,
+      "step": 2695
+    },
+    {
+      "epoch": 0.003524169186264629,
+      "grad_norm": 1.5049877166748047,
+      "learning_rate": 3.6566566566566566e-05,
+      "loss": 2.283,
+      "step": 2696
+    },
+    {
+      "epoch": 0.003525476370680899,
+      "grad_norm": 1.2927007675170898,
+      "learning_rate": 3.6561561561561566e-05,
+      "loss": 1.2643,
+      "step": 2697
+    },
+    {
+      "epoch": 0.0035267835550971695,
+      "grad_norm": 1.5404701232910156,
+      "learning_rate": 3.655655655655656e-05,
+      "loss": 2.034,
+      "step": 2698
+    },
+    {
+      "epoch": 0.00352809073951344,
+      "grad_norm": 1.2446562051773071,
+      "learning_rate": 3.655155155155156e-05,
+      "loss": 1.5165,
+      "step": 2699
+    },
+    {
+      "epoch": 0.00352939792392971,
+      "grad_norm": 1.0212162733078003,
+      "learning_rate": 3.654654654654655e-05,
+      "loss": 1.2044,
+      "step": 2700
+    },
+    {
+      "epoch": 0.0035307051083459805,
+      "grad_norm": 0.6918138861656189,
+      "learning_rate": 3.654154154154154e-05,
+      "loss": 0.2673,
+      "step": 2701
+    },
+    {
+      "epoch": 0.003532012292762251,
+      "grad_norm": 1.3889102935791016,
+      "learning_rate": 3.653653653653654e-05,
+      "loss": 1.4752,
+      "step": 2702
+    },
+    {
+      "epoch": 0.0035333194771785207,
+      "grad_norm": 1.3112328052520752,
+      "learning_rate": 3.653153153153153e-05,
+      "loss": 1.5286,
+      "step": 2703
+    },
+    {
+      "epoch": 0.003534626661594791,
+      "grad_norm": 1.3342585563659668,
+      "learning_rate": 3.652652652652653e-05,
+      "loss": 1.82,
+      "step": 2704
+    },
+    {
+      "epoch": 0.0035359338460110613,
+      "grad_norm": 1.5786573886871338,
+      "learning_rate": 3.652152152152152e-05,
+      "loss": 2.0618,
+      "step": 2705
+    },
+    {
+      "epoch": 0.0035372410304273317,
+      "grad_norm": 1.3687831163406372,
+      "learning_rate": 3.6516516516516516e-05,
+      "loss": 1.3909,
+      "step": 2706
+    },
+    {
+      "epoch": 0.003538548214843602,
+      "grad_norm": 0.765046238899231,
+      "learning_rate": 3.6511511511511516e-05,
+      "loss": 0.3056,
+      "step": 2707
+    },
+    {
+      "epoch": 0.0035398553992598723,
+      "grad_norm": 1.2466877698898315,
+      "learning_rate": 3.650650650650651e-05,
+      "loss": 1.6116,
+      "step": 2708
+    },
+    {
+      "epoch": 0.0035411625836761426,
+      "grad_norm": 1.3500930070877075,
+      "learning_rate": 3.65015015015015e-05,
+      "loss": 2.1569,
+      "step": 2709
+    },
+    {
+      "epoch": 0.0035424697680924125,
+      "grad_norm": 1.6886346340179443,
+      "learning_rate": 3.6496496496496495e-05,
+      "loss": 2.0208,
+      "step": 2710
+    },
+    {
+      "epoch": 0.003543776952508683,
+      "grad_norm": 1.1580898761749268,
+      "learning_rate": 3.6491491491491494e-05,
+      "loss": 1.0356,
+      "step": 2711
+    },
+    {
+      "epoch": 0.003545084136924953,
+      "grad_norm": 1.8278592824935913,
+      "learning_rate": 3.648648648648649e-05,
+      "loss": 1.4979,
+      "step": 2712
+    },
+    {
+      "epoch": 0.0035463913213412235,
+      "grad_norm": 1.3436801433563232,
+      "learning_rate": 3.648148148148148e-05,
+      "loss": 1.5156,
+      "step": 2713
+    },
+    {
+      "epoch": 0.0035476985057574938,
+      "grad_norm": 1.5520777702331543,
+      "learning_rate": 3.647647647647648e-05,
+      "loss": 1.7169,
+      "step": 2714
+    },
+    {
+      "epoch": 0.003549005690173764,
+      "grad_norm": 1.1051979064941406,
+      "learning_rate": 3.647147147147147e-05,
+      "loss": 1.486,
+      "step": 2715
+    },
+    {
+      "epoch": 0.0035503128745900344,
+      "grad_norm": 1.2409307956695557,
+      "learning_rate": 3.646646646646647e-05,
+      "loss": 1.1388,
+      "step": 2716
+    },
+    {
+      "epoch": 0.0035516200590063047,
+      "grad_norm": 1.4446899890899658,
+      "learning_rate": 3.6461461461461466e-05,
+      "loss": 1.5068,
+      "step": 2717
+    },
+    {
+      "epoch": 0.0035529272434225746,
+      "grad_norm": 1.3706861734390259,
+      "learning_rate": 3.645645645645646e-05,
+      "loss": 2.3063,
+      "step": 2718
+    },
+    {
+      "epoch": 0.003554234427838845,
+      "grad_norm": 1.4194670915603638,
+      "learning_rate": 3.645145145145145e-05,
+      "loss": 0.9802,
+      "step": 2719
+    },
+    {
+      "epoch": 0.0035555416122551153,
+      "grad_norm": 1.2788985967636108,
+      "learning_rate": 3.6446446446446445e-05,
+      "loss": 2.1441,
+      "step": 2720
+    },
+    {
+      "epoch": 0.0035568487966713856,
+      "grad_norm": 1.4382998943328857,
+      "learning_rate": 3.6441441441441444e-05,
+      "loss": 1.6782,
+      "step": 2721
+    },
+    {
+      "epoch": 0.003558155981087656,
+      "grad_norm": 1.1949737071990967,
+      "learning_rate": 3.643643643643644e-05,
+      "loss": 1.4638,
+      "step": 2722
+    },
+    {
+      "epoch": 0.0035594631655039262,
+      "grad_norm": 1.1757441759109497,
+      "learning_rate": 3.643143143143144e-05,
+      "loss": 0.9753,
+      "step": 2723
+    },
+    {
+      "epoch": 0.0035607703499201965,
+      "grad_norm": 1.1326805353164673,
+      "learning_rate": 3.642642642642643e-05,
+      "loss": 1.2462,
+      "step": 2724
+    },
+    {
+      "epoch": 0.003562077534336467,
+      "grad_norm": 1.3143575191497803,
+      "learning_rate": 3.642142142142142e-05,
+      "loss": 2.2638,
+      "step": 2725
+    },
+    {
+      "epoch": 0.0035633847187527368,
+      "grad_norm": 1.3827975988388062,
+      "learning_rate": 3.6416416416416416e-05,
+      "loss": 1.6345,
+      "step": 2726
+    },
+    {
+      "epoch": 0.003564691903169007,
+      "grad_norm": 1.6178183555603027,
+      "learning_rate": 3.641141141141141e-05,
+      "loss": 1.6856,
+      "step": 2727
+    },
+    {
+      "epoch": 0.0035659990875852774,
+      "grad_norm": 1.2727773189544678,
+      "learning_rate": 3.640640640640641e-05,
+      "loss": 2.107,
+      "step": 2728
+    },
+    {
+      "epoch": 0.0035673062720015477,
+      "grad_norm": 1.663008213043213,
+      "learning_rate": 3.64014014014014e-05,
+      "loss": 1.4046,
+      "step": 2729
+    },
+    {
+      "epoch": 0.003568613456417818,
+      "grad_norm": 1.891373872756958,
+      "learning_rate": 3.63963963963964e-05,
+      "loss": 2.3615,
+      "step": 2730
+    },
+    {
+      "epoch": 0.0035699206408340883,
+      "grad_norm": 1.1734168529510498,
+      "learning_rate": 3.6391391391391394e-05,
+      "loss": 1.5666,
+      "step": 2731
+    },
+    {
+      "epoch": 0.0035712278252503587,
+      "grad_norm": 1.3985590934753418,
+      "learning_rate": 3.638638638638639e-05,
+      "loss": 1.9032,
+      "step": 2732
+    },
+    {
+      "epoch": 0.0035725350096666286,
+      "grad_norm": 1.5874643325805664,
+      "learning_rate": 3.638138138138139e-05,
+      "loss": 2.4631,
+      "step": 2733
+    },
+    {
+      "epoch": 0.003573842194082899,
+      "grad_norm": 1.22110116481781,
+      "learning_rate": 3.6376376376376373e-05,
+      "loss": 1.1453,
+      "step": 2734
+    },
+    {
+      "epoch": 0.003575149378499169,
+      "grad_norm": 1.2256441116333008,
+      "learning_rate": 3.637137137137137e-05,
+      "loss": 1.534,
+      "step": 2735
+    },
+    {
+      "epoch": 0.0035764565629154395,
+      "grad_norm": 0.8718202114105225,
+      "learning_rate": 3.6366366366366366e-05,
+      "loss": 0.775,
+      "step": 2736
+    },
+    {
+      "epoch": 0.00357776374733171,
+      "grad_norm": 1.309079885482788,
+      "learning_rate": 3.636136136136136e-05,
+      "loss": 1.7057,
+      "step": 2737
+    },
+    {
+      "epoch": 0.00357907093174798,
+      "grad_norm": 1.9876086711883545,
+      "learning_rate": 3.635635635635636e-05,
+      "loss": 1.9161,
+      "step": 2738
+    },
+    {
+      "epoch": 0.0035803781161642505,
+      "grad_norm": 1.2189862728118896,
+      "learning_rate": 3.635135135135135e-05,
+      "loss": 1.2285,
+      "step": 2739
+    },
+    {
+      "epoch": 0.003581685300580521,
+      "grad_norm": 1.5076738595962524,
+      "learning_rate": 3.634634634634635e-05,
+      "loss": 2.0042,
+      "step": 2740
+    },
+    {
+      "epoch": 0.0035829924849967907,
+      "grad_norm": 1.1978598833084106,
+      "learning_rate": 3.6341341341341345e-05,
+      "loss": 1.579,
+      "step": 2741
+    },
+    {
+      "epoch": 0.003584299669413061,
+      "grad_norm": 1.1972904205322266,
+      "learning_rate": 3.633633633633634e-05,
+      "loss": 1.0246,
+      "step": 2742
+    },
+    {
+      "epoch": 0.0035856068538293313,
+      "grad_norm": 1.270995855331421,
+      "learning_rate": 3.633133133133133e-05,
+      "loss": 1.4218,
+      "step": 2743
+    },
+    {
+      "epoch": 0.0035869140382456016,
+      "grad_norm": 1.7138490676879883,
+      "learning_rate": 3.6326326326326323e-05,
+      "loss": 1.6782,
+      "step": 2744
+    },
+    {
+      "epoch": 0.003588221222661872,
+      "grad_norm": 1.25938880443573,
+      "learning_rate": 3.632132132132132e-05,
+      "loss": 1.6351,
+      "step": 2745
+    },
+    {
+      "epoch": 0.0035895284070781423,
+      "grad_norm": 1.6466416120529175,
+      "learning_rate": 3.6316316316316316e-05,
+      "loss": 2.3664,
+      "step": 2746
+    },
+    {
+      "epoch": 0.0035908355914944126,
+      "grad_norm": 1.4402780532836914,
+      "learning_rate": 3.6311311311311316e-05,
+      "loss": 1.0121,
+      "step": 2747
+    },
+    {
+      "epoch": 0.003592142775910683,
+      "grad_norm": 1.7264593839645386,
+      "learning_rate": 3.630630630630631e-05,
+      "loss": 1.7267,
+      "step": 2748
+    },
+    {
+      "epoch": 0.003593449960326953,
+      "grad_norm": 1.3810210227966309,
+      "learning_rate": 3.63013013013013e-05,
+      "loss": 0.8964,
+      "step": 2749
+    },
+    {
+      "epoch": 0.003594757144743223,
+      "grad_norm": 1.1196056604385376,
+      "learning_rate": 3.62962962962963e-05,
+      "loss": 1.3622,
+      "step": 2750
+    },
+    {
+      "epoch": 0.0035960643291594934,
+      "grad_norm": 1.2765989303588867,
+      "learning_rate": 3.629129129129129e-05,
+      "loss": 1.3108,
+      "step": 2751
+    },
+    {
+      "epoch": 0.0035973715135757638,
+      "grad_norm": 1.0934408903121948,
+      "learning_rate": 3.628628628628629e-05,
+      "loss": 0.7677,
+      "step": 2752
+    },
+    {
+      "epoch": 0.003598678697992034,
+      "grad_norm": 1.5729115009307861,
+      "learning_rate": 3.628128128128128e-05,
+      "loss": 1.6689,
+      "step": 2753
+    },
+    {
+      "epoch": 0.0035999858824083044,
+      "grad_norm": 1.9439693689346313,
+      "learning_rate": 3.627627627627628e-05,
+      "loss": 1.9945,
+      "step": 2754
+    },
+    {
+      "epoch": 0.0036012930668245747,
+      "grad_norm": 1.241098165512085,
+      "learning_rate": 3.627127127127127e-05,
+      "loss": 1.37,
+      "step": 2755
+    },
+    {
+      "epoch": 0.0036026002512408446,
+      "grad_norm": 0.9213352203369141,
+      "learning_rate": 3.6266266266266266e-05,
+      "loss": 0.6678,
+      "step": 2756
+    },
+    {
+      "epoch": 0.003603907435657115,
+      "grad_norm": 1.3839374780654907,
+      "learning_rate": 3.6261261261261266e-05,
+      "loss": 2.1659,
+      "step": 2757
+    },
+    {
+      "epoch": 0.0036052146200733852,
+      "grad_norm": 1.65111243724823,
+      "learning_rate": 3.625625625625626e-05,
+      "loss": 1.5738,
+      "step": 2758
+    },
+    {
+      "epoch": 0.0036065218044896556,
+      "grad_norm": 1.4861472845077515,
+      "learning_rate": 3.625125125125125e-05,
+      "loss": 1.9727,
+      "step": 2759
+    },
+    {
+      "epoch": 0.003607828988905926,
+      "grad_norm": 1.513171911239624,
+      "learning_rate": 3.6246246246246245e-05,
+      "loss": 1.5101,
+      "step": 2760
+    },
+    {
+      "epoch": 0.003609136173322196,
+      "grad_norm": 1.6812933683395386,
+      "learning_rate": 3.6241241241241245e-05,
+      "loss": 1.2803,
+      "step": 2761
+    },
+    {
+      "epoch": 0.0036104433577384665,
+      "grad_norm": 1.1660507917404175,
+      "learning_rate": 3.623623623623624e-05,
+      "loss": 1.5766,
+      "step": 2762
+    },
+    {
+      "epoch": 0.003611750542154737,
+      "grad_norm": 1.432906150817871,
+      "learning_rate": 3.623123123123123e-05,
+      "loss": 1.6748,
+      "step": 2763
+    },
+    {
+      "epoch": 0.0036130577265710067,
+      "grad_norm": 1.5281808376312256,
+      "learning_rate": 3.622622622622623e-05,
+      "loss": 2.4214,
+      "step": 2764
+    },
+    {
+      "epoch": 0.003614364910987277,
+      "grad_norm": 1.2281867265701294,
+      "learning_rate": 3.622122122122122e-05,
+      "loss": 1.9802,
+      "step": 2765
+    },
+    {
+      "epoch": 0.0036156720954035474,
+      "grad_norm": 1.4438848495483398,
+      "learning_rate": 3.621621621621622e-05,
+      "loss": 1.9872,
+      "step": 2766
+    },
+    {
+      "epoch": 0.0036169792798198177,
+      "grad_norm": 1.3221749067306519,
+      "learning_rate": 3.6211211211211216e-05,
+      "loss": 1.5433,
+      "step": 2767
+    },
+    {
+      "epoch": 0.003618286464236088,
+      "grad_norm": 1.3427280187606812,
+      "learning_rate": 3.620620620620621e-05,
+      "loss": 2.0838,
+      "step": 2768
+    },
+    {
+      "epoch": 0.0036195936486523583,
+      "grad_norm": 0.9607228636741638,
+      "learning_rate": 3.62012012012012e-05,
+      "loss": 0.5642,
+      "step": 2769
+    },
+    {
+      "epoch": 0.0036209008330686287,
+      "grad_norm": 1.0648623704910278,
+      "learning_rate": 3.6196196196196195e-05,
+      "loss": 0.7764,
+      "step": 2770
+    },
+    {
+      "epoch": 0.0036222080174848985,
+      "grad_norm": 1.641813039779663,
+      "learning_rate": 3.6191191191191195e-05,
+      "loss": 1.4402,
+      "step": 2771
+    },
+    {
+      "epoch": 0.003623515201901169,
+      "grad_norm": 1.8091514110565186,
+      "learning_rate": 3.618618618618619e-05,
+      "loss": 1.7152,
+      "step": 2772
+    },
+    {
+      "epoch": 0.003624822386317439,
+      "grad_norm": 0.9133872389793396,
+      "learning_rate": 3.618118118118118e-05,
+      "loss": 0.7042,
+      "step": 2773
+    },
+    {
+      "epoch": 0.0036261295707337095,
+      "grad_norm": 1.1771806478500366,
+      "learning_rate": 3.617617617617618e-05,
+      "loss": 1.3096,
+      "step": 2774
+    },
+    {
+      "epoch": 0.00362743675514998,
+      "grad_norm": 1.4197189807891846,
+      "learning_rate": 3.6171171171171173e-05,
+      "loss": 1.7609,
+      "step": 2775
+    },
+    {
+      "epoch": 0.00362874393956625,
+      "grad_norm": 1.3153395652770996,
+      "learning_rate": 3.6166166166166166e-05,
+      "loss": 1.5192,
+      "step": 2776
+    },
+    {
+      "epoch": 0.0036300511239825205,
+      "grad_norm": 1.0451775789260864,
+      "learning_rate": 3.616116116116116e-05,
+      "loss": 0.6944,
+      "step": 2777
+    },
+    {
+      "epoch": 0.0036313583083987908,
+      "grad_norm": 1.5979881286621094,
+      "learning_rate": 3.615615615615616e-05,
+      "loss": 1.407,
+      "step": 2778
+    },
+    {
+      "epoch": 0.0036326654928150607,
+      "grad_norm": 1.4447836875915527,
+      "learning_rate": 3.615115115115115e-05,
+      "loss": 2.0622,
+      "step": 2779
+    },
+    {
+      "epoch": 0.003633972677231331,
+      "grad_norm": 1.6700429916381836,
+      "learning_rate": 3.6146146146146145e-05,
+      "loss": 1.6142,
+      "step": 2780
+    },
+    {
+      "epoch": 0.0036352798616476013,
+      "grad_norm": 1.444746732711792,
+      "learning_rate": 3.6141141141141145e-05,
+      "loss": 1.8693,
+      "step": 2781
+    },
+    {
+      "epoch": 0.0036365870460638716,
+      "grad_norm": 1.4580326080322266,
+      "learning_rate": 3.613613613613614e-05,
+      "loss": 1.7837,
+      "step": 2782
+    },
+    {
+      "epoch": 0.003637894230480142,
+      "grad_norm": 1.6224457025527954,
+      "learning_rate": 3.613113113113114e-05,
+      "loss": 1.2407,
+      "step": 2783
+    },
+    {
+      "epoch": 0.0036392014148964123,
+      "grad_norm": 1.0638720989227295,
+      "learning_rate": 3.6126126126126124e-05,
+      "loss": 1.3155,
+      "step": 2784
+    },
+    {
+      "epoch": 0.0036405085993126826,
+      "grad_norm": 1.4752346277236938,
+      "learning_rate": 3.6121121121121124e-05,
+      "loss": 1.9314,
+      "step": 2785
+    },
+    {
+      "epoch": 0.003641815783728953,
+      "grad_norm": 1.3079760074615479,
+      "learning_rate": 3.6116116116116116e-05,
+      "loss": 2.2984,
+      "step": 2786
+    },
+    {
+      "epoch": 0.003643122968145223,
+      "grad_norm": 1.5435187816619873,
+      "learning_rate": 3.611111111111111e-05,
+      "loss": 1.1026,
+      "step": 2787
+    },
+    {
+      "epoch": 0.003644430152561493,
+      "grad_norm": 1.0407167673110962,
+      "learning_rate": 3.610610610610611e-05,
+      "loss": 1.2093,
+      "step": 2788
+    },
+    {
+      "epoch": 0.0036457373369777634,
+      "grad_norm": 1.2366989850997925,
+      "learning_rate": 3.61011011011011e-05,
+      "loss": 1.6081,
+      "step": 2789
+    },
+    {
+      "epoch": 0.0036470445213940337,
+      "grad_norm": 1.3279781341552734,
+      "learning_rate": 3.60960960960961e-05,
+      "loss": 2.413,
+      "step": 2790
+    },
+    {
+      "epoch": 0.003648351705810304,
+      "grad_norm": 1.4654514789581299,
+      "learning_rate": 3.6091091091091095e-05,
+      "loss": 2.3588,
+      "step": 2791
+    },
+    {
+      "epoch": 0.0036496588902265744,
+      "grad_norm": 1.2349086999893188,
+      "learning_rate": 3.608608608608609e-05,
+      "loss": 2.0003,
+      "step": 2792
+    },
+    {
+      "epoch": 0.0036509660746428447,
+      "grad_norm": 1.4695086479187012,
+      "learning_rate": 3.608108108108108e-05,
+      "loss": 1.8307,
+      "step": 2793
+    },
+    {
+      "epoch": 0.0036522732590591146,
+      "grad_norm": 1.233309268951416,
+      "learning_rate": 3.6076076076076074e-05,
+      "loss": 1.8326,
+      "step": 2794
+    },
+    {
+      "epoch": 0.003653580443475385,
+      "grad_norm": 1.3636963367462158,
+      "learning_rate": 3.6071071071071074e-05,
+      "loss": 1.708,
+      "step": 2795
+    },
+    {
+      "epoch": 0.0036548876278916552,
+      "grad_norm": 1.919586420059204,
+      "learning_rate": 3.6066066066066067e-05,
+      "loss": 1.4533,
+      "step": 2796
+    },
+    {
+      "epoch": 0.0036561948123079255,
+      "grad_norm": 1.2418015003204346,
+      "learning_rate": 3.6061061061061066e-05,
+      "loss": 1.5807,
+      "step": 2797
+    },
+    {
+      "epoch": 0.003657501996724196,
+      "grad_norm": 1.5402629375457764,
+      "learning_rate": 3.605605605605606e-05,
+      "loss": 1.5292,
+      "step": 2798
+    },
+    {
+      "epoch": 0.003658809181140466,
+      "grad_norm": 1.3969857692718506,
+      "learning_rate": 3.605105105105105e-05,
+      "loss": 1.9023,
+      "step": 2799
+    },
+    {
+      "epoch": 0.0036601163655567365,
+      "grad_norm": 1.4541535377502441,
+      "learning_rate": 3.604604604604605e-05,
+      "loss": 0.908,
+      "step": 2800
+    },
+    {
+      "epoch": 0.003661423549973007,
+      "grad_norm": 1.1810957193374634,
+      "learning_rate": 3.604104104104104e-05,
+      "loss": 1.1768,
+      "step": 2801
+    },
+    {
+      "epoch": 0.0036627307343892767,
+      "grad_norm": 1.2556079626083374,
+      "learning_rate": 3.603603603603604e-05,
+      "loss": 1.2211,
+      "step": 2802
+    },
+    {
+      "epoch": 0.003664037918805547,
+      "grad_norm": 1.527541995048523,
+      "learning_rate": 3.603103103103103e-05,
+      "loss": 1.0255,
+      "step": 2803
+    },
+    {
+      "epoch": 0.0036653451032218174,
+      "grad_norm": 1.1544946432113647,
+      "learning_rate": 3.602602602602603e-05,
+      "loss": 0.9866,
+      "step": 2804
+    },
+    {
+      "epoch": 0.0036666522876380877,
+      "grad_norm": 1.3906079530715942,
+      "learning_rate": 3.6021021021021024e-05,
+      "loss": 1.6058,
+      "step": 2805
+    },
+    {
+      "epoch": 0.003667959472054358,
+      "grad_norm": 1.4735944271087646,
+      "learning_rate": 3.601601601601602e-05,
+      "loss": 1.331,
+      "step": 2806
+    },
+    {
+      "epoch": 0.0036692666564706283,
+      "grad_norm": 2.028404951095581,
+      "learning_rate": 3.6011011011011016e-05,
+      "loss": 1.7958,
+      "step": 2807
+    },
+    {
+      "epoch": 0.0036705738408868986,
+      "grad_norm": 1.1880191564559937,
+      "learning_rate": 3.600600600600601e-05,
+      "loss": 1.7556,
+      "step": 2808
+    },
+    {
+      "epoch": 0.003671881025303169,
+      "grad_norm": 1.5469419956207275,
+      "learning_rate": 3.6001001001001e-05,
+      "loss": 1.8183,
+      "step": 2809
+    },
+    {
+      "epoch": 0.003673188209719439,
+      "grad_norm": 1.3799777030944824,
+      "learning_rate": 3.5995995995995995e-05,
+      "loss": 1.1578,
+      "step": 2810
+    },
+    {
+      "epoch": 0.003674495394135709,
+      "grad_norm": 1.8059953451156616,
+      "learning_rate": 3.599099099099099e-05,
+      "loss": 2.4007,
+      "step": 2811
+    },
+    {
+      "epoch": 0.0036758025785519795,
+      "grad_norm": 0.8728529214859009,
+      "learning_rate": 3.598598598598599e-05,
+      "loss": 0.446,
+      "step": 2812
+    },
+    {
+      "epoch": 0.00367710976296825,
+      "grad_norm": 1.1514288187026978,
+      "learning_rate": 3.598098098098098e-05,
+      "loss": 1.115,
+      "step": 2813
+    },
+    {
+      "epoch": 0.00367841694738452,
+      "grad_norm": 1.515260934829712,
+      "learning_rate": 3.597597597597598e-05,
+      "loss": 0.9303,
+      "step": 2814
+    },
+    {
+      "epoch": 0.0036797241318007904,
+      "grad_norm": 1.3464422225952148,
+      "learning_rate": 3.5970970970970974e-05,
+      "loss": 2.0264,
+      "step": 2815
+    },
+    {
+      "epoch": 0.0036810313162170608,
+      "grad_norm": 1.3606653213500977,
+      "learning_rate": 3.596596596596597e-05,
+      "loss": 1.3713,
+      "step": 2816
+    },
+    {
+      "epoch": 0.0036823385006333306,
+      "grad_norm": 1.72804594039917,
+      "learning_rate": 3.5960960960960966e-05,
+      "loss": 1.8349,
+      "step": 2817
+    },
+    {
+      "epoch": 0.003683645685049601,
+      "grad_norm": 1.645738124847412,
+      "learning_rate": 3.595595595595595e-05,
+      "loss": 1.5562,
+      "step": 2818
+    },
+    {
+      "epoch": 0.0036849528694658713,
+      "grad_norm": 1.2921960353851318,
+      "learning_rate": 3.595095095095095e-05,
+      "loss": 1.9368,
+      "step": 2819
+    },
+    {
+      "epoch": 0.0036862600538821416,
+      "grad_norm": 0.7433218359947205,
+      "learning_rate": 3.5945945945945945e-05,
+      "loss": 0.411,
+      "step": 2820
+    },
+    {
+      "epoch": 0.003687567238298412,
+      "grad_norm": 1.2635174989700317,
+      "learning_rate": 3.5940940940940945e-05,
+      "loss": 1.3199,
+      "step": 2821
+    },
+    {
+      "epoch": 0.0036888744227146822,
+      "grad_norm": 1.3492826223373413,
+      "learning_rate": 3.593593593593594e-05,
+      "loss": 1.7473,
+      "step": 2822
+    },
+    {
+      "epoch": 0.0036901816071309526,
+      "grad_norm": 1.0525577068328857,
+      "learning_rate": 3.593093093093093e-05,
+      "loss": 0.7882,
+      "step": 2823
+    },
+    {
+      "epoch": 0.003691488791547223,
+      "grad_norm": 1.1368989944458008,
+      "learning_rate": 3.592592592592593e-05,
+      "loss": 0.5619,
+      "step": 2824
+    },
+    {
+      "epoch": 0.0036927959759634928,
+      "grad_norm": 1.4940296411514282,
+      "learning_rate": 3.5920920920920924e-05,
+      "loss": 1.8725,
+      "step": 2825
+    },
+    {
+      "epoch": 0.003694103160379763,
+      "grad_norm": 1.6705666780471802,
+      "learning_rate": 3.591591591591592e-05,
+      "loss": 1.8083,
+      "step": 2826
+    },
+    {
+      "epoch": 0.0036954103447960334,
+      "grad_norm": 1.4401953220367432,
+      "learning_rate": 3.591091091091091e-05,
+      "loss": 2.0942,
+      "step": 2827
+    },
+    {
+      "epoch": 0.0036967175292123037,
+      "grad_norm": 1.499419927597046,
+      "learning_rate": 3.590590590590591e-05,
+      "loss": 1.9172,
+      "step": 2828
+    },
+    {
+      "epoch": 0.003698024713628574,
+      "grad_norm": 1.3575751781463623,
+      "learning_rate": 3.59009009009009e-05,
+      "loss": 1.9664,
+      "step": 2829
+    },
+    {
+      "epoch": 0.0036993318980448444,
+      "grad_norm": 1.7696120738983154,
+      "learning_rate": 3.5895895895895895e-05,
+      "loss": 2.1035,
+      "step": 2830
+    },
+    {
+      "epoch": 0.0037006390824611147,
+      "grad_norm": 1.2392507791519165,
+      "learning_rate": 3.5890890890890895e-05,
+      "loss": 2.0127,
+      "step": 2831
+    },
+    {
+      "epoch": 0.003701946266877385,
+      "grad_norm": 2.090320110321045,
+      "learning_rate": 3.588588588588589e-05,
+      "loss": 1.7287,
+      "step": 2832
+    },
+    {
+      "epoch": 0.003703253451293655,
+      "grad_norm": 1.0863144397735596,
+      "learning_rate": 3.588088088088089e-05,
+      "loss": 0.8817,
+      "step": 2833
+    },
+    {
+      "epoch": 0.003704560635709925,
+      "grad_norm": 2.586513042449951,
+      "learning_rate": 3.5875875875875874e-05,
+      "loss": 1.0531,
+      "step": 2834
+    },
+    {
+      "epoch": 0.0037058678201261955,
+      "grad_norm": 0.8365259766578674,
+      "learning_rate": 3.5870870870870874e-05,
+      "loss": 0.6141,
+      "step": 2835
+    },
+    {
+      "epoch": 0.003707175004542466,
+      "grad_norm": 1.3868286609649658,
+      "learning_rate": 3.586586586586587e-05,
+      "loss": 2.0799,
+      "step": 2836
+    },
+    {
+      "epoch": 0.003708482188958736,
+      "grad_norm": 1.5161962509155273,
+      "learning_rate": 3.586086086086086e-05,
+      "loss": 1.1679,
+      "step": 2837
+    },
+    {
+      "epoch": 0.0037097893733750065,
+      "grad_norm": 1.219996452331543,
+      "learning_rate": 3.585585585585586e-05,
+      "loss": 1.0706,
+      "step": 2838
+    },
+    {
+      "epoch": 0.003711096557791277,
+      "grad_norm": 1.1823912858963013,
+      "learning_rate": 3.585085085085085e-05,
+      "loss": 1.1517,
+      "step": 2839
+    },
+    {
+      "epoch": 0.0037124037422075467,
+      "grad_norm": 1.2970173358917236,
+      "learning_rate": 3.584584584584585e-05,
+      "loss": 1.4897,
+      "step": 2840
+    },
+    {
+      "epoch": 0.003713710926623817,
+      "grad_norm": 1.574537754058838,
+      "learning_rate": 3.5840840840840845e-05,
+      "loss": 2.0799,
+      "step": 2841
+    },
+    {
+      "epoch": 0.0037150181110400873,
+      "grad_norm": 1.5579519271850586,
+      "learning_rate": 3.583583583583583e-05,
+      "loss": 2.3632,
+      "step": 2842
+    },
+    {
+      "epoch": 0.0037163252954563577,
+      "grad_norm": 1.4396135807037354,
+      "learning_rate": 3.583083083083083e-05,
+      "loss": 1.3123,
+      "step": 2843
+    },
+    {
+      "epoch": 0.003717632479872628,
+      "grad_norm": 1.3910770416259766,
+      "learning_rate": 3.5825825825825824e-05,
+      "loss": 2.1197,
+      "step": 2844
+    },
+    {
+      "epoch": 0.0037189396642888983,
+      "grad_norm": 1.433112382888794,
+      "learning_rate": 3.5820820820820824e-05,
+      "loss": 2.147,
+      "step": 2845
+    },
+    {
+      "epoch": 0.0037202468487051686,
+      "grad_norm": 1.4482258558273315,
+      "learning_rate": 3.581581581581582e-05,
+      "loss": 2.0403,
+      "step": 2846
+    },
+    {
+      "epoch": 0.003721554033121439,
+      "grad_norm": 1.544071912765503,
+      "learning_rate": 3.581081081081081e-05,
+      "loss": 1.6307,
+      "step": 2847
+    },
+    {
+      "epoch": 0.003722861217537709,
+      "grad_norm": 2.0137648582458496,
+      "learning_rate": 3.580580580580581e-05,
+      "loss": 2.4535,
+      "step": 2848
+    },
+    {
+      "epoch": 0.003724168401953979,
+      "grad_norm": 1.228350281715393,
+      "learning_rate": 3.58008008008008e-05,
+      "loss": 1.6079,
+      "step": 2849
+    },
+    {
+      "epoch": 0.0037254755863702495,
+      "grad_norm": 1.3425854444503784,
+      "learning_rate": 3.57957957957958e-05,
+      "loss": 1.899,
+      "step": 2850
+    },
+    {
+      "epoch": 0.0037267827707865198,
+      "grad_norm": 1.5645263195037842,
+      "learning_rate": 3.579079079079079e-05,
+      "loss": 2.0451,
+      "step": 2851
+    },
+    {
+      "epoch": 0.00372808995520279,
+      "grad_norm": 1.3970853090286255,
+      "learning_rate": 3.578578578578579e-05,
+      "loss": 1.112,
+      "step": 2852
+    },
+    {
+      "epoch": 0.0037293971396190604,
+      "grad_norm": 1.226244568824768,
+      "learning_rate": 3.578078078078078e-05,
+      "loss": 1.5567,
+      "step": 2853
+    },
+    {
+      "epoch": 0.0037307043240353307,
+      "grad_norm": 1.329813003540039,
+      "learning_rate": 3.5775775775775774e-05,
+      "loss": 1.6624,
+      "step": 2854
+    },
+    {
+      "epoch": 0.003732011508451601,
+      "grad_norm": 1.1417771577835083,
+      "learning_rate": 3.5770770770770774e-05,
+      "loss": 0.8495,
+      "step": 2855
+    },
+    {
+      "epoch": 0.003733318692867871,
+      "grad_norm": 1.6802831888198853,
+      "learning_rate": 3.576576576576577e-05,
+      "loss": 2.3375,
+      "step": 2856
+    },
+    {
+      "epoch": 0.0037346258772841413,
+      "grad_norm": 1.2436169385910034,
+      "learning_rate": 3.576076076076077e-05,
+      "loss": 1.2446,
+      "step": 2857
+    },
+    {
+      "epoch": 0.0037359330617004116,
+      "grad_norm": 1.562708854675293,
+      "learning_rate": 3.575575575575576e-05,
+      "loss": 1.8955,
+      "step": 2858
+    },
+    {
+      "epoch": 0.003737240246116682,
+      "grad_norm": 1.2857745885849,
+      "learning_rate": 3.575075075075075e-05,
+      "loss": 1.5187,
+      "step": 2859
+    },
+    {
+      "epoch": 0.0037385474305329522,
+      "grad_norm": 1.4674456119537354,
+      "learning_rate": 3.5745745745745746e-05,
+      "loss": 1.6598,
+      "step": 2860
+    },
+    {
+      "epoch": 0.0037398546149492225,
+      "grad_norm": 1.1386479139328003,
+      "learning_rate": 3.574074074074074e-05,
+      "loss": 1.0302,
+      "step": 2861
+    },
+    {
+      "epoch": 0.003741161799365493,
+      "grad_norm": 1.6778892278671265,
+      "learning_rate": 3.573573573573574e-05,
+      "loss": 2.0375,
+      "step": 2862
+    },
+    {
+      "epoch": 0.0037424689837817627,
+      "grad_norm": 1.5058590173721313,
+      "learning_rate": 3.573073073073073e-05,
+      "loss": 2.3527,
+      "step": 2863
+    },
+    {
+      "epoch": 0.003743776168198033,
+      "grad_norm": 1.368266224861145,
+      "learning_rate": 3.572572572572573e-05,
+      "loss": 1.7428,
+      "step": 2864
+    },
+    {
+      "epoch": 0.0037450833526143034,
+      "grad_norm": 1.563900113105774,
+      "learning_rate": 3.5720720720720724e-05,
+      "loss": 2.1114,
+      "step": 2865
+    },
+    {
+      "epoch": 0.0037463905370305737,
+      "grad_norm": 1.4526491165161133,
+      "learning_rate": 3.571571571571572e-05,
+      "loss": 1.9585,
+      "step": 2866
+    },
+    {
+      "epoch": 0.003747697721446844,
+      "grad_norm": 1.0346988439559937,
+      "learning_rate": 3.571071071071071e-05,
+      "loss": 1.139,
+      "step": 2867
+    },
+    {
+      "epoch": 0.0037490049058631143,
+      "grad_norm": 1.4406800270080566,
+      "learning_rate": 3.57057057057057e-05,
+      "loss": 1.7987,
+      "step": 2868
+    },
+    {
+      "epoch": 0.0037503120902793847,
+      "grad_norm": 1.4611284732818604,
+      "learning_rate": 3.57007007007007e-05,
+      "loss": 2.1828,
+      "step": 2869
+    },
+    {
+      "epoch": 0.003751619274695655,
+      "grad_norm": 1.3067735433578491,
+      "learning_rate": 3.5695695695695696e-05,
+      "loss": 1.2217,
+      "step": 2870
+    },
+    {
+      "epoch": 0.003752926459111925,
+      "grad_norm": 1.5834803581237793,
+      "learning_rate": 3.5690690690690695e-05,
+      "loss": 2.2173,
+      "step": 2871
+    },
+    {
+      "epoch": 0.003754233643528195,
+      "grad_norm": 0.972912073135376,
+      "learning_rate": 3.568568568568569e-05,
+      "loss": 0.6823,
+      "step": 2872
+    },
+    {
+      "epoch": 0.0037555408279444655,
+      "grad_norm": 1.2433712482452393,
+      "learning_rate": 3.568068068068068e-05,
+      "loss": 0.9695,
+      "step": 2873
+    },
+    {
+      "epoch": 0.003756848012360736,
+      "grad_norm": 1.2157670259475708,
+      "learning_rate": 3.567567567567568e-05,
+      "loss": 1.383,
+      "step": 2874
+    },
+    {
+      "epoch": 0.003758155196777006,
+      "grad_norm": 1.4538652896881104,
+      "learning_rate": 3.5670670670670674e-05,
+      "loss": 1.6423,
+      "step": 2875
+    },
+    {
+      "epoch": 0.0037594623811932765,
+      "grad_norm": 1.2139878273010254,
+      "learning_rate": 3.566566566566567e-05,
+      "loss": 1.9016,
+      "step": 2876
+    },
+    {
+      "epoch": 0.003760769565609547,
+      "grad_norm": 2.926220417022705,
+      "learning_rate": 3.566066066066066e-05,
+      "loss": 1.1788,
+      "step": 2877
+    },
+    {
+      "epoch": 0.0037620767500258167,
+      "grad_norm": 1.4777657985687256,
+      "learning_rate": 3.565565565565566e-05,
+      "loss": 1.718,
+      "step": 2878
+    },
+    {
+      "epoch": 0.003763383934442087,
+      "grad_norm": 1.435696005821228,
+      "learning_rate": 3.565065065065065e-05,
+      "loss": 1.3769,
+      "step": 2879
+    },
+    {
+      "epoch": 0.0037646911188583573,
+      "grad_norm": 2.1821699142456055,
+      "learning_rate": 3.5645645645645646e-05,
+      "loss": 2.5161,
+      "step": 2880
+    },
+    {
+      "epoch": 0.0037659983032746276,
+      "grad_norm": 1.6220372915267944,
+      "learning_rate": 3.5640640640640646e-05,
+      "loss": 1.7052,
+      "step": 2881
+    },
+    {
+      "epoch": 0.003767305487690898,
+      "grad_norm": 1.359977126121521,
+      "learning_rate": 3.563563563563564e-05,
+      "loss": 1.5287,
+      "step": 2882
+    },
+    {
+      "epoch": 0.0037686126721071683,
+      "grad_norm": 1.1792558431625366,
+      "learning_rate": 3.563063063063063e-05,
+      "loss": 1.7222,
+      "step": 2883
+    },
+    {
+      "epoch": 0.0037699198565234386,
+      "grad_norm": 1.3171054124832153,
+      "learning_rate": 3.5625625625625624e-05,
+      "loss": 1.5711,
+      "step": 2884
+    },
+    {
+      "epoch": 0.003771227040939709,
+      "grad_norm": 1.126084327697754,
+      "learning_rate": 3.562062062062062e-05,
+      "loss": 1.275,
+      "step": 2885
+    },
+    {
+      "epoch": 0.003772534225355979,
+      "grad_norm": 1.3692059516906738,
+      "learning_rate": 3.561561561561562e-05,
+      "loss": 1.3445,
+      "step": 2886
+    },
+    {
+      "epoch": 0.003773841409772249,
+      "grad_norm": 1.535939335823059,
+      "learning_rate": 3.561061061061061e-05,
+      "loss": 1.7141,
+      "step": 2887
+    },
+    {
+      "epoch": 0.0037751485941885194,
+      "grad_norm": 1.2123743295669556,
+      "learning_rate": 3.560560560560561e-05,
+      "loss": 1.0331,
+      "step": 2888
+    },
+    {
+      "epoch": 0.0037764557786047898,
+      "grad_norm": 1.5320062637329102,
+      "learning_rate": 3.56006006006006e-05,
+      "loss": 1.7823,
+      "step": 2889
+    },
+    {
+      "epoch": 0.00377776296302106,
+      "grad_norm": 1.646065354347229,
+      "learning_rate": 3.5595595595595596e-05,
+      "loss": 2.5011,
+      "step": 2890
+    },
+    {
+      "epoch": 0.0037790701474373304,
+      "grad_norm": 1.158921241760254,
+      "learning_rate": 3.5590590590590596e-05,
+      "loss": 1.8345,
+      "step": 2891
+    },
+    {
+      "epoch": 0.0037803773318536007,
+      "grad_norm": 1.1694645881652832,
+      "learning_rate": 3.558558558558558e-05,
+      "loss": 0.6898,
+      "step": 2892
+    },
+    {
+      "epoch": 0.003781684516269871,
+      "grad_norm": 1.1073052883148193,
+      "learning_rate": 3.558058058058058e-05,
+      "loss": 0.7602,
+      "step": 2893
+    },
+    {
+      "epoch": 0.003782991700686141,
+      "grad_norm": 0.90475994348526,
+      "learning_rate": 3.5575575575575575e-05,
+      "loss": 0.5407,
+      "step": 2894
+    },
+    {
+      "epoch": 0.0037842988851024112,
+      "grad_norm": 1.2424514293670654,
+      "learning_rate": 3.5570570570570574e-05,
+      "loss": 1.3455,
+      "step": 2895
+    },
+    {
+      "epoch": 0.0037856060695186816,
+      "grad_norm": 1.515891194343567,
+      "learning_rate": 3.556556556556557e-05,
+      "loss": 1.6309,
+      "step": 2896
+    },
+    {
+      "epoch": 0.003786913253934952,
+      "grad_norm": 1.9598431587219238,
+      "learning_rate": 3.556056056056056e-05,
+      "loss": 2.246,
+      "step": 2897
+    },
+    {
+      "epoch": 0.003788220438351222,
+      "grad_norm": 1.3922761678695679,
+      "learning_rate": 3.555555555555556e-05,
+      "loss": 1.8156,
+      "step": 2898
+    },
+    {
+      "epoch": 0.0037895276227674925,
+      "grad_norm": 1.8645504713058472,
+      "learning_rate": 3.555055055055055e-05,
+      "loss": 1.2815,
+      "step": 2899
+    },
+    {
+      "epoch": 0.003790834807183763,
+      "grad_norm": 1.2784292697906494,
+      "learning_rate": 3.554554554554555e-05,
+      "loss": 1.4871,
+      "step": 2900
+    },
+    {
+      "epoch": 0.0037921419916000327,
+      "grad_norm": 2.3278822898864746,
+      "learning_rate": 3.554054054054054e-05,
+      "loss": 2.0599,
+      "step": 2901
+    },
+    {
+      "epoch": 0.003793449176016303,
+      "grad_norm": 1.6007696390151978,
+      "learning_rate": 3.553553553553554e-05,
+      "loss": 1.4745,
+      "step": 2902
+    },
+    {
+      "epoch": 0.0037947563604325734,
+      "grad_norm": 0.9891722798347473,
+      "learning_rate": 3.553053053053053e-05,
+      "loss": 0.5051,
+      "step": 2903
+    },
+    {
+      "epoch": 0.0037960635448488437,
+      "grad_norm": 1.9795669317245483,
+      "learning_rate": 3.5525525525525525e-05,
+      "loss": 1.3656,
+      "step": 2904
+    },
+    {
+      "epoch": 0.003797370729265114,
+      "grad_norm": 1.1306062936782837,
+      "learning_rate": 3.5520520520520524e-05,
+      "loss": 0.9597,
+      "step": 2905
+    },
+    {
+      "epoch": 0.0037986779136813843,
+      "grad_norm": 1.4530857801437378,
+      "learning_rate": 3.551551551551552e-05,
+      "loss": 1.0015,
+      "step": 2906
+    },
+    {
+      "epoch": 0.0037999850980976546,
+      "grad_norm": 0.9873569011688232,
+      "learning_rate": 3.551051051051052e-05,
+      "loss": 0.8135,
+      "step": 2907
+    },
+    {
+      "epoch": 0.003801292282513925,
+      "grad_norm": 1.5969102382659912,
+      "learning_rate": 3.550550550550551e-05,
+      "loss": 2.0807,
+      "step": 2908
+    },
+    {
+      "epoch": 0.003802599466930195,
+      "grad_norm": 1.4553649425506592,
+      "learning_rate": 3.55005005005005e-05,
+      "loss": 1.7513,
+      "step": 2909
+    },
+    {
+      "epoch": 0.003803906651346465,
+      "grad_norm": 1.1900486946105957,
+      "learning_rate": 3.5495495495495496e-05,
+      "loss": 0.907,
+      "step": 2910
+    },
+    {
+      "epoch": 0.0038052138357627355,
+      "grad_norm": 1.461517572402954,
+      "learning_rate": 3.549049049049049e-05,
+      "loss": 1.133,
+      "step": 2911
+    },
+    {
+      "epoch": 0.003806521020179006,
+      "grad_norm": 1.1475123167037964,
+      "learning_rate": 3.548548548548549e-05,
+      "loss": 1.7598,
+      "step": 2912
+    },
+    {
+      "epoch": 0.003807828204595276,
+      "grad_norm": 1.4428800344467163,
+      "learning_rate": 3.548048048048048e-05,
+      "loss": 1.418,
+      "step": 2913
+    },
+    {
+      "epoch": 0.0038091353890115465,
+      "grad_norm": 1.0116337537765503,
+      "learning_rate": 3.547547547547548e-05,
+      "loss": 1.119,
+      "step": 2914
+    },
+    {
+      "epoch": 0.0038104425734278168,
+      "grad_norm": 1.1862832307815552,
+      "learning_rate": 3.5470470470470474e-05,
+      "loss": 1.4619,
+      "step": 2915
+    },
+    {
+      "epoch": 0.003811749757844087,
+      "grad_norm": 1.0510801076889038,
+      "learning_rate": 3.546546546546547e-05,
+      "loss": 0.9031,
+      "step": 2916
+    },
+    {
+      "epoch": 0.003813056942260357,
+      "grad_norm": 1.3959060907363892,
+      "learning_rate": 3.546046046046046e-05,
+      "loss": 1.18,
+      "step": 2917
+    },
+    {
+      "epoch": 0.0038143641266766273,
+      "grad_norm": 1.1961666345596313,
+      "learning_rate": 3.545545545545545e-05,
+      "loss": 1.5436,
+      "step": 2918
+    },
+    {
+      "epoch": 0.0038156713110928976,
+      "grad_norm": 0.9673799276351929,
+      "learning_rate": 3.545045045045045e-05,
+      "loss": 0.6468,
+      "step": 2919
+    },
+    {
+      "epoch": 0.003816978495509168,
+      "grad_norm": 1.7895686626434326,
+      "learning_rate": 3.5445445445445446e-05,
+      "loss": 1.5691,
+      "step": 2920
+    },
+    {
+      "epoch": 0.0038182856799254383,
+      "grad_norm": 1.2765990495681763,
+      "learning_rate": 3.544044044044044e-05,
+      "loss": 1.6989,
+      "step": 2921
+    },
+    {
+      "epoch": 0.0038195928643417086,
+      "grad_norm": 1.4323266744613647,
+      "learning_rate": 3.543543543543544e-05,
+      "loss": 1.2459,
+      "step": 2922
+    },
+    {
+      "epoch": 0.003820900048757979,
+      "grad_norm": 1.136162519454956,
+      "learning_rate": 3.543043043043043e-05,
+      "loss": 1.2106,
+      "step": 2923
+    },
+    {
+      "epoch": 0.0038222072331742488,
+      "grad_norm": 1.1361809968948364,
+      "learning_rate": 3.542542542542543e-05,
+      "loss": 1.0098,
+      "step": 2924
+    },
+    {
+      "epoch": 0.003823514417590519,
+      "grad_norm": 1.5866011381149292,
+      "learning_rate": 3.5420420420420424e-05,
+      "loss": 1.4895,
+      "step": 2925
+    },
+    {
+      "epoch": 0.0038248216020067894,
+      "grad_norm": 1.3634051084518433,
+      "learning_rate": 3.541541541541542e-05,
+      "loss": 1.7779,
+      "step": 2926
+    },
+    {
+      "epoch": 0.0038261287864230597,
+      "grad_norm": 1.8048988580703735,
+      "learning_rate": 3.541041041041041e-05,
+      "loss": 1.2679,
+      "step": 2927
+    },
+    {
+      "epoch": 0.00382743597083933,
+      "grad_norm": 1.0359121561050415,
+      "learning_rate": 3.5405405405405403e-05,
+      "loss": 1.1603,
+      "step": 2928
+    },
+    {
+      "epoch": 0.0038287431552556004,
+      "grad_norm": 1.5208593606948853,
+      "learning_rate": 3.54004004004004e-05,
+      "loss": 1.3646,
+      "step": 2929
+    },
+    {
+      "epoch": 0.0038300503396718707,
+      "grad_norm": 1.8672988414764404,
+      "learning_rate": 3.5395395395395396e-05,
+      "loss": 2.1305,
+      "step": 2930
+    },
+    {
+      "epoch": 0.003831357524088141,
+      "grad_norm": 2.531644582748413,
+      "learning_rate": 3.5390390390390396e-05,
+      "loss": 1.4276,
+      "step": 2931
+    },
+    {
+      "epoch": 0.003832664708504411,
+      "grad_norm": 1.2581934928894043,
+      "learning_rate": 3.538538538538539e-05,
+      "loss": 1.6208,
+      "step": 2932
+    },
+    {
+      "epoch": 0.0038339718929206812,
+      "grad_norm": 1.4779516458511353,
+      "learning_rate": 3.538038038038038e-05,
+      "loss": 2.0919,
+      "step": 2933
+    },
+    {
+      "epoch": 0.0038352790773369515,
+      "grad_norm": 1.491503357887268,
+      "learning_rate": 3.5375375375375375e-05,
+      "loss": 1.5249,
+      "step": 2934
+    },
+    {
+      "epoch": 0.003836586261753222,
+      "grad_norm": 1.1134963035583496,
+      "learning_rate": 3.537037037037037e-05,
+      "loss": 0.8671,
+      "step": 2935
+    },
+    {
+      "epoch": 0.003837893446169492,
+      "grad_norm": 1.422563910484314,
+      "learning_rate": 3.536536536536537e-05,
+      "loss": 1.5566,
+      "step": 2936
+    },
+    {
+      "epoch": 0.0038392006305857625,
+      "grad_norm": 1.415735125541687,
+      "learning_rate": 3.536036036036036e-05,
+      "loss": 2.059,
+      "step": 2937
+    },
+    {
+      "epoch": 0.003840507815002033,
+      "grad_norm": 1.7905042171478271,
+      "learning_rate": 3.535535535535536e-05,
+      "loss": 2.2439,
+      "step": 2938
+    },
+    {
+      "epoch": 0.003841814999418303,
+      "grad_norm": 1.212609887123108,
+      "learning_rate": 3.535035035035035e-05,
+      "loss": 1.416,
+      "step": 2939
+    },
+    {
+      "epoch": 0.003843122183834573,
+      "grad_norm": 1.7317209243774414,
+      "learning_rate": 3.5345345345345346e-05,
+      "loss": 2.1068,
+      "step": 2940
+    },
+    {
+      "epoch": 0.0038444293682508433,
+      "grad_norm": 1.191105842590332,
+      "learning_rate": 3.5340340340340346e-05,
+      "loss": 3.2061,
+      "step": 2941
+    },
+    {
+      "epoch": 0.0038457365526671137,
+      "grad_norm": 1.4557204246520996,
+      "learning_rate": 3.533533533533533e-05,
+      "loss": 1.6439,
+      "step": 2942
+    },
+    {
+      "epoch": 0.003847043737083384,
+      "grad_norm": 1.3944965600967407,
+      "learning_rate": 3.533033033033033e-05,
+      "loss": 1.6022,
+      "step": 2943
+    },
+    {
+      "epoch": 0.0038483509214996543,
+      "grad_norm": 1.6579053401947021,
+      "learning_rate": 3.5325325325325325e-05,
+      "loss": 1.6073,
+      "step": 2944
+    },
+    {
+      "epoch": 0.0038496581059159246,
+      "grad_norm": 0.5907145738601685,
+      "learning_rate": 3.5320320320320325e-05,
+      "loss": 0.1369,
+      "step": 2945
+    },
+    {
+      "epoch": 0.003850965290332195,
+      "grad_norm": 1.2994279861450195,
+      "learning_rate": 3.531531531531532e-05,
+      "loss": 1.1092,
+      "step": 2946
+    },
+    {
+      "epoch": 0.003852272474748465,
+      "grad_norm": 0.9247709512710571,
+      "learning_rate": 3.531031031031031e-05,
+      "loss": 0.808,
+      "step": 2947
+    },
+    {
+      "epoch": 0.003853579659164735,
+      "grad_norm": 1.1819360256195068,
+      "learning_rate": 3.530530530530531e-05,
+      "loss": 1.312,
+      "step": 2948
+    },
+    {
+      "epoch": 0.0038548868435810055,
+      "grad_norm": 1.82845139503479,
+      "learning_rate": 3.53003003003003e-05,
+      "loss": 1.0634,
+      "step": 2949
+    },
+    {
+      "epoch": 0.003856194027997276,
+      "grad_norm": 1.565563678741455,
+      "learning_rate": 3.5295295295295296e-05,
+      "loss": 1.4934,
+      "step": 2950
+    },
+    {
+      "epoch": 0.003857501212413546,
+      "grad_norm": 1.7253789901733398,
+      "learning_rate": 3.529029029029029e-05,
+      "loss": 1.8555,
+      "step": 2951
+    },
+    {
+      "epoch": 0.0038588083968298164,
+      "grad_norm": 1.1565011739730835,
+      "learning_rate": 3.528528528528528e-05,
+      "loss": 1.6121,
+      "step": 2952
+    },
+    {
+      "epoch": 0.0038601155812460868,
+      "grad_norm": 1.220159888267517,
+      "learning_rate": 3.528028028028028e-05,
+      "loss": 1.0855,
+      "step": 2953
+    },
+    {
+      "epoch": 0.003861422765662357,
+      "grad_norm": 1.5695409774780273,
+      "learning_rate": 3.5275275275275275e-05,
+      "loss": 1.5202,
+      "step": 2954
+    },
+    {
+      "epoch": 0.003862729950078627,
+      "grad_norm": 1.3466715812683105,
+      "learning_rate": 3.5270270270270275e-05,
+      "loss": 1.599,
+      "step": 2955
+    },
+    {
+      "epoch": 0.0038640371344948973,
+      "grad_norm": 1.3318231105804443,
+      "learning_rate": 3.526526526526527e-05,
+      "loss": 1.6561,
+      "step": 2956
+    },
+    {
+      "epoch": 0.0038653443189111676,
+      "grad_norm": 1.2460094690322876,
+      "learning_rate": 3.526026026026026e-05,
+      "loss": 0.9661,
+      "step": 2957
+    },
+    {
+      "epoch": 0.003866651503327438,
+      "grad_norm": 1.6551042795181274,
+      "learning_rate": 3.525525525525526e-05,
+      "loss": 1.3082,
+      "step": 2958
+    },
+    {
+      "epoch": 0.0038679586877437082,
+      "grad_norm": 1.1861748695373535,
+      "learning_rate": 3.5250250250250247e-05,
+      "loss": 0.9147,
+      "step": 2959
+    },
+    {
+      "epoch": 0.0038692658721599786,
+      "grad_norm": 1.1155003309249878,
+      "learning_rate": 3.5245245245245246e-05,
+      "loss": 0.9237,
+      "step": 2960
+    },
+    {
+      "epoch": 0.003870573056576249,
+      "grad_norm": 0.7666558027267456,
+      "learning_rate": 3.524024024024024e-05,
+      "loss": 0.8335,
+      "step": 2961
+    },
+    {
+      "epoch": 0.003871880240992519,
+      "grad_norm": 0.6921489238739014,
+      "learning_rate": 3.523523523523524e-05,
+      "loss": 0.1911,
+      "step": 2962
+    },
+    {
+      "epoch": 0.003873187425408789,
+      "grad_norm": 1.9374140501022339,
+      "learning_rate": 3.523023023023023e-05,
+      "loss": 2.3305,
+      "step": 2963
+    },
+    {
+      "epoch": 0.0038744946098250594,
+      "grad_norm": 1.7464078664779663,
+      "learning_rate": 3.5225225225225225e-05,
+      "loss": 1.7274,
+      "step": 2964
+    },
+    {
+      "epoch": 0.0038758017942413297,
+      "grad_norm": 1.217767357826233,
+      "learning_rate": 3.5220220220220225e-05,
+      "loss": 1.6785,
+      "step": 2965
+    },
+    {
+      "epoch": 0.0038771089786576,
+      "grad_norm": 1.2671685218811035,
+      "learning_rate": 3.521521521521522e-05,
+      "loss": 1.5265,
+      "step": 2966
+    },
+    {
+      "epoch": 0.0038784161630738704,
+      "grad_norm": 1.0266705751419067,
+      "learning_rate": 3.521021021021021e-05,
+      "loss": 0.9164,
+      "step": 2967
+    },
+    {
+      "epoch": 0.0038797233474901407,
+      "grad_norm": 1.292940378189087,
+      "learning_rate": 3.5205205205205204e-05,
+      "loss": 0.5265,
+      "step": 2968
+    },
+    {
+      "epoch": 0.003881030531906411,
+      "grad_norm": 0.9241542220115662,
+      "learning_rate": 3.5200200200200203e-05,
+      "loss": 0.8437,
+      "step": 2969
+    },
+    {
+      "epoch": 0.003882337716322681,
+      "grad_norm": 1.2688102722167969,
+      "learning_rate": 3.5195195195195196e-05,
+      "loss": 1.6983,
+      "step": 2970
+    },
+    {
+      "epoch": 0.003883644900738951,
+      "grad_norm": 1.606455683708191,
+      "learning_rate": 3.519019019019019e-05,
+      "loss": 1.3938,
+      "step": 2971
+    },
+    {
+      "epoch": 0.0038849520851552215,
+      "grad_norm": 1.3137727975845337,
+      "learning_rate": 3.518518518518519e-05,
+      "loss": 1.4471,
+      "step": 2972
+    },
+    {
+      "epoch": 0.003886259269571492,
+      "grad_norm": 1.777220368385315,
+      "learning_rate": 3.518018018018018e-05,
+      "loss": 1.3869,
+      "step": 2973
+    },
+    {
+      "epoch": 0.003887566453987762,
+      "grad_norm": 1.5107009410858154,
+      "learning_rate": 3.517517517517518e-05,
+      "loss": 1.6195,
+      "step": 2974
+    },
+    {
+      "epoch": 0.0038888736384040325,
+      "grad_norm": 1.5090281963348389,
+      "learning_rate": 3.517017017017017e-05,
+      "loss": 2.2084,
+      "step": 2975
+    },
+    {
+      "epoch": 0.003890180822820303,
+      "grad_norm": 1.4425992965698242,
+      "learning_rate": 3.516516516516517e-05,
+      "loss": 1.3324,
+      "step": 2976
+    },
+    {
+      "epoch": 0.003891488007236573,
+      "grad_norm": 1.4445717334747314,
+      "learning_rate": 3.516016016016016e-05,
+      "loss": 1.5308,
+      "step": 2977
+    },
+    {
+      "epoch": 0.003892795191652843,
+      "grad_norm": 2.252941370010376,
+      "learning_rate": 3.5155155155155154e-05,
+      "loss": 1.895,
+      "step": 2978
+    },
+    {
+      "epoch": 0.0038941023760691133,
+      "grad_norm": 1.4449974298477173,
+      "learning_rate": 3.5150150150150154e-05,
+      "loss": 1.1624,
+      "step": 2979
+    },
+    {
+      "epoch": 0.0038954095604853837,
+      "grad_norm": 1.9070172309875488,
+      "learning_rate": 3.5145145145145146e-05,
+      "loss": 2.3463,
+      "step": 2980
+    },
+    {
+      "epoch": 0.003896716744901654,
+      "grad_norm": 1.5298019647598267,
+      "learning_rate": 3.5140140140140146e-05,
+      "loss": 2.3904,
+      "step": 2981
+    },
+    {
+      "epoch": 0.0038980239293179243,
+      "grad_norm": 1.27940034866333,
+      "learning_rate": 3.513513513513514e-05,
+      "loss": 1.4201,
+      "step": 2982
+    },
+    {
+      "epoch": 0.0038993311137341946,
+      "grad_norm": 0.7541519999504089,
+      "learning_rate": 3.513013013013013e-05,
+      "loss": 0.5796,
+      "step": 2983
+    },
+    {
+      "epoch": 0.003900638298150465,
+      "grad_norm": 1.3703696727752686,
+      "learning_rate": 3.5125125125125125e-05,
+      "loss": 1.2865,
+      "step": 2984
+    },
+    {
+      "epoch": 0.0039019454825667352,
+      "grad_norm": 0.9735272526741028,
+      "learning_rate": 3.512012012012012e-05,
+      "loss": 0.6398,
+      "step": 2985
+    },
+    {
+      "epoch": 0.003903252666983005,
+      "grad_norm": 1.7789387702941895,
+      "learning_rate": 3.511511511511512e-05,
+      "loss": 1.622,
+      "step": 2986
+    },
+    {
+      "epoch": 0.0039045598513992755,
+      "grad_norm": 1.66666841506958,
+      "learning_rate": 3.511011011011011e-05,
+      "loss": 2.2939,
+      "step": 2987
+    },
+    {
+      "epoch": 0.0039058670358155458,
+      "grad_norm": 1.3295263051986694,
+      "learning_rate": 3.5105105105105104e-05,
+      "loss": 1.5108,
+      "step": 2988
+    },
+    {
+      "epoch": 0.003907174220231816,
+      "grad_norm": 1.5785349607467651,
+      "learning_rate": 3.5100100100100104e-05,
+      "loss": 2.1871,
+      "step": 2989
+    },
+    {
+      "epoch": 0.003908481404648086,
+      "grad_norm": 1.3709182739257812,
+      "learning_rate": 3.5095095095095097e-05,
+      "loss": 2.0262,
+      "step": 2990
+    },
+    {
+      "epoch": 0.003909788589064356,
+      "grad_norm": 1.9108771085739136,
+      "learning_rate": 3.5090090090090096e-05,
+      "loss": 1.4337,
+      "step": 2991
+    },
+    {
+      "epoch": 0.003911095773480627,
+      "grad_norm": 1.0523524284362793,
+      "learning_rate": 3.508508508508508e-05,
+      "loss": 1.1247,
+      "step": 2992
+    },
+    {
+      "epoch": 0.003912402957896897,
+      "grad_norm": 1.1967765092849731,
+      "learning_rate": 3.508008008008008e-05,
+      "loss": 1.1047,
+      "step": 2993
+    },
+    {
+      "epoch": 0.003913710142313167,
+      "grad_norm": 1.2453207969665527,
+      "learning_rate": 3.5075075075075075e-05,
+      "loss": 1.3788,
+      "step": 2994
+    },
+    {
+      "epoch": 0.003915017326729438,
+      "grad_norm": 1.536301851272583,
+      "learning_rate": 3.507007007007007e-05,
+      "loss": 1.5193,
+      "step": 2995
+    },
+    {
+      "epoch": 0.003916324511145708,
+      "grad_norm": 1.2442865371704102,
+      "learning_rate": 3.506506506506507e-05,
+      "loss": 1.1786,
+      "step": 2996
+    },
+    {
+      "epoch": 0.003917631695561978,
+      "grad_norm": 1.2153270244598389,
+      "learning_rate": 3.506006006006006e-05,
+      "loss": 0.9422,
+      "step": 2997
+    },
+    {
+      "epoch": 0.0039189388799782485,
+      "grad_norm": 1.9064691066741943,
+      "learning_rate": 3.505505505505506e-05,
+      "loss": 2.5219,
+      "step": 2998
+    },
+    {
+      "epoch": 0.003920246064394519,
+      "grad_norm": 1.600610375404358,
+      "learning_rate": 3.5050050050050054e-05,
+      "loss": 2.2259,
+      "step": 2999
+    },
+    {
+      "epoch": 0.003921553248810789,
+      "grad_norm": 1.5644943714141846,
+      "learning_rate": 3.504504504504505e-05,
+      "loss": 1.5639,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0039228604332270595,
+      "grad_norm": 1.4519202709197998,
+      "learning_rate": 3.504004004004004e-05,
+      "loss": 1.8575,
+      "step": 3001
+    },
+    {
+      "epoch": 0.00392416761764333,
+      "grad_norm": 1.3234783411026,
+      "learning_rate": 3.503503503503503e-05,
+      "loss": 1.228,
+      "step": 3002
+    },
+    {
+      "epoch": 0.0039254748020596,
+      "grad_norm": 2.7118053436279297,
+      "learning_rate": 3.503003003003003e-05,
+      "loss": 1.6916,
+      "step": 3003
+    },
+    {
+      "epoch": 0.0039267819864758705,
+      "grad_norm": 1.2207688093185425,
+      "learning_rate": 3.5025025025025025e-05,
+      "loss": 0.8615,
+      "step": 3004
+    },
+    {
+      "epoch": 0.00392808917089214,
+      "grad_norm": 1.4358417987823486,
+      "learning_rate": 3.5020020020020025e-05,
+      "loss": 2.0254,
+      "step": 3005
+    },
+    {
+      "epoch": 0.00392939635530841,
+      "grad_norm": 1.334745168685913,
+      "learning_rate": 3.501501501501502e-05,
+      "loss": 1.5698,
+      "step": 3006
+    },
+    {
+      "epoch": 0.0039307035397246805,
+      "grad_norm": 2.159534215927124,
+      "learning_rate": 3.501001001001001e-05,
+      "loss": 2.3129,
+      "step": 3007
+    },
+    {
+      "epoch": 0.003932010724140951,
+      "grad_norm": 1.4581501483917236,
+      "learning_rate": 3.500500500500501e-05,
+      "loss": 1.5028,
+      "step": 3008
+    },
+    {
+      "epoch": 0.003933317908557221,
+      "grad_norm": 1.3912407159805298,
+      "learning_rate": 3.5e-05,
+      "loss": 1.6765,
+      "step": 3009
+    },
+    {
+      "epoch": 0.0039346250929734915,
+      "grad_norm": 1.3085557222366333,
+      "learning_rate": 3.4994994994995e-05,
+      "loss": 1.4739,
+      "step": 3010
+    },
+    {
+      "epoch": 0.003935932277389762,
+      "grad_norm": 1.3022725582122803,
+      "learning_rate": 3.498998998998999e-05,
+      "loss": 1.8638,
+      "step": 3011
+    },
+    {
+      "epoch": 0.003937239461806032,
+      "grad_norm": 1.1369415521621704,
+      "learning_rate": 3.498498498498499e-05,
+      "loss": 0.967,
+      "step": 3012
+    },
+    {
+      "epoch": 0.0039385466462223025,
+      "grad_norm": 1.153390645980835,
+      "learning_rate": 3.497997997997998e-05,
+      "loss": 0.8395,
+      "step": 3013
+    },
+    {
+      "epoch": 0.003939853830638573,
+      "grad_norm": 1.1280264854431152,
+      "learning_rate": 3.4974974974974975e-05,
+      "loss": 1.2198,
+      "step": 3014
+    },
+    {
+      "epoch": 0.003941161015054843,
+      "grad_norm": 1.5527092218399048,
+      "learning_rate": 3.4969969969969975e-05,
+      "loss": 1.9394,
+      "step": 3015
+    },
+    {
+      "epoch": 0.003942468199471113,
+      "grad_norm": 1.3961762189865112,
+      "learning_rate": 3.496496496496497e-05,
+      "loss": 1.434,
+      "step": 3016
+    },
+    {
+      "epoch": 0.003943775383887384,
+      "grad_norm": 1.4734688997268677,
+      "learning_rate": 3.495995995995996e-05,
+      "loss": 1.8537,
+      "step": 3017
+    },
+    {
+      "epoch": 0.003945082568303654,
+      "grad_norm": 1.604814887046814,
+      "learning_rate": 3.4954954954954954e-05,
+      "loss": 1.837,
+      "step": 3018
+    },
+    {
+      "epoch": 0.003946389752719924,
+      "grad_norm": 1.250744104385376,
+      "learning_rate": 3.4949949949949954e-05,
+      "loss": 1.4434,
+      "step": 3019
+    },
+    {
+      "epoch": 0.003947696937136194,
+      "grad_norm": 1.3366793394088745,
+      "learning_rate": 3.494494494494495e-05,
+      "loss": 1.5313,
+      "step": 3020
+    },
+    {
+      "epoch": 0.003949004121552464,
+      "grad_norm": 1.4234669208526611,
+      "learning_rate": 3.493993993993994e-05,
+      "loss": 1.7876,
+      "step": 3021
+    },
+    {
+      "epoch": 0.0039503113059687345,
+      "grad_norm": 1.407591700553894,
+      "learning_rate": 3.493493493493494e-05,
+      "loss": 1.7432,
+      "step": 3022
+    },
+    {
+      "epoch": 0.003951618490385005,
+      "grad_norm": 1.1962890625,
+      "learning_rate": 3.492992992992993e-05,
+      "loss": 1.2931,
+      "step": 3023
+    },
+    {
+      "epoch": 0.003952925674801275,
+      "grad_norm": 1.4079426527023315,
+      "learning_rate": 3.492492492492493e-05,
+      "loss": 1.2157,
+      "step": 3024
+    },
+    {
+      "epoch": 0.0039542328592175454,
+      "grad_norm": 1.1582657098770142,
+      "learning_rate": 3.491991991991992e-05,
+      "loss": 1.189,
+      "step": 3025
+    },
+    {
+      "epoch": 0.003955540043633816,
+      "grad_norm": 1.0695112943649292,
+      "learning_rate": 3.491491491491491e-05,
+      "loss": 1.257,
+      "step": 3026
+    },
+    {
+      "epoch": 0.003956847228050086,
+      "grad_norm": 1.7699209451675415,
+      "learning_rate": 3.490990990990991e-05,
+      "loss": 2.228,
+      "step": 3027
+    },
+    {
+      "epoch": 0.003958154412466356,
+      "grad_norm": 1.3629157543182373,
+      "learning_rate": 3.4904904904904904e-05,
+      "loss": 1.7004,
+      "step": 3028
+    },
+    {
+      "epoch": 0.003959461596882627,
+      "grad_norm": 1.7020158767700195,
+      "learning_rate": 3.4899899899899904e-05,
+      "loss": 1.9054,
+      "step": 3029
+    },
+    {
+      "epoch": 0.003960768781298897,
+      "grad_norm": 1.7787907123565674,
+      "learning_rate": 3.48948948948949e-05,
+      "loss": 1.8323,
+      "step": 3030
+    },
+    {
+      "epoch": 0.003962075965715167,
+      "grad_norm": 0.9018983840942383,
+      "learning_rate": 3.488988988988989e-05,
+      "loss": 0.4986,
+      "step": 3031
+    },
+    {
+      "epoch": 0.003963383150131438,
+      "grad_norm": 1.6388546228408813,
+      "learning_rate": 3.488488488488489e-05,
+      "loss": 2.0628,
+      "step": 3032
+    },
+    {
+      "epoch": 0.003964690334547708,
+      "grad_norm": 1.5269041061401367,
+      "learning_rate": 3.487987987987988e-05,
+      "loss": 1.7616,
+      "step": 3033
+    },
+    {
+      "epoch": 0.003965997518963978,
+      "grad_norm": 1.1879149675369263,
+      "learning_rate": 3.4874874874874875e-05,
+      "loss": 1.7169,
+      "step": 3034
+    },
+    {
+      "epoch": 0.003967304703380248,
+      "grad_norm": 1.5579456090927124,
+      "learning_rate": 3.486986986986987e-05,
+      "loss": 2.2649,
+      "step": 3035
+    },
+    {
+      "epoch": 0.003968611887796518,
+      "grad_norm": 1.4833892583847046,
+      "learning_rate": 3.486486486486487e-05,
+      "loss": 1.9602,
+      "step": 3036
+    },
+    {
+      "epoch": 0.003969919072212788,
+      "grad_norm": 1.7539290189743042,
+      "learning_rate": 3.485985985985986e-05,
+      "loss": 1.184,
+      "step": 3037
+    },
+    {
+      "epoch": 0.003971226256629059,
+      "grad_norm": 1.360100507736206,
+      "learning_rate": 3.4854854854854854e-05,
+      "loss": 1.3044,
+      "step": 3038
+    },
+    {
+      "epoch": 0.003972533441045329,
+      "grad_norm": 1.4995378255844116,
+      "learning_rate": 3.4849849849849854e-05,
+      "loss": 1.8729,
+      "step": 3039
+    },
+    {
+      "epoch": 0.003973840625461599,
+      "grad_norm": 1.7745767831802368,
+      "learning_rate": 3.484484484484485e-05,
+      "loss": 2.6789,
+      "step": 3040
+    },
+    {
+      "epoch": 0.00397514780987787,
+      "grad_norm": 1.0317915678024292,
+      "learning_rate": 3.483983983983985e-05,
+      "loss": 0.8017,
+      "step": 3041
+    },
+    {
+      "epoch": 0.00397645499429414,
+      "grad_norm": 0.1933841109275818,
+      "learning_rate": 3.483483483483483e-05,
+      "loss": 0.0057,
+      "step": 3042
+    },
+    {
+      "epoch": 0.00397776217871041,
+      "grad_norm": 1.515081524848938,
+      "learning_rate": 3.482982982982983e-05,
+      "loss": 2.0483,
+      "step": 3043
+    },
+    {
+      "epoch": 0.003979069363126681,
+      "grad_norm": 1.5368002653121948,
+      "learning_rate": 3.4824824824824826e-05,
+      "loss": 1.7273,
+      "step": 3044
+    },
+    {
+      "epoch": 0.003980376547542951,
+      "grad_norm": 1.3938548564910889,
+      "learning_rate": 3.481981981981982e-05,
+      "loss": 1.3612,
+      "step": 3045
+    },
+    {
+      "epoch": 0.003981683731959221,
+      "grad_norm": 1.7698230743408203,
+      "learning_rate": 3.481481481481482e-05,
+      "loss": 1.8493,
+      "step": 3046
+    },
+    {
+      "epoch": 0.003982990916375492,
+      "grad_norm": 1.4085091352462769,
+      "learning_rate": 3.480980980980981e-05,
+      "loss": 1.4263,
+      "step": 3047
+    },
+    {
+      "epoch": 0.003984298100791762,
+      "grad_norm": 1.4019111394882202,
+      "learning_rate": 3.480480480480481e-05,
+      "loss": 1.6011,
+      "step": 3048
+    },
+    {
+      "epoch": 0.003985605285208032,
+      "grad_norm": 1.314015507698059,
+      "learning_rate": 3.4799799799799804e-05,
+      "loss": 1.9534,
+      "step": 3049
+    },
+    {
+      "epoch": 0.003986912469624302,
+      "grad_norm": 1.6126222610473633,
+      "learning_rate": 3.47947947947948e-05,
+      "loss": 1.156,
+      "step": 3050
+    },
+    {
+      "epoch": 0.003988219654040572,
+      "grad_norm": 1.3849620819091797,
+      "learning_rate": 3.478978978978979e-05,
+      "loss": 1.9176,
+      "step": 3051
+    },
+    {
+      "epoch": 0.003989526838456842,
+      "grad_norm": 1.3734973669052124,
+      "learning_rate": 3.478478478478478e-05,
+      "loss": 1.4637,
+      "step": 3052
+    },
+    {
+      "epoch": 0.003990834022873113,
+      "grad_norm": 1.1997824907302856,
+      "learning_rate": 3.477977977977978e-05,
+      "loss": 2.1383,
+      "step": 3053
+    },
+    {
+      "epoch": 0.003992141207289383,
+      "grad_norm": 1.3193588256835938,
+      "learning_rate": 3.4774774774774776e-05,
+      "loss": 1.5641,
+      "step": 3054
+    },
+    {
+      "epoch": 0.003993448391705653,
+      "grad_norm": 1.2594304084777832,
+      "learning_rate": 3.4769769769769775e-05,
+      "loss": 2.0804,
+      "step": 3055
+    },
+    {
+      "epoch": 0.003994755576121924,
+      "grad_norm": 1.2876545190811157,
+      "learning_rate": 3.476476476476477e-05,
+      "loss": 1.4786,
+      "step": 3056
+    },
+    {
+      "epoch": 0.003996062760538194,
+      "grad_norm": 1.0934370756149292,
+      "learning_rate": 3.475975975975976e-05,
+      "loss": 1.1057,
+      "step": 3057
+    },
+    {
+      "epoch": 0.003997369944954464,
+      "grad_norm": 1.6201649904251099,
+      "learning_rate": 3.475475475475476e-05,
+      "loss": 2.1845,
+      "step": 3058
+    },
+    {
+      "epoch": 0.003998677129370735,
+      "grad_norm": 1.499373197555542,
+      "learning_rate": 3.474974974974975e-05,
+      "loss": 0.9499,
+      "step": 3059
+    },
+    {
+      "epoch": 0.003999984313787005,
+      "grad_norm": 1.2120426893234253,
+      "learning_rate": 3.474474474474475e-05,
+      "loss": 1.3205,
+      "step": 3060
+    },
+    {
+      "epoch": 0.004001291498203275,
+      "grad_norm": 1.762945532798767,
+      "learning_rate": 3.473973973973974e-05,
+      "loss": 1.7909,
+      "step": 3061
+    },
+    {
+      "epoch": 0.0040025986826195455,
+      "grad_norm": 1.3389142751693726,
+      "learning_rate": 3.473473473473473e-05,
+      "loss": 0.8638,
+      "step": 3062
+    },
+    {
+      "epoch": 0.004003905867035816,
+      "grad_norm": 1.581483244895935,
+      "learning_rate": 3.472972972972973e-05,
+      "loss": 1.8528,
+      "step": 3063
+    },
+    {
+      "epoch": 0.004005213051452086,
+      "grad_norm": 1.3933796882629395,
+      "learning_rate": 3.4724724724724726e-05,
+      "loss": 1.6575,
+      "step": 3064
+    },
+    {
+      "epoch": 0.0040065202358683565,
+      "grad_norm": 1.7124098539352417,
+      "learning_rate": 3.4719719719719725e-05,
+      "loss": 1.4276,
+      "step": 3065
+    },
+    {
+      "epoch": 0.004007827420284626,
+      "grad_norm": 1.312466025352478,
+      "learning_rate": 3.471471471471472e-05,
+      "loss": 1.6419,
+      "step": 3066
+    },
+    {
+      "epoch": 0.004009134604700896,
+      "grad_norm": 1.2958272695541382,
+      "learning_rate": 3.470970970970971e-05,
+      "loss": 0.92,
+      "step": 3067
+    },
+    {
+      "epoch": 0.004010441789117167,
+      "grad_norm": 1.3370827436447144,
+      "learning_rate": 3.4704704704704704e-05,
+      "loss": 1.1879,
+      "step": 3068
+    },
+    {
+      "epoch": 0.004011748973533437,
+      "grad_norm": 1.1844183206558228,
+      "learning_rate": 3.46996996996997e-05,
+      "loss": 1.3838,
+      "step": 3069
+    },
+    {
+      "epoch": 0.004013056157949707,
+      "grad_norm": 1.4316681623458862,
+      "learning_rate": 3.46946946946947e-05,
+      "loss": 1.5224,
+      "step": 3070
+    },
+    {
+      "epoch": 0.0040143633423659775,
+      "grad_norm": 1.2743200063705444,
+      "learning_rate": 3.468968968968969e-05,
+      "loss": 1.1874,
+      "step": 3071
+    },
+    {
+      "epoch": 0.004015670526782248,
+      "grad_norm": 1.3280013799667358,
+      "learning_rate": 3.468468468468469e-05,
+      "loss": 1.2895,
+      "step": 3072
+    },
+    {
+      "epoch": 0.004016977711198518,
+      "grad_norm": 1.2835335731506348,
+      "learning_rate": 3.467967967967968e-05,
+      "loss": 1.6655,
+      "step": 3073
+    },
+    {
+      "epoch": 0.0040182848956147885,
+      "grad_norm": 1.6130801439285278,
+      "learning_rate": 3.4674674674674676e-05,
+      "loss": 1.6615,
+      "step": 3074
+    },
+    {
+      "epoch": 0.004019592080031059,
+      "grad_norm": 1.727862000465393,
+      "learning_rate": 3.466966966966967e-05,
+      "loss": 2.0495,
+      "step": 3075
+    },
+    {
+      "epoch": 0.004020899264447329,
+      "grad_norm": 1.2147480249404907,
+      "learning_rate": 3.466466466466466e-05,
+      "loss": 1.8953,
+      "step": 3076
+    },
+    {
+      "epoch": 0.0040222064488635995,
+      "grad_norm": 1.3226546049118042,
+      "learning_rate": 3.465965965965966e-05,
+      "loss": 1.3459,
+      "step": 3077
+    },
+    {
+      "epoch": 0.00402351363327987,
+      "grad_norm": 1.3533437252044678,
+      "learning_rate": 3.4654654654654654e-05,
+      "loss": 1.3074,
+      "step": 3078
+    },
+    {
+      "epoch": 0.00402482081769614,
+      "grad_norm": 1.3049466609954834,
+      "learning_rate": 3.4649649649649654e-05,
+      "loss": 1.6788,
+      "step": 3079
+    },
+    {
+      "epoch": 0.00402612800211241,
+      "grad_norm": 1.2676786184310913,
+      "learning_rate": 3.464464464464465e-05,
+      "loss": 1.4402,
+      "step": 3080
+    },
+    {
+      "epoch": 0.00402743518652868,
+      "grad_norm": 1.3660720586776733,
+      "learning_rate": 3.463963963963964e-05,
+      "loss": 1.0751,
+      "step": 3081
+    },
+    {
+      "epoch": 0.00402874237094495,
+      "grad_norm": 1.3864120244979858,
+      "learning_rate": 3.463463463463464e-05,
+      "loss": 2.0504,
+      "step": 3082
+    },
+    {
+      "epoch": 0.0040300495553612205,
+      "grad_norm": 1.05370032787323,
+      "learning_rate": 3.4629629629629626e-05,
+      "loss": 0.975,
+      "step": 3083
+    },
+    {
+      "epoch": 0.004031356739777491,
+      "grad_norm": 1.6520171165466309,
+      "learning_rate": 3.4624624624624626e-05,
+      "loss": 1.9085,
+      "step": 3084
+    },
+    {
+      "epoch": 0.004032663924193761,
+      "grad_norm": 1.611242651939392,
+      "learning_rate": 3.461961961961962e-05,
+      "loss": 1.4971,
+      "step": 3085
+    },
+    {
+      "epoch": 0.0040339711086100315,
+      "grad_norm": 1.697516918182373,
+      "learning_rate": 3.461461461461462e-05,
+      "loss": 1.9102,
+      "step": 3086
+    },
+    {
+      "epoch": 0.004035278293026302,
+      "grad_norm": 1.4752655029296875,
+      "learning_rate": 3.460960960960961e-05,
+      "loss": 2.2565,
+      "step": 3087
+    },
+    {
+      "epoch": 0.004036585477442572,
+      "grad_norm": 1.4701422452926636,
+      "learning_rate": 3.4604604604604605e-05,
+      "loss": 1.1016,
+      "step": 3088
+    },
+    {
+      "epoch": 0.004037892661858842,
+      "grad_norm": 1.2365832328796387,
+      "learning_rate": 3.4599599599599604e-05,
+      "loss": 0.7387,
+      "step": 3089
+    },
+    {
+      "epoch": 0.004039199846275113,
+      "grad_norm": 1.4117200374603271,
+      "learning_rate": 3.45945945945946e-05,
+      "loss": 1.2398,
+      "step": 3090
+    },
+    {
+      "epoch": 0.004040507030691383,
+      "grad_norm": 1.5530434846878052,
+      "learning_rate": 3.45895895895896e-05,
+      "loss": 2.0828,
+      "step": 3091
+    },
+    {
+      "epoch": 0.004041814215107653,
+      "grad_norm": 1.4450762271881104,
+      "learning_rate": 3.458458458458458e-05,
+      "loss": 1.3394,
+      "step": 3092
+    },
+    {
+      "epoch": 0.004043121399523924,
+      "grad_norm": 1.2139427661895752,
+      "learning_rate": 3.457957957957958e-05,
+      "loss": 0.9542,
+      "step": 3093
+    },
+    {
+      "epoch": 0.004044428583940194,
+      "grad_norm": 1.1256664991378784,
+      "learning_rate": 3.4574574574574576e-05,
+      "loss": 1.0699,
+      "step": 3094
+    },
+    {
+      "epoch": 0.004045735768356464,
+      "grad_norm": 1.8318194150924683,
+      "learning_rate": 3.456956956956957e-05,
+      "loss": 1.1218,
+      "step": 3095
+    },
+    {
+      "epoch": 0.004047042952772734,
+      "grad_norm": 1.2693408727645874,
+      "learning_rate": 3.456456456456457e-05,
+      "loss": 0.7905,
+      "step": 3096
+    },
+    {
+      "epoch": 0.004048350137189004,
+      "grad_norm": 1.5982357263565063,
+      "learning_rate": 3.455955955955956e-05,
+      "loss": 1.5264,
+      "step": 3097
+    },
+    {
+      "epoch": 0.0040496573216052744,
+      "grad_norm": 1.6894598007202148,
+      "learning_rate": 3.4554554554554555e-05,
+      "loss": 1.9282,
+      "step": 3098
+    },
+    {
+      "epoch": 0.004050964506021545,
+      "grad_norm": 1.5593653917312622,
+      "learning_rate": 3.4549549549549554e-05,
+      "loss": 2.0765,
+      "step": 3099
+    },
+    {
+      "epoch": 0.004052271690437815,
+      "grad_norm": 1.7873270511627197,
+      "learning_rate": 3.454454454454454e-05,
+      "loss": 3.0185,
+      "step": 3100
+    },
+    {
+      "epoch": 0.004053578874854085,
+      "grad_norm": 1.2681888341903687,
+      "learning_rate": 3.453953953953954e-05,
+      "loss": 1.9051,
+      "step": 3101
+    },
+    {
+      "epoch": 0.004054886059270356,
+      "grad_norm": 1.4057801961898804,
+      "learning_rate": 3.453453453453453e-05,
+      "loss": 1.9863,
+      "step": 3102
+    },
+    {
+      "epoch": 0.004056193243686626,
+      "grad_norm": 1.254677176475525,
+      "learning_rate": 3.452952952952953e-05,
+      "loss": 1.3979,
+      "step": 3103
+    },
+    {
+      "epoch": 0.004057500428102896,
+      "grad_norm": 1.981992483139038,
+      "learning_rate": 3.4524524524524526e-05,
+      "loss": 2.0712,
+      "step": 3104
+    },
+    {
+      "epoch": 0.004058807612519167,
+      "grad_norm": 1.528877854347229,
+      "learning_rate": 3.451951951951952e-05,
+      "loss": 1.2017,
+      "step": 3105
+    },
+    {
+      "epoch": 0.004060114796935437,
+      "grad_norm": 1.304294466972351,
+      "learning_rate": 3.451451451451452e-05,
+      "loss": 1.0527,
+      "step": 3106
+    },
+    {
+      "epoch": 0.004061421981351707,
+      "grad_norm": 1.300398826599121,
+      "learning_rate": 3.450950950950951e-05,
+      "loss": 1.3148,
+      "step": 3107
+    },
+    {
+      "epoch": 0.004062729165767978,
+      "grad_norm": 1.3835780620574951,
+      "learning_rate": 3.4504504504504505e-05,
+      "loss": 2.315,
+      "step": 3108
+    },
+    {
+      "epoch": 0.004064036350184248,
+      "grad_norm": 1.720723271369934,
+      "learning_rate": 3.44994994994995e-05,
+      "loss": 2.2243,
+      "step": 3109
+    },
+    {
+      "epoch": 0.004065343534600518,
+      "grad_norm": 1.434697151184082,
+      "learning_rate": 3.44944944944945e-05,
+      "loss": 1.5962,
+      "step": 3110
+    },
+    {
+      "epoch": 0.004066650719016789,
+      "grad_norm": 1.6379166841506958,
+      "learning_rate": 3.448948948948949e-05,
+      "loss": 1.8953,
+      "step": 3111
+    },
+    {
+      "epoch": 0.004067957903433058,
+      "grad_norm": 1.7029049396514893,
+      "learning_rate": 3.448448448448448e-05,
+      "loss": 1.8709,
+      "step": 3112
+    },
+    {
+      "epoch": 0.004069265087849328,
+      "grad_norm": 1.326456904411316,
+      "learning_rate": 3.447947947947948e-05,
+      "loss": 1.9167,
+      "step": 3113
+    },
+    {
+      "epoch": 0.004070572272265599,
+      "grad_norm": 1.3319567441940308,
+      "learning_rate": 3.4474474474474476e-05,
+      "loss": 1.4808,
+      "step": 3114
+    },
+    {
+      "epoch": 0.004071879456681869,
+      "grad_norm": 1.4006801843643188,
+      "learning_rate": 3.4469469469469476e-05,
+      "loss": 1.7092,
+      "step": 3115
+    },
+    {
+      "epoch": 0.004073186641098139,
+      "grad_norm": 1.5987683534622192,
+      "learning_rate": 3.446446446446447e-05,
+      "loss": 0.9802,
+      "step": 3116
+    },
+    {
+      "epoch": 0.00407449382551441,
+      "grad_norm": 1.2016615867614746,
+      "learning_rate": 3.445945945945946e-05,
+      "loss": 1.6036,
+      "step": 3117
+    },
+    {
+      "epoch": 0.00407580100993068,
+      "grad_norm": 2.1535258293151855,
+      "learning_rate": 3.4454454454454455e-05,
+      "loss": 1.0551,
+      "step": 3118
+    },
+    {
+      "epoch": 0.00407710819434695,
+      "grad_norm": 1.4059028625488281,
+      "learning_rate": 3.444944944944945e-05,
+      "loss": 2.5003,
+      "step": 3119
+    },
+    {
+      "epoch": 0.004078415378763221,
+      "grad_norm": 1.2504137754440308,
+      "learning_rate": 3.444444444444445e-05,
+      "loss": 0.8702,
+      "step": 3120
+    },
+    {
+      "epoch": 0.004079722563179491,
+      "grad_norm": 1.584952712059021,
+      "learning_rate": 3.443943943943944e-05,
+      "loss": 1.7499,
+      "step": 3121
+    },
+    {
+      "epoch": 0.004081029747595761,
+      "grad_norm": 1.5660609006881714,
+      "learning_rate": 3.443443443443444e-05,
+      "loss": 1.4305,
+      "step": 3122
+    },
+    {
+      "epoch": 0.0040823369320120316,
+      "grad_norm": 1.24374258518219,
+      "learning_rate": 3.442942942942943e-05,
+      "loss": 1.0693,
+      "step": 3123
+    },
+    {
+      "epoch": 0.004083644116428302,
+      "grad_norm": 1.12322199344635,
+      "learning_rate": 3.4424424424424426e-05,
+      "loss": 1.5566,
+      "step": 3124
+    },
+    {
+      "epoch": 0.004084951300844572,
+      "grad_norm": 1.2935782670974731,
+      "learning_rate": 3.441941941941942e-05,
+      "loss": 1.6242,
+      "step": 3125
+    },
+    {
+      "epoch": 0.0040862584852608425,
+      "grad_norm": 1.4487347602844238,
+      "learning_rate": 3.441441441441441e-05,
+      "loss": 1.8707,
+      "step": 3126
+    },
+    {
+      "epoch": 0.004087565669677112,
+      "grad_norm": 1.3822920322418213,
+      "learning_rate": 3.440940940940941e-05,
+      "loss": 1.7481,
+      "step": 3127
+    },
+    {
+      "epoch": 0.004088872854093382,
+      "grad_norm": 1.5649157762527466,
+      "learning_rate": 3.4404404404404405e-05,
+      "loss": 1.4279,
+      "step": 3128
+    },
+    {
+      "epoch": 0.004090180038509653,
+      "grad_norm": 1.9539108276367188,
+      "learning_rate": 3.4399399399399405e-05,
+      "loss": 1.9719,
+      "step": 3129
+    },
+    {
+      "epoch": 0.004091487222925923,
+      "grad_norm": 1.4738495349884033,
+      "learning_rate": 3.43943943943944e-05,
+      "loss": 2.0457,
+      "step": 3130
+    },
+    {
+      "epoch": 0.004092794407342193,
+      "grad_norm": 1.4030342102050781,
+      "learning_rate": 3.438938938938939e-05,
+      "loss": 1.2634,
+      "step": 3131
+    },
+    {
+      "epoch": 0.004094101591758464,
+      "grad_norm": 1.2907785177230835,
+      "learning_rate": 3.438438438438439e-05,
+      "loss": 2.5312,
+      "step": 3132
+    },
+    {
+      "epoch": 0.004095408776174734,
+      "grad_norm": 1.8926842212677002,
+      "learning_rate": 3.4379379379379376e-05,
+      "loss": 2.179,
+      "step": 3133
+    },
+    {
+      "epoch": 0.004096715960591004,
+      "grad_norm": 1.294417381286621,
+      "learning_rate": 3.4374374374374376e-05,
+      "loss": 1.7783,
+      "step": 3134
+    },
+    {
+      "epoch": 0.0040980231450072745,
+      "grad_norm": 1.3879483938217163,
+      "learning_rate": 3.436936936936937e-05,
+      "loss": 1.8237,
+      "step": 3135
+    },
+    {
+      "epoch": 0.004099330329423545,
+      "grad_norm": 2.7954463958740234,
+      "learning_rate": 3.436436436436436e-05,
+      "loss": 1.6506,
+      "step": 3136
+    },
+    {
+      "epoch": 0.004100637513839815,
+      "grad_norm": 1.4958326816558838,
+      "learning_rate": 3.435935935935936e-05,
+      "loss": 1.6873,
+      "step": 3137
+    },
+    {
+      "epoch": 0.0041019446982560855,
+      "grad_norm": 1.283188819885254,
+      "learning_rate": 3.4354354354354355e-05,
+      "loss": 1.4407,
+      "step": 3138
+    },
+    {
+      "epoch": 0.004103251882672356,
+      "grad_norm": 1.489578127861023,
+      "learning_rate": 3.4349349349349355e-05,
+      "loss": 1.7305,
+      "step": 3139
+    },
+    {
+      "epoch": 0.004104559067088626,
+      "grad_norm": 1.0773628950119019,
+      "learning_rate": 3.434434434434435e-05,
+      "loss": 0.9776,
+      "step": 3140
+    },
+    {
+      "epoch": 0.0041058662515048965,
+      "grad_norm": 1.4204884767532349,
+      "learning_rate": 3.433933933933934e-05,
+      "loss": 1.4471,
+      "step": 3141
+    },
+    {
+      "epoch": 0.004107173435921166,
+      "grad_norm": 1.15123450756073,
+      "learning_rate": 3.4334334334334334e-05,
+      "loss": 1.6587,
+      "step": 3142
+    },
+    {
+      "epoch": 0.004108480620337436,
+      "grad_norm": 1.0152323246002197,
+      "learning_rate": 3.4329329329329327e-05,
+      "loss": 0.8515,
+      "step": 3143
+    },
+    {
+      "epoch": 0.0041097878047537065,
+      "grad_norm": 1.4164080619812012,
+      "learning_rate": 3.4324324324324326e-05,
+      "loss": 1.9048,
+      "step": 3144
+    },
+    {
+      "epoch": 0.004111094989169977,
+      "grad_norm": 2.8175857067108154,
+      "learning_rate": 3.431931931931932e-05,
+      "loss": 1.7619,
+      "step": 3145
+    },
+    {
+      "epoch": 0.004112402173586247,
+      "grad_norm": 0.877864420413971,
+      "learning_rate": 3.431431431431432e-05,
+      "loss": 0.5015,
+      "step": 3146
+    },
+    {
+      "epoch": 0.0041137093580025175,
+      "grad_norm": 1.6641517877578735,
+      "learning_rate": 3.430930930930931e-05,
+      "loss": 2.2866,
+      "step": 3147
+    },
+    {
+      "epoch": 0.004115016542418788,
+      "grad_norm": 1.5778363943099976,
+      "learning_rate": 3.4304304304304305e-05,
+      "loss": 1.8798,
+      "step": 3148
+    },
+    {
+      "epoch": 0.004116323726835058,
+      "grad_norm": 1.7861371040344238,
+      "learning_rate": 3.4299299299299305e-05,
+      "loss": 1.8125,
+      "step": 3149
+    },
+    {
+      "epoch": 0.0041176309112513285,
+      "grad_norm": 1.2221319675445557,
+      "learning_rate": 3.429429429429429e-05,
+      "loss": 1.624,
+      "step": 3150
+    },
+    {
+      "epoch": 0.004118938095667599,
+      "grad_norm": 1.026269555091858,
+      "learning_rate": 3.428928928928929e-05,
+      "loss": 1.1159,
+      "step": 3151
+    },
+    {
+      "epoch": 0.004120245280083869,
+      "grad_norm": 1.3376885652542114,
+      "learning_rate": 3.4284284284284284e-05,
+      "loss": 0.7073,
+      "step": 3152
+    },
+    {
+      "epoch": 0.004121552464500139,
+      "grad_norm": 1.1651440858840942,
+      "learning_rate": 3.427927927927928e-05,
+      "loss": 1.2125,
+      "step": 3153
+    },
+    {
+      "epoch": 0.00412285964891641,
+      "grad_norm": 0.6324484944343567,
+      "learning_rate": 3.4274274274274276e-05,
+      "loss": 0.3312,
+      "step": 3154
+    },
+    {
+      "epoch": 0.00412416683333268,
+      "grad_norm": 1.293595314025879,
+      "learning_rate": 3.426926926926927e-05,
+      "loss": 1.1177,
+      "step": 3155
+    },
+    {
+      "epoch": 0.00412547401774895,
+      "grad_norm": 1.4104477167129517,
+      "learning_rate": 3.426426426426427e-05,
+      "loss": 2.1442,
+      "step": 3156
+    },
+    {
+      "epoch": 0.00412678120216522,
+      "grad_norm": 1.9766932725906372,
+      "learning_rate": 3.425925925925926e-05,
+      "loss": 2.4962,
+      "step": 3157
+    },
+    {
+      "epoch": 0.00412808838658149,
+      "grad_norm": 1.645484447479248,
+      "learning_rate": 3.4254254254254255e-05,
+      "loss": 2.0274,
+      "step": 3158
+    },
+    {
+      "epoch": 0.0041293955709977605,
+      "grad_norm": 1.3759992122650146,
+      "learning_rate": 3.424924924924925e-05,
+      "loss": 2.0334,
+      "step": 3159
+    },
+    {
+      "epoch": 0.004130702755414031,
+      "grad_norm": 1.5312098264694214,
+      "learning_rate": 3.424424424424425e-05,
+      "loss": 2.0213,
+      "step": 3160
+    },
+    {
+      "epoch": 0.004132009939830301,
+      "grad_norm": 1.2432340383529663,
+      "learning_rate": 3.423923923923924e-05,
+      "loss": 1.0481,
+      "step": 3161
+    },
+    {
+      "epoch": 0.004133317124246571,
+      "grad_norm": 1.3274372816085815,
+      "learning_rate": 3.4234234234234234e-05,
+      "loss": 1.1019,
+      "step": 3162
+    },
+    {
+      "epoch": 0.004134624308662842,
+      "grad_norm": 1.2473013401031494,
+      "learning_rate": 3.4229229229229233e-05,
+      "loss": 1.3554,
+      "step": 3163
+    },
+    {
+      "epoch": 0.004135931493079112,
+      "grad_norm": 1.5646729469299316,
+      "learning_rate": 3.4224224224224226e-05,
+      "loss": 1.1946,
+      "step": 3164
+    },
+    {
+      "epoch": 0.004137238677495382,
+      "grad_norm": 1.3847192525863647,
+      "learning_rate": 3.4219219219219226e-05,
+      "loss": 1.6244,
+      "step": 3165
+    },
+    {
+      "epoch": 0.004138545861911653,
+      "grad_norm": 1.6332430839538574,
+      "learning_rate": 3.421421421421422e-05,
+      "loss": 1.3991,
+      "step": 3166
+    },
+    {
+      "epoch": 0.004139853046327923,
+      "grad_norm": 1.4769597053527832,
+      "learning_rate": 3.4209209209209205e-05,
+      "loss": 1.2606,
+      "step": 3167
+    },
+    {
+      "epoch": 0.004141160230744193,
+      "grad_norm": 1.281429409980774,
+      "learning_rate": 3.4204204204204205e-05,
+      "loss": 2.0524,
+      "step": 3168
+    },
+    {
+      "epoch": 0.004142467415160464,
+      "grad_norm": 1.7496025562286377,
+      "learning_rate": 3.41991991991992e-05,
+      "loss": 2.0225,
+      "step": 3169
+    },
+    {
+      "epoch": 0.004143774599576734,
+      "grad_norm": 1.6076065301895142,
+      "learning_rate": 3.41941941941942e-05,
+      "loss": 1.9878,
+      "step": 3170
+    },
+    {
+      "epoch": 0.004145081783993004,
+      "grad_norm": 1.2490872144699097,
+      "learning_rate": 3.418918918918919e-05,
+      "loss": 1.701,
+      "step": 3171
+    },
+    {
+      "epoch": 0.004146388968409275,
+      "grad_norm": 1.1832423210144043,
+      "learning_rate": 3.4184184184184184e-05,
+      "loss": 0.7495,
+      "step": 3172
+    },
+    {
+      "epoch": 0.004147696152825544,
+      "grad_norm": 1.2155017852783203,
+      "learning_rate": 3.4179179179179183e-05,
+      "loss": 0.518,
+      "step": 3173
+    },
+    {
+      "epoch": 0.004149003337241814,
+      "grad_norm": 1.5045818090438843,
+      "learning_rate": 3.4174174174174176e-05,
+      "loss": 1.9706,
+      "step": 3174
+    },
+    {
+      "epoch": 0.004150310521658085,
+      "grad_norm": 1.5498178005218506,
+      "learning_rate": 3.416916916916917e-05,
+      "loss": 2.0638,
+      "step": 3175
+    },
+    {
+      "epoch": 0.004151617706074355,
+      "grad_norm": 1.4913337230682373,
+      "learning_rate": 3.416416416416416e-05,
+      "loss": 1.9748,
+      "step": 3176
+    },
+    {
+      "epoch": 0.004152924890490625,
+      "grad_norm": 1.4130157232284546,
+      "learning_rate": 3.415915915915916e-05,
+      "loss": 1.2864,
+      "step": 3177
+    },
+    {
+      "epoch": 0.004154232074906896,
+      "grad_norm": 1.3120231628417969,
+      "learning_rate": 3.4154154154154155e-05,
+      "loss": 1.5615,
+      "step": 3178
+    },
+    {
+      "epoch": 0.004155539259323166,
+      "grad_norm": 1.3780614137649536,
+      "learning_rate": 3.414914914914915e-05,
+      "loss": 1.6048,
+      "step": 3179
+    },
+    {
+      "epoch": 0.004156846443739436,
+      "grad_norm": 1.500356674194336,
+      "learning_rate": 3.414414414414415e-05,
+      "loss": 1.752,
+      "step": 3180
+    },
+    {
+      "epoch": 0.004158153628155707,
+      "grad_norm": 1.8537654876708984,
+      "learning_rate": 3.413913913913914e-05,
+      "loss": 2.0021,
+      "step": 3181
+    },
+    {
+      "epoch": 0.004159460812571977,
+      "grad_norm": 1.1014785766601562,
+      "learning_rate": 3.413413413413414e-05,
+      "loss": 0.9029,
+      "step": 3182
+    },
+    {
+      "epoch": 0.004160767996988247,
+      "grad_norm": 1.6115854978561401,
+      "learning_rate": 3.412912912912913e-05,
+      "loss": 1.9145,
+      "step": 3183
+    },
+    {
+      "epoch": 0.004162075181404518,
+      "grad_norm": 1.4508918523788452,
+      "learning_rate": 3.4124124124124127e-05,
+      "loss": 1.8623,
+      "step": 3184
+    },
+    {
+      "epoch": 0.004163382365820788,
+      "grad_norm": 2.020355463027954,
+      "learning_rate": 3.411911911911912e-05,
+      "loss": 2.1354,
+      "step": 3185
+    },
+    {
+      "epoch": 0.004164689550237058,
+      "grad_norm": 1.0732336044311523,
+      "learning_rate": 3.411411411411411e-05,
+      "loss": 0.839,
+      "step": 3186
+    },
+    {
+      "epoch": 0.0041659967346533286,
+      "grad_norm": 1.4184250831604004,
+      "learning_rate": 3.410910910910911e-05,
+      "loss": 1.6644,
+      "step": 3187
+    },
+    {
+      "epoch": 0.004167303919069598,
+      "grad_norm": 1.5065339803695679,
+      "learning_rate": 3.4104104104104105e-05,
+      "loss": 2.3329,
+      "step": 3188
+    },
+    {
+      "epoch": 0.004168611103485868,
+      "grad_norm": 0.9889479279518127,
+      "learning_rate": 3.4099099099099105e-05,
+      "loss": 0.778,
+      "step": 3189
+    },
+    {
+      "epoch": 0.004169918287902139,
+      "grad_norm": 1.3000985383987427,
+      "learning_rate": 3.40940940940941e-05,
+      "loss": 1.5824,
+      "step": 3190
+    },
+    {
+      "epoch": 0.004171225472318409,
+      "grad_norm": 1.4913486242294312,
+      "learning_rate": 3.408908908908909e-05,
+      "loss": 1.8632,
+      "step": 3191
+    },
+    {
+      "epoch": 0.004172532656734679,
+      "grad_norm": 1.344981074333191,
+      "learning_rate": 3.4084084084084084e-05,
+      "loss": 2.0219,
+      "step": 3192
+    },
+    {
+      "epoch": 0.00417383984115095,
+      "grad_norm": 1.1597093343734741,
+      "learning_rate": 3.407907907907908e-05,
+      "loss": 0.9599,
+      "step": 3193
+    },
+    {
+      "epoch": 0.00417514702556722,
+      "grad_norm": 1.2788474559783936,
+      "learning_rate": 3.4074074074074077e-05,
+      "loss": 2.1464,
+      "step": 3194
+    },
+    {
+      "epoch": 0.00417645420998349,
+      "grad_norm": 1.2413365840911865,
+      "learning_rate": 3.406906906906907e-05,
+      "loss": 1.0914,
+      "step": 3195
+    },
+    {
+      "epoch": 0.004177761394399761,
+      "grad_norm": 0.7834155559539795,
+      "learning_rate": 3.406406406406407e-05,
+      "loss": 0.8921,
+      "step": 3196
+    },
+    {
+      "epoch": 0.004179068578816031,
+      "grad_norm": 1.6392751932144165,
+      "learning_rate": 3.405905905905906e-05,
+      "loss": 1.9249,
+      "step": 3197
+    },
+    {
+      "epoch": 0.004180375763232301,
+      "grad_norm": 1.2931698560714722,
+      "learning_rate": 3.4054054054054055e-05,
+      "loss": 0.8984,
+      "step": 3198
+    },
+    {
+      "epoch": 0.0041816829476485715,
+      "grad_norm": 1.1856334209442139,
+      "learning_rate": 3.4049049049049055e-05,
+      "loss": 1.0242,
+      "step": 3199
+    },
+    {
+      "epoch": 0.004182990132064842,
+      "grad_norm": 0.8567159175872803,
+      "learning_rate": 3.404404404404404e-05,
+      "loss": 0.79,
+      "step": 3200
+    },
+    {
+      "epoch": 0.004184297316481112,
+      "grad_norm": 1.289857268333435,
+      "learning_rate": 3.403903903903904e-05,
+      "loss": 1.1631,
+      "step": 3201
+    },
+    {
+      "epoch": 0.0041856045008973825,
+      "grad_norm": 0.9272838830947876,
+      "learning_rate": 3.4034034034034034e-05,
+      "loss": 0.8526,
+      "step": 3202
+    },
+    {
+      "epoch": 0.004186911685313652,
+      "grad_norm": 1.2670031785964966,
+      "learning_rate": 3.4029029029029034e-05,
+      "loss": 1.5983,
+      "step": 3203
+    },
+    {
+      "epoch": 0.004188218869729922,
+      "grad_norm": 1.0945837497711182,
+      "learning_rate": 3.402402402402403e-05,
+      "loss": 1.2652,
+      "step": 3204
+    },
+    {
+      "epoch": 0.004189526054146193,
+      "grad_norm": 1.8214190006256104,
+      "learning_rate": 3.401901901901902e-05,
+      "loss": 1.5028,
+      "step": 3205
+    },
+    {
+      "epoch": 0.004190833238562463,
+      "grad_norm": 1.5018476247787476,
+      "learning_rate": 3.401401401401402e-05,
+      "loss": 1.9864,
+      "step": 3206
+    },
+    {
+      "epoch": 0.004192140422978733,
+      "grad_norm": 0.9617783427238464,
+      "learning_rate": 3.400900900900901e-05,
+      "loss": 0.8595,
+      "step": 3207
+    },
+    {
+      "epoch": 0.0041934476073950035,
+      "grad_norm": 1.7697923183441162,
+      "learning_rate": 3.4004004004004005e-05,
+      "loss": 2.0644,
+      "step": 3208
+    },
+    {
+      "epoch": 0.004194754791811274,
+      "grad_norm": 1.155632734298706,
+      "learning_rate": 3.3998998998999e-05,
+      "loss": 1.3313,
+      "step": 3209
+    },
+    {
+      "epoch": 0.004196061976227544,
+      "grad_norm": 1.4709161520004272,
+      "learning_rate": 3.399399399399399e-05,
+      "loss": 2.0404,
+      "step": 3210
+    },
+    {
+      "epoch": 0.0041973691606438145,
+      "grad_norm": 1.3467426300048828,
+      "learning_rate": 3.398898898898899e-05,
+      "loss": 1.6204,
+      "step": 3211
+    },
+    {
+      "epoch": 0.004198676345060085,
+      "grad_norm": 1.3717788457870483,
+      "learning_rate": 3.3983983983983984e-05,
+      "loss": 1.9546,
+      "step": 3212
+    },
+    {
+      "epoch": 0.004199983529476355,
+      "grad_norm": 1.667248249053955,
+      "learning_rate": 3.3978978978978984e-05,
+      "loss": 1.7271,
+      "step": 3213
+    },
+    {
+      "epoch": 0.0042012907138926255,
+      "grad_norm": 1.4660571813583374,
+      "learning_rate": 3.397397397397398e-05,
+      "loss": 1.9649,
+      "step": 3214
+    },
+    {
+      "epoch": 0.004202597898308896,
+      "grad_norm": 1.184264898300171,
+      "learning_rate": 3.396896896896897e-05,
+      "loss": 0.6494,
+      "step": 3215
+    },
+    {
+      "epoch": 0.004203905082725166,
+      "grad_norm": 1.6719837188720703,
+      "learning_rate": 3.396396396396396e-05,
+      "loss": 1.545,
+      "step": 3216
+    },
+    {
+      "epoch": 0.004205212267141436,
+      "grad_norm": 1.8655800819396973,
+      "learning_rate": 3.3958958958958956e-05,
+      "loss": 2.1093,
+      "step": 3217
+    },
+    {
+      "epoch": 0.004206519451557707,
+      "grad_norm": 1.2787548303604126,
+      "learning_rate": 3.3953953953953955e-05,
+      "loss": 1.5456,
+      "step": 3218
+    },
+    {
+      "epoch": 0.004207826635973976,
+      "grad_norm": 1.7180787324905396,
+      "learning_rate": 3.394894894894895e-05,
+      "loss": 2.2231,
+      "step": 3219
+    },
+    {
+      "epoch": 0.0042091338203902465,
+      "grad_norm": 1.2762776613235474,
+      "learning_rate": 3.394394394394395e-05,
+      "loss": 0.6889,
+      "step": 3220
+    },
+    {
+      "epoch": 0.004210441004806517,
+      "grad_norm": 1.3487834930419922,
+      "learning_rate": 3.393893893893894e-05,
+      "loss": 1.3389,
+      "step": 3221
+    },
+    {
+      "epoch": 0.004211748189222787,
+      "grad_norm": 1.691834807395935,
+      "learning_rate": 3.3933933933933934e-05,
+      "loss": 1.4602,
+      "step": 3222
+    },
+    {
+      "epoch": 0.0042130553736390575,
+      "grad_norm": 1.3026552200317383,
+      "learning_rate": 3.3928928928928934e-05,
+      "loss": 1.8215,
+      "step": 3223
+    },
+    {
+      "epoch": 0.004214362558055328,
+      "grad_norm": 1.352425456047058,
+      "learning_rate": 3.392392392392393e-05,
+      "loss": 1.8341,
+      "step": 3224
+    },
+    {
+      "epoch": 0.004215669742471598,
+      "grad_norm": 1.3101794719696045,
+      "learning_rate": 3.391891891891892e-05,
+      "loss": 1.2853,
+      "step": 3225
+    },
+    {
+      "epoch": 0.004216976926887868,
+      "grad_norm": 1.9043660163879395,
+      "learning_rate": 3.391391391391391e-05,
+      "loss": 1.9089,
+      "step": 3226
+    },
+    {
+      "epoch": 0.004218284111304139,
+      "grad_norm": 1.0830366611480713,
+      "learning_rate": 3.390890890890891e-05,
+      "loss": 1.1584,
+      "step": 3227
+    },
+    {
+      "epoch": 0.004219591295720409,
+      "grad_norm": 1.6860272884368896,
+      "learning_rate": 3.3903903903903905e-05,
+      "loss": 2.0777,
+      "step": 3228
+    },
+    {
+      "epoch": 0.004220898480136679,
+      "grad_norm": 1.2777775526046753,
+      "learning_rate": 3.38988988988989e-05,
+      "loss": 1.1453,
+      "step": 3229
+    },
+    {
+      "epoch": 0.00422220566455295,
+      "grad_norm": 1.0225402116775513,
+      "learning_rate": 3.38938938938939e-05,
+      "loss": 0.6292,
+      "step": 3230
+    },
+    {
+      "epoch": 0.00422351284896922,
+      "grad_norm": 1.339294672012329,
+      "learning_rate": 3.388888888888889e-05,
+      "loss": 1.9023,
+      "step": 3231
+    },
+    {
+      "epoch": 0.00422482003338549,
+      "grad_norm": 1.397996425628662,
+      "learning_rate": 3.388388388388389e-05,
+      "loss": 2.3066,
+      "step": 3232
+    },
+    {
+      "epoch": 0.004226127217801761,
+      "grad_norm": 1.2374407052993774,
+      "learning_rate": 3.387887887887888e-05,
+      "loss": 1.6232,
+      "step": 3233
+    },
+    {
+      "epoch": 0.00422743440221803,
+      "grad_norm": 1.3591307401657104,
+      "learning_rate": 3.387387387387388e-05,
+      "loss": 1.4205,
+      "step": 3234
+    },
+    {
+      "epoch": 0.0042287415866343004,
+      "grad_norm": 1.0532375574111938,
+      "learning_rate": 3.386886886886887e-05,
+      "loss": 0.7899,
+      "step": 3235
+    },
+    {
+      "epoch": 0.004230048771050571,
+      "grad_norm": 1.201082706451416,
+      "learning_rate": 3.386386386386386e-05,
+      "loss": 1.2536,
+      "step": 3236
+    },
+    {
+      "epoch": 0.004231355955466841,
+      "grad_norm": 1.2788898944854736,
+      "learning_rate": 3.385885885885886e-05,
+      "loss": 1.0324,
+      "step": 3237
+    },
+    {
+      "epoch": 0.004232663139883111,
+      "grad_norm": 1.1261671781539917,
+      "learning_rate": 3.3853853853853856e-05,
+      "loss": 1.1954,
+      "step": 3238
+    },
+    {
+      "epoch": 0.004233970324299382,
+      "grad_norm": 1.313024878501892,
+      "learning_rate": 3.3848848848848855e-05,
+      "loss": 1.2528,
+      "step": 3239
+    },
+    {
+      "epoch": 0.004235277508715652,
+      "grad_norm": 1.2694529294967651,
+      "learning_rate": 3.384384384384385e-05,
+      "loss": 1.7541,
+      "step": 3240
+    },
+    {
+      "epoch": 0.004236584693131922,
+      "grad_norm": 1.2748205661773682,
+      "learning_rate": 3.3838838838838834e-05,
+      "loss": 0.8612,
+      "step": 3241
+    },
+    {
+      "epoch": 0.004237891877548193,
+      "grad_norm": 1.4546661376953125,
+      "learning_rate": 3.3833833833833834e-05,
+      "loss": 2.3487,
+      "step": 3242
+    },
+    {
+      "epoch": 0.004239199061964463,
+      "grad_norm": 1.2582027912139893,
+      "learning_rate": 3.382882882882883e-05,
+      "loss": 1.1754,
+      "step": 3243
+    },
+    {
+      "epoch": 0.004240506246380733,
+      "grad_norm": 1.301694393157959,
+      "learning_rate": 3.382382382382383e-05,
+      "loss": 1.7073,
+      "step": 3244
+    },
+    {
+      "epoch": 0.004241813430797004,
+      "grad_norm": 0.9866232872009277,
+      "learning_rate": 3.381881881881882e-05,
+      "loss": 0.6707,
+      "step": 3245
+    },
+    {
+      "epoch": 0.004243120615213274,
+      "grad_norm": 1.110988974571228,
+      "learning_rate": 3.381381381381381e-05,
+      "loss": 1.0956,
+      "step": 3246
+    },
+    {
+      "epoch": 0.004244427799629544,
+      "grad_norm": 1.3637092113494873,
+      "learning_rate": 3.380880880880881e-05,
+      "loss": 0.7852,
+      "step": 3247
+    },
+    {
+      "epoch": 0.004245734984045815,
+      "grad_norm": 1.5957266092300415,
+      "learning_rate": 3.3803803803803806e-05,
+      "loss": 1.8584,
+      "step": 3248
+    },
+    {
+      "epoch": 0.004247042168462084,
+      "grad_norm": 1.7779754400253296,
+      "learning_rate": 3.3798798798798805e-05,
+      "loss": 1.4819,
+      "step": 3249
+    },
+    {
+      "epoch": 0.004248349352878354,
+      "grad_norm": 1.4387362003326416,
+      "learning_rate": 3.379379379379379e-05,
+      "loss": 2.0644,
+      "step": 3250
+    },
+    {
+      "epoch": 0.004249656537294625,
+      "grad_norm": 1.837379813194275,
+      "learning_rate": 3.378878878878879e-05,
+      "loss": 1.7624,
+      "step": 3251
+    },
+    {
+      "epoch": 0.004250963721710895,
+      "grad_norm": 1.4112223386764526,
+      "learning_rate": 3.3783783783783784e-05,
+      "loss": 2.0642,
+      "step": 3252
+    },
+    {
+      "epoch": 0.004252270906127165,
+      "grad_norm": 1.2607730627059937,
+      "learning_rate": 3.377877877877878e-05,
+      "loss": 1.8346,
+      "step": 3253
+    },
+    {
+      "epoch": 0.004253578090543436,
+      "grad_norm": 0.8135144710540771,
+      "learning_rate": 3.377377377377378e-05,
+      "loss": 0.492,
+      "step": 3254
+    },
+    {
+      "epoch": 0.004254885274959706,
+      "grad_norm": 0.8086853623390198,
+      "learning_rate": 3.376876876876877e-05,
+      "loss": 0.3675,
+      "step": 3255
+    },
+    {
+      "epoch": 0.004256192459375976,
+      "grad_norm": 1.8897724151611328,
+      "learning_rate": 3.376376376376377e-05,
+      "loss": 2.101,
+      "step": 3256
+    },
+    {
+      "epoch": 0.004257499643792247,
+      "grad_norm": 1.2320036888122559,
+      "learning_rate": 3.375875875875876e-05,
+      "loss": 1.6978,
+      "step": 3257
+    },
+    {
+      "epoch": 0.004258806828208517,
+      "grad_norm": 1.3521469831466675,
+      "learning_rate": 3.3753753753753756e-05,
+      "loss": 2.1054,
+      "step": 3258
+    },
+    {
+      "epoch": 0.004260114012624787,
+      "grad_norm": 1.7384158372879028,
+      "learning_rate": 3.374874874874875e-05,
+      "loss": 1.6784,
+      "step": 3259
+    },
+    {
+      "epoch": 0.0042614211970410576,
+      "grad_norm": 1.2262991666793823,
+      "learning_rate": 3.374374374374374e-05,
+      "loss": 0.7249,
+      "step": 3260
+    },
+    {
+      "epoch": 0.004262728381457328,
+      "grad_norm": 1.2205599546432495,
+      "learning_rate": 3.373873873873874e-05,
+      "loss": 1.0238,
+      "step": 3261
+    },
+    {
+      "epoch": 0.004264035565873598,
+      "grad_norm": 1.2266802787780762,
+      "learning_rate": 3.3733733733733734e-05,
+      "loss": 1.8706,
+      "step": 3262
+    },
+    {
+      "epoch": 0.0042653427502898685,
+      "grad_norm": 1.1375057697296143,
+      "learning_rate": 3.3728728728728734e-05,
+      "loss": 1.1974,
+      "step": 3263
+    },
+    {
+      "epoch": 0.004266649934706139,
+      "grad_norm": 1.2505751848220825,
+      "learning_rate": 3.372372372372373e-05,
+      "loss": 1.7302,
+      "step": 3264
+    },
+    {
+      "epoch": 0.004267957119122408,
+      "grad_norm": 1.3683843612670898,
+      "learning_rate": 3.371871871871872e-05,
+      "loss": 0.8942,
+      "step": 3265
+    },
+    {
+      "epoch": 0.004269264303538679,
+      "grad_norm": 1.20112943649292,
+      "learning_rate": 3.371371371371371e-05,
+      "loss": 0.7064,
+      "step": 3266
+    },
+    {
+      "epoch": 0.004270571487954949,
+      "grad_norm": 2.156282901763916,
+      "learning_rate": 3.3708708708708706e-05,
+      "loss": 1.684,
+      "step": 3267
+    },
+    {
+      "epoch": 0.004271878672371219,
+      "grad_norm": 2.5173799991607666,
+      "learning_rate": 3.3703703703703706e-05,
+      "loss": 1.4599,
+      "step": 3268
+    },
+    {
+      "epoch": 0.00427318585678749,
+      "grad_norm": 1.5606098175048828,
+      "learning_rate": 3.36986986986987e-05,
+      "loss": 1.7264,
+      "step": 3269
+    },
+    {
+      "epoch": 0.00427449304120376,
+      "grad_norm": 1.5135796070098877,
+      "learning_rate": 3.36936936936937e-05,
+      "loss": 1.7718,
+      "step": 3270
+    },
+    {
+      "epoch": 0.00427580022562003,
+      "grad_norm": 1.558154821395874,
+      "learning_rate": 3.368868868868869e-05,
+      "loss": 2.2755,
+      "step": 3271
+    },
+    {
+      "epoch": 0.0042771074100363005,
+      "grad_norm": 1.5072457790374756,
+      "learning_rate": 3.3683683683683684e-05,
+      "loss": 2.3179,
+      "step": 3272
+    },
+    {
+      "epoch": 0.004278414594452571,
+      "grad_norm": 1.589341402053833,
+      "learning_rate": 3.3678678678678684e-05,
+      "loss": 1.4053,
+      "step": 3273
+    },
+    {
+      "epoch": 0.004279721778868841,
+      "grad_norm": 1.2742806673049927,
+      "learning_rate": 3.367367367367368e-05,
+      "loss": 1.7169,
+      "step": 3274
+    },
+    {
+      "epoch": 0.0042810289632851115,
+      "grad_norm": 1.6226623058319092,
+      "learning_rate": 3.366866866866867e-05,
+      "loss": 1.8231,
+      "step": 3275
+    },
+    {
+      "epoch": 0.004282336147701382,
+      "grad_norm": 1.4614248275756836,
+      "learning_rate": 3.366366366366366e-05,
+      "loss": 1.2733,
+      "step": 3276
+    },
+    {
+      "epoch": 0.004283643332117652,
+      "grad_norm": 1.62065589427948,
+      "learning_rate": 3.3658658658658656e-05,
+      "loss": 2.0657,
+      "step": 3277
+    },
+    {
+      "epoch": 0.0042849505165339224,
+      "grad_norm": 0.9981331825256348,
+      "learning_rate": 3.3653653653653656e-05,
+      "loss": 0.5218,
+      "step": 3278
+    },
+    {
+      "epoch": 0.004286257700950193,
+      "grad_norm": 1.293732762336731,
+      "learning_rate": 3.364864864864865e-05,
+      "loss": 0.7363,
+      "step": 3279
+    },
+    {
+      "epoch": 0.004287564885366462,
+      "grad_norm": 1.4088441133499146,
+      "learning_rate": 3.364364364364365e-05,
+      "loss": 1.575,
+      "step": 3280
+    },
+    {
+      "epoch": 0.0042888720697827325,
+      "grad_norm": 1.236334204673767,
+      "learning_rate": 3.363863863863864e-05,
+      "loss": 1.2568,
+      "step": 3281
+    },
+    {
+      "epoch": 0.004290179254199003,
+      "grad_norm": 1.4390239715576172,
+      "learning_rate": 3.3633633633633635e-05,
+      "loss": 1.6851,
+      "step": 3282
+    },
+    {
+      "epoch": 0.004291486438615273,
+      "grad_norm": 1.4568650722503662,
+      "learning_rate": 3.362862862862863e-05,
+      "loss": 1.4387,
+      "step": 3283
+    },
+    {
+      "epoch": 0.0042927936230315435,
+      "grad_norm": 4.666598320007324,
+      "learning_rate": 3.362362362362362e-05,
+      "loss": 0.9166,
+      "step": 3284
+    },
+    {
+      "epoch": 0.004294100807447814,
+      "grad_norm": 1.0137736797332764,
+      "learning_rate": 3.361861861861862e-05,
+      "loss": 0.6776,
+      "step": 3285
+    },
+    {
+      "epoch": 0.004295407991864084,
+      "grad_norm": 1.152185320854187,
+      "learning_rate": 3.361361361361361e-05,
+      "loss": 0.7615,
+      "step": 3286
+    },
+    {
+      "epoch": 0.0042967151762803545,
+      "grad_norm": 1.7524219751358032,
+      "learning_rate": 3.360860860860861e-05,
+      "loss": 1.7527,
+      "step": 3287
+    },
+    {
+      "epoch": 0.004298022360696625,
+      "grad_norm": 1.6522284746170044,
+      "learning_rate": 3.3603603603603606e-05,
+      "loss": 1.6983,
+      "step": 3288
+    },
+    {
+      "epoch": 0.004299329545112895,
+      "grad_norm": 1.0490833520889282,
+      "learning_rate": 3.35985985985986e-05,
+      "loss": 0.7763,
+      "step": 3289
+    },
+    {
+      "epoch": 0.004300636729529165,
+      "grad_norm": 1.316153645515442,
+      "learning_rate": 3.35935935935936e-05,
+      "loss": 1.8923,
+      "step": 3290
+    },
+    {
+      "epoch": 0.004301943913945436,
+      "grad_norm": 1.271294355392456,
+      "learning_rate": 3.3588588588588585e-05,
+      "loss": 1.9226,
+      "step": 3291
+    },
+    {
+      "epoch": 0.004303251098361706,
+      "grad_norm": 1.0413357019424438,
+      "learning_rate": 3.3583583583583585e-05,
+      "loss": 0.7019,
+      "step": 3292
+    },
+    {
+      "epoch": 0.004304558282777976,
+      "grad_norm": 1.7482075691223145,
+      "learning_rate": 3.357857857857858e-05,
+      "loss": 2.6391,
+      "step": 3293
+    },
+    {
+      "epoch": 0.004305865467194247,
+      "grad_norm": 1.340573787689209,
+      "learning_rate": 3.357357357357358e-05,
+      "loss": 1.456,
+      "step": 3294
+    },
+    {
+      "epoch": 0.004307172651610516,
+      "grad_norm": 2.253429651260376,
+      "learning_rate": 3.356856856856857e-05,
+      "loss": 2.4903,
+      "step": 3295
+    },
+    {
+      "epoch": 0.0043084798360267865,
+      "grad_norm": 1.2621557712554932,
+      "learning_rate": 3.356356356356356e-05,
+      "loss": 1.3964,
+      "step": 3296
+    },
+    {
+      "epoch": 0.004309787020443057,
+      "grad_norm": 1.6297005414962769,
+      "learning_rate": 3.355855855855856e-05,
+      "loss": 1.9243,
+      "step": 3297
+    },
+    {
+      "epoch": 0.004311094204859327,
+      "grad_norm": 1.3744324445724487,
+      "learning_rate": 3.3553553553553556e-05,
+      "loss": 2.098,
+      "step": 3298
+    },
+    {
+      "epoch": 0.004312401389275597,
+      "grad_norm": 1.4279913902282715,
+      "learning_rate": 3.3548548548548556e-05,
+      "loss": 1.425,
+      "step": 3299
+    },
+    {
+      "epoch": 0.004313708573691868,
+      "grad_norm": 1.4083722829818726,
+      "learning_rate": 3.354354354354354e-05,
+      "loss": 1.9666,
+      "step": 3300
+    },
+    {
+      "epoch": 0.004315015758108138,
+      "grad_norm": 1.3334280252456665,
+      "learning_rate": 3.353853853853854e-05,
+      "loss": 1.5063,
+      "step": 3301
+    },
+    {
+      "epoch": 0.004316322942524408,
+      "grad_norm": 1.5585932731628418,
+      "learning_rate": 3.3533533533533535e-05,
+      "loss": 1.0183,
+      "step": 3302
+    },
+    {
+      "epoch": 0.004317630126940679,
+      "grad_norm": 1.3571521043777466,
+      "learning_rate": 3.352852852852853e-05,
+      "loss": 1.1959,
+      "step": 3303
+    },
+    {
+      "epoch": 0.004318937311356949,
+      "grad_norm": 1.616234302520752,
+      "learning_rate": 3.352352352352353e-05,
+      "loss": 1.7933,
+      "step": 3304
+    },
+    {
+      "epoch": 0.004320244495773219,
+      "grad_norm": 1.3724408149719238,
+      "learning_rate": 3.351851851851852e-05,
+      "loss": 2.1286,
+      "step": 3305
+    },
+    {
+      "epoch": 0.00432155168018949,
+      "grad_norm": 1.6403170824050903,
+      "learning_rate": 3.351351351351352e-05,
+      "loss": 2.1878,
+      "step": 3306
+    },
+    {
+      "epoch": 0.00432285886460576,
+      "grad_norm": 2.0880966186523438,
+      "learning_rate": 3.350850850850851e-05,
+      "loss": 1.6815,
+      "step": 3307
+    },
+    {
+      "epoch": 0.00432416604902203,
+      "grad_norm": 1.3008496761322021,
+      "learning_rate": 3.3503503503503506e-05,
+      "loss": 1.5747,
+      "step": 3308
+    },
+    {
+      "epoch": 0.004325473233438301,
+      "grad_norm": 1.3305301666259766,
+      "learning_rate": 3.34984984984985e-05,
+      "loss": 1.6022,
+      "step": 3309
+    },
+    {
+      "epoch": 0.00432678041785457,
+      "grad_norm": 1.1590991020202637,
+      "learning_rate": 3.349349349349349e-05,
+      "loss": 1.7054,
+      "step": 3310
+    },
+    {
+      "epoch": 0.00432808760227084,
+      "grad_norm": 1.166357398033142,
+      "learning_rate": 3.348848848848849e-05,
+      "loss": 0.9754,
+      "step": 3311
+    },
+    {
+      "epoch": 0.004329394786687111,
+      "grad_norm": 1.5032774209976196,
+      "learning_rate": 3.3483483483483485e-05,
+      "loss": 1.571,
+      "step": 3312
+    },
+    {
+      "epoch": 0.004330701971103381,
+      "grad_norm": 1.7500120401382446,
+      "learning_rate": 3.347847847847848e-05,
+      "loss": 1.8743,
+      "step": 3313
+    },
+    {
+      "epoch": 0.004332009155519651,
+      "grad_norm": 1.407330870628357,
+      "learning_rate": 3.347347347347348e-05,
+      "loss": 1.3002,
+      "step": 3314
+    },
+    {
+      "epoch": 0.004333316339935922,
+      "grad_norm": 1.081375002861023,
+      "learning_rate": 3.346846846846847e-05,
+      "loss": 0.9548,
+      "step": 3315
+    },
+    {
+      "epoch": 0.004334623524352192,
+      "grad_norm": 1.3675044775009155,
+      "learning_rate": 3.3463463463463463e-05,
+      "loss": 2.3542,
+      "step": 3316
+    },
+    {
+      "epoch": 0.004335930708768462,
+      "grad_norm": 1.3908326625823975,
+      "learning_rate": 3.3458458458458456e-05,
+      "loss": 0.9364,
+      "step": 3317
+    },
+    {
+      "epoch": 0.004337237893184733,
+      "grad_norm": 1.415856957435608,
+      "learning_rate": 3.3453453453453456e-05,
+      "loss": 2.1198,
+      "step": 3318
+    },
+    {
+      "epoch": 0.004338545077601003,
+      "grad_norm": 1.2785530090332031,
+      "learning_rate": 3.344844844844845e-05,
+      "loss": 1.4829,
+      "step": 3319
+    },
+    {
+      "epoch": 0.004339852262017273,
+      "grad_norm": 1.3288816213607788,
+      "learning_rate": 3.344344344344344e-05,
+      "loss": 1.6281,
+      "step": 3320
+    },
+    {
+      "epoch": 0.004341159446433544,
+      "grad_norm": 1.5631892681121826,
+      "learning_rate": 3.343843843843844e-05,
+      "loss": 1.8434,
+      "step": 3321
+    },
+    {
+      "epoch": 0.004342466630849814,
+      "grad_norm": 1.0766639709472656,
+      "learning_rate": 3.3433433433433435e-05,
+      "loss": 1.1216,
+      "step": 3322
+    },
+    {
+      "epoch": 0.004343773815266084,
+      "grad_norm": 1.1968775987625122,
+      "learning_rate": 3.3428428428428435e-05,
+      "loss": 1.413,
+      "step": 3323
+    },
+    {
+      "epoch": 0.0043450809996823546,
+      "grad_norm": 1.5687758922576904,
+      "learning_rate": 3.342342342342342e-05,
+      "loss": 1.8507,
+      "step": 3324
+    },
+    {
+      "epoch": 0.004346388184098625,
+      "grad_norm": 2.053462266921997,
+      "learning_rate": 3.341841841841842e-05,
+      "loss": 1.6499,
+      "step": 3325
+    },
+    {
+      "epoch": 0.004347695368514894,
+      "grad_norm": 1.3537338972091675,
+      "learning_rate": 3.3413413413413413e-05,
+      "loss": 1.8942,
+      "step": 3326
+    },
+    {
+      "epoch": 0.004349002552931165,
+      "grad_norm": 1.4953231811523438,
+      "learning_rate": 3.3408408408408406e-05,
+      "loss": 2.0954,
+      "step": 3327
+    },
+    {
+      "epoch": 0.004350309737347435,
+      "grad_norm": 1.3626375198364258,
+      "learning_rate": 3.3403403403403406e-05,
+      "loss": 2.3014,
+      "step": 3328
+    },
+    {
+      "epoch": 0.004351616921763705,
+      "grad_norm": 1.4396573305130005,
+      "learning_rate": 3.33983983983984e-05,
+      "loss": 1.9623,
+      "step": 3329
+    },
+    {
+      "epoch": 0.004352924106179976,
+      "grad_norm": 1.4227871894836426,
+      "learning_rate": 3.33933933933934e-05,
+      "loss": 2.3085,
+      "step": 3330
+    },
+    {
+      "epoch": 0.004354231290596246,
+      "grad_norm": 1.4760898351669312,
+      "learning_rate": 3.338838838838839e-05,
+      "loss": 2.0705,
+      "step": 3331
+    },
+    {
+      "epoch": 0.004355538475012516,
+      "grad_norm": 1.1615900993347168,
+      "learning_rate": 3.3383383383383385e-05,
+      "loss": 1.596,
+      "step": 3332
+    },
+    {
+      "epoch": 0.0043568456594287866,
+      "grad_norm": 1.2561218738555908,
+      "learning_rate": 3.337837837837838e-05,
+      "loss": 1.4359,
+      "step": 3333
+    },
+    {
+      "epoch": 0.004358152843845057,
+      "grad_norm": 1.221392273902893,
+      "learning_rate": 3.337337337337337e-05,
+      "loss": 2.0782,
+      "step": 3334
+    },
+    {
+      "epoch": 0.004359460028261327,
+      "grad_norm": 0.7159159779548645,
+      "learning_rate": 3.336836836836837e-05,
+      "loss": 0.3704,
+      "step": 3335
+    },
+    {
+      "epoch": 0.0043607672126775975,
+      "grad_norm": 1.4985312223434448,
+      "learning_rate": 3.3363363363363364e-05,
+      "loss": 1.024,
+      "step": 3336
+    },
+    {
+      "epoch": 0.004362074397093868,
+      "grad_norm": 0.9608471989631653,
+      "learning_rate": 3.335835835835836e-05,
+      "loss": 0.9954,
+      "step": 3337
+    },
+    {
+      "epoch": 0.004363381581510138,
+      "grad_norm": 1.5233590602874756,
+      "learning_rate": 3.3353353353353356e-05,
+      "loss": 1.5183,
+      "step": 3338
+    },
+    {
+      "epoch": 0.0043646887659264085,
+      "grad_norm": 1.4853579998016357,
+      "learning_rate": 3.334834834834835e-05,
+      "loss": 2.3944,
+      "step": 3339
+    },
+    {
+      "epoch": 0.004365995950342679,
+      "grad_norm": 1.2855236530303955,
+      "learning_rate": 3.334334334334335e-05,
+      "loss": 1.3541,
+      "step": 3340
+    },
+    {
+      "epoch": 0.004367303134758948,
+      "grad_norm": 1.33084237575531,
+      "learning_rate": 3.3338338338338335e-05,
+      "loss": 0.9641,
+      "step": 3341
+    },
+    {
+      "epoch": 0.004368610319175219,
+      "grad_norm": 1.3391188383102417,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.2186,
+      "step": 3342
+    },
+    {
+      "epoch": 0.004369917503591489,
+      "grad_norm": 0.987775444984436,
+      "learning_rate": 3.332832832832833e-05,
+      "loss": 0.4806,
+      "step": 3343
+    },
+    {
+      "epoch": 0.004371224688007759,
+      "grad_norm": 0.8732923269271851,
+      "learning_rate": 3.332332332332333e-05,
+      "loss": 0.7215,
+      "step": 3344
+    },
+    {
+      "epoch": 0.0043725318724240295,
+      "grad_norm": 1.6878703832626343,
+      "learning_rate": 3.331831831831832e-05,
+      "loss": 2.1202,
+      "step": 3345
+    },
+    {
+      "epoch": 0.0043738390568403,
+      "grad_norm": 1.4609410762786865,
+      "learning_rate": 3.3313313313313314e-05,
+      "loss": 2.1002,
+      "step": 3346
+    },
+    {
+      "epoch": 0.00437514624125657,
+      "grad_norm": 1.1153719425201416,
+      "learning_rate": 3.330830830830831e-05,
+      "loss": 1.3023,
+      "step": 3347
+    },
+    {
+      "epoch": 0.0043764534256728405,
+      "grad_norm": 1.8145288228988647,
+      "learning_rate": 3.3303303303303306e-05,
+      "loss": 2.278,
+      "step": 3348
+    },
+    {
+      "epoch": 0.004377760610089111,
+      "grad_norm": 1.4297199249267578,
+      "learning_rate": 3.32982982982983e-05,
+      "loss": 1.8058,
+      "step": 3349
+    },
+    {
+      "epoch": 0.004379067794505381,
+      "grad_norm": 1.3090025186538696,
+      "learning_rate": 3.329329329329329e-05,
+      "loss": 1.1813,
+      "step": 3350
+    },
+    {
+      "epoch": 0.0043803749789216515,
+      "grad_norm": 1.2335373163223267,
+      "learning_rate": 3.3288288288288285e-05,
+      "loss": 1.2796,
+      "step": 3351
+    },
+    {
+      "epoch": 0.004381682163337922,
+      "grad_norm": 1.1985572576522827,
+      "learning_rate": 3.3283283283283285e-05,
+      "loss": 0.8543,
+      "step": 3352
+    },
+    {
+      "epoch": 0.004382989347754192,
+      "grad_norm": 1.4354188442230225,
+      "learning_rate": 3.327827827827828e-05,
+      "loss": 1.2683,
+      "step": 3353
+    },
+    {
+      "epoch": 0.004384296532170462,
+      "grad_norm": 1.200337290763855,
+      "learning_rate": 3.327327327327328e-05,
+      "loss": 1.436,
+      "step": 3354
+    },
+    {
+      "epoch": 0.004385603716586733,
+      "grad_norm": 1.7506444454193115,
+      "learning_rate": 3.326826826826827e-05,
+      "loss": 0.4337,
+      "step": 3355
+    },
+    {
+      "epoch": 0.004386910901003002,
+      "grad_norm": 1.461457371711731,
+      "learning_rate": 3.3263263263263264e-05,
+      "loss": 2.5194,
+      "step": 3356
+    },
+    {
+      "epoch": 0.0043882180854192725,
+      "grad_norm": 1.6282626390457153,
+      "learning_rate": 3.3258258258258263e-05,
+      "loss": 1.9809,
+      "step": 3357
+    },
+    {
+      "epoch": 0.004389525269835543,
+      "grad_norm": 1.9164149761199951,
+      "learning_rate": 3.325325325325325e-05,
+      "loss": 1.6283,
+      "step": 3358
+    },
+    {
+      "epoch": 0.004390832454251813,
+      "grad_norm": 1.4207667112350464,
+      "learning_rate": 3.324824824824825e-05,
+      "loss": 2.0383,
+      "step": 3359
+    },
+    {
+      "epoch": 0.0043921396386680835,
+      "grad_norm": 1.2669339179992676,
+      "learning_rate": 3.324324324324324e-05,
+      "loss": 1.948,
+      "step": 3360
+    },
+    {
+      "epoch": 0.004393446823084354,
+      "grad_norm": 1.7127854824066162,
+      "learning_rate": 3.323823823823824e-05,
+      "loss": 1.6426,
+      "step": 3361
+    },
+    {
+      "epoch": 0.004394754007500624,
+      "grad_norm": 1.2539429664611816,
+      "learning_rate": 3.3233233233233235e-05,
+      "loss": 1.2145,
+      "step": 3362
+    },
+    {
+      "epoch": 0.004396061191916894,
+      "grad_norm": 1.2124152183532715,
+      "learning_rate": 3.322822822822823e-05,
+      "loss": 1.136,
+      "step": 3363
+    },
+    {
+      "epoch": 0.004397368376333165,
+      "grad_norm": 1.1915217638015747,
+      "learning_rate": 3.322322322322323e-05,
+      "loss": 0.9334,
+      "step": 3364
+    },
+    {
+      "epoch": 0.004398675560749435,
+      "grad_norm": 1.3883477449417114,
+      "learning_rate": 3.321821821821822e-05,
+      "loss": 1.6982,
+      "step": 3365
+    },
+    {
+      "epoch": 0.004399982745165705,
+      "grad_norm": 1.7634721994400024,
+      "learning_rate": 3.3213213213213214e-05,
+      "loss": 2.5574,
+      "step": 3366
+    },
+    {
+      "epoch": 0.004401289929581976,
+      "grad_norm": 1.6605018377304077,
+      "learning_rate": 3.320820820820821e-05,
+      "loss": 0.9758,
+      "step": 3367
+    },
+    {
+      "epoch": 0.004402597113998246,
+      "grad_norm": 1.1735637187957764,
+      "learning_rate": 3.3203203203203206e-05,
+      "loss": 1.8441,
+      "step": 3368
+    },
+    {
+      "epoch": 0.004403904298414516,
+      "grad_norm": 1.4276081323623657,
+      "learning_rate": 3.31981981981982e-05,
+      "loss": 0.6021,
+      "step": 3369
+    },
+    {
+      "epoch": 0.004405211482830787,
+      "grad_norm": 1.272661805152893,
+      "learning_rate": 3.319319319319319e-05,
+      "loss": 0.6821,
+      "step": 3370
+    },
+    {
+      "epoch": 0.004406518667247057,
+      "grad_norm": 1.701146125793457,
+      "learning_rate": 3.318818818818819e-05,
+      "loss": 2.18,
+      "step": 3371
+    },
+    {
+      "epoch": 0.004407825851663326,
+      "grad_norm": 1.2687122821807861,
+      "learning_rate": 3.3183183183183185e-05,
+      "loss": 2.6545,
+      "step": 3372
+    },
+    {
+      "epoch": 0.004409133036079597,
+      "grad_norm": 1.394423246383667,
+      "learning_rate": 3.3178178178178185e-05,
+      "loss": 1.1564,
+      "step": 3373
+    },
+    {
+      "epoch": 0.004410440220495867,
+      "grad_norm": 1.3281351327896118,
+      "learning_rate": 3.317317317317317e-05,
+      "loss": 2.4069,
+      "step": 3374
+    },
+    {
+      "epoch": 0.004411747404912137,
+      "grad_norm": 1.655739665031433,
+      "learning_rate": 3.316816816816817e-05,
+      "loss": 1.6975,
+      "step": 3375
+    },
+    {
+      "epoch": 0.004413054589328408,
+      "grad_norm": 1.5157688856124878,
+      "learning_rate": 3.3163163163163164e-05,
+      "loss": 2.0832,
+      "step": 3376
+    },
+    {
+      "epoch": 0.004414361773744678,
+      "grad_norm": 1.7725648880004883,
+      "learning_rate": 3.315815815815816e-05,
+      "loss": 2.0922,
+      "step": 3377
+    },
+    {
+      "epoch": 0.004415668958160948,
+      "grad_norm": 1.6835054159164429,
+      "learning_rate": 3.3153153153153157e-05,
+      "loss": 2.2988,
+      "step": 3378
+    },
+    {
+      "epoch": 0.004416976142577219,
+      "grad_norm": 1.4247983694076538,
+      "learning_rate": 3.314814814814815e-05,
+      "loss": 2.4292,
+      "step": 3379
+    },
+    {
+      "epoch": 0.004418283326993489,
+      "grad_norm": 1.38420569896698,
+      "learning_rate": 3.314314314314315e-05,
+      "loss": 1.3384,
+      "step": 3380
+    },
+    {
+      "epoch": 0.004419590511409759,
+      "grad_norm": 1.6431100368499756,
+      "learning_rate": 3.313813813813814e-05,
+      "loss": 1.793,
+      "step": 3381
+    },
+    {
+      "epoch": 0.00442089769582603,
+      "grad_norm": 1.1732089519500732,
+      "learning_rate": 3.3133133133133135e-05,
+      "loss": 1.9091,
+      "step": 3382
+    },
+    {
+      "epoch": 0.0044222048802423,
+      "grad_norm": 1.453823447227478,
+      "learning_rate": 3.312812812812813e-05,
+      "loss": 1.9784,
+      "step": 3383
+    },
+    {
+      "epoch": 0.00442351206465857,
+      "grad_norm": 1.1105390787124634,
+      "learning_rate": 3.312312312312312e-05,
+      "loss": 1.3568,
+      "step": 3384
+    },
+    {
+      "epoch": 0.004424819249074841,
+      "grad_norm": 1.4305601119995117,
+      "learning_rate": 3.311811811811812e-05,
+      "loss": 1.3435,
+      "step": 3385
+    },
+    {
+      "epoch": 0.004426126433491111,
+      "grad_norm": 1.4516966342926025,
+      "learning_rate": 3.3113113113113114e-05,
+      "loss": 2.0252,
+      "step": 3386
+    },
+    {
+      "epoch": 0.00442743361790738,
+      "grad_norm": 1.2145485877990723,
+      "learning_rate": 3.310810810810811e-05,
+      "loss": 1.3338,
+      "step": 3387
+    },
+    {
+      "epoch": 0.004428740802323651,
+      "grad_norm": 1.2250754833221436,
+      "learning_rate": 3.3103103103103107e-05,
+      "loss": 1.1094,
+      "step": 3388
+    },
+    {
+      "epoch": 0.004430047986739921,
+      "grad_norm": 1.67503821849823,
+      "learning_rate": 3.30980980980981e-05,
+      "loss": 1.8609,
+      "step": 3389
+    },
+    {
+      "epoch": 0.004431355171156191,
+      "grad_norm": 1.7395014762878418,
+      "learning_rate": 3.30930930930931e-05,
+      "loss": 1.6519,
+      "step": 3390
+    },
+    {
+      "epoch": 0.004432662355572462,
+      "grad_norm": 1.1626067161560059,
+      "learning_rate": 3.3088088088088086e-05,
+      "loss": 1.201,
+      "step": 3391
+    },
+    {
+      "epoch": 0.004433969539988732,
+      "grad_norm": 1.3127105236053467,
+      "learning_rate": 3.3083083083083085e-05,
+      "loss": 0.9864,
+      "step": 3392
+    },
+    {
+      "epoch": 0.004435276724405002,
+      "grad_norm": 1.2742059230804443,
+      "learning_rate": 3.307807807807808e-05,
+      "loss": 0.9642,
+      "step": 3393
+    },
+    {
+      "epoch": 0.004436583908821273,
+      "grad_norm": 1.3982608318328857,
+      "learning_rate": 3.307307307307307e-05,
+      "loss": 1.5488,
+      "step": 3394
+    },
+    {
+      "epoch": 0.004437891093237543,
+      "grad_norm": 1.1694340705871582,
+      "learning_rate": 3.306806806806807e-05,
+      "loss": 0.8605,
+      "step": 3395
+    },
+    {
+      "epoch": 0.004439198277653813,
+      "grad_norm": 1.5203526020050049,
+      "learning_rate": 3.3063063063063064e-05,
+      "loss": 1.4136,
+      "step": 3396
+    },
+    {
+      "epoch": 0.0044405054620700836,
+      "grad_norm": 1.333490252494812,
+      "learning_rate": 3.3058058058058064e-05,
+      "loss": 1.4852,
+      "step": 3397
+    },
+    {
+      "epoch": 0.004441812646486354,
+      "grad_norm": 1.961647868156433,
+      "learning_rate": 3.305305305305306e-05,
+      "loss": 1.8321,
+      "step": 3398
+    },
+    {
+      "epoch": 0.004443119830902624,
+      "grad_norm": 1.3402601480484009,
+      "learning_rate": 3.304804804804805e-05,
+      "loss": 1.7504,
+      "step": 3399
+    },
+    {
+      "epoch": 0.0044444270153188945,
+      "grad_norm": 1.2401872873306274,
+      "learning_rate": 3.304304304304304e-05,
+      "loss": 1.4617,
+      "step": 3400
+    },
+    {
+      "epoch": 0.004445734199735165,
+      "grad_norm": 1.4922242164611816,
+      "learning_rate": 3.3038038038038036e-05,
+      "loss": 2.0191,
+      "step": 3401
+    },
+    {
+      "epoch": 0.004447041384151434,
+      "grad_norm": 1.4765310287475586,
+      "learning_rate": 3.3033033033033035e-05,
+      "loss": 2.2143,
+      "step": 3402
+    },
+    {
+      "epoch": 0.004448348568567705,
+      "grad_norm": 1.0875645875930786,
+      "learning_rate": 3.302802802802803e-05,
+      "loss": 1.3754,
+      "step": 3403
+    },
+    {
+      "epoch": 0.004449655752983975,
+      "grad_norm": 1.5556226968765259,
+      "learning_rate": 3.302302302302303e-05,
+      "loss": 2.1509,
+      "step": 3404
+    },
+    {
+      "epoch": 0.004450962937400245,
+      "grad_norm": 1.300756573677063,
+      "learning_rate": 3.301801801801802e-05,
+      "loss": 1.1539,
+      "step": 3405
+    },
+    {
+      "epoch": 0.004452270121816516,
+      "grad_norm": 1.3705846071243286,
+      "learning_rate": 3.3013013013013014e-05,
+      "loss": 2.1289,
+      "step": 3406
+    },
+    {
+      "epoch": 0.004453577306232786,
+      "grad_norm": 1.1773649454116821,
+      "learning_rate": 3.3008008008008014e-05,
+      "loss": 0.8426,
+      "step": 3407
+    },
+    {
+      "epoch": 0.004454884490649056,
+      "grad_norm": 1.3171160221099854,
+      "learning_rate": 3.3003003003003e-05,
+      "loss": 2.055,
+      "step": 3408
+    },
+    {
+      "epoch": 0.0044561916750653265,
+      "grad_norm": 1.347678780555725,
+      "learning_rate": 3.2997997997998e-05,
+      "loss": 1.2144,
+      "step": 3409
+    },
+    {
+      "epoch": 0.004457498859481597,
+      "grad_norm": 1.4931813478469849,
+      "learning_rate": 3.299299299299299e-05,
+      "loss": 2.225,
+      "step": 3410
+    },
+    {
+      "epoch": 0.004458806043897867,
+      "grad_norm": 1.2681080102920532,
+      "learning_rate": 3.298798798798799e-05,
+      "loss": 0.7586,
+      "step": 3411
+    },
+    {
+      "epoch": 0.0044601132283141375,
+      "grad_norm": 1.4055017232894897,
+      "learning_rate": 3.2982982982982985e-05,
+      "loss": 1.9602,
+      "step": 3412
+    },
+    {
+      "epoch": 0.004461420412730408,
+      "grad_norm": 1.3664710521697998,
+      "learning_rate": 3.297797797797798e-05,
+      "loss": 1.046,
+      "step": 3413
+    },
+    {
+      "epoch": 0.004462727597146678,
+      "grad_norm": 1.422303318977356,
+      "learning_rate": 3.297297297297298e-05,
+      "loss": 1.9314,
+      "step": 3414
+    },
+    {
+      "epoch": 0.0044640347815629484,
+      "grad_norm": 1.23494291305542,
+      "learning_rate": 3.296796796796797e-05,
+      "loss": 1.0492,
+      "step": 3415
+    },
+    {
+      "epoch": 0.004465341965979219,
+      "grad_norm": 1.3020633459091187,
+      "learning_rate": 3.2962962962962964e-05,
+      "loss": 1.6133,
+      "step": 3416
+    },
+    {
+      "epoch": 0.004466649150395488,
+      "grad_norm": 1.1204026937484741,
+      "learning_rate": 3.295795795795796e-05,
+      "loss": 0.965,
+      "step": 3417
+    },
+    {
+      "epoch": 0.0044679563348117585,
+      "grad_norm": 1.7060054540634155,
+      "learning_rate": 3.295295295295296e-05,
+      "loss": 2.0568,
+      "step": 3418
+    },
+    {
+      "epoch": 0.004469263519228029,
+      "grad_norm": 1.1027138233184814,
+      "learning_rate": 3.294794794794795e-05,
+      "loss": 1.0146,
+      "step": 3419
+    },
+    {
+      "epoch": 0.004470570703644299,
+      "grad_norm": 1.352363109588623,
+      "learning_rate": 3.294294294294294e-05,
+      "loss": 1.689,
+      "step": 3420
+    },
+    {
+      "epoch": 0.0044718778880605695,
+      "grad_norm": 1.1835700273513794,
+      "learning_rate": 3.293793793793794e-05,
+      "loss": 1.4542,
+      "step": 3421
+    },
+    {
+      "epoch": 0.00447318507247684,
+      "grad_norm": 1.5091197490692139,
+      "learning_rate": 3.2932932932932935e-05,
+      "loss": 1.1687,
+      "step": 3422
+    },
+    {
+      "epoch": 0.00447449225689311,
+      "grad_norm": 1.0750128030776978,
+      "learning_rate": 3.292792792792793e-05,
+      "loss": 0.8174,
+      "step": 3423
+    },
+    {
+      "epoch": 0.0044757994413093805,
+      "grad_norm": 1.0716252326965332,
+      "learning_rate": 3.292292292292292e-05,
+      "loss": 0.7839,
+      "step": 3424
+    },
+    {
+      "epoch": 0.004477106625725651,
+      "grad_norm": 1.0311143398284912,
+      "learning_rate": 3.2917917917917914e-05,
+      "loss": 1.1022,
+      "step": 3425
+    },
+    {
+      "epoch": 0.004478413810141921,
+      "grad_norm": 1.1251130104064941,
+      "learning_rate": 3.2912912912912914e-05,
+      "loss": 1.2031,
+      "step": 3426
+    },
+    {
+      "epoch": 0.004479720994558191,
+      "grad_norm": 1.8101786375045776,
+      "learning_rate": 3.290790790790791e-05,
+      "loss": 2.4296,
+      "step": 3427
+    },
+    {
+      "epoch": 0.004481028178974462,
+      "grad_norm": 1.139562964439392,
+      "learning_rate": 3.290290290290291e-05,
+      "loss": 1.0931,
+      "step": 3428
+    },
+    {
+      "epoch": 0.004482335363390732,
+      "grad_norm": 1.1558988094329834,
+      "learning_rate": 3.28978978978979e-05,
+      "loss": 1.1789,
+      "step": 3429
+    },
+    {
+      "epoch": 0.004483642547807002,
+      "grad_norm": 1.529935598373413,
+      "learning_rate": 3.289289289289289e-05,
+      "loss": 1.5161,
+      "step": 3430
+    },
+    {
+      "epoch": 0.004484949732223273,
+      "grad_norm": 1.63473379611969,
+      "learning_rate": 3.288788788788789e-05,
+      "loss": 1.9498,
+      "step": 3431
+    },
+    {
+      "epoch": 0.004486256916639543,
+      "grad_norm": 1.4359771013259888,
+      "learning_rate": 3.2882882882882886e-05,
+      "loss": 2.0989,
+      "step": 3432
+    },
+    {
+      "epoch": 0.0044875641010558125,
+      "grad_norm": 1.386170744895935,
+      "learning_rate": 3.287787787787788e-05,
+      "loss": 1.3985,
+      "step": 3433
+    },
+    {
+      "epoch": 0.004488871285472083,
+      "grad_norm": 1.3101948499679565,
+      "learning_rate": 3.287287287287287e-05,
+      "loss": 1.6363,
+      "step": 3434
+    },
+    {
+      "epoch": 0.004490178469888353,
+      "grad_norm": 1.2498805522918701,
+      "learning_rate": 3.286786786786787e-05,
+      "loss": 1.3618,
+      "step": 3435
+    },
+    {
+      "epoch": 0.004491485654304623,
+      "grad_norm": 1.4738324880599976,
+      "learning_rate": 3.2862862862862864e-05,
+      "loss": 2.1685,
+      "step": 3436
+    },
+    {
+      "epoch": 0.004492792838720894,
+      "grad_norm": 1.1388912200927734,
+      "learning_rate": 3.285785785785786e-05,
+      "loss": 1.8293,
+      "step": 3437
+    },
+    {
+      "epoch": 0.004494100023137164,
+      "grad_norm": 2.583237409591675,
+      "learning_rate": 3.285285285285286e-05,
+      "loss": 1.8825,
+      "step": 3438
+    },
+    {
+      "epoch": 0.004495407207553434,
+      "grad_norm": 1.7235517501831055,
+      "learning_rate": 3.284784784784785e-05,
+      "loss": 1.7428,
+      "step": 3439
+    },
+    {
+      "epoch": 0.004496714391969705,
+      "grad_norm": 0.9737971425056458,
+      "learning_rate": 3.284284284284285e-05,
+      "loss": 0.8065,
+      "step": 3440
+    },
+    {
+      "epoch": 0.004498021576385975,
+      "grad_norm": 1.4729934930801392,
+      "learning_rate": 3.2837837837837836e-05,
+      "loss": 2.2304,
+      "step": 3441
+    },
+    {
+      "epoch": 0.004499328760802245,
+      "grad_norm": 1.3615418672561646,
+      "learning_rate": 3.2832832832832836e-05,
+      "loss": 2.0553,
+      "step": 3442
+    },
+    {
+      "epoch": 0.004500635945218516,
+      "grad_norm": 1.806220531463623,
+      "learning_rate": 3.282782782782783e-05,
+      "loss": 1.0335,
+      "step": 3443
+    },
+    {
+      "epoch": 0.004501943129634786,
+      "grad_norm": 1.4712011814117432,
+      "learning_rate": 3.282282282282282e-05,
+      "loss": 1.758,
+      "step": 3444
+    },
+    {
+      "epoch": 0.004503250314051056,
+      "grad_norm": 1.4298021793365479,
+      "learning_rate": 3.281781781781782e-05,
+      "loss": 1.5658,
+      "step": 3445
+    },
+    {
+      "epoch": 0.004504557498467327,
+      "grad_norm": 1.4044040441513062,
+      "learning_rate": 3.2812812812812814e-05,
+      "loss": 1.7129,
+      "step": 3446
+    },
+    {
+      "epoch": 0.004505864682883597,
+      "grad_norm": 1.265467643737793,
+      "learning_rate": 3.2807807807807814e-05,
+      "loss": 0.7943,
+      "step": 3447
+    },
+    {
+      "epoch": 0.004507171867299866,
+      "grad_norm": 1.6177741289138794,
+      "learning_rate": 3.280280280280281e-05,
+      "loss": 2.0238,
+      "step": 3448
+    },
+    {
+      "epoch": 0.004508479051716137,
+      "grad_norm": 1.6964904069900513,
+      "learning_rate": 3.27977977977978e-05,
+      "loss": 1.8657,
+      "step": 3449
+    },
+    {
+      "epoch": 0.004509786236132407,
+      "grad_norm": 1.5637845993041992,
+      "learning_rate": 3.279279279279279e-05,
+      "loss": 0.8604,
+      "step": 3450
+    },
+    {
+      "epoch": 0.004511093420548677,
+      "grad_norm": 1.3037234544754028,
+      "learning_rate": 3.2787787787787786e-05,
+      "loss": 0.8777,
+      "step": 3451
+    },
+    {
+      "epoch": 0.004512400604964948,
+      "grad_norm": 1.340825080871582,
+      "learning_rate": 3.2782782782782786e-05,
+      "loss": 1.7216,
+      "step": 3452
+    },
+    {
+      "epoch": 0.004513707789381218,
+      "grad_norm": 1.760250210762024,
+      "learning_rate": 3.277777777777778e-05,
+      "loss": 1.1055,
+      "step": 3453
+    },
+    {
+      "epoch": 0.004515014973797488,
+      "grad_norm": 1.3634790182113647,
+      "learning_rate": 3.277277277277278e-05,
+      "loss": 1.3246,
+      "step": 3454
+    },
+    {
+      "epoch": 0.004516322158213759,
+      "grad_norm": 1.2608551979064941,
+      "learning_rate": 3.276776776776777e-05,
+      "loss": 1.5469,
+      "step": 3455
+    },
+    {
+      "epoch": 0.004517629342630029,
+      "grad_norm": 1.1032097339630127,
+      "learning_rate": 3.2762762762762764e-05,
+      "loss": 1.0837,
+      "step": 3456
+    },
+    {
+      "epoch": 0.004518936527046299,
+      "grad_norm": 1.5008866786956787,
+      "learning_rate": 3.275775775775776e-05,
+      "loss": 2.0486,
+      "step": 3457
+    },
+    {
+      "epoch": 0.00452024371146257,
+      "grad_norm": 1.2154933214187622,
+      "learning_rate": 3.275275275275275e-05,
+      "loss": 1.2138,
+      "step": 3458
+    },
+    {
+      "epoch": 0.00452155089587884,
+      "grad_norm": 1.4545056819915771,
+      "learning_rate": 3.274774774774775e-05,
+      "loss": 1.6876,
+      "step": 3459
+    },
+    {
+      "epoch": 0.00452285808029511,
+      "grad_norm": 1.179267406463623,
+      "learning_rate": 3.274274274274274e-05,
+      "loss": 1.3852,
+      "step": 3460
+    },
+    {
+      "epoch": 0.0045241652647113806,
+      "grad_norm": 2.0691568851470947,
+      "learning_rate": 3.2737737737737736e-05,
+      "loss": 1.8317,
+      "step": 3461
+    },
+    {
+      "epoch": 0.004525472449127651,
+      "grad_norm": 1.3251851797103882,
+      "learning_rate": 3.2732732732732736e-05,
+      "loss": 1.5431,
+      "step": 3462
+    },
+    {
+      "epoch": 0.00452677963354392,
+      "grad_norm": 1.3910210132598877,
+      "learning_rate": 3.272772772772773e-05,
+      "loss": 1.5923,
+      "step": 3463
+    },
+    {
+      "epoch": 0.004528086817960191,
+      "grad_norm": 1.3338592052459717,
+      "learning_rate": 3.272272272272273e-05,
+      "loss": 2.1364,
+      "step": 3464
+    },
+    {
+      "epoch": 0.004529394002376461,
+      "grad_norm": 1.3412494659423828,
+      "learning_rate": 3.271771771771772e-05,
+      "loss": 2.2985,
+      "step": 3465
+    },
+    {
+      "epoch": 0.004530701186792731,
+      "grad_norm": 1.5217441320419312,
+      "learning_rate": 3.2712712712712714e-05,
+      "loss": 2.0965,
+      "step": 3466
+    },
+    {
+      "epoch": 0.004532008371209002,
+      "grad_norm": 1.8436414003372192,
+      "learning_rate": 3.270770770770771e-05,
+      "loss": 1.7896,
+      "step": 3467
+    },
+    {
+      "epoch": 0.004533315555625272,
+      "grad_norm": 2.0944390296936035,
+      "learning_rate": 3.27027027027027e-05,
+      "loss": 1.7704,
+      "step": 3468
+    },
+    {
+      "epoch": 0.004534622740041542,
+      "grad_norm": 0.7646641731262207,
+      "learning_rate": 3.26976976976977e-05,
+      "loss": 0.5873,
+      "step": 3469
+    },
+    {
+      "epoch": 0.0045359299244578126,
+      "grad_norm": 1.2727854251861572,
+      "learning_rate": 3.269269269269269e-05,
+      "loss": 1.878,
+      "step": 3470
+    },
+    {
+      "epoch": 0.004537237108874083,
+      "grad_norm": 1.2774443626403809,
+      "learning_rate": 3.268768768768769e-05,
+      "loss": 1.0336,
+      "step": 3471
+    },
+    {
+      "epoch": 0.004538544293290353,
+      "grad_norm": 1.2159315347671509,
+      "learning_rate": 3.2682682682682686e-05,
+      "loss": 1.0907,
+      "step": 3472
+    },
+    {
+      "epoch": 0.0045398514777066235,
+      "grad_norm": 1.7308366298675537,
+      "learning_rate": 3.267767767767768e-05,
+      "loss": 2.1376,
+      "step": 3473
+    },
+    {
+      "epoch": 0.004541158662122894,
+      "grad_norm": 1.140066146850586,
+      "learning_rate": 3.267267267267267e-05,
+      "loss": 1.0983,
+      "step": 3474
+    },
+    {
+      "epoch": 0.004542465846539164,
+      "grad_norm": 1.2653149366378784,
+      "learning_rate": 3.2667667667667665e-05,
+      "loss": 1.1812,
+      "step": 3475
+    },
+    {
+      "epoch": 0.0045437730309554345,
+      "grad_norm": 1.255092740058899,
+      "learning_rate": 3.2662662662662665e-05,
+      "loss": 1.4035,
+      "step": 3476
+    },
+    {
+      "epoch": 0.004545080215371705,
+      "grad_norm": 1.206146240234375,
+      "learning_rate": 3.265765765765766e-05,
+      "loss": 1.0583,
+      "step": 3477
+    },
+    {
+      "epoch": 0.004546387399787975,
+      "grad_norm": 1.6553977727890015,
+      "learning_rate": 3.265265265265266e-05,
+      "loss": 2.0165,
+      "step": 3478
+    },
+    {
+      "epoch": 0.004547694584204245,
+      "grad_norm": 1.2937085628509521,
+      "learning_rate": 3.264764764764765e-05,
+      "loss": 1.6769,
+      "step": 3479
+    },
+    {
+      "epoch": 0.004549001768620515,
+      "grad_norm": 1.4032535552978516,
+      "learning_rate": 3.264264264264264e-05,
+      "loss": 1.5941,
+      "step": 3480
+    },
+    {
+      "epoch": 0.004550308953036785,
+      "grad_norm": 1.5537962913513184,
+      "learning_rate": 3.263763763763764e-05,
+      "loss": 1.8676,
+      "step": 3481
+    },
+    {
+      "epoch": 0.0045516161374530555,
+      "grad_norm": 1.173814058303833,
+      "learning_rate": 3.263263263263263e-05,
+      "loss": 1.1474,
+      "step": 3482
+    },
+    {
+      "epoch": 0.004552923321869326,
+      "grad_norm": 1.5229007005691528,
+      "learning_rate": 3.262762762762763e-05,
+      "loss": 1.3243,
+      "step": 3483
+    },
+    {
+      "epoch": 0.004554230506285596,
+      "grad_norm": 1.4470778703689575,
+      "learning_rate": 3.262262262262262e-05,
+      "loss": 1.3942,
+      "step": 3484
+    },
+    {
+      "epoch": 0.0045555376907018665,
+      "grad_norm": 1.4965757131576538,
+      "learning_rate": 3.261761761761762e-05,
+      "loss": 2.4147,
+      "step": 3485
+    },
+    {
+      "epoch": 0.004556844875118137,
+      "grad_norm": 1.1950100660324097,
+      "learning_rate": 3.2612612612612615e-05,
+      "loss": 1.7802,
+      "step": 3486
+    },
+    {
+      "epoch": 0.004558152059534407,
+      "grad_norm": 1.4015123844146729,
+      "learning_rate": 3.260760760760761e-05,
+      "loss": 1.0182,
+      "step": 3487
+    },
+    {
+      "epoch": 0.0045594592439506774,
+      "grad_norm": 1.5516818761825562,
+      "learning_rate": 3.260260260260261e-05,
+      "loss": 2.1813,
+      "step": 3488
+    },
+    {
+      "epoch": 0.004560766428366948,
+      "grad_norm": 1.3698395490646362,
+      "learning_rate": 3.25975975975976e-05,
+      "loss": 1.8472,
+      "step": 3489
+    },
+    {
+      "epoch": 0.004562073612783218,
+      "grad_norm": 1.3282825946807861,
+      "learning_rate": 3.25925925925926e-05,
+      "loss": 1.4206,
+      "step": 3490
+    },
+    {
+      "epoch": 0.004563380797199488,
+      "grad_norm": 0.13517466187477112,
+      "learning_rate": 3.2587587587587586e-05,
+      "loss": 0.0042,
+      "step": 3491
+    },
+    {
+      "epoch": 0.004564687981615759,
+      "grad_norm": 0.906393826007843,
+      "learning_rate": 3.258258258258258e-05,
+      "loss": 0.8476,
+      "step": 3492
+    },
+    {
+      "epoch": 0.004565995166032029,
+      "grad_norm": 0.5521675944328308,
+      "learning_rate": 3.257757757757758e-05,
+      "loss": 0.266,
+      "step": 3493
+    },
+    {
+      "epoch": 0.0045673023504482985,
+      "grad_norm": 0.5499749779701233,
+      "learning_rate": 3.257257257257257e-05,
+      "loss": 0.1884,
+      "step": 3494
+    },
+    {
+      "epoch": 0.004568609534864569,
+      "grad_norm": 1.4766079187393188,
+      "learning_rate": 3.256756756756757e-05,
+      "loss": 2.1978,
+      "step": 3495
+    },
+    {
+      "epoch": 0.004569916719280839,
+      "grad_norm": 1.052686333656311,
+      "learning_rate": 3.2562562562562565e-05,
+      "loss": 0.9017,
+      "step": 3496
+    },
+    {
+      "epoch": 0.0045712239036971095,
+      "grad_norm": 1.6090459823608398,
+      "learning_rate": 3.255755755755756e-05,
+      "loss": 1.3107,
+      "step": 3497
+    },
+    {
+      "epoch": 0.00457253108811338,
+      "grad_norm": 0.7887752056121826,
+      "learning_rate": 3.255255255255256e-05,
+      "loss": 0.2895,
+      "step": 3498
+    },
+    {
+      "epoch": 0.00457383827252965,
+      "grad_norm": 1.415866732597351,
+      "learning_rate": 3.2547547547547544e-05,
+      "loss": 1.8296,
+      "step": 3499
+    },
+    {
+      "epoch": 0.00457514545694592,
+      "grad_norm": 2.5468523502349854,
+      "learning_rate": 3.254254254254254e-05,
+      "loss": 0.8814,
+      "step": 3500
+    },
+    {
+      "epoch": 0.004576452641362191,
+      "grad_norm": 1.5901374816894531,
+      "learning_rate": 3.2537537537537536e-05,
+      "loss": 1.6674,
+      "step": 3501
+    },
+    {
+      "epoch": 0.004577759825778461,
+      "grad_norm": 1.4655040502548218,
+      "learning_rate": 3.2532532532532536e-05,
+      "loss": 1.0636,
+      "step": 3502
+    },
+    {
+      "epoch": 0.004579067010194731,
+      "grad_norm": 1.1705151796340942,
+      "learning_rate": 3.252752752752753e-05,
+      "loss": 1.404,
+      "step": 3503
+    },
+    {
+      "epoch": 0.004580374194611002,
+      "grad_norm": 0.05538543686270714,
+      "learning_rate": 3.252252252252252e-05,
+      "loss": 0.0017,
+      "step": 3504
+    },
+    {
+      "epoch": 0.004581681379027272,
+      "grad_norm": 1.3375270366668701,
+      "learning_rate": 3.251751751751752e-05,
+      "loss": 1.7335,
+      "step": 3505
+    },
+    {
+      "epoch": 0.004582988563443542,
+      "grad_norm": 1.8972514867782593,
+      "learning_rate": 3.2512512512512515e-05,
+      "loss": 2.7016,
+      "step": 3506
+    },
+    {
+      "epoch": 0.004584295747859813,
+      "grad_norm": 1.516108751296997,
+      "learning_rate": 3.250750750750751e-05,
+      "loss": 2.2544,
+      "step": 3507
+    },
+    {
+      "epoch": 0.004585602932276083,
+      "grad_norm": 1.5116583108901978,
+      "learning_rate": 3.25025025025025e-05,
+      "loss": 2.1838,
+      "step": 3508
+    },
+    {
+      "epoch": 0.004586910116692352,
+      "grad_norm": 1.7643729448318481,
+      "learning_rate": 3.24974974974975e-05,
+      "loss": 2.0868,
+      "step": 3509
+    },
+    {
+      "epoch": 0.004588217301108623,
+      "grad_norm": 1.5786253213882446,
+      "learning_rate": 3.2492492492492493e-05,
+      "loss": 1.3002,
+      "step": 3510
+    },
+    {
+      "epoch": 0.004589524485524893,
+      "grad_norm": 1.1401959657669067,
+      "learning_rate": 3.2487487487487486e-05,
+      "loss": 0.5973,
+      "step": 3511
+    },
+    {
+      "epoch": 0.004590831669941163,
+      "grad_norm": 1.2223362922668457,
+      "learning_rate": 3.2482482482482486e-05,
+      "loss": 1.2558,
+      "step": 3512
+    },
+    {
+      "epoch": 0.004592138854357434,
+      "grad_norm": 1.3012408018112183,
+      "learning_rate": 3.247747747747748e-05,
+      "loss": 0.677,
+      "step": 3513
+    },
+    {
+      "epoch": 0.004593446038773704,
+      "grad_norm": 1.3909122943878174,
+      "learning_rate": 3.247247247247248e-05,
+      "loss": 1.6644,
+      "step": 3514
+    },
+    {
+      "epoch": 0.004594753223189974,
+      "grad_norm": 2.256901264190674,
+      "learning_rate": 3.246746746746747e-05,
+      "loss": 2.5553,
+      "step": 3515
+    },
+    {
+      "epoch": 0.004596060407606245,
+      "grad_norm": 2.2257118225097656,
+      "learning_rate": 3.2462462462462465e-05,
+      "loss": 1.8497,
+      "step": 3516
+    },
+    {
+      "epoch": 0.004597367592022515,
+      "grad_norm": 1.582715630531311,
+      "learning_rate": 3.245745745745746e-05,
+      "loss": 1.9562,
+      "step": 3517
+    },
+    {
+      "epoch": 0.004598674776438785,
+      "grad_norm": 1.6081463098526,
+      "learning_rate": 3.245245245245245e-05,
+      "loss": 2.0058,
+      "step": 3518
+    },
+    {
+      "epoch": 0.004599981960855056,
+      "grad_norm": 1.0268296003341675,
+      "learning_rate": 3.244744744744745e-05,
+      "loss": 0.9628,
+      "step": 3519
+    },
+    {
+      "epoch": 0.004601289145271326,
+      "grad_norm": 1.3526577949523926,
+      "learning_rate": 3.2442442442442443e-05,
+      "loss": 1.2543,
+      "step": 3520
+    },
+    {
+      "epoch": 0.004602596329687596,
+      "grad_norm": 1.5042239427566528,
+      "learning_rate": 3.243743743743744e-05,
+      "loss": 1.9348,
+      "step": 3521
+    },
+    {
+      "epoch": 0.004603903514103867,
+      "grad_norm": 0.9436313509941101,
+      "learning_rate": 3.2432432432432436e-05,
+      "loss": 0.5382,
+      "step": 3522
+    },
+    {
+      "epoch": 0.004605210698520137,
+      "grad_norm": 1.2370840311050415,
+      "learning_rate": 3.242742742742743e-05,
+      "loss": 1.3594,
+      "step": 3523
+    },
+    {
+      "epoch": 0.004606517882936406,
+      "grad_norm": 1.7518489360809326,
+      "learning_rate": 3.242242242242242e-05,
+      "loss": 2.3725,
+      "step": 3524
+    },
+    {
+      "epoch": 0.004607825067352677,
+      "grad_norm": 1.3632359504699707,
+      "learning_rate": 3.2417417417417415e-05,
+      "loss": 1.9033,
+      "step": 3525
+    },
+    {
+      "epoch": 0.004609132251768947,
+      "grad_norm": 1.0352603197097778,
+      "learning_rate": 3.2412412412412415e-05,
+      "loss": 1.4626,
+      "step": 3526
+    },
+    {
+      "epoch": 0.004610439436185217,
+      "grad_norm": 1.4573428630828857,
+      "learning_rate": 3.240740740740741e-05,
+      "loss": 1.9619,
+      "step": 3527
+    },
+    {
+      "epoch": 0.004611746620601488,
+      "grad_norm": 1.4323865175247192,
+      "learning_rate": 3.240240240240241e-05,
+      "loss": 1.3638,
+      "step": 3528
+    },
+    {
+      "epoch": 0.004613053805017758,
+      "grad_norm": 1.2883960008621216,
+      "learning_rate": 3.23973973973974e-05,
+      "loss": 1.3254,
+      "step": 3529
+    },
+    {
+      "epoch": 0.004614360989434028,
+      "grad_norm": 1.414891242980957,
+      "learning_rate": 3.2392392392392394e-05,
+      "loss": 1.5099,
+      "step": 3530
+    },
+    {
+      "epoch": 0.004615668173850299,
+      "grad_norm": 1.8671427965164185,
+      "learning_rate": 3.238738738738739e-05,
+      "loss": 2.2505,
+      "step": 3531
+    },
+    {
+      "epoch": 0.004616975358266569,
+      "grad_norm": 1.387880563735962,
+      "learning_rate": 3.238238238238238e-05,
+      "loss": 1.294,
+      "step": 3532
+    },
+    {
+      "epoch": 0.004618282542682839,
+      "grad_norm": 1.739189624786377,
+      "learning_rate": 3.237737737737738e-05,
+      "loss": 1.7146,
+      "step": 3533
+    },
+    {
+      "epoch": 0.0046195897270991096,
+      "grad_norm": 1.6699799299240112,
+      "learning_rate": 3.237237237237237e-05,
+      "loss": 1.5322,
+      "step": 3534
+    },
+    {
+      "epoch": 0.00462089691151538,
+      "grad_norm": 1.3850643634796143,
+      "learning_rate": 3.2367367367367365e-05,
+      "loss": 1.4124,
+      "step": 3535
+    },
+    {
+      "epoch": 0.00462220409593165,
+      "grad_norm": 1.475756049156189,
+      "learning_rate": 3.2362362362362365e-05,
+      "loss": 1.7417,
+      "step": 3536
+    },
+    {
+      "epoch": 0.0046235112803479205,
+      "grad_norm": 1.7742390632629395,
+      "learning_rate": 3.235735735735736e-05,
+      "loss": 2.1823,
+      "step": 3537
+    },
+    {
+      "epoch": 0.004624818464764191,
+      "grad_norm": 1.3474591970443726,
+      "learning_rate": 3.235235235235236e-05,
+      "loss": 2.022,
+      "step": 3538
+    },
+    {
+      "epoch": 0.004626125649180461,
+      "grad_norm": 1.6687538623809814,
+      "learning_rate": 3.234734734734735e-05,
+      "loss": 1.6908,
+      "step": 3539
+    },
+    {
+      "epoch": 0.004627432833596731,
+      "grad_norm": 1.7065433263778687,
+      "learning_rate": 3.2342342342342344e-05,
+      "loss": 1.3587,
+      "step": 3540
+    },
+    {
+      "epoch": 0.004628740018013001,
+      "grad_norm": 1.5683473348617554,
+      "learning_rate": 3.2337337337337337e-05,
+      "loss": 2.0135,
+      "step": 3541
+    },
+    {
+      "epoch": 0.004630047202429271,
+      "grad_norm": 1.1581127643585205,
+      "learning_rate": 3.233233233233233e-05,
+      "loss": 1.3921,
+      "step": 3542
+    },
+    {
+      "epoch": 0.0046313543868455416,
+      "grad_norm": 1.6355133056640625,
+      "learning_rate": 3.232732732732733e-05,
+      "loss": 2.0211,
+      "step": 3543
+    },
+    {
+      "epoch": 0.004632661571261812,
+      "grad_norm": 1.376306414604187,
+      "learning_rate": 3.232232232232232e-05,
+      "loss": 1.4126,
+      "step": 3544
+    },
+    {
+      "epoch": 0.004633968755678082,
+      "grad_norm": 1.3422397375106812,
+      "learning_rate": 3.231731731731732e-05,
+      "loss": 1.8796,
+      "step": 3545
+    },
+    {
+      "epoch": 0.0046352759400943525,
+      "grad_norm": 1.2714289426803589,
+      "learning_rate": 3.2312312312312315e-05,
+      "loss": 1.6788,
+      "step": 3546
+    },
+    {
+      "epoch": 0.004636583124510623,
+      "grad_norm": 1.6274192333221436,
+      "learning_rate": 3.230730730730731e-05,
+      "loss": 1.5433,
+      "step": 3547
+    },
+    {
+      "epoch": 0.004637890308926893,
+      "grad_norm": 1.3880807161331177,
+      "learning_rate": 3.230230230230231e-05,
+      "loss": 1.9823,
+      "step": 3548
+    },
+    {
+      "epoch": 0.0046391974933431635,
+      "grad_norm": 1.8717379570007324,
+      "learning_rate": 3.2297297297297294e-05,
+      "loss": 1.9903,
+      "step": 3549
+    },
+    {
+      "epoch": 0.004640504677759434,
+      "grad_norm": 1.597456932067871,
+      "learning_rate": 3.2292292292292294e-05,
+      "loss": 2.4108,
+      "step": 3550
+    },
+    {
+      "epoch": 0.004641811862175704,
+      "grad_norm": 1.4068193435668945,
+      "learning_rate": 3.228728728728729e-05,
+      "loss": 1.4105,
+      "step": 3551
+    },
+    {
+      "epoch": 0.0046431190465919744,
+      "grad_norm": 1.40187406539917,
+      "learning_rate": 3.2282282282282286e-05,
+      "loss": 0.9164,
+      "step": 3552
+    },
+    {
+      "epoch": 0.004644426231008245,
+      "grad_norm": 2.1876447200775146,
+      "learning_rate": 3.227727727727728e-05,
+      "loss": 1.091,
+      "step": 3553
+    },
+    {
+      "epoch": 0.004645733415424515,
+      "grad_norm": 1.9667879343032837,
+      "learning_rate": 3.227227227227227e-05,
+      "loss": 1.9551,
+      "step": 3554
+    },
+    {
+      "epoch": 0.0046470405998407845,
+      "grad_norm": 1.5102306604385376,
+      "learning_rate": 3.226726726726727e-05,
+      "loss": 2.0046,
+      "step": 3555
+    },
+    {
+      "epoch": 0.004648347784257055,
+      "grad_norm": 1.2902199029922485,
+      "learning_rate": 3.2262262262262265e-05,
+      "loss": 1.3951,
+      "step": 3556
+    },
+    {
+      "epoch": 0.004649654968673325,
+      "grad_norm": 1.3399486541748047,
+      "learning_rate": 3.225725725725726e-05,
+      "loss": 1.261,
+      "step": 3557
+    },
+    {
+      "epoch": 0.0046509621530895955,
+      "grad_norm": 1.5210719108581543,
+      "learning_rate": 3.225225225225225e-05,
+      "loss": 1.8902,
+      "step": 3558
+    },
+    {
+      "epoch": 0.004652269337505866,
+      "grad_norm": 1.2718628644943237,
+      "learning_rate": 3.224724724724725e-05,
+      "loss": 2.3412,
+      "step": 3559
+    },
+    {
+      "epoch": 0.004653576521922136,
+      "grad_norm": 1.510949969291687,
+      "learning_rate": 3.2242242242242244e-05,
+      "loss": 1.5819,
+      "step": 3560
+    },
+    {
+      "epoch": 0.0046548837063384065,
+      "grad_norm": 1.675874948501587,
+      "learning_rate": 3.223723723723724e-05,
+      "loss": 2.3774,
+      "step": 3561
+    },
+    {
+      "epoch": 0.004656190890754677,
+      "grad_norm": 1.0188419818878174,
+      "learning_rate": 3.2232232232232236e-05,
+      "loss": 1.1733,
+      "step": 3562
+    },
+    {
+      "epoch": 0.004657498075170947,
+      "grad_norm": 1.365590214729309,
+      "learning_rate": 3.222722722722723e-05,
+      "loss": 0.9324,
+      "step": 3563
+    },
+    {
+      "epoch": 0.004658805259587217,
+      "grad_norm": 1.4135323762893677,
+      "learning_rate": 3.222222222222223e-05,
+      "loss": 1.287,
+      "step": 3564
+    },
+    {
+      "epoch": 0.004660112444003488,
+      "grad_norm": 1.0653365850448608,
+      "learning_rate": 3.221721721721722e-05,
+      "loss": 0.9668,
+      "step": 3565
+    },
+    {
+      "epoch": 0.004661419628419758,
+      "grad_norm": 0.10443060100078583,
+      "learning_rate": 3.221221221221221e-05,
+      "loss": 0.0032,
+      "step": 3566
+    },
+    {
+      "epoch": 0.004662726812836028,
+      "grad_norm": 1.355697512626648,
+      "learning_rate": 3.220720720720721e-05,
+      "loss": 1.0522,
+      "step": 3567
+    },
+    {
+      "epoch": 0.004664033997252299,
+      "grad_norm": 1.4275299310684204,
+      "learning_rate": 3.22022022022022e-05,
+      "loss": 1.1095,
+      "step": 3568
+    },
+    {
+      "epoch": 0.004665341181668569,
+      "grad_norm": 1.2596920728683472,
+      "learning_rate": 3.21971971971972e-05,
+      "loss": 1.3123,
+      "step": 3569
+    },
+    {
+      "epoch": 0.0046666483660848385,
+      "grad_norm": 1.2201354503631592,
+      "learning_rate": 3.2192192192192194e-05,
+      "loss": 1.8335,
+      "step": 3570
+    },
+    {
+      "epoch": 0.004667955550501109,
+      "grad_norm": 1.846868634223938,
+      "learning_rate": 3.218718718718719e-05,
+      "loss": 2.503,
+      "step": 3571
+    },
+    {
+      "epoch": 0.004669262734917379,
+      "grad_norm": 1.5024113655090332,
+      "learning_rate": 3.2182182182182187e-05,
+      "loss": 2.0594,
+      "step": 3572
+    },
+    {
+      "epoch": 0.004670569919333649,
+      "grad_norm": 1.4716944694519043,
+      "learning_rate": 3.217717717717718e-05,
+      "loss": 1.9545,
+      "step": 3573
+    },
+    {
+      "epoch": 0.00467187710374992,
+      "grad_norm": 1.7237690687179565,
+      "learning_rate": 3.217217217217217e-05,
+      "loss": 1.3138,
+      "step": 3574
+    },
+    {
+      "epoch": 0.00467318428816619,
+      "grad_norm": 1.5967888832092285,
+      "learning_rate": 3.2167167167167165e-05,
+      "loss": 1.9015,
+      "step": 3575
+    },
+    {
+      "epoch": 0.00467449147258246,
+      "grad_norm": 1.5429633855819702,
+      "learning_rate": 3.2162162162162165e-05,
+      "loss": 1.1042,
+      "step": 3576
+    },
+    {
+      "epoch": 0.004675798656998731,
+      "grad_norm": 1.5322041511535645,
+      "learning_rate": 3.215715715715716e-05,
+      "loss": 1.2036,
+      "step": 3577
+    },
+    {
+      "epoch": 0.004677105841415001,
+      "grad_norm": 1.1402878761291504,
+      "learning_rate": 3.215215215215215e-05,
+      "loss": 1.2375,
+      "step": 3578
+    },
+    {
+      "epoch": 0.004678413025831271,
+      "grad_norm": 1.4293047189712524,
+      "learning_rate": 3.214714714714715e-05,
+      "loss": 1.2667,
+      "step": 3579
+    },
+    {
+      "epoch": 0.004679720210247542,
+      "grad_norm": 1.0040302276611328,
+      "learning_rate": 3.2142142142142144e-05,
+      "loss": 1.1114,
+      "step": 3580
+    },
+    {
+      "epoch": 0.004681027394663812,
+      "grad_norm": 1.0025014877319336,
+      "learning_rate": 3.2137137137137144e-05,
+      "loss": 0.4373,
+      "step": 3581
+    },
+    {
+      "epoch": 0.004682334579080082,
+      "grad_norm": 1.371025562286377,
+      "learning_rate": 3.213213213213213e-05,
+      "loss": 1.2502,
+      "step": 3582
+    },
+    {
+      "epoch": 0.004683641763496353,
+      "grad_norm": 2.119929075241089,
+      "learning_rate": 3.212712712712713e-05,
+      "loss": 1.5517,
+      "step": 3583
+    },
+    {
+      "epoch": 0.004684948947912623,
+      "grad_norm": 1.9257744550704956,
+      "learning_rate": 3.212212212212212e-05,
+      "loss": 2.072,
+      "step": 3584
+    },
+    {
+      "epoch": 0.004686256132328893,
+      "grad_norm": 1.2487419843673706,
+      "learning_rate": 3.2117117117117116e-05,
+      "loss": 2.1152,
+      "step": 3585
+    },
+    {
+      "epoch": 0.004687563316745163,
+      "grad_norm": 1.3217498064041138,
+      "learning_rate": 3.2112112112112115e-05,
+      "loss": 1.0878,
+      "step": 3586
+    },
+    {
+      "epoch": 0.004688870501161433,
+      "grad_norm": 1.6573771238327026,
+      "learning_rate": 3.210710710710711e-05,
+      "loss": 1.7779,
+      "step": 3587
+    },
+    {
+      "epoch": 0.004690177685577703,
+      "grad_norm": 0.759430468082428,
+      "learning_rate": 3.210210210210211e-05,
+      "loss": 0.4198,
+      "step": 3588
+    },
+    {
+      "epoch": 0.004691484869993974,
+      "grad_norm": 1.3947277069091797,
+      "learning_rate": 3.20970970970971e-05,
+      "loss": 1.327,
+      "step": 3589
+    },
+    {
+      "epoch": 0.004692792054410244,
+      "grad_norm": 1.5001299381256104,
+      "learning_rate": 3.2092092092092094e-05,
+      "loss": 1.3219,
+      "step": 3590
+    },
+    {
+      "epoch": 0.004694099238826514,
+      "grad_norm": 2.0483205318450928,
+      "learning_rate": 3.208708708708709e-05,
+      "loss": 1.928,
+      "step": 3591
+    },
+    {
+      "epoch": 0.004695406423242785,
+      "grad_norm": 1.5803836584091187,
+      "learning_rate": 3.208208208208208e-05,
+      "loss": 1.4387,
+      "step": 3592
+    },
+    {
+      "epoch": 0.004696713607659055,
+      "grad_norm": 2.045901298522949,
+      "learning_rate": 3.207707707707708e-05,
+      "loss": 1.7783,
+      "step": 3593
+    },
+    {
+      "epoch": 0.004698020792075325,
+      "grad_norm": 1.3705579042434692,
+      "learning_rate": 3.207207207207207e-05,
+      "loss": 1.5561,
+      "step": 3594
+    },
+    {
+      "epoch": 0.004699327976491596,
+      "grad_norm": 1.4956648349761963,
+      "learning_rate": 3.206706706706707e-05,
+      "loss": 1.3289,
+      "step": 3595
+    },
+    {
+      "epoch": 0.004700635160907866,
+      "grad_norm": 1.0795695781707764,
+      "learning_rate": 3.2062062062062065e-05,
+      "loss": 1.2993,
+      "step": 3596
+    },
+    {
+      "epoch": 0.004701942345324136,
+      "grad_norm": 1.5036541223526,
+      "learning_rate": 3.205705705705706e-05,
+      "loss": 1.8776,
+      "step": 3597
+    },
+    {
+      "epoch": 0.0047032495297404065,
+      "grad_norm": 1.3187249898910522,
+      "learning_rate": 3.205205205205206e-05,
+      "loss": 1.9673,
+      "step": 3598
+    },
+    {
+      "epoch": 0.004704556714156677,
+      "grad_norm": 1.795992374420166,
+      "learning_rate": 3.2047047047047044e-05,
+      "loss": 2.3845,
+      "step": 3599
+    },
+    {
+      "epoch": 0.004705863898572947,
+      "grad_norm": 1.8545953035354614,
+      "learning_rate": 3.2042042042042044e-05,
+      "loss": 1.6007,
+      "step": 3600
+    },
+    {
+      "epoch": 0.004707171082989217,
+      "grad_norm": 1.5052769184112549,
+      "learning_rate": 3.203703703703704e-05,
+      "loss": 1.8397,
+      "step": 3601
+    },
+    {
+      "epoch": 0.004708478267405487,
+      "grad_norm": 1.6787859201431274,
+      "learning_rate": 3.203203203203203e-05,
+      "loss": 1.3834,
+      "step": 3602
+    },
+    {
+      "epoch": 0.004709785451821757,
+      "grad_norm": 1.3500570058822632,
+      "learning_rate": 3.202702702702703e-05,
+      "loss": 2.0835,
+      "step": 3603
+    },
+    {
+      "epoch": 0.004711092636238028,
+      "grad_norm": 1.5070812702178955,
+      "learning_rate": 3.202202202202202e-05,
+      "loss": 1.8374,
+      "step": 3604
+    },
+    {
+      "epoch": 0.004712399820654298,
+      "grad_norm": 1.3793407678604126,
+      "learning_rate": 3.201701701701702e-05,
+      "loss": 1.2709,
+      "step": 3605
+    },
+    {
+      "epoch": 0.004713707005070568,
+      "grad_norm": 1.1693944931030273,
+      "learning_rate": 3.2012012012012015e-05,
+      "loss": 1.6078,
+      "step": 3606
+    },
+    {
+      "epoch": 0.0047150141894868386,
+      "grad_norm": 1.7247902154922485,
+      "learning_rate": 3.200700700700701e-05,
+      "loss": 1.6249,
+      "step": 3607
+    },
+    {
+      "epoch": 0.004716321373903109,
+      "grad_norm": 1.0528442859649658,
+      "learning_rate": 3.2002002002002e-05,
+      "loss": 1.1878,
+      "step": 3608
+    },
+    {
+      "epoch": 0.004717628558319379,
+      "grad_norm": 1.234477162361145,
+      "learning_rate": 3.1996996996996994e-05,
+      "loss": 1.7553,
+      "step": 3609
+    },
+    {
+      "epoch": 0.0047189357427356495,
+      "grad_norm": 2.0904176235198975,
+      "learning_rate": 3.1991991991991994e-05,
+      "loss": 1.2615,
+      "step": 3610
+    },
+    {
+      "epoch": 0.00472024292715192,
+      "grad_norm": 1.2952682971954346,
+      "learning_rate": 3.198698698698699e-05,
+      "loss": 1.9065,
+      "step": 3611
+    },
+    {
+      "epoch": 0.00472155011156819,
+      "grad_norm": 1.2490006685256958,
+      "learning_rate": 3.198198198198199e-05,
+      "loss": 1.2951,
+      "step": 3612
+    },
+    {
+      "epoch": 0.0047228572959844605,
+      "grad_norm": 1.257523536682129,
+      "learning_rate": 3.197697697697698e-05,
+      "loss": 0.7895,
+      "step": 3613
+    },
+    {
+      "epoch": 0.004724164480400731,
+      "grad_norm": 1.4111295938491821,
+      "learning_rate": 3.197197197197197e-05,
+      "loss": 1.8755,
+      "step": 3614
+    },
+    {
+      "epoch": 0.004725471664817001,
+      "grad_norm": 1.5956237316131592,
+      "learning_rate": 3.1966966966966966e-05,
+      "loss": 1.6004,
+      "step": 3615
+    },
+    {
+      "epoch": 0.004726778849233271,
+      "grad_norm": 1.302992343902588,
+      "learning_rate": 3.196196196196196e-05,
+      "loss": 1.6595,
+      "step": 3616
+    },
+    {
+      "epoch": 0.004728086033649541,
+      "grad_norm": 1.135660171508789,
+      "learning_rate": 3.195695695695696e-05,
+      "loss": 0.7909,
+      "step": 3617
+    },
+    {
+      "epoch": 0.004729393218065811,
+      "grad_norm": 1.2870028018951416,
+      "learning_rate": 3.195195195195195e-05,
+      "loss": 2.1736,
+      "step": 3618
+    },
+    {
+      "epoch": 0.0047307004024820815,
+      "grad_norm": 1.578142523765564,
+      "learning_rate": 3.194694694694695e-05,
+      "loss": 1.8881,
+      "step": 3619
+    },
+    {
+      "epoch": 0.004732007586898352,
+      "grad_norm": 1.1805518865585327,
+      "learning_rate": 3.1941941941941944e-05,
+      "loss": 1.0268,
+      "step": 3620
+    },
+    {
+      "epoch": 0.004733314771314622,
+      "grad_norm": 1.5735121965408325,
+      "learning_rate": 3.193693693693694e-05,
+      "loss": 2.8798,
+      "step": 3621
+    },
+    {
+      "epoch": 0.0047346219557308925,
+      "grad_norm": 1.1697373390197754,
+      "learning_rate": 3.193193193193194e-05,
+      "loss": 1.304,
+      "step": 3622
+    },
+    {
+      "epoch": 0.004735929140147163,
+      "grad_norm": 1.3963468074798584,
+      "learning_rate": 3.192692692692693e-05,
+      "loss": 1.3128,
+      "step": 3623
+    },
+    {
+      "epoch": 0.004737236324563433,
+      "grad_norm": 1.4289218187332153,
+      "learning_rate": 3.192192192192192e-05,
+      "loss": 1.2493,
+      "step": 3624
+    },
+    {
+      "epoch": 0.0047385435089797034,
+      "grad_norm": 1.563373327255249,
+      "learning_rate": 3.1916916916916916e-05,
+      "loss": 1.5941,
+      "step": 3625
+    },
+    {
+      "epoch": 0.004739850693395974,
+      "grad_norm": 1.355543613433838,
+      "learning_rate": 3.1911911911911916e-05,
+      "loss": 1.5227,
+      "step": 3626
+    },
+    {
+      "epoch": 0.004741157877812244,
+      "grad_norm": 1.559887170791626,
+      "learning_rate": 3.190690690690691e-05,
+      "loss": 1.7534,
+      "step": 3627
+    },
+    {
+      "epoch": 0.004742465062228514,
+      "grad_norm": 1.4405226707458496,
+      "learning_rate": 3.19019019019019e-05,
+      "loss": 1.7339,
+      "step": 3628
+    },
+    {
+      "epoch": 0.004743772246644785,
+      "grad_norm": 1.5233449935913086,
+      "learning_rate": 3.18968968968969e-05,
+      "loss": 1.9298,
+      "step": 3629
+    },
+    {
+      "epoch": 0.004745079431061055,
+      "grad_norm": 1.7641301155090332,
+      "learning_rate": 3.1891891891891894e-05,
+      "loss": 2.0674,
+      "step": 3630
+    },
+    {
+      "epoch": 0.0047463866154773245,
+      "grad_norm": 1.5913950204849243,
+      "learning_rate": 3.1886886886886894e-05,
+      "loss": 1.7385,
+      "step": 3631
+    },
+    {
+      "epoch": 0.004747693799893595,
+      "grad_norm": 1.918821930885315,
+      "learning_rate": 3.188188188188188e-05,
+      "loss": 1.9494,
+      "step": 3632
+    },
+    {
+      "epoch": 0.004749000984309865,
+      "grad_norm": 1.1846126317977905,
+      "learning_rate": 3.187687687687688e-05,
+      "loss": 0.5799,
+      "step": 3633
+    },
+    {
+      "epoch": 0.0047503081687261355,
+      "grad_norm": 1.2971302270889282,
+      "learning_rate": 3.187187187187187e-05,
+      "loss": 1.1448,
+      "step": 3634
+    },
+    {
+      "epoch": 0.004751615353142406,
+      "grad_norm": 1.4343546628952026,
+      "learning_rate": 3.1866866866866866e-05,
+      "loss": 1.1654,
+      "step": 3635
+    },
+    {
+      "epoch": 0.004752922537558676,
+      "grad_norm": 1.3027766942977905,
+      "learning_rate": 3.1861861861861866e-05,
+      "loss": 1.7297,
+      "step": 3636
+    },
+    {
+      "epoch": 0.004754229721974946,
+      "grad_norm": 1.6701722145080566,
+      "learning_rate": 3.185685685685686e-05,
+      "loss": 2.4018,
+      "step": 3637
+    },
+    {
+      "epoch": 0.004755536906391217,
+      "grad_norm": 1.5261790752410889,
+      "learning_rate": 3.185185185185185e-05,
+      "loss": 1.7573,
+      "step": 3638
+    },
+    {
+      "epoch": 0.004756844090807487,
+      "grad_norm": 1.0334054231643677,
+      "learning_rate": 3.184684684684685e-05,
+      "loss": 1.0505,
+      "step": 3639
+    },
+    {
+      "epoch": 0.004758151275223757,
+      "grad_norm": 1.7205411195755005,
+      "learning_rate": 3.184184184184184e-05,
+      "loss": 1.6653,
+      "step": 3640
+    },
+    {
+      "epoch": 0.004759458459640028,
+      "grad_norm": 2.0306613445281982,
+      "learning_rate": 3.183683683683684e-05,
+      "loss": 1.8649,
+      "step": 3641
+    },
+    {
+      "epoch": 0.004760765644056298,
+      "grad_norm": 1.4870742559432983,
+      "learning_rate": 3.183183183183183e-05,
+      "loss": 1.248,
+      "step": 3642
+    },
+    {
+      "epoch": 0.004762072828472568,
+      "grad_norm": 1.8396729230880737,
+      "learning_rate": 3.182682682682683e-05,
+      "loss": 1.3978,
+      "step": 3643
+    },
+    {
+      "epoch": 0.004763380012888839,
+      "grad_norm": 1.380837082862854,
+      "learning_rate": 3.182182182182182e-05,
+      "loss": 2.1392,
+      "step": 3644
+    },
+    {
+      "epoch": 0.004764687197305109,
+      "grad_norm": 1.2719833850860596,
+      "learning_rate": 3.1816816816816816e-05,
+      "loss": 1.4195,
+      "step": 3645
+    },
+    {
+      "epoch": 0.004765994381721379,
+      "grad_norm": 1.3935086727142334,
+      "learning_rate": 3.1811811811811816e-05,
+      "loss": 1.043,
+      "step": 3646
+    },
+    {
+      "epoch": 0.004767301566137649,
+      "grad_norm": 1.6418629884719849,
+      "learning_rate": 3.180680680680681e-05,
+      "loss": 1.8635,
+      "step": 3647
+    },
+    {
+      "epoch": 0.004768608750553919,
+      "grad_norm": 1.3307095766067505,
+      "learning_rate": 3.180180180180181e-05,
+      "loss": 1.0823,
+      "step": 3648
+    },
+    {
+      "epoch": 0.004769915934970189,
+      "grad_norm": 1.3457857370376587,
+      "learning_rate": 3.1796796796796795e-05,
+      "loss": 1.2447,
+      "step": 3649
+    },
+    {
+      "epoch": 0.00477122311938646,
+      "grad_norm": 1.6677151918411255,
+      "learning_rate": 3.1791791791791794e-05,
+      "loss": 1.3809,
+      "step": 3650
+    },
+    {
+      "epoch": 0.00477253030380273,
+      "grad_norm": 1.169090986251831,
+      "learning_rate": 3.178678678678679e-05,
+      "loss": 1.66,
+      "step": 3651
+    },
+    {
+      "epoch": 0.004773837488219,
+      "grad_norm": 1.3457608222961426,
+      "learning_rate": 3.178178178178178e-05,
+      "loss": 2.1165,
+      "step": 3652
+    },
+    {
+      "epoch": 0.004775144672635271,
+      "grad_norm": 1.3707876205444336,
+      "learning_rate": 3.177677677677678e-05,
+      "loss": 1.6441,
+      "step": 3653
+    },
+    {
+      "epoch": 0.004776451857051541,
+      "grad_norm": 1.6748955249786377,
+      "learning_rate": 3.177177177177177e-05,
+      "loss": 1.1621,
+      "step": 3654
+    },
+    {
+      "epoch": 0.004777759041467811,
+      "grad_norm": 1.5809892416000366,
+      "learning_rate": 3.176676676676677e-05,
+      "loss": 2.0714,
+      "step": 3655
+    },
+    {
+      "epoch": 0.004779066225884082,
+      "grad_norm": 1.1592813730239868,
+      "learning_rate": 3.1761761761761766e-05,
+      "loss": 1.115,
+      "step": 3656
+    },
+    {
+      "epoch": 0.004780373410300352,
+      "grad_norm": 1.308276891708374,
+      "learning_rate": 3.175675675675676e-05,
+      "loss": 1.2811,
+      "step": 3657
+    },
+    {
+      "epoch": 0.004781680594716622,
+      "grad_norm": 1.197853684425354,
+      "learning_rate": 3.175175175175175e-05,
+      "loss": 0.5902,
+      "step": 3658
+    },
+    {
+      "epoch": 0.004782987779132893,
+      "grad_norm": 1.8574713468551636,
+      "learning_rate": 3.1746746746746745e-05,
+      "loss": 0.6372,
+      "step": 3659
+    },
+    {
+      "epoch": 0.004784294963549163,
+      "grad_norm": 1.5223958492279053,
+      "learning_rate": 3.1741741741741744e-05,
+      "loss": 1.1674,
+      "step": 3660
+    },
+    {
+      "epoch": 0.004785602147965433,
+      "grad_norm": 1.2148301601409912,
+      "learning_rate": 3.173673673673674e-05,
+      "loss": 0.8975,
+      "step": 3661
+    },
+    {
+      "epoch": 0.004786909332381703,
+      "grad_norm": 1.20482337474823,
+      "learning_rate": 3.173173173173174e-05,
+      "loss": 1.448,
+      "step": 3662
+    },
+    {
+      "epoch": 0.004788216516797973,
+      "grad_norm": 1.188102126121521,
+      "learning_rate": 3.172672672672673e-05,
+      "loss": 0.8226,
+      "step": 3663
+    },
+    {
+      "epoch": 0.004789523701214243,
+      "grad_norm": 1.66988205909729,
+      "learning_rate": 3.172172172172172e-05,
+      "loss": 1.573,
+      "step": 3664
+    },
+    {
+      "epoch": 0.004790830885630514,
+      "grad_norm": 1.4177364110946655,
+      "learning_rate": 3.1716716716716716e-05,
+      "loss": 1.8125,
+      "step": 3665
+    },
+    {
+      "epoch": 0.004792138070046784,
+      "grad_norm": 1.2240238189697266,
+      "learning_rate": 3.171171171171171e-05,
+      "loss": 1.6531,
+      "step": 3666
+    },
+    {
+      "epoch": 0.004793445254463054,
+      "grad_norm": 3.8797104358673096,
+      "learning_rate": 3.170670670670671e-05,
+      "loss": 0.6702,
+      "step": 3667
+    },
+    {
+      "epoch": 0.004794752438879325,
+      "grad_norm": 1.4226473569869995,
+      "learning_rate": 3.17017017017017e-05,
+      "loss": 1.8049,
+      "step": 3668
+    },
+    {
+      "epoch": 0.004796059623295595,
+      "grad_norm": 1.0726206302642822,
+      "learning_rate": 3.16966966966967e-05,
+      "loss": 1.1429,
+      "step": 3669
+    },
+    {
+      "epoch": 0.004797366807711865,
+      "grad_norm": 1.2767502069473267,
+      "learning_rate": 3.1691691691691695e-05,
+      "loss": 1.4751,
+      "step": 3670
+    },
+    {
+      "epoch": 0.0047986739921281356,
+      "grad_norm": 1.7479175329208374,
+      "learning_rate": 3.168668668668669e-05,
+      "loss": 2.0488,
+      "step": 3671
+    },
+    {
+      "epoch": 0.004799981176544406,
+      "grad_norm": 1.8567023277282715,
+      "learning_rate": 3.168168168168169e-05,
+      "loss": 1.838,
+      "step": 3672
+    },
+    {
+      "epoch": 0.004801288360960676,
+      "grad_norm": 1.4364548921585083,
+      "learning_rate": 3.167667667667668e-05,
+      "loss": 1.4075,
+      "step": 3673
+    },
+    {
+      "epoch": 0.0048025955453769465,
+      "grad_norm": 1.7608340978622437,
+      "learning_rate": 3.167167167167167e-05,
+      "loss": 2.6817,
+      "step": 3674
+    },
+    {
+      "epoch": 0.004803902729793217,
+      "grad_norm": 1.4127233028411865,
+      "learning_rate": 3.1666666666666666e-05,
+      "loss": 2.2005,
+      "step": 3675
+    },
+    {
+      "epoch": 0.004805209914209487,
+      "grad_norm": 0.3165508508682251,
+      "learning_rate": 3.166166166166166e-05,
+      "loss": 0.0099,
+      "step": 3676
+    },
+    {
+      "epoch": 0.004806517098625757,
+      "grad_norm": 1.2403323650360107,
+      "learning_rate": 3.165665665665666e-05,
+      "loss": 1.4472,
+      "step": 3677
+    },
+    {
+      "epoch": 0.004807824283042027,
+      "grad_norm": 1.2452306747436523,
+      "learning_rate": 3.165165165165165e-05,
+      "loss": 1.9553,
+      "step": 3678
+    },
+    {
+      "epoch": 0.004809131467458297,
+      "grad_norm": 1.3816317319869995,
+      "learning_rate": 3.164664664664665e-05,
+      "loss": 2.1776,
+      "step": 3679
+    },
+    {
+      "epoch": 0.0048104386518745676,
+      "grad_norm": 1.5297333002090454,
+      "learning_rate": 3.1641641641641645e-05,
+      "loss": 1.9598,
+      "step": 3680
+    },
+    {
+      "epoch": 0.004811745836290838,
+      "grad_norm": 1.2205753326416016,
+      "learning_rate": 3.163663663663664e-05,
+      "loss": 1.242,
+      "step": 3681
+    },
+    {
+      "epoch": 0.004813053020707108,
+      "grad_norm": 1.56426203250885,
+      "learning_rate": 3.163163163163163e-05,
+      "loss": 1.8114,
+      "step": 3682
+    },
+    {
+      "epoch": 0.0048143602051233785,
+      "grad_norm": 1.4199020862579346,
+      "learning_rate": 3.1626626626626623e-05,
+      "loss": 0.9275,
+      "step": 3683
+    },
+    {
+      "epoch": 0.004815667389539649,
+      "grad_norm": 1.230777621269226,
+      "learning_rate": 3.162162162162162e-05,
+      "loss": 1.6598,
+      "step": 3684
+    },
+    {
+      "epoch": 0.004816974573955919,
+      "grad_norm": 1.2032729387283325,
+      "learning_rate": 3.1616616616616616e-05,
+      "loss": 0.9291,
+      "step": 3685
+    },
+    {
+      "epoch": 0.0048182817583721895,
+      "grad_norm": 1.23746919631958,
+      "learning_rate": 3.1611611611611616e-05,
+      "loss": 1.0412,
+      "step": 3686
+    },
+    {
+      "epoch": 0.00481958894278846,
+      "grad_norm": 0.8638294339179993,
+      "learning_rate": 3.160660660660661e-05,
+      "loss": 0.5696,
+      "step": 3687
+    },
+    {
+      "epoch": 0.00482089612720473,
+      "grad_norm": 1.9952846765518188,
+      "learning_rate": 3.16016016016016e-05,
+      "loss": 1.8462,
+      "step": 3688
+    },
+    {
+      "epoch": 0.0048222033116210004,
+      "grad_norm": 1.2512773275375366,
+      "learning_rate": 3.15965965965966e-05,
+      "loss": 1.6932,
+      "step": 3689
+    },
+    {
+      "epoch": 0.004823510496037271,
+      "grad_norm": 1.3405901193618774,
+      "learning_rate": 3.159159159159159e-05,
+      "loss": 1.5655,
+      "step": 3690
+    },
+    {
+      "epoch": 0.004824817680453541,
+      "grad_norm": 1.4530911445617676,
+      "learning_rate": 3.158658658658659e-05,
+      "loss": 1.5605,
+      "step": 3691
+    },
+    {
+      "epoch": 0.004826124864869811,
+      "grad_norm": 1.6099843978881836,
+      "learning_rate": 3.158158158158158e-05,
+      "loss": 1.8457,
+      "step": 3692
+    },
+    {
+      "epoch": 0.004827432049286081,
+      "grad_norm": 2.0444841384887695,
+      "learning_rate": 3.157657657657658e-05,
+      "loss": 1.4762,
+      "step": 3693
+    },
+    {
+      "epoch": 0.004828739233702351,
+      "grad_norm": 1.3107658624649048,
+      "learning_rate": 3.157157157157157e-05,
+      "loss": 1.5953,
+      "step": 3694
+    },
+    {
+      "epoch": 0.0048300464181186215,
+      "grad_norm": 1.321577548980713,
+      "learning_rate": 3.1566566566566566e-05,
+      "loss": 1.5414,
+      "step": 3695
+    },
+    {
+      "epoch": 0.004831353602534892,
+      "grad_norm": 1.3919023275375366,
+      "learning_rate": 3.1561561561561566e-05,
+      "loss": 1.47,
+      "step": 3696
+    },
+    {
+      "epoch": 0.004832660786951162,
+      "grad_norm": 0.9926742315292358,
+      "learning_rate": 3.155655655655656e-05,
+      "loss": 0.7795,
+      "step": 3697
+    },
+    {
+      "epoch": 0.0048339679713674324,
+      "grad_norm": 1.6390317678451538,
+      "learning_rate": 3.155155155155155e-05,
+      "loss": 2.107,
+      "step": 3698
+    },
+    {
+      "epoch": 0.004835275155783703,
+      "grad_norm": 1.1866517066955566,
+      "learning_rate": 3.1546546546546545e-05,
+      "loss": 1.1853,
+      "step": 3699
+    },
+    {
+      "epoch": 0.004836582340199973,
+      "grad_norm": 1.3226901292800903,
+      "learning_rate": 3.1541541541541545e-05,
+      "loss": 0.849,
+      "step": 3700
+    },
+    {
+      "epoch": 0.004837889524616243,
+      "grad_norm": 1.383776307106018,
+      "learning_rate": 3.153653653653654e-05,
+      "loss": 1.7018,
+      "step": 3701
+    },
+    {
+      "epoch": 0.004839196709032514,
+      "grad_norm": 1.0069390535354614,
+      "learning_rate": 3.153153153153153e-05,
+      "loss": 1.0944,
+      "step": 3702
+    },
+    {
+      "epoch": 0.004840503893448784,
+      "grad_norm": 1.4307210445404053,
+      "learning_rate": 3.152652652652653e-05,
+      "loss": 1.5844,
+      "step": 3703
+    },
+    {
+      "epoch": 0.004841811077865054,
+      "grad_norm": 1.555961012840271,
+      "learning_rate": 3.152152152152152e-05,
+      "loss": 1.5134,
+      "step": 3704
+    },
+    {
+      "epoch": 0.004843118262281325,
+      "grad_norm": 1.47184157371521,
+      "learning_rate": 3.151651651651652e-05,
+      "loss": 1.8004,
+      "step": 3705
+    },
+    {
+      "epoch": 0.004844425446697595,
+      "grad_norm": 1.3547722101211548,
+      "learning_rate": 3.1511511511511516e-05,
+      "loss": 1.0907,
+      "step": 3706
+    },
+    {
+      "epoch": 0.004845732631113865,
+      "grad_norm": 1.2149510383605957,
+      "learning_rate": 3.150650650650651e-05,
+      "loss": 1.4492,
+      "step": 3707
+    },
+    {
+      "epoch": 0.004847039815530135,
+      "grad_norm": 0.9765141606330872,
+      "learning_rate": 3.15015015015015e-05,
+      "loss": 0.9773,
+      "step": 3708
+    },
+    {
+      "epoch": 0.004848346999946405,
+      "grad_norm": 1.2797836065292358,
+      "learning_rate": 3.1496496496496495e-05,
+      "loss": 1.6592,
+      "step": 3709
+    },
+    {
+      "epoch": 0.004849654184362675,
+      "grad_norm": 1.8713387250900269,
+      "learning_rate": 3.1491491491491495e-05,
+      "loss": 1.0499,
+      "step": 3710
+    },
+    {
+      "epoch": 0.004850961368778946,
+      "grad_norm": 1.3799238204956055,
+      "learning_rate": 3.148648648648649e-05,
+      "loss": 1.8018,
+      "step": 3711
+    },
+    {
+      "epoch": 0.004852268553195216,
+      "grad_norm": 1.3621039390563965,
+      "learning_rate": 3.148148148148148e-05,
+      "loss": 1.6709,
+      "step": 3712
+    },
+    {
+      "epoch": 0.004853575737611486,
+      "grad_norm": 1.3098262548446655,
+      "learning_rate": 3.147647647647648e-05,
+      "loss": 1.834,
+      "step": 3713
+    },
+    {
+      "epoch": 0.004854882922027757,
+      "grad_norm": 1.3079845905303955,
+      "learning_rate": 3.1471471471471473e-05,
+      "loss": 0.872,
+      "step": 3714
+    },
+    {
+      "epoch": 0.004856190106444027,
+      "grad_norm": 1.3932185173034668,
+      "learning_rate": 3.1466466466466466e-05,
+      "loss": 1.6597,
+      "step": 3715
+    },
+    {
+      "epoch": 0.004857497290860297,
+      "grad_norm": 1.3278216123580933,
+      "learning_rate": 3.146146146146146e-05,
+      "loss": 2.0909,
+      "step": 3716
+    },
+    {
+      "epoch": 0.004858804475276568,
+      "grad_norm": 1.4096519947052002,
+      "learning_rate": 3.145645645645646e-05,
+      "loss": 0.7408,
+      "step": 3717
+    },
+    {
+      "epoch": 0.004860111659692838,
+      "grad_norm": 1.1825799942016602,
+      "learning_rate": 3.145145145145145e-05,
+      "loss": 1.1875,
+      "step": 3718
+    },
+    {
+      "epoch": 0.004861418844109108,
+      "grad_norm": 1.4682544469833374,
+      "learning_rate": 3.1446446446446445e-05,
+      "loss": 1.3739,
+      "step": 3719
+    },
+    {
+      "epoch": 0.004862726028525379,
+      "grad_norm": 1.7104785442352295,
+      "learning_rate": 3.1441441441441445e-05,
+      "loss": 1.7796,
+      "step": 3720
+    },
+    {
+      "epoch": 0.004864033212941649,
+      "grad_norm": 1.4157119989395142,
+      "learning_rate": 3.143643643643644e-05,
+      "loss": 1.901,
+      "step": 3721
+    },
+    {
+      "epoch": 0.004865340397357919,
+      "grad_norm": 1.4188841581344604,
+      "learning_rate": 3.143143143143144e-05,
+      "loss": 1.3268,
+      "step": 3722
+    },
+    {
+      "epoch": 0.004866647581774189,
+      "grad_norm": 1.194572925567627,
+      "learning_rate": 3.1426426426426424e-05,
+      "loss": 1.6056,
+      "step": 3723
+    },
+    {
+      "epoch": 0.004867954766190459,
+      "grad_norm": 1.4595437049865723,
+      "learning_rate": 3.1421421421421424e-05,
+      "loss": 1.6247,
+      "step": 3724
+    },
+    {
+      "epoch": 0.004869261950606729,
+      "grad_norm": 1.4265269041061401,
+      "learning_rate": 3.1416416416416416e-05,
+      "loss": 1.7015,
+      "step": 3725
+    },
+    {
+      "epoch": 0.004870569135023,
+      "grad_norm": 1.2855182886123657,
+      "learning_rate": 3.141141141141141e-05,
+      "loss": 1.7,
+      "step": 3726
+    },
+    {
+      "epoch": 0.00487187631943927,
+      "grad_norm": 1.772713541984558,
+      "learning_rate": 3.140640640640641e-05,
+      "loss": 1.9666,
+      "step": 3727
+    },
+    {
+      "epoch": 0.00487318350385554,
+      "grad_norm": 1.5232847929000854,
+      "learning_rate": 3.14014014014014e-05,
+      "loss": 2.5218,
+      "step": 3728
+    },
+    {
+      "epoch": 0.004874490688271811,
+      "grad_norm": 1.0619542598724365,
+      "learning_rate": 3.13963963963964e-05,
+      "loss": 0.6395,
+      "step": 3729
+    },
+    {
+      "epoch": 0.004875797872688081,
+      "grad_norm": 1.0204658508300781,
+      "learning_rate": 3.1391391391391395e-05,
+      "loss": 0.4004,
+      "step": 3730
+    },
+    {
+      "epoch": 0.004877105057104351,
+      "grad_norm": 1.7120537757873535,
+      "learning_rate": 3.138638638638639e-05,
+      "loss": 1.862,
+      "step": 3731
+    },
+    {
+      "epoch": 0.004878412241520622,
+      "grad_norm": 1.0097644329071045,
+      "learning_rate": 3.138138138138138e-05,
+      "loss": 1.0182,
+      "step": 3732
+    },
+    {
+      "epoch": 0.004879719425936892,
+      "grad_norm": 1.4513212442398071,
+      "learning_rate": 3.1376376376376374e-05,
+      "loss": 1.5499,
+      "step": 3733
+    },
+    {
+      "epoch": 0.004881026610353162,
+      "grad_norm": 1.503116250038147,
+      "learning_rate": 3.1371371371371374e-05,
+      "loss": 1.3664,
+      "step": 3734
+    },
+    {
+      "epoch": 0.0048823337947694325,
+      "grad_norm": 1.4559953212738037,
+      "learning_rate": 3.1366366366366367e-05,
+      "loss": 1.7309,
+      "step": 3735
+    },
+    {
+      "epoch": 0.004883640979185703,
+      "grad_norm": 1.5689241886138916,
+      "learning_rate": 3.1361361361361366e-05,
+      "loss": 1.1958,
+      "step": 3736
+    },
+    {
+      "epoch": 0.004884948163601973,
+      "grad_norm": 1.7262475490570068,
+      "learning_rate": 3.135635635635636e-05,
+      "loss": 1.6301,
+      "step": 3737
+    },
+    {
+      "epoch": 0.004886255348018243,
+      "grad_norm": 1.2928582429885864,
+      "learning_rate": 3.135135135135135e-05,
+      "loss": 1.4227,
+      "step": 3738
+    },
+    {
+      "epoch": 0.004887562532434513,
+      "grad_norm": 1.6690127849578857,
+      "learning_rate": 3.134634634634635e-05,
+      "loss": 2.2065,
+      "step": 3739
+    },
+    {
+      "epoch": 0.004888869716850783,
+      "grad_norm": 1.4808510541915894,
+      "learning_rate": 3.134134134134134e-05,
+      "loss": 2.2455,
+      "step": 3740
+    },
+    {
+      "epoch": 0.004890176901267054,
+      "grad_norm": 1.6367934942245483,
+      "learning_rate": 3.133633633633634e-05,
+      "loss": 1.6816,
+      "step": 3741
+    },
+    {
+      "epoch": 0.004891484085683324,
+      "grad_norm": 1.6810646057128906,
+      "learning_rate": 3.133133133133133e-05,
+      "loss": 1.7209,
+      "step": 3742
+    },
+    {
+      "epoch": 0.004892791270099594,
+      "grad_norm": 2.6315197944641113,
+      "learning_rate": 3.132632632632633e-05,
+      "loss": 1.2619,
+      "step": 3743
+    },
+    {
+      "epoch": 0.0048940984545158646,
+      "grad_norm": 1.4508322477340698,
+      "learning_rate": 3.1321321321321324e-05,
+      "loss": 1.5334,
+      "step": 3744
+    },
+    {
+      "epoch": 0.004895405638932135,
+      "grad_norm": 1.4310745000839233,
+      "learning_rate": 3.131631631631632e-05,
+      "loss": 1.4105,
+      "step": 3745
+    },
+    {
+      "epoch": 0.004896712823348405,
+      "grad_norm": 1.3787332773208618,
+      "learning_rate": 3.1311311311311316e-05,
+      "loss": 1.2904,
+      "step": 3746
+    },
+    {
+      "epoch": 0.0048980200077646755,
+      "grad_norm": 1.22126305103302,
+      "learning_rate": 3.130630630630631e-05,
+      "loss": 0.8807,
+      "step": 3747
+    },
+    {
+      "epoch": 0.004899327192180946,
+      "grad_norm": 1.1711541414260864,
+      "learning_rate": 3.13013013013013e-05,
+      "loss": 1.3723,
+      "step": 3748
+    },
+    {
+      "epoch": 0.004900634376597216,
+      "grad_norm": 1.491410732269287,
+      "learning_rate": 3.1296296296296295e-05,
+      "loss": 1.9876,
+      "step": 3749
+    },
+    {
+      "epoch": 0.0049019415610134865,
+      "grad_norm": 1.7137730121612549,
+      "learning_rate": 3.129129129129129e-05,
+      "loss": 1.7691,
+      "step": 3750
+    },
+    {
+      "epoch": 0.004903248745429757,
+      "grad_norm": 1.3348352909088135,
+      "learning_rate": 3.128628628628629e-05,
+      "loss": 1.4505,
+      "step": 3751
+    },
+    {
+      "epoch": 0.004904555929846027,
+      "grad_norm": 1.2195537090301514,
+      "learning_rate": 3.128128128128128e-05,
+      "loss": 2.026,
+      "step": 3752
+    },
+    {
+      "epoch": 0.004905863114262297,
+      "grad_norm": 1.2924495935440063,
+      "learning_rate": 3.127627627627628e-05,
+      "loss": 1.5301,
+      "step": 3753
+    },
+    {
+      "epoch": 0.004907170298678567,
+      "grad_norm": 1.3727775812149048,
+      "learning_rate": 3.1271271271271274e-05,
+      "loss": 1.6014,
+      "step": 3754
+    },
+    {
+      "epoch": 0.004908477483094837,
+      "grad_norm": 1.3832722902297974,
+      "learning_rate": 3.126626626626627e-05,
+      "loss": 2.8051,
+      "step": 3755
+    },
+    {
+      "epoch": 0.0049097846675111075,
+      "grad_norm": 1.4717427492141724,
+      "learning_rate": 3.1261261261261266e-05,
+      "loss": 1.5982,
+      "step": 3756
+    },
+    {
+      "epoch": 0.004911091851927378,
+      "grad_norm": 1.6524131298065186,
+      "learning_rate": 3.125625625625625e-05,
+      "loss": 2.446,
+      "step": 3757
+    },
+    {
+      "epoch": 0.004912399036343648,
+      "grad_norm": 1.3562718629837036,
+      "learning_rate": 3.125125125125125e-05,
+      "loss": 1.6417,
+      "step": 3758
+    },
+    {
+      "epoch": 0.0049137062207599185,
+      "grad_norm": 1.5755422115325928,
+      "learning_rate": 3.1246246246246245e-05,
+      "loss": 2.2693,
+      "step": 3759
+    },
+    {
+      "epoch": 0.004915013405176189,
+      "grad_norm": 1.5252537727355957,
+      "learning_rate": 3.1241241241241245e-05,
+      "loss": 0.9476,
+      "step": 3760
+    },
+    {
+      "epoch": 0.004916320589592459,
+      "grad_norm": 1.4433828592300415,
+      "learning_rate": 3.123623623623624e-05,
+      "loss": 1.6035,
+      "step": 3761
+    },
+    {
+      "epoch": 0.0049176277740087294,
+      "grad_norm": 1.7493857145309448,
+      "learning_rate": 3.123123123123123e-05,
+      "loss": 1.965,
+      "step": 3762
+    },
+    {
+      "epoch": 0.004918934958425,
+      "grad_norm": 1.332976222038269,
+      "learning_rate": 3.122622622622623e-05,
+      "loss": 1.108,
+      "step": 3763
+    },
+    {
+      "epoch": 0.00492024214284127,
+      "grad_norm": 1.3209744691848755,
+      "learning_rate": 3.1221221221221224e-05,
+      "loss": 2.0076,
+      "step": 3764
+    },
+    {
+      "epoch": 0.00492154932725754,
+      "grad_norm": 1.3752130270004272,
+      "learning_rate": 3.121621621621622e-05,
+      "loss": 1.6558,
+      "step": 3765
+    },
+    {
+      "epoch": 0.004922856511673811,
+      "grad_norm": 1.3825314044952393,
+      "learning_rate": 3.121121121121121e-05,
+      "loss": 1.165,
+      "step": 3766
+    },
+    {
+      "epoch": 0.004924163696090081,
+      "grad_norm": 1.4443551301956177,
+      "learning_rate": 3.120620620620621e-05,
+      "loss": 0.7859,
+      "step": 3767
+    },
+    {
+      "epoch": 0.004925470880506351,
+      "grad_norm": 1.6220439672470093,
+      "learning_rate": 3.12012012012012e-05,
+      "loss": 1.8355,
+      "step": 3768
+    },
+    {
+      "epoch": 0.004926778064922621,
+      "grad_norm": 1.9596610069274902,
+      "learning_rate": 3.1196196196196195e-05,
+      "loss": 1.9683,
+      "step": 3769
+    },
+    {
+      "epoch": 0.004928085249338891,
+      "grad_norm": 1.4999781847000122,
+      "learning_rate": 3.1191191191191195e-05,
+      "loss": 1.9104,
+      "step": 3770
+    },
+    {
+      "epoch": 0.0049293924337551615,
+      "grad_norm": 1.3378304243087769,
+      "learning_rate": 3.118618618618619e-05,
+      "loss": 2.2265,
+      "step": 3771
+    },
+    {
+      "epoch": 0.004930699618171432,
+      "grad_norm": 0.8426940441131592,
+      "learning_rate": 3.118118118118119e-05,
+      "loss": 0.5809,
+      "step": 3772
+    },
+    {
+      "epoch": 0.004932006802587702,
+      "grad_norm": 1.7287280559539795,
+      "learning_rate": 3.1176176176176174e-05,
+      "loss": 1.426,
+      "step": 3773
+    },
+    {
+      "epoch": 0.004933313987003972,
+      "grad_norm": 1.3054864406585693,
+      "learning_rate": 3.1171171171171174e-05,
+      "loss": 1.1987,
+      "step": 3774
+    },
+    {
+      "epoch": 0.004934621171420243,
+      "grad_norm": 1.7415351867675781,
+      "learning_rate": 3.116616616616617e-05,
+      "loss": 1.4178,
+      "step": 3775
+    },
+    {
+      "epoch": 0.004935928355836513,
+      "grad_norm": 0.9457842111587524,
+      "learning_rate": 3.116116116116116e-05,
+      "loss": 0.5989,
+      "step": 3776
+    },
+    {
+      "epoch": 0.004937235540252783,
+      "grad_norm": 1.7196027040481567,
+      "learning_rate": 3.115615615615616e-05,
+      "loss": 1.4657,
+      "step": 3777
+    },
+    {
+      "epoch": 0.004938542724669054,
+      "grad_norm": 1.1638261079788208,
+      "learning_rate": 3.115115115115115e-05,
+      "loss": 0.7442,
+      "step": 3778
+    },
+    {
+      "epoch": 0.004939849909085324,
+      "grad_norm": 1.0973879098892212,
+      "learning_rate": 3.114614614614615e-05,
+      "loss": 0.8039,
+      "step": 3779
+    },
+    {
+      "epoch": 0.004941157093501594,
+      "grad_norm": 0.8544688820838928,
+      "learning_rate": 3.1141141141141145e-05,
+      "loss": 0.4965,
+      "step": 3780
+    },
+    {
+      "epoch": 0.004942464277917865,
+      "grad_norm": 1.5183149576187134,
+      "learning_rate": 3.113613613613614e-05,
+      "loss": 1.8016,
+      "step": 3781
+    },
+    {
+      "epoch": 0.004943771462334135,
+      "grad_norm": 1.3627920150756836,
+      "learning_rate": 3.113113113113113e-05,
+      "loss": 2.1404,
+      "step": 3782
+    },
+    {
+      "epoch": 0.004945078646750405,
+      "grad_norm": 0.16800549626350403,
+      "learning_rate": 3.1126126126126124e-05,
+      "loss": 0.0057,
+      "step": 3783
+    },
+    {
+      "epoch": 0.004946385831166675,
+      "grad_norm": 1.2639939785003662,
+      "learning_rate": 3.1121121121121124e-05,
+      "loss": 2.1521,
+      "step": 3784
+    },
+    {
+      "epoch": 0.004947693015582945,
+      "grad_norm": 1.469838261604309,
+      "learning_rate": 3.111611611611612e-05,
+      "loss": 1.5785,
+      "step": 3785
+    },
+    {
+      "epoch": 0.004949000199999215,
+      "grad_norm": 1.517969012260437,
+      "learning_rate": 3.111111111111111e-05,
+      "loss": 1.6104,
+      "step": 3786
+    },
+    {
+      "epoch": 0.004950307384415486,
+      "grad_norm": 1.6162807941436768,
+      "learning_rate": 3.110610610610611e-05,
+      "loss": 1.5645,
+      "step": 3787
+    },
+    {
+      "epoch": 0.004951614568831756,
+      "grad_norm": 1.280788540840149,
+      "learning_rate": 3.11011011011011e-05,
+      "loss": 1.5355,
+      "step": 3788
+    },
+    {
+      "epoch": 0.004952921753248026,
+      "grad_norm": 0.8779072761535645,
+      "learning_rate": 3.10960960960961e-05,
+      "loss": 0.3087,
+      "step": 3789
+    },
+    {
+      "epoch": 0.004954228937664297,
+      "grad_norm": 1.1055281162261963,
+      "learning_rate": 3.109109109109109e-05,
+      "loss": 1.4564,
+      "step": 3790
+    },
+    {
+      "epoch": 0.004955536122080567,
+      "grad_norm": 0.8993090391159058,
+      "learning_rate": 3.108608608608609e-05,
+      "loss": 0.5528,
+      "step": 3791
+    },
+    {
+      "epoch": 0.004956843306496837,
+      "grad_norm": 1.3963969945907593,
+      "learning_rate": 3.108108108108108e-05,
+      "loss": 1.9061,
+      "step": 3792
+    },
+    {
+      "epoch": 0.004958150490913108,
+      "grad_norm": 1.6435760259628296,
+      "learning_rate": 3.1076076076076074e-05,
+      "loss": 0.9245,
+      "step": 3793
+    },
+    {
+      "epoch": 0.004959457675329378,
+      "grad_norm": 0.9147536158561707,
+      "learning_rate": 3.1071071071071074e-05,
+      "loss": 1.7488,
+      "step": 3794
+    },
+    {
+      "epoch": 0.004960764859745648,
+      "grad_norm": 1.349532961845398,
+      "learning_rate": 3.106606606606607e-05,
+      "loss": 1.8887,
+      "step": 3795
+    },
+    {
+      "epoch": 0.004962072044161919,
+      "grad_norm": 1.2240136861801147,
+      "learning_rate": 3.106106106106107e-05,
+      "loss": 1.2803,
+      "step": 3796
+    },
+    {
+      "epoch": 0.004963379228578189,
+      "grad_norm": 1.3309961557388306,
+      "learning_rate": 3.105605605605606e-05,
+      "loss": 1.8965,
+      "step": 3797
+    },
+    {
+      "epoch": 0.004964686412994459,
+      "grad_norm": 1.4561996459960938,
+      "learning_rate": 3.105105105105105e-05,
+      "loss": 1.5273,
+      "step": 3798
+    },
+    {
+      "epoch": 0.0049659935974107295,
+      "grad_norm": 1.5692342519760132,
+      "learning_rate": 3.1046046046046046e-05,
+      "loss": 2.5776,
+      "step": 3799
+    },
+    {
+      "epoch": 0.004967300781826999,
+      "grad_norm": 1.372957706451416,
+      "learning_rate": 3.104104104104104e-05,
+      "loss": 1.6934,
+      "step": 3800
+    },
+    {
+      "epoch": 0.004968607966243269,
+      "grad_norm": 1.2698099613189697,
+      "learning_rate": 3.103603603603604e-05,
+      "loss": 1.9306,
+      "step": 3801
+    },
+    {
+      "epoch": 0.00496991515065954,
+      "grad_norm": 1.3815041780471802,
+      "learning_rate": 3.103103103103103e-05,
+      "loss": 1.4298,
+      "step": 3802
+    },
+    {
+      "epoch": 0.00497122233507581,
+      "grad_norm": 1.3558297157287598,
+      "learning_rate": 3.102602602602603e-05,
+      "loss": 1.6784,
+      "step": 3803
+    },
+    {
+      "epoch": 0.00497252951949208,
+      "grad_norm": 1.3270268440246582,
+      "learning_rate": 3.1021021021021024e-05,
+      "loss": 1.2281,
+      "step": 3804
+    },
+    {
+      "epoch": 0.004973836703908351,
+      "grad_norm": 1.6843928098678589,
+      "learning_rate": 3.101601601601602e-05,
+      "loss": 1.5563,
+      "step": 3805
+    },
+    {
+      "epoch": 0.004975143888324621,
+      "grad_norm": 1.5665204524993896,
+      "learning_rate": 3.101101101101102e-05,
+      "loss": 2.3969,
+      "step": 3806
+    },
+    {
+      "epoch": 0.004976451072740891,
+      "grad_norm": 1.5264755487442017,
+      "learning_rate": 3.1006006006006e-05,
+      "loss": 1.8079,
+      "step": 3807
+    },
+    {
+      "epoch": 0.0049777582571571615,
+      "grad_norm": 1.4550938606262207,
+      "learning_rate": 3.1001001001001e-05,
+      "loss": 1.1257,
+      "step": 3808
+    },
+    {
+      "epoch": 0.004979065441573432,
+      "grad_norm": 0.832181990146637,
+      "learning_rate": 3.0995995995995996e-05,
+      "loss": 0.4516,
+      "step": 3809
+    },
+    {
+      "epoch": 0.004980372625989702,
+      "grad_norm": 1.3691387176513672,
+      "learning_rate": 3.0990990990990995e-05,
+      "loss": 1.7998,
+      "step": 3810
+    },
+    {
+      "epoch": 0.0049816798104059725,
+      "grad_norm": 1.7858459949493408,
+      "learning_rate": 3.098598598598599e-05,
+      "loss": 1.1019,
+      "step": 3811
+    },
+    {
+      "epoch": 0.004982986994822243,
+      "grad_norm": 1.4970571994781494,
+      "learning_rate": 3.098098098098098e-05,
+      "loss": 1.5699,
+      "step": 3812
+    },
+    {
+      "epoch": 0.004984294179238513,
+      "grad_norm": 1.4318580627441406,
+      "learning_rate": 3.097597597597598e-05,
+      "loss": 1.3201,
+      "step": 3813
+    },
+    {
+      "epoch": 0.0049856013636547835,
+      "grad_norm": 1.5248149633407593,
+      "learning_rate": 3.0970970970970974e-05,
+      "loss": 1.4487,
+      "step": 3814
+    },
+    {
+      "epoch": 0.004986908548071053,
+      "grad_norm": 1.4313851594924927,
+      "learning_rate": 3.096596596596597e-05,
+      "loss": 2.0793,
+      "step": 3815
+    },
+    {
+      "epoch": 0.004988215732487323,
+      "grad_norm": 1.2950103282928467,
+      "learning_rate": 3.096096096096096e-05,
+      "loss": 0.9913,
+      "step": 3816
+    },
+    {
+      "epoch": 0.0049895229169035936,
+      "grad_norm": 1.6443979740142822,
+      "learning_rate": 3.095595595595595e-05,
+      "loss": 2.0131,
+      "step": 3817
+    },
+    {
+      "epoch": 0.004990830101319864,
+      "grad_norm": 1.1953177452087402,
+      "learning_rate": 3.095095095095095e-05,
+      "loss": 1.2366,
+      "step": 3818
+    },
+    {
+      "epoch": 0.004992137285736134,
+      "grad_norm": 1.3844408988952637,
+      "learning_rate": 3.0945945945945946e-05,
+      "loss": 1.9406,
+      "step": 3819
+    },
+    {
+      "epoch": 0.0049934444701524045,
+      "grad_norm": 1.5423098802566528,
+      "learning_rate": 3.0940940940940946e-05,
+      "loss": 1.7356,
+      "step": 3820
+    },
+    {
+      "epoch": 0.004994751654568675,
+      "grad_norm": 1.2574340105056763,
+      "learning_rate": 3.093593593593594e-05,
+      "loss": 1.601,
+      "step": 3821
+    },
+    {
+      "epoch": 0.004996058838984945,
+      "grad_norm": 1.2597051858901978,
+      "learning_rate": 3.093093093093093e-05,
+      "loss": 0.9394,
+      "step": 3822
+    },
+    {
+      "epoch": 0.0049973660234012155,
+      "grad_norm": 1.4392980337142944,
+      "learning_rate": 3.0925925925925924e-05,
+      "loss": 2.0329,
+      "step": 3823
+    },
+    {
+      "epoch": 0.004998673207817486,
+      "grad_norm": 1.226604700088501,
+      "learning_rate": 3.092092092092092e-05,
+      "loss": 1.5339,
+      "step": 3824
+    },
+    {
+      "epoch": 0.004999980392233756,
+      "grad_norm": 1.097657561302185,
+      "learning_rate": 3.091591591591592e-05,
+      "loss": 1.2831,
+      "step": 3825
+    },
+    {
+      "epoch": 0.0050012875766500264,
+      "grad_norm": 1.4463257789611816,
+      "learning_rate": 3.091091091091091e-05,
+      "loss": 1.8838,
+      "step": 3826
+    },
+    {
+      "epoch": 0.005002594761066297,
+      "grad_norm": 1.5350286960601807,
+      "learning_rate": 3.090590590590591e-05,
+      "loss": 1.7532,
+      "step": 3827
+    },
+    {
+      "epoch": 0.005003901945482567,
+      "grad_norm": 1.0023425817489624,
+      "learning_rate": 3.09009009009009e-05,
+      "loss": 1.1523,
+      "step": 3828
+    },
+    {
+      "epoch": 0.005005209129898837,
+      "grad_norm": 1.4945123195648193,
+      "learning_rate": 3.0895895895895896e-05,
+      "loss": 1.6207,
+      "step": 3829
+    },
+    {
+      "epoch": 0.005006516314315107,
+      "grad_norm": 1.4526740312576294,
+      "learning_rate": 3.0890890890890896e-05,
+      "loss": 1.1736,
+      "step": 3830
+    },
+    {
+      "epoch": 0.005007823498731377,
+      "grad_norm": 1.4821656942367554,
+      "learning_rate": 3.088588588588588e-05,
+      "loss": 1.9191,
+      "step": 3831
+    },
+    {
+      "epoch": 0.0050091306831476475,
+      "grad_norm": 1.2776916027069092,
+      "learning_rate": 3.088088088088088e-05,
+      "loss": 1.2258,
+      "step": 3832
+    },
+    {
+      "epoch": 0.005010437867563918,
+      "grad_norm": 1.4201115369796753,
+      "learning_rate": 3.0875875875875875e-05,
+      "loss": 1.1617,
+      "step": 3833
+    },
+    {
+      "epoch": 0.005011745051980188,
+      "grad_norm": 1.3112481832504272,
+      "learning_rate": 3.0870870870870874e-05,
+      "loss": 1.7392,
+      "step": 3834
+    },
+    {
+      "epoch": 0.0050130522363964584,
+      "grad_norm": 1.392569899559021,
+      "learning_rate": 3.086586586586587e-05,
+      "loss": 1.5906,
+      "step": 3835
+    },
+    {
+      "epoch": 0.005014359420812729,
+      "grad_norm": 1.2668886184692383,
+      "learning_rate": 3.086086086086086e-05,
+      "loss": 1.3269,
+      "step": 3836
+    },
+    {
+      "epoch": 0.005015666605228999,
+      "grad_norm": 1.610604166984558,
+      "learning_rate": 3.085585585585586e-05,
+      "loss": 1.5205,
+      "step": 3837
+    },
+    {
+      "epoch": 0.005016973789645269,
+      "grad_norm": 1.1114451885223389,
+      "learning_rate": 3.085085085085085e-05,
+      "loss": 1.3334,
+      "step": 3838
+    },
+    {
+      "epoch": 0.00501828097406154,
+      "grad_norm": 1.4294980764389038,
+      "learning_rate": 3.084584584584585e-05,
+      "loss": 1.698,
+      "step": 3839
+    },
+    {
+      "epoch": 0.00501958815847781,
+      "grad_norm": 1.447871208190918,
+      "learning_rate": 3.084084084084084e-05,
+      "loss": 1.7142,
+      "step": 3840
+    },
+    {
+      "epoch": 0.00502089534289408,
+      "grad_norm": 1.4239797592163086,
+      "learning_rate": 3.083583583583584e-05,
+      "loss": 2.4205,
+      "step": 3841
+    },
+    {
+      "epoch": 0.005022202527310351,
+      "grad_norm": 1.1823289394378662,
+      "learning_rate": 3.083083083083083e-05,
+      "loss": 1.2524,
+      "step": 3842
+    },
+    {
+      "epoch": 0.005023509711726621,
+      "grad_norm": 1.530429720878601,
+      "learning_rate": 3.0825825825825825e-05,
+      "loss": 1.7162,
+      "step": 3843
+    },
+    {
+      "epoch": 0.005024816896142891,
+      "grad_norm": 1.387465000152588,
+      "learning_rate": 3.0820820820820824e-05,
+      "loss": 1.7207,
+      "step": 3844
+    },
+    {
+      "epoch": 0.005026124080559161,
+      "grad_norm": 1.108802080154419,
+      "learning_rate": 3.081581581581582e-05,
+      "loss": 0.9126,
+      "step": 3845
+    },
+    {
+      "epoch": 0.005027431264975431,
+      "grad_norm": 1.5083847045898438,
+      "learning_rate": 3.081081081081082e-05,
+      "loss": 1.7968,
+      "step": 3846
+    },
+    {
+      "epoch": 0.005028738449391701,
+      "grad_norm": 0.9123502969741821,
+      "learning_rate": 3.080580580580581e-05,
+      "loss": 0.8725,
+      "step": 3847
+    },
+    {
+      "epoch": 0.005030045633807972,
+      "grad_norm": 1.357458472251892,
+      "learning_rate": 3.08008008008008e-05,
+      "loss": 1.9534,
+      "step": 3848
+    },
+    {
+      "epoch": 0.005031352818224242,
+      "grad_norm": 1.5899296998977661,
+      "learning_rate": 3.0795795795795796e-05,
+      "loss": 1.5007,
+      "step": 3849
+    },
+    {
+      "epoch": 0.005032660002640512,
+      "grad_norm": 1.6267447471618652,
+      "learning_rate": 3.079079079079079e-05,
+      "loss": 1.8514,
+      "step": 3850
+    },
+    {
+      "epoch": 0.005033967187056783,
+      "grad_norm": 1.4869794845581055,
+      "learning_rate": 3.078578578578579e-05,
+      "loss": 2.3902,
+      "step": 3851
+    },
+    {
+      "epoch": 0.005035274371473053,
+      "grad_norm": 1.597355604171753,
+      "learning_rate": 3.078078078078078e-05,
+      "loss": 1.6907,
+      "step": 3852
+    },
+    {
+      "epoch": 0.005036581555889323,
+      "grad_norm": 1.4246262311935425,
+      "learning_rate": 3.077577577577578e-05,
+      "loss": 1.8536,
+      "step": 3853
+    },
+    {
+      "epoch": 0.005037888740305594,
+      "grad_norm": 1.2486509084701538,
+      "learning_rate": 3.0770770770770774e-05,
+      "loss": 1.6303,
+      "step": 3854
+    },
+    {
+      "epoch": 0.005039195924721864,
+      "grad_norm": 1.5275346040725708,
+      "learning_rate": 3.076576576576577e-05,
+      "loss": 1.2414,
+      "step": 3855
+    },
+    {
+      "epoch": 0.005040503109138134,
+      "grad_norm": 1.6860066652297974,
+      "learning_rate": 3.076076076076076e-05,
+      "loss": 0.8457,
+      "step": 3856
+    },
+    {
+      "epoch": 0.005041810293554405,
+      "grad_norm": 1.3903907537460327,
+      "learning_rate": 3.075575575575575e-05,
+      "loss": 1.9508,
+      "step": 3857
+    },
+    {
+      "epoch": 0.005043117477970675,
+      "grad_norm": 1.3862698078155518,
+      "learning_rate": 3.075075075075075e-05,
+      "loss": 1.4247,
+      "step": 3858
+    },
+    {
+      "epoch": 0.005044424662386945,
+      "grad_norm": 1.5670406818389893,
+      "learning_rate": 3.0745745745745746e-05,
+      "loss": 2.008,
+      "step": 3859
+    },
+    {
+      "epoch": 0.005045731846803216,
+      "grad_norm": 1.2622721195220947,
+      "learning_rate": 3.074074074074074e-05,
+      "loss": 1.7833,
+      "step": 3860
+    },
+    {
+      "epoch": 0.005047039031219485,
+      "grad_norm": 1.5731350183486938,
+      "learning_rate": 3.073573573573574e-05,
+      "loss": 1.7571,
+      "step": 3861
+    },
+    {
+      "epoch": 0.005048346215635755,
+      "grad_norm": 0.49037688970565796,
+      "learning_rate": 3.073073073073073e-05,
+      "loss": 0.2146,
+      "step": 3862
+    },
+    {
+      "epoch": 0.005049653400052026,
+      "grad_norm": 1.4120863676071167,
+      "learning_rate": 3.072572572572573e-05,
+      "loss": 1.5387,
+      "step": 3863
+    },
+    {
+      "epoch": 0.005050960584468296,
+      "grad_norm": 1.233785629272461,
+      "learning_rate": 3.0720720720720725e-05,
+      "loss": 0.707,
+      "step": 3864
+    },
+    {
+      "epoch": 0.005052267768884566,
+      "grad_norm": 1.447662353515625,
+      "learning_rate": 3.071571571571572e-05,
+      "loss": 1.9236,
+      "step": 3865
+    },
+    {
+      "epoch": 0.005053574953300837,
+      "grad_norm": 1.3165266513824463,
+      "learning_rate": 3.071071071071071e-05,
+      "loss": 1.3135,
+      "step": 3866
+    },
+    {
+      "epoch": 0.005054882137717107,
+      "grad_norm": 1.6165521144866943,
+      "learning_rate": 3.0705705705705703e-05,
+      "loss": 1.7681,
+      "step": 3867
+    },
+    {
+      "epoch": 0.005056189322133377,
+      "grad_norm": 1.3999338150024414,
+      "learning_rate": 3.07007007007007e-05,
+      "loss": 1.2878,
+      "step": 3868
+    },
+    {
+      "epoch": 0.005057496506549648,
+      "grad_norm": 1.3656322956085205,
+      "learning_rate": 3.0695695695695696e-05,
+      "loss": 1.5706,
+      "step": 3869
+    },
+    {
+      "epoch": 0.005058803690965918,
+      "grad_norm": 0.7025394439697266,
+      "learning_rate": 3.0690690690690696e-05,
+      "loss": 0.4072,
+      "step": 3870
+    },
+    {
+      "epoch": 0.005060110875382188,
+      "grad_norm": 1.5745431184768677,
+      "learning_rate": 3.068568568568569e-05,
+      "loss": 0.9511,
+      "step": 3871
+    },
+    {
+      "epoch": 0.0050614180597984585,
+      "grad_norm": 1.4109649658203125,
+      "learning_rate": 3.068068068068068e-05,
+      "loss": 1.5932,
+      "step": 3872
+    },
+    {
+      "epoch": 0.005062725244214729,
+      "grad_norm": 1.449183464050293,
+      "learning_rate": 3.0675675675675675e-05,
+      "loss": 1.2166,
+      "step": 3873
+    },
+    {
+      "epoch": 0.005064032428630999,
+      "grad_norm": 1.4604486227035522,
+      "learning_rate": 3.067067067067067e-05,
+      "loss": 1.7653,
+      "step": 3874
+    },
+    {
+      "epoch": 0.0050653396130472695,
+      "grad_norm": 1.6821430921554565,
+      "learning_rate": 3.066566566566567e-05,
+      "loss": 2.0105,
+      "step": 3875
+    },
+    {
+      "epoch": 0.005066646797463539,
+      "grad_norm": 1.26608145236969,
+      "learning_rate": 3.066066066066066e-05,
+      "loss": 1.1588,
+      "step": 3876
+    },
+    {
+      "epoch": 0.005067953981879809,
+      "grad_norm": 1.2332102060317993,
+      "learning_rate": 3.065565565565566e-05,
+      "loss": 1.6295,
+      "step": 3877
+    },
+    {
+      "epoch": 0.00506926116629608,
+      "grad_norm": 1.4106889963150024,
+      "learning_rate": 3.065065065065065e-05,
+      "loss": 1.5531,
+      "step": 3878
+    },
+    {
+      "epoch": 0.00507056835071235,
+      "grad_norm": 1.3613563776016235,
+      "learning_rate": 3.0645645645645646e-05,
+      "loss": 1.8212,
+      "step": 3879
+    },
+    {
+      "epoch": 0.00507187553512862,
+      "grad_norm": 1.2397220134735107,
+      "learning_rate": 3.0640640640640646e-05,
+      "loss": 1.5072,
+      "step": 3880
+    },
+    {
+      "epoch": 0.0050731827195448905,
+      "grad_norm": 1.7757266759872437,
+      "learning_rate": 3.063563563563563e-05,
+      "loss": 1.938,
+      "step": 3881
+    },
+    {
+      "epoch": 0.005074489903961161,
+      "grad_norm": 1.2141960859298706,
+      "learning_rate": 3.063063063063063e-05,
+      "loss": 1.2716,
+      "step": 3882
+    },
+    {
+      "epoch": 0.005075797088377431,
+      "grad_norm": 1.608217477798462,
+      "learning_rate": 3.0625625625625625e-05,
+      "loss": 2.4804,
+      "step": 3883
+    },
+    {
+      "epoch": 0.0050771042727937015,
+      "grad_norm": 1.523977518081665,
+      "learning_rate": 3.0620620620620625e-05,
+      "loss": 2.1668,
+      "step": 3884
+    },
+    {
+      "epoch": 0.005078411457209972,
+      "grad_norm": 1.8462940454483032,
+      "learning_rate": 3.061561561561562e-05,
+      "loss": 1.2455,
+      "step": 3885
+    },
+    {
+      "epoch": 0.005079718641626242,
+      "grad_norm": 1.3278788328170776,
+      "learning_rate": 3.061061061061061e-05,
+      "loss": 0.7154,
+      "step": 3886
+    },
+    {
+      "epoch": 0.0050810258260425125,
+      "grad_norm": 1.8919024467468262,
+      "learning_rate": 3.060560560560561e-05,
+      "loss": 2.3122,
+      "step": 3887
+    },
+    {
+      "epoch": 0.005082333010458783,
+      "grad_norm": 1.6344910860061646,
+      "learning_rate": 3.06006006006006e-05,
+      "loss": 1.9378,
+      "step": 3888
+    },
+    {
+      "epoch": 0.005083640194875053,
+      "grad_norm": 1.780098557472229,
+      "learning_rate": 3.05955955955956e-05,
+      "loss": 1.373,
+      "step": 3889
+    },
+    {
+      "epoch": 0.005084947379291323,
+      "grad_norm": 1.385277271270752,
+      "learning_rate": 3.059059059059059e-05,
+      "loss": 1.1848,
+      "step": 3890
+    },
+    {
+      "epoch": 0.005086254563707593,
+      "grad_norm": 1.8491382598876953,
+      "learning_rate": 3.058558558558558e-05,
+      "loss": 2.0659,
+      "step": 3891
+    },
+    {
+      "epoch": 0.005087561748123863,
+      "grad_norm": 1.6266995668411255,
+      "learning_rate": 3.058058058058058e-05,
+      "loss": 2.664,
+      "step": 3892
+    },
+    {
+      "epoch": 0.0050888689325401335,
+      "grad_norm": 1.288470983505249,
+      "learning_rate": 3.0575575575575575e-05,
+      "loss": 1.5972,
+      "step": 3893
+    },
+    {
+      "epoch": 0.005090176116956404,
+      "grad_norm": 1.3688236474990845,
+      "learning_rate": 3.0570570570570575e-05,
+      "loss": 2.3009,
+      "step": 3894
+    },
+    {
+      "epoch": 0.005091483301372674,
+      "grad_norm": 1.4017184972763062,
+      "learning_rate": 3.056556556556557e-05,
+      "loss": 1.8817,
+      "step": 3895
+    },
+    {
+      "epoch": 0.0050927904857889445,
+      "grad_norm": 1.2632732391357422,
+      "learning_rate": 3.056056056056056e-05,
+      "loss": 1.4689,
+      "step": 3896
+    },
+    {
+      "epoch": 0.005094097670205215,
+      "grad_norm": 1.6259170770645142,
+      "learning_rate": 3.055555555555556e-05,
+      "loss": 2.1099,
+      "step": 3897
+    },
+    {
+      "epoch": 0.005095404854621485,
+      "grad_norm": 1.5173264741897583,
+      "learning_rate": 3.0550550550550547e-05,
+      "loss": 2.0938,
+      "step": 3898
+    },
+    {
+      "epoch": 0.0050967120390377554,
+      "grad_norm": 1.8067376613616943,
+      "learning_rate": 3.0545545545545546e-05,
+      "loss": 1.9913,
+      "step": 3899
+    },
+    {
+      "epoch": 0.005098019223454026,
+      "grad_norm": 1.4699007272720337,
+      "learning_rate": 3.054054054054054e-05,
+      "loss": 1.688,
+      "step": 3900
+    },
+    {
+      "epoch": 0.005099326407870296,
+      "grad_norm": 1.1567562818527222,
+      "learning_rate": 3.053553553553554e-05,
+      "loss": 0.7406,
+      "step": 3901
+    },
+    {
+      "epoch": 0.005100633592286566,
+      "grad_norm": 1.2628108263015747,
+      "learning_rate": 3.053053053053053e-05,
+      "loss": 1.4789,
+      "step": 3902
+    },
+    {
+      "epoch": 0.005101940776702837,
+      "grad_norm": 1.5220507383346558,
+      "learning_rate": 3.0525525525525525e-05,
+      "loss": 1.174,
+      "step": 3903
+    },
+    {
+      "epoch": 0.005103247961119107,
+      "grad_norm": 1.101322889328003,
+      "learning_rate": 3.0520520520520525e-05,
+      "loss": 0.755,
+      "step": 3904
+    },
+    {
+      "epoch": 0.005104555145535377,
+      "grad_norm": 2.2422983646392822,
+      "learning_rate": 3.0515515515515518e-05,
+      "loss": 1.4174,
+      "step": 3905
+    },
+    {
+      "epoch": 0.005105862329951648,
+      "grad_norm": 1.2769927978515625,
+      "learning_rate": 3.051051051051051e-05,
+      "loss": 1.4848,
+      "step": 3906
+    },
+    {
+      "epoch": 0.005107169514367917,
+      "grad_norm": 1.8421123027801514,
+      "learning_rate": 3.0505505505505504e-05,
+      "loss": 1.4738,
+      "step": 3907
+    },
+    {
+      "epoch": 0.0051084766987841874,
+      "grad_norm": 1.2219547033309937,
+      "learning_rate": 3.05005005005005e-05,
+      "loss": 1.4379,
+      "step": 3908
+    },
+    {
+      "epoch": 0.005109783883200458,
+      "grad_norm": 0.988002598285675,
+      "learning_rate": 3.0495495495495496e-05,
+      "loss": 0.7816,
+      "step": 3909
+    },
+    {
+      "epoch": 0.005111091067616728,
+      "grad_norm": 1.7818750143051147,
+      "learning_rate": 3.0490490490490493e-05,
+      "loss": 2.6364,
+      "step": 3910
+    },
+    {
+      "epoch": 0.005112398252032998,
+      "grad_norm": 1.7232840061187744,
+      "learning_rate": 3.048548548548549e-05,
+      "loss": 2.1144,
+      "step": 3911
+    },
+    {
+      "epoch": 0.005113705436449269,
+      "grad_norm": 1.665440559387207,
+      "learning_rate": 3.0480480480480482e-05,
+      "loss": 1.1679,
+      "step": 3912
+    },
+    {
+      "epoch": 0.005115012620865539,
+      "grad_norm": 1.7987512350082397,
+      "learning_rate": 3.047547547547548e-05,
+      "loss": 1.5124,
+      "step": 3913
+    },
+    {
+      "epoch": 0.005116319805281809,
+      "grad_norm": 1.1339677572250366,
+      "learning_rate": 3.0470470470470475e-05,
+      "loss": 1.081,
+      "step": 3914
+    },
+    {
+      "epoch": 0.00511762698969808,
+      "grad_norm": 2.1898550987243652,
+      "learning_rate": 3.0465465465465464e-05,
+      "loss": 2.11,
+      "step": 3915
+    },
+    {
+      "epoch": 0.00511893417411435,
+      "grad_norm": 1.8821930885314941,
+      "learning_rate": 3.046046046046046e-05,
+      "loss": 1.3807,
+      "step": 3916
+    },
+    {
+      "epoch": 0.00512024135853062,
+      "grad_norm": 1.2945269346237183,
+      "learning_rate": 3.0455455455455457e-05,
+      "loss": 1.6068,
+      "step": 3917
+    },
+    {
+      "epoch": 0.005121548542946891,
+      "grad_norm": 1.5113170146942139,
+      "learning_rate": 3.045045045045045e-05,
+      "loss": 2.0159,
+      "step": 3918
+    },
+    {
+      "epoch": 0.005122855727363161,
+      "grad_norm": 1.1999839544296265,
+      "learning_rate": 3.0445445445445446e-05,
+      "loss": 1.6002,
+      "step": 3919
+    },
+    {
+      "epoch": 0.005124162911779431,
+      "grad_norm": 1.1943840980529785,
+      "learning_rate": 3.0440440440440443e-05,
+      "loss": 0.9942,
+      "step": 3920
+    },
+    {
+      "epoch": 0.005125470096195702,
+      "grad_norm": 1.087363362312317,
+      "learning_rate": 3.043543543543544e-05,
+      "loss": 1.0976,
+      "step": 3921
+    },
+    {
+      "epoch": 0.005126777280611971,
+      "grad_norm": 1.3017834424972534,
+      "learning_rate": 3.0430430430430436e-05,
+      "loss": 1.6221,
+      "step": 3922
+    },
+    {
+      "epoch": 0.005128084465028241,
+      "grad_norm": 1.8021808862686157,
+      "learning_rate": 3.0425425425425425e-05,
+      "loss": 1.5643,
+      "step": 3923
+    },
+    {
+      "epoch": 0.005129391649444512,
+      "grad_norm": 1.5512696504592896,
+      "learning_rate": 3.042042042042042e-05,
+      "loss": 2.137,
+      "step": 3924
+    },
+    {
+      "epoch": 0.005130698833860782,
+      "grad_norm": 1.3120859861373901,
+      "learning_rate": 3.0415415415415415e-05,
+      "loss": 1.4936,
+      "step": 3925
+    },
+    {
+      "epoch": 0.005132006018277052,
+      "grad_norm": 1.7148271799087524,
+      "learning_rate": 3.041041041041041e-05,
+      "loss": 2.2575,
+      "step": 3926
+    },
+    {
+      "epoch": 0.005133313202693323,
+      "grad_norm": 1.3548051118850708,
+      "learning_rate": 3.0405405405405407e-05,
+      "loss": 1.9609,
+      "step": 3927
+    },
+    {
+      "epoch": 0.005134620387109593,
+      "grad_norm": 1.4902853965759277,
+      "learning_rate": 3.0400400400400404e-05,
+      "loss": 2.4328,
+      "step": 3928
+    },
+    {
+      "epoch": 0.005135927571525863,
+      "grad_norm": 1.2725756168365479,
+      "learning_rate": 3.03953953953954e-05,
+      "loss": 1.5157,
+      "step": 3929
+    },
+    {
+      "epoch": 0.005137234755942134,
+      "grad_norm": 1.4815192222595215,
+      "learning_rate": 3.0390390390390393e-05,
+      "loss": 2.1088,
+      "step": 3930
+    },
+    {
+      "epoch": 0.005138541940358404,
+      "grad_norm": 1.92054283618927,
+      "learning_rate": 3.0385385385385386e-05,
+      "loss": 2.4672,
+      "step": 3931
+    },
+    {
+      "epoch": 0.005139849124774674,
+      "grad_norm": 1.292859435081482,
+      "learning_rate": 3.038038038038038e-05,
+      "loss": 1.2631,
+      "step": 3932
+    },
+    {
+      "epoch": 0.005141156309190945,
+      "grad_norm": 1.4453579187393188,
+      "learning_rate": 3.0375375375375375e-05,
+      "loss": 1.8643,
+      "step": 3933
+    },
+    {
+      "epoch": 0.005142463493607215,
+      "grad_norm": 1.1290963888168335,
+      "learning_rate": 3.037037037037037e-05,
+      "loss": 1.4817,
+      "step": 3934
+    },
+    {
+      "epoch": 0.005143770678023485,
+      "grad_norm": 2.485969066619873,
+      "learning_rate": 3.0365365365365368e-05,
+      "loss": 1.5641,
+      "step": 3935
+    },
+    {
+      "epoch": 0.0051450778624397555,
+      "grad_norm": 1.0863666534423828,
+      "learning_rate": 3.036036036036036e-05,
+      "loss": 1.3209,
+      "step": 3936
+    },
+    {
+      "epoch": 0.005146385046856025,
+      "grad_norm": 1.381125807762146,
+      "learning_rate": 3.0355355355355357e-05,
+      "loss": 1.8747,
+      "step": 3937
+    },
+    {
+      "epoch": 0.005147692231272295,
+      "grad_norm": 1.3585411310195923,
+      "learning_rate": 3.0350350350350354e-05,
+      "loss": 1.6075,
+      "step": 3938
+    },
+    {
+      "epoch": 0.005148999415688566,
+      "grad_norm": 1.486929178237915,
+      "learning_rate": 3.034534534534535e-05,
+      "loss": 2.1813,
+      "step": 3939
+    },
+    {
+      "epoch": 0.005150306600104836,
+      "grad_norm": 1.236401915550232,
+      "learning_rate": 3.034034034034034e-05,
+      "loss": 1.0431,
+      "step": 3940
+    },
+    {
+      "epoch": 0.005151613784521106,
+      "grad_norm": 1.7662668228149414,
+      "learning_rate": 3.0335335335335336e-05,
+      "loss": 0.4126,
+      "step": 3941
+    },
+    {
+      "epoch": 0.005152920968937377,
+      "grad_norm": 0.39661329984664917,
+      "learning_rate": 3.0330330330330332e-05,
+      "loss": 0.0985,
+      "step": 3942
+    },
+    {
+      "epoch": 0.005154228153353647,
+      "grad_norm": 1.2617276906967163,
+      "learning_rate": 3.0325325325325325e-05,
+      "loss": 1.93,
+      "step": 3943
+    },
+    {
+      "epoch": 0.005155535337769917,
+      "grad_norm": 1.3448374271392822,
+      "learning_rate": 3.032032032032032e-05,
+      "loss": 1.7189,
+      "step": 3944
+    },
+    {
+      "epoch": 0.0051568425221861875,
+      "grad_norm": 1.467301607131958,
+      "learning_rate": 3.0315315315315318e-05,
+      "loss": 1.8288,
+      "step": 3945
+    },
+    {
+      "epoch": 0.005158149706602458,
+      "grad_norm": 1.7216575145721436,
+      "learning_rate": 3.0310310310310314e-05,
+      "loss": 2.1672,
+      "step": 3946
+    },
+    {
+      "epoch": 0.005159456891018728,
+      "grad_norm": 1.4049893617630005,
+      "learning_rate": 3.030530530530531e-05,
+      "loss": 2.1515,
+      "step": 3947
+    },
+    {
+      "epoch": 0.0051607640754349985,
+      "grad_norm": 1.3189424276351929,
+      "learning_rate": 3.03003003003003e-05,
+      "loss": 1.5532,
+      "step": 3948
+    },
+    {
+      "epoch": 0.005162071259851269,
+      "grad_norm": 1.5764214992523193,
+      "learning_rate": 3.0295295295295297e-05,
+      "loss": 2.4288,
+      "step": 3949
+    },
+    {
+      "epoch": 0.005163378444267539,
+      "grad_norm": 1.7084904909133911,
+      "learning_rate": 3.029029029029029e-05,
+      "loss": 2.2538,
+      "step": 3950
+    },
+    {
+      "epoch": 0.0051646856286838095,
+      "grad_norm": 1.0062123537063599,
+      "learning_rate": 3.0285285285285286e-05,
+      "loss": 0.7487,
+      "step": 3951
+    },
+    {
+      "epoch": 0.005165992813100079,
+      "grad_norm": 1.389182686805725,
+      "learning_rate": 3.0280280280280282e-05,
+      "loss": 1.6571,
+      "step": 3952
+    },
+    {
+      "epoch": 0.005167299997516349,
+      "grad_norm": 1.282210350036621,
+      "learning_rate": 3.027527527527528e-05,
+      "loss": 1.687,
+      "step": 3953
+    },
+    {
+      "epoch": 0.0051686071819326196,
+      "grad_norm": 1.5306994915008545,
+      "learning_rate": 3.0270270270270272e-05,
+      "loss": 0.9264,
+      "step": 3954
+    },
+    {
+      "epoch": 0.00516991436634889,
+      "grad_norm": 1.62705397605896,
+      "learning_rate": 3.0265265265265268e-05,
+      "loss": 0.8769,
+      "step": 3955
+    },
+    {
+      "epoch": 0.00517122155076516,
+      "grad_norm": 1.1297496557235718,
+      "learning_rate": 3.0260260260260258e-05,
+      "loss": 1.5768,
+      "step": 3956
+    },
+    {
+      "epoch": 0.0051725287351814305,
+      "grad_norm": 1.0778602361679077,
+      "learning_rate": 3.0255255255255254e-05,
+      "loss": 1.4979,
+      "step": 3957
+    },
+    {
+      "epoch": 0.005173835919597701,
+      "grad_norm": 1.2032550573349,
+      "learning_rate": 3.025025025025025e-05,
+      "loss": 1.5283,
+      "step": 3958
+    },
+    {
+      "epoch": 0.005175143104013971,
+      "grad_norm": 1.4122638702392578,
+      "learning_rate": 3.0245245245245247e-05,
+      "loss": 2.0337,
+      "step": 3959
+    },
+    {
+      "epoch": 0.0051764502884302415,
+      "grad_norm": 1.2657437324523926,
+      "learning_rate": 3.0240240240240243e-05,
+      "loss": 1.2058,
+      "step": 3960
+    },
+    {
+      "epoch": 0.005177757472846512,
+      "grad_norm": 1.3228867053985596,
+      "learning_rate": 3.0235235235235236e-05,
+      "loss": 1.4182,
+      "step": 3961
+    },
+    {
+      "epoch": 0.005179064657262782,
+      "grad_norm": 1.4114117622375488,
+      "learning_rate": 3.0230230230230232e-05,
+      "loss": 1.5989,
+      "step": 3962
+    },
+    {
+      "epoch": 0.005180371841679052,
+      "grad_norm": 1.5709216594696045,
+      "learning_rate": 3.022522522522523e-05,
+      "loss": 1.7789,
+      "step": 3963
+    },
+    {
+      "epoch": 0.005181679026095323,
+      "grad_norm": 1.1502093076705933,
+      "learning_rate": 3.022022022022022e-05,
+      "loss": 1.4445,
+      "step": 3964
+    },
+    {
+      "epoch": 0.005182986210511593,
+      "grad_norm": 1.081997036933899,
+      "learning_rate": 3.0215215215215215e-05,
+      "loss": 0.4628,
+      "step": 3965
+    },
+    {
+      "epoch": 0.005184293394927863,
+      "grad_norm": 1.8606919050216675,
+      "learning_rate": 3.021021021021021e-05,
+      "loss": 1.6158,
+      "step": 3966
+    },
+    {
+      "epoch": 0.005185600579344134,
+      "grad_norm": 1.4685664176940918,
+      "learning_rate": 3.0205205205205208e-05,
+      "loss": 1.7,
+      "step": 3967
+    },
+    {
+      "epoch": 0.005186907763760403,
+      "grad_norm": 2.036262273788452,
+      "learning_rate": 3.02002002002002e-05,
+      "loss": 1.9996,
+      "step": 3968
+    },
+    {
+      "epoch": 0.0051882149481766735,
+      "grad_norm": 1.3020395040512085,
+      "learning_rate": 3.0195195195195197e-05,
+      "loss": 2.2164,
+      "step": 3969
+    },
+    {
+      "epoch": 0.005189522132592944,
+      "grad_norm": 1.8807430267333984,
+      "learning_rate": 3.0190190190190193e-05,
+      "loss": 1.4596,
+      "step": 3970
+    },
+    {
+      "epoch": 0.005190829317009214,
+      "grad_norm": 1.3569239377975464,
+      "learning_rate": 3.018518518518519e-05,
+      "loss": 1.2008,
+      "step": 3971
+    },
+    {
+      "epoch": 0.0051921365014254844,
+      "grad_norm": 1.5020999908447266,
+      "learning_rate": 3.0180180180180183e-05,
+      "loss": 2.0178,
+      "step": 3972
+    },
+    {
+      "epoch": 0.005193443685841755,
+      "grad_norm": 1.818529486656189,
+      "learning_rate": 3.0175175175175176e-05,
+      "loss": 2.2893,
+      "step": 3973
+    },
+    {
+      "epoch": 0.005194750870258025,
+      "grad_norm": 1.7429556846618652,
+      "learning_rate": 3.017017017017017e-05,
+      "loss": 2.0205,
+      "step": 3974
+    },
+    {
+      "epoch": 0.005196058054674295,
+      "grad_norm": 1.3398441076278687,
+      "learning_rate": 3.0165165165165165e-05,
+      "loss": 2.0438,
+      "step": 3975
+    },
+    {
+      "epoch": 0.005197365239090566,
+      "grad_norm": 1.6080973148345947,
+      "learning_rate": 3.016016016016016e-05,
+      "loss": 1.9848,
+      "step": 3976
+    },
+    {
+      "epoch": 0.005198672423506836,
+      "grad_norm": 1.59352445602417,
+      "learning_rate": 3.0155155155155158e-05,
+      "loss": 1.3002,
+      "step": 3977
+    },
+    {
+      "epoch": 0.005199979607923106,
+      "grad_norm": 1.3189865350723267,
+      "learning_rate": 3.0150150150150154e-05,
+      "loss": 1.5112,
+      "step": 3978
+    },
+    {
+      "epoch": 0.005201286792339377,
+      "grad_norm": 0.9983789324760437,
+      "learning_rate": 3.0145145145145147e-05,
+      "loss": 1.1272,
+      "step": 3979
+    },
+    {
+      "epoch": 0.005202593976755647,
+      "grad_norm": 1.3214930295944214,
+      "learning_rate": 3.0140140140140143e-05,
+      "loss": 1.3803,
+      "step": 3980
+    },
+    {
+      "epoch": 0.005203901161171917,
+      "grad_norm": 1.3970783948898315,
+      "learning_rate": 3.0135135135135133e-05,
+      "loss": 2.0669,
+      "step": 3981
+    },
+    {
+      "epoch": 0.005205208345588188,
+      "grad_norm": 1.3851169347763062,
+      "learning_rate": 3.013013013013013e-05,
+      "loss": 1.6852,
+      "step": 3982
+    },
+    {
+      "epoch": 0.005206515530004457,
+      "grad_norm": 2.470238208770752,
+      "learning_rate": 3.0125125125125126e-05,
+      "loss": 1.6011,
+      "step": 3983
+    },
+    {
+      "epoch": 0.005207822714420727,
+      "grad_norm": 1.4750674962997437,
+      "learning_rate": 3.0120120120120122e-05,
+      "loss": 1.4124,
+      "step": 3984
+    },
+    {
+      "epoch": 0.005209129898836998,
+      "grad_norm": 1.652298092842102,
+      "learning_rate": 3.011511511511512e-05,
+      "loss": 2.0957,
+      "step": 3985
+    },
+    {
+      "epoch": 0.005210437083253268,
+      "grad_norm": 1.6263561248779297,
+      "learning_rate": 3.011011011011011e-05,
+      "loss": 1.7622,
+      "step": 3986
+    },
+    {
+      "epoch": 0.005211744267669538,
+      "grad_norm": 1.4267874956130981,
+      "learning_rate": 3.0105105105105108e-05,
+      "loss": 1.8071,
+      "step": 3987
+    },
+    {
+      "epoch": 0.005213051452085809,
+      "grad_norm": 1.535159707069397,
+      "learning_rate": 3.0100100100100104e-05,
+      "loss": 1.5081,
+      "step": 3988
+    },
+    {
+      "epoch": 0.005214358636502079,
+      "grad_norm": 1.272867202758789,
+      "learning_rate": 3.0095095095095094e-05,
+      "loss": 1.2476,
+      "step": 3989
+    },
+    {
+      "epoch": 0.005215665820918349,
+      "grad_norm": 1.3325471878051758,
+      "learning_rate": 3.009009009009009e-05,
+      "loss": 1.3297,
+      "step": 3990
+    },
+    {
+      "epoch": 0.00521697300533462,
+      "grad_norm": 0.10624521970748901,
+      "learning_rate": 3.0085085085085086e-05,
+      "loss": 0.0035,
+      "step": 3991
+    },
+    {
+      "epoch": 0.00521828018975089,
+      "grad_norm": 1.4129120111465454,
+      "learning_rate": 3.008008008008008e-05,
+      "loss": 1.4697,
+      "step": 3992
+    },
+    {
+      "epoch": 0.00521958737416716,
+      "grad_norm": 1.5587170124053955,
+      "learning_rate": 3.0075075075075076e-05,
+      "loss": 2.1331,
+      "step": 3993
+    },
+    {
+      "epoch": 0.005220894558583431,
+      "grad_norm": 1.2794324159622192,
+      "learning_rate": 3.0070070070070072e-05,
+      "loss": 1.6604,
+      "step": 3994
+    },
+    {
+      "epoch": 0.005222201742999701,
+      "grad_norm": 1.8744522333145142,
+      "learning_rate": 3.006506506506507e-05,
+      "loss": 2.1301,
+      "step": 3995
+    },
+    {
+      "epoch": 0.005223508927415971,
+      "grad_norm": 1.3067083358764648,
+      "learning_rate": 3.0060060060060065e-05,
+      "loss": 1.0791,
+      "step": 3996
+    },
+    {
+      "epoch": 0.005224816111832242,
+      "grad_norm": 1.36367928981781,
+      "learning_rate": 3.0055055055055058e-05,
+      "loss": 1.7601,
+      "step": 3997
+    },
+    {
+      "epoch": 0.005226123296248511,
+      "grad_norm": 1.0701515674591064,
+      "learning_rate": 3.005005005005005e-05,
+      "loss": 0.9166,
+      "step": 3998
+    },
+    {
+      "epoch": 0.005227430480664781,
+      "grad_norm": 1.371238350868225,
+      "learning_rate": 3.0045045045045044e-05,
+      "loss": 1.3296,
+      "step": 3999
+    },
+    {
+      "epoch": 0.005228737665081052,
+      "grad_norm": 1.2824819087982178,
+      "learning_rate": 3.004004004004004e-05,
+      "loss": 1.1878,
+      "step": 4000
+    },
+    {
+      "epoch": 0.005230044849497322,
+      "grad_norm": 1.2916369438171387,
+      "learning_rate": 3.0035035035035036e-05,
+      "loss": 1.0952,
+      "step": 4001
+    },
+    {
+      "epoch": 0.005231352033913592,
+      "grad_norm": 1.1732114553451538,
+      "learning_rate": 3.0030030030030033e-05,
+      "loss": 0.8397,
+      "step": 4002
+    },
+    {
+      "epoch": 0.005232659218329863,
+      "grad_norm": 1.5335127115249634,
+      "learning_rate": 3.002502502502503e-05,
+      "loss": 1.597,
+      "step": 4003
+    },
+    {
+      "epoch": 0.005233966402746133,
+      "grad_norm": 1.3678678274154663,
+      "learning_rate": 3.0020020020020022e-05,
+      "loss": 1.8033,
+      "step": 4004
+    },
+    {
+      "epoch": 0.005235273587162403,
+      "grad_norm": 1.5905109643936157,
+      "learning_rate": 3.001501501501502e-05,
+      "loss": 1.4287,
+      "step": 4005
+    },
+    {
+      "epoch": 0.005236580771578674,
+      "grad_norm": 0.9627315998077393,
+      "learning_rate": 3.0010010010010008e-05,
+      "loss": 0.7091,
+      "step": 4006
+    },
+    {
+      "epoch": 0.005237887955994944,
+      "grad_norm": 1.4774770736694336,
+      "learning_rate": 3.0005005005005004e-05,
+      "loss": 1.8635,
+      "step": 4007
+    },
+    {
+      "epoch": 0.005239195140411214,
+      "grad_norm": 1.2251297235488892,
+      "learning_rate": 3e-05,
+      "loss": 1.4553,
+      "step": 4008
+    },
+    {
+      "epoch": 0.0052405023248274845,
+      "grad_norm": 3.726456642150879,
+      "learning_rate": 2.9994994994994997e-05,
+      "loss": 1.3996,
+      "step": 4009
+    },
+    {
+      "epoch": 0.005241809509243755,
+      "grad_norm": 1.1985440254211426,
+      "learning_rate": 2.998998998998999e-05,
+      "loss": 1.5065,
+      "step": 4010
+    },
+    {
+      "epoch": 0.005243116693660025,
+      "grad_norm": 2.101668119430542,
+      "learning_rate": 2.9984984984984986e-05,
+      "loss": 1.9513,
+      "step": 4011
+    },
+    {
+      "epoch": 0.0052444238780762955,
+      "grad_norm": 1.4120612144470215,
+      "learning_rate": 2.9979979979979983e-05,
+      "loss": 1.9982,
+      "step": 4012
+    },
+    {
+      "epoch": 0.005245731062492566,
+      "grad_norm": 1.6634056568145752,
+      "learning_rate": 2.997497497497498e-05,
+      "loss": 1.1196,
+      "step": 4013
+    },
+    {
+      "epoch": 0.005247038246908835,
+      "grad_norm": 0.8341153264045715,
+      "learning_rate": 2.996996996996997e-05,
+      "loss": 0.7328,
+      "step": 4014
+    },
+    {
+      "epoch": 0.005248345431325106,
+      "grad_norm": 1.411283254623413,
+      "learning_rate": 2.9964964964964965e-05,
+      "loss": 2.2131,
+      "step": 4015
+    },
+    {
+      "epoch": 0.005249652615741376,
+      "grad_norm": 1.5974432229995728,
+      "learning_rate": 2.995995995995996e-05,
+      "loss": 2.1508,
+      "step": 4016
+    },
+    {
+      "epoch": 0.005250959800157646,
+      "grad_norm": 1.3599892854690552,
+      "learning_rate": 2.9954954954954954e-05,
+      "loss": 1.2834,
+      "step": 4017
+    },
+    {
+      "epoch": 0.0052522669845739165,
+      "grad_norm": 1.2266435623168945,
+      "learning_rate": 2.994994994994995e-05,
+      "loss": 1.2547,
+      "step": 4018
+    },
+    {
+      "epoch": 0.005253574168990187,
+      "grad_norm": 1.8032587766647339,
+      "learning_rate": 2.9944944944944947e-05,
+      "loss": 2.104,
+      "step": 4019
+    },
+    {
+      "epoch": 0.005254881353406457,
+      "grad_norm": 1.5179250240325928,
+      "learning_rate": 2.9939939939939944e-05,
+      "loss": 1.7133,
+      "step": 4020
+    },
+    {
+      "epoch": 0.0052561885378227275,
+      "grad_norm": 1.4687103033065796,
+      "learning_rate": 2.993493493493494e-05,
+      "loss": 2.1631,
+      "step": 4021
+    },
+    {
+      "epoch": 0.005257495722238998,
+      "grad_norm": 1.5236855745315552,
+      "learning_rate": 2.9929929929929933e-05,
+      "loss": 1.541,
+      "step": 4022
+    },
+    {
+      "epoch": 0.005258802906655268,
+      "grad_norm": 1.0213513374328613,
+      "learning_rate": 2.9924924924924922e-05,
+      "loss": 1.0391,
+      "step": 4023
+    },
+    {
+      "epoch": 0.0052601100910715385,
+      "grad_norm": 1.277165174484253,
+      "learning_rate": 2.991991991991992e-05,
+      "loss": 1.428,
+      "step": 4024
+    },
+    {
+      "epoch": 0.005261417275487809,
+      "grad_norm": 1.41391122341156,
+      "learning_rate": 2.9914914914914915e-05,
+      "loss": 2.1036,
+      "step": 4025
+    },
+    {
+      "epoch": 0.005262724459904079,
+      "grad_norm": 1.6977667808532715,
+      "learning_rate": 2.990990990990991e-05,
+      "loss": 2.0315,
+      "step": 4026
+    },
+    {
+      "epoch": 0.005264031644320349,
+      "grad_norm": 0.6335197687149048,
+      "learning_rate": 2.9904904904904908e-05,
+      "loss": 0.3484,
+      "step": 4027
+    },
+    {
+      "epoch": 0.00526533882873662,
+      "grad_norm": 1.1168594360351562,
+      "learning_rate": 2.98998998998999e-05,
+      "loss": 1.5472,
+      "step": 4028
+    },
+    {
+      "epoch": 0.005266646013152889,
+      "grad_norm": 1.384838342666626,
+      "learning_rate": 2.9894894894894897e-05,
+      "loss": 2.3141,
+      "step": 4029
+    },
+    {
+      "epoch": 0.0052679531975691595,
+      "grad_norm": 1.5127724409103394,
+      "learning_rate": 2.9889889889889894e-05,
+      "loss": 1.2288,
+      "step": 4030
+    },
+    {
+      "epoch": 0.00526926038198543,
+      "grad_norm": 1.2457209825515747,
+      "learning_rate": 2.9884884884884883e-05,
+      "loss": 1.2959,
+      "step": 4031
+    },
+    {
+      "epoch": 0.0052705675664017,
+      "grad_norm": 1.4981776475906372,
+      "learning_rate": 2.987987987987988e-05,
+      "loss": 1.6573,
+      "step": 4032
+    },
+    {
+      "epoch": 0.0052718747508179705,
+      "grad_norm": 1.3369247913360596,
+      "learning_rate": 2.9874874874874876e-05,
+      "loss": 1.6071,
+      "step": 4033
+    },
+    {
+      "epoch": 0.005273181935234241,
+      "grad_norm": 1.5028166770935059,
+      "learning_rate": 2.9869869869869872e-05,
+      "loss": 1.2902,
+      "step": 4034
+    },
+    {
+      "epoch": 0.005274489119650511,
+      "grad_norm": 1.516501545906067,
+      "learning_rate": 2.9864864864864865e-05,
+      "loss": 1.4015,
+      "step": 4035
+    },
+    {
+      "epoch": 0.005275796304066781,
+      "grad_norm": 1.6612046957015991,
+      "learning_rate": 2.985985985985986e-05,
+      "loss": 0.9504,
+      "step": 4036
+    },
+    {
+      "epoch": 0.005277103488483052,
+      "grad_norm": 1.5363661050796509,
+      "learning_rate": 2.9854854854854858e-05,
+      "loss": 1.9785,
+      "step": 4037
+    },
+    {
+      "epoch": 0.005278410672899322,
+      "grad_norm": 1.4342427253723145,
+      "learning_rate": 2.9849849849849854e-05,
+      "loss": 2.3875,
+      "step": 4038
+    },
+    {
+      "epoch": 0.005279717857315592,
+      "grad_norm": 1.5097599029541016,
+      "learning_rate": 2.9844844844844844e-05,
+      "loss": 2.2282,
+      "step": 4039
+    },
+    {
+      "epoch": 0.005281025041731863,
+      "grad_norm": 1.5163568258285522,
+      "learning_rate": 2.983983983983984e-05,
+      "loss": 1.9805,
+      "step": 4040
+    },
+    {
+      "epoch": 0.005282332226148133,
+      "grad_norm": 1.1682405471801758,
+      "learning_rate": 2.9834834834834837e-05,
+      "loss": 1.2371,
+      "step": 4041
+    },
+    {
+      "epoch": 0.005283639410564403,
+      "grad_norm": 2.2486281394958496,
+      "learning_rate": 2.982982982982983e-05,
+      "loss": 2.3603,
+      "step": 4042
+    },
+    {
+      "epoch": 0.005284946594980674,
+      "grad_norm": 0.978038489818573,
+      "learning_rate": 2.9824824824824826e-05,
+      "loss": 0.8301,
+      "step": 4043
+    },
+    {
+      "epoch": 0.005286253779396943,
+      "grad_norm": 1.8141127824783325,
+      "learning_rate": 2.9819819819819822e-05,
+      "loss": 1.9677,
+      "step": 4044
+    },
+    {
+      "epoch": 0.0052875609638132134,
+      "grad_norm": 1.414255142211914,
+      "learning_rate": 2.981481481481482e-05,
+      "loss": 1.7135,
+      "step": 4045
+    },
+    {
+      "epoch": 0.005288868148229484,
+      "grad_norm": 1.3979578018188477,
+      "learning_rate": 2.9809809809809812e-05,
+      "loss": 2.2297,
+      "step": 4046
+    },
+    {
+      "epoch": 0.005290175332645754,
+      "grad_norm": 1.3528640270233154,
+      "learning_rate": 2.9804804804804808e-05,
+      "loss": 1.078,
+      "step": 4047
+    },
+    {
+      "epoch": 0.005291482517062024,
+      "grad_norm": 1.5461640357971191,
+      "learning_rate": 2.9799799799799798e-05,
+      "loss": 1.3527,
+      "step": 4048
+    },
+    {
+      "epoch": 0.005292789701478295,
+      "grad_norm": 2.2187678813934326,
+      "learning_rate": 2.9794794794794794e-05,
+      "loss": 1.8574,
+      "step": 4049
+    },
+    {
+      "epoch": 0.005294096885894565,
+      "grad_norm": 1.5476106405258179,
+      "learning_rate": 2.978978978978979e-05,
+      "loss": 1.1891,
+      "step": 4050
+    },
+    {
+      "epoch": 0.005295404070310835,
+      "grad_norm": 1.8600287437438965,
+      "learning_rate": 2.9784784784784787e-05,
+      "loss": 2.1869,
+      "step": 4051
+    },
+    {
+      "epoch": 0.005296711254727106,
+      "grad_norm": 1.8778189420700073,
+      "learning_rate": 2.9779779779779783e-05,
+      "loss": 2.7495,
+      "step": 4052
+    },
+    {
+      "epoch": 0.005298018439143376,
+      "grad_norm": 1.6939892768859863,
+      "learning_rate": 2.9774774774774776e-05,
+      "loss": 1.8657,
+      "step": 4053
+    },
+    {
+      "epoch": 0.005299325623559646,
+      "grad_norm": 1.19851553440094,
+      "learning_rate": 2.9769769769769772e-05,
+      "loss": 1.2315,
+      "step": 4054
+    },
+    {
+      "epoch": 0.005300632807975917,
+      "grad_norm": 1.6432510614395142,
+      "learning_rate": 2.976476476476477e-05,
+      "loss": 1.6485,
+      "step": 4055
+    },
+    {
+      "epoch": 0.005301939992392187,
+      "grad_norm": 1.3820788860321045,
+      "learning_rate": 2.975975975975976e-05,
+      "loss": 1.4788,
+      "step": 4056
+    },
+    {
+      "epoch": 0.005303247176808457,
+      "grad_norm": 1.3465074300765991,
+      "learning_rate": 2.9754754754754755e-05,
+      "loss": 1.3539,
+      "step": 4057
+    },
+    {
+      "epoch": 0.005304554361224728,
+      "grad_norm": 1.276322364807129,
+      "learning_rate": 2.974974974974975e-05,
+      "loss": 1.7147,
+      "step": 4058
+    },
+    {
+      "epoch": 0.005305861545640997,
+      "grad_norm": 1.1834782361984253,
+      "learning_rate": 2.9744744744744747e-05,
+      "loss": 1.0521,
+      "step": 4059
+    },
+    {
+      "epoch": 0.005307168730057267,
+      "grad_norm": 1.3292689323425293,
+      "learning_rate": 2.973973973973974e-05,
+      "loss": 1.4359,
+      "step": 4060
+    },
+    {
+      "epoch": 0.005308475914473538,
+      "grad_norm": 1.8679982423782349,
+      "learning_rate": 2.9734734734734737e-05,
+      "loss": 2.3744,
+      "step": 4061
+    },
+    {
+      "epoch": 0.005309783098889808,
+      "grad_norm": 1.484821081161499,
+      "learning_rate": 2.9729729729729733e-05,
+      "loss": 2.414,
+      "step": 4062
+    },
+    {
+      "epoch": 0.005311090283306078,
+      "grad_norm": 1.2264702320098877,
+      "learning_rate": 2.972472472472473e-05,
+      "loss": 1.3791,
+      "step": 4063
+    },
+    {
+      "epoch": 0.005312397467722349,
+      "grad_norm": 0.956763744354248,
+      "learning_rate": 2.971971971971972e-05,
+      "loss": 0.898,
+      "step": 4064
+    },
+    {
+      "epoch": 0.005313704652138619,
+      "grad_norm": 1.3402897119522095,
+      "learning_rate": 2.9714714714714715e-05,
+      "loss": 0.8604,
+      "step": 4065
+    },
+    {
+      "epoch": 0.005315011836554889,
+      "grad_norm": 1.4688493013381958,
+      "learning_rate": 2.970970970970971e-05,
+      "loss": 1.133,
+      "step": 4066
+    },
+    {
+      "epoch": 0.00531631902097116,
+      "grad_norm": 1.2965465784072876,
+      "learning_rate": 2.9704704704704705e-05,
+      "loss": 0.6013,
+      "step": 4067
+    },
+    {
+      "epoch": 0.00531762620538743,
+      "grad_norm": 1.416363000869751,
+      "learning_rate": 2.96996996996997e-05,
+      "loss": 1.1599,
+      "step": 4068
+    },
+    {
+      "epoch": 0.0053189333898037,
+      "grad_norm": 1.3424878120422363,
+      "learning_rate": 2.9694694694694698e-05,
+      "loss": 1.4798,
+      "step": 4069
+    },
+    {
+      "epoch": 0.005320240574219971,
+      "grad_norm": 1.6130867004394531,
+      "learning_rate": 2.9689689689689694e-05,
+      "loss": 1.8301,
+      "step": 4070
+    },
+    {
+      "epoch": 0.005321547758636241,
+      "grad_norm": 1.2378065586090088,
+      "learning_rate": 2.9684684684684687e-05,
+      "loss": 0.845,
+      "step": 4071
+    },
+    {
+      "epoch": 0.005322854943052511,
+      "grad_norm": 1.9770609140396118,
+      "learning_rate": 2.967967967967968e-05,
+      "loss": 1.5093,
+      "step": 4072
+    },
+    {
+      "epoch": 0.0053241621274687815,
+      "grad_norm": 2.9150550365448,
+      "learning_rate": 2.9674674674674673e-05,
+      "loss": 1.4756,
+      "step": 4073
+    },
+    {
+      "epoch": 0.005325469311885052,
+      "grad_norm": 1.7420226335525513,
+      "learning_rate": 2.966966966966967e-05,
+      "loss": 1.8158,
+      "step": 4074
+    },
+    {
+      "epoch": 0.005326776496301321,
+      "grad_norm": 1.405499815940857,
+      "learning_rate": 2.9664664664664666e-05,
+      "loss": 2.0278,
+      "step": 4075
+    },
+    {
+      "epoch": 0.005328083680717592,
+      "grad_norm": 1.2924232482910156,
+      "learning_rate": 2.9659659659659662e-05,
+      "loss": 1.3962,
+      "step": 4076
+    },
+    {
+      "epoch": 0.005329390865133862,
+      "grad_norm": 1.2252588272094727,
+      "learning_rate": 2.9654654654654658e-05,
+      "loss": 1.3396,
+      "step": 4077
+    },
+    {
+      "epoch": 0.005330698049550132,
+      "grad_norm": 1.4513376951217651,
+      "learning_rate": 2.964964964964965e-05,
+      "loss": 1.3578,
+      "step": 4078
+    },
+    {
+      "epoch": 0.005332005233966403,
+      "grad_norm": 1.252532720565796,
+      "learning_rate": 2.9644644644644648e-05,
+      "loss": 1.1181,
+      "step": 4079
+    },
+    {
+      "epoch": 0.005333312418382673,
+      "grad_norm": 1.4835253953933716,
+      "learning_rate": 2.9639639639639644e-05,
+      "loss": 1.0088,
+      "step": 4080
+    },
+    {
+      "epoch": 0.005334619602798943,
+      "grad_norm": 1.24605131149292,
+      "learning_rate": 2.9634634634634634e-05,
+      "loss": 1.5904,
+      "step": 4081
+    },
+    {
+      "epoch": 0.0053359267872152135,
+      "grad_norm": 1.7479842901229858,
+      "learning_rate": 2.962962962962963e-05,
+      "loss": 1.2117,
+      "step": 4082
+    },
+    {
+      "epoch": 0.005337233971631484,
+      "grad_norm": 1.2948482036590576,
+      "learning_rate": 2.9624624624624626e-05,
+      "loss": 0.5717,
+      "step": 4083
+    },
+    {
+      "epoch": 0.005338541156047754,
+      "grad_norm": 1.61135733127594,
+      "learning_rate": 2.961961961961962e-05,
+      "loss": 1.9354,
+      "step": 4084
+    },
+    {
+      "epoch": 0.0053398483404640245,
+      "grad_norm": 1.5231976509094238,
+      "learning_rate": 2.9614614614614616e-05,
+      "loss": 1.3906,
+      "step": 4085
+    },
+    {
+      "epoch": 0.005341155524880295,
+      "grad_norm": 1.6687772274017334,
+      "learning_rate": 2.9609609609609612e-05,
+      "loss": 2.0485,
+      "step": 4086
+    },
+    {
+      "epoch": 0.005342462709296565,
+      "grad_norm": 1.0947699546813965,
+      "learning_rate": 2.960460460460461e-05,
+      "loss": 0.5931,
+      "step": 4087
+    },
+    {
+      "epoch": 0.0053437698937128355,
+      "grad_norm": 2.3770132064819336,
+      "learning_rate": 2.9599599599599605e-05,
+      "loss": 2.0798,
+      "step": 4088
+    },
+    {
+      "epoch": 0.005345077078129106,
+      "grad_norm": 1.5586649179458618,
+      "learning_rate": 2.9594594594594594e-05,
+      "loss": 1.3866,
+      "step": 4089
+    },
+    {
+      "epoch": 0.005346384262545375,
+      "grad_norm": 1.7561349868774414,
+      "learning_rate": 2.958958958958959e-05,
+      "loss": 1.9537,
+      "step": 4090
+    },
+    {
+      "epoch": 0.0053476914469616455,
+      "grad_norm": 1.410833716392517,
+      "learning_rate": 2.9584584584584584e-05,
+      "loss": 1.6152,
+      "step": 4091
+    },
+    {
+      "epoch": 0.005348998631377916,
+      "grad_norm": 1.448752522468567,
+      "learning_rate": 2.957957957957958e-05,
+      "loss": 1.7743,
+      "step": 4092
+    },
+    {
+      "epoch": 0.005350305815794186,
+      "grad_norm": 1.158316731452942,
+      "learning_rate": 2.9574574574574576e-05,
+      "loss": 1.6676,
+      "step": 4093
+    },
+    {
+      "epoch": 0.0053516130002104565,
+      "grad_norm": 1.1778324842453003,
+      "learning_rate": 2.9569569569569573e-05,
+      "loss": 0.8498,
+      "step": 4094
+    },
+    {
+      "epoch": 0.005352920184626727,
+      "grad_norm": 1.2874423265457153,
+      "learning_rate": 2.956456456456457e-05,
+      "loss": 1.2325,
+      "step": 4095
+    },
+    {
+      "epoch": 0.005354227369042997,
+      "grad_norm": 0.8247398138046265,
+      "learning_rate": 2.9559559559559562e-05,
+      "loss": 0.5015,
+      "step": 4096
+    },
+    {
+      "epoch": 0.0053555345534592675,
+      "grad_norm": 1.594558596611023,
+      "learning_rate": 2.955455455455455e-05,
+      "loss": 1.68,
+      "step": 4097
+    },
+    {
+      "epoch": 0.005356841737875538,
+      "grad_norm": 1.0848050117492676,
+      "learning_rate": 2.9549549549549548e-05,
+      "loss": 1.1137,
+      "step": 4098
+    },
+    {
+      "epoch": 0.005358148922291808,
+      "grad_norm": 1.2536247968673706,
+      "learning_rate": 2.9544544544544544e-05,
+      "loss": 0.9907,
+      "step": 4099
+    },
+    {
+      "epoch": 0.005359456106708078,
+      "grad_norm": 1.460402011871338,
+      "learning_rate": 2.953953953953954e-05,
+      "loss": 0.9259,
+      "step": 4100
+    },
+    {
+      "epoch": 0.005360763291124349,
+      "grad_norm": 1.4824951887130737,
+      "learning_rate": 2.9534534534534537e-05,
+      "loss": 1.2259,
+      "step": 4101
+    },
+    {
+      "epoch": 0.005362070475540619,
+      "grad_norm": 1.071153998374939,
+      "learning_rate": 2.952952952952953e-05,
+      "loss": 0.1523,
+      "step": 4102
+    },
+    {
+      "epoch": 0.005363377659956889,
+      "grad_norm": 1.4466205835342407,
+      "learning_rate": 2.9524524524524526e-05,
+      "loss": 2.0074,
+      "step": 4103
+    },
+    {
+      "epoch": 0.00536468484437316,
+      "grad_norm": 1.38466215133667,
+      "learning_rate": 2.9519519519519523e-05,
+      "loss": 1.756,
+      "step": 4104
+    },
+    {
+      "epoch": 0.005365992028789429,
+      "grad_norm": 1.249909520149231,
+      "learning_rate": 2.951451451451452e-05,
+      "loss": 1.5122,
+      "step": 4105
+    },
+    {
+      "epoch": 0.0053672992132056995,
+      "grad_norm": 1.2940884828567505,
+      "learning_rate": 2.950950950950951e-05,
+      "loss": 1.9788,
+      "step": 4106
+    },
+    {
+      "epoch": 0.00536860639762197,
+      "grad_norm": 1.1578996181488037,
+      "learning_rate": 2.9504504504504505e-05,
+      "loss": 0.9572,
+      "step": 4107
+    },
+    {
+      "epoch": 0.00536991358203824,
+      "grad_norm": 1.5114784240722656,
+      "learning_rate": 2.94994994994995e-05,
+      "loss": 1.6153,
+      "step": 4108
+    },
+    {
+      "epoch": 0.0053712207664545104,
+      "grad_norm": 1.6149907112121582,
+      "learning_rate": 2.9494494494494494e-05,
+      "loss": 1.8905,
+      "step": 4109
+    },
+    {
+      "epoch": 0.005372527950870781,
+      "grad_norm": 1.962986707687378,
+      "learning_rate": 2.948948948948949e-05,
+      "loss": 1.548,
+      "step": 4110
+    },
+    {
+      "epoch": 0.005373835135287051,
+      "grad_norm": 1.7411539554595947,
+      "learning_rate": 2.9484484484484487e-05,
+      "loss": 1.7577,
+      "step": 4111
+    },
+    {
+      "epoch": 0.005375142319703321,
+      "grad_norm": 1.4991580247879028,
+      "learning_rate": 2.9479479479479484e-05,
+      "loss": 2.3463,
+      "step": 4112
+    },
+    {
+      "epoch": 0.005376449504119592,
+      "grad_norm": 1.5015610456466675,
+      "learning_rate": 2.947447447447448e-05,
+      "loss": 1.2522,
+      "step": 4113
+    },
+    {
+      "epoch": 0.005377756688535862,
+      "grad_norm": 1.8257182836532593,
+      "learning_rate": 2.946946946946947e-05,
+      "loss": 1.3818,
+      "step": 4114
+    },
+    {
+      "epoch": 0.005379063872952132,
+      "grad_norm": 1.2060786485671997,
+      "learning_rate": 2.9464464464464462e-05,
+      "loss": 0.6944,
+      "step": 4115
+    },
+    {
+      "epoch": 0.005380371057368403,
+      "grad_norm": 1.5658361911773682,
+      "learning_rate": 2.945945945945946e-05,
+      "loss": 2.0436,
+      "step": 4116
+    },
+    {
+      "epoch": 0.005381678241784673,
+      "grad_norm": 1.2601805925369263,
+      "learning_rate": 2.9454454454454455e-05,
+      "loss": 2.3827,
+      "step": 4117
+    },
+    {
+      "epoch": 0.005382985426200943,
+      "grad_norm": 1.7508758306503296,
+      "learning_rate": 2.944944944944945e-05,
+      "loss": 2.2073,
+      "step": 4118
+    },
+    {
+      "epoch": 0.005384292610617214,
+      "grad_norm": 1.2439030408859253,
+      "learning_rate": 2.9444444444444448e-05,
+      "loss": 1.0292,
+      "step": 4119
+    },
+    {
+      "epoch": 0.005385599795033484,
+      "grad_norm": 1.2200229167938232,
+      "learning_rate": 2.943943943943944e-05,
+      "loss": 1.8708,
+      "step": 4120
+    },
+    {
+      "epoch": 0.005386906979449753,
+      "grad_norm": 1.2839365005493164,
+      "learning_rate": 2.9434434434434437e-05,
+      "loss": 1.2836,
+      "step": 4121
+    },
+    {
+      "epoch": 0.005388214163866024,
+      "grad_norm": 1.2074635028839111,
+      "learning_rate": 2.9429429429429427e-05,
+      "loss": 0.9737,
+      "step": 4122
+    },
+    {
+      "epoch": 0.005389521348282294,
+      "grad_norm": 1.5010725259780884,
+      "learning_rate": 2.9424424424424423e-05,
+      "loss": 1.5979,
+      "step": 4123
+    },
+    {
+      "epoch": 0.005390828532698564,
+      "grad_norm": 0.8919773697853088,
+      "learning_rate": 2.941941941941942e-05,
+      "loss": 0.5451,
+      "step": 4124
+    },
+    {
+      "epoch": 0.005392135717114835,
+      "grad_norm": 1.288129448890686,
+      "learning_rate": 2.9414414414414416e-05,
+      "loss": 1.4396,
+      "step": 4125
+    },
+    {
+      "epoch": 0.005393442901531105,
+      "grad_norm": 1.2675108909606934,
+      "learning_rate": 2.9409409409409412e-05,
+      "loss": 1.4536,
+      "step": 4126
+    },
+    {
+      "epoch": 0.005394750085947375,
+      "grad_norm": 1.5927096605300903,
+      "learning_rate": 2.9404404404404405e-05,
+      "loss": 1.936,
+      "step": 4127
+    },
+    {
+      "epoch": 0.005396057270363646,
+      "grad_norm": 1.2619012594223022,
+      "learning_rate": 2.93993993993994e-05,
+      "loss": 1.8308,
+      "step": 4128
+    },
+    {
+      "epoch": 0.005397364454779916,
+      "grad_norm": 1.7294118404388428,
+      "learning_rate": 2.9394394394394398e-05,
+      "loss": 1.3222,
+      "step": 4129
+    },
+    {
+      "epoch": 0.005398671639196186,
+      "grad_norm": 1.121827483177185,
+      "learning_rate": 2.9389389389389394e-05,
+      "loss": 0.8356,
+      "step": 4130
+    },
+    {
+      "epoch": 0.005399978823612457,
+      "grad_norm": 1.3670960664749146,
+      "learning_rate": 2.9384384384384384e-05,
+      "loss": 1.6921,
+      "step": 4131
+    },
+    {
+      "epoch": 0.005401286008028727,
+      "grad_norm": 2.0093159675598145,
+      "learning_rate": 2.937937937937938e-05,
+      "loss": 1.7097,
+      "step": 4132
+    },
+    {
+      "epoch": 0.005402593192444997,
+      "grad_norm": 1.3391540050506592,
+      "learning_rate": 2.9374374374374373e-05,
+      "loss": 1.9497,
+      "step": 4133
+    },
+    {
+      "epoch": 0.0054039003768612676,
+      "grad_norm": 1.8273593187332153,
+      "learning_rate": 2.936936936936937e-05,
+      "loss": 2.3481,
+      "step": 4134
+    },
+    {
+      "epoch": 0.005405207561277538,
+      "grad_norm": 1.2771556377410889,
+      "learning_rate": 2.9364364364364366e-05,
+      "loss": 1.688,
+      "step": 4135
+    },
+    {
+      "epoch": 0.005406514745693807,
+      "grad_norm": 0.42988118529319763,
+      "learning_rate": 2.9359359359359362e-05,
+      "loss": 0.1023,
+      "step": 4136
+    },
+    {
+      "epoch": 0.005407821930110078,
+      "grad_norm": 1.3533769845962524,
+      "learning_rate": 2.935435435435436e-05,
+      "loss": 1.7218,
+      "step": 4137
+    },
+    {
+      "epoch": 0.005409129114526348,
+      "grad_norm": 0.5712968707084656,
+      "learning_rate": 2.934934934934935e-05,
+      "loss": 0.2834,
+      "step": 4138
+    },
+    {
+      "epoch": 0.005410436298942618,
+      "grad_norm": 0.3944425880908966,
+      "learning_rate": 2.9344344344344345e-05,
+      "loss": 0.0638,
+      "step": 4139
+    },
+    {
+      "epoch": 0.005411743483358889,
+      "grad_norm": 1.5170212984085083,
+      "learning_rate": 2.9339339339339338e-05,
+      "loss": 2.0775,
+      "step": 4140
+    },
+    {
+      "epoch": 0.005413050667775159,
+      "grad_norm": 1.4504379034042358,
+      "learning_rate": 2.9334334334334334e-05,
+      "loss": 2.2362,
+      "step": 4141
+    },
+    {
+      "epoch": 0.005414357852191429,
+      "grad_norm": 1.2232084274291992,
+      "learning_rate": 2.932932932932933e-05,
+      "loss": 1.5776,
+      "step": 4142
+    },
+    {
+      "epoch": 0.0054156650366077,
+      "grad_norm": 1.309281826019287,
+      "learning_rate": 2.9324324324324327e-05,
+      "loss": 1.6392,
+      "step": 4143
+    },
+    {
+      "epoch": 0.00541697222102397,
+      "grad_norm": 1.3173363208770752,
+      "learning_rate": 2.9319319319319323e-05,
+      "loss": 1.2395,
+      "step": 4144
+    },
+    {
+      "epoch": 0.00541827940544024,
+      "grad_norm": 1.339113712310791,
+      "learning_rate": 2.9314314314314316e-05,
+      "loss": 1.2524,
+      "step": 4145
+    },
+    {
+      "epoch": 0.0054195865898565105,
+      "grad_norm": 1.7485679388046265,
+      "learning_rate": 2.9309309309309312e-05,
+      "loss": 1.9277,
+      "step": 4146
+    },
+    {
+      "epoch": 0.005420893774272781,
+      "grad_norm": 1.398622989654541,
+      "learning_rate": 2.9304304304304302e-05,
+      "loss": 1.9966,
+      "step": 4147
+    },
+    {
+      "epoch": 0.005422200958689051,
+      "grad_norm": 1.513893485069275,
+      "learning_rate": 2.92992992992993e-05,
+      "loss": 1.0748,
+      "step": 4148
+    },
+    {
+      "epoch": 0.0054235081431053215,
+      "grad_norm": 1.7466142177581787,
+      "learning_rate": 2.9294294294294295e-05,
+      "loss": 1.5952,
+      "step": 4149
+    },
+    {
+      "epoch": 0.005424815327521592,
+      "grad_norm": 1.5772430896759033,
+      "learning_rate": 2.928928928928929e-05,
+      "loss": 1.8181,
+      "step": 4150
+    },
+    {
+      "epoch": 0.005426122511937861,
+      "grad_norm": 1.0802656412124634,
+      "learning_rate": 2.9284284284284284e-05,
+      "loss": 1.1314,
+      "step": 4151
+    },
+    {
+      "epoch": 0.005427429696354132,
+      "grad_norm": 1.4351768493652344,
+      "learning_rate": 2.927927927927928e-05,
+      "loss": 1.4993,
+      "step": 4152
+    },
+    {
+      "epoch": 0.005428736880770402,
+      "grad_norm": 1.5392018556594849,
+      "learning_rate": 2.9274274274274277e-05,
+      "loss": 1.7791,
+      "step": 4153
+    },
+    {
+      "epoch": 0.005430044065186672,
+      "grad_norm": 1.0799862146377563,
+      "learning_rate": 2.9269269269269273e-05,
+      "loss": 0.9802,
+      "step": 4154
+    },
+    {
+      "epoch": 0.0054313512496029425,
+      "grad_norm": 1.434422254562378,
+      "learning_rate": 2.926426426426427e-05,
+      "loss": 2.1232,
+      "step": 4155
+    },
+    {
+      "epoch": 0.005432658434019213,
+      "grad_norm": 1.0858607292175293,
+      "learning_rate": 2.925925925925926e-05,
+      "loss": 1.0806,
+      "step": 4156
+    },
+    {
+      "epoch": 0.005433965618435483,
+      "grad_norm": 1.2151899337768555,
+      "learning_rate": 2.9254254254254255e-05,
+      "loss": 1.9562,
+      "step": 4157
+    },
+    {
+      "epoch": 0.0054352728028517535,
+      "grad_norm": 1.5449118614196777,
+      "learning_rate": 2.924924924924925e-05,
+      "loss": 2.4299,
+      "step": 4158
+    },
+    {
+      "epoch": 0.005436579987268024,
+      "grad_norm": 2.363358497619629,
+      "learning_rate": 2.9244244244244245e-05,
+      "loss": 2.1465,
+      "step": 4159
+    },
+    {
+      "epoch": 0.005437887171684294,
+      "grad_norm": 1.40199875831604,
+      "learning_rate": 2.923923923923924e-05,
+      "loss": 2.1431,
+      "step": 4160
+    },
+    {
+      "epoch": 0.0054391943561005645,
+      "grad_norm": 1.7016685009002686,
+      "learning_rate": 2.9234234234234238e-05,
+      "loss": 2.1111,
+      "step": 4161
+    },
+    {
+      "epoch": 0.005440501540516835,
+      "grad_norm": 1.0990418195724487,
+      "learning_rate": 2.9229229229229234e-05,
+      "loss": 1.3881,
+      "step": 4162
+    },
+    {
+      "epoch": 0.005441808724933105,
+      "grad_norm": 1.534283995628357,
+      "learning_rate": 2.9224224224224227e-05,
+      "loss": 2.1458,
+      "step": 4163
+    },
+    {
+      "epoch": 0.005443115909349375,
+      "grad_norm": 1.4265902042388916,
+      "learning_rate": 2.921921921921922e-05,
+      "loss": 1.9369,
+      "step": 4164
+    },
+    {
+      "epoch": 0.005444423093765646,
+      "grad_norm": 1.3839584589004517,
+      "learning_rate": 2.9214214214214213e-05,
+      "loss": 2.3003,
+      "step": 4165
+    },
+    {
+      "epoch": 0.005445730278181916,
+      "grad_norm": 1.3635835647583008,
+      "learning_rate": 2.920920920920921e-05,
+      "loss": 1.6396,
+      "step": 4166
+    },
+    {
+      "epoch": 0.0054470374625981855,
+      "grad_norm": 1.3067107200622559,
+      "learning_rate": 2.9204204204204206e-05,
+      "loss": 1.3275,
+      "step": 4167
+    },
+    {
+      "epoch": 0.005448344647014456,
+      "grad_norm": 1.3304104804992676,
+      "learning_rate": 2.9199199199199202e-05,
+      "loss": 2.1103,
+      "step": 4168
+    },
+    {
+      "epoch": 0.005449651831430726,
+      "grad_norm": 1.729176640510559,
+      "learning_rate": 2.9194194194194198e-05,
+      "loss": 2.6383,
+      "step": 4169
+    },
+    {
+      "epoch": 0.0054509590158469965,
+      "grad_norm": 1.0136545896530151,
+      "learning_rate": 2.918918918918919e-05,
+      "loss": 0.7643,
+      "step": 4170
+    },
+    {
+      "epoch": 0.005452266200263267,
+      "grad_norm": 1.3618992567062378,
+      "learning_rate": 2.9184184184184188e-05,
+      "loss": 1.682,
+      "step": 4171
+    },
+    {
+      "epoch": 0.005453573384679537,
+      "grad_norm": 1.5882495641708374,
+      "learning_rate": 2.9179179179179177e-05,
+      "loss": 1.287,
+      "step": 4172
+    },
+    {
+      "epoch": 0.005454880569095807,
+      "grad_norm": 1.342618703842163,
+      "learning_rate": 2.9174174174174174e-05,
+      "loss": 2.5067,
+      "step": 4173
+    },
+    {
+      "epoch": 0.005456187753512078,
+      "grad_norm": 1.3568830490112305,
+      "learning_rate": 2.916916916916917e-05,
+      "loss": 1.8072,
+      "step": 4174
+    },
+    {
+      "epoch": 0.005457494937928348,
+      "grad_norm": 1.7406281232833862,
+      "learning_rate": 2.9164164164164166e-05,
+      "loss": 1.2055,
+      "step": 4175
+    },
+    {
+      "epoch": 0.005458802122344618,
+      "grad_norm": 1.276370882987976,
+      "learning_rate": 2.915915915915916e-05,
+      "loss": 1.3629,
+      "step": 4176
+    },
+    {
+      "epoch": 0.005460109306760889,
+      "grad_norm": 2.038780450820923,
+      "learning_rate": 2.9154154154154156e-05,
+      "loss": 2.0477,
+      "step": 4177
+    },
+    {
+      "epoch": 0.005461416491177159,
+      "grad_norm": 1.6250640153884888,
+      "learning_rate": 2.9149149149149152e-05,
+      "loss": 1.3783,
+      "step": 4178
+    },
+    {
+      "epoch": 0.005462723675593429,
+      "grad_norm": 1.26302969455719,
+      "learning_rate": 2.914414414414415e-05,
+      "loss": 1.4241,
+      "step": 4179
+    },
+    {
+      "epoch": 0.0054640308600097,
+      "grad_norm": 1.429371953010559,
+      "learning_rate": 2.9139139139139145e-05,
+      "loss": 1.1196,
+      "step": 4180
+    },
+    {
+      "epoch": 0.00546533804442597,
+      "grad_norm": 1.1238417625427246,
+      "learning_rate": 2.9134134134134134e-05,
+      "loss": 1.5248,
+      "step": 4181
+    },
+    {
+      "epoch": 0.0054666452288422394,
+      "grad_norm": 1.5224025249481201,
+      "learning_rate": 2.912912912912913e-05,
+      "loss": 1.2612,
+      "step": 4182
+    },
+    {
+      "epoch": 0.00546795241325851,
+      "grad_norm": 1.0428560972213745,
+      "learning_rate": 2.9124124124124124e-05,
+      "loss": 1.1933,
+      "step": 4183
+    },
+    {
+      "epoch": 0.00546925959767478,
+      "grad_norm": 1.5289918184280396,
+      "learning_rate": 2.911911911911912e-05,
+      "loss": 1.3711,
+      "step": 4184
+    },
+    {
+      "epoch": 0.00547056678209105,
+      "grad_norm": 0.9582993984222412,
+      "learning_rate": 2.9114114114114116e-05,
+      "loss": 0.677,
+      "step": 4185
+    },
+    {
+      "epoch": 0.005471873966507321,
+      "grad_norm": 1.3050169944763184,
+      "learning_rate": 2.9109109109109113e-05,
+      "loss": 1.4705,
+      "step": 4186
+    },
+    {
+      "epoch": 0.005473181150923591,
+      "grad_norm": 1.6039645671844482,
+      "learning_rate": 2.910410410410411e-05,
+      "loss": 1.9096,
+      "step": 4187
+    },
+    {
+      "epoch": 0.005474488335339861,
+      "grad_norm": 1.6438899040222168,
+      "learning_rate": 2.9099099099099102e-05,
+      "loss": 1.4444,
+      "step": 4188
+    },
+    {
+      "epoch": 0.005475795519756132,
+      "grad_norm": 1.5755305290222168,
+      "learning_rate": 2.909409409409409e-05,
+      "loss": 1.6197,
+      "step": 4189
+    },
+    {
+      "epoch": 0.005477102704172402,
+      "grad_norm": 1.314052700996399,
+      "learning_rate": 2.9089089089089088e-05,
+      "loss": 1.4682,
+      "step": 4190
+    },
+    {
+      "epoch": 0.005478409888588672,
+      "grad_norm": 1.8639711141586304,
+      "learning_rate": 2.9084084084084084e-05,
+      "loss": 1.6511,
+      "step": 4191
+    },
+    {
+      "epoch": 0.005479717073004943,
+      "grad_norm": 1.738631248474121,
+      "learning_rate": 2.907907907907908e-05,
+      "loss": 1.3725,
+      "step": 4192
+    },
+    {
+      "epoch": 0.005481024257421213,
+      "grad_norm": 1.657043695449829,
+      "learning_rate": 2.9074074074074077e-05,
+      "loss": 1.486,
+      "step": 4193
+    },
+    {
+      "epoch": 0.005482331441837483,
+      "grad_norm": 1.3719614744186401,
+      "learning_rate": 2.906906906906907e-05,
+      "loss": 0.9165,
+      "step": 4194
+    },
+    {
+      "epoch": 0.005483638626253754,
+      "grad_norm": 1.5638031959533691,
+      "learning_rate": 2.9064064064064066e-05,
+      "loss": 2.1465,
+      "step": 4195
+    },
+    {
+      "epoch": 0.005484945810670024,
+      "grad_norm": 1.9525598287582397,
+      "learning_rate": 2.9059059059059063e-05,
+      "loss": 1.6353,
+      "step": 4196
+    },
+    {
+      "epoch": 0.005486252995086293,
+      "grad_norm": 1.0231860876083374,
+      "learning_rate": 2.9054054054054052e-05,
+      "loss": 0.8683,
+      "step": 4197
+    },
+    {
+      "epoch": 0.005487560179502564,
+      "grad_norm": 1.6397080421447754,
+      "learning_rate": 2.904904904904905e-05,
+      "loss": 1.6128,
+      "step": 4198
+    },
+    {
+      "epoch": 0.005488867363918834,
+      "grad_norm": 1.5479964017868042,
+      "learning_rate": 2.9044044044044045e-05,
+      "loss": 1.811,
+      "step": 4199
+    },
+    {
+      "epoch": 0.005490174548335104,
+      "grad_norm": 1.086164951324463,
+      "learning_rate": 2.903903903903904e-05,
+      "loss": 0.9965,
+      "step": 4200
+    },
+    {
+      "epoch": 0.005491481732751375,
+      "grad_norm": 2.092073440551758,
+      "learning_rate": 2.9034034034034034e-05,
+      "loss": 1.7999,
+      "step": 4201
+    },
+    {
+      "epoch": 0.005492788917167645,
+      "grad_norm": 1.7532830238342285,
+      "learning_rate": 2.902902902902903e-05,
+      "loss": 1.7091,
+      "step": 4202
+    },
+    {
+      "epoch": 0.005494096101583915,
+      "grad_norm": 1.3610460758209229,
+      "learning_rate": 2.9024024024024027e-05,
+      "loss": 1.7028,
+      "step": 4203
+    },
+    {
+      "epoch": 0.005495403286000186,
+      "grad_norm": 1.7217621803283691,
+      "learning_rate": 2.9019019019019023e-05,
+      "loss": 1.829,
+      "step": 4204
+    },
+    {
+      "epoch": 0.005496710470416456,
+      "grad_norm": 1.506975769996643,
+      "learning_rate": 2.9014014014014013e-05,
+      "loss": 0.7152,
+      "step": 4205
+    },
+    {
+      "epoch": 0.005498017654832726,
+      "grad_norm": 1.250571846961975,
+      "learning_rate": 2.900900900900901e-05,
+      "loss": 1.9064,
+      "step": 4206
+    },
+    {
+      "epoch": 0.0054993248392489966,
+      "grad_norm": 1.1317660808563232,
+      "learning_rate": 2.9004004004004002e-05,
+      "loss": 1.2174,
+      "step": 4207
+    },
+    {
+      "epoch": 0.005500632023665267,
+      "grad_norm": 1.7498847246170044,
+      "learning_rate": 2.8998998998999e-05,
+      "loss": 2.427,
+      "step": 4208
+    },
+    {
+      "epoch": 0.005501939208081537,
+      "grad_norm": 2.119338035583496,
+      "learning_rate": 2.8993993993993995e-05,
+      "loss": 0.9906,
+      "step": 4209
+    },
+    {
+      "epoch": 0.0055032463924978075,
+      "grad_norm": 1.1812905073165894,
+      "learning_rate": 2.898898898898899e-05,
+      "loss": 1.0145,
+      "step": 4210
+    },
+    {
+      "epoch": 0.005504553576914078,
+      "grad_norm": 1.4701709747314453,
+      "learning_rate": 2.8983983983983988e-05,
+      "loss": 1.3504,
+      "step": 4211
+    },
+    {
+      "epoch": 0.005505860761330347,
+      "grad_norm": 1.2842655181884766,
+      "learning_rate": 2.897897897897898e-05,
+      "loss": 1.6702,
+      "step": 4212
+    },
+    {
+      "epoch": 0.005507167945746618,
+      "grad_norm": 0.834423840045929,
+      "learning_rate": 2.8973973973973977e-05,
+      "loss": 0.4888,
+      "step": 4213
+    },
+    {
+      "epoch": 0.005508475130162888,
+      "grad_norm": 1.1556131839752197,
+      "learning_rate": 2.8968968968968967e-05,
+      "loss": 1.1316,
+      "step": 4214
+    },
+    {
+      "epoch": 0.005509782314579158,
+      "grad_norm": 0.7447503209114075,
+      "learning_rate": 2.8963963963963963e-05,
+      "loss": 0.3288,
+      "step": 4215
+    },
+    {
+      "epoch": 0.005511089498995429,
+      "grad_norm": 1.8828060626983643,
+      "learning_rate": 2.895895895895896e-05,
+      "loss": 1.4985,
+      "step": 4216
+    },
+    {
+      "epoch": 0.005512396683411699,
+      "grad_norm": 1.4596177339553833,
+      "learning_rate": 2.8953953953953956e-05,
+      "loss": 1.853,
+      "step": 4217
+    },
+    {
+      "epoch": 0.005513703867827969,
+      "grad_norm": 1.5201256275177002,
+      "learning_rate": 2.8948948948948952e-05,
+      "loss": 2.2352,
+      "step": 4218
+    },
+    {
+      "epoch": 0.0055150110522442395,
+      "grad_norm": 1.391777515411377,
+      "learning_rate": 2.8943943943943945e-05,
+      "loss": 1.2746,
+      "step": 4219
+    },
+    {
+      "epoch": 0.00551631823666051,
+      "grad_norm": 0.950370192527771,
+      "learning_rate": 2.893893893893894e-05,
+      "loss": 0.7342,
+      "step": 4220
+    },
+    {
+      "epoch": 0.00551762542107678,
+      "grad_norm": 1.4749772548675537,
+      "learning_rate": 2.8933933933933938e-05,
+      "loss": 1.5455,
+      "step": 4221
+    },
+    {
+      "epoch": 0.0055189326054930505,
+      "grad_norm": 1.1712368726730347,
+      "learning_rate": 2.8928928928928928e-05,
+      "loss": 1.4744,
+      "step": 4222
+    },
+    {
+      "epoch": 0.005520239789909321,
+      "grad_norm": 1.3482296466827393,
+      "learning_rate": 2.8923923923923924e-05,
+      "loss": 0.9975,
+      "step": 4223
+    },
+    {
+      "epoch": 0.005521546974325591,
+      "grad_norm": 1.4106783866882324,
+      "learning_rate": 2.891891891891892e-05,
+      "loss": 2.0068,
+      "step": 4224
+    },
+    {
+      "epoch": 0.0055228541587418615,
+      "grad_norm": 1.4403020143508911,
+      "learning_rate": 2.8913913913913913e-05,
+      "loss": 1.6688,
+      "step": 4225
+    },
+    {
+      "epoch": 0.005524161343158132,
+      "grad_norm": 1.0724825859069824,
+      "learning_rate": 2.890890890890891e-05,
+      "loss": 1.0229,
+      "step": 4226
+    },
+    {
+      "epoch": 0.005525468527574402,
+      "grad_norm": 1.4714053869247437,
+      "learning_rate": 2.8903903903903906e-05,
+      "loss": 1.3696,
+      "step": 4227
+    },
+    {
+      "epoch": 0.0055267757119906715,
+      "grad_norm": 1.674062728881836,
+      "learning_rate": 2.8898898898898902e-05,
+      "loss": 2.6816,
+      "step": 4228
+    },
+    {
+      "epoch": 0.005528082896406942,
+      "grad_norm": 1.3610457181930542,
+      "learning_rate": 2.88938938938939e-05,
+      "loss": 1.1134,
+      "step": 4229
+    },
+    {
+      "epoch": 0.005529390080823212,
+      "grad_norm": 1.7344157695770264,
+      "learning_rate": 2.8888888888888888e-05,
+      "loss": 2.3445,
+      "step": 4230
+    },
+    {
+      "epoch": 0.0055306972652394825,
+      "grad_norm": 1.5322017669677734,
+      "learning_rate": 2.8883883883883885e-05,
+      "loss": 1.6136,
+      "step": 4231
+    },
+    {
+      "epoch": 0.005532004449655753,
+      "grad_norm": 1.3956323862075806,
+      "learning_rate": 2.8878878878878878e-05,
+      "loss": 1.3615,
+      "step": 4232
+    },
+    {
+      "epoch": 0.005533311634072023,
+      "grad_norm": 1.4928953647613525,
+      "learning_rate": 2.8873873873873874e-05,
+      "loss": 2.2725,
+      "step": 4233
+    },
+    {
+      "epoch": 0.0055346188184882935,
+      "grad_norm": 1.3267854452133179,
+      "learning_rate": 2.886886886886887e-05,
+      "loss": 1.254,
+      "step": 4234
+    },
+    {
+      "epoch": 0.005535926002904564,
+      "grad_norm": 2.1886637210845947,
+      "learning_rate": 2.8863863863863867e-05,
+      "loss": 1.348,
+      "step": 4235
+    },
+    {
+      "epoch": 0.005537233187320834,
+      "grad_norm": 1.8258168697357178,
+      "learning_rate": 2.8858858858858863e-05,
+      "loss": 2.3233,
+      "step": 4236
+    },
+    {
+      "epoch": 0.005538540371737104,
+      "grad_norm": 1.5227609872817993,
+      "learning_rate": 2.8853853853853856e-05,
+      "loss": 1.8894,
+      "step": 4237
+    },
+    {
+      "epoch": 0.005539847556153375,
+      "grad_norm": 1.358924150466919,
+      "learning_rate": 2.8848848848848852e-05,
+      "loss": 1.664,
+      "step": 4238
+    },
+    {
+      "epoch": 0.005541154740569645,
+      "grad_norm": 1.5230141878128052,
+      "learning_rate": 2.8843843843843842e-05,
+      "loss": 2.2424,
+      "step": 4239
+    },
+    {
+      "epoch": 0.005542461924985915,
+      "grad_norm": 1.3749489784240723,
+      "learning_rate": 2.883883883883884e-05,
+      "loss": 1.9691,
+      "step": 4240
+    },
+    {
+      "epoch": 0.005543769109402186,
+      "grad_norm": 1.808670997619629,
+      "learning_rate": 2.8833833833833835e-05,
+      "loss": 1.9372,
+      "step": 4241
+    },
+    {
+      "epoch": 0.005545076293818456,
+      "grad_norm": 1.6644052267074585,
+      "learning_rate": 2.882882882882883e-05,
+      "loss": 2.5546,
+      "step": 4242
+    },
+    {
+      "epoch": 0.0055463834782347255,
+      "grad_norm": 1.5874078273773193,
+      "learning_rate": 2.8823823823823824e-05,
+      "loss": 1.574,
+      "step": 4243
+    },
+    {
+      "epoch": 0.005547690662650996,
+      "grad_norm": 2.2482244968414307,
+      "learning_rate": 2.881881881881882e-05,
+      "loss": 1.4651,
+      "step": 4244
+    },
+    {
+      "epoch": 0.005548997847067266,
+      "grad_norm": 2.618257761001587,
+      "learning_rate": 2.8813813813813817e-05,
+      "loss": 2.2826,
+      "step": 4245
+    },
+    {
+      "epoch": 0.005550305031483536,
+      "grad_norm": 1.252557635307312,
+      "learning_rate": 2.8808808808808813e-05,
+      "loss": 1.1494,
+      "step": 4246
+    },
+    {
+      "epoch": 0.005551612215899807,
+      "grad_norm": 1.520745873451233,
+      "learning_rate": 2.8803803803803803e-05,
+      "loss": 1.0301,
+      "step": 4247
+    },
+    {
+      "epoch": 0.005552919400316077,
+      "grad_norm": 1.0597611665725708,
+      "learning_rate": 2.87987987987988e-05,
+      "loss": 0.4603,
+      "step": 4248
+    },
+    {
+      "epoch": 0.005554226584732347,
+      "grad_norm": 1.3895390033721924,
+      "learning_rate": 2.8793793793793795e-05,
+      "loss": 1.552,
+      "step": 4249
+    },
+    {
+      "epoch": 0.005555533769148618,
+      "grad_norm": 1.363867998123169,
+      "learning_rate": 2.878878878878879e-05,
+      "loss": 2.0237,
+      "step": 4250
+    },
+    {
+      "epoch": 0.005556840953564888,
+      "grad_norm": 1.2985750436782837,
+      "learning_rate": 2.8783783783783785e-05,
+      "loss": 2.0791,
+      "step": 4251
+    },
+    {
+      "epoch": 0.005558148137981158,
+      "grad_norm": 1.1552602052688599,
+      "learning_rate": 2.877877877877878e-05,
+      "loss": 1.2024,
+      "step": 4252
+    },
+    {
+      "epoch": 0.005559455322397429,
+      "grad_norm": 2.141451835632324,
+      "learning_rate": 2.8773773773773777e-05,
+      "loss": 2.597,
+      "step": 4253
+    },
+    {
+      "epoch": 0.005560762506813699,
+      "grad_norm": 1.4751765727996826,
+      "learning_rate": 2.8768768768768774e-05,
+      "loss": 2.3004,
+      "step": 4254
+    },
+    {
+      "epoch": 0.005562069691229969,
+      "grad_norm": 1.3879584074020386,
+      "learning_rate": 2.8763763763763763e-05,
+      "loss": 1.1176,
+      "step": 4255
+    },
+    {
+      "epoch": 0.00556337687564624,
+      "grad_norm": 1.524890422821045,
+      "learning_rate": 2.875875875875876e-05,
+      "loss": 1.6836,
+      "step": 4256
+    },
+    {
+      "epoch": 0.00556468406006251,
+      "grad_norm": 1.3520563840866089,
+      "learning_rate": 2.8753753753753753e-05,
+      "loss": 1.7082,
+      "step": 4257
+    },
+    {
+      "epoch": 0.005565991244478779,
+      "grad_norm": 1.2732563018798828,
+      "learning_rate": 2.874874874874875e-05,
+      "loss": 1.3663,
+      "step": 4258
+    },
+    {
+      "epoch": 0.00556729842889505,
+      "grad_norm": 1.8641862869262695,
+      "learning_rate": 2.8743743743743745e-05,
+      "loss": 1.5034,
+      "step": 4259
+    },
+    {
+      "epoch": 0.00556860561331132,
+      "grad_norm": 1.4239312410354614,
+      "learning_rate": 2.8738738738738742e-05,
+      "loss": 2.1412,
+      "step": 4260
+    },
+    {
+      "epoch": 0.00556991279772759,
+      "grad_norm": 1.4045958518981934,
+      "learning_rate": 2.8733733733733735e-05,
+      "loss": 2.1093,
+      "step": 4261
+    },
+    {
+      "epoch": 0.005571219982143861,
+      "grad_norm": 0.9058089852333069,
+      "learning_rate": 2.872872872872873e-05,
+      "loss": 0.4339,
+      "step": 4262
+    },
+    {
+      "epoch": 0.005572527166560131,
+      "grad_norm": 1.3097726106643677,
+      "learning_rate": 2.8723723723723728e-05,
+      "loss": 1.7667,
+      "step": 4263
+    },
+    {
+      "epoch": 0.005573834350976401,
+      "grad_norm": 1.2843130826950073,
+      "learning_rate": 2.8718718718718717e-05,
+      "loss": 1.5219,
+      "step": 4264
+    },
+    {
+      "epoch": 0.005575141535392672,
+      "grad_norm": 1.5980618000030518,
+      "learning_rate": 2.8713713713713713e-05,
+      "loss": 1.8795,
+      "step": 4265
+    },
+    {
+      "epoch": 0.005576448719808942,
+      "grad_norm": 1.4557859897613525,
+      "learning_rate": 2.870870870870871e-05,
+      "loss": 1.1557,
+      "step": 4266
+    },
+    {
+      "epoch": 0.005577755904225212,
+      "grad_norm": 1.171390175819397,
+      "learning_rate": 2.8703703703703706e-05,
+      "loss": 1.6091,
+      "step": 4267
+    },
+    {
+      "epoch": 0.005579063088641483,
+      "grad_norm": 1.2827117443084717,
+      "learning_rate": 2.86986986986987e-05,
+      "loss": 1.6291,
+      "step": 4268
+    },
+    {
+      "epoch": 0.005580370273057753,
+      "grad_norm": 1.3634698390960693,
+      "learning_rate": 2.8693693693693696e-05,
+      "loss": 1.6091,
+      "step": 4269
+    },
+    {
+      "epoch": 0.005581677457474023,
+      "grad_norm": 1.2937051057815552,
+      "learning_rate": 2.8688688688688692e-05,
+      "loss": 1.0203,
+      "step": 4270
+    },
+    {
+      "epoch": 0.0055829846418902936,
+      "grad_norm": 1.020224690437317,
+      "learning_rate": 2.8683683683683688e-05,
+      "loss": 1.0269,
+      "step": 4271
+    },
+    {
+      "epoch": 0.005584291826306564,
+      "grad_norm": 1.0403454303741455,
+      "learning_rate": 2.8678678678678678e-05,
+      "loss": 0.9339,
+      "step": 4272
+    },
+    {
+      "epoch": 0.005585599010722834,
+      "grad_norm": 1.27135169506073,
+      "learning_rate": 2.8673673673673674e-05,
+      "loss": 1.3564,
+      "step": 4273
+    },
+    {
+      "epoch": 0.005586906195139104,
+      "grad_norm": 2.0754501819610596,
+      "learning_rate": 2.866866866866867e-05,
+      "loss": 1.8619,
+      "step": 4274
+    },
+    {
+      "epoch": 0.005588213379555374,
+      "grad_norm": 1.5041426420211792,
+      "learning_rate": 2.8663663663663664e-05,
+      "loss": 1.6024,
+      "step": 4275
+    },
+    {
+      "epoch": 0.005589520563971644,
+      "grad_norm": 1.2857162952423096,
+      "learning_rate": 2.865865865865866e-05,
+      "loss": 1.4205,
+      "step": 4276
+    },
+    {
+      "epoch": 0.005590827748387915,
+      "grad_norm": 1.620589256286621,
+      "learning_rate": 2.8653653653653656e-05,
+      "loss": 2.2664,
+      "step": 4277
+    },
+    {
+      "epoch": 0.005592134932804185,
+      "grad_norm": 1.1413763761520386,
+      "learning_rate": 2.8648648648648653e-05,
+      "loss": 1.7671,
+      "step": 4278
+    },
+    {
+      "epoch": 0.005593442117220455,
+      "grad_norm": 1.5563832521438599,
+      "learning_rate": 2.8643643643643646e-05,
+      "loss": 2.1173,
+      "step": 4279
+    },
+    {
+      "epoch": 0.005594749301636726,
+      "grad_norm": 2.1280593872070312,
+      "learning_rate": 2.863863863863864e-05,
+      "loss": 1.7272,
+      "step": 4280
+    },
+    {
+      "epoch": 0.005596056486052996,
+      "grad_norm": 1.2858173847198486,
+      "learning_rate": 2.863363363363363e-05,
+      "loss": 1.802,
+      "step": 4281
+    },
+    {
+      "epoch": 0.005597363670469266,
+      "grad_norm": 0.7956651449203491,
+      "learning_rate": 2.8628628628628628e-05,
+      "loss": 0.2504,
+      "step": 4282
+    },
+    {
+      "epoch": 0.0055986708548855365,
+      "grad_norm": 1.6259052753448486,
+      "learning_rate": 2.8623623623623624e-05,
+      "loss": 2.0568,
+      "step": 4283
+    },
+    {
+      "epoch": 0.005599978039301807,
+      "grad_norm": 1.622543215751648,
+      "learning_rate": 2.861861861861862e-05,
+      "loss": 2.2747,
+      "step": 4284
+    },
+    {
+      "epoch": 0.005601285223718077,
+      "grad_norm": 2.033503532409668,
+      "learning_rate": 2.8613613613613617e-05,
+      "loss": 2.1424,
+      "step": 4285
+    },
+    {
+      "epoch": 0.0056025924081343475,
+      "grad_norm": 1.8267107009887695,
+      "learning_rate": 2.860860860860861e-05,
+      "loss": 1.7011,
+      "step": 4286
+    },
+    {
+      "epoch": 0.005603899592550618,
+      "grad_norm": 0.9420508146286011,
+      "learning_rate": 2.8603603603603606e-05,
+      "loss": 0.5231,
+      "step": 4287
+    },
+    {
+      "epoch": 0.005605206776966888,
+      "grad_norm": 1.6445866823196411,
+      "learning_rate": 2.8598598598598603e-05,
+      "loss": 2.1724,
+      "step": 4288
+    },
+    {
+      "epoch": 0.005606513961383158,
+      "grad_norm": 1.1826239824295044,
+      "learning_rate": 2.8593593593593592e-05,
+      "loss": 1.0689,
+      "step": 4289
+    },
+    {
+      "epoch": 0.005607821145799428,
+      "grad_norm": 1.9088611602783203,
+      "learning_rate": 2.858858858858859e-05,
+      "loss": 0.8771,
+      "step": 4290
+    },
+    {
+      "epoch": 0.005609128330215698,
+      "grad_norm": 1.6188392639160156,
+      "learning_rate": 2.8583583583583585e-05,
+      "loss": 2.1099,
+      "step": 4291
+    },
+    {
+      "epoch": 0.0056104355146319685,
+      "grad_norm": 1.1412091255187988,
+      "learning_rate": 2.857857857857858e-05,
+      "loss": 1.1894,
+      "step": 4292
+    },
+    {
+      "epoch": 0.005611742699048239,
+      "grad_norm": 1.5537430047988892,
+      "learning_rate": 2.8573573573573574e-05,
+      "loss": 1.9099,
+      "step": 4293
+    },
+    {
+      "epoch": 0.005613049883464509,
+      "grad_norm": 1.3783318996429443,
+      "learning_rate": 2.856856856856857e-05,
+      "loss": 1.783,
+      "step": 4294
+    },
+    {
+      "epoch": 0.0056143570678807795,
+      "grad_norm": 1.431734323501587,
+      "learning_rate": 2.8563563563563567e-05,
+      "loss": 1.6736,
+      "step": 4295
+    },
+    {
+      "epoch": 0.00561566425229705,
+      "grad_norm": 1.30081045627594,
+      "learning_rate": 2.8558558558558563e-05,
+      "loss": 1.4242,
+      "step": 4296
+    },
+    {
+      "epoch": 0.00561697143671332,
+      "grad_norm": 1.451080322265625,
+      "learning_rate": 2.8553553553553553e-05,
+      "loss": 1.5964,
+      "step": 4297
+    },
+    {
+      "epoch": 0.0056182786211295905,
+      "grad_norm": 0.6339301466941833,
+      "learning_rate": 2.854854854854855e-05,
+      "loss": 0.3634,
+      "step": 4298
+    },
+    {
+      "epoch": 0.005619585805545861,
+      "grad_norm": 1.157254934310913,
+      "learning_rate": 2.8543543543543542e-05,
+      "loss": 1.0451,
+      "step": 4299
+    },
+    {
+      "epoch": 0.005620892989962131,
+      "grad_norm": 1.2376325130462646,
+      "learning_rate": 2.853853853853854e-05,
+      "loss": 1.4992,
+      "step": 4300
+    },
+    {
+      "epoch": 0.005622200174378401,
+      "grad_norm": 1.460959553718567,
+      "learning_rate": 2.8533533533533535e-05,
+      "loss": 1.2902,
+      "step": 4301
+    },
+    {
+      "epoch": 0.005623507358794672,
+      "grad_norm": 1.383216142654419,
+      "learning_rate": 2.852852852852853e-05,
+      "loss": 0.7648,
+      "step": 4302
+    },
+    {
+      "epoch": 0.005624814543210942,
+      "grad_norm": 1.5059325695037842,
+      "learning_rate": 2.8523523523523528e-05,
+      "loss": 2.2083,
+      "step": 4303
+    },
+    {
+      "epoch": 0.0056261217276272115,
+      "grad_norm": 1.6451513767242432,
+      "learning_rate": 2.851851851851852e-05,
+      "loss": 2.0027,
+      "step": 4304
+    },
+    {
+      "epoch": 0.005627428912043482,
+      "grad_norm": 1.2840052843093872,
+      "learning_rate": 2.8513513513513514e-05,
+      "loss": 1.2143,
+      "step": 4305
+    },
+    {
+      "epoch": 0.005628736096459752,
+      "grad_norm": 1.1847655773162842,
+      "learning_rate": 2.8508508508508507e-05,
+      "loss": 1.0142,
+      "step": 4306
+    },
+    {
+      "epoch": 0.0056300432808760225,
+      "grad_norm": 1.5710846185684204,
+      "learning_rate": 2.8503503503503503e-05,
+      "loss": 1.5712,
+      "step": 4307
+    },
+    {
+      "epoch": 0.005631350465292293,
+      "grad_norm": 1.763695240020752,
+      "learning_rate": 2.84984984984985e-05,
+      "loss": 1.453,
+      "step": 4308
+    },
+    {
+      "epoch": 0.005632657649708563,
+      "grad_norm": 1.671021819114685,
+      "learning_rate": 2.8493493493493496e-05,
+      "loss": 2.1698,
+      "step": 4309
+    },
+    {
+      "epoch": 0.005633964834124833,
+      "grad_norm": 1.6881579160690308,
+      "learning_rate": 2.8488488488488492e-05,
+      "loss": 1.3343,
+      "step": 4310
+    },
+    {
+      "epoch": 0.005635272018541104,
+      "grad_norm": 1.267035722732544,
+      "learning_rate": 2.8483483483483485e-05,
+      "loss": 1.7547,
+      "step": 4311
+    },
+    {
+      "epoch": 0.005636579202957374,
+      "grad_norm": 1.0128989219665527,
+      "learning_rate": 2.847847847847848e-05,
+      "loss": 0.7723,
+      "step": 4312
+    },
+    {
+      "epoch": 0.005637886387373644,
+      "grad_norm": 2.1057817935943604,
+      "learning_rate": 2.8473473473473478e-05,
+      "loss": 1.3939,
+      "step": 4313
+    },
+    {
+      "epoch": 0.005639193571789915,
+      "grad_norm": 1.370462417602539,
+      "learning_rate": 2.8468468468468467e-05,
+      "loss": 1.1006,
+      "step": 4314
+    },
+    {
+      "epoch": 0.005640500756206185,
+      "grad_norm": 1.355556845664978,
+      "learning_rate": 2.8463463463463464e-05,
+      "loss": 2.0037,
+      "step": 4315
+    },
+    {
+      "epoch": 0.005641807940622455,
+      "grad_norm": 1.6993882656097412,
+      "learning_rate": 2.845845845845846e-05,
+      "loss": 2.234,
+      "step": 4316
+    },
+    {
+      "epoch": 0.005643115125038726,
+      "grad_norm": 1.3555034399032593,
+      "learning_rate": 2.8453453453453453e-05,
+      "loss": 1.0993,
+      "step": 4317
+    },
+    {
+      "epoch": 0.005644422309454996,
+      "grad_norm": 1.1703836917877197,
+      "learning_rate": 2.844844844844845e-05,
+      "loss": 0.714,
+      "step": 4318
+    },
+    {
+      "epoch": 0.0056457294938712654,
+      "grad_norm": 1.392417311668396,
+      "learning_rate": 2.8443443443443446e-05,
+      "loss": 1.5793,
+      "step": 4319
+    },
+    {
+      "epoch": 0.005647036678287536,
+      "grad_norm": 1.4685407876968384,
+      "learning_rate": 2.8438438438438442e-05,
+      "loss": 1.6679,
+      "step": 4320
+    },
+    {
+      "epoch": 0.005648343862703806,
+      "grad_norm": 1.6190531253814697,
+      "learning_rate": 2.843343343343344e-05,
+      "loss": 1.1182,
+      "step": 4321
+    },
+    {
+      "epoch": 0.005649651047120076,
+      "grad_norm": 1.1069111824035645,
+      "learning_rate": 2.8428428428428428e-05,
+      "loss": 1.5961,
+      "step": 4322
+    },
+    {
+      "epoch": 0.005650958231536347,
+      "grad_norm": 1.4361752271652222,
+      "learning_rate": 2.8423423423423425e-05,
+      "loss": 1.9221,
+      "step": 4323
+    },
+    {
+      "epoch": 0.005652265415952617,
+      "grad_norm": 1.2325265407562256,
+      "learning_rate": 2.8418418418418418e-05,
+      "loss": 1.1489,
+      "step": 4324
+    },
+    {
+      "epoch": 0.005653572600368887,
+      "grad_norm": 1.468827247619629,
+      "learning_rate": 2.8413413413413414e-05,
+      "loss": 1.8344,
+      "step": 4325
+    },
+    {
+      "epoch": 0.005654879784785158,
+      "grad_norm": 0.9098957777023315,
+      "learning_rate": 2.840840840840841e-05,
+      "loss": 0.6928,
+      "step": 4326
+    },
+    {
+      "epoch": 0.005656186969201428,
+      "grad_norm": 1.965370774269104,
+      "learning_rate": 2.8403403403403407e-05,
+      "loss": 2.1212,
+      "step": 4327
+    },
+    {
+      "epoch": 0.005657494153617698,
+      "grad_norm": 1.2574388980865479,
+      "learning_rate": 2.8398398398398403e-05,
+      "loss": 1.8527,
+      "step": 4328
+    },
+    {
+      "epoch": 0.005658801338033969,
+      "grad_norm": 1.4474751949310303,
+      "learning_rate": 2.8393393393393396e-05,
+      "loss": 1.8719,
+      "step": 4329
+    },
+    {
+      "epoch": 0.005660108522450239,
+      "grad_norm": 1.3107092380523682,
+      "learning_rate": 2.8388388388388386e-05,
+      "loss": 1.2382,
+      "step": 4330
+    },
+    {
+      "epoch": 0.005661415706866509,
+      "grad_norm": 0.8240578770637512,
+      "learning_rate": 2.8383383383383382e-05,
+      "loss": 0.5257,
+      "step": 4331
+    },
+    {
+      "epoch": 0.00566272289128278,
+      "grad_norm": 1.7899569272994995,
+      "learning_rate": 2.8378378378378378e-05,
+      "loss": 1.4259,
+      "step": 4332
+    },
+    {
+      "epoch": 0.00566403007569905,
+      "grad_norm": 1.1241390705108643,
+      "learning_rate": 2.8373373373373375e-05,
+      "loss": 1.0369,
+      "step": 4333
+    },
+    {
+      "epoch": 0.00566533726011532,
+      "grad_norm": 0.9048229455947876,
+      "learning_rate": 2.836836836836837e-05,
+      "loss": 0.6278,
+      "step": 4334
+    },
+    {
+      "epoch": 0.00566664444453159,
+      "grad_norm": 1.5691156387329102,
+      "learning_rate": 2.8363363363363364e-05,
+      "loss": 2.1564,
+      "step": 4335
+    },
+    {
+      "epoch": 0.00566795162894786,
+      "grad_norm": 1.4230084419250488,
+      "learning_rate": 2.835835835835836e-05,
+      "loss": 1.561,
+      "step": 4336
+    },
+    {
+      "epoch": 0.00566925881336413,
+      "grad_norm": 1.230269193649292,
+      "learning_rate": 2.8353353353353357e-05,
+      "loss": 1.5952,
+      "step": 4337
+    },
+    {
+      "epoch": 0.005670565997780401,
+      "grad_norm": 1.7398109436035156,
+      "learning_rate": 2.8348348348348346e-05,
+      "loss": 2.6506,
+      "step": 4338
+    },
+    {
+      "epoch": 0.005671873182196671,
+      "grad_norm": 1.3690277338027954,
+      "learning_rate": 2.8343343343343343e-05,
+      "loss": 1.7461,
+      "step": 4339
+    },
+    {
+      "epoch": 0.005673180366612941,
+      "grad_norm": 1.5525764226913452,
+      "learning_rate": 2.833833833833834e-05,
+      "loss": 1.3336,
+      "step": 4340
+    },
+    {
+      "epoch": 0.005674487551029212,
+      "grad_norm": 1.516066074371338,
+      "learning_rate": 2.8333333333333335e-05,
+      "loss": 2.1793,
+      "step": 4341
+    },
+    {
+      "epoch": 0.005675794735445482,
+      "grad_norm": 0.9120578765869141,
+      "learning_rate": 2.832832832832833e-05,
+      "loss": 0.6903,
+      "step": 4342
+    },
+    {
+      "epoch": 0.005677101919861752,
+      "grad_norm": 1.4972691535949707,
+      "learning_rate": 2.8323323323323325e-05,
+      "loss": 1.9634,
+      "step": 4343
+    },
+    {
+      "epoch": 0.0056784091042780226,
+      "grad_norm": 1.2909897565841675,
+      "learning_rate": 2.831831831831832e-05,
+      "loss": 1.168,
+      "step": 4344
+    },
+    {
+      "epoch": 0.005679716288694293,
+      "grad_norm": 1.3689055442810059,
+      "learning_rate": 2.8313313313313317e-05,
+      "loss": 1.7877,
+      "step": 4345
+    },
+    {
+      "epoch": 0.005681023473110563,
+      "grad_norm": 1.524366855621338,
+      "learning_rate": 2.8308308308308314e-05,
+      "loss": 1.8707,
+      "step": 4346
+    },
+    {
+      "epoch": 0.0056823306575268335,
+      "grad_norm": 1.2219198942184448,
+      "learning_rate": 2.8303303303303303e-05,
+      "loss": 1.7695,
+      "step": 4347
+    },
+    {
+      "epoch": 0.005683637841943104,
+      "grad_norm": 1.3683356046676636,
+      "learning_rate": 2.8298298298298296e-05,
+      "loss": 1.5319,
+      "step": 4348
+    },
+    {
+      "epoch": 0.005684945026359374,
+      "grad_norm": 1.316664218902588,
+      "learning_rate": 2.8293293293293293e-05,
+      "loss": 0.8933,
+      "step": 4349
+    },
+    {
+      "epoch": 0.005686252210775644,
+      "grad_norm": 1.30536949634552,
+      "learning_rate": 2.828828828828829e-05,
+      "loss": 2.406,
+      "step": 4350
+    },
+    {
+      "epoch": 0.005687559395191914,
+      "grad_norm": 1.403899908065796,
+      "learning_rate": 2.8283283283283285e-05,
+      "loss": 1.5705,
+      "step": 4351
+    },
+    {
+      "epoch": 0.005688866579608184,
+      "grad_norm": 1.6594147682189941,
+      "learning_rate": 2.8278278278278282e-05,
+      "loss": 1.6269,
+      "step": 4352
+    },
+    {
+      "epoch": 0.005690173764024455,
+      "grad_norm": 1.6887184381484985,
+      "learning_rate": 2.8273273273273275e-05,
+      "loss": 1.9601,
+      "step": 4353
+    },
+    {
+      "epoch": 0.005691480948440725,
+      "grad_norm": 1.6300891637802124,
+      "learning_rate": 2.826826826826827e-05,
+      "loss": 1.0804,
+      "step": 4354
+    },
+    {
+      "epoch": 0.005692788132856995,
+      "grad_norm": 1.4092605113983154,
+      "learning_rate": 2.826326326326326e-05,
+      "loss": 1.5573,
+      "step": 4355
+    },
+    {
+      "epoch": 0.0056940953172732655,
+      "grad_norm": 1.1786856651306152,
+      "learning_rate": 2.8258258258258257e-05,
+      "loss": 1.6038,
+      "step": 4356
+    },
+    {
+      "epoch": 0.005695402501689536,
+      "grad_norm": 1.8649697303771973,
+      "learning_rate": 2.8253253253253253e-05,
+      "loss": 2.0618,
+      "step": 4357
+    },
+    {
+      "epoch": 0.005696709686105806,
+      "grad_norm": 1.3943796157836914,
+      "learning_rate": 2.824824824824825e-05,
+      "loss": 1.5221,
+      "step": 4358
+    },
+    {
+      "epoch": 0.0056980168705220765,
+      "grad_norm": 1.2776241302490234,
+      "learning_rate": 2.8243243243243246e-05,
+      "loss": 1.3954,
+      "step": 4359
+    },
+    {
+      "epoch": 0.005699324054938347,
+      "grad_norm": 1.4063420295715332,
+      "learning_rate": 2.823823823823824e-05,
+      "loss": 1.6836,
+      "step": 4360
+    },
+    {
+      "epoch": 0.005700631239354617,
+      "grad_norm": 0.9964683651924133,
+      "learning_rate": 2.8233233233233236e-05,
+      "loss": 0.8423,
+      "step": 4361
+    },
+    {
+      "epoch": 0.0057019384237708874,
+      "grad_norm": 1.0518639087677002,
+      "learning_rate": 2.8228228228228232e-05,
+      "loss": 1.7396,
+      "step": 4362
+    },
+    {
+      "epoch": 0.005703245608187158,
+      "grad_norm": 1.2594558000564575,
+      "learning_rate": 2.822322322322322e-05,
+      "loss": 1.2041,
+      "step": 4363
+    },
+    {
+      "epoch": 0.005704552792603428,
+      "grad_norm": 1.2609777450561523,
+      "learning_rate": 2.8218218218218218e-05,
+      "loss": 1.1199,
+      "step": 4364
+    },
+    {
+      "epoch": 0.0057058599770196975,
+      "grad_norm": 1.4198492765426636,
+      "learning_rate": 2.8213213213213214e-05,
+      "loss": 1.3797,
+      "step": 4365
+    },
+    {
+      "epoch": 0.005707167161435968,
+      "grad_norm": 1.3430931568145752,
+      "learning_rate": 2.820820820820821e-05,
+      "loss": 2.4575,
+      "step": 4366
+    },
+    {
+      "epoch": 0.005708474345852238,
+      "grad_norm": 1.625356912612915,
+      "learning_rate": 2.8203203203203204e-05,
+      "loss": 1.9986,
+      "step": 4367
+    },
+    {
+      "epoch": 0.0057097815302685085,
+      "grad_norm": 1.8073177337646484,
+      "learning_rate": 2.81981981981982e-05,
+      "loss": 1.8805,
+      "step": 4368
+    },
+    {
+      "epoch": 0.005711088714684779,
+      "grad_norm": 1.7810263633728027,
+      "learning_rate": 2.8193193193193196e-05,
+      "loss": 2.1128,
+      "step": 4369
+    },
+    {
+      "epoch": 0.005712395899101049,
+      "grad_norm": 1.6765902042388916,
+      "learning_rate": 2.8188188188188193e-05,
+      "loss": 2.1138,
+      "step": 4370
+    },
+    {
+      "epoch": 0.0057137030835173195,
+      "grad_norm": 3.0937254428863525,
+      "learning_rate": 2.8183183183183186e-05,
+      "loss": 1.2992,
+      "step": 4371
+    },
+    {
+      "epoch": 0.00571501026793359,
+      "grad_norm": 1.497421145439148,
+      "learning_rate": 2.817817817817818e-05,
+      "loss": 1.939,
+      "step": 4372
+    },
+    {
+      "epoch": 0.00571631745234986,
+      "grad_norm": 1.119702696800232,
+      "learning_rate": 2.817317317317317e-05,
+      "loss": 1.503,
+      "step": 4373
+    },
+    {
+      "epoch": 0.00571762463676613,
+      "grad_norm": 1.2608596086502075,
+      "learning_rate": 2.8168168168168168e-05,
+      "loss": 1.6345,
+      "step": 4374
+    },
+    {
+      "epoch": 0.005718931821182401,
+      "grad_norm": 1.039914846420288,
+      "learning_rate": 2.8163163163163164e-05,
+      "loss": 0.6204,
+      "step": 4375
+    },
+    {
+      "epoch": 0.005720239005598671,
+      "grad_norm": 1.533408761024475,
+      "learning_rate": 2.815815815815816e-05,
+      "loss": 2.1288,
+      "step": 4376
+    },
+    {
+      "epoch": 0.005721546190014941,
+      "grad_norm": 1.259765863418579,
+      "learning_rate": 2.8153153153153157e-05,
+      "loss": 1.1722,
+      "step": 4377
+    },
+    {
+      "epoch": 0.005722853374431212,
+      "grad_norm": 1.0609039068222046,
+      "learning_rate": 2.814814814814815e-05,
+      "loss": 0.8753,
+      "step": 4378
+    },
+    {
+      "epoch": 0.005724160558847482,
+      "grad_norm": 1.2843860387802124,
+      "learning_rate": 2.8143143143143146e-05,
+      "loss": 0.8733,
+      "step": 4379
+    },
+    {
+      "epoch": 0.005725467743263752,
+      "grad_norm": 1.213637113571167,
+      "learning_rate": 2.8138138138138136e-05,
+      "loss": 1.2974,
+      "step": 4380
+    },
+    {
+      "epoch": 0.005726774927680022,
+      "grad_norm": 1.5685056447982788,
+      "learning_rate": 2.8133133133133132e-05,
+      "loss": 1.8454,
+      "step": 4381
+    },
+    {
+      "epoch": 0.005728082112096292,
+      "grad_norm": 2.216892719268799,
+      "learning_rate": 2.812812812812813e-05,
+      "loss": 2.2829,
+      "step": 4382
+    },
+    {
+      "epoch": 0.005729389296512562,
+      "grad_norm": 1.7946327924728394,
+      "learning_rate": 2.8123123123123125e-05,
+      "loss": 1.8978,
+      "step": 4383
+    },
+    {
+      "epoch": 0.005730696480928833,
+      "grad_norm": 1.4264217615127563,
+      "learning_rate": 2.811811811811812e-05,
+      "loss": 1.702,
+      "step": 4384
+    },
+    {
+      "epoch": 0.005732003665345103,
+      "grad_norm": 1.2557882070541382,
+      "learning_rate": 2.8113113113113114e-05,
+      "loss": 1.5248,
+      "step": 4385
+    },
+    {
+      "epoch": 0.005733310849761373,
+      "grad_norm": 1.3069298267364502,
+      "learning_rate": 2.810810810810811e-05,
+      "loss": 0.8517,
+      "step": 4386
+    },
+    {
+      "epoch": 0.005734618034177644,
+      "grad_norm": 1.961179494857788,
+      "learning_rate": 2.8103103103103107e-05,
+      "loss": 1.8361,
+      "step": 4387
+    },
+    {
+      "epoch": 0.005735925218593914,
+      "grad_norm": 1.753015398979187,
+      "learning_rate": 2.8098098098098097e-05,
+      "loss": 1.8327,
+      "step": 4388
+    },
+    {
+      "epoch": 0.005737232403010184,
+      "grad_norm": 1.191309928894043,
+      "learning_rate": 2.8093093093093093e-05,
+      "loss": 1.0144,
+      "step": 4389
+    },
+    {
+      "epoch": 0.005738539587426455,
+      "grad_norm": 1.3492724895477295,
+      "learning_rate": 2.808808808808809e-05,
+      "loss": 1.7254,
+      "step": 4390
+    },
+    {
+      "epoch": 0.005739846771842725,
+      "grad_norm": 0.9211691617965698,
+      "learning_rate": 2.8083083083083082e-05,
+      "loss": 1.6037,
+      "step": 4391
+    },
+    {
+      "epoch": 0.005741153956258995,
+      "grad_norm": 1.2470166683197021,
+      "learning_rate": 2.807807807807808e-05,
+      "loss": 1.4074,
+      "step": 4392
+    },
+    {
+      "epoch": 0.005742461140675266,
+      "grad_norm": 1.496669054031372,
+      "learning_rate": 2.8073073073073075e-05,
+      "loss": 1.5938,
+      "step": 4393
+    },
+    {
+      "epoch": 0.005743768325091536,
+      "grad_norm": 1.090317726135254,
+      "learning_rate": 2.806806806806807e-05,
+      "loss": 0.9844,
+      "step": 4394
+    },
+    {
+      "epoch": 0.005745075509507806,
+      "grad_norm": 0.9228986501693726,
+      "learning_rate": 2.8063063063063068e-05,
+      "loss": 0.4387,
+      "step": 4395
+    },
+    {
+      "epoch": 0.005746382693924076,
+      "grad_norm": 1.1389049291610718,
+      "learning_rate": 2.805805805805806e-05,
+      "loss": 1.1934,
+      "step": 4396
+    },
+    {
+      "epoch": 0.005747689878340346,
+      "grad_norm": 1.2950663566589355,
+      "learning_rate": 2.8053053053053054e-05,
+      "loss": 1.3754,
+      "step": 4397
+    },
+    {
+      "epoch": 0.005748997062756616,
+      "grad_norm": 1.5343849658966064,
+      "learning_rate": 2.8048048048048047e-05,
+      "loss": 1.658,
+      "step": 4398
+    },
+    {
+      "epoch": 0.005750304247172887,
+      "grad_norm": 2.4840612411499023,
+      "learning_rate": 2.8043043043043043e-05,
+      "loss": 1.6069,
+      "step": 4399
+    },
+    {
+      "epoch": 0.005751611431589157,
+      "grad_norm": 2.0089261531829834,
+      "learning_rate": 2.803803803803804e-05,
+      "loss": 1.913,
+      "step": 4400
+    },
+    {
+      "epoch": 0.005752918616005427,
+      "grad_norm": 1.366276502609253,
+      "learning_rate": 2.8033033033033036e-05,
+      "loss": 1.6578,
+      "step": 4401
+    },
+    {
+      "epoch": 0.005754225800421698,
+      "grad_norm": 1.2364778518676758,
+      "learning_rate": 2.8028028028028032e-05,
+      "loss": 1.4142,
+      "step": 4402
+    },
+    {
+      "epoch": 0.005755532984837968,
+      "grad_norm": 2.2674903869628906,
+      "learning_rate": 2.8023023023023025e-05,
+      "loss": 0.8659,
+      "step": 4403
+    },
+    {
+      "epoch": 0.005756840169254238,
+      "grad_norm": 1.2411550283432007,
+      "learning_rate": 2.801801801801802e-05,
+      "loss": 1.6902,
+      "step": 4404
+    },
+    {
+      "epoch": 0.005758147353670509,
+      "grad_norm": 1.667500615119934,
+      "learning_rate": 2.801301301301301e-05,
+      "loss": 1.8532,
+      "step": 4405
+    },
+    {
+      "epoch": 0.005759454538086779,
+      "grad_norm": 1.3346153497695923,
+      "learning_rate": 2.8008008008008007e-05,
+      "loss": 1.8025,
+      "step": 4406
+    },
+    {
+      "epoch": 0.005760761722503049,
+      "grad_norm": 1.5389180183410645,
+      "learning_rate": 2.8003003003003004e-05,
+      "loss": 1.8479,
+      "step": 4407
+    },
+    {
+      "epoch": 0.0057620689069193196,
+      "grad_norm": 1.5254756212234497,
+      "learning_rate": 2.7997997997998e-05,
+      "loss": 0.8879,
+      "step": 4408
+    },
+    {
+      "epoch": 0.00576337609133559,
+      "grad_norm": 1.568851113319397,
+      "learning_rate": 2.7992992992992993e-05,
+      "loss": 2.001,
+      "step": 4409
+    },
+    {
+      "epoch": 0.00576468327575186,
+      "grad_norm": 1.691598892211914,
+      "learning_rate": 2.798798798798799e-05,
+      "loss": 2.1614,
+      "step": 4410
+    },
+    {
+      "epoch": 0.00576599046016813,
+      "grad_norm": 1.1894441843032837,
+      "learning_rate": 2.7982982982982986e-05,
+      "loss": 1.1592,
+      "step": 4411
+    },
+    {
+      "epoch": 0.0057672976445844,
+      "grad_norm": 1.465369701385498,
+      "learning_rate": 2.7977977977977982e-05,
+      "loss": 1.8953,
+      "step": 4412
+    },
+    {
+      "epoch": 0.00576860482900067,
+      "grad_norm": 1.7106328010559082,
+      "learning_rate": 2.7972972972972972e-05,
+      "loss": 2.0521,
+      "step": 4413
+    },
+    {
+      "epoch": 0.005769912013416941,
+      "grad_norm": 1.6243577003479004,
+      "learning_rate": 2.7967967967967968e-05,
+      "loss": 2.023,
+      "step": 4414
+    },
+    {
+      "epoch": 0.005771219197833211,
+      "grad_norm": 1.486568808555603,
+      "learning_rate": 2.7962962962962965e-05,
+      "loss": 1.7869,
+      "step": 4415
+    },
+    {
+      "epoch": 0.005772526382249481,
+      "grad_norm": 1.2151774168014526,
+      "learning_rate": 2.7957957957957958e-05,
+      "loss": 1.1366,
+      "step": 4416
+    },
+    {
+      "epoch": 0.0057738335666657516,
+      "grad_norm": 1.6414657831192017,
+      "learning_rate": 2.7952952952952954e-05,
+      "loss": 1.623,
+      "step": 4417
+    },
+    {
+      "epoch": 0.005775140751082022,
+      "grad_norm": 1.2551473379135132,
+      "learning_rate": 2.794794794794795e-05,
+      "loss": 1.9135,
+      "step": 4418
+    },
+    {
+      "epoch": 0.005776447935498292,
+      "grad_norm": 1.2534723281860352,
+      "learning_rate": 2.7942942942942947e-05,
+      "loss": 1.1503,
+      "step": 4419
+    },
+    {
+      "epoch": 0.0057777551199145625,
+      "grad_norm": 1.264475703239441,
+      "learning_rate": 2.7937937937937943e-05,
+      "loss": 1.489,
+      "step": 4420
+    },
+    {
+      "epoch": 0.005779062304330833,
+      "grad_norm": 1.4543061256408691,
+      "learning_rate": 2.7932932932932936e-05,
+      "loss": 1.2806,
+      "step": 4421
+    },
+    {
+      "epoch": 0.005780369488747103,
+      "grad_norm": 1.9797331094741821,
+      "learning_rate": 2.7927927927927926e-05,
+      "loss": 2.2045,
+      "step": 4422
+    },
+    {
+      "epoch": 0.0057816766731633735,
+      "grad_norm": 1.5653047561645508,
+      "learning_rate": 2.7922922922922922e-05,
+      "loss": 2.22,
+      "step": 4423
+    },
+    {
+      "epoch": 0.005782983857579644,
+      "grad_norm": 1.6056344509124756,
+      "learning_rate": 2.7917917917917918e-05,
+      "loss": 1.6999,
+      "step": 4424
+    },
+    {
+      "epoch": 0.005784291041995914,
+      "grad_norm": 1.3474721908569336,
+      "learning_rate": 2.7912912912912915e-05,
+      "loss": 1.4429,
+      "step": 4425
+    },
+    {
+      "epoch": 0.005785598226412184,
+      "grad_norm": 1.5668607950210571,
+      "learning_rate": 2.790790790790791e-05,
+      "loss": 2.2455,
+      "step": 4426
+    },
+    {
+      "epoch": 0.005786905410828454,
+      "grad_norm": 1.3745261430740356,
+      "learning_rate": 2.7902902902902904e-05,
+      "loss": 1.7882,
+      "step": 4427
+    },
+    {
+      "epoch": 0.005788212595244724,
+      "grad_norm": 1.276803731918335,
+      "learning_rate": 2.78978978978979e-05,
+      "loss": 1.1888,
+      "step": 4428
+    },
+    {
+      "epoch": 0.0057895197796609945,
+      "grad_norm": 1.0936208963394165,
+      "learning_rate": 2.7892892892892897e-05,
+      "loss": 1.1354,
+      "step": 4429
+    },
+    {
+      "epoch": 0.005790826964077265,
+      "grad_norm": 2.078791379928589,
+      "learning_rate": 2.7887887887887886e-05,
+      "loss": 0.6965,
+      "step": 4430
+    },
+    {
+      "epoch": 0.005792134148493535,
+      "grad_norm": 1.3754019737243652,
+      "learning_rate": 2.7882882882882883e-05,
+      "loss": 1.8339,
+      "step": 4431
+    },
+    {
+      "epoch": 0.0057934413329098055,
+      "grad_norm": 1.533116340637207,
+      "learning_rate": 2.787787787787788e-05,
+      "loss": 1.7588,
+      "step": 4432
+    },
+    {
+      "epoch": 0.005794748517326076,
+      "grad_norm": 1.350574016571045,
+      "learning_rate": 2.7872872872872875e-05,
+      "loss": 1.6867,
+      "step": 4433
+    },
+    {
+      "epoch": 0.005796055701742346,
+      "grad_norm": 1.7762374877929688,
+      "learning_rate": 2.786786786786787e-05,
+      "loss": 1.9082,
+      "step": 4434
+    },
+    {
+      "epoch": 0.0057973628861586165,
+      "grad_norm": 1.438230276107788,
+      "learning_rate": 2.7862862862862865e-05,
+      "loss": 0.9092,
+      "step": 4435
+    },
+    {
+      "epoch": 0.005798670070574887,
+      "grad_norm": 1.3388832807540894,
+      "learning_rate": 2.785785785785786e-05,
+      "loss": 1.4801,
+      "step": 4436
+    },
+    {
+      "epoch": 0.005799977254991157,
+      "grad_norm": 1.4860488176345825,
+      "learning_rate": 2.7852852852852857e-05,
+      "loss": 1.1813,
+      "step": 4437
+    },
+    {
+      "epoch": 0.005801284439407427,
+      "grad_norm": 1.6598443984985352,
+      "learning_rate": 2.7847847847847847e-05,
+      "loss": 1.9864,
+      "step": 4438
+    },
+    {
+      "epoch": 0.005802591623823698,
+      "grad_norm": 1.204086184501648,
+      "learning_rate": 2.7842842842842843e-05,
+      "loss": 0.8697,
+      "step": 4439
+    },
+    {
+      "epoch": 0.005803898808239968,
+      "grad_norm": 1.6929036378860474,
+      "learning_rate": 2.7837837837837836e-05,
+      "loss": 1.8201,
+      "step": 4440
+    },
+    {
+      "epoch": 0.005805205992656238,
+      "grad_norm": 1.5603952407836914,
+      "learning_rate": 2.7832832832832833e-05,
+      "loss": 2.4203,
+      "step": 4441
+    },
+    {
+      "epoch": 0.005806513177072508,
+      "grad_norm": 1.3910847902297974,
+      "learning_rate": 2.782782782782783e-05,
+      "loss": 2.3053,
+      "step": 4442
+    },
+    {
+      "epoch": 0.005807820361488778,
+      "grad_norm": 1.4856058359146118,
+      "learning_rate": 2.7822822822822825e-05,
+      "loss": 1.1012,
+      "step": 4443
+    },
+    {
+      "epoch": 0.0058091275459050485,
+      "grad_norm": 1.56679368019104,
+      "learning_rate": 2.7817817817817822e-05,
+      "loss": 1.7009,
+      "step": 4444
+    },
+    {
+      "epoch": 0.005810434730321319,
+      "grad_norm": 1.274522066116333,
+      "learning_rate": 2.7812812812812815e-05,
+      "loss": 1.601,
+      "step": 4445
+    },
+    {
+      "epoch": 0.005811741914737589,
+      "grad_norm": 1.2207599878311157,
+      "learning_rate": 2.7807807807807808e-05,
+      "loss": 1.7167,
+      "step": 4446
+    },
+    {
+      "epoch": 0.005813049099153859,
+      "grad_norm": 1.5133830308914185,
+      "learning_rate": 2.78028028028028e-05,
+      "loss": 2.0625,
+      "step": 4447
+    },
+    {
+      "epoch": 0.00581435628357013,
+      "grad_norm": 1.3491390943527222,
+      "learning_rate": 2.7797797797797797e-05,
+      "loss": 1.6825,
+      "step": 4448
+    },
+    {
+      "epoch": 0.0058156634679864,
+      "grad_norm": 1.54470956325531,
+      "learning_rate": 2.7792792792792793e-05,
+      "loss": 1.3598,
+      "step": 4449
+    },
+    {
+      "epoch": 0.00581697065240267,
+      "grad_norm": 1.432422399520874,
+      "learning_rate": 2.778778778778779e-05,
+      "loss": 2.3029,
+      "step": 4450
+    },
+    {
+      "epoch": 0.005818277836818941,
+      "grad_norm": 1.44601571559906,
+      "learning_rate": 2.7782782782782786e-05,
+      "loss": 1.6436,
+      "step": 4451
+    },
+    {
+      "epoch": 0.005819585021235211,
+      "grad_norm": 1.3847241401672363,
+      "learning_rate": 2.777777777777778e-05,
+      "loss": 1.8569,
+      "step": 4452
+    },
+    {
+      "epoch": 0.005820892205651481,
+      "grad_norm": 2.3063082695007324,
+      "learning_rate": 2.7772772772772775e-05,
+      "loss": 2.1198,
+      "step": 4453
+    },
+    {
+      "epoch": 0.005822199390067752,
+      "grad_norm": 1.100419521331787,
+      "learning_rate": 2.7767767767767772e-05,
+      "loss": 1.0104,
+      "step": 4454
+    },
+    {
+      "epoch": 0.005823506574484022,
+      "grad_norm": 1.781338095664978,
+      "learning_rate": 2.776276276276276e-05,
+      "loss": 1.8361,
+      "step": 4455
+    },
+    {
+      "epoch": 0.005824813758900292,
+      "grad_norm": 1.6767401695251465,
+      "learning_rate": 2.7757757757757758e-05,
+      "loss": 2.2702,
+      "step": 4456
+    },
+    {
+      "epoch": 0.005826120943316562,
+      "grad_norm": 1.363655924797058,
+      "learning_rate": 2.7752752752752754e-05,
+      "loss": 2.0243,
+      "step": 4457
+    },
+    {
+      "epoch": 0.005827428127732832,
+      "grad_norm": 1.2314109802246094,
+      "learning_rate": 2.7747747747747747e-05,
+      "loss": 1.621,
+      "step": 4458
+    },
+    {
+      "epoch": 0.005828735312149102,
+      "grad_norm": 1.2720845937728882,
+      "learning_rate": 2.7742742742742743e-05,
+      "loss": 1.4834,
+      "step": 4459
+    },
+    {
+      "epoch": 0.005830042496565373,
+      "grad_norm": 1.8392155170440674,
+      "learning_rate": 2.773773773773774e-05,
+      "loss": 1.8025,
+      "step": 4460
+    },
+    {
+      "epoch": 0.005831349680981643,
+      "grad_norm": 1.2127138376235962,
+      "learning_rate": 2.7732732732732736e-05,
+      "loss": 0.9916,
+      "step": 4461
+    },
+    {
+      "epoch": 0.005832656865397913,
+      "grad_norm": 1.182537317276001,
+      "learning_rate": 2.7727727727727733e-05,
+      "loss": 0.6776,
+      "step": 4462
+    },
+    {
+      "epoch": 0.005833964049814184,
+      "grad_norm": 1.3282581567764282,
+      "learning_rate": 2.7722722722722722e-05,
+      "loss": 1.4722,
+      "step": 4463
+    },
+    {
+      "epoch": 0.005835271234230454,
+      "grad_norm": 1.129106044769287,
+      "learning_rate": 2.771771771771772e-05,
+      "loss": 0.701,
+      "step": 4464
+    },
+    {
+      "epoch": 0.005836578418646724,
+      "grad_norm": 1.3795222043991089,
+      "learning_rate": 2.771271271271271e-05,
+      "loss": 1.1251,
+      "step": 4465
+    },
+    {
+      "epoch": 0.005837885603062995,
+      "grad_norm": 1.7393888235092163,
+      "learning_rate": 2.7707707707707708e-05,
+      "loss": 2.0822,
+      "step": 4466
+    },
+    {
+      "epoch": 0.005839192787479265,
+      "grad_norm": 1.11077880859375,
+      "learning_rate": 2.7702702702702704e-05,
+      "loss": 1.2817,
+      "step": 4467
+    },
+    {
+      "epoch": 0.005840499971895535,
+      "grad_norm": 1.0539791584014893,
+      "learning_rate": 2.76976976976977e-05,
+      "loss": 1.0883,
+      "step": 4468
+    },
+    {
+      "epoch": 0.005841807156311806,
+      "grad_norm": 1.2682170867919922,
+      "learning_rate": 2.7692692692692697e-05,
+      "loss": 1.0641,
+      "step": 4469
+    },
+    {
+      "epoch": 0.005843114340728076,
+      "grad_norm": 1.7622284889221191,
+      "learning_rate": 2.768768768768769e-05,
+      "loss": 1.5599,
+      "step": 4470
+    },
+    {
+      "epoch": 0.005844421525144346,
+      "grad_norm": 1.6490631103515625,
+      "learning_rate": 2.7682682682682683e-05,
+      "loss": 1.1861,
+      "step": 4471
+    },
+    {
+      "epoch": 0.005845728709560616,
+      "grad_norm": 1.6690438985824585,
+      "learning_rate": 2.7677677677677676e-05,
+      "loss": 2.0037,
+      "step": 4472
+    },
+    {
+      "epoch": 0.005847035893976886,
+      "grad_norm": 1.7246538400650024,
+      "learning_rate": 2.7672672672672672e-05,
+      "loss": 2.2619,
+      "step": 4473
+    },
+    {
+      "epoch": 0.005848343078393156,
+      "grad_norm": 1.3424654006958008,
+      "learning_rate": 2.766766766766767e-05,
+      "loss": 1.17,
+      "step": 4474
+    },
+    {
+      "epoch": 0.005849650262809427,
+      "grad_norm": 1.2491614818572998,
+      "learning_rate": 2.7662662662662665e-05,
+      "loss": 0.934,
+      "step": 4475
+    },
+    {
+      "epoch": 0.005850957447225697,
+      "grad_norm": 1.7632230520248413,
+      "learning_rate": 2.7657657657657658e-05,
+      "loss": 1.8941,
+      "step": 4476
+    },
+    {
+      "epoch": 0.005852264631641967,
+      "grad_norm": 1.9646333456039429,
+      "learning_rate": 2.7652652652652654e-05,
+      "loss": 3.0237,
+      "step": 4477
+    },
+    {
+      "epoch": 0.005853571816058238,
+      "grad_norm": 1.4480732679367065,
+      "learning_rate": 2.764764764764765e-05,
+      "loss": 1.9373,
+      "step": 4478
+    },
+    {
+      "epoch": 0.005854879000474508,
+      "grad_norm": 1.6283597946166992,
+      "learning_rate": 2.7642642642642647e-05,
+      "loss": 2.3415,
+      "step": 4479
+    },
+    {
+      "epoch": 0.005856186184890778,
+      "grad_norm": 1.4436355829238892,
+      "learning_rate": 2.7637637637637637e-05,
+      "loss": 1.6378,
+      "step": 4480
+    },
+    {
+      "epoch": 0.0058574933693070486,
+      "grad_norm": 2.1179215908050537,
+      "learning_rate": 2.7632632632632633e-05,
+      "loss": 2.1936,
+      "step": 4481
+    },
+    {
+      "epoch": 0.005858800553723319,
+      "grad_norm": 0.7755745053291321,
+      "learning_rate": 2.762762762762763e-05,
+      "loss": 0.4512,
+      "step": 4482
+    },
+    {
+      "epoch": 0.005860107738139589,
+      "grad_norm": 1.0014432668685913,
+      "learning_rate": 2.7622622622622622e-05,
+      "loss": 0.8922,
+      "step": 4483
+    },
+    {
+      "epoch": 0.0058614149225558595,
+      "grad_norm": 1.4644931554794312,
+      "learning_rate": 2.761761761761762e-05,
+      "loss": 1.912,
+      "step": 4484
+    },
+    {
+      "epoch": 0.00586272210697213,
+      "grad_norm": 0.9821320176124573,
+      "learning_rate": 2.7612612612612615e-05,
+      "loss": 0.2876,
+      "step": 4485
+    },
+    {
+      "epoch": 0.0058640292913884,
+      "grad_norm": 1.4810662269592285,
+      "learning_rate": 2.760760760760761e-05,
+      "loss": 2.044,
+      "step": 4486
+    },
+    {
+      "epoch": 0.0058653364758046705,
+      "grad_norm": 1.6798914670944214,
+      "learning_rate": 2.7602602602602608e-05,
+      "loss": 2.2338,
+      "step": 4487
+    },
+    {
+      "epoch": 0.00586664366022094,
+      "grad_norm": 1.3654935359954834,
+      "learning_rate": 2.7597597597597597e-05,
+      "loss": 0.8379,
+      "step": 4488
+    },
+    {
+      "epoch": 0.00586795084463721,
+      "grad_norm": 1.7100900411605835,
+      "learning_rate": 2.7592592592592594e-05,
+      "loss": 1.4746,
+      "step": 4489
+    },
+    {
+      "epoch": 0.005869258029053481,
+      "grad_norm": 1.4110320806503296,
+      "learning_rate": 2.7587587587587587e-05,
+      "loss": 1.0247,
+      "step": 4490
+    },
+    {
+      "epoch": 0.005870565213469751,
+      "grad_norm": 1.33967125415802,
+      "learning_rate": 2.7582582582582583e-05,
+      "loss": 1.7065,
+      "step": 4491
+    },
+    {
+      "epoch": 0.005871872397886021,
+      "grad_norm": 0.8779923319816589,
+      "learning_rate": 2.757757757757758e-05,
+      "loss": 0.6428,
+      "step": 4492
+    },
+    {
+      "epoch": 0.0058731795823022915,
+      "grad_norm": 1.3973801136016846,
+      "learning_rate": 2.7572572572572576e-05,
+      "loss": 2.2172,
+      "step": 4493
+    },
+    {
+      "epoch": 0.005874486766718562,
+      "grad_norm": 1.2693867683410645,
+      "learning_rate": 2.7567567567567572e-05,
+      "loss": 1.1047,
+      "step": 4494
+    },
+    {
+      "epoch": 0.005875793951134832,
+      "grad_norm": 1.0805323123931885,
+      "learning_rate": 2.7562562562562565e-05,
+      "loss": 1.0982,
+      "step": 4495
+    },
+    {
+      "epoch": 0.0058771011355511025,
+      "grad_norm": 1.073543667793274,
+      "learning_rate": 2.7557557557557555e-05,
+      "loss": 0.9905,
+      "step": 4496
+    },
+    {
+      "epoch": 0.005878408319967373,
+      "grad_norm": 1.2854206562042236,
+      "learning_rate": 2.755255255255255e-05,
+      "loss": 1.682,
+      "step": 4497
+    },
+    {
+      "epoch": 0.005879715504383643,
+      "grad_norm": 0.41551318764686584,
+      "learning_rate": 2.7547547547547547e-05,
+      "loss": 0.1455,
+      "step": 4498
+    },
+    {
+      "epoch": 0.0058810226887999134,
+      "grad_norm": 1.1960279941558838,
+      "learning_rate": 2.7542542542542544e-05,
+      "loss": 1.3656,
+      "step": 4499
+    },
+    {
+      "epoch": 0.005882329873216184,
+      "grad_norm": 1.5446255207061768,
+      "learning_rate": 2.753753753753754e-05,
+      "loss": 1.8419,
+      "step": 4500
+    },
+    {
+      "epoch": 0.005883637057632454,
+      "grad_norm": 1.7863303422927856,
+      "learning_rate": 2.7532532532532533e-05,
+      "loss": 1.5654,
+      "step": 4501
+    },
+    {
+      "epoch": 0.005884944242048724,
+      "grad_norm": 0.920793354511261,
+      "learning_rate": 2.752752752752753e-05,
+      "loss": 0.9692,
+      "step": 4502
+    },
+    {
+      "epoch": 0.005886251426464994,
+      "grad_norm": 1.3322253227233887,
+      "learning_rate": 2.7522522522522526e-05,
+      "loss": 1.2718,
+      "step": 4503
+    },
+    {
+      "epoch": 0.005887558610881264,
+      "grad_norm": 1.1357336044311523,
+      "learning_rate": 2.7517517517517522e-05,
+      "loss": 0.6377,
+      "step": 4504
+    },
+    {
+      "epoch": 0.0058888657952975345,
+      "grad_norm": 1.69141685962677,
+      "learning_rate": 2.7512512512512512e-05,
+      "loss": 2.459,
+      "step": 4505
+    },
+    {
+      "epoch": 0.005890172979713805,
+      "grad_norm": 1.773729920387268,
+      "learning_rate": 2.7507507507507508e-05,
+      "loss": 2.0479,
+      "step": 4506
+    },
+    {
+      "epoch": 0.005891480164130075,
+      "grad_norm": 1.4994827508926392,
+      "learning_rate": 2.7502502502502504e-05,
+      "loss": 1.3825,
+      "step": 4507
+    },
+    {
+      "epoch": 0.0058927873485463455,
+      "grad_norm": 1.5043987035751343,
+      "learning_rate": 2.7497497497497497e-05,
+      "loss": 2.1334,
+      "step": 4508
+    },
+    {
+      "epoch": 0.005894094532962616,
+      "grad_norm": 1.6048840284347534,
+      "learning_rate": 2.7492492492492494e-05,
+      "loss": 1.9128,
+      "step": 4509
+    },
+    {
+      "epoch": 0.005895401717378886,
+      "grad_norm": 1.3038866519927979,
+      "learning_rate": 2.748748748748749e-05,
+      "loss": 1.9801,
+      "step": 4510
+    },
+    {
+      "epoch": 0.005896708901795156,
+      "grad_norm": 1.335004448890686,
+      "learning_rate": 2.7482482482482487e-05,
+      "loss": 1.9478,
+      "step": 4511
+    },
+    {
+      "epoch": 0.005898016086211427,
+      "grad_norm": 1.8973201513290405,
+      "learning_rate": 2.7477477477477483e-05,
+      "loss": 2.3909,
+      "step": 4512
+    },
+    {
+      "epoch": 0.005899323270627697,
+      "grad_norm": 1.5345458984375,
+      "learning_rate": 2.7472472472472473e-05,
+      "loss": 1.7096,
+      "step": 4513
+    },
+    {
+      "epoch": 0.005900630455043967,
+      "grad_norm": 1.5915439128875732,
+      "learning_rate": 2.7467467467467465e-05,
+      "loss": 1.9676,
+      "step": 4514
+    },
+    {
+      "epoch": 0.005901937639460238,
+      "grad_norm": 1.367655873298645,
+      "learning_rate": 2.7462462462462462e-05,
+      "loss": 1.1287,
+      "step": 4515
+    },
+    {
+      "epoch": 0.005903244823876508,
+      "grad_norm": 1.6291239261627197,
+      "learning_rate": 2.7457457457457458e-05,
+      "loss": 2.0746,
+      "step": 4516
+    },
+    {
+      "epoch": 0.005904552008292778,
+      "grad_norm": 1.0266294479370117,
+      "learning_rate": 2.7452452452452455e-05,
+      "loss": 0.7633,
+      "step": 4517
+    },
+    {
+      "epoch": 0.005905859192709048,
+      "grad_norm": 1.2061597108840942,
+      "learning_rate": 2.744744744744745e-05,
+      "loss": 1.5891,
+      "step": 4518
+    },
+    {
+      "epoch": 0.005907166377125318,
+      "grad_norm": 1.4866387844085693,
+      "learning_rate": 2.7442442442442444e-05,
+      "loss": 2.1518,
+      "step": 4519
+    },
+    {
+      "epoch": 0.005908473561541588,
+      "grad_norm": 1.4174457788467407,
+      "learning_rate": 2.743743743743744e-05,
+      "loss": 1.2991,
+      "step": 4520
+    },
+    {
+      "epoch": 0.005909780745957859,
+      "grad_norm": 1.3831840753555298,
+      "learning_rate": 2.743243243243243e-05,
+      "loss": 1.7736,
+      "step": 4521
+    },
+    {
+      "epoch": 0.005911087930374129,
+      "grad_norm": 1.1448744535446167,
+      "learning_rate": 2.7427427427427426e-05,
+      "loss": 0.6044,
+      "step": 4522
+    },
+    {
+      "epoch": 0.005912395114790399,
+      "grad_norm": 0.48047974705696106,
+      "learning_rate": 2.7422422422422423e-05,
+      "loss": 0.1546,
+      "step": 4523
+    },
+    {
+      "epoch": 0.00591370229920667,
+      "grad_norm": 1.7978545427322388,
+      "learning_rate": 2.741741741741742e-05,
+      "loss": 1.7776,
+      "step": 4524
+    },
+    {
+      "epoch": 0.00591500948362294,
+      "grad_norm": 1.378781795501709,
+      "learning_rate": 2.7412412412412415e-05,
+      "loss": 1.5883,
+      "step": 4525
+    },
+    {
+      "epoch": 0.00591631666803921,
+      "grad_norm": 1.2129422426223755,
+      "learning_rate": 2.7407407407407408e-05,
+      "loss": 1.2862,
+      "step": 4526
+    },
+    {
+      "epoch": 0.005917623852455481,
+      "grad_norm": 1.1361254453659058,
+      "learning_rate": 2.7402402402402405e-05,
+      "loss": 1.6522,
+      "step": 4527
+    },
+    {
+      "epoch": 0.005918931036871751,
+      "grad_norm": 1.295478343963623,
+      "learning_rate": 2.73973973973974e-05,
+      "loss": 1.5767,
+      "step": 4528
+    },
+    {
+      "epoch": 0.005920238221288021,
+      "grad_norm": 1.3233872652053833,
+      "learning_rate": 2.7392392392392397e-05,
+      "loss": 1.4133,
+      "step": 4529
+    },
+    {
+      "epoch": 0.005921545405704292,
+      "grad_norm": 1.4945601224899292,
+      "learning_rate": 2.7387387387387387e-05,
+      "loss": 1.6032,
+      "step": 4530
+    },
+    {
+      "epoch": 0.005922852590120562,
+      "grad_norm": 1.1618452072143555,
+      "learning_rate": 2.7382382382382383e-05,
+      "loss": 0.75,
+      "step": 4531
+    },
+    {
+      "epoch": 0.005924159774536832,
+      "grad_norm": 1.0901761054992676,
+      "learning_rate": 2.7377377377377376e-05,
+      "loss": 0.8995,
+      "step": 4532
+    },
+    {
+      "epoch": 0.005925466958953102,
+      "grad_norm": 2.597090721130371,
+      "learning_rate": 2.7372372372372373e-05,
+      "loss": 1.0034,
+      "step": 4533
+    },
+    {
+      "epoch": 0.005926774143369372,
+      "grad_norm": 1.5070717334747314,
+      "learning_rate": 2.736736736736737e-05,
+      "loss": 1.9582,
+      "step": 4534
+    },
+    {
+      "epoch": 0.005928081327785642,
+      "grad_norm": 1.6143525838851929,
+      "learning_rate": 2.7362362362362365e-05,
+      "loss": 1.5081,
+      "step": 4535
+    },
+    {
+      "epoch": 0.005929388512201913,
+      "grad_norm": 1.516845703125,
+      "learning_rate": 2.7357357357357362e-05,
+      "loss": 1.5981,
+      "step": 4536
+    },
+    {
+      "epoch": 0.005930695696618183,
+      "grad_norm": 1.2037991285324097,
+      "learning_rate": 2.7352352352352355e-05,
+      "loss": 1.0692,
+      "step": 4537
+    },
+    {
+      "epoch": 0.005932002881034453,
+      "grad_norm": 1.3893437385559082,
+      "learning_rate": 2.7347347347347348e-05,
+      "loss": 1.3165,
+      "step": 4538
+    },
+    {
+      "epoch": 0.005933310065450724,
+      "grad_norm": 2.6239469051361084,
+      "learning_rate": 2.734234234234234e-05,
+      "loss": 1.1894,
+      "step": 4539
+    },
+    {
+      "epoch": 0.005934617249866994,
+      "grad_norm": 1.4084033966064453,
+      "learning_rate": 2.7337337337337337e-05,
+      "loss": 1.9334,
+      "step": 4540
+    },
+    {
+      "epoch": 0.005935924434283264,
+      "grad_norm": 1.241999864578247,
+      "learning_rate": 2.7332332332332333e-05,
+      "loss": 1.4015,
+      "step": 4541
+    },
+    {
+      "epoch": 0.005937231618699535,
+      "grad_norm": 1.1996272802352905,
+      "learning_rate": 2.732732732732733e-05,
+      "loss": 1.5568,
+      "step": 4542
+    },
+    {
+      "epoch": 0.005938538803115805,
+      "grad_norm": 1.0937309265136719,
+      "learning_rate": 2.7322322322322326e-05,
+      "loss": 1.0675,
+      "step": 4543
+    },
+    {
+      "epoch": 0.005939845987532075,
+      "grad_norm": 1.1635380983352661,
+      "learning_rate": 2.731731731731732e-05,
+      "loss": 0.7926,
+      "step": 4544
+    },
+    {
+      "epoch": 0.0059411531719483456,
+      "grad_norm": 1.3224263191223145,
+      "learning_rate": 2.7312312312312315e-05,
+      "loss": 1.6908,
+      "step": 4545
+    },
+    {
+      "epoch": 0.005942460356364616,
+      "grad_norm": 1.0710914134979248,
+      "learning_rate": 2.7307307307307305e-05,
+      "loss": 0.8195,
+      "step": 4546
+    },
+    {
+      "epoch": 0.005943767540780886,
+      "grad_norm": 1.6145724058151245,
+      "learning_rate": 2.73023023023023e-05,
+      "loss": 2.5763,
+      "step": 4547
+    },
+    {
+      "epoch": 0.0059450747251971565,
+      "grad_norm": 2.071474552154541,
+      "learning_rate": 2.7297297297297298e-05,
+      "loss": 1.8341,
+      "step": 4548
+    },
+    {
+      "epoch": 0.005946381909613426,
+      "grad_norm": 2.232714891433716,
+      "learning_rate": 2.7292292292292294e-05,
+      "loss": 2.384,
+      "step": 4549
+    },
+    {
+      "epoch": 0.005947689094029696,
+      "grad_norm": 1.0968847274780273,
+      "learning_rate": 2.7287287287287287e-05,
+      "loss": 1.4613,
+      "step": 4550
+    },
+    {
+      "epoch": 0.005948996278445967,
+      "grad_norm": 1.1714986562728882,
+      "learning_rate": 2.7282282282282283e-05,
+      "loss": 1.5086,
+      "step": 4551
+    },
+    {
+      "epoch": 0.005950303462862237,
+      "grad_norm": 1.8250237703323364,
+      "learning_rate": 2.727727727727728e-05,
+      "loss": 1.8768,
+      "step": 4552
+    },
+    {
+      "epoch": 0.005951610647278507,
+      "grad_norm": 1.4997811317443848,
+      "learning_rate": 2.7272272272272276e-05,
+      "loss": 1.4614,
+      "step": 4553
+    },
+    {
+      "epoch": 0.0059529178316947776,
+      "grad_norm": 1.518946647644043,
+      "learning_rate": 2.7267267267267273e-05,
+      "loss": 1.9356,
+      "step": 4554
+    },
+    {
+      "epoch": 0.005954225016111048,
+      "grad_norm": 1.0030150413513184,
+      "learning_rate": 2.7262262262262262e-05,
+      "loss": 0.9248,
+      "step": 4555
+    },
+    {
+      "epoch": 0.005955532200527318,
+      "grad_norm": 1.0351202487945557,
+      "learning_rate": 2.725725725725726e-05,
+      "loss": 0.8596,
+      "step": 4556
+    },
+    {
+      "epoch": 0.0059568393849435885,
+      "grad_norm": 1.254581093788147,
+      "learning_rate": 2.725225225225225e-05,
+      "loss": 1.7417,
+      "step": 4557
+    },
+    {
+      "epoch": 0.005958146569359859,
+      "grad_norm": 1.4694186449050903,
+      "learning_rate": 2.7247247247247248e-05,
+      "loss": 1.6252,
+      "step": 4558
+    },
+    {
+      "epoch": 0.005959453753776129,
+      "grad_norm": 0.9192073941230774,
+      "learning_rate": 2.7242242242242244e-05,
+      "loss": 0.7666,
+      "step": 4559
+    },
+    {
+      "epoch": 0.0059607609381923995,
+      "grad_norm": 1.2314153909683228,
+      "learning_rate": 2.723723723723724e-05,
+      "loss": 1.1803,
+      "step": 4560
+    },
+    {
+      "epoch": 0.00596206812260867,
+      "grad_norm": 1.2800151109695435,
+      "learning_rate": 2.7232232232232237e-05,
+      "loss": 1.5586,
+      "step": 4561
+    },
+    {
+      "epoch": 0.00596337530702494,
+      "grad_norm": 1.3003185987472534,
+      "learning_rate": 2.722722722722723e-05,
+      "loss": 1.7923,
+      "step": 4562
+    },
+    {
+      "epoch": 0.0059646824914412104,
+      "grad_norm": 1.7046586275100708,
+      "learning_rate": 2.7222222222222223e-05,
+      "loss": 2.2115,
+      "step": 4563
+    },
+    {
+      "epoch": 0.00596598967585748,
+      "grad_norm": 1.3891723155975342,
+      "learning_rate": 2.7217217217217216e-05,
+      "loss": 1.9827,
+      "step": 4564
+    },
+    {
+      "epoch": 0.00596729686027375,
+      "grad_norm": 1.2739899158477783,
+      "learning_rate": 2.7212212212212212e-05,
+      "loss": 1.1418,
+      "step": 4565
+    },
+    {
+      "epoch": 0.0059686040446900205,
+      "grad_norm": 1.645696759223938,
+      "learning_rate": 2.720720720720721e-05,
+      "loss": 1.2792,
+      "step": 4566
+    },
+    {
+      "epoch": 0.005969911229106291,
+      "grad_norm": 1.3250584602355957,
+      "learning_rate": 2.7202202202202205e-05,
+      "loss": 2.1625,
+      "step": 4567
+    },
+    {
+      "epoch": 0.005971218413522561,
+      "grad_norm": 1.8971846103668213,
+      "learning_rate": 2.7197197197197198e-05,
+      "loss": 1.4785,
+      "step": 4568
+    },
+    {
+      "epoch": 0.0059725255979388315,
+      "grad_norm": 1.7817602157592773,
+      "learning_rate": 2.7192192192192194e-05,
+      "loss": 1.4573,
+      "step": 4569
+    },
+    {
+      "epoch": 0.005973832782355102,
+      "grad_norm": 1.141405701637268,
+      "learning_rate": 2.718718718718719e-05,
+      "loss": 1.5392,
+      "step": 4570
+    },
+    {
+      "epoch": 0.005975139966771372,
+      "grad_norm": 1.2977555990219116,
+      "learning_rate": 2.718218218218218e-05,
+      "loss": 1.388,
+      "step": 4571
+    },
+    {
+      "epoch": 0.0059764471511876424,
+      "grad_norm": 1.260256290435791,
+      "learning_rate": 2.7177177177177177e-05,
+      "loss": 1.2962,
+      "step": 4572
+    },
+    {
+      "epoch": 0.005977754335603913,
+      "grad_norm": 2.004807233810425,
+      "learning_rate": 2.7172172172172173e-05,
+      "loss": 1.5993,
+      "step": 4573
+    },
+    {
+      "epoch": 0.005979061520020183,
+      "grad_norm": 1.2926323413848877,
+      "learning_rate": 2.716716716716717e-05,
+      "loss": 0.8663,
+      "step": 4574
+    },
+    {
+      "epoch": 0.005980368704436453,
+      "grad_norm": 1.4041225910186768,
+      "learning_rate": 2.7162162162162162e-05,
+      "loss": 0.643,
+      "step": 4575
+    },
+    {
+      "epoch": 0.005981675888852724,
+      "grad_norm": 1.1777406930923462,
+      "learning_rate": 2.715715715715716e-05,
+      "loss": 1.0346,
+      "step": 4576
+    },
+    {
+      "epoch": 0.005982983073268994,
+      "grad_norm": 1.823625087738037,
+      "learning_rate": 2.7152152152152155e-05,
+      "loss": 1.5389,
+      "step": 4577
+    },
+    {
+      "epoch": 0.005984290257685264,
+      "grad_norm": 1.3238357305526733,
+      "learning_rate": 2.714714714714715e-05,
+      "loss": 1.8398,
+      "step": 4578
+    },
+    {
+      "epoch": 0.005985597442101534,
+      "grad_norm": 1.365293264389038,
+      "learning_rate": 2.714214214214214e-05,
+      "loss": 1.6712,
+      "step": 4579
+    },
+    {
+      "epoch": 0.005986904626517804,
+      "grad_norm": 1.9305909872055054,
+      "learning_rate": 2.7137137137137137e-05,
+      "loss": 2.3235,
+      "step": 4580
+    },
+    {
+      "epoch": 0.0059882118109340745,
+      "grad_norm": 1.6678999662399292,
+      "learning_rate": 2.7132132132132134e-05,
+      "loss": 2.4073,
+      "step": 4581
+    },
+    {
+      "epoch": 0.005989518995350345,
+      "grad_norm": 1.6170896291732788,
+      "learning_rate": 2.7127127127127127e-05,
+      "loss": 1.0221,
+      "step": 4582
+    },
+    {
+      "epoch": 0.005990826179766615,
+      "grad_norm": 1.307937741279602,
+      "learning_rate": 2.7122122122122123e-05,
+      "loss": 1.5406,
+      "step": 4583
+    },
+    {
+      "epoch": 0.005992133364182885,
+      "grad_norm": 1.301514983177185,
+      "learning_rate": 2.711711711711712e-05,
+      "loss": 1.049,
+      "step": 4584
+    },
+    {
+      "epoch": 0.005993440548599156,
+      "grad_norm": 1.1849579811096191,
+      "learning_rate": 2.7112112112112116e-05,
+      "loss": 1.3121,
+      "step": 4585
+    },
+    {
+      "epoch": 0.005994747733015426,
+      "grad_norm": 1.4573951959609985,
+      "learning_rate": 2.710710710710711e-05,
+      "loss": 0.5689,
+      "step": 4586
+    },
+    {
+      "epoch": 0.005996054917431696,
+      "grad_norm": 1.2303340435028076,
+      "learning_rate": 2.7102102102102105e-05,
+      "loss": 1.5054,
+      "step": 4587
+    },
+    {
+      "epoch": 0.005997362101847967,
+      "grad_norm": 1.458243727684021,
+      "learning_rate": 2.7097097097097095e-05,
+      "loss": 1.9335,
+      "step": 4588
+    },
+    {
+      "epoch": 0.005998669286264237,
+      "grad_norm": 1.421000599861145,
+      "learning_rate": 2.709209209209209e-05,
+      "loss": 2.1278,
+      "step": 4589
+    },
+    {
+      "epoch": 0.005999976470680507,
+      "grad_norm": 1.4406183958053589,
+      "learning_rate": 2.7087087087087087e-05,
+      "loss": 1.1345,
+      "step": 4590
+    },
+    {
+      "epoch": 0.006001283655096778,
+      "grad_norm": 1.403174877166748,
+      "learning_rate": 2.7082082082082084e-05,
+      "loss": 1.8551,
+      "step": 4591
+    },
+    {
+      "epoch": 0.006002590839513048,
+      "grad_norm": 1.3452682495117188,
+      "learning_rate": 2.707707707707708e-05,
+      "loss": 1.1318,
+      "step": 4592
+    },
+    {
+      "epoch": 0.006003898023929318,
+      "grad_norm": 1.1725605726242065,
+      "learning_rate": 2.7072072072072073e-05,
+      "loss": 0.9315,
+      "step": 4593
+    },
+    {
+      "epoch": 0.006005205208345589,
+      "grad_norm": 1.4707680940628052,
+      "learning_rate": 2.706706706706707e-05,
+      "loss": 1.3352,
+      "step": 4594
+    },
+    {
+      "epoch": 0.006006512392761858,
+      "grad_norm": 1.1738170385360718,
+      "learning_rate": 2.7062062062062066e-05,
+      "loss": 0.9229,
+      "step": 4595
+    },
+    {
+      "epoch": 0.006007819577178128,
+      "grad_norm": 1.3580758571624756,
+      "learning_rate": 2.7057057057057055e-05,
+      "loss": 1.503,
+      "step": 4596
+    },
+    {
+      "epoch": 0.006009126761594399,
+      "grad_norm": 1.994407296180725,
+      "learning_rate": 2.7052052052052052e-05,
+      "loss": 2.0532,
+      "step": 4597
+    },
+    {
+      "epoch": 0.006010433946010669,
+      "grad_norm": 1.0120824575424194,
+      "learning_rate": 2.7047047047047048e-05,
+      "loss": 1.0223,
+      "step": 4598
+    },
+    {
+      "epoch": 0.006011741130426939,
+      "grad_norm": 1.233608603477478,
+      "learning_rate": 2.7042042042042044e-05,
+      "loss": 1.1796,
+      "step": 4599
+    },
+    {
+      "epoch": 0.00601304831484321,
+      "grad_norm": 1.1348155736923218,
+      "learning_rate": 2.7037037037037037e-05,
+      "loss": 1.5782,
+      "step": 4600
+    },
+    {
+      "epoch": 0.00601435549925948,
+      "grad_norm": 1.2070088386535645,
+      "learning_rate": 2.7032032032032034e-05,
+      "loss": 0.8292,
+      "step": 4601
+    },
+    {
+      "epoch": 0.00601566268367575,
+      "grad_norm": 2.4224276542663574,
+      "learning_rate": 2.702702702702703e-05,
+      "loss": 1.9593,
+      "step": 4602
+    },
+    {
+      "epoch": 0.006016969868092021,
+      "grad_norm": 1.4776989221572876,
+      "learning_rate": 2.7022022022022027e-05,
+      "loss": 1.8447,
+      "step": 4603
+    },
+    {
+      "epoch": 0.006018277052508291,
+      "grad_norm": 0.7324312925338745,
+      "learning_rate": 2.7017017017017016e-05,
+      "loss": 0.5582,
+      "step": 4604
+    },
+    {
+      "epoch": 0.006019584236924561,
+      "grad_norm": 1.5802336931228638,
+      "learning_rate": 2.7012012012012012e-05,
+      "loss": 0.5591,
+      "step": 4605
+    },
+    {
+      "epoch": 0.006020891421340832,
+      "grad_norm": 1.1445679664611816,
+      "learning_rate": 2.7007007007007005e-05,
+      "loss": 1.6053,
+      "step": 4606
+    },
+    {
+      "epoch": 0.006022198605757102,
+      "grad_norm": 1.8522350788116455,
+      "learning_rate": 2.7002002002002002e-05,
+      "loss": 1.807,
+      "step": 4607
+    },
+    {
+      "epoch": 0.006023505790173372,
+      "grad_norm": 1.3248999118804932,
+      "learning_rate": 2.6996996996996998e-05,
+      "loss": 1.8805,
+      "step": 4608
+    },
+    {
+      "epoch": 0.0060248129745896425,
+      "grad_norm": 1.551998257637024,
+      "learning_rate": 2.6991991991991995e-05,
+      "loss": 1.4324,
+      "step": 4609
+    },
+    {
+      "epoch": 0.006026120159005912,
+      "grad_norm": 1.2199177742004395,
+      "learning_rate": 2.698698698698699e-05,
+      "loss": 1.2793,
+      "step": 4610
+    },
+    {
+      "epoch": 0.006027427343422182,
+      "grad_norm": 1.4984536170959473,
+      "learning_rate": 2.6981981981981984e-05,
+      "loss": 1.6202,
+      "step": 4611
+    },
+    {
+      "epoch": 0.006028734527838453,
+      "grad_norm": 1.0211639404296875,
+      "learning_rate": 2.697697697697698e-05,
+      "loss": 0.714,
+      "step": 4612
+    },
+    {
+      "epoch": 0.006030041712254723,
+      "grad_norm": 1.2947882413864136,
+      "learning_rate": 2.697197197197197e-05,
+      "loss": 1.5083,
+      "step": 4613
+    },
+    {
+      "epoch": 0.006031348896670993,
+      "grad_norm": 1.4325261116027832,
+      "learning_rate": 2.6966966966966966e-05,
+      "loss": 1.7186,
+      "step": 4614
+    },
+    {
+      "epoch": 0.006032656081087264,
+      "grad_norm": 1.4929618835449219,
+      "learning_rate": 2.6961961961961963e-05,
+      "loss": 1.9007,
+      "step": 4615
+    },
+    {
+      "epoch": 0.006033963265503534,
+      "grad_norm": 1.3063008785247803,
+      "learning_rate": 2.695695695695696e-05,
+      "loss": 1.3873,
+      "step": 4616
+    },
+    {
+      "epoch": 0.006035270449919804,
+      "grad_norm": 1.7007031440734863,
+      "learning_rate": 2.6951951951951955e-05,
+      "loss": 1.0742,
+      "step": 4617
+    },
+    {
+      "epoch": 0.0060365776343360746,
+      "grad_norm": 1.254239797592163,
+      "learning_rate": 2.6946946946946948e-05,
+      "loss": 1.7879,
+      "step": 4618
+    },
+    {
+      "epoch": 0.006037884818752345,
+      "grad_norm": 1.3672372102737427,
+      "learning_rate": 2.6941941941941945e-05,
+      "loss": 1.5704,
+      "step": 4619
+    },
+    {
+      "epoch": 0.006039192003168615,
+      "grad_norm": 1.713386058807373,
+      "learning_rate": 2.693693693693694e-05,
+      "loss": 1.3197,
+      "step": 4620
+    },
+    {
+      "epoch": 0.0060404991875848855,
+      "grad_norm": 1.9801901578903198,
+      "learning_rate": 2.693193193193193e-05,
+      "loss": 2.1761,
+      "step": 4621
+    },
+    {
+      "epoch": 0.006041806372001156,
+      "grad_norm": 1.2543176412582397,
+      "learning_rate": 2.6926926926926927e-05,
+      "loss": 1.8595,
+      "step": 4622
+    },
+    {
+      "epoch": 0.006043113556417426,
+      "grad_norm": 1.2384915351867676,
+      "learning_rate": 2.6921921921921923e-05,
+      "loss": 1.2742,
+      "step": 4623
+    },
+    {
+      "epoch": 0.0060444207408336965,
+      "grad_norm": 1.288662314414978,
+      "learning_rate": 2.6916916916916916e-05,
+      "loss": 1.6448,
+      "step": 4624
+    },
+    {
+      "epoch": 0.006045727925249966,
+      "grad_norm": 1.4387112855911255,
+      "learning_rate": 2.6911911911911913e-05,
+      "loss": 1.4686,
+      "step": 4625
+    },
+    {
+      "epoch": 0.006047035109666236,
+      "grad_norm": 1.2616511583328247,
+      "learning_rate": 2.690690690690691e-05,
+      "loss": 0.9611,
+      "step": 4626
+    },
+    {
+      "epoch": 0.0060483422940825066,
+      "grad_norm": 1.7393677234649658,
+      "learning_rate": 2.6901901901901905e-05,
+      "loss": 2.1443,
+      "step": 4627
+    },
+    {
+      "epoch": 0.006049649478498777,
+      "grad_norm": 1.5628941059112549,
+      "learning_rate": 2.68968968968969e-05,
+      "loss": 1.5941,
+      "step": 4628
+    },
+    {
+      "epoch": 0.006050956662915047,
+      "grad_norm": 1.4653555154800415,
+      "learning_rate": 2.689189189189189e-05,
+      "loss": 2.0648,
+      "step": 4629
+    },
+    {
+      "epoch": 0.0060522638473313175,
+      "grad_norm": 1.7371406555175781,
+      "learning_rate": 2.6886886886886888e-05,
+      "loss": 1.7457,
+      "step": 4630
+    },
+    {
+      "epoch": 0.006053571031747588,
+      "grad_norm": 1.8325577974319458,
+      "learning_rate": 2.688188188188188e-05,
+      "loss": 1.7509,
+      "step": 4631
+    },
+    {
+      "epoch": 0.006054878216163858,
+      "grad_norm": 1.78169846534729,
+      "learning_rate": 2.6876876876876877e-05,
+      "loss": 1.2025,
+      "step": 4632
+    },
+    {
+      "epoch": 0.0060561854005801285,
+      "grad_norm": 1.4576302766799927,
+      "learning_rate": 2.6871871871871873e-05,
+      "loss": 1.7571,
+      "step": 4633
+    },
+    {
+      "epoch": 0.006057492584996399,
+      "grad_norm": 1.1141631603240967,
+      "learning_rate": 2.686686686686687e-05,
+      "loss": 1.0541,
+      "step": 4634
+    },
+    {
+      "epoch": 0.006058799769412669,
+      "grad_norm": 2.8849167823791504,
+      "learning_rate": 2.6861861861861866e-05,
+      "loss": 2.8289,
+      "step": 4635
+    },
+    {
+      "epoch": 0.0060601069538289394,
+      "grad_norm": 1.549849271774292,
+      "learning_rate": 2.685685685685686e-05,
+      "loss": 2.1104,
+      "step": 4636
+    },
+    {
+      "epoch": 0.00606141413824521,
+      "grad_norm": 1.2333029508590698,
+      "learning_rate": 2.6851851851851855e-05,
+      "loss": 1.7095,
+      "step": 4637
+    },
+    {
+      "epoch": 0.00606272132266148,
+      "grad_norm": 1.3913273811340332,
+      "learning_rate": 2.6846846846846845e-05,
+      "loss": 1.5645,
+      "step": 4638
+    },
+    {
+      "epoch": 0.00606402850707775,
+      "grad_norm": 1.425921082496643,
+      "learning_rate": 2.684184184184184e-05,
+      "loss": 1.0759,
+      "step": 4639
+    },
+    {
+      "epoch": 0.00606533569149402,
+      "grad_norm": 1.4491631984710693,
+      "learning_rate": 2.6836836836836838e-05,
+      "loss": 2.0763,
+      "step": 4640
+    },
+    {
+      "epoch": 0.00606664287591029,
+      "grad_norm": 1.2254585027694702,
+      "learning_rate": 2.6831831831831834e-05,
+      "loss": 0.9514,
+      "step": 4641
+    },
+    {
+      "epoch": 0.0060679500603265605,
+      "grad_norm": 1.8128684759140015,
+      "learning_rate": 2.6826826826826827e-05,
+      "loss": 1.523,
+      "step": 4642
+    },
+    {
+      "epoch": 0.006069257244742831,
+      "grad_norm": 1.1692755222320557,
+      "learning_rate": 2.6821821821821823e-05,
+      "loss": 1.5476,
+      "step": 4643
+    },
+    {
+      "epoch": 0.006070564429159101,
+      "grad_norm": 1.3306920528411865,
+      "learning_rate": 2.681681681681682e-05,
+      "loss": 1.6868,
+      "step": 4644
+    },
+    {
+      "epoch": 0.0060718716135753715,
+      "grad_norm": 1.2907334566116333,
+      "learning_rate": 2.6811811811811816e-05,
+      "loss": 1.4488,
+      "step": 4645
+    },
+    {
+      "epoch": 0.006073178797991642,
+      "grad_norm": 1.276984691619873,
+      "learning_rate": 2.6806806806806806e-05,
+      "loss": 1.8493,
+      "step": 4646
+    },
+    {
+      "epoch": 0.006074485982407912,
+      "grad_norm": 1.512281894683838,
+      "learning_rate": 2.6801801801801802e-05,
+      "loss": 1.6367,
+      "step": 4647
+    },
+    {
+      "epoch": 0.006075793166824182,
+      "grad_norm": 1.0937358140945435,
+      "learning_rate": 2.67967967967968e-05,
+      "loss": 1.2592,
+      "step": 4648
+    },
+    {
+      "epoch": 0.006077100351240453,
+      "grad_norm": 1.6638611555099487,
+      "learning_rate": 2.679179179179179e-05,
+      "loss": 2.0514,
+      "step": 4649
+    },
+    {
+      "epoch": 0.006078407535656723,
+      "grad_norm": 0.8934481143951416,
+      "learning_rate": 2.6786786786786788e-05,
+      "loss": 0.5937,
+      "step": 4650
+    },
+    {
+      "epoch": 0.006079714720072993,
+      "grad_norm": 1.3837021589279175,
+      "learning_rate": 2.6781781781781784e-05,
+      "loss": 2.5888,
+      "step": 4651
+    },
+    {
+      "epoch": 0.006081021904489264,
+      "grad_norm": 1.5535849332809448,
+      "learning_rate": 2.677677677677678e-05,
+      "loss": 1.5833,
+      "step": 4652
+    },
+    {
+      "epoch": 0.006082329088905534,
+      "grad_norm": 1.2867777347564697,
+      "learning_rate": 2.6771771771771777e-05,
+      "loss": 1.5797,
+      "step": 4653
+    },
+    {
+      "epoch": 0.006083636273321804,
+      "grad_norm": 1.2508496046066284,
+      "learning_rate": 2.6766766766766766e-05,
+      "loss": 1.9757,
+      "step": 4654
+    },
+    {
+      "epoch": 0.006084943457738075,
+      "grad_norm": 1.4031577110290527,
+      "learning_rate": 2.676176176176176e-05,
+      "loss": 2.1357,
+      "step": 4655
+    },
+    {
+      "epoch": 0.006086250642154344,
+      "grad_norm": 1.4137791395187378,
+      "learning_rate": 2.6756756756756756e-05,
+      "loss": 1.7447,
+      "step": 4656
+    },
+    {
+      "epoch": 0.006087557826570614,
+      "grad_norm": 1.3070495128631592,
+      "learning_rate": 2.6751751751751752e-05,
+      "loss": 1.7686,
+      "step": 4657
+    },
+    {
+      "epoch": 0.006088865010986885,
+      "grad_norm": 1.3221222162246704,
+      "learning_rate": 2.674674674674675e-05,
+      "loss": 1.5775,
+      "step": 4658
+    },
+    {
+      "epoch": 0.006090172195403155,
+      "grad_norm": 1.4027259349822998,
+      "learning_rate": 2.6741741741741745e-05,
+      "loss": 0.7041,
+      "step": 4659
+    },
+    {
+      "epoch": 0.006091479379819425,
+      "grad_norm": 1.4853458404541016,
+      "learning_rate": 2.6736736736736738e-05,
+      "loss": 1.7396,
+      "step": 4660
+    },
+    {
+      "epoch": 0.006092786564235696,
+      "grad_norm": 1.2362545728683472,
+      "learning_rate": 2.6731731731731734e-05,
+      "loss": 1.9568,
+      "step": 4661
+    },
+    {
+      "epoch": 0.006094093748651966,
+      "grad_norm": 1.1557559967041016,
+      "learning_rate": 2.672672672672673e-05,
+      "loss": 0.9835,
+      "step": 4662
+    },
+    {
+      "epoch": 0.006095400933068236,
+      "grad_norm": 1.5388582944869995,
+      "learning_rate": 2.672172172172172e-05,
+      "loss": 2.3253,
+      "step": 4663
+    },
+    {
+      "epoch": 0.006096708117484507,
+      "grad_norm": 1.6413805484771729,
+      "learning_rate": 2.6716716716716717e-05,
+      "loss": 1.8878,
+      "step": 4664
+    },
+    {
+      "epoch": 0.006098015301900777,
+      "grad_norm": 1.4095503091812134,
+      "learning_rate": 2.6711711711711713e-05,
+      "loss": 1.3684,
+      "step": 4665
+    },
+    {
+      "epoch": 0.006099322486317047,
+      "grad_norm": 1.2404285669326782,
+      "learning_rate": 2.670670670670671e-05,
+      "loss": 1.4802,
+      "step": 4666
+    },
+    {
+      "epoch": 0.006100629670733318,
+      "grad_norm": 1.4483082294464111,
+      "learning_rate": 2.6701701701701702e-05,
+      "loss": 1.6315,
+      "step": 4667
+    },
+    {
+      "epoch": 0.006101936855149588,
+      "grad_norm": 1.6716853380203247,
+      "learning_rate": 2.66966966966967e-05,
+      "loss": 1.8891,
+      "step": 4668
+    },
+    {
+      "epoch": 0.006103244039565858,
+      "grad_norm": 1.571984052658081,
+      "learning_rate": 2.6691691691691695e-05,
+      "loss": 1.9052,
+      "step": 4669
+    },
+    {
+      "epoch": 0.006104551223982129,
+      "grad_norm": 1.5272374153137207,
+      "learning_rate": 2.668668668668669e-05,
+      "loss": 2.3145,
+      "step": 4670
+    },
+    {
+      "epoch": 0.006105858408398398,
+      "grad_norm": 1.0078963041305542,
+      "learning_rate": 2.668168168168168e-05,
+      "loss": 0.7804,
+      "step": 4671
+    },
+    {
+      "epoch": 0.006107165592814668,
+      "grad_norm": 1.2091896533966064,
+      "learning_rate": 2.6676676676676677e-05,
+      "loss": 1.2498,
+      "step": 4672
+    },
+    {
+      "epoch": 0.006108472777230939,
+      "grad_norm": 0.7078223824501038,
+      "learning_rate": 2.667167167167167e-05,
+      "loss": 0.2627,
+      "step": 4673
+    },
+    {
+      "epoch": 0.006109779961647209,
+      "grad_norm": 1.321449637413025,
+      "learning_rate": 2.6666666666666667e-05,
+      "loss": 2.0435,
+      "step": 4674
+    },
+    {
+      "epoch": 0.006111087146063479,
+      "grad_norm": 1.4888925552368164,
+      "learning_rate": 2.6661661661661663e-05,
+      "loss": 1.4589,
+      "step": 4675
+    },
+    {
+      "epoch": 0.00611239433047975,
+      "grad_norm": 1.9104163646697998,
+      "learning_rate": 2.665665665665666e-05,
+      "loss": 1.9126,
+      "step": 4676
+    },
+    {
+      "epoch": 0.00611370151489602,
+      "grad_norm": 1.5757044553756714,
+      "learning_rate": 2.6651651651651656e-05,
+      "loss": 0.9498,
+      "step": 4677
+    },
+    {
+      "epoch": 0.00611500869931229,
+      "grad_norm": 1.9705997705459595,
+      "learning_rate": 2.664664664664665e-05,
+      "loss": 2.2523,
+      "step": 4678
+    },
+    {
+      "epoch": 0.006116315883728561,
+      "grad_norm": 1.508890151977539,
+      "learning_rate": 2.664164164164164e-05,
+      "loss": 1.5121,
+      "step": 4679
+    },
+    {
+      "epoch": 0.006117623068144831,
+      "grad_norm": 1.1426305770874023,
+      "learning_rate": 2.6636636636636635e-05,
+      "loss": 1.3935,
+      "step": 4680
+    },
+    {
+      "epoch": 0.006118930252561101,
+      "grad_norm": 1.3537492752075195,
+      "learning_rate": 2.663163163163163e-05,
+      "loss": 1.3065,
+      "step": 4681
+    },
+    {
+      "epoch": 0.0061202374369773715,
+      "grad_norm": 1.2592811584472656,
+      "learning_rate": 2.6626626626626627e-05,
+      "loss": 1.2459,
+      "step": 4682
+    },
+    {
+      "epoch": 0.006121544621393642,
+      "grad_norm": 2.3574771881103516,
+      "learning_rate": 2.6621621621621624e-05,
+      "loss": 1.7161,
+      "step": 4683
+    },
+    {
+      "epoch": 0.006122851805809912,
+      "grad_norm": 1.7480440139770508,
+      "learning_rate": 2.661661661661662e-05,
+      "loss": 1.3793,
+      "step": 4684
+    },
+    {
+      "epoch": 0.0061241589902261825,
+      "grad_norm": 1.2695801258087158,
+      "learning_rate": 2.6611611611611613e-05,
+      "loss": 0.6321,
+      "step": 4685
+    },
+    {
+      "epoch": 0.006125466174642452,
+      "grad_norm": 1.4838573932647705,
+      "learning_rate": 2.660660660660661e-05,
+      "loss": 1.1046,
+      "step": 4686
+    },
+    {
+      "epoch": 0.006126773359058722,
+      "grad_norm": 1.0667942762374878,
+      "learning_rate": 2.6601601601601606e-05,
+      "loss": 0.5684,
+      "step": 4687
+    },
+    {
+      "epoch": 0.006128080543474993,
+      "grad_norm": 1.012521743774414,
+      "learning_rate": 2.6596596596596595e-05,
+      "loss": 0.6907,
+      "step": 4688
+    },
+    {
+      "epoch": 0.006129387727891263,
+      "grad_norm": 1.9696877002716064,
+      "learning_rate": 2.659159159159159e-05,
+      "loss": 1.4446,
+      "step": 4689
+    },
+    {
+      "epoch": 0.006130694912307533,
+      "grad_norm": 1.455737590789795,
+      "learning_rate": 2.6586586586586588e-05,
+      "loss": 1.2188,
+      "step": 4690
+    },
+    {
+      "epoch": 0.0061320020967238036,
+      "grad_norm": 1.4361563920974731,
+      "learning_rate": 2.6581581581581584e-05,
+      "loss": 2.273,
+      "step": 4691
+    },
+    {
+      "epoch": 0.006133309281140074,
+      "grad_norm": 0.7802792191505432,
+      "learning_rate": 2.6576576576576577e-05,
+      "loss": 0.3718,
+      "step": 4692
+    },
+    {
+      "epoch": 0.006134616465556344,
+      "grad_norm": 1.371790885925293,
+      "learning_rate": 2.6571571571571574e-05,
+      "loss": 1.2948,
+      "step": 4693
+    },
+    {
+      "epoch": 0.0061359236499726145,
+      "grad_norm": 1.4105048179626465,
+      "learning_rate": 2.656656656656657e-05,
+      "loss": 2.0964,
+      "step": 4694
+    },
+    {
+      "epoch": 0.006137230834388885,
+      "grad_norm": 1.4307628870010376,
+      "learning_rate": 2.6561561561561566e-05,
+      "loss": 1.6141,
+      "step": 4695
+    },
+    {
+      "epoch": 0.006138538018805155,
+      "grad_norm": 1.3322772979736328,
+      "learning_rate": 2.6556556556556556e-05,
+      "loss": 1.1061,
+      "step": 4696
+    },
+    {
+      "epoch": 0.0061398452032214255,
+      "grad_norm": 1.5543593168258667,
+      "learning_rate": 2.6551551551551552e-05,
+      "loss": 1.6365,
+      "step": 4697
+    },
+    {
+      "epoch": 0.006141152387637696,
+      "grad_norm": 1.6536176204681396,
+      "learning_rate": 2.6546546546546545e-05,
+      "loss": 2.2407,
+      "step": 4698
+    },
+    {
+      "epoch": 0.006142459572053966,
+      "grad_norm": 1.528714895248413,
+      "learning_rate": 2.6541541541541542e-05,
+      "loss": 1.7994,
+      "step": 4699
+    },
+    {
+      "epoch": 0.0061437667564702364,
+      "grad_norm": 1.4893373250961304,
+      "learning_rate": 2.6536536536536538e-05,
+      "loss": 2.0433,
+      "step": 4700
+    },
+    {
+      "epoch": 0.006145073940886507,
+      "grad_norm": 1.20215904712677,
+      "learning_rate": 2.6531531531531534e-05,
+      "loss": 1.0625,
+      "step": 4701
+    },
+    {
+      "epoch": 0.006146381125302776,
+      "grad_norm": 1.4095020294189453,
+      "learning_rate": 2.652652652652653e-05,
+      "loss": 2.0433,
+      "step": 4702
+    },
+    {
+      "epoch": 0.0061476883097190465,
+      "grad_norm": 1.470470666885376,
+      "learning_rate": 2.6521521521521524e-05,
+      "loss": 2.1789,
+      "step": 4703
+    },
+    {
+      "epoch": 0.006148995494135317,
+      "grad_norm": 1.3548117876052856,
+      "learning_rate": 2.6516516516516517e-05,
+      "loss": 1.6584,
+      "step": 4704
+    },
+    {
+      "epoch": 0.006150302678551587,
+      "grad_norm": 1.16972815990448,
+      "learning_rate": 2.651151151151151e-05,
+      "loss": 1.3044,
+      "step": 4705
+    },
+    {
+      "epoch": 0.0061516098629678575,
+      "grad_norm": 0.9667633175849915,
+      "learning_rate": 2.6506506506506506e-05,
+      "loss": 0.4639,
+      "step": 4706
+    },
+    {
+      "epoch": 0.006152917047384128,
+      "grad_norm": 1.4115394353866577,
+      "learning_rate": 2.6501501501501502e-05,
+      "loss": 2.8034,
+      "step": 4707
+    },
+    {
+      "epoch": 0.006154224231800398,
+      "grad_norm": 1.7266497611999512,
+      "learning_rate": 2.64964964964965e-05,
+      "loss": 1.6869,
+      "step": 4708
+    },
+    {
+      "epoch": 0.0061555314162166684,
+      "grad_norm": 1.3582348823547363,
+      "learning_rate": 2.6491491491491495e-05,
+      "loss": 2.3296,
+      "step": 4709
+    },
+    {
+      "epoch": 0.006156838600632939,
+      "grad_norm": 1.5247173309326172,
+      "learning_rate": 2.6486486486486488e-05,
+      "loss": 2.3206,
+      "step": 4710
+    },
+    {
+      "epoch": 0.006158145785049209,
+      "grad_norm": 1.723049283027649,
+      "learning_rate": 2.6481481481481485e-05,
+      "loss": 2.6322,
+      "step": 4711
+    },
+    {
+      "epoch": 0.006159452969465479,
+      "grad_norm": 1.2405083179473877,
+      "learning_rate": 2.6476476476476474e-05,
+      "loss": 0.9819,
+      "step": 4712
+    },
+    {
+      "epoch": 0.00616076015388175,
+      "grad_norm": 1.5001581907272339,
+      "learning_rate": 2.647147147147147e-05,
+      "loss": 1.8055,
+      "step": 4713
+    },
+    {
+      "epoch": 0.00616206733829802,
+      "grad_norm": 1.7793477773666382,
+      "learning_rate": 2.6466466466466467e-05,
+      "loss": 1.7875,
+      "step": 4714
+    },
+    {
+      "epoch": 0.00616337452271429,
+      "grad_norm": 1.378769040107727,
+      "learning_rate": 2.6461461461461463e-05,
+      "loss": 1.542,
+      "step": 4715
+    },
+    {
+      "epoch": 0.006164681707130561,
+      "grad_norm": 1.3598569631576538,
+      "learning_rate": 2.6456456456456456e-05,
+      "loss": 1.9346,
+      "step": 4716
+    },
+    {
+      "epoch": 0.00616598889154683,
+      "grad_norm": 1.567480206489563,
+      "learning_rate": 2.6451451451451453e-05,
+      "loss": 1.581,
+      "step": 4717
+    },
+    {
+      "epoch": 0.0061672960759631005,
+      "grad_norm": 1.4699760675430298,
+      "learning_rate": 2.644644644644645e-05,
+      "loss": 1.1294,
+      "step": 4718
+    },
+    {
+      "epoch": 0.006168603260379371,
+      "grad_norm": 1.3948906660079956,
+      "learning_rate": 2.6441441441441445e-05,
+      "loss": 1.6606,
+      "step": 4719
+    },
+    {
+      "epoch": 0.006169910444795641,
+      "grad_norm": 1.3031389713287354,
+      "learning_rate": 2.643643643643644e-05,
+      "loss": 1.6123,
+      "step": 4720
+    },
+    {
+      "epoch": 0.006171217629211911,
+      "grad_norm": 1.3700019121170044,
+      "learning_rate": 2.643143143143143e-05,
+      "loss": 1.7261,
+      "step": 4721
+    },
+    {
+      "epoch": 0.006172524813628182,
+      "grad_norm": 1.0076428651809692,
+      "learning_rate": 2.6426426426426428e-05,
+      "loss": 0.9012,
+      "step": 4722
+    },
+    {
+      "epoch": 0.006173831998044452,
+      "grad_norm": 1.5477211475372314,
+      "learning_rate": 2.642142142142142e-05,
+      "loss": 2.0498,
+      "step": 4723
+    },
+    {
+      "epoch": 0.006175139182460722,
+      "grad_norm": 1.4300508499145508,
+      "learning_rate": 2.6416416416416417e-05,
+      "loss": 1.2934,
+      "step": 4724
+    },
+    {
+      "epoch": 0.006176446366876993,
+      "grad_norm": 1.2553033828735352,
+      "learning_rate": 2.6411411411411413e-05,
+      "loss": 2.0741,
+      "step": 4725
+    },
+    {
+      "epoch": 0.006177753551293263,
+      "grad_norm": 1.4733145236968994,
+      "learning_rate": 2.640640640640641e-05,
+      "loss": 1.1827,
+      "step": 4726
+    },
+    {
+      "epoch": 0.006179060735709533,
+      "grad_norm": 1.5104353427886963,
+      "learning_rate": 2.6401401401401406e-05,
+      "loss": 2.1088,
+      "step": 4727
+    },
+    {
+      "epoch": 0.006180367920125804,
+      "grad_norm": 1.4224801063537598,
+      "learning_rate": 2.63963963963964e-05,
+      "loss": 1.6722,
+      "step": 4728
+    },
+    {
+      "epoch": 0.006181675104542074,
+      "grad_norm": 1.436845064163208,
+      "learning_rate": 2.639139139139139e-05,
+      "loss": 1.3594,
+      "step": 4729
+    },
+    {
+      "epoch": 0.006182982288958344,
+      "grad_norm": 1.2946275472640991,
+      "learning_rate": 2.6386386386386385e-05,
+      "loss": 2.2457,
+      "step": 4730
+    },
+    {
+      "epoch": 0.006184289473374615,
+      "grad_norm": 1.219205617904663,
+      "learning_rate": 2.638138138138138e-05,
+      "loss": 1.8026,
+      "step": 4731
+    },
+    {
+      "epoch": 0.006185596657790884,
+      "grad_norm": 0.9623035192489624,
+      "learning_rate": 2.6376376376376378e-05,
+      "loss": 0.8416,
+      "step": 4732
+    },
+    {
+      "epoch": 0.006186903842207154,
+      "grad_norm": 1.3593882322311401,
+      "learning_rate": 2.6371371371371374e-05,
+      "loss": 0.9589,
+      "step": 4733
+    },
+    {
+      "epoch": 0.006188211026623425,
+      "grad_norm": 1.2340073585510254,
+      "learning_rate": 2.6366366366366367e-05,
+      "loss": 1.8316,
+      "step": 4734
+    },
+    {
+      "epoch": 0.006189518211039695,
+      "grad_norm": 1.385628581047058,
+      "learning_rate": 2.6361361361361363e-05,
+      "loss": 1.5554,
+      "step": 4735
+    },
+    {
+      "epoch": 0.006190825395455965,
+      "grad_norm": 1.6764763593673706,
+      "learning_rate": 2.635635635635636e-05,
+      "loss": 1.6712,
+      "step": 4736
+    },
+    {
+      "epoch": 0.006192132579872236,
+      "grad_norm": 1.4023232460021973,
+      "learning_rate": 2.635135135135135e-05,
+      "loss": 2.1597,
+      "step": 4737
+    },
+    {
+      "epoch": 0.006193439764288506,
+      "grad_norm": 1.5834654569625854,
+      "learning_rate": 2.6346346346346346e-05,
+      "loss": 1.6959,
+      "step": 4738
+    },
+    {
+      "epoch": 0.006194746948704776,
+      "grad_norm": 1.2830349206924438,
+      "learning_rate": 2.6341341341341342e-05,
+      "loss": 1.6158,
+      "step": 4739
+    },
+    {
+      "epoch": 0.006196054133121047,
+      "grad_norm": 1.3062167167663574,
+      "learning_rate": 2.633633633633634e-05,
+      "loss": 1.2823,
+      "step": 4740
+    },
+    {
+      "epoch": 0.006197361317537317,
+      "grad_norm": 1.2177425622940063,
+      "learning_rate": 2.633133133133133e-05,
+      "loss": 1.1528,
+      "step": 4741
+    },
+    {
+      "epoch": 0.006198668501953587,
+      "grad_norm": 2.1241137981414795,
+      "learning_rate": 2.6326326326326328e-05,
+      "loss": 1.9215,
+      "step": 4742
+    },
+    {
+      "epoch": 0.006199975686369858,
+      "grad_norm": 1.5425007343292236,
+      "learning_rate": 2.6321321321321324e-05,
+      "loss": 1.096,
+      "step": 4743
+    },
+    {
+      "epoch": 0.006201282870786128,
+      "grad_norm": 1.4108089208602905,
+      "learning_rate": 2.631631631631632e-05,
+      "loss": 1.434,
+      "step": 4744
+    },
+    {
+      "epoch": 0.006202590055202398,
+      "grad_norm": 1.071708083152771,
+      "learning_rate": 2.6311311311311317e-05,
+      "loss": 1.1933,
+      "step": 4745
+    },
+    {
+      "epoch": 0.0062038972396186685,
+      "grad_norm": 1.2898845672607422,
+      "learning_rate": 2.6306306306306306e-05,
+      "loss": 1.157,
+      "step": 4746
+    },
+    {
+      "epoch": 0.006205204424034938,
+      "grad_norm": 1.6358304023742676,
+      "learning_rate": 2.63013013013013e-05,
+      "loss": 1.9617,
+      "step": 4747
+    },
+    {
+      "epoch": 0.006206511608451208,
+      "grad_norm": 1.2863661050796509,
+      "learning_rate": 2.6296296296296296e-05,
+      "loss": 1.5833,
+      "step": 4748
+    },
+    {
+      "epoch": 0.006207818792867479,
+      "grad_norm": 1.2845563888549805,
+      "learning_rate": 2.6291291291291292e-05,
+      "loss": 1.2708,
+      "step": 4749
+    },
+    {
+      "epoch": 0.006209125977283749,
+      "grad_norm": 1.194750428199768,
+      "learning_rate": 2.628628628628629e-05,
+      "loss": 0.9808,
+      "step": 4750
+    },
+    {
+      "epoch": 0.006210433161700019,
+      "grad_norm": 1.1590017080307007,
+      "learning_rate": 2.6281281281281285e-05,
+      "loss": 0.9612,
+      "step": 4751
+    },
+    {
+      "epoch": 0.00621174034611629,
+      "grad_norm": 1.5183595418930054,
+      "learning_rate": 2.6276276276276278e-05,
+      "loss": 2.0858,
+      "step": 4752
+    },
+    {
+      "epoch": 0.00621304753053256,
+      "grad_norm": 1.4522358179092407,
+      "learning_rate": 2.6271271271271274e-05,
+      "loss": 1.9005,
+      "step": 4753
+    },
+    {
+      "epoch": 0.00621435471494883,
+      "grad_norm": 1.2549878358840942,
+      "learning_rate": 2.6266266266266264e-05,
+      "loss": 1.6393,
+      "step": 4754
+    },
+    {
+      "epoch": 0.0062156618993651006,
+      "grad_norm": 1.3866510391235352,
+      "learning_rate": 2.626126126126126e-05,
+      "loss": 2.1553,
+      "step": 4755
+    },
+    {
+      "epoch": 0.006216969083781371,
+      "grad_norm": 1.5478570461273193,
+      "learning_rate": 2.6256256256256256e-05,
+      "loss": 1.9862,
+      "step": 4756
+    },
+    {
+      "epoch": 0.006218276268197641,
+      "grad_norm": 1.1481629610061646,
+      "learning_rate": 2.6251251251251253e-05,
+      "loss": 0.5631,
+      "step": 4757
+    },
+    {
+      "epoch": 0.0062195834526139115,
+      "grad_norm": 1.2481003999710083,
+      "learning_rate": 2.624624624624625e-05,
+      "loss": 1.3347,
+      "step": 4758
+    },
+    {
+      "epoch": 0.006220890637030182,
+      "grad_norm": 1.6747281551361084,
+      "learning_rate": 2.6241241241241242e-05,
+      "loss": 1.3549,
+      "step": 4759
+    },
+    {
+      "epoch": 0.006222197821446452,
+      "grad_norm": 1.1186561584472656,
+      "learning_rate": 2.623623623623624e-05,
+      "loss": 1.4718,
+      "step": 4760
+    },
+    {
+      "epoch": 0.0062235050058627225,
+      "grad_norm": 1.4656668901443481,
+      "learning_rate": 2.6231231231231235e-05,
+      "loss": 1.9173,
+      "step": 4761
+    },
+    {
+      "epoch": 0.006224812190278993,
+      "grad_norm": 1.300826907157898,
+      "learning_rate": 2.6226226226226224e-05,
+      "loss": 1.6429,
+      "step": 4762
+    },
+    {
+      "epoch": 0.006226119374695262,
+      "grad_norm": 0.24033747613430023,
+      "learning_rate": 2.622122122122122e-05,
+      "loss": 0.0077,
+      "step": 4763
+    },
+    {
+      "epoch": 0.0062274265591115326,
+      "grad_norm": 1.4789477586746216,
+      "learning_rate": 2.6216216216216217e-05,
+      "loss": 1.7467,
+      "step": 4764
+    },
+    {
+      "epoch": 0.006228733743527803,
+      "grad_norm": 1.1243022680282593,
+      "learning_rate": 2.621121121121121e-05,
+      "loss": 1.2833,
+      "step": 4765
+    },
+    {
+      "epoch": 0.006230040927944073,
+      "grad_norm": 1.5849765539169312,
+      "learning_rate": 2.6206206206206207e-05,
+      "loss": 1.0956,
+      "step": 4766
+    },
+    {
+      "epoch": 0.0062313481123603435,
+      "grad_norm": 1.0976423025131226,
+      "learning_rate": 2.6201201201201203e-05,
+      "loss": 0.4623,
+      "step": 4767
+    },
+    {
+      "epoch": 0.006232655296776614,
+      "grad_norm": 1.1707427501678467,
+      "learning_rate": 2.61961961961962e-05,
+      "loss": 1.0877,
+      "step": 4768
+    },
+    {
+      "epoch": 0.006233962481192884,
+      "grad_norm": 1.9566854238510132,
+      "learning_rate": 2.6191191191191196e-05,
+      "loss": 1.9983,
+      "step": 4769
+    },
+    {
+      "epoch": 0.0062352696656091545,
+      "grad_norm": 1.1422438621520996,
+      "learning_rate": 2.618618618618619e-05,
+      "loss": 1.1963,
+      "step": 4770
+    },
+    {
+      "epoch": 0.006236576850025425,
+      "grad_norm": 1.3081384897232056,
+      "learning_rate": 2.618118118118118e-05,
+      "loss": 1.1439,
+      "step": 4771
+    },
+    {
+      "epoch": 0.006237884034441695,
+      "grad_norm": 1.4375317096710205,
+      "learning_rate": 2.6176176176176175e-05,
+      "loss": 1.3243,
+      "step": 4772
+    },
+    {
+      "epoch": 0.0062391912188579654,
+      "grad_norm": 1.3601716756820679,
+      "learning_rate": 2.617117117117117e-05,
+      "loss": 1.2135,
+      "step": 4773
+    },
+    {
+      "epoch": 0.006240498403274236,
+      "grad_norm": 1.3637810945510864,
+      "learning_rate": 2.6166166166166167e-05,
+      "loss": 1.4246,
+      "step": 4774
+    },
+    {
+      "epoch": 0.006241805587690506,
+      "grad_norm": 1.7627609968185425,
+      "learning_rate": 2.6161161161161164e-05,
+      "loss": 2.0621,
+      "step": 4775
+    },
+    {
+      "epoch": 0.006243112772106776,
+      "grad_norm": 1.3657066822052002,
+      "learning_rate": 2.615615615615616e-05,
+      "loss": 1.4517,
+      "step": 4776
+    },
+    {
+      "epoch": 0.006244419956523047,
+      "grad_norm": 1.3763835430145264,
+      "learning_rate": 2.6151151151151153e-05,
+      "loss": 1.3847,
+      "step": 4777
+    },
+    {
+      "epoch": 0.006245727140939316,
+      "grad_norm": 1.742016077041626,
+      "learning_rate": 2.614614614614615e-05,
+      "loss": 2.1943,
+      "step": 4778
+    },
+    {
+      "epoch": 0.0062470343253555865,
+      "grad_norm": 1.7974493503570557,
+      "learning_rate": 2.614114114114114e-05,
+      "loss": 1.4833,
+      "step": 4779
+    },
+    {
+      "epoch": 0.006248341509771857,
+      "grad_norm": 1.2447599172592163,
+      "learning_rate": 2.6136136136136135e-05,
+      "loss": 2.0969,
+      "step": 4780
+    },
+    {
+      "epoch": 0.006249648694188127,
+      "grad_norm": 0.9583659768104553,
+      "learning_rate": 2.613113113113113e-05,
+      "loss": 0.7769,
+      "step": 4781
+    },
+    {
+      "epoch": 0.0062509558786043974,
+      "grad_norm": 1.1657263040542603,
+      "learning_rate": 2.6126126126126128e-05,
+      "loss": 1.541,
+      "step": 4782
+    },
+    {
+      "epoch": 0.006252263063020668,
+      "grad_norm": 1.9231082201004028,
+      "learning_rate": 2.612112112112112e-05,
+      "loss": 2.1935,
+      "step": 4783
+    },
+    {
+      "epoch": 0.006253570247436938,
+      "grad_norm": 1.4425759315490723,
+      "learning_rate": 2.6116116116116117e-05,
+      "loss": 1.7725,
+      "step": 4784
+    },
+    {
+      "epoch": 0.006254877431853208,
+      "grad_norm": 1.478497862815857,
+      "learning_rate": 2.6111111111111114e-05,
+      "loss": 1.1695,
+      "step": 4785
+    },
+    {
+      "epoch": 0.006256184616269479,
+      "grad_norm": 1.6033686399459839,
+      "learning_rate": 2.610610610610611e-05,
+      "loss": 2.0271,
+      "step": 4786
+    },
+    {
+      "epoch": 0.006257491800685749,
+      "grad_norm": 1.5563942193984985,
+      "learning_rate": 2.61011011011011e-05,
+      "loss": 1.2953,
+      "step": 4787
+    },
+    {
+      "epoch": 0.006258798985102019,
+      "grad_norm": 1.2652244567871094,
+      "learning_rate": 2.6096096096096096e-05,
+      "loss": 1.3393,
+      "step": 4788
+    },
+    {
+      "epoch": 0.00626010616951829,
+      "grad_norm": 1.2679587602615356,
+      "learning_rate": 2.6091091091091092e-05,
+      "loss": 2.203,
+      "step": 4789
+    },
+    {
+      "epoch": 0.00626141335393456,
+      "grad_norm": 1.5652753114700317,
+      "learning_rate": 2.6086086086086085e-05,
+      "loss": 1.9241,
+      "step": 4790
+    },
+    {
+      "epoch": 0.00626272053835083,
+      "grad_norm": 2.295814037322998,
+      "learning_rate": 2.6081081081081082e-05,
+      "loss": 1.9224,
+      "step": 4791
+    },
+    {
+      "epoch": 0.006264027722767101,
+      "grad_norm": 1.451837182044983,
+      "learning_rate": 2.6076076076076078e-05,
+      "loss": 1.5074,
+      "step": 4792
+    },
+    {
+      "epoch": 0.00626533490718337,
+      "grad_norm": 1.3265202045440674,
+      "learning_rate": 2.6071071071071074e-05,
+      "loss": 1.5743,
+      "step": 4793
+    },
+    {
+      "epoch": 0.00626664209159964,
+      "grad_norm": 1.3379676342010498,
+      "learning_rate": 2.606606606606607e-05,
+      "loss": 1.5378,
+      "step": 4794
+    },
+    {
+      "epoch": 0.006267949276015911,
+      "grad_norm": 1.5971381664276123,
+      "learning_rate": 2.6061061061061064e-05,
+      "loss": 1.8594,
+      "step": 4795
+    },
+    {
+      "epoch": 0.006269256460432181,
+      "grad_norm": 1.527430772781372,
+      "learning_rate": 2.6056056056056057e-05,
+      "loss": 1.242,
+      "step": 4796
+    },
+    {
+      "epoch": 0.006270563644848451,
+      "grad_norm": 1.3615148067474365,
+      "learning_rate": 2.605105105105105e-05,
+      "loss": 1.0028,
+      "step": 4797
+    },
+    {
+      "epoch": 0.006271870829264722,
+      "grad_norm": 1.3173282146453857,
+      "learning_rate": 2.6046046046046046e-05,
+      "loss": 1.6724,
+      "step": 4798
+    },
+    {
+      "epoch": 0.006273178013680992,
+      "grad_norm": 1.5709670782089233,
+      "learning_rate": 2.6041041041041042e-05,
+      "loss": 2.07,
+      "step": 4799
+    },
+    {
+      "epoch": 0.006274485198097262,
+      "grad_norm": 1.2711657285690308,
+      "learning_rate": 2.603603603603604e-05,
+      "loss": 0.9115,
+      "step": 4800
+    },
+    {
+      "epoch": 0.006275792382513533,
+      "grad_norm": 1.4653055667877197,
+      "learning_rate": 2.6031031031031032e-05,
+      "loss": 2.0412,
+      "step": 4801
+    },
+    {
+      "epoch": 0.006277099566929803,
+      "grad_norm": 1.5319633483886719,
+      "learning_rate": 2.6026026026026028e-05,
+      "loss": 2.5698,
+      "step": 4802
+    },
+    {
+      "epoch": 0.006278406751346073,
+      "grad_norm": 1.1998448371887207,
+      "learning_rate": 2.6021021021021025e-05,
+      "loss": 1.594,
+      "step": 4803
+    },
+    {
+      "epoch": 0.006279713935762344,
+      "grad_norm": 1.4664112329483032,
+      "learning_rate": 2.6016016016016014e-05,
+      "loss": 1.788,
+      "step": 4804
+    },
+    {
+      "epoch": 0.006281021120178614,
+      "grad_norm": 1.8106110095977783,
+      "learning_rate": 2.601101101101101e-05,
+      "loss": 1.8143,
+      "step": 4805
+    },
+    {
+      "epoch": 0.006282328304594884,
+      "grad_norm": 1.168310284614563,
+      "learning_rate": 2.6006006006006007e-05,
+      "loss": 0.9455,
+      "step": 4806
+    },
+    {
+      "epoch": 0.006283635489011155,
+      "grad_norm": 1.5982136726379395,
+      "learning_rate": 2.6001001001001003e-05,
+      "loss": 2.5495,
+      "step": 4807
+    },
+    {
+      "epoch": 0.006284942673427425,
+      "grad_norm": 1.0001338720321655,
+      "learning_rate": 2.5995995995995996e-05,
+      "loss": 0.9693,
+      "step": 4808
+    },
+    {
+      "epoch": 0.006286249857843694,
+      "grad_norm": 0.9833245277404785,
+      "learning_rate": 2.5990990990990993e-05,
+      "loss": 1.1484,
+      "step": 4809
+    },
+    {
+      "epoch": 0.006287557042259965,
+      "grad_norm": 1.5022048950195312,
+      "learning_rate": 2.598598598598599e-05,
+      "loss": 1.7803,
+      "step": 4810
+    },
+    {
+      "epoch": 0.006288864226676235,
+      "grad_norm": 2.3877978324890137,
+      "learning_rate": 2.5980980980980985e-05,
+      "loss": 2.1045,
+      "step": 4811
+    },
+    {
+      "epoch": 0.006290171411092505,
+      "grad_norm": 1.317478060722351,
+      "learning_rate": 2.5975975975975975e-05,
+      "loss": 0.7678,
+      "step": 4812
+    },
+    {
+      "epoch": 0.006291478595508776,
+      "grad_norm": 1.5463601350784302,
+      "learning_rate": 2.597097097097097e-05,
+      "loss": 1.5506,
+      "step": 4813
+    },
+    {
+      "epoch": 0.006292785779925046,
+      "grad_norm": 1.450846552848816,
+      "learning_rate": 2.5965965965965968e-05,
+      "loss": 1.2573,
+      "step": 4814
+    },
+    {
+      "epoch": 0.006294092964341316,
+      "grad_norm": 1.6809941530227661,
+      "learning_rate": 2.596096096096096e-05,
+      "loss": 1.1448,
+      "step": 4815
+    },
+    {
+      "epoch": 0.006295400148757587,
+      "grad_norm": 1.632978916168213,
+      "learning_rate": 2.5955955955955957e-05,
+      "loss": 1.5462,
+      "step": 4816
+    },
+    {
+      "epoch": 0.006296707333173857,
+      "grad_norm": 1.1055536270141602,
+      "learning_rate": 2.5950950950950953e-05,
+      "loss": 0.7556,
+      "step": 4817
+    },
+    {
+      "epoch": 0.006298014517590127,
+      "grad_norm": 1.5149821043014526,
+      "learning_rate": 2.594594594594595e-05,
+      "loss": 1.5842,
+      "step": 4818
+    },
+    {
+      "epoch": 0.0062993217020063975,
+      "grad_norm": 1.4802197217941284,
+      "learning_rate": 2.5940940940940943e-05,
+      "loss": 1.4025,
+      "step": 4819
+    },
+    {
+      "epoch": 0.006300628886422668,
+      "grad_norm": 1.1901905536651611,
+      "learning_rate": 2.5935935935935936e-05,
+      "loss": 1.6937,
+      "step": 4820
+    },
+    {
+      "epoch": 0.006301936070838938,
+      "grad_norm": 1.5244590044021606,
+      "learning_rate": 2.593093093093093e-05,
+      "loss": 2.2286,
+      "step": 4821
+    },
+    {
+      "epoch": 0.0063032432552552085,
+      "grad_norm": 1.301283836364746,
+      "learning_rate": 2.5925925925925925e-05,
+      "loss": 1.5872,
+      "step": 4822
+    },
+    {
+      "epoch": 0.006304550439671479,
+      "grad_norm": 1.2024562358856201,
+      "learning_rate": 2.592092092092092e-05,
+      "loss": 1.6721,
+      "step": 4823
+    },
+    {
+      "epoch": 0.006305857624087748,
+      "grad_norm": 1.7497867345809937,
+      "learning_rate": 2.5915915915915918e-05,
+      "loss": 1.7156,
+      "step": 4824
+    },
+    {
+      "epoch": 0.006307164808504019,
+      "grad_norm": 1.3617902994155884,
+      "learning_rate": 2.5910910910910914e-05,
+      "loss": 1.5817,
+      "step": 4825
+    },
+    {
+      "epoch": 0.006308471992920289,
+      "grad_norm": 4.858547687530518,
+      "learning_rate": 2.5905905905905907e-05,
+      "loss": 1.8558,
+      "step": 4826
+    },
+    {
+      "epoch": 0.006309779177336559,
+      "grad_norm": 1.7486077547073364,
+      "learning_rate": 2.5900900900900903e-05,
+      "loss": 1.6573,
+      "step": 4827
+    },
+    {
+      "epoch": 0.0063110863617528296,
+      "grad_norm": 1.2262052297592163,
+      "learning_rate": 2.58958958958959e-05,
+      "loss": 1.2231,
+      "step": 4828
+    },
+    {
+      "epoch": 0.0063123935461691,
+      "grad_norm": 1.5300824642181396,
+      "learning_rate": 2.589089089089089e-05,
+      "loss": 1.1345,
+      "step": 4829
+    },
+    {
+      "epoch": 0.00631370073058537,
+      "grad_norm": 1.0661228895187378,
+      "learning_rate": 2.5885885885885886e-05,
+      "loss": 0.9679,
+      "step": 4830
+    },
+    {
+      "epoch": 0.0063150079150016405,
+      "grad_norm": 1.3968257904052734,
+      "learning_rate": 2.5880880880880882e-05,
+      "loss": 2.2058,
+      "step": 4831
+    },
+    {
+      "epoch": 0.006316315099417911,
+      "grad_norm": 1.3589729070663452,
+      "learning_rate": 2.587587587587588e-05,
+      "loss": 1.5297,
+      "step": 4832
+    },
+    {
+      "epoch": 0.006317622283834181,
+      "grad_norm": 1.2875456809997559,
+      "learning_rate": 2.587087087087087e-05,
+      "loss": 1.3669,
+      "step": 4833
+    },
+    {
+      "epoch": 0.0063189294682504515,
+      "grad_norm": 1.3697768449783325,
+      "learning_rate": 2.5865865865865868e-05,
+      "loss": 1.6222,
+      "step": 4834
+    },
+    {
+      "epoch": 0.006320236652666722,
+      "grad_norm": 1.2453858852386475,
+      "learning_rate": 2.5860860860860864e-05,
+      "loss": 0.9858,
+      "step": 4835
+    },
+    {
+      "epoch": 0.006321543837082992,
+      "grad_norm": 1.3395682573318481,
+      "learning_rate": 2.585585585585586e-05,
+      "loss": 1.144,
+      "step": 4836
+    },
+    {
+      "epoch": 0.006322851021499262,
+      "grad_norm": 1.460461974143982,
+      "learning_rate": 2.585085085085085e-05,
+      "loss": 1.3625,
+      "step": 4837
+    },
+    {
+      "epoch": 0.006324158205915533,
+      "grad_norm": 1.592047929763794,
+      "learning_rate": 2.5845845845845846e-05,
+      "loss": 1.9715,
+      "step": 4838
+    },
+    {
+      "epoch": 0.006325465390331802,
+      "grad_norm": 1.5000596046447754,
+      "learning_rate": 2.584084084084084e-05,
+      "loss": 1.9899,
+      "step": 4839
+    },
+    {
+      "epoch": 0.0063267725747480725,
+      "grad_norm": 1.725395917892456,
+      "learning_rate": 2.5835835835835836e-05,
+      "loss": 1.2786,
+      "step": 4840
+    },
+    {
+      "epoch": 0.006328079759164343,
+      "grad_norm": 1.9039555788040161,
+      "learning_rate": 2.5830830830830832e-05,
+      "loss": 1.1943,
+      "step": 4841
+    },
+    {
+      "epoch": 0.006329386943580613,
+      "grad_norm": 1.6448601484298706,
+      "learning_rate": 2.582582582582583e-05,
+      "loss": 1.863,
+      "step": 4842
+    },
+    {
+      "epoch": 0.0063306941279968835,
+      "grad_norm": 1.3886311054229736,
+      "learning_rate": 2.5820820820820825e-05,
+      "loss": 1.4866,
+      "step": 4843
+    },
+    {
+      "epoch": 0.006332001312413154,
+      "grad_norm": 1.284842848777771,
+      "learning_rate": 2.5815815815815818e-05,
+      "loss": 1.745,
+      "step": 4844
+    },
+    {
+      "epoch": 0.006333308496829424,
+      "grad_norm": 1.7960773706436157,
+      "learning_rate": 2.581081081081081e-05,
+      "loss": 2.2688,
+      "step": 4845
+    },
+    {
+      "epoch": 0.0063346156812456944,
+      "grad_norm": 1.4355748891830444,
+      "learning_rate": 2.5805805805805804e-05,
+      "loss": 1.2826,
+      "step": 4846
+    },
+    {
+      "epoch": 0.006335922865661965,
+      "grad_norm": 1.1046466827392578,
+      "learning_rate": 2.58008008008008e-05,
+      "loss": 1.015,
+      "step": 4847
+    },
+    {
+      "epoch": 0.006337230050078235,
+      "grad_norm": 1.656449317932129,
+      "learning_rate": 2.5795795795795796e-05,
+      "loss": 2.0886,
+      "step": 4848
+    },
+    {
+      "epoch": 0.006338537234494505,
+      "grad_norm": 1.306458592414856,
+      "learning_rate": 2.5790790790790793e-05,
+      "loss": 1.816,
+      "step": 4849
+    },
+    {
+      "epoch": 0.006339844418910776,
+      "grad_norm": 1.7660913467407227,
+      "learning_rate": 2.578578578578579e-05,
+      "loss": 1.5738,
+      "step": 4850
+    },
+    {
+      "epoch": 0.006341151603327046,
+      "grad_norm": 1.4156774282455444,
+      "learning_rate": 2.5780780780780782e-05,
+      "loss": 1.293,
+      "step": 4851
+    },
+    {
+      "epoch": 0.006342458787743316,
+      "grad_norm": 0.594032883644104,
+      "learning_rate": 2.577577577577578e-05,
+      "loss": 0.2397,
+      "step": 4852
+    },
+    {
+      "epoch": 0.006343765972159587,
+      "grad_norm": 1.6140273809432983,
+      "learning_rate": 2.5770770770770775e-05,
+      "loss": 2.1572,
+      "step": 4853
+    },
+    {
+      "epoch": 0.006345073156575856,
+      "grad_norm": 1.7880245447158813,
+      "learning_rate": 2.5765765765765764e-05,
+      "loss": 1.7242,
+      "step": 4854
+    },
+    {
+      "epoch": 0.0063463803409921265,
+      "grad_norm": 1.6952918767929077,
+      "learning_rate": 2.576076076076076e-05,
+      "loss": 0.6082,
+      "step": 4855
+    },
+    {
+      "epoch": 0.006347687525408397,
+      "grad_norm": 1.4121230840682983,
+      "learning_rate": 2.5755755755755757e-05,
+      "loss": 2.2601,
+      "step": 4856
+    },
+    {
+      "epoch": 0.006348994709824667,
+      "grad_norm": 1.4291114807128906,
+      "learning_rate": 2.575075075075075e-05,
+      "loss": 1.1732,
+      "step": 4857
+    },
+    {
+      "epoch": 0.006350301894240937,
+      "grad_norm": 1.3873541355133057,
+      "learning_rate": 2.5745745745745747e-05,
+      "loss": 1.4346,
+      "step": 4858
+    },
+    {
+      "epoch": 0.006351609078657208,
+      "grad_norm": 1.324655532836914,
+      "learning_rate": 2.5740740740740743e-05,
+      "loss": 1.7449,
+      "step": 4859
+    },
+    {
+      "epoch": 0.006352916263073478,
+      "grad_norm": 1.070568323135376,
+      "learning_rate": 2.573573573573574e-05,
+      "loss": 1.3591,
+      "step": 4860
+    },
+    {
+      "epoch": 0.006354223447489748,
+      "grad_norm": 1.368545413017273,
+      "learning_rate": 2.5730730730730736e-05,
+      "loss": 1.4135,
+      "step": 4861
+    },
+    {
+      "epoch": 0.006355530631906019,
+      "grad_norm": 1.4678174257278442,
+      "learning_rate": 2.5725725725725725e-05,
+      "loss": 2.1047,
+      "step": 4862
+    },
+    {
+      "epoch": 0.006356837816322289,
+      "grad_norm": 1.5521612167358398,
+      "learning_rate": 2.572072072072072e-05,
+      "loss": 1.203,
+      "step": 4863
+    },
+    {
+      "epoch": 0.006358145000738559,
+      "grad_norm": 1.4992181062698364,
+      "learning_rate": 2.5715715715715715e-05,
+      "loss": 1.3157,
+      "step": 4864
+    },
+    {
+      "epoch": 0.00635945218515483,
+      "grad_norm": 1.8472744226455688,
+      "learning_rate": 2.571071071071071e-05,
+      "loss": 1.7399,
+      "step": 4865
+    },
+    {
+      "epoch": 0.0063607593695711,
+      "grad_norm": 1.6695661544799805,
+      "learning_rate": 2.5705705705705707e-05,
+      "loss": 1.8314,
+      "step": 4866
+    },
+    {
+      "epoch": 0.00636206655398737,
+      "grad_norm": 1.4658576250076294,
+      "learning_rate": 2.5700700700700704e-05,
+      "loss": 1.7412,
+      "step": 4867
+    },
+    {
+      "epoch": 0.006363373738403641,
+      "grad_norm": 1.4932149648666382,
+      "learning_rate": 2.56956956956957e-05,
+      "loss": 2.2262,
+      "step": 4868
+    },
+    {
+      "epoch": 0.006364680922819911,
+      "grad_norm": 1.1450793743133545,
+      "learning_rate": 2.5690690690690693e-05,
+      "loss": 1.1826,
+      "step": 4869
+    },
+    {
+      "epoch": 0.00636598810723618,
+      "grad_norm": 1.6078826189041138,
+      "learning_rate": 2.5685685685685686e-05,
+      "loss": 2.0047,
+      "step": 4870
+    },
+    {
+      "epoch": 0.006367295291652451,
+      "grad_norm": 1.4217129945755005,
+      "learning_rate": 2.568068068068068e-05,
+      "loss": 1.8428,
+      "step": 4871
+    },
+    {
+      "epoch": 0.006368602476068721,
+      "grad_norm": 1.8352524042129517,
+      "learning_rate": 2.5675675675675675e-05,
+      "loss": 2.7877,
+      "step": 4872
+    },
+    {
+      "epoch": 0.006369909660484991,
+      "grad_norm": 1.4695007801055908,
+      "learning_rate": 2.567067067067067e-05,
+      "loss": 2.1189,
+      "step": 4873
+    },
+    {
+      "epoch": 0.006371216844901262,
+      "grad_norm": 1.0273864269256592,
+      "learning_rate": 2.5665665665665668e-05,
+      "loss": 0.9118,
+      "step": 4874
+    },
+    {
+      "epoch": 0.006372524029317532,
+      "grad_norm": 1.62455415725708,
+      "learning_rate": 2.566066066066066e-05,
+      "loss": 1.5222,
+      "step": 4875
+    },
+    {
+      "epoch": 0.006373831213733802,
+      "grad_norm": 1.2171602249145508,
+      "learning_rate": 2.5655655655655657e-05,
+      "loss": 1.8902,
+      "step": 4876
+    },
+    {
+      "epoch": 0.006375138398150073,
+      "grad_norm": 1.5469013452529907,
+      "learning_rate": 2.5650650650650654e-05,
+      "loss": 0.7982,
+      "step": 4877
+    },
+    {
+      "epoch": 0.006376445582566343,
+      "grad_norm": 2.2952897548675537,
+      "learning_rate": 2.564564564564565e-05,
+      "loss": 1.3783,
+      "step": 4878
+    },
+    {
+      "epoch": 0.006377752766982613,
+      "grad_norm": 1.0861114263534546,
+      "learning_rate": 2.564064064064064e-05,
+      "loss": 1.5308,
+      "step": 4879
+    },
+    {
+      "epoch": 0.006379059951398884,
+      "grad_norm": 1.454387903213501,
+      "learning_rate": 2.5635635635635636e-05,
+      "loss": 2.1193,
+      "step": 4880
+    },
+    {
+      "epoch": 0.006380367135815154,
+      "grad_norm": 1.1714078187942505,
+      "learning_rate": 2.5630630630630632e-05,
+      "loss": 1.6009,
+      "step": 4881
+    },
+    {
+      "epoch": 0.006381674320231424,
+      "grad_norm": 1.1439074277877808,
+      "learning_rate": 2.5625625625625625e-05,
+      "loss": 0.8435,
+      "step": 4882
+    },
+    {
+      "epoch": 0.0063829815046476945,
+      "grad_norm": 1.246873140335083,
+      "learning_rate": 2.562062062062062e-05,
+      "loss": 1.9046,
+      "step": 4883
+    },
+    {
+      "epoch": 0.006384288689063965,
+      "grad_norm": 1.4038609266281128,
+      "learning_rate": 2.5615615615615618e-05,
+      "loss": 1.6063,
+      "step": 4884
+    },
+    {
+      "epoch": 0.006385595873480234,
+      "grad_norm": 1.8748316764831543,
+      "learning_rate": 2.5610610610610614e-05,
+      "loss": 1.8987,
+      "step": 4885
+    },
+    {
+      "epoch": 0.006386903057896505,
+      "grad_norm": 1.275752305984497,
+      "learning_rate": 2.560560560560561e-05,
+      "loss": 1.463,
+      "step": 4886
+    },
+    {
+      "epoch": 0.006388210242312775,
+      "grad_norm": 0.5025787949562073,
+      "learning_rate": 2.56006006006006e-05,
+      "loss": 0.2664,
+      "step": 4887
+    },
+    {
+      "epoch": 0.006389517426729045,
+      "grad_norm": 1.2886857986450195,
+      "learning_rate": 2.5595595595595597e-05,
+      "loss": 1.8038,
+      "step": 4888
+    },
+    {
+      "epoch": 0.006390824611145316,
+      "grad_norm": 1.4118934869766235,
+      "learning_rate": 2.559059059059059e-05,
+      "loss": 1.6147,
+      "step": 4889
+    },
+    {
+      "epoch": 0.006392131795561586,
+      "grad_norm": 1.5031754970550537,
+      "learning_rate": 2.5585585585585586e-05,
+      "loss": 1.6839,
+      "step": 4890
+    },
+    {
+      "epoch": 0.006393438979977856,
+      "grad_norm": 1.3891412019729614,
+      "learning_rate": 2.5580580580580582e-05,
+      "loss": 1.4146,
+      "step": 4891
+    },
+    {
+      "epoch": 0.0063947461643941265,
+      "grad_norm": 1.548589825630188,
+      "learning_rate": 2.557557557557558e-05,
+      "loss": 1.7612,
+      "step": 4892
+    },
+    {
+      "epoch": 0.006396053348810397,
+      "grad_norm": 1.6407426595687866,
+      "learning_rate": 2.5570570570570572e-05,
+      "loss": 1.8205,
+      "step": 4893
+    },
+    {
+      "epoch": 0.006397360533226667,
+      "grad_norm": 0.9312735795974731,
+      "learning_rate": 2.5565565565565568e-05,
+      "loss": 0.2967,
+      "step": 4894
+    },
+    {
+      "epoch": 0.0063986677176429375,
+      "grad_norm": 1.3159929513931274,
+      "learning_rate": 2.5560560560560558e-05,
+      "loss": 1.5412,
+      "step": 4895
+    },
+    {
+      "epoch": 0.006399974902059208,
+      "grad_norm": 1.110274076461792,
+      "learning_rate": 2.5555555555555554e-05,
+      "loss": 0.9801,
+      "step": 4896
+    },
+    {
+      "epoch": 0.006401282086475478,
+      "grad_norm": 1.428620457649231,
+      "learning_rate": 2.555055055055055e-05,
+      "loss": 1.1076,
+      "step": 4897
+    },
+    {
+      "epoch": 0.0064025892708917485,
+      "grad_norm": 1.1856319904327393,
+      "learning_rate": 2.5545545545545547e-05,
+      "loss": 2.0598,
+      "step": 4898
+    },
+    {
+      "epoch": 0.006403896455308019,
+      "grad_norm": 2.0306930541992188,
+      "learning_rate": 2.5540540540540543e-05,
+      "loss": 2.2419,
+      "step": 4899
+    },
+    {
+      "epoch": 0.006405203639724288,
+      "grad_norm": 1.234919786453247,
+      "learning_rate": 2.5535535535535536e-05,
+      "loss": 1.8929,
+      "step": 4900
+    },
+    {
+      "epoch": 0.0064065108241405586,
+      "grad_norm": 1.6334352493286133,
+      "learning_rate": 2.5530530530530532e-05,
+      "loss": 1.1904,
+      "step": 4901
+    },
+    {
+      "epoch": 0.006407818008556829,
+      "grad_norm": 1.4979437589645386,
+      "learning_rate": 2.552552552552553e-05,
+      "loss": 2.3443,
+      "step": 4902
+    },
+    {
+      "epoch": 0.006409125192973099,
+      "grad_norm": 1.7883096933364868,
+      "learning_rate": 2.5520520520520525e-05,
+      "loss": 1.1907,
+      "step": 4903
+    },
+    {
+      "epoch": 0.0064104323773893695,
+      "grad_norm": 1.690582036972046,
+      "learning_rate": 2.5515515515515515e-05,
+      "loss": 0.6268,
+      "step": 4904
+    },
+    {
+      "epoch": 0.00641173956180564,
+      "grad_norm": 2.273069381713867,
+      "learning_rate": 2.551051051051051e-05,
+      "loss": 1.8646,
+      "step": 4905
+    },
+    {
+      "epoch": 0.00641304674622191,
+      "grad_norm": 1.207201600074768,
+      "learning_rate": 2.5505505505505508e-05,
+      "loss": 0.9243,
+      "step": 4906
+    },
+    {
+      "epoch": 0.0064143539306381805,
+      "grad_norm": 1.1845802068710327,
+      "learning_rate": 2.55005005005005e-05,
+      "loss": 1.3383,
+      "step": 4907
+    },
+    {
+      "epoch": 0.006415661115054451,
+      "grad_norm": 0.7869453430175781,
+      "learning_rate": 2.5495495495495497e-05,
+      "loss": 0.5253,
+      "step": 4908
+    },
+    {
+      "epoch": 0.006416968299470721,
+      "grad_norm": 1.4020757675170898,
+      "learning_rate": 2.5490490490490493e-05,
+      "loss": 2.1087,
+      "step": 4909
+    },
+    {
+      "epoch": 0.0064182754838869914,
+      "grad_norm": 1.2631078958511353,
+      "learning_rate": 2.548548548548549e-05,
+      "loss": 1.7588,
+      "step": 4910
+    },
+    {
+      "epoch": 0.006419582668303262,
+      "grad_norm": 2.405306339263916,
+      "learning_rate": 2.5480480480480483e-05,
+      "loss": 1.8134,
+      "step": 4911
+    },
+    {
+      "epoch": 0.006420889852719532,
+      "grad_norm": 1.167773723602295,
+      "learning_rate": 2.5475475475475476e-05,
+      "loss": 1.1651,
+      "step": 4912
+    },
+    {
+      "epoch": 0.006422197037135802,
+      "grad_norm": 0.7471924424171448,
+      "learning_rate": 2.547047047047047e-05,
+      "loss": 0.5299,
+      "step": 4913
+    },
+    {
+      "epoch": 0.006423504221552073,
+      "grad_norm": 1.1760526895523071,
+      "learning_rate": 2.5465465465465465e-05,
+      "loss": 0.9048,
+      "step": 4914
+    },
+    {
+      "epoch": 0.006424811405968343,
+      "grad_norm": 1.419487476348877,
+      "learning_rate": 2.546046046046046e-05,
+      "loss": 1.6001,
+      "step": 4915
+    },
+    {
+      "epoch": 0.0064261185903846125,
+      "grad_norm": 1.5802972316741943,
+      "learning_rate": 2.5455455455455458e-05,
+      "loss": 2.4647,
+      "step": 4916
+    },
+    {
+      "epoch": 0.006427425774800883,
+      "grad_norm": 1.7615938186645508,
+      "learning_rate": 2.5450450450450454e-05,
+      "loss": 2.0673,
+      "step": 4917
+    },
+    {
+      "epoch": 0.006428732959217153,
+      "grad_norm": 1.6433693170547485,
+      "learning_rate": 2.5445445445445447e-05,
+      "loss": 1.5008,
+      "step": 4918
+    },
+    {
+      "epoch": 0.0064300401436334234,
+      "grad_norm": 1.3868213891983032,
+      "learning_rate": 2.5440440440440443e-05,
+      "loss": 1.7478,
+      "step": 4919
+    },
+    {
+      "epoch": 0.006431347328049694,
+      "grad_norm": 2.6759636402130127,
+      "learning_rate": 2.5435435435435433e-05,
+      "loss": 1.6087,
+      "step": 4920
+    },
+    {
+      "epoch": 0.006432654512465964,
+      "grad_norm": 1.375841498374939,
+      "learning_rate": 2.543043043043043e-05,
+      "loss": 1.8341,
+      "step": 4921
+    },
+    {
+      "epoch": 0.006433961696882234,
+      "grad_norm": 1.2199252843856812,
+      "learning_rate": 2.5425425425425426e-05,
+      "loss": 1.856,
+      "step": 4922
+    },
+    {
+      "epoch": 0.006435268881298505,
+      "grad_norm": 1.4435184001922607,
+      "learning_rate": 2.5420420420420422e-05,
+      "loss": 2.1369,
+      "step": 4923
+    },
+    {
+      "epoch": 0.006436576065714775,
+      "grad_norm": 1.5833723545074463,
+      "learning_rate": 2.541541541541542e-05,
+      "loss": 1.9347,
+      "step": 4924
+    },
+    {
+      "epoch": 0.006437883250131045,
+      "grad_norm": 1.127346396446228,
+      "learning_rate": 2.541041041041041e-05,
+      "loss": 1.004,
+      "step": 4925
+    },
+    {
+      "epoch": 0.006439190434547316,
+      "grad_norm": 1.5756001472473145,
+      "learning_rate": 2.5405405405405408e-05,
+      "loss": 2.2927,
+      "step": 4926
+    },
+    {
+      "epoch": 0.006440497618963586,
+      "grad_norm": 1.2051646709442139,
+      "learning_rate": 2.5400400400400404e-05,
+      "loss": 1.3271,
+      "step": 4927
+    },
+    {
+      "epoch": 0.006441804803379856,
+      "grad_norm": 1.378963828086853,
+      "learning_rate": 2.53953953953954e-05,
+      "loss": 1.2765,
+      "step": 4928
+    },
+    {
+      "epoch": 0.006443111987796127,
+      "grad_norm": 1.7246993780136108,
+      "learning_rate": 2.539039039039039e-05,
+      "loss": 1.6229,
+      "step": 4929
+    },
+    {
+      "epoch": 0.006444419172212397,
+      "grad_norm": 1.5072771310806274,
+      "learning_rate": 2.5385385385385386e-05,
+      "loss": 1.5982,
+      "step": 4930
+    },
+    {
+      "epoch": 0.006445726356628666,
+      "grad_norm": 1.236986517906189,
+      "learning_rate": 2.538038038038038e-05,
+      "loss": 1.7623,
+      "step": 4931
+    },
+    {
+      "epoch": 0.006447033541044937,
+      "grad_norm": 1.3008828163146973,
+      "learning_rate": 2.5375375375375376e-05,
+      "loss": 1.7933,
+      "step": 4932
+    },
+    {
+      "epoch": 0.006448340725461207,
+      "grad_norm": 1.3070518970489502,
+      "learning_rate": 2.5370370370370372e-05,
+      "loss": 1.323,
+      "step": 4933
+    },
+    {
+      "epoch": 0.006449647909877477,
+      "grad_norm": 1.4692625999450684,
+      "learning_rate": 2.536536536536537e-05,
+      "loss": 1.976,
+      "step": 4934
+    },
+    {
+      "epoch": 0.006450955094293748,
+      "grad_norm": 1.6098899841308594,
+      "learning_rate": 2.5360360360360365e-05,
+      "loss": 1.4649,
+      "step": 4935
+    },
+    {
+      "epoch": 0.006452262278710018,
+      "grad_norm": 1.1645052433013916,
+      "learning_rate": 2.5355355355355358e-05,
+      "loss": 1.052,
+      "step": 4936
+    },
+    {
+      "epoch": 0.006453569463126288,
+      "grad_norm": 1.4120129346847534,
+      "learning_rate": 2.535035035035035e-05,
+      "loss": 1.7336,
+      "step": 4937
+    },
+    {
+      "epoch": 0.006454876647542559,
+      "grad_norm": 1.3779908418655396,
+      "learning_rate": 2.5345345345345344e-05,
+      "loss": 1.6335,
+      "step": 4938
+    },
+    {
+      "epoch": 0.006456183831958829,
+      "grad_norm": 1.5670130252838135,
+      "learning_rate": 2.534034034034034e-05,
+      "loss": 1.5972,
+      "step": 4939
+    },
+    {
+      "epoch": 0.006457491016375099,
+      "grad_norm": 1.3322279453277588,
+      "learning_rate": 2.5335335335335336e-05,
+      "loss": 1.469,
+      "step": 4940
+    },
+    {
+      "epoch": 0.00645879820079137,
+      "grad_norm": 1.30517578125,
+      "learning_rate": 2.5330330330330333e-05,
+      "loss": 1.7536,
+      "step": 4941
+    },
+    {
+      "epoch": 0.00646010538520764,
+      "grad_norm": 1.4458529949188232,
+      "learning_rate": 2.532532532532533e-05,
+      "loss": 1.3623,
+      "step": 4942
+    },
+    {
+      "epoch": 0.00646141256962391,
+      "grad_norm": 1.7416597604751587,
+      "learning_rate": 2.5320320320320322e-05,
+      "loss": 1.5861,
+      "step": 4943
+    },
+    {
+      "epoch": 0.006462719754040181,
+      "grad_norm": 1.4250835180282593,
+      "learning_rate": 2.531531531531532e-05,
+      "loss": 1.3175,
+      "step": 4944
+    },
+    {
+      "epoch": 0.006464026938456451,
+      "grad_norm": 1.243314266204834,
+      "learning_rate": 2.5310310310310308e-05,
+      "loss": 1.6925,
+      "step": 4945
+    },
+    {
+      "epoch": 0.00646533412287272,
+      "grad_norm": 1.3566879034042358,
+      "learning_rate": 2.5305305305305304e-05,
+      "loss": 1.238,
+      "step": 4946
+    },
+    {
+      "epoch": 0.006466641307288991,
+      "grad_norm": 1.457491397857666,
+      "learning_rate": 2.53003003003003e-05,
+      "loss": 2.0554,
+      "step": 4947
+    },
+    {
+      "epoch": 0.006467948491705261,
+      "grad_norm": 1.1077502965927124,
+      "learning_rate": 2.5295295295295297e-05,
+      "loss": 0.8012,
+      "step": 4948
+    },
+    {
+      "epoch": 0.006469255676121531,
+      "grad_norm": 1.1061164140701294,
+      "learning_rate": 2.529029029029029e-05,
+      "loss": 1.2927,
+      "step": 4949
+    },
+    {
+      "epoch": 0.006470562860537802,
+      "grad_norm": 1.4009140729904175,
+      "learning_rate": 2.5285285285285286e-05,
+      "loss": 1.2944,
+      "step": 4950
+    },
+    {
+      "epoch": 0.006471870044954072,
+      "grad_norm": 1.2200093269348145,
+      "learning_rate": 2.5280280280280283e-05,
+      "loss": 1.6784,
+      "step": 4951
+    },
+    {
+      "epoch": 0.006473177229370342,
+      "grad_norm": 1.3305288553237915,
+      "learning_rate": 2.527527527527528e-05,
+      "loss": 1.6161,
+      "step": 4952
+    },
+    {
+      "epoch": 0.006474484413786613,
+      "grad_norm": 1.3028401136398315,
+      "learning_rate": 2.527027027027027e-05,
+      "loss": 1.7749,
+      "step": 4953
+    },
+    {
+      "epoch": 0.006475791598202883,
+      "grad_norm": 1.2685480117797852,
+      "learning_rate": 2.5265265265265265e-05,
+      "loss": 1.374,
+      "step": 4954
+    },
+    {
+      "epoch": 0.006477098782619153,
+      "grad_norm": 1.9004288911819458,
+      "learning_rate": 2.526026026026026e-05,
+      "loss": 2.4514,
+      "step": 4955
+    },
+    {
+      "epoch": 0.0064784059670354235,
+      "grad_norm": 1.0486550331115723,
+      "learning_rate": 2.5255255255255254e-05,
+      "loss": 0.8672,
+      "step": 4956
+    },
+    {
+      "epoch": 0.006479713151451694,
+      "grad_norm": 1.1374530792236328,
+      "learning_rate": 2.525025025025025e-05,
+      "loss": 2.0519,
+      "step": 4957
+    },
+    {
+      "epoch": 0.006481020335867964,
+      "grad_norm": 1.672104835510254,
+      "learning_rate": 2.5245245245245247e-05,
+      "loss": 2.3479,
+      "step": 4958
+    },
+    {
+      "epoch": 0.0064823275202842345,
+      "grad_norm": 1.0620357990264893,
+      "learning_rate": 2.5240240240240244e-05,
+      "loss": 0.902,
+      "step": 4959
+    },
+    {
+      "epoch": 0.006483634704700505,
+      "grad_norm": 1.3465262651443481,
+      "learning_rate": 2.523523523523524e-05,
+      "loss": 1.313,
+      "step": 4960
+    },
+    {
+      "epoch": 0.006484941889116774,
+      "grad_norm": 1.746870994567871,
+      "learning_rate": 2.5230230230230233e-05,
+      "loss": 0.9014,
+      "step": 4961
+    },
+    {
+      "epoch": 0.006486249073533045,
+      "grad_norm": 1.2340298891067505,
+      "learning_rate": 2.5225225225225222e-05,
+      "loss": 1.2933,
+      "step": 4962
+    },
+    {
+      "epoch": 0.006487556257949315,
+      "grad_norm": 1.265334963798523,
+      "learning_rate": 2.522022022022022e-05,
+      "loss": 1.1322,
+      "step": 4963
+    },
+    {
+      "epoch": 0.006488863442365585,
+      "grad_norm": 1.4624158143997192,
+      "learning_rate": 2.5215215215215215e-05,
+      "loss": 1.9008,
+      "step": 4964
+    },
+    {
+      "epoch": 0.0064901706267818556,
+      "grad_norm": NaN,
+      "learning_rate": 2.521021021021021e-05,
+      "loss": 1.7095,
+      "step": 4965
+    },
+    {
+      "epoch": 0.006491477811198126,
+      "grad_norm": 1.2724614143371582,
+      "learning_rate": 2.521021021021021e-05,
+      "loss": 1.6443,
+      "step": 4966
+    },
+    {
+      "epoch": 0.006492784995614396,
+      "grad_norm": 1.1773241758346558,
+      "learning_rate": 2.5205205205205208e-05,
+      "loss": 1.8938,
+      "step": 4967
+    },
+    {
+      "epoch": 0.0064940921800306665,
+      "grad_norm": 1.0180754661560059,
+      "learning_rate": 2.52002002002002e-05,
+      "loss": 1.0012,
+      "step": 4968
+    },
+    {
+      "epoch": 0.006495399364446937,
+      "grad_norm": 1.4138820171356201,
+      "learning_rate": 2.5195195195195197e-05,
+      "loss": 1.243,
+      "step": 4969
+    },
+    {
+      "epoch": 0.006496706548863207,
+      "grad_norm": 1.5600658655166626,
+      "learning_rate": 2.5190190190190194e-05,
+      "loss": 0.4998,
+      "step": 4970
+    },
+    {
+      "epoch": 0.0064980137332794775,
+      "grad_norm": 1.2568681240081787,
+      "learning_rate": 2.5185185185185183e-05,
+      "loss": 1.3767,
+      "step": 4971
+    },
+    {
+      "epoch": 0.006499320917695748,
+      "grad_norm": 1.1647648811340332,
+      "learning_rate": 2.518018018018018e-05,
+      "loss": 1.5304,
+      "step": 4972
+    },
+    {
+      "epoch": 0.006500628102112018,
+      "grad_norm": 1.4320952892303467,
+      "learning_rate": 2.5175175175175176e-05,
+      "loss": 1.781,
+      "step": 4973
+    },
+    {
+      "epoch": 0.006501935286528288,
+      "grad_norm": 1.5594635009765625,
+      "learning_rate": 2.5170170170170172e-05,
+      "loss": 1.8533,
+      "step": 4974
+    },
+    {
+      "epoch": 0.006503242470944559,
+      "grad_norm": 0.7217073440551758,
+      "learning_rate": 2.5165165165165165e-05,
+      "loss": 0.4733,
+      "step": 4975
+    },
+    {
+      "epoch": 0.006504549655360829,
+      "grad_norm": 1.2945690155029297,
+      "learning_rate": 2.516016016016016e-05,
+      "loss": 1.0965,
+      "step": 4976
+    },
+    {
+      "epoch": 0.0065058568397770985,
+      "grad_norm": 1.7763279676437378,
+      "learning_rate": 2.5155155155155158e-05,
+      "loss": 1.6898,
+      "step": 4977
+    },
+    {
+      "epoch": 0.006507164024193369,
+      "grad_norm": 0.8919986486434937,
+      "learning_rate": 2.5150150150150154e-05,
+      "loss": 0.4967,
+      "step": 4978
+    },
+    {
+      "epoch": 0.006508471208609639,
+      "grad_norm": 1.4215621948242188,
+      "learning_rate": 2.5145145145145144e-05,
+      "loss": 1.4244,
+      "step": 4979
+    },
+    {
+      "epoch": 0.0065097783930259095,
+      "grad_norm": 1.0939370393753052,
+      "learning_rate": 2.514014014014014e-05,
+      "loss": 0.8388,
+      "step": 4980
+    },
+    {
+      "epoch": 0.00651108557744218,
+      "grad_norm": 1.5513668060302734,
+      "learning_rate": 2.5135135135135133e-05,
+      "loss": 1.7782,
+      "step": 4981
+    },
+    {
+      "epoch": 0.00651239276185845,
+      "grad_norm": 1.142903447151184,
+      "learning_rate": 2.513013013013013e-05,
+      "loss": 1.1686,
+      "step": 4982
+    },
+    {
+      "epoch": 0.0065136999462747204,
+      "grad_norm": 1.2259658575057983,
+      "learning_rate": 2.5125125125125126e-05,
+      "loss": 1.8008,
+      "step": 4983
+    },
+    {
+      "epoch": 0.006515007130690991,
+      "grad_norm": 1.2704530954360962,
+      "learning_rate": 2.5120120120120122e-05,
+      "loss": 0.8705,
+      "step": 4984
+    },
+    {
+      "epoch": 0.006516314315107261,
+      "grad_norm": 1.5700956583023071,
+      "learning_rate": 2.511511511511512e-05,
+      "loss": 1.9146,
+      "step": 4985
+    },
+    {
+      "epoch": 0.006517621499523531,
+      "grad_norm": 1.3781633377075195,
+      "learning_rate": 2.5110110110110112e-05,
+      "loss": 1.8104,
+      "step": 4986
+    },
+    {
+      "epoch": 0.006518928683939802,
+      "grad_norm": 0.9846285581588745,
+      "learning_rate": 2.5105105105105108e-05,
+      "loss": 0.959,
+      "step": 4987
+    },
+    {
+      "epoch": 0.006520235868356072,
+      "grad_norm": 1.3729844093322754,
+      "learning_rate": 2.5100100100100098e-05,
+      "loss": 2.2913,
+      "step": 4988
+    },
+    {
+      "epoch": 0.006521543052772342,
+      "grad_norm": 1.2394644021987915,
+      "learning_rate": 2.5095095095095094e-05,
+      "loss": 1.2796,
+      "step": 4989
+    },
+    {
+      "epoch": 0.006522850237188613,
+      "grad_norm": 1.0035998821258545,
+      "learning_rate": 2.509009009009009e-05,
+      "loss": 0.5478,
+      "step": 4990
+    },
+    {
+      "epoch": 0.006524157421604883,
+      "grad_norm": 1.3576663732528687,
+      "learning_rate": 2.5085085085085087e-05,
+      "loss": 1.5293,
+      "step": 4991
+    },
+    {
+      "epoch": 0.0065254646060211524,
+      "grad_norm": 1.475887656211853,
+      "learning_rate": 2.5080080080080083e-05,
+      "loss": 1.9641,
+      "step": 4992
+    },
+    {
+      "epoch": 0.006526771790437423,
+      "grad_norm": 1.615504264831543,
+      "learning_rate": 2.5075075075075076e-05,
+      "loss": 1.3756,
+      "step": 4993
+    },
+    {
+      "epoch": 0.006528078974853693,
+      "grad_norm": 1.7244857549667358,
+      "learning_rate": 2.5070070070070072e-05,
+      "loss": 1.6952,
+      "step": 4994
+    },
+    {
+      "epoch": 0.006529386159269963,
+      "grad_norm": 1.358705997467041,
+      "learning_rate": 2.506506506506507e-05,
+      "loss": 1.14,
+      "step": 4995
+    },
+    {
+      "epoch": 0.006530693343686234,
+      "grad_norm": 1.355635643005371,
+      "learning_rate": 2.506006006006006e-05,
+      "loss": 2.2007,
+      "step": 4996
+    },
+    {
+      "epoch": 0.006532000528102504,
+      "grad_norm": 1.8061715364456177,
+      "learning_rate": 2.5055055055055055e-05,
+      "loss": 1.2522,
+      "step": 4997
+    },
+    {
+      "epoch": 0.006533307712518774,
+      "grad_norm": 1.2004097700119019,
+      "learning_rate": 2.505005005005005e-05,
+      "loss": 1.237,
+      "step": 4998
+    },
+    {
+      "epoch": 0.006534614896935045,
+      "grad_norm": 1.231264591217041,
+      "learning_rate": 2.5045045045045044e-05,
+      "loss": 1.3373,
+      "step": 4999
+    },
+    {
+      "epoch": 0.006535922081351315,
+      "grad_norm": 1.4087964296340942,
+      "learning_rate": 2.504004004004004e-05,
+      "loss": 1.9755,
+      "step": 5000
+    },
+    {
+      "epoch": 0.006537229265767585,
+      "grad_norm": 1.475189208984375,
+      "learning_rate": 2.5035035035035037e-05,
+      "loss": 2.0295,
+      "step": 5001
+    },
+    {
+      "epoch": 0.006538536450183856,
+      "grad_norm": 1.3248839378356934,
+      "learning_rate": 2.5030030030030033e-05,
+      "loss": 0.9695,
+      "step": 5002
+    },
+    {
+      "epoch": 0.006539843634600126,
+      "grad_norm": 1.342484951019287,
+      "learning_rate": 2.502502502502503e-05,
+      "loss": 1.1201,
+      "step": 5003
+    },
+    {
+      "epoch": 0.006541150819016396,
+      "grad_norm": 1.1737421751022339,
+      "learning_rate": 2.502002002002002e-05,
+      "loss": 0.908,
+      "step": 5004
+    },
+    {
+      "epoch": 0.006542458003432667,
+      "grad_norm": 1.5990629196166992,
+      "learning_rate": 2.5015015015015015e-05,
+      "loss": 1.8102,
+      "step": 5005
+    },
+    {
+      "epoch": 0.006543765187848937,
+      "grad_norm": 1.4940165281295776,
+      "learning_rate": 2.501001001001001e-05,
+      "loss": 2.1764,
+      "step": 5006
+    },
+    {
+      "epoch": 0.006545072372265206,
+      "grad_norm": 1.5747052431106567,
+      "learning_rate": 2.5005005005005005e-05,
+      "loss": 1.3669,
+      "step": 5007
+    },
+    {
+      "epoch": 0.006546379556681477,
+      "grad_norm": 1.2544509172439575,
+      "learning_rate": 2.5e-05,
+      "loss": 2.0216,
+      "step": 5008
+    },
+    {
+      "epoch": 0.006547686741097747,
+      "grad_norm": 1.444571614265442,
+      "learning_rate": 2.4994994994994998e-05,
+      "loss": 1.3421,
+      "step": 5009
+    },
+    {
+      "epoch": 0.006548993925514017,
+      "grad_norm": 1.832253336906433,
+      "learning_rate": 2.4989989989989994e-05,
+      "loss": 1.8445,
+      "step": 5010
+    },
+    {
+      "epoch": 0.006550301109930288,
+      "grad_norm": 1.6326618194580078,
+      "learning_rate": 2.4984984984984987e-05,
+      "loss": 2.0725,
+      "step": 5011
+    },
+    {
+      "epoch": 0.006551608294346558,
+      "grad_norm": 1.2522037029266357,
+      "learning_rate": 2.497997997997998e-05,
+      "loss": 1.3564,
+      "step": 5012
+    },
+    {
+      "epoch": 0.006552915478762828,
+      "grad_norm": 1.241577386856079,
+      "learning_rate": 2.4974974974974976e-05,
+      "loss": 1.1683,
+      "step": 5013
+    },
+    {
+      "epoch": 0.006554222663179099,
+      "grad_norm": 1.422757863998413,
+      "learning_rate": 2.4969969969969973e-05,
+      "loss": 1.9089,
+      "step": 5014
+    },
+    {
+      "epoch": 0.006555529847595369,
+      "grad_norm": 1.174750566482544,
+      "learning_rate": 2.4964964964964966e-05,
+      "loss": 1.392,
+      "step": 5015
+    },
+    {
+      "epoch": 0.006556837032011639,
+      "grad_norm": 1.2710038423538208,
+      "learning_rate": 2.4959959959959962e-05,
+      "loss": 1.7177,
+      "step": 5016
+    },
+    {
+      "epoch": 0.00655814421642791,
+      "grad_norm": 1.2263295650482178,
+      "learning_rate": 2.4954954954954958e-05,
+      "loss": 0.9955,
+      "step": 5017
+    },
+    {
+      "epoch": 0.00655945140084418,
+      "grad_norm": 1.4414699077606201,
+      "learning_rate": 2.494994994994995e-05,
+      "loss": 1.8978,
+      "step": 5018
+    },
+    {
+      "epoch": 0.00656075858526045,
+      "grad_norm": 1.666784644126892,
+      "learning_rate": 2.4944944944944944e-05,
+      "loss": 1.7701,
+      "step": 5019
+    },
+    {
+      "epoch": 0.0065620657696767205,
+      "grad_norm": 0.9503254890441895,
+      "learning_rate": 2.493993993993994e-05,
+      "loss": 0.7524,
+      "step": 5020
+    },
+    {
+      "epoch": 0.006563372954092991,
+      "grad_norm": 1.2896714210510254,
+      "learning_rate": 2.4934934934934937e-05,
+      "loss": 1.9921,
+      "step": 5021
+    },
+    {
+      "epoch": 0.006564680138509261,
+      "grad_norm": 1.671697735786438,
+      "learning_rate": 2.4929929929929933e-05,
+      "loss": 1.6243,
+      "step": 5022
+    },
+    {
+      "epoch": 0.006565987322925531,
+      "grad_norm": 1.1275125741958618,
+      "learning_rate": 2.4924924924924926e-05,
+      "loss": 1.0101,
+      "step": 5023
+    },
+    {
+      "epoch": 0.006567294507341801,
+      "grad_norm": 0.9089884757995605,
+      "learning_rate": 2.491991991991992e-05,
+      "loss": 0.833,
+      "step": 5024
+    },
+    {
+      "epoch": 0.006568601691758071,
+      "grad_norm": 1.2082183361053467,
+      "learning_rate": 2.4914914914914916e-05,
+      "loss": 1.1387,
+      "step": 5025
+    },
+    {
+      "epoch": 0.006569908876174342,
+      "grad_norm": 1.5970537662506104,
+      "learning_rate": 2.4909909909909912e-05,
+      "loss": 1.5774,
+      "step": 5026
+    },
+    {
+      "epoch": 0.006571216060590612,
+      "grad_norm": 1.2484818696975708,
+      "learning_rate": 2.4904904904904905e-05,
+      "loss": 1.8442,
+      "step": 5027
+    },
+    {
+      "epoch": 0.006572523245006882,
+      "grad_norm": 1.304917812347412,
+      "learning_rate": 2.48998998998999e-05,
+      "loss": 1.5965,
+      "step": 5028
+    },
+    {
+      "epoch": 0.0065738304294231525,
+      "grad_norm": 1.471189260482788,
+      "learning_rate": 2.4894894894894898e-05,
+      "loss": 1.7932,
+      "step": 5029
+    },
+    {
+      "epoch": 0.006575137613839423,
+      "grad_norm": 1.0791022777557373,
+      "learning_rate": 2.488988988988989e-05,
+      "loss": 0.993,
+      "step": 5030
+    },
+    {
+      "epoch": 0.006576444798255693,
+      "grad_norm": 1.3162100315093994,
+      "learning_rate": 2.4884884884884884e-05,
+      "loss": 1.112,
+      "step": 5031
+    },
+    {
+      "epoch": 0.0065777519826719635,
+      "grad_norm": 1.6941759586334229,
+      "learning_rate": 2.487987987987988e-05,
+      "loss": 1.8479,
+      "step": 5032
+    },
+    {
+      "epoch": 0.006579059167088234,
+      "grad_norm": 1.3153090476989746,
+      "learning_rate": 2.4874874874874876e-05,
+      "loss": 0.8602,
+      "step": 5033
+    },
+    {
+      "epoch": 0.006580366351504504,
+      "grad_norm": 1.450746774673462,
+      "learning_rate": 2.4869869869869873e-05,
+      "loss": 1.7963,
+      "step": 5034
+    },
+    {
+      "epoch": 0.0065816735359207745,
+      "grad_norm": 1.4164656400680542,
+      "learning_rate": 2.486486486486487e-05,
+      "loss": 1.3487,
+      "step": 5035
+    },
+    {
+      "epoch": 0.006582980720337045,
+      "grad_norm": 1.9562982320785522,
+      "learning_rate": 2.485985985985986e-05,
+      "loss": 1.6999,
+      "step": 5036
+    },
+    {
+      "epoch": 0.006584287904753315,
+      "grad_norm": 1.4145323038101196,
+      "learning_rate": 2.4854854854854855e-05,
+      "loss": 1.1745,
+      "step": 5037
+    },
+    {
+      "epoch": 0.0065855950891695846,
+      "grad_norm": 1.483902096748352,
+      "learning_rate": 2.484984984984985e-05,
+      "loss": 1.9891,
+      "step": 5038
+    },
+    {
+      "epoch": 0.006586902273585855,
+      "grad_norm": 1.603550672531128,
+      "learning_rate": 2.4844844844844848e-05,
+      "loss": 1.0221,
+      "step": 5039
+    },
+    {
+      "epoch": 0.006588209458002125,
+      "grad_norm": 1.2919052839279175,
+      "learning_rate": 2.483983983983984e-05,
+      "loss": 1.9601,
+      "step": 5040
+    },
+    {
+      "epoch": 0.0065895166424183955,
+      "grad_norm": 1.522194504737854,
+      "learning_rate": 2.4834834834834837e-05,
+      "loss": 1.9809,
+      "step": 5041
+    },
+    {
+      "epoch": 0.006590823826834666,
+      "grad_norm": 1.428803563117981,
+      "learning_rate": 2.482982982982983e-05,
+      "loss": 1.2295,
+      "step": 5042
+    },
+    {
+      "epoch": 0.006592131011250936,
+      "grad_norm": 1.3971636295318604,
+      "learning_rate": 2.4824824824824826e-05,
+      "loss": 1.3907,
+      "step": 5043
+    },
+    {
+      "epoch": 0.0065934381956672065,
+      "grad_norm": 1.3137544393539429,
+      "learning_rate": 2.481981981981982e-05,
+      "loss": 1.2308,
+      "step": 5044
+    },
+    {
+      "epoch": 0.006594745380083477,
+      "grad_norm": 1.5255560874938965,
+      "learning_rate": 2.4814814814814816e-05,
+      "loss": 1.9137,
+      "step": 5045
+    },
+    {
+      "epoch": 0.006596052564499747,
+      "grad_norm": 1.3475407361984253,
+      "learning_rate": 2.4809809809809812e-05,
+      "loss": 1.823,
+      "step": 5046
+    },
+    {
+      "epoch": 0.006597359748916017,
+      "grad_norm": 1.286723256111145,
+      "learning_rate": 2.480480480480481e-05,
+      "loss": 2.0667,
+      "step": 5047
+    },
+    {
+      "epoch": 0.006598666933332288,
+      "grad_norm": 0.6623641848564148,
+      "learning_rate": 2.47997997997998e-05,
+      "loss": 0.3608,
+      "step": 5048
+    },
+    {
+      "epoch": 0.006599974117748558,
+      "grad_norm": 1.6288570165634155,
+      "learning_rate": 2.4794794794794794e-05,
+      "loss": 1.9467,
+      "step": 5049
+    },
+    {
+      "epoch": 0.006601281302164828,
+      "grad_norm": 1.5388872623443604,
+      "learning_rate": 2.478978978978979e-05,
+      "loss": 2.1304,
+      "step": 5050
+    },
+    {
+      "epoch": 0.006602588486581099,
+      "grad_norm": 1.5069769620895386,
+      "learning_rate": 2.4784784784784787e-05,
+      "loss": 1.3846,
+      "step": 5051
+    },
+    {
+      "epoch": 0.006603895670997369,
+      "grad_norm": 1.654964566230774,
+      "learning_rate": 2.477977977977978e-05,
+      "loss": 2.0563,
+      "step": 5052
+    },
+    {
+      "epoch": 0.0066052028554136385,
+      "grad_norm": 1.4049372673034668,
+      "learning_rate": 2.4774774774774777e-05,
+      "loss": 2.1372,
+      "step": 5053
+    },
+    {
+      "epoch": 0.006606510039829909,
+      "grad_norm": 1.0520540475845337,
+      "learning_rate": 2.476976976976977e-05,
+      "loss": 1.5033,
+      "step": 5054
+    },
+    {
+      "epoch": 0.006607817224246179,
+      "grad_norm": 1.339145541191101,
+      "learning_rate": 2.4764764764764766e-05,
+      "loss": 2.0102,
+      "step": 5055
+    },
+    {
+      "epoch": 0.0066091244086624494,
+      "grad_norm": 1.092621922492981,
+      "learning_rate": 2.475975975975976e-05,
+      "loss": 1.2941,
+      "step": 5056
+    },
+    {
+      "epoch": 0.00661043159307872,
+      "grad_norm": 1.20045804977417,
+      "learning_rate": 2.4754754754754755e-05,
+      "loss": 1.1276,
+      "step": 5057
+    },
+    {
+      "epoch": 0.00661173877749499,
+      "grad_norm": 1.1402517557144165,
+      "learning_rate": 2.474974974974975e-05,
+      "loss": 1.2156,
+      "step": 5058
+    },
+    {
+      "epoch": 0.00661304596191126,
+      "grad_norm": 1.3097063302993774,
+      "learning_rate": 2.4744744744744748e-05,
+      "loss": 1.1777,
+      "step": 5059
+    },
+    {
+      "epoch": 0.006614353146327531,
+      "grad_norm": 1.542409896850586,
+      "learning_rate": 2.473973973973974e-05,
+      "loss": 2.0511,
+      "step": 5060
+    },
+    {
+      "epoch": 0.006615660330743801,
+      "grad_norm": 1.1858668327331543,
+      "learning_rate": 2.4734734734734734e-05,
+      "loss": 1.4308,
+      "step": 5061
+    },
+    {
+      "epoch": 0.006616967515160071,
+      "grad_norm": 1.5697602033615112,
+      "learning_rate": 2.472972972972973e-05,
+      "loss": 2.1216,
+      "step": 5062
+    },
+    {
+      "epoch": 0.006618274699576342,
+      "grad_norm": 1.562796711921692,
+      "learning_rate": 2.4724724724724727e-05,
+      "loss": 1.6165,
+      "step": 5063
+    },
+    {
+      "epoch": 0.006619581883992612,
+      "grad_norm": 1.6844980716705322,
+      "learning_rate": 2.4719719719719723e-05,
+      "loss": 1.495,
+      "step": 5064
+    },
+    {
+      "epoch": 0.006620889068408882,
+      "grad_norm": 1.525625228881836,
+      "learning_rate": 2.4714714714714716e-05,
+      "loss": 1.8215,
+      "step": 5065
+    },
+    {
+      "epoch": 0.006622196252825153,
+      "grad_norm": 1.1809179782867432,
+      "learning_rate": 2.4709709709709712e-05,
+      "loss": 0.985,
+      "step": 5066
+    },
+    {
+      "epoch": 0.006623503437241423,
+      "grad_norm": 1.042462706565857,
+      "learning_rate": 2.4704704704704705e-05,
+      "loss": 0.7159,
+      "step": 5067
+    },
+    {
+      "epoch": 0.006624810621657693,
+      "grad_norm": 1.4611902236938477,
+      "learning_rate": 2.46996996996997e-05,
+      "loss": 1.2362,
+      "step": 5068
+    },
+    {
+      "epoch": 0.006626117806073963,
+      "grad_norm": 1.5762981176376343,
+      "learning_rate": 2.4694694694694695e-05,
+      "loss": 2.1567,
+      "step": 5069
+    },
+    {
+      "epoch": 0.006627424990490233,
+      "grad_norm": 1.4686235189437866,
+      "learning_rate": 2.468968968968969e-05,
+      "loss": 1.6726,
+      "step": 5070
+    },
+    {
+      "epoch": 0.006628732174906503,
+      "grad_norm": 2.0353047847747803,
+      "learning_rate": 2.4684684684684687e-05,
+      "loss": 1.6787,
+      "step": 5071
+    },
+    {
+      "epoch": 0.006630039359322774,
+      "grad_norm": 1.57929527759552,
+      "learning_rate": 2.467967967967968e-05,
+      "loss": 1.4602,
+      "step": 5072
+    },
+    {
+      "epoch": 0.006631346543739044,
+      "grad_norm": 1.5543593168258667,
+      "learning_rate": 2.4674674674674673e-05,
+      "loss": 1.5885,
+      "step": 5073
+    },
+    {
+      "epoch": 0.006632653728155314,
+      "grad_norm": 1.700930118560791,
+      "learning_rate": 2.466966966966967e-05,
+      "loss": 1.8008,
+      "step": 5074
+    },
+    {
+      "epoch": 0.006633960912571585,
+      "grad_norm": 1.712970495223999,
+      "learning_rate": 2.4664664664664666e-05,
+      "loss": 1.3838,
+      "step": 5075
+    },
+    {
+      "epoch": 0.006635268096987855,
+      "grad_norm": 1.1867934465408325,
+      "learning_rate": 2.4659659659659662e-05,
+      "loss": 1.2664,
+      "step": 5076
+    },
+    {
+      "epoch": 0.006636575281404125,
+      "grad_norm": 1.5568827390670776,
+      "learning_rate": 2.4654654654654655e-05,
+      "loss": 1.434,
+      "step": 5077
+    },
+    {
+      "epoch": 0.006637882465820396,
+      "grad_norm": 1.5676674842834473,
+      "learning_rate": 2.464964964964965e-05,
+      "loss": 1.2699,
+      "step": 5078
+    },
+    {
+      "epoch": 0.006639189650236666,
+      "grad_norm": 0.41681233048439026,
+      "learning_rate": 2.4644644644644645e-05,
+      "loss": 0.1206,
+      "step": 5079
+    },
+    {
+      "epoch": 0.006640496834652936,
+      "grad_norm": 1.2142876386642456,
+      "learning_rate": 2.463963963963964e-05,
+      "loss": 0.7038,
+      "step": 5080
+    },
+    {
+      "epoch": 0.006641804019069207,
+      "grad_norm": 1.270497441291809,
+      "learning_rate": 2.4634634634634634e-05,
+      "loss": 1.4523,
+      "step": 5081
+    },
+    {
+      "epoch": 0.006643111203485477,
+      "grad_norm": 1.4971543550491333,
+      "learning_rate": 2.462962962962963e-05,
+      "loss": 1.1888,
+      "step": 5082
+    },
+    {
+      "epoch": 0.006644418387901747,
+      "grad_norm": 1.293985366821289,
+      "learning_rate": 2.4624624624624627e-05,
+      "loss": 1.2956,
+      "step": 5083
+    },
+    {
+      "epoch": 0.006645725572318017,
+      "grad_norm": 1.5881441831588745,
+      "learning_rate": 2.4619619619619623e-05,
+      "loss": 1.9275,
+      "step": 5084
+    },
+    {
+      "epoch": 0.006647032756734287,
+      "grad_norm": 1.3860828876495361,
+      "learning_rate": 2.4614614614614616e-05,
+      "loss": 1.8438,
+      "step": 5085
+    },
+    {
+      "epoch": 0.006648339941150557,
+      "grad_norm": 1.360149621963501,
+      "learning_rate": 2.460960960960961e-05,
+      "loss": 2.1291,
+      "step": 5086
+    },
+    {
+      "epoch": 0.006649647125566828,
+      "grad_norm": 1.1819478273391724,
+      "learning_rate": 2.4604604604604605e-05,
+      "loss": 1.6203,
+      "step": 5087
+    },
+    {
+      "epoch": 0.006650954309983098,
+      "grad_norm": 1.1349557638168335,
+      "learning_rate": 2.4599599599599602e-05,
+      "loss": 1.3898,
+      "step": 5088
+    },
+    {
+      "epoch": 0.006652261494399368,
+      "grad_norm": 0.636745035648346,
+      "learning_rate": 2.4594594594594598e-05,
+      "loss": 0.2116,
+      "step": 5089
+    },
+    {
+      "epoch": 0.006653568678815639,
+      "grad_norm": 1.1753062009811401,
+      "learning_rate": 2.458958958958959e-05,
+      "loss": 1.3629,
+      "step": 5090
+    },
+    {
+      "epoch": 0.006654875863231909,
+      "grad_norm": 1.6626254320144653,
+      "learning_rate": 2.4584584584584584e-05,
+      "loss": 1.9128,
+      "step": 5091
+    },
+    {
+      "epoch": 0.006656183047648179,
+      "grad_norm": 1.3711360692977905,
+      "learning_rate": 2.457957957957958e-05,
+      "loss": 2.0705,
+      "step": 5092
+    },
+    {
+      "epoch": 0.0066574902320644495,
+      "grad_norm": 1.469247817993164,
+      "learning_rate": 2.4574574574574577e-05,
+      "loss": 1.62,
+      "step": 5093
+    },
+    {
+      "epoch": 0.00665879741648072,
+      "grad_norm": 1.2287954092025757,
+      "learning_rate": 2.456956956956957e-05,
+      "loss": 0.8911,
+      "step": 5094
+    },
+    {
+      "epoch": 0.00666010460089699,
+      "grad_norm": 1.504233956336975,
+      "learning_rate": 2.4564564564564566e-05,
+      "loss": 1.7041,
+      "step": 5095
+    },
+    {
+      "epoch": 0.0066614117853132605,
+      "grad_norm": 1.5229817628860474,
+      "learning_rate": 2.4559559559559562e-05,
+      "loss": 2.0711,
+      "step": 5096
+    },
+    {
+      "epoch": 0.006662718969729531,
+      "grad_norm": 1.098416805267334,
+      "learning_rate": 2.4554554554554555e-05,
+      "loss": 0.9679,
+      "step": 5097
+    },
+    {
+      "epoch": 0.006664026154145801,
+      "grad_norm": 1.5018385648727417,
+      "learning_rate": 2.454954954954955e-05,
+      "loss": 1.9584,
+      "step": 5098
+    },
+    {
+      "epoch": 0.006665333338562071,
+      "grad_norm": 1.1300309896469116,
+      "learning_rate": 2.4544544544544545e-05,
+      "loss": 1.2923,
+      "step": 5099
+    },
+    {
+      "epoch": 0.006666640522978341,
+      "grad_norm": 1.455424427986145,
+      "learning_rate": 2.453953953953954e-05,
+      "loss": 1.2827,
+      "step": 5100
+    },
+    {
+      "epoch": 0.006667947707394611,
+      "grad_norm": 1.7407639026641846,
+      "learning_rate": 2.4534534534534538e-05,
+      "loss": 1.5773,
+      "step": 5101
+    },
+    {
+      "epoch": 0.0066692548918108815,
+      "grad_norm": 1.4770019054412842,
+      "learning_rate": 2.452952952952953e-05,
+      "loss": 1.4785,
+      "step": 5102
+    },
+    {
+      "epoch": 0.006670562076227152,
+      "grad_norm": 1.602638840675354,
+      "learning_rate": 2.4524524524524527e-05,
+      "loss": 1.6096,
+      "step": 5103
+    },
+    {
+      "epoch": 0.006671869260643422,
+      "grad_norm": 1.342702865600586,
+      "learning_rate": 2.451951951951952e-05,
+      "loss": 1.9198,
+      "step": 5104
+    },
+    {
+      "epoch": 0.0066731764450596925,
+      "grad_norm": 1.185347080230713,
+      "learning_rate": 2.4514514514514516e-05,
+      "loss": 1.4612,
+      "step": 5105
+    },
+    {
+      "epoch": 0.006674483629475963,
+      "grad_norm": 1.6696664094924927,
+      "learning_rate": 2.450950950950951e-05,
+      "loss": 1.796,
+      "step": 5106
+    },
+    {
+      "epoch": 0.006675790813892233,
+      "grad_norm": 1.4966129064559937,
+      "learning_rate": 2.4504504504504506e-05,
+      "loss": 1.6412,
+      "step": 5107
+    },
+    {
+      "epoch": 0.0066770979983085035,
+      "grad_norm": 1.330275297164917,
+      "learning_rate": 2.4499499499499502e-05,
+      "loss": 1.8247,
+      "step": 5108
+    },
+    {
+      "epoch": 0.006678405182724774,
+      "grad_norm": 1.4164527654647827,
+      "learning_rate": 2.4494494494494495e-05,
+      "loss": 2.5233,
+      "step": 5109
+    },
+    {
+      "epoch": 0.006679712367141044,
+      "grad_norm": 1.1726564168930054,
+      "learning_rate": 2.4489489489489488e-05,
+      "loss": 0.7163,
+      "step": 5110
+    },
+    {
+      "epoch": 0.006681019551557314,
+      "grad_norm": 1.380189061164856,
+      "learning_rate": 2.4484484484484484e-05,
+      "loss": 1.5322,
+      "step": 5111
+    },
+    {
+      "epoch": 0.006682326735973585,
+      "grad_norm": 1.8298122882843018,
+      "learning_rate": 2.447947947947948e-05,
+      "loss": 1.429,
+      "step": 5112
+    },
+    {
+      "epoch": 0.006683633920389855,
+      "grad_norm": 1.3876371383666992,
+      "learning_rate": 2.4474474474474477e-05,
+      "loss": 1.7321,
+      "step": 5113
+    },
+    {
+      "epoch": 0.0066849411048061245,
+      "grad_norm": 1.3032163381576538,
+      "learning_rate": 2.446946946946947e-05,
+      "loss": 1.7335,
+      "step": 5114
+    },
+    {
+      "epoch": 0.006686248289222395,
+      "grad_norm": 1.3089029788970947,
+      "learning_rate": 2.4464464464464466e-05,
+      "loss": 2.2648,
+      "step": 5115
+    },
+    {
+      "epoch": 0.006687555473638665,
+      "grad_norm": 1.3884716033935547,
+      "learning_rate": 2.445945945945946e-05,
+      "loss": 2.0857,
+      "step": 5116
+    },
+    {
+      "epoch": 0.0066888626580549355,
+      "grad_norm": 4.880166053771973,
+      "learning_rate": 2.4454454454454456e-05,
+      "loss": 2.4471,
+      "step": 5117
+    },
+    {
+      "epoch": 0.006690169842471206,
+      "grad_norm": 1.8406802415847778,
+      "learning_rate": 2.4449449449449452e-05,
+      "loss": 1.3502,
+      "step": 5118
+    },
+    {
+      "epoch": 0.006691477026887476,
+      "grad_norm": 1.9580419063568115,
+      "learning_rate": 2.4444444444444445e-05,
+      "loss": 2.099,
+      "step": 5119
+    },
+    {
+      "epoch": 0.0066927842113037464,
+      "grad_norm": 1.3233329057693481,
+      "learning_rate": 2.443943943943944e-05,
+      "loss": 2.0223,
+      "step": 5120
+    },
+    {
+      "epoch": 0.006694091395720017,
+      "grad_norm": 1.6957024335861206,
+      "learning_rate": 2.4434434434434438e-05,
+      "loss": 1.5188,
+      "step": 5121
+    },
+    {
+      "epoch": 0.006695398580136287,
+      "grad_norm": 1.67441987991333,
+      "learning_rate": 2.442942942942943e-05,
+      "loss": 1.993,
+      "step": 5122
+    },
+    {
+      "epoch": 0.006696705764552557,
+      "grad_norm": 1.4810354709625244,
+      "learning_rate": 2.4424424424424424e-05,
+      "loss": 1.3279,
+      "step": 5123
+    },
+    {
+      "epoch": 0.006698012948968828,
+      "grad_norm": 1.278114914894104,
+      "learning_rate": 2.441941941941942e-05,
+      "loss": 1.7005,
+      "step": 5124
+    },
+    {
+      "epoch": 0.006699320133385098,
+      "grad_norm": 1.5316447019577026,
+      "learning_rate": 2.4414414414414416e-05,
+      "loss": 2.0416,
+      "step": 5125
+    },
+    {
+      "epoch": 0.006700627317801368,
+      "grad_norm": 1.6597137451171875,
+      "learning_rate": 2.4409409409409413e-05,
+      "loss": 1.8956,
+      "step": 5126
+    },
+    {
+      "epoch": 0.006701934502217639,
+      "grad_norm": 1.1998282670974731,
+      "learning_rate": 2.4404404404404406e-05,
+      "loss": 1.832,
+      "step": 5127
+    },
+    {
+      "epoch": 0.006703241686633909,
+      "grad_norm": 1.3698290586471558,
+      "learning_rate": 2.43993993993994e-05,
+      "loss": 1.3752,
+      "step": 5128
+    },
+    {
+      "epoch": 0.006704548871050179,
+      "grad_norm": 2.061058759689331,
+      "learning_rate": 2.4394394394394395e-05,
+      "loss": 1.4608,
+      "step": 5129
+    },
+    {
+      "epoch": 0.006705856055466449,
+      "grad_norm": 1.6419496536254883,
+      "learning_rate": 2.438938938938939e-05,
+      "loss": 1.7256,
+      "step": 5130
+    },
+    {
+      "epoch": 0.006707163239882719,
+      "grad_norm": 1.5738807916641235,
+      "learning_rate": 2.4384384384384384e-05,
+      "loss": 2.1804,
+      "step": 5131
+    },
+    {
+      "epoch": 0.006708470424298989,
+      "grad_norm": 1.8309868574142456,
+      "learning_rate": 2.437937937937938e-05,
+      "loss": 2.2131,
+      "step": 5132
+    },
+    {
+      "epoch": 0.00670977760871526,
+      "grad_norm": 1.867318034172058,
+      "learning_rate": 2.4374374374374377e-05,
+      "loss": 0.8591,
+      "step": 5133
+    },
+    {
+      "epoch": 0.00671108479313153,
+      "grad_norm": 1.4604066610336304,
+      "learning_rate": 2.436936936936937e-05,
+      "loss": 1.7765,
+      "step": 5134
+    },
+    {
+      "epoch": 0.0067123919775478,
+      "grad_norm": 1.4297016859054565,
+      "learning_rate": 2.4364364364364363e-05,
+      "loss": 1.468,
+      "step": 5135
+    },
+    {
+      "epoch": 0.006713699161964071,
+      "grad_norm": 1.8569707870483398,
+      "learning_rate": 2.435935935935936e-05,
+      "loss": 1.3892,
+      "step": 5136
+    },
+    {
+      "epoch": 0.006715006346380341,
+      "grad_norm": 1.2977726459503174,
+      "learning_rate": 2.4354354354354356e-05,
+      "loss": 1.6461,
+      "step": 5137
+    },
+    {
+      "epoch": 0.006716313530796611,
+      "grad_norm": 1.305626630783081,
+      "learning_rate": 2.4349349349349352e-05,
+      "loss": 1.3621,
+      "step": 5138
+    },
+    {
+      "epoch": 0.006717620715212882,
+      "grad_norm": 0.8450940847396851,
+      "learning_rate": 2.4344344344344345e-05,
+      "loss": 0.26,
+      "step": 5139
+    },
+    {
+      "epoch": 0.006718927899629152,
+      "grad_norm": 0.980656623840332,
+      "learning_rate": 2.433933933933934e-05,
+      "loss": 0.6029,
+      "step": 5140
+    },
+    {
+      "epoch": 0.006720235084045422,
+      "grad_norm": 1.511507272720337,
+      "learning_rate": 2.4334334334334334e-05,
+      "loss": 1.8077,
+      "step": 5141
+    },
+    {
+      "epoch": 0.006721542268461693,
+      "grad_norm": 1.2866464853286743,
+      "learning_rate": 2.432932932932933e-05,
+      "loss": 0.9525,
+      "step": 5142
+    },
+    {
+      "epoch": 0.006722849452877963,
+      "grad_norm": 1.3022022247314453,
+      "learning_rate": 2.4324324324324327e-05,
+      "loss": 1.7384,
+      "step": 5143
+    },
+    {
+      "epoch": 0.006724156637294233,
+      "grad_norm": 1.2001067399978638,
+      "learning_rate": 2.431931931931932e-05,
+      "loss": 1.0443,
+      "step": 5144
+    },
+    {
+      "epoch": 0.006725463821710503,
+      "grad_norm": 1.3391985893249512,
+      "learning_rate": 2.4314314314314316e-05,
+      "loss": 1.7276,
+      "step": 5145
+    },
+    {
+      "epoch": 0.006726771006126773,
+      "grad_norm": 1.224219799041748,
+      "learning_rate": 2.430930930930931e-05,
+      "loss": 1.3168,
+      "step": 5146
+    },
+    {
+      "epoch": 0.006728078190543043,
+      "grad_norm": 1.505295753479004,
+      "learning_rate": 2.4304304304304306e-05,
+      "loss": 1.9178,
+      "step": 5147
+    },
+    {
+      "epoch": 0.006729385374959314,
+      "grad_norm": 1.2445213794708252,
+      "learning_rate": 2.42992992992993e-05,
+      "loss": 0.7848,
+      "step": 5148
+    },
+    {
+      "epoch": 0.006730692559375584,
+      "grad_norm": 1.7164814472198486,
+      "learning_rate": 2.4294294294294295e-05,
+      "loss": 1.7821,
+      "step": 5149
+    },
+    {
+      "epoch": 0.006731999743791854,
+      "grad_norm": 1.3765486478805542,
+      "learning_rate": 2.428928928928929e-05,
+      "loss": 1.4387,
+      "step": 5150
+    },
+    {
+      "epoch": 0.006733306928208125,
+      "grad_norm": 1.3376928567886353,
+      "learning_rate": 2.4284284284284288e-05,
+      "loss": 1.0995,
+      "step": 5151
+    },
+    {
+      "epoch": 0.006734614112624395,
+      "grad_norm": 1.464873194694519,
+      "learning_rate": 2.427927927927928e-05,
+      "loss": 2.3517,
+      "step": 5152
+    },
+    {
+      "epoch": 0.006735921297040665,
+      "grad_norm": 1.0756274461746216,
+      "learning_rate": 2.4274274274274274e-05,
+      "loss": 0.8097,
+      "step": 5153
+    },
+    {
+      "epoch": 0.006737228481456936,
+      "grad_norm": 1.6639760732650757,
+      "learning_rate": 2.426926926926927e-05,
+      "loss": 2.1544,
+      "step": 5154
+    },
+    {
+      "epoch": 0.006738535665873206,
+      "grad_norm": 1.5416715145111084,
+      "learning_rate": 2.4264264264264267e-05,
+      "loss": 1.4057,
+      "step": 5155
+    },
+    {
+      "epoch": 0.006739842850289476,
+      "grad_norm": 1.0215994119644165,
+      "learning_rate": 2.425925925925926e-05,
+      "loss": 0.8767,
+      "step": 5156
+    },
+    {
+      "epoch": 0.0067411500347057465,
+      "grad_norm": 1.477290391921997,
+      "learning_rate": 2.4254254254254256e-05,
+      "loss": 0.6737,
+      "step": 5157
+    },
+    {
+      "epoch": 0.006742457219122017,
+      "grad_norm": 2.3970861434936523,
+      "learning_rate": 2.4249249249249252e-05,
+      "loss": 0.9407,
+      "step": 5158
+    },
+    {
+      "epoch": 0.006743764403538287,
+      "grad_norm": 1.318664312362671,
+      "learning_rate": 2.4244244244244245e-05,
+      "loss": 1.3682,
+      "step": 5159
+    },
+    {
+      "epoch": 0.006745071587954557,
+      "grad_norm": 0.9410414099693298,
+      "learning_rate": 2.4239239239239238e-05,
+      "loss": 0.5658,
+      "step": 5160
+    },
+    {
+      "epoch": 0.006746378772370827,
+      "grad_norm": 1.8351000547409058,
+      "learning_rate": 2.4234234234234235e-05,
+      "loss": 1.3316,
+      "step": 5161
+    },
+    {
+      "epoch": 0.006747685956787097,
+      "grad_norm": 1.3685272932052612,
+      "learning_rate": 2.422922922922923e-05,
+      "loss": 1.4648,
+      "step": 5162
+    },
+    {
+      "epoch": 0.006748993141203368,
+      "grad_norm": 2.0369739532470703,
+      "learning_rate": 2.4224224224224227e-05,
+      "loss": 1.4238,
+      "step": 5163
+    },
+    {
+      "epoch": 0.006750300325619638,
+      "grad_norm": 1.4605495929718018,
+      "learning_rate": 2.421921921921922e-05,
+      "loss": 1.4949,
+      "step": 5164
+    },
+    {
+      "epoch": 0.006751607510035908,
+      "grad_norm": 1.170142412185669,
+      "learning_rate": 2.4214214214214213e-05,
+      "loss": 1.7896,
+      "step": 5165
+    },
+    {
+      "epoch": 0.0067529146944521785,
+      "grad_norm": 1.7659274339675903,
+      "learning_rate": 2.420920920920921e-05,
+      "loss": 2.137,
+      "step": 5166
+    },
+    {
+      "epoch": 0.006754221878868449,
+      "grad_norm": 1.7806605100631714,
+      "learning_rate": 2.4204204204204206e-05,
+      "loss": 1.8833,
+      "step": 5167
+    },
+    {
+      "epoch": 0.006755529063284719,
+      "grad_norm": 1.5453237295150757,
+      "learning_rate": 2.41991991991992e-05,
+      "loss": 1.5863,
+      "step": 5168
+    },
+    {
+      "epoch": 0.0067568362477009895,
+      "grad_norm": 1.303763747215271,
+      "learning_rate": 2.4194194194194195e-05,
+      "loss": 1.699,
+      "step": 5169
+    },
+    {
+      "epoch": 0.00675814343211726,
+      "grad_norm": 1.0735180377960205,
+      "learning_rate": 2.418918918918919e-05,
+      "loss": 0.9076,
+      "step": 5170
+    },
+    {
+      "epoch": 0.00675945061653353,
+      "grad_norm": 1.2449947595596313,
+      "learning_rate": 2.4184184184184185e-05,
+      "loss": 1.303,
+      "step": 5171
+    },
+    {
+      "epoch": 0.0067607578009498005,
+      "grad_norm": 1.1637952327728271,
+      "learning_rate": 2.417917917917918e-05,
+      "loss": 0.8086,
+      "step": 5172
+    },
+    {
+      "epoch": 0.006762064985366071,
+      "grad_norm": 0.913434624671936,
+      "learning_rate": 2.4174174174174174e-05,
+      "loss": 0.494,
+      "step": 5173
+    },
+    {
+      "epoch": 0.006763372169782341,
+      "grad_norm": 1.3578027486801147,
+      "learning_rate": 2.416916916916917e-05,
+      "loss": 1.5264,
+      "step": 5174
+    },
+    {
+      "epoch": 0.006764679354198611,
+      "grad_norm": 2.490394353866577,
+      "learning_rate": 2.4164164164164167e-05,
+      "loss": 1.9425,
+      "step": 5175
+    },
+    {
+      "epoch": 0.006765986538614881,
+      "grad_norm": 1.3357962369918823,
+      "learning_rate": 2.4159159159159163e-05,
+      "loss": 1.3698,
+      "step": 5176
+    },
+    {
+      "epoch": 0.006767293723031151,
+      "grad_norm": 1.372023344039917,
+      "learning_rate": 2.4154154154154156e-05,
+      "loss": 1.5659,
+      "step": 5177
+    },
+    {
+      "epoch": 0.0067686009074474215,
+      "grad_norm": 1.4323053359985352,
+      "learning_rate": 2.414914914914915e-05,
+      "loss": 1.236,
+      "step": 5178
+    },
+    {
+      "epoch": 0.006769908091863692,
+      "grad_norm": 1.0950192213058472,
+      "learning_rate": 2.4144144144144145e-05,
+      "loss": 0.9072,
+      "step": 5179
+    },
+    {
+      "epoch": 0.006771215276279962,
+      "grad_norm": 1.1497838497161865,
+      "learning_rate": 2.4139139139139142e-05,
+      "loss": 1.6327,
+      "step": 5180
+    },
+    {
+      "epoch": 0.0067725224606962325,
+      "grad_norm": 1.3500725030899048,
+      "learning_rate": 2.4134134134134135e-05,
+      "loss": 1.7046,
+      "step": 5181
+    },
+    {
+      "epoch": 0.006773829645112503,
+      "grad_norm": 1.3124586343765259,
+      "learning_rate": 2.412912912912913e-05,
+      "loss": 1.926,
+      "step": 5182
+    },
+    {
+      "epoch": 0.006775136829528773,
+      "grad_norm": 1.9600379467010498,
+      "learning_rate": 2.4124124124124124e-05,
+      "loss": 1.1162,
+      "step": 5183
+    },
+    {
+      "epoch": 0.006776444013945043,
+      "grad_norm": 1.104040265083313,
+      "learning_rate": 2.411911911911912e-05,
+      "loss": 1.259,
+      "step": 5184
+    },
+    {
+      "epoch": 0.006777751198361314,
+      "grad_norm": 2.6584110260009766,
+      "learning_rate": 2.4114114114114113e-05,
+      "loss": 1.3717,
+      "step": 5185
+    },
+    {
+      "epoch": 0.006779058382777584,
+      "grad_norm": 1.3104841709136963,
+      "learning_rate": 2.410910910910911e-05,
+      "loss": 1.4223,
+      "step": 5186
+    },
+    {
+      "epoch": 0.006780365567193854,
+      "grad_norm": 1.9476407766342163,
+      "learning_rate": 2.4104104104104106e-05,
+      "loss": 2.6116,
+      "step": 5187
+    },
+    {
+      "epoch": 0.006781672751610125,
+      "grad_norm": 1.5181946754455566,
+      "learning_rate": 2.4099099099099102e-05,
+      "loss": 1.6749,
+      "step": 5188
+    },
+    {
+      "epoch": 0.006782979936026395,
+      "grad_norm": 1.1891509294509888,
+      "learning_rate": 2.4094094094094095e-05,
+      "loss": 1.3746,
+      "step": 5189
+    },
+    {
+      "epoch": 0.006784287120442665,
+      "grad_norm": 1.602076530456543,
+      "learning_rate": 2.408908908908909e-05,
+      "loss": 1.7917,
+      "step": 5190
+    },
+    {
+      "epoch": 0.006785594304858935,
+      "grad_norm": 1.6089667081832886,
+      "learning_rate": 2.4084084084084085e-05,
+      "loss": 1.9196,
+      "step": 5191
+    },
+    {
+      "epoch": 0.006786901489275205,
+      "grad_norm": 1.3522752523422241,
+      "learning_rate": 2.407907907907908e-05,
+      "loss": 1.7237,
+      "step": 5192
+    },
+    {
+      "epoch": 0.0067882086736914754,
+      "grad_norm": 1.4763245582580566,
+      "learning_rate": 2.4074074074074074e-05,
+      "loss": 1.8595,
+      "step": 5193
+    },
+    {
+      "epoch": 0.006789515858107746,
+      "grad_norm": 1.2325096130371094,
+      "learning_rate": 2.406906906906907e-05,
+      "loss": 1.7698,
+      "step": 5194
+    },
+    {
+      "epoch": 0.006790823042524016,
+      "grad_norm": 1.6049537658691406,
+      "learning_rate": 2.4064064064064067e-05,
+      "loss": 1.5132,
+      "step": 5195
+    },
+    {
+      "epoch": 0.006792130226940286,
+      "grad_norm": 1.3449797630310059,
+      "learning_rate": 2.405905905905906e-05,
+      "loss": 1.4475,
+      "step": 5196
+    },
+    {
+      "epoch": 0.006793437411356557,
+      "grad_norm": 1.044716477394104,
+      "learning_rate": 2.4054054054054056e-05,
+      "loss": 0.7884,
+      "step": 5197
+    },
+    {
+      "epoch": 0.006794744595772827,
+      "grad_norm": 1.5163275003433228,
+      "learning_rate": 2.404904904904905e-05,
+      "loss": 2.0895,
+      "step": 5198
+    },
+    {
+      "epoch": 0.006796051780189097,
+      "grad_norm": 1.4344747066497803,
+      "learning_rate": 2.4044044044044045e-05,
+      "loss": 1.424,
+      "step": 5199
+    },
+    {
+      "epoch": 0.006797358964605368,
+      "grad_norm": 1.398292064666748,
+      "learning_rate": 2.4039039039039042e-05,
+      "loss": 1.3409,
+      "step": 5200
+    },
+    {
+      "epoch": 0.006798666149021638,
+      "grad_norm": 1.415850281715393,
+      "learning_rate": 2.4034034034034035e-05,
+      "loss": 1.1875,
+      "step": 5201
+    },
+    {
+      "epoch": 0.006799973333437908,
+      "grad_norm": 1.379431128501892,
+      "learning_rate": 2.4029029029029028e-05,
+      "loss": 2.2387,
+      "step": 5202
+    },
+    {
+      "epoch": 0.006801280517854179,
+      "grad_norm": 1.3395307064056396,
+      "learning_rate": 2.4024024024024024e-05,
+      "loss": 1.3599,
+      "step": 5203
+    },
+    {
+      "epoch": 0.006802587702270449,
+      "grad_norm": 1.7280420064926147,
+      "learning_rate": 2.401901901901902e-05,
+      "loss": 2.3027,
+      "step": 5204
+    },
+    {
+      "epoch": 0.006803894886686719,
+      "grad_norm": 1.204862117767334,
+      "learning_rate": 2.4014014014014017e-05,
+      "loss": 1.2689,
+      "step": 5205
+    },
+    {
+      "epoch": 0.006805202071102989,
+      "grad_norm": 1.3023409843444824,
+      "learning_rate": 2.400900900900901e-05,
+      "loss": 0.9795,
+      "step": 5206
+    },
+    {
+      "epoch": 0.006806509255519259,
+      "grad_norm": 1.4317615032196045,
+      "learning_rate": 2.4004004004004006e-05,
+      "loss": 1.8955,
+      "step": 5207
+    },
+    {
+      "epoch": 0.006807816439935529,
+      "grad_norm": 1.499489665031433,
+      "learning_rate": 2.3998998998999e-05,
+      "loss": 1.3194,
+      "step": 5208
+    },
+    {
+      "epoch": 0.0068091236243518,
+      "grad_norm": 1.8017634153366089,
+      "learning_rate": 2.3993993993993996e-05,
+      "loss": 0.6345,
+      "step": 5209
+    },
+    {
+      "epoch": 0.00681043080876807,
+      "grad_norm": 1.3291469812393188,
+      "learning_rate": 2.398898898898899e-05,
+      "loss": 2.503,
+      "step": 5210
+    },
+    {
+      "epoch": 0.00681173799318434,
+      "grad_norm": 1.6420637369155884,
+      "learning_rate": 2.3983983983983985e-05,
+      "loss": 1.6537,
+      "step": 5211
+    },
+    {
+      "epoch": 0.006813045177600611,
+      "grad_norm": 1.1422632932662964,
+      "learning_rate": 2.397897897897898e-05,
+      "loss": 0.8569,
+      "step": 5212
+    },
+    {
+      "epoch": 0.006814352362016881,
+      "grad_norm": 1.584867238998413,
+      "learning_rate": 2.3973973973973978e-05,
+      "loss": 1.8266,
+      "step": 5213
+    },
+    {
+      "epoch": 0.006815659546433151,
+      "grad_norm": 1.628583312034607,
+      "learning_rate": 2.396896896896897e-05,
+      "loss": 1.9057,
+      "step": 5214
+    },
+    {
+      "epoch": 0.006816966730849422,
+      "grad_norm": 1.4069159030914307,
+      "learning_rate": 2.3963963963963964e-05,
+      "loss": 1.3815,
+      "step": 5215
+    },
+    {
+      "epoch": 0.006818273915265692,
+      "grad_norm": 1.4100984334945679,
+      "learning_rate": 2.395895895895896e-05,
+      "loss": 1.5416,
+      "step": 5216
+    },
+    {
+      "epoch": 0.006819581099681962,
+      "grad_norm": 1.6543521881103516,
+      "learning_rate": 2.3953953953953956e-05,
+      "loss": 1.4182,
+      "step": 5217
+    },
+    {
+      "epoch": 0.0068208882840982326,
+      "grad_norm": 1.8731374740600586,
+      "learning_rate": 2.394894894894895e-05,
+      "loss": 1.626,
+      "step": 5218
+    },
+    {
+      "epoch": 0.006822195468514503,
+      "grad_norm": 1.1168686151504517,
+      "learning_rate": 2.3943943943943946e-05,
+      "loss": 1.1684,
+      "step": 5219
+    },
+    {
+      "epoch": 0.006823502652930773,
+      "grad_norm": 2.152836561203003,
+      "learning_rate": 2.393893893893894e-05,
+      "loss": 1.7469,
+      "step": 5220
+    },
+    {
+      "epoch": 0.006824809837347043,
+      "grad_norm": 1.309692621231079,
+      "learning_rate": 2.3933933933933935e-05,
+      "loss": 2.3861,
+      "step": 5221
+    },
+    {
+      "epoch": 0.006826117021763313,
+      "grad_norm": 1.384637475013733,
+      "learning_rate": 2.392892892892893e-05,
+      "loss": 1.7207,
+      "step": 5222
+    },
+    {
+      "epoch": 0.006827424206179583,
+      "grad_norm": 1.1091439723968506,
+      "learning_rate": 2.3923923923923924e-05,
+      "loss": 0.8857,
+      "step": 5223
+    },
+    {
+      "epoch": 0.006828731390595854,
+      "grad_norm": 1.4374399185180664,
+      "learning_rate": 2.391891891891892e-05,
+      "loss": 1.7694,
+      "step": 5224
+    },
+    {
+      "epoch": 0.006830038575012124,
+      "grad_norm": 1.1748125553131104,
+      "learning_rate": 2.3913913913913917e-05,
+      "loss": 1.0725,
+      "step": 5225
+    },
+    {
+      "epoch": 0.006831345759428394,
+      "grad_norm": 1.4612393379211426,
+      "learning_rate": 2.390890890890891e-05,
+      "loss": 2.0632,
+      "step": 5226
+    },
+    {
+      "epoch": 0.006832652943844665,
+      "grad_norm": 1.6743860244750977,
+      "learning_rate": 2.3903903903903903e-05,
+      "loss": 2.0016,
+      "step": 5227
+    },
+    {
+      "epoch": 0.006833960128260935,
+      "grad_norm": 1.3318228721618652,
+      "learning_rate": 2.38988988988989e-05,
+      "loss": 1.3722,
+      "step": 5228
+    },
+    {
+      "epoch": 0.006835267312677205,
+      "grad_norm": 1.2059401273727417,
+      "learning_rate": 2.3893893893893896e-05,
+      "loss": 1.4461,
+      "step": 5229
+    },
+    {
+      "epoch": 0.0068365744970934755,
+      "grad_norm": 1.4194608926773071,
+      "learning_rate": 2.3888888888888892e-05,
+      "loss": 0.9823,
+      "step": 5230
+    },
+    {
+      "epoch": 0.006837881681509746,
+      "grad_norm": 1.3727227449417114,
+      "learning_rate": 2.3883883883883885e-05,
+      "loss": 1.3697,
+      "step": 5231
+    },
+    {
+      "epoch": 0.006839188865926016,
+      "grad_norm": 1.330737590789795,
+      "learning_rate": 2.387887887887888e-05,
+      "loss": 1.8372,
+      "step": 5232
+    },
+    {
+      "epoch": 0.0068404960503422865,
+      "grad_norm": 1.6119452714920044,
+      "learning_rate": 2.3873873873873874e-05,
+      "loss": 1.4178,
+      "step": 5233
+    },
+    {
+      "epoch": 0.006841803234758557,
+      "grad_norm": 0.9142619371414185,
+      "learning_rate": 2.386886886886887e-05,
+      "loss": 0.7687,
+      "step": 5234
+    },
+    {
+      "epoch": 0.006843110419174827,
+      "grad_norm": 1.5699083805084229,
+      "learning_rate": 2.3863863863863864e-05,
+      "loss": 2.1005,
+      "step": 5235
+    },
+    {
+      "epoch": 0.0068444176035910975,
+      "grad_norm": 0.9492635130882263,
+      "learning_rate": 2.385885885885886e-05,
+      "loss": 0.5727,
+      "step": 5236
+    },
+    {
+      "epoch": 0.006845724788007367,
+      "grad_norm": 1.4195315837860107,
+      "learning_rate": 2.3853853853853856e-05,
+      "loss": 1.2223,
+      "step": 5237
+    },
+    {
+      "epoch": 0.006847031972423637,
+      "grad_norm": 1.7125307321548462,
+      "learning_rate": 2.384884884884885e-05,
+      "loss": 2.1594,
+      "step": 5238
+    },
+    {
+      "epoch": 0.0068483391568399075,
+      "grad_norm": 1.25047767162323,
+      "learning_rate": 2.3843843843843842e-05,
+      "loss": 1.5188,
+      "step": 5239
+    },
+    {
+      "epoch": 0.006849646341256178,
+      "grad_norm": 1.0839757919311523,
+      "learning_rate": 2.383883883883884e-05,
+      "loss": 1.1239,
+      "step": 5240
+    },
+    {
+      "epoch": 0.006850953525672448,
+      "grad_norm": 1.98062264919281,
+      "learning_rate": 2.3833833833833835e-05,
+      "loss": 1.8429,
+      "step": 5241
+    },
+    {
+      "epoch": 0.0068522607100887185,
+      "grad_norm": 1.238759994506836,
+      "learning_rate": 2.382882882882883e-05,
+      "loss": 1.253,
+      "step": 5242
+    },
+    {
+      "epoch": 0.006853567894504989,
+      "grad_norm": 1.1805882453918457,
+      "learning_rate": 2.3823823823823824e-05,
+      "loss": 0.9404,
+      "step": 5243
+    },
+    {
+      "epoch": 0.006854875078921259,
+      "grad_norm": 1.286556363105774,
+      "learning_rate": 2.381881881881882e-05,
+      "loss": 2.3354,
+      "step": 5244
+    },
+    {
+      "epoch": 0.0068561822633375295,
+      "grad_norm": 1.128224492073059,
+      "learning_rate": 2.3813813813813814e-05,
+      "loss": 0.7638,
+      "step": 5245
+    },
+    {
+      "epoch": 0.0068574894477538,
+      "grad_norm": 1.5756337642669678,
+      "learning_rate": 2.380880880880881e-05,
+      "loss": 1.9557,
+      "step": 5246
+    },
+    {
+      "epoch": 0.00685879663217007,
+      "grad_norm": 1.6223574876785278,
+      "learning_rate": 2.3803803803803803e-05,
+      "loss": 1.2095,
+      "step": 5247
+    },
+    {
+      "epoch": 0.00686010381658634,
+      "grad_norm": 1.7080355882644653,
+      "learning_rate": 2.37987987987988e-05,
+      "loss": 1.8025,
+      "step": 5248
+    },
+    {
+      "epoch": 0.006861411001002611,
+      "grad_norm": 1.4979538917541504,
+      "learning_rate": 2.3793793793793796e-05,
+      "loss": 2.3975,
+      "step": 5249
+    },
+    {
+      "epoch": 0.006862718185418881,
+      "grad_norm": 1.3513624668121338,
+      "learning_rate": 2.3788788788788792e-05,
+      "loss": 1.1838,
+      "step": 5250
+    },
+    {
+      "epoch": 0.006864025369835151,
+      "grad_norm": 1.6387608051300049,
+      "learning_rate": 2.3783783783783785e-05,
+      "loss": 1.9729,
+      "step": 5251
+    },
+    {
+      "epoch": 0.006865332554251421,
+      "grad_norm": 1.4821265935897827,
+      "learning_rate": 2.3778778778778778e-05,
+      "loss": 2.2966,
+      "step": 5252
+    },
+    {
+      "epoch": 0.006866639738667691,
+      "grad_norm": 2.081787347793579,
+      "learning_rate": 2.3773773773773775e-05,
+      "loss": 1.0834,
+      "step": 5253
+    },
+    {
+      "epoch": 0.0068679469230839615,
+      "grad_norm": 1.3557729721069336,
+      "learning_rate": 2.376876876876877e-05,
+      "loss": 1.3505,
+      "step": 5254
+    },
+    {
+      "epoch": 0.006869254107500232,
+      "grad_norm": 3.2460150718688965,
+      "learning_rate": 2.3763763763763767e-05,
+      "loss": 1.7588,
+      "step": 5255
+    },
+    {
+      "epoch": 0.006870561291916502,
+      "grad_norm": 1.2762162685394287,
+      "learning_rate": 2.375875875875876e-05,
+      "loss": 0.9961,
+      "step": 5256
+    },
+    {
+      "epoch": 0.006871868476332772,
+      "grad_norm": 1.6455742120742798,
+      "learning_rate": 2.3753753753753753e-05,
+      "loss": 2.046,
+      "step": 5257
+    },
+    {
+      "epoch": 0.006873175660749043,
+      "grad_norm": 1.4383970499038696,
+      "learning_rate": 2.374874874874875e-05,
+      "loss": 1.7005,
+      "step": 5258
+    },
+    {
+      "epoch": 0.006874482845165313,
+      "grad_norm": 1.5427228212356567,
+      "learning_rate": 2.3743743743743746e-05,
+      "loss": 1.618,
+      "step": 5259
+    },
+    {
+      "epoch": 0.006875790029581583,
+      "grad_norm": 2.3443117141723633,
+      "learning_rate": 2.373873873873874e-05,
+      "loss": 1.9373,
+      "step": 5260
+    },
+    {
+      "epoch": 0.006877097213997854,
+      "grad_norm": 1.485619306564331,
+      "learning_rate": 2.3733733733733735e-05,
+      "loss": 1.7801,
+      "step": 5261
+    },
+    {
+      "epoch": 0.006878404398414124,
+      "grad_norm": 1.1171085834503174,
+      "learning_rate": 2.372872872872873e-05,
+      "loss": 1.3716,
+      "step": 5262
+    },
+    {
+      "epoch": 0.006879711582830394,
+      "grad_norm": 1.2093690633773804,
+      "learning_rate": 2.3723723723723725e-05,
+      "loss": 1.5662,
+      "step": 5263
+    },
+    {
+      "epoch": 0.006881018767246665,
+      "grad_norm": 1.4352233409881592,
+      "learning_rate": 2.3718718718718718e-05,
+      "loss": 1.0029,
+      "step": 5264
+    },
+    {
+      "epoch": 0.006882325951662935,
+      "grad_norm": 1.3214361667633057,
+      "learning_rate": 2.3713713713713714e-05,
+      "loss": 1.1591,
+      "step": 5265
+    },
+    {
+      "epoch": 0.006883633136079205,
+      "grad_norm": 1.196938157081604,
+      "learning_rate": 2.370870870870871e-05,
+      "loss": 1.2002,
+      "step": 5266
+    },
+    {
+      "epoch": 0.006884940320495475,
+      "grad_norm": 1.4648559093475342,
+      "learning_rate": 2.3703703703703707e-05,
+      "loss": 1.7282,
+      "step": 5267
+    },
+    {
+      "epoch": 0.006886247504911745,
+      "grad_norm": 1.4855016469955444,
+      "learning_rate": 2.36986986986987e-05,
+      "loss": 0.8336,
+      "step": 5268
+    },
+    {
+      "epoch": 0.006887554689328015,
+      "grad_norm": 1.4503592252731323,
+      "learning_rate": 2.3693693693693696e-05,
+      "loss": 1.9362,
+      "step": 5269
+    },
+    {
+      "epoch": 0.006888861873744286,
+      "grad_norm": 1.2709898948669434,
+      "learning_rate": 2.368868868868869e-05,
+      "loss": 1.6369,
+      "step": 5270
+    },
+    {
+      "epoch": 0.006890169058160556,
+      "grad_norm": 1.4264291524887085,
+      "learning_rate": 2.3683683683683685e-05,
+      "loss": 1.5667,
+      "step": 5271
+    },
+    {
+      "epoch": 0.006891476242576826,
+      "grad_norm": 1.366345763206482,
+      "learning_rate": 2.3678678678678678e-05,
+      "loss": 0.793,
+      "step": 5272
+    },
+    {
+      "epoch": 0.006892783426993097,
+      "grad_norm": 1.314568281173706,
+      "learning_rate": 2.3673673673673675e-05,
+      "loss": 1.4215,
+      "step": 5273
+    },
+    {
+      "epoch": 0.006894090611409367,
+      "grad_norm": 1.0779858827590942,
+      "learning_rate": 2.366866866866867e-05,
+      "loss": 0.924,
+      "step": 5274
+    },
+    {
+      "epoch": 0.006895397795825637,
+      "grad_norm": 1.0787593126296997,
+      "learning_rate": 2.3663663663663664e-05,
+      "loss": 1.0664,
+      "step": 5275
+    },
+    {
+      "epoch": 0.006896704980241908,
+      "grad_norm": 1.2487764358520508,
+      "learning_rate": 2.365865865865866e-05,
+      "loss": 1.3019,
+      "step": 5276
+    },
+    {
+      "epoch": 0.006898012164658178,
+      "grad_norm": 1.8630363941192627,
+      "learning_rate": 2.3653653653653653e-05,
+      "loss": 1.6486,
+      "step": 5277
+    },
+    {
+      "epoch": 0.006899319349074448,
+      "grad_norm": 1.3853257894515991,
+      "learning_rate": 2.364864864864865e-05,
+      "loss": 1.409,
+      "step": 5278
+    },
+    {
+      "epoch": 0.006900626533490719,
+      "grad_norm": 1.6463223695755005,
+      "learning_rate": 2.3643643643643646e-05,
+      "loss": 2.378,
+      "step": 5279
+    },
+    {
+      "epoch": 0.006901933717906989,
+      "grad_norm": 1.7271759510040283,
+      "learning_rate": 2.3638638638638642e-05,
+      "loss": 2.2023,
+      "step": 5280
+    },
+    {
+      "epoch": 0.006903240902323259,
+      "grad_norm": 1.247131586074829,
+      "learning_rate": 2.3633633633633635e-05,
+      "loss": 1.1618,
+      "step": 5281
+    },
+    {
+      "epoch": 0.0069045480867395296,
+      "grad_norm": 1.2421094179153442,
+      "learning_rate": 2.362862862862863e-05,
+      "loss": 0.8686,
+      "step": 5282
+    },
+    {
+      "epoch": 0.006905855271155799,
+      "grad_norm": 1.5895910263061523,
+      "learning_rate": 2.3623623623623625e-05,
+      "loss": 1.5564,
+      "step": 5283
+    },
+    {
+      "epoch": 0.006907162455572069,
+      "grad_norm": 1.050890564918518,
+      "learning_rate": 2.361861861861862e-05,
+      "loss": 1.2069,
+      "step": 5284
+    },
+    {
+      "epoch": 0.00690846963998834,
+      "grad_norm": 1.9188271760940552,
+      "learning_rate": 2.3613613613613614e-05,
+      "loss": 2.6112,
+      "step": 5285
+    },
+    {
+      "epoch": 0.00690977682440461,
+      "grad_norm": 1.5348882675170898,
+      "learning_rate": 2.360860860860861e-05,
+      "loss": 1.276,
+      "step": 5286
+    },
+    {
+      "epoch": 0.00691108400882088,
+      "grad_norm": 1.7857577800750732,
+      "learning_rate": 2.3603603603603607e-05,
+      "loss": 2.0682,
+      "step": 5287
+    },
+    {
+      "epoch": 0.006912391193237151,
+      "grad_norm": 1.9310287237167358,
+      "learning_rate": 2.35985985985986e-05,
+      "loss": 1.7848,
+      "step": 5288
+    },
+    {
+      "epoch": 0.006913698377653421,
+      "grad_norm": 1.3942519426345825,
+      "learning_rate": 2.3593593593593593e-05,
+      "loss": 1.181,
+      "step": 5289
+    },
+    {
+      "epoch": 0.006915005562069691,
+      "grad_norm": 1.8630304336547852,
+      "learning_rate": 2.358858858858859e-05,
+      "loss": 1.5979,
+      "step": 5290
+    },
+    {
+      "epoch": 0.006916312746485962,
+      "grad_norm": 0.9505850672721863,
+      "learning_rate": 2.3583583583583585e-05,
+      "loss": 0.6404,
+      "step": 5291
+    },
+    {
+      "epoch": 0.006917619930902232,
+      "grad_norm": 1.2821629047393799,
+      "learning_rate": 2.3578578578578582e-05,
+      "loss": 1.7606,
+      "step": 5292
+    },
+    {
+      "epoch": 0.006918927115318502,
+      "grad_norm": 1.5782756805419922,
+      "learning_rate": 2.3573573573573575e-05,
+      "loss": 2.249,
+      "step": 5293
+    },
+    {
+      "epoch": 0.0069202342997347725,
+      "grad_norm": 1.5444504022598267,
+      "learning_rate": 2.3568568568568568e-05,
+      "loss": 1.579,
+      "step": 5294
+    },
+    {
+      "epoch": 0.006921541484151043,
+      "grad_norm": 1.3233392238616943,
+      "learning_rate": 2.3563563563563564e-05,
+      "loss": 1.3529,
+      "step": 5295
+    },
+    {
+      "epoch": 0.006922848668567313,
+      "grad_norm": 1.5965698957443237,
+      "learning_rate": 2.355855855855856e-05,
+      "loss": 1.186,
+      "step": 5296
+    },
+    {
+      "epoch": 0.0069241558529835835,
+      "grad_norm": 1.3758896589279175,
+      "learning_rate": 2.3553553553553553e-05,
+      "loss": 2.0034,
+      "step": 5297
+    },
+    {
+      "epoch": 0.006925463037399853,
+      "grad_norm": 1.380207896232605,
+      "learning_rate": 2.354854854854855e-05,
+      "loss": 1.5438,
+      "step": 5298
+    },
+    {
+      "epoch": 0.006926770221816123,
+      "grad_norm": 1.3084547519683838,
+      "learning_rate": 2.3543543543543546e-05,
+      "loss": 1.5186,
+      "step": 5299
+    },
+    {
+      "epoch": 0.006928077406232394,
+      "grad_norm": 1.7564197778701782,
+      "learning_rate": 2.353853853853854e-05,
+      "loss": 2.3773,
+      "step": 5300
+    },
+    {
+      "epoch": 0.006929384590648664,
+      "grad_norm": 1.6349116563796997,
+      "learning_rate": 2.3533533533533532e-05,
+      "loss": 1.9448,
+      "step": 5301
+    },
+    {
+      "epoch": 0.006930691775064934,
+      "grad_norm": 1.5414271354675293,
+      "learning_rate": 2.352852852852853e-05,
+      "loss": 1.7767,
+      "step": 5302
+    },
+    {
+      "epoch": 0.0069319989594812045,
+      "grad_norm": 1.4953641891479492,
+      "learning_rate": 2.3523523523523525e-05,
+      "loss": 1.5118,
+      "step": 5303
+    },
+    {
+      "epoch": 0.006933306143897475,
+      "grad_norm": 1.4771270751953125,
+      "learning_rate": 2.351851851851852e-05,
+      "loss": 1.1124,
+      "step": 5304
+    },
+    {
+      "epoch": 0.006934613328313745,
+      "grad_norm": 2.345992088317871,
+      "learning_rate": 2.3513513513513518e-05,
+      "loss": 1.6587,
+      "step": 5305
+    },
+    {
+      "epoch": 0.0069359205127300155,
+      "grad_norm": 1.3581690788269043,
+      "learning_rate": 2.3508508508508507e-05,
+      "loss": 1.8011,
+      "step": 5306
+    },
+    {
+      "epoch": 0.006937227697146286,
+      "grad_norm": 1.1031707525253296,
+      "learning_rate": 2.3503503503503504e-05,
+      "loss": 2.225,
+      "step": 5307
+    },
+    {
+      "epoch": 0.006938534881562556,
+      "grad_norm": 1.3592791557312012,
+      "learning_rate": 2.34984984984985e-05,
+      "loss": 1.2387,
+      "step": 5308
+    },
+    {
+      "epoch": 0.0069398420659788265,
+      "grad_norm": 1.4060096740722656,
+      "learning_rate": 2.3493493493493496e-05,
+      "loss": 1.9524,
+      "step": 5309
+    },
+    {
+      "epoch": 0.006941149250395097,
+      "grad_norm": 1.6784847974777222,
+      "learning_rate": 2.348848848848849e-05,
+      "loss": 2.2446,
+      "step": 5310
+    },
+    {
+      "epoch": 0.006942456434811367,
+      "grad_norm": 1.0280356407165527,
+      "learning_rate": 2.3483483483483486e-05,
+      "loss": 0.6934,
+      "step": 5311
+    },
+    {
+      "epoch": 0.006943763619227637,
+      "grad_norm": 1.6229842901229858,
+      "learning_rate": 2.347847847847848e-05,
+      "loss": 1.6324,
+      "step": 5312
+    },
+    {
+      "epoch": 0.006945070803643907,
+      "grad_norm": 1.5115021467208862,
+      "learning_rate": 2.3473473473473475e-05,
+      "loss": 1.7724,
+      "step": 5313
+    },
+    {
+      "epoch": 0.006946377988060177,
+      "grad_norm": 2.3112611770629883,
+      "learning_rate": 2.3468468468468468e-05,
+      "loss": 2.5698,
+      "step": 5314
+    },
+    {
+      "epoch": 0.0069476851724764475,
+      "grad_norm": 1.0140069723129272,
+      "learning_rate": 2.3463463463463464e-05,
+      "loss": 1.1625,
+      "step": 5315
+    },
+    {
+      "epoch": 0.006948992356892718,
+      "grad_norm": 1.4211794137954712,
+      "learning_rate": 2.345845845845846e-05,
+      "loss": 1.6272,
+      "step": 5316
+    },
+    {
+      "epoch": 0.006950299541308988,
+      "grad_norm": 1.6229989528656006,
+      "learning_rate": 2.3453453453453457e-05,
+      "loss": 2.3475,
+      "step": 5317
+    },
+    {
+      "epoch": 0.0069516067257252585,
+      "grad_norm": 1.6087658405303955,
+      "learning_rate": 2.344844844844845e-05,
+      "loss": 1.9261,
+      "step": 5318
+    },
+    {
+      "epoch": 0.006952913910141529,
+      "grad_norm": 2.3902406692504883,
+      "learning_rate": 2.3443443443443443e-05,
+      "loss": 1.7826,
+      "step": 5319
+    },
+    {
+      "epoch": 0.006954221094557799,
+      "grad_norm": 1.2412396669387817,
+      "learning_rate": 2.343843843843844e-05,
+      "loss": 1.4133,
+      "step": 5320
+    },
+    {
+      "epoch": 0.006955528278974069,
+      "grad_norm": 1.0065289735794067,
+      "learning_rate": 2.3433433433433436e-05,
+      "loss": 0.733,
+      "step": 5321
+    },
+    {
+      "epoch": 0.00695683546339034,
+      "grad_norm": 1.4324588775634766,
+      "learning_rate": 2.342842842842843e-05,
+      "loss": 2.3137,
+      "step": 5322
+    },
+    {
+      "epoch": 0.00695814264780661,
+      "grad_norm": 1.1168471574783325,
+      "learning_rate": 2.3423423423423425e-05,
+      "loss": 0.8206,
+      "step": 5323
+    },
+    {
+      "epoch": 0.00695944983222288,
+      "grad_norm": 1.3522891998291016,
+      "learning_rate": 2.3418418418418418e-05,
+      "loss": 1.9269,
+      "step": 5324
+    },
+    {
+      "epoch": 0.006960757016639151,
+      "grad_norm": 1.4346336126327515,
+      "learning_rate": 2.3413413413413414e-05,
+      "loss": 1.7796,
+      "step": 5325
+    },
+    {
+      "epoch": 0.006962064201055421,
+      "grad_norm": 1.5707212686538696,
+      "learning_rate": 2.3408408408408407e-05,
+      "loss": 1.6761,
+      "step": 5326
+    },
+    {
+      "epoch": 0.006963371385471691,
+      "grad_norm": 1.4819737672805786,
+      "learning_rate": 2.3403403403403404e-05,
+      "loss": 1.8922,
+      "step": 5327
+    },
+    {
+      "epoch": 0.006964678569887961,
+      "grad_norm": 2.123586893081665,
+      "learning_rate": 2.33983983983984e-05,
+      "loss": 2.1495,
+      "step": 5328
+    },
+    {
+      "epoch": 0.006965985754304231,
+      "grad_norm": 1.2238904237747192,
+      "learning_rate": 2.3393393393393396e-05,
+      "loss": 1.0987,
+      "step": 5329
+    },
+    {
+      "epoch": 0.0069672929387205014,
+      "grad_norm": 1.2786980867385864,
+      "learning_rate": 2.338838838838839e-05,
+      "loss": 1.808,
+      "step": 5330
+    },
+    {
+      "epoch": 0.006968600123136772,
+      "grad_norm": 1.3246428966522217,
+      "learning_rate": 2.3383383383383382e-05,
+      "loss": 0.8551,
+      "step": 5331
+    },
+    {
+      "epoch": 0.006969907307553042,
+      "grad_norm": 1.4565104246139526,
+      "learning_rate": 2.337837837837838e-05,
+      "loss": 1.5376,
+      "step": 5332
+    },
+    {
+      "epoch": 0.006971214491969312,
+      "grad_norm": 1.217069149017334,
+      "learning_rate": 2.3373373373373375e-05,
+      "loss": 1.3149,
+      "step": 5333
+    },
+    {
+      "epoch": 0.006972521676385583,
+      "grad_norm": 1.7385847568511963,
+      "learning_rate": 2.336836836836837e-05,
+      "loss": 2.3363,
+      "step": 5334
+    },
+    {
+      "epoch": 0.006973828860801853,
+      "grad_norm": 1.4229004383087158,
+      "learning_rate": 2.3363363363363364e-05,
+      "loss": 2.1541,
+      "step": 5335
+    },
+    {
+      "epoch": 0.006975136045218123,
+      "grad_norm": 1.3839478492736816,
+      "learning_rate": 2.335835835835836e-05,
+      "loss": 1.4482,
+      "step": 5336
+    },
+    {
+      "epoch": 0.006976443229634394,
+      "grad_norm": 1.4325567483901978,
+      "learning_rate": 2.3353353353353354e-05,
+      "loss": 1.8986,
+      "step": 5337
+    },
+    {
+      "epoch": 0.006977750414050664,
+      "grad_norm": 1.6140114068984985,
+      "learning_rate": 2.334834834834835e-05,
+      "loss": 2.382,
+      "step": 5338
+    },
+    {
+      "epoch": 0.006979057598466934,
+      "grad_norm": 1.3417935371398926,
+      "learning_rate": 2.3343343343343343e-05,
+      "loss": 1.7142,
+      "step": 5339
+    },
+    {
+      "epoch": 0.006980364782883205,
+      "grad_norm": 1.4469348192214966,
+      "learning_rate": 2.333833833833834e-05,
+      "loss": 2.0766,
+      "step": 5340
+    },
+    {
+      "epoch": 0.006981671967299475,
+      "grad_norm": 1.7070165872573853,
+      "learning_rate": 2.3333333333333336e-05,
+      "loss": 1.5902,
+      "step": 5341
+    },
+    {
+      "epoch": 0.006982979151715745,
+      "grad_norm": 1.429005742073059,
+      "learning_rate": 2.3328328328328332e-05,
+      "loss": 2.7314,
+      "step": 5342
+    },
+    {
+      "epoch": 0.006984286336132016,
+      "grad_norm": 1.2844817638397217,
+      "learning_rate": 2.3323323323323322e-05,
+      "loss": 1.5424,
+      "step": 5343
+    },
+    {
+      "epoch": 0.006985593520548285,
+      "grad_norm": 1.5157930850982666,
+      "learning_rate": 2.3318318318318318e-05,
+      "loss": 1.9293,
+      "step": 5344
+    },
+    {
+      "epoch": 0.006986900704964555,
+      "grad_norm": 1.5238776206970215,
+      "learning_rate": 2.3313313313313314e-05,
+      "loss": 2.2804,
+      "step": 5345
+    },
+    {
+      "epoch": 0.006988207889380826,
+      "grad_norm": 1.3895963430404663,
+      "learning_rate": 2.330830830830831e-05,
+      "loss": 1.8329,
+      "step": 5346
+    },
+    {
+      "epoch": 0.006989515073797096,
+      "grad_norm": 1.2151991128921509,
+      "learning_rate": 2.3303303303303304e-05,
+      "loss": 1.4373,
+      "step": 5347
+    },
+    {
+      "epoch": 0.006990822258213366,
+      "grad_norm": 1.227550745010376,
+      "learning_rate": 2.32982982982983e-05,
+      "loss": 0.888,
+      "step": 5348
+    },
+    {
+      "epoch": 0.006992129442629637,
+      "grad_norm": 0.4632551968097687,
+      "learning_rate": 2.3293293293293293e-05,
+      "loss": 0.162,
+      "step": 5349
+    },
+    {
+      "epoch": 0.006993436627045907,
+      "grad_norm": 1.4738856554031372,
+      "learning_rate": 2.328828828828829e-05,
+      "loss": 2.0875,
+      "step": 5350
+    },
+    {
+      "epoch": 0.006994743811462177,
+      "grad_norm": 1.1825004816055298,
+      "learning_rate": 2.3283283283283282e-05,
+      "loss": 2.2923,
+      "step": 5351
+    },
+    {
+      "epoch": 0.006996050995878448,
+      "grad_norm": 1.3689936399459839,
+      "learning_rate": 2.327827827827828e-05,
+      "loss": 1.1477,
+      "step": 5352
+    },
+    {
+      "epoch": 0.006997358180294718,
+      "grad_norm": 1.4123591184616089,
+      "learning_rate": 2.3273273273273275e-05,
+      "loss": 1.3957,
+      "step": 5353
+    },
+    {
+      "epoch": 0.006998665364710988,
+      "grad_norm": 1.4242956638336182,
+      "learning_rate": 2.326826826826827e-05,
+      "loss": 1.8722,
+      "step": 5354
+    },
+    {
+      "epoch": 0.0069999725491272586,
+      "grad_norm": 2.2539846897125244,
+      "learning_rate": 2.3263263263263265e-05,
+      "loss": 1.6856,
+      "step": 5355
+    },
+    {
+      "epoch": 0.007001279733543529,
+      "grad_norm": 1.5431872606277466,
+      "learning_rate": 2.3258258258258258e-05,
+      "loss": 1.9248,
+      "step": 5356
+    },
+    {
+      "epoch": 0.007002586917959799,
+      "grad_norm": 1.7273964881896973,
+      "learning_rate": 2.3253253253253254e-05,
+      "loss": 1.4408,
+      "step": 5357
+    },
+    {
+      "epoch": 0.0070038941023760695,
+      "grad_norm": 1.4280543327331543,
+      "learning_rate": 2.324824824824825e-05,
+      "loss": 1.6886,
+      "step": 5358
+    },
+    {
+      "epoch": 0.007005201286792339,
+      "grad_norm": 1.3539223670959473,
+      "learning_rate": 2.3243243243243247e-05,
+      "loss": 1.5024,
+      "step": 5359
+    },
+    {
+      "epoch": 0.007006508471208609,
+      "grad_norm": 1.4311639070510864,
+      "learning_rate": 2.323823823823824e-05,
+      "loss": 1.2303,
+      "step": 5360
+    },
+    {
+      "epoch": 0.00700781565562488,
+      "grad_norm": 1.3627809286117554,
+      "learning_rate": 2.3233233233233233e-05,
+      "loss": 1.3654,
+      "step": 5361
+    },
+    {
+      "epoch": 0.00700912284004115,
+      "grad_norm": 1.172925353050232,
+      "learning_rate": 2.322822822822823e-05,
+      "loss": 1.1899,
+      "step": 5362
+    },
+    {
+      "epoch": 0.00701043002445742,
+      "grad_norm": 1.3862049579620361,
+      "learning_rate": 2.3223223223223225e-05,
+      "loss": 1.6526,
+      "step": 5363
+    },
+    {
+      "epoch": 0.007011737208873691,
+      "grad_norm": 1.3054598569869995,
+      "learning_rate": 2.3218218218218218e-05,
+      "loss": 1.7045,
+      "step": 5364
+    },
+    {
+      "epoch": 0.007013044393289961,
+      "grad_norm": 1.1205755472183228,
+      "learning_rate": 2.3213213213213215e-05,
+      "loss": 1.0522,
+      "step": 5365
+    },
+    {
+      "epoch": 0.007014351577706231,
+      "grad_norm": 1.3412173986434937,
+      "learning_rate": 2.320820820820821e-05,
+      "loss": 1.7568,
+      "step": 5366
+    },
+    {
+      "epoch": 0.0070156587621225015,
+      "grad_norm": 1.6863069534301758,
+      "learning_rate": 2.3203203203203204e-05,
+      "loss": 1.946,
+      "step": 5367
+    },
+    {
+      "epoch": 0.007016965946538772,
+      "grad_norm": 1.4588853120803833,
+      "learning_rate": 2.3198198198198197e-05,
+      "loss": 1.8594,
+      "step": 5368
+    },
+    {
+      "epoch": 0.007018273130955042,
+      "grad_norm": 1.40575110912323,
+      "learning_rate": 2.3193193193193193e-05,
+      "loss": 1.8196,
+      "step": 5369
+    },
+    {
+      "epoch": 0.0070195803153713125,
+      "grad_norm": 1.4614986181259155,
+      "learning_rate": 2.318818818818819e-05,
+      "loss": 1.4843,
+      "step": 5370
+    },
+    {
+      "epoch": 0.007020887499787583,
+      "grad_norm": 1.3923163414001465,
+      "learning_rate": 2.3183183183183186e-05,
+      "loss": 1.6127,
+      "step": 5371
+    },
+    {
+      "epoch": 0.007022194684203853,
+      "grad_norm": 1.5587844848632812,
+      "learning_rate": 2.317817817817818e-05,
+      "loss": 1.657,
+      "step": 5372
+    },
+    {
+      "epoch": 0.0070235018686201234,
+      "grad_norm": 1.5257054567337036,
+      "learning_rate": 2.3173173173173175e-05,
+      "loss": 1.9167,
+      "step": 5373
+    },
+    {
+      "epoch": 0.007024809053036393,
+      "grad_norm": 1.826261281967163,
+      "learning_rate": 2.316816816816817e-05,
+      "loss": 1.7265,
+      "step": 5374
+    },
+    {
+      "epoch": 0.007026116237452663,
+      "grad_norm": 1.3140952587127686,
+      "learning_rate": 2.3163163163163165e-05,
+      "loss": 1.191,
+      "step": 5375
+    },
+    {
+      "epoch": 0.0070274234218689335,
+      "grad_norm": 1.4552007913589478,
+      "learning_rate": 2.3158158158158158e-05,
+      "loss": 1.5734,
+      "step": 5376
+    },
+    {
+      "epoch": 0.007028730606285204,
+      "grad_norm": 1.7459192276000977,
+      "learning_rate": 2.3153153153153154e-05,
+      "loss": 1.9004,
+      "step": 5377
+    },
+    {
+      "epoch": 0.007030037790701474,
+      "grad_norm": 1.5222671031951904,
+      "learning_rate": 2.314814814814815e-05,
+      "loss": 1.4285,
+      "step": 5378
+    },
+    {
+      "epoch": 0.0070313449751177445,
+      "grad_norm": 1.5322662591934204,
+      "learning_rate": 2.3143143143143143e-05,
+      "loss": 1.9817,
+      "step": 5379
+    },
+    {
+      "epoch": 0.007032652159534015,
+      "grad_norm": 1.3815932273864746,
+      "learning_rate": 2.3138138138138136e-05,
+      "loss": 3.1581,
+      "step": 5380
+    },
+    {
+      "epoch": 0.007033959343950285,
+      "grad_norm": 1.3526736497879028,
+      "learning_rate": 2.3133133133133133e-05,
+      "loss": 1.192,
+      "step": 5381
+    },
+    {
+      "epoch": 0.0070352665283665555,
+      "grad_norm": 1.377832055091858,
+      "learning_rate": 2.312812812812813e-05,
+      "loss": 2.0278,
+      "step": 5382
+    },
+    {
+      "epoch": 0.007036573712782826,
+      "grad_norm": 1.7369847297668457,
+      "learning_rate": 2.3123123123123125e-05,
+      "loss": 2.0927,
+      "step": 5383
+    },
+    {
+      "epoch": 0.007037880897199096,
+      "grad_norm": 1.3556615114212036,
+      "learning_rate": 2.3118118118118122e-05,
+      "loss": 1.396,
+      "step": 5384
+    },
+    {
+      "epoch": 0.007039188081615366,
+      "grad_norm": 1.6186351776123047,
+      "learning_rate": 2.3113113113113115e-05,
+      "loss": 1.4979,
+      "step": 5385
+    },
+    {
+      "epoch": 0.007040495266031637,
+      "grad_norm": 1.3546936511993408,
+      "learning_rate": 2.3108108108108108e-05,
+      "loss": 1.7772,
+      "step": 5386
+    },
+    {
+      "epoch": 0.007041802450447907,
+      "grad_norm": 1.3359774351119995,
+      "learning_rate": 2.3103103103103104e-05,
+      "loss": 1.6407,
+      "step": 5387
+    },
+    {
+      "epoch": 0.007043109634864177,
+      "grad_norm": 1.5607740879058838,
+      "learning_rate": 2.30980980980981e-05,
+      "loss": 1.6428,
+      "step": 5388
+    },
+    {
+      "epoch": 0.007044416819280448,
+      "grad_norm": 1.4687775373458862,
+      "learning_rate": 2.3093093093093093e-05,
+      "loss": 1.2689,
+      "step": 5389
+    },
+    {
+      "epoch": 0.007045724003696717,
+      "grad_norm": 1.3042832612991333,
+      "learning_rate": 2.308808808808809e-05,
+      "loss": 1.21,
+      "step": 5390
+    },
+    {
+      "epoch": 0.0070470311881129875,
+      "grad_norm": 1.397918701171875,
+      "learning_rate": 2.3083083083083086e-05,
+      "loss": 0.8934,
+      "step": 5391
+    },
+    {
+      "epoch": 0.007048338372529258,
+      "grad_norm": 1.1901936531066895,
+      "learning_rate": 2.307807807807808e-05,
+      "loss": 0.5063,
+      "step": 5392
+    },
+    {
+      "epoch": 0.007049645556945528,
+      "grad_norm": 1.1835185289382935,
+      "learning_rate": 2.3073073073073072e-05,
+      "loss": 0.6675,
+      "step": 5393
+    },
+    {
+      "epoch": 0.007050952741361798,
+      "grad_norm": 1.3358733654022217,
+      "learning_rate": 2.306806806806807e-05,
+      "loss": 1.456,
+      "step": 5394
+    },
+    {
+      "epoch": 0.007052259925778069,
+      "grad_norm": 1.4109305143356323,
+      "learning_rate": 2.3063063063063065e-05,
+      "loss": 1.1308,
+      "step": 5395
+    },
+    {
+      "epoch": 0.007053567110194339,
+      "grad_norm": 1.788608193397522,
+      "learning_rate": 2.305805805805806e-05,
+      "loss": 2.2589,
+      "step": 5396
+    },
+    {
+      "epoch": 0.007054874294610609,
+      "grad_norm": 1.2691082954406738,
+      "learning_rate": 2.3053053053053054e-05,
+      "loss": 1.414,
+      "step": 5397
+    },
+    {
+      "epoch": 0.00705618147902688,
+      "grad_norm": 1.2861449718475342,
+      "learning_rate": 2.3048048048048047e-05,
+      "loss": 1.5826,
+      "step": 5398
+    },
+    {
+      "epoch": 0.00705748866344315,
+      "grad_norm": 1.6571067571640015,
+      "learning_rate": 2.3043043043043044e-05,
+      "loss": 2.3649,
+      "step": 5399
+    },
+    {
+      "epoch": 0.00705879584785942,
+      "grad_norm": 1.5016894340515137,
+      "learning_rate": 2.303803803803804e-05,
+      "loss": 2.0084,
+      "step": 5400
+    },
+    {
+      "epoch": 0.007060103032275691,
+      "grad_norm": 1.6949189901351929,
+      "learning_rate": 2.3033033033033033e-05,
+      "loss": 1.639,
+      "step": 5401
+    },
+    {
+      "epoch": 0.007061410216691961,
+      "grad_norm": 1.7690269947052002,
+      "learning_rate": 2.302802802802803e-05,
+      "loss": 1.3272,
+      "step": 5402
+    },
+    {
+      "epoch": 0.007062717401108231,
+      "grad_norm": 1.1733324527740479,
+      "learning_rate": 2.3023023023023026e-05,
+      "loss": 1.771,
+      "step": 5403
+    },
+    {
+      "epoch": 0.007064024585524502,
+      "grad_norm": 0.9945363998413086,
+      "learning_rate": 2.301801801801802e-05,
+      "loss": 0.7623,
+      "step": 5404
+    },
+    {
+      "epoch": 0.007065331769940771,
+      "grad_norm": 1.6948235034942627,
+      "learning_rate": 2.301301301301301e-05,
+      "loss": 2.1988,
+      "step": 5405
+    },
+    {
+      "epoch": 0.007066638954357041,
+      "grad_norm": 0.9905794858932495,
+      "learning_rate": 2.3008008008008008e-05,
+      "loss": 0.2773,
+      "step": 5406
+    },
+    {
+      "epoch": 0.007067946138773312,
+      "grad_norm": 1.2560864686965942,
+      "learning_rate": 2.3003003003003004e-05,
+      "loss": 1.2608,
+      "step": 5407
+    },
+    {
+      "epoch": 0.007069253323189582,
+      "grad_norm": 1.1086689233779907,
+      "learning_rate": 2.2997997997998e-05,
+      "loss": 0.6752,
+      "step": 5408
+    },
+    {
+      "epoch": 0.007070560507605852,
+      "grad_norm": 1.1074962615966797,
+      "learning_rate": 2.2992992992992997e-05,
+      "loss": 1.2429,
+      "step": 5409
+    },
+    {
+      "epoch": 0.007071867692022123,
+      "grad_norm": 1.1979008913040161,
+      "learning_rate": 2.298798798798799e-05,
+      "loss": 0.9813,
+      "step": 5410
+    },
+    {
+      "epoch": 0.007073174876438393,
+      "grad_norm": 1.4056764841079712,
+      "learning_rate": 2.2982982982982983e-05,
+      "loss": 1.4439,
+      "step": 5411
+    },
+    {
+      "epoch": 0.007074482060854663,
+      "grad_norm": 1.2614349126815796,
+      "learning_rate": 2.297797797797798e-05,
+      "loss": 1.3821,
+      "step": 5412
+    },
+    {
+      "epoch": 0.007075789245270934,
+      "grad_norm": 1.2867610454559326,
+      "learning_rate": 2.2972972972972976e-05,
+      "loss": 1.2718,
+      "step": 5413
+    },
+    {
+      "epoch": 0.007077096429687204,
+      "grad_norm": 1.3679802417755127,
+      "learning_rate": 2.296796796796797e-05,
+      "loss": 1.4612,
+      "step": 5414
+    },
+    {
+      "epoch": 0.007078403614103474,
+      "grad_norm": 1.2212283611297607,
+      "learning_rate": 2.2962962962962965e-05,
+      "loss": 1.4899,
+      "step": 5415
+    },
+    {
+      "epoch": 0.007079710798519745,
+      "grad_norm": 1.5309759378433228,
+      "learning_rate": 2.2957957957957958e-05,
+      "loss": 1.5519,
+      "step": 5416
+    },
+    {
+      "epoch": 0.007081017982936015,
+      "grad_norm": 2.320669174194336,
+      "learning_rate": 2.2952952952952954e-05,
+      "loss": 2.3762,
+      "step": 5417
+    },
+    {
+      "epoch": 0.007082325167352285,
+      "grad_norm": 2.0511152744293213,
+      "learning_rate": 2.2947947947947947e-05,
+      "loss": 2.1822,
+      "step": 5418
+    },
+    {
+      "epoch": 0.0070836323517685556,
+      "grad_norm": 1.1022928953170776,
+      "learning_rate": 2.2942942942942944e-05,
+      "loss": 1.0725,
+      "step": 5419
+    },
+    {
+      "epoch": 0.007084939536184825,
+      "grad_norm": 1.3031129837036133,
+      "learning_rate": 2.293793793793794e-05,
+      "loss": 1.6002,
+      "step": 5420
+    },
+    {
+      "epoch": 0.007086246720601095,
+      "grad_norm": 1.4437355995178223,
+      "learning_rate": 2.2932932932932936e-05,
+      "loss": 2.3003,
+      "step": 5421
+    },
+    {
+      "epoch": 0.007087553905017366,
+      "grad_norm": 1.2483431100845337,
+      "learning_rate": 2.292792792792793e-05,
+      "loss": 0.9758,
+      "step": 5422
+    },
+    {
+      "epoch": 0.007088861089433636,
+      "grad_norm": 1.3714990615844727,
+      "learning_rate": 2.2922922922922922e-05,
+      "loss": 2.4992,
+      "step": 5423
+    },
+    {
+      "epoch": 0.007090168273849906,
+      "grad_norm": 1.3193145990371704,
+      "learning_rate": 2.291791791791792e-05,
+      "loss": 1.6544,
+      "step": 5424
+    },
+    {
+      "epoch": 0.007091475458266177,
+      "grad_norm": 1.3587726354599,
+      "learning_rate": 2.2912912912912915e-05,
+      "loss": 1.6289,
+      "step": 5425
+    },
+    {
+      "epoch": 0.007092782642682447,
+      "grad_norm": 1.4263043403625488,
+      "learning_rate": 2.2907907907907908e-05,
+      "loss": 2.2381,
+      "step": 5426
+    },
+    {
+      "epoch": 0.007094089827098717,
+      "grad_norm": 1.4066418409347534,
+      "learning_rate": 2.2902902902902904e-05,
+      "loss": 2.0508,
+      "step": 5427
+    },
+    {
+      "epoch": 0.0070953970115149876,
+      "grad_norm": 1.551438331604004,
+      "learning_rate": 2.28978978978979e-05,
+      "loss": 1.8078,
+      "step": 5428
+    },
+    {
+      "epoch": 0.007096704195931258,
+      "grad_norm": 1.0236139297485352,
+      "learning_rate": 2.2892892892892894e-05,
+      "loss": 0.9164,
+      "step": 5429
+    },
+    {
+      "epoch": 0.007098011380347528,
+      "grad_norm": 1.161881685256958,
+      "learning_rate": 2.2887887887887887e-05,
+      "loss": 1.1839,
+      "step": 5430
+    },
+    {
+      "epoch": 0.0070993185647637985,
+      "grad_norm": 1.0907707214355469,
+      "learning_rate": 2.2882882882882883e-05,
+      "loss": 0.6145,
+      "step": 5431
+    },
+    {
+      "epoch": 0.007100625749180069,
+      "grad_norm": 0.6749468445777893,
+      "learning_rate": 2.287787787787788e-05,
+      "loss": 0.5603,
+      "step": 5432
+    },
+    {
+      "epoch": 0.007101932933596339,
+      "grad_norm": 1.5087850093841553,
+      "learning_rate": 2.2872872872872876e-05,
+      "loss": 2.2435,
+      "step": 5433
+    },
+    {
+      "epoch": 0.0071032401180126095,
+      "grad_norm": 1.0757808685302734,
+      "learning_rate": 2.286786786786787e-05,
+      "loss": 0.9049,
+      "step": 5434
+    },
+    {
+      "epoch": 0.007104547302428879,
+      "grad_norm": 1.811165690422058,
+      "learning_rate": 2.2862862862862862e-05,
+      "loss": 2.7105,
+      "step": 5435
+    },
+    {
+      "epoch": 0.007105854486845149,
+      "grad_norm": 1.920759677886963,
+      "learning_rate": 2.2857857857857858e-05,
+      "loss": 2.4386,
+      "step": 5436
+    },
+    {
+      "epoch": 0.00710716167126142,
+      "grad_norm": 1.0408014059066772,
+      "learning_rate": 2.2852852852852854e-05,
+      "loss": 0.9336,
+      "step": 5437
+    },
+    {
+      "epoch": 0.00710846885567769,
+      "grad_norm": 1.4687211513519287,
+      "learning_rate": 2.284784784784785e-05,
+      "loss": 2.1742,
+      "step": 5438
+    },
+    {
+      "epoch": 0.00710977604009396,
+      "grad_norm": 2.6432600021362305,
+      "learning_rate": 2.2842842842842844e-05,
+      "loss": 1.9289,
+      "step": 5439
+    },
+    {
+      "epoch": 0.0071110832245102305,
+      "grad_norm": 1.649970293045044,
+      "learning_rate": 2.283783783783784e-05,
+      "loss": 1.7329,
+      "step": 5440
+    },
+    {
+      "epoch": 0.007112390408926501,
+      "grad_norm": 1.2329566478729248,
+      "learning_rate": 2.2832832832832833e-05,
+      "loss": 1.5533,
+      "step": 5441
+    },
+    {
+      "epoch": 0.007113697593342771,
+      "grad_norm": 1.4443047046661377,
+      "learning_rate": 2.282782782782783e-05,
+      "loss": 1.7416,
+      "step": 5442
+    },
+    {
+      "epoch": 0.0071150047777590415,
+      "grad_norm": 1.2948952913284302,
+      "learning_rate": 2.2822822822822822e-05,
+      "loss": 2.1151,
+      "step": 5443
+    },
+    {
+      "epoch": 0.007116311962175312,
+      "grad_norm": 1.8255547285079956,
+      "learning_rate": 2.281781781781782e-05,
+      "loss": 1.8391,
+      "step": 5444
+    },
+    {
+      "epoch": 0.007117619146591582,
+      "grad_norm": 1.6529966592788696,
+      "learning_rate": 2.2812812812812815e-05,
+      "loss": 2.2177,
+      "step": 5445
+    },
+    {
+      "epoch": 0.0071189263310078525,
+      "grad_norm": 1.3106571435928345,
+      "learning_rate": 2.280780780780781e-05,
+      "loss": 1.3415,
+      "step": 5446
+    },
+    {
+      "epoch": 0.007120233515424123,
+      "grad_norm": 1.209762454032898,
+      "learning_rate": 2.2802802802802805e-05,
+      "loss": 1.7385,
+      "step": 5447
+    },
+    {
+      "epoch": 0.007121540699840393,
+      "grad_norm": 1.5051172971725464,
+      "learning_rate": 2.2797797797797797e-05,
+      "loss": 1.2906,
+      "step": 5448
+    },
+    {
+      "epoch": 0.007122847884256663,
+      "grad_norm": 1.3002114295959473,
+      "learning_rate": 2.2792792792792794e-05,
+      "loss": 1.3257,
+      "step": 5449
+    },
+    {
+      "epoch": 0.007124155068672934,
+      "grad_norm": 1.5771403312683105,
+      "learning_rate": 2.278778778778779e-05,
+      "loss": 1.8958,
+      "step": 5450
+    },
+    {
+      "epoch": 0.007125462253089203,
+      "grad_norm": 1.506068229675293,
+      "learning_rate": 2.2782782782782783e-05,
+      "loss": 1.7915,
+      "step": 5451
+    },
+    {
+      "epoch": 0.0071267694375054735,
+      "grad_norm": 1.1082429885864258,
+      "learning_rate": 2.277777777777778e-05,
+      "loss": 0.9029,
+      "step": 5452
+    },
+    {
+      "epoch": 0.007128076621921744,
+      "grad_norm": 1.2704278230667114,
+      "learning_rate": 2.2772772772772773e-05,
+      "loss": 0.961,
+      "step": 5453
+    },
+    {
+      "epoch": 0.007129383806338014,
+      "grad_norm": 1.5671616792678833,
+      "learning_rate": 2.276776776776777e-05,
+      "loss": 1.5356,
+      "step": 5454
+    },
+    {
+      "epoch": 0.0071306909907542845,
+      "grad_norm": 0.8593857288360596,
+      "learning_rate": 2.2762762762762762e-05,
+      "loss": 0.327,
+      "step": 5455
+    },
+    {
+      "epoch": 0.007131998175170555,
+      "grad_norm": 0.8990904688835144,
+      "learning_rate": 2.2757757757757758e-05,
+      "loss": 0.6839,
+      "step": 5456
+    },
+    {
+      "epoch": 0.007133305359586825,
+      "grad_norm": 1.606475830078125,
+      "learning_rate": 2.2752752752752755e-05,
+      "loss": 1.4865,
+      "step": 5457
+    },
+    {
+      "epoch": 0.007134612544003095,
+      "grad_norm": 1.265170693397522,
+      "learning_rate": 2.274774774774775e-05,
+      "loss": 1.2553,
+      "step": 5458
+    },
+    {
+      "epoch": 0.007135919728419366,
+      "grad_norm": 1.5988065004348755,
+      "learning_rate": 2.2742742742742744e-05,
+      "loss": 1.5838,
+      "step": 5459
+    },
+    {
+      "epoch": 0.007137226912835636,
+      "grad_norm": 1.3256183862686157,
+      "learning_rate": 2.2737737737737737e-05,
+      "loss": 2.3243,
+      "step": 5460
+    },
+    {
+      "epoch": 0.007138534097251906,
+      "grad_norm": 1.9125336408615112,
+      "learning_rate": 2.2732732732732733e-05,
+      "loss": 1.1488,
+      "step": 5461
+    },
+    {
+      "epoch": 0.007139841281668177,
+      "grad_norm": 1.871177315711975,
+      "learning_rate": 2.272772772772773e-05,
+      "loss": 1.3848,
+      "step": 5462
+    },
+    {
+      "epoch": 0.007141148466084447,
+      "grad_norm": 1.4954242706298828,
+      "learning_rate": 2.2722722722722726e-05,
+      "loss": 1.9139,
+      "step": 5463
+    },
+    {
+      "epoch": 0.007142455650500717,
+      "grad_norm": 1.3875888586044312,
+      "learning_rate": 2.271771771771772e-05,
+      "loss": 1.7327,
+      "step": 5464
+    },
+    {
+      "epoch": 0.007143762834916988,
+      "grad_norm": 2.3641231060028076,
+      "learning_rate": 2.2712712712712715e-05,
+      "loss": 2.4267,
+      "step": 5465
+    },
+    {
+      "epoch": 0.007145070019333257,
+      "grad_norm": 1.6929700374603271,
+      "learning_rate": 2.2707707707707708e-05,
+      "loss": 1.7361,
+      "step": 5466
+    },
+    {
+      "epoch": 0.007146377203749527,
+      "grad_norm": 1.2979222536087036,
+      "learning_rate": 2.2702702702702705e-05,
+      "loss": 2.0836,
+      "step": 5467
+    },
+    {
+      "epoch": 0.007147684388165798,
+      "grad_norm": 1.3312759399414062,
+      "learning_rate": 2.2697697697697698e-05,
+      "loss": 2.3369,
+      "step": 5468
+    },
+    {
+      "epoch": 0.007148991572582068,
+      "grad_norm": 1.306383490562439,
+      "learning_rate": 2.2692692692692694e-05,
+      "loss": 1.1612,
+      "step": 5469
+    },
+    {
+      "epoch": 0.007150298756998338,
+      "grad_norm": 1.497333288192749,
+      "learning_rate": 2.268768768768769e-05,
+      "loss": 2.3773,
+      "step": 5470
+    },
+    {
+      "epoch": 0.007151605941414609,
+      "grad_norm": 1.2466745376586914,
+      "learning_rate": 2.2682682682682683e-05,
+      "loss": 1.3352,
+      "step": 5471
+    },
+    {
+      "epoch": 0.007152913125830879,
+      "grad_norm": 1.4587796926498413,
+      "learning_rate": 2.2677677677677676e-05,
+      "loss": 0.8688,
+      "step": 5472
+    },
+    {
+      "epoch": 0.007154220310247149,
+      "grad_norm": 1.250903606414795,
+      "learning_rate": 2.2672672672672673e-05,
+      "loss": 1.3538,
+      "step": 5473
+    },
+    {
+      "epoch": 0.00715552749466342,
+      "grad_norm": 1.394606113433838,
+      "learning_rate": 2.266766766766767e-05,
+      "loss": 1.4583,
+      "step": 5474
+    },
+    {
+      "epoch": 0.00715683467907969,
+      "grad_norm": 1.630262017250061,
+      "learning_rate": 2.2662662662662665e-05,
+      "loss": 1.5777,
+      "step": 5475
+    },
+    {
+      "epoch": 0.00715814186349596,
+      "grad_norm": 1.566664695739746,
+      "learning_rate": 2.265765765765766e-05,
+      "loss": 1.9633,
+      "step": 5476
+    },
+    {
+      "epoch": 0.007159449047912231,
+      "grad_norm": 1.4662564992904663,
+      "learning_rate": 2.2652652652652655e-05,
+      "loss": 1.4438,
+      "step": 5477
+    },
+    {
+      "epoch": 0.007160756232328501,
+      "grad_norm": 1.5617560148239136,
+      "learning_rate": 2.2647647647647648e-05,
+      "loss": 1.6556,
+      "step": 5478
+    },
+    {
+      "epoch": 0.007162063416744771,
+      "grad_norm": 1.4973769187927246,
+      "learning_rate": 2.2642642642642644e-05,
+      "loss": 2.5385,
+      "step": 5479
+    },
+    {
+      "epoch": 0.007163370601161042,
+      "grad_norm": 1.4423776865005493,
+      "learning_rate": 2.2637637637637637e-05,
+      "loss": 1.8542,
+      "step": 5480
+    },
+    {
+      "epoch": 0.007164677785577311,
+      "grad_norm": 1.5184099674224854,
+      "learning_rate": 2.2632632632632633e-05,
+      "loss": 1.2773,
+      "step": 5481
+    },
+    {
+      "epoch": 0.007165984969993581,
+      "grad_norm": 1.2675577402114868,
+      "learning_rate": 2.262762762762763e-05,
+      "loss": 2.1396,
+      "step": 5482
+    },
+    {
+      "epoch": 0.007167292154409852,
+      "grad_norm": 1.3148930072784424,
+      "learning_rate": 2.2622622622622626e-05,
+      "loss": 1.6735,
+      "step": 5483
+    },
+    {
+      "epoch": 0.007168599338826122,
+      "grad_norm": 1.704709768295288,
+      "learning_rate": 2.261761761761762e-05,
+      "loss": 1.9475,
+      "step": 5484
+    },
+    {
+      "epoch": 0.007169906523242392,
+      "grad_norm": 1.5002299547195435,
+      "learning_rate": 2.2612612612612612e-05,
+      "loss": 1.9891,
+      "step": 5485
+    },
+    {
+      "epoch": 0.007171213707658663,
+      "grad_norm": 1.306270956993103,
+      "learning_rate": 2.260760760760761e-05,
+      "loss": 2.4084,
+      "step": 5486
+    },
+    {
+      "epoch": 0.007172520892074933,
+      "grad_norm": 1.3288187980651855,
+      "learning_rate": 2.2602602602602605e-05,
+      "loss": 1.5222,
+      "step": 5487
+    },
+    {
+      "epoch": 0.007173828076491203,
+      "grad_norm": 0.9205089807510376,
+      "learning_rate": 2.2597597597597598e-05,
+      "loss": 0.5275,
+      "step": 5488
+    },
+    {
+      "epoch": 0.007175135260907474,
+      "grad_norm": 1.4213277101516724,
+      "learning_rate": 2.2592592592592594e-05,
+      "loss": 1.7643,
+      "step": 5489
+    },
+    {
+      "epoch": 0.007176442445323744,
+      "grad_norm": 1.1442643404006958,
+      "learning_rate": 2.2587587587587587e-05,
+      "loss": 0.9043,
+      "step": 5490
+    },
+    {
+      "epoch": 0.007177749629740014,
+      "grad_norm": 1.8178114891052246,
+      "learning_rate": 2.2582582582582583e-05,
+      "loss": 2.5482,
+      "step": 5491
+    },
+    {
+      "epoch": 0.0071790568141562846,
+      "grad_norm": 1.181736707687378,
+      "learning_rate": 2.257757757757758e-05,
+      "loss": 1.2869,
+      "step": 5492
+    },
+    {
+      "epoch": 0.007180363998572555,
+      "grad_norm": 1.4033716917037964,
+      "learning_rate": 2.2572572572572573e-05,
+      "loss": 1.7895,
+      "step": 5493
+    },
+    {
+      "epoch": 0.007181671182988825,
+      "grad_norm": 1.1112866401672363,
+      "learning_rate": 2.256756756756757e-05,
+      "loss": 1.1602,
+      "step": 5494
+    },
+    {
+      "epoch": 0.0071829783674050955,
+      "grad_norm": 1.2490532398223877,
+      "learning_rate": 2.2562562562562566e-05,
+      "loss": 0.9747,
+      "step": 5495
+    },
+    {
+      "epoch": 0.007184285551821366,
+      "grad_norm": 1.004259705543518,
+      "learning_rate": 2.255755755755756e-05,
+      "loss": 1.1974,
+      "step": 5496
+    },
+    {
+      "epoch": 0.007185592736237635,
+      "grad_norm": 1.4341120719909668,
+      "learning_rate": 2.255255255255255e-05,
+      "loss": 1.4908,
+      "step": 5497
+    },
+    {
+      "epoch": 0.007186899920653906,
+      "grad_norm": 1.3087522983551025,
+      "learning_rate": 2.2547547547547548e-05,
+      "loss": 1.4094,
+      "step": 5498
+    },
+    {
+      "epoch": 0.007188207105070176,
+      "grad_norm": 1.2443864345550537,
+      "learning_rate": 2.2542542542542544e-05,
+      "loss": 1.0241,
+      "step": 5499
+    },
+    {
+      "epoch": 0.007189514289486446,
+      "grad_norm": 1.448864459991455,
+      "learning_rate": 2.253753753753754e-05,
+      "loss": 1.6063,
+      "step": 5500
+    },
+    {
+      "epoch": 0.007190821473902717,
+      "grad_norm": 1.3612865209579468,
+      "learning_rate": 2.2532532532532534e-05,
+      "loss": 2.0153,
+      "step": 5501
+    },
+    {
+      "epoch": 0.007192128658318987,
+      "grad_norm": 1.4161765575408936,
+      "learning_rate": 2.252752752752753e-05,
+      "loss": 2.01,
+      "step": 5502
+    },
+    {
+      "epoch": 0.007193435842735257,
+      "grad_norm": 0.526180624961853,
+      "learning_rate": 2.2522522522522523e-05,
+      "loss": 0.3856,
+      "step": 5503
+    },
+    {
+      "epoch": 0.0071947430271515275,
+      "grad_norm": 1.4418623447418213,
+      "learning_rate": 2.251751751751752e-05,
+      "loss": 1.9625,
+      "step": 5504
+    },
+    {
+      "epoch": 0.007196050211567798,
+      "grad_norm": 1.1156922578811646,
+      "learning_rate": 2.2512512512512512e-05,
+      "loss": 1.687,
+      "step": 5505
+    },
+    {
+      "epoch": 0.007197357395984068,
+      "grad_norm": 1.538995385169983,
+      "learning_rate": 2.250750750750751e-05,
+      "loss": 1.7003,
+      "step": 5506
+    },
+    {
+      "epoch": 0.0071986645804003385,
+      "grad_norm": 1.3299857378005981,
+      "learning_rate": 2.2502502502502505e-05,
+      "loss": 1.4268,
+      "step": 5507
+    },
+    {
+      "epoch": 0.007199971764816609,
+      "grad_norm": 1.332794189453125,
+      "learning_rate": 2.2497497497497498e-05,
+      "loss": 0.9311,
+      "step": 5508
+    },
+    {
+      "epoch": 0.007201278949232879,
+      "grad_norm": 1.3024089336395264,
+      "learning_rate": 2.249249249249249e-05,
+      "loss": 1.7021,
+      "step": 5509
+    },
+    {
+      "epoch": 0.0072025861336491494,
+      "grad_norm": 1.540079116821289,
+      "learning_rate": 2.2487487487487487e-05,
+      "loss": 1.6418,
+      "step": 5510
+    },
+    {
+      "epoch": 0.00720389331806542,
+      "grad_norm": 1.0778299570083618,
+      "learning_rate": 2.2482482482482484e-05,
+      "loss": 1.0402,
+      "step": 5511
+    },
+    {
+      "epoch": 0.007205200502481689,
+      "grad_norm": 1.2043464183807373,
+      "learning_rate": 2.247747747747748e-05,
+      "loss": 2.3608,
+      "step": 5512
+    },
+    {
+      "epoch": 0.0072065076868979595,
+      "grad_norm": 1.5667375326156616,
+      "learning_rate": 2.2472472472472473e-05,
+      "loss": 1.8256,
+      "step": 5513
+    },
+    {
+      "epoch": 0.00720781487131423,
+      "grad_norm": 1.2886743545532227,
+      "learning_rate": 2.246746746746747e-05,
+      "loss": 1.9129,
+      "step": 5514
+    },
+    {
+      "epoch": 0.0072091220557305,
+      "grad_norm": 2.9183692932128906,
+      "learning_rate": 2.2462462462462462e-05,
+      "loss": 1.9956,
+      "step": 5515
+    },
+    {
+      "epoch": 0.0072104292401467705,
+      "grad_norm": 1.7904489040374756,
+      "learning_rate": 2.245745745745746e-05,
+      "loss": 2.1863,
+      "step": 5516
+    },
+    {
+      "epoch": 0.007211736424563041,
+      "grad_norm": 1.366678237915039,
+      "learning_rate": 2.2452452452452455e-05,
+      "loss": 1.4414,
+      "step": 5517
+    },
+    {
+      "epoch": 0.007213043608979311,
+      "grad_norm": 1.5434590578079224,
+      "learning_rate": 2.2447447447447448e-05,
+      "loss": 1.9691,
+      "step": 5518
+    },
+    {
+      "epoch": 0.0072143507933955815,
+      "grad_norm": 1.4171375036239624,
+      "learning_rate": 2.2442442442442444e-05,
+      "loss": 1.7875,
+      "step": 5519
+    },
+    {
+      "epoch": 0.007215657977811852,
+      "grad_norm": 1.9429006576538086,
+      "learning_rate": 2.243743743743744e-05,
+      "loss": 2.3652,
+      "step": 5520
+    },
+    {
+      "epoch": 0.007216965162228122,
+      "grad_norm": 1.8096050024032593,
+      "learning_rate": 2.2432432432432434e-05,
+      "loss": 1.9208,
+      "step": 5521
+    },
+    {
+      "epoch": 0.007218272346644392,
+      "grad_norm": 1.5811213254928589,
+      "learning_rate": 2.2427427427427427e-05,
+      "loss": 1.8206,
+      "step": 5522
+    },
+    {
+      "epoch": 0.007219579531060663,
+      "grad_norm": 1.2713427543640137,
+      "learning_rate": 2.2422422422422423e-05,
+      "loss": 1.5803,
+      "step": 5523
+    },
+    {
+      "epoch": 0.007220886715476933,
+      "grad_norm": 1.561004877090454,
+      "learning_rate": 2.241741741741742e-05,
+      "loss": 1.4073,
+      "step": 5524
+    },
+    {
+      "epoch": 0.007222193899893203,
+      "grad_norm": 1.3025965690612793,
+      "learning_rate": 2.2412412412412416e-05,
+      "loss": 1.5616,
+      "step": 5525
+    },
+    {
+      "epoch": 0.007223501084309474,
+      "grad_norm": 1.2230075597763062,
+      "learning_rate": 2.240740740740741e-05,
+      "loss": 0.6175,
+      "step": 5526
+    },
+    {
+      "epoch": 0.007224808268725743,
+      "grad_norm": 0.9776055216789246,
+      "learning_rate": 2.24024024024024e-05,
+      "loss": 0.8579,
+      "step": 5527
+    },
+    {
+      "epoch": 0.0072261154531420135,
+      "grad_norm": 0.8296329379081726,
+      "learning_rate": 2.2397397397397398e-05,
+      "loss": 0.5644,
+      "step": 5528
+    },
+    {
+      "epoch": 0.007227422637558284,
+      "grad_norm": 1.4586780071258545,
+      "learning_rate": 2.2392392392392394e-05,
+      "loss": 1.9926,
+      "step": 5529
+    },
+    {
+      "epoch": 0.007228729821974554,
+      "grad_norm": 1.0929583311080933,
+      "learning_rate": 2.2387387387387387e-05,
+      "loss": 1.0397,
+      "step": 5530
+    },
+    {
+      "epoch": 0.007230037006390824,
+      "grad_norm": 1.3129661083221436,
+      "learning_rate": 2.2382382382382384e-05,
+      "loss": 1.7398,
+      "step": 5531
+    },
+    {
+      "epoch": 0.007231344190807095,
+      "grad_norm": 1.472814679145813,
+      "learning_rate": 2.237737737737738e-05,
+      "loss": 1.4665,
+      "step": 5532
+    },
+    {
+      "epoch": 0.007232651375223365,
+      "grad_norm": 1.0274913311004639,
+      "learning_rate": 2.2372372372372373e-05,
+      "loss": 0.7996,
+      "step": 5533
+    },
+    {
+      "epoch": 0.007233958559639635,
+      "grad_norm": 1.3445357084274292,
+      "learning_rate": 2.2367367367367366e-05,
+      "loss": 1.7511,
+      "step": 5534
+    },
+    {
+      "epoch": 0.007235265744055906,
+      "grad_norm": 1.4070007801055908,
+      "learning_rate": 2.2362362362362362e-05,
+      "loss": 1.4836,
+      "step": 5535
+    },
+    {
+      "epoch": 0.007236572928472176,
+      "grad_norm": 1.311958909034729,
+      "learning_rate": 2.235735735735736e-05,
+      "loss": 2.4003,
+      "step": 5536
+    },
+    {
+      "epoch": 0.007237880112888446,
+      "grad_norm": 0.788358747959137,
+      "learning_rate": 2.2352352352352355e-05,
+      "loss": 0.4855,
+      "step": 5537
+    },
+    {
+      "epoch": 0.007239187297304717,
+      "grad_norm": 1.282039999961853,
+      "learning_rate": 2.2347347347347348e-05,
+      "loss": 1.2066,
+      "step": 5538
+    },
+    {
+      "epoch": 0.007240494481720987,
+      "grad_norm": 1.0228936672210693,
+      "learning_rate": 2.2342342342342344e-05,
+      "loss": 0.7015,
+      "step": 5539
+    },
+    {
+      "epoch": 0.007241801666137257,
+      "grad_norm": 1.1706244945526123,
+      "learning_rate": 2.2337337337337337e-05,
+      "loss": 1.103,
+      "step": 5540
+    },
+    {
+      "epoch": 0.007243108850553528,
+      "grad_norm": 1.3907966613769531,
+      "learning_rate": 2.2332332332332334e-05,
+      "loss": 1.5434,
+      "step": 5541
+    },
+    {
+      "epoch": 0.007244416034969797,
+      "grad_norm": 1.5349805355072021,
+      "learning_rate": 2.2327327327327327e-05,
+      "loss": 1.7228,
+      "step": 5542
+    },
+    {
+      "epoch": 0.007245723219386067,
+      "grad_norm": 1.6820098161697388,
+      "learning_rate": 2.2322322322322323e-05,
+      "loss": 2.2776,
+      "step": 5543
+    },
+    {
+      "epoch": 0.007247030403802338,
+      "grad_norm": 1.5607692003250122,
+      "learning_rate": 2.231731731731732e-05,
+      "loss": 2.0336,
+      "step": 5544
+    },
+    {
+      "epoch": 0.007248337588218608,
+      "grad_norm": 1.3325567245483398,
+      "learning_rate": 2.2312312312312312e-05,
+      "loss": 1.5904,
+      "step": 5545
+    },
+    {
+      "epoch": 0.007249644772634878,
+      "grad_norm": 1.4808861017227173,
+      "learning_rate": 2.230730730730731e-05,
+      "loss": 1.7103,
+      "step": 5546
+    },
+    {
+      "epoch": 0.007250951957051149,
+      "grad_norm": 1.8867589235305786,
+      "learning_rate": 2.2302302302302302e-05,
+      "loss": 2.8215,
+      "step": 5547
+    },
+    {
+      "epoch": 0.007252259141467419,
+      "grad_norm": 1.5563528537750244,
+      "learning_rate": 2.2297297297297298e-05,
+      "loss": 1.9708,
+      "step": 5548
+    },
+    {
+      "epoch": 0.007253566325883689,
+      "grad_norm": 1.508362054824829,
+      "learning_rate": 2.2292292292292295e-05,
+      "loss": 1.8718,
+      "step": 5549
+    },
+    {
+      "epoch": 0.00725487351029996,
+      "grad_norm": 1.5985064506530762,
+      "learning_rate": 2.228728728728729e-05,
+      "loss": 1.6418,
+      "step": 5550
+    },
+    {
+      "epoch": 0.00725618069471623,
+      "grad_norm": 1.2306033372879028,
+      "learning_rate": 2.2282282282282284e-05,
+      "loss": 1.2528,
+      "step": 5551
+    },
+    {
+      "epoch": 0.0072574878791325,
+      "grad_norm": 1.0870139598846436,
+      "learning_rate": 2.2277277277277277e-05,
+      "loss": 0.8151,
+      "step": 5552
+    },
+    {
+      "epoch": 0.007258795063548771,
+      "grad_norm": 1.71553373336792,
+      "learning_rate": 2.2272272272272273e-05,
+      "loss": 2.1821,
+      "step": 5553
+    },
+    {
+      "epoch": 0.007260102247965041,
+      "grad_norm": 1.0520764589309692,
+      "learning_rate": 2.226726726726727e-05,
+      "loss": 0.5181,
+      "step": 5554
+    },
+    {
+      "epoch": 0.007261409432381311,
+      "grad_norm": 1.250373363494873,
+      "learning_rate": 2.2262262262262263e-05,
+      "loss": 1.3384,
+      "step": 5555
+    },
+    {
+      "epoch": 0.0072627166167975815,
+      "grad_norm": 1.1105518341064453,
+      "learning_rate": 2.225725725725726e-05,
+      "loss": 0.8758,
+      "step": 5556
+    },
+    {
+      "epoch": 0.007264023801213852,
+      "grad_norm": 1.8298919200897217,
+      "learning_rate": 2.2252252252252255e-05,
+      "loss": 1.601,
+      "step": 5557
+    },
+    {
+      "epoch": 0.007265330985630121,
+      "grad_norm": 1.2859467267990112,
+      "learning_rate": 2.2247247247247248e-05,
+      "loss": 1.5951,
+      "step": 5558
+    },
+    {
+      "epoch": 0.007266638170046392,
+      "grad_norm": 1.4340909719467163,
+      "learning_rate": 2.224224224224224e-05,
+      "loss": 1.3768,
+      "step": 5559
+    },
+    {
+      "epoch": 0.007267945354462662,
+      "grad_norm": 1.757991909980774,
+      "learning_rate": 2.2237237237237238e-05,
+      "loss": 2.4527,
+      "step": 5560
+    },
+    {
+      "epoch": 0.007269252538878932,
+      "grad_norm": 1.1364606618881226,
+      "learning_rate": 2.2232232232232234e-05,
+      "loss": 0.8462,
+      "step": 5561
+    },
+    {
+      "epoch": 0.007270559723295203,
+      "grad_norm": 1.5076777935028076,
+      "learning_rate": 2.222722722722723e-05,
+      "loss": 1.6313,
+      "step": 5562
+    },
+    {
+      "epoch": 0.007271866907711473,
+      "grad_norm": 1.467132329940796,
+      "learning_rate": 2.2222222222222223e-05,
+      "loss": 2.1217,
+      "step": 5563
+    },
+    {
+      "epoch": 0.007273174092127743,
+      "grad_norm": 1.1722973585128784,
+      "learning_rate": 2.2217217217217216e-05,
+      "loss": 1.7225,
+      "step": 5564
+    },
+    {
+      "epoch": 0.0072744812765440136,
+      "grad_norm": 1.3336807489395142,
+      "learning_rate": 2.2212212212212213e-05,
+      "loss": 1.2281,
+      "step": 5565
+    },
+    {
+      "epoch": 0.007275788460960284,
+      "grad_norm": 1.8226176500320435,
+      "learning_rate": 2.220720720720721e-05,
+      "loss": 1.3974,
+      "step": 5566
+    },
+    {
+      "epoch": 0.007277095645376554,
+      "grad_norm": 0.8084770441055298,
+      "learning_rate": 2.2202202202202202e-05,
+      "loss": 0.4369,
+      "step": 5567
+    },
+    {
+      "epoch": 0.0072784028297928245,
+      "grad_norm": 1.2906761169433594,
+      "learning_rate": 2.21971971971972e-05,
+      "loss": 1.7873,
+      "step": 5568
+    },
+    {
+      "epoch": 0.007279710014209095,
+      "grad_norm": 1.5542234182357788,
+      "learning_rate": 2.2192192192192195e-05,
+      "loss": 1.605,
+      "step": 5569
+    },
+    {
+      "epoch": 0.007281017198625365,
+      "grad_norm": 1.4804691076278687,
+      "learning_rate": 2.2187187187187188e-05,
+      "loss": 2.0715,
+      "step": 5570
+    },
+    {
+      "epoch": 0.0072823243830416355,
+      "grad_norm": 1.4808045625686646,
+      "learning_rate": 2.2182182182182184e-05,
+      "loss": 1.9569,
+      "step": 5571
+    },
+    {
+      "epoch": 0.007283631567457906,
+      "grad_norm": 1.3109846115112305,
+      "learning_rate": 2.2177177177177177e-05,
+      "loss": 1.1442,
+      "step": 5572
+    },
+    {
+      "epoch": 0.007284938751874175,
+      "grad_norm": 1.2424572706222534,
+      "learning_rate": 2.2172172172172173e-05,
+      "loss": 1.4589,
+      "step": 5573
+    },
+    {
+      "epoch": 0.007286245936290446,
+      "grad_norm": 1.8530339002609253,
+      "learning_rate": 2.216716716716717e-05,
+      "loss": 1.9876,
+      "step": 5574
+    },
+    {
+      "epoch": 0.007287553120706716,
+      "grad_norm": 1.5281034708023071,
+      "learning_rate": 2.2162162162162166e-05,
+      "loss": 1.5921,
+      "step": 5575
+    },
+    {
+      "epoch": 0.007288860305122986,
+      "grad_norm": 1.3007786273956299,
+      "learning_rate": 2.2157157157157156e-05,
+      "loss": 1.546,
+      "step": 5576
+    },
+    {
+      "epoch": 0.0072901674895392565,
+      "grad_norm": 1.1685339212417603,
+      "learning_rate": 2.2152152152152152e-05,
+      "loss": 1.5568,
+      "step": 5577
+    },
+    {
+      "epoch": 0.007291474673955527,
+      "grad_norm": 1.3943432569503784,
+      "learning_rate": 2.214714714714715e-05,
+      "loss": 1.5659,
+      "step": 5578
+    },
+    {
+      "epoch": 0.007292781858371797,
+      "grad_norm": 2.1932291984558105,
+      "learning_rate": 2.2142142142142145e-05,
+      "loss": 2.0952,
+      "step": 5579
+    },
+    {
+      "epoch": 0.0072940890427880675,
+      "grad_norm": 1.5557178258895874,
+      "learning_rate": 2.2137137137137138e-05,
+      "loss": 1.2778,
+      "step": 5580
+    },
+    {
+      "epoch": 0.007295396227204338,
+      "grad_norm": 1.2777018547058105,
+      "learning_rate": 2.2132132132132134e-05,
+      "loss": 1.5653,
+      "step": 5581
+    },
+    {
+      "epoch": 0.007296703411620608,
+      "grad_norm": 1.1967073678970337,
+      "learning_rate": 2.2127127127127127e-05,
+      "loss": 1.2725,
+      "step": 5582
+    },
+    {
+      "epoch": 0.0072980105960368784,
+      "grad_norm": 1.455257773399353,
+      "learning_rate": 2.2122122122122123e-05,
+      "loss": 1.6506,
+      "step": 5583
+    },
+    {
+      "epoch": 0.007299317780453149,
+      "grad_norm": 1.3821836709976196,
+      "learning_rate": 2.2117117117117116e-05,
+      "loss": 1.7495,
+      "step": 5584
+    },
+    {
+      "epoch": 0.007300624964869419,
+      "grad_norm": 1.9764699935913086,
+      "learning_rate": 2.2112112112112113e-05,
+      "loss": 2.162,
+      "step": 5585
+    },
+    {
+      "epoch": 0.007301932149285689,
+      "grad_norm": 0.9757213592529297,
+      "learning_rate": 2.210710710710711e-05,
+      "loss": 0.6293,
+      "step": 5586
+    },
+    {
+      "epoch": 0.00730323933370196,
+      "grad_norm": 1.7481871843338013,
+      "learning_rate": 2.2102102102102105e-05,
+      "loss": 1.3732,
+      "step": 5587
+    },
+    {
+      "epoch": 0.007304546518118229,
+      "grad_norm": 1.3009567260742188,
+      "learning_rate": 2.20970970970971e-05,
+      "loss": 2.4943,
+      "step": 5588
+    },
+    {
+      "epoch": 0.0073058537025344995,
+      "grad_norm": 1.1610480546951294,
+      "learning_rate": 2.209209209209209e-05,
+      "loss": 1.461,
+      "step": 5589
+    },
+    {
+      "epoch": 0.00730716088695077,
+      "grad_norm": 1.6352801322937012,
+      "learning_rate": 2.2087087087087088e-05,
+      "loss": 2.1292,
+      "step": 5590
+    },
+    {
+      "epoch": 0.00730846807136704,
+      "grad_norm": 1.4568318128585815,
+      "learning_rate": 2.2082082082082084e-05,
+      "loss": 1.4006,
+      "step": 5591
+    },
+    {
+      "epoch": 0.0073097752557833105,
+      "grad_norm": 1.2338224649429321,
+      "learning_rate": 2.2077077077077077e-05,
+      "loss": 1.3811,
+      "step": 5592
+    },
+    {
+      "epoch": 0.007311082440199581,
+      "grad_norm": 0.8923683762550354,
+      "learning_rate": 2.2072072072072073e-05,
+      "loss": 0.5123,
+      "step": 5593
+    },
+    {
+      "epoch": 0.007312389624615851,
+      "grad_norm": 1.9392383098602295,
+      "learning_rate": 2.206706706706707e-05,
+      "loss": 1.8415,
+      "step": 5594
+    },
+    {
+      "epoch": 0.007313696809032121,
+      "grad_norm": 1.855711579322815,
+      "learning_rate": 2.2062062062062063e-05,
+      "loss": 1.4282,
+      "step": 5595
+    },
+    {
+      "epoch": 0.007315003993448392,
+      "grad_norm": 1.4232537746429443,
+      "learning_rate": 2.205705705705706e-05,
+      "loss": 1.0717,
+      "step": 5596
+    },
+    {
+      "epoch": 0.007316311177864662,
+      "grad_norm": 1.1656882762908936,
+      "learning_rate": 2.2052052052052052e-05,
+      "loss": 1.2185,
+      "step": 5597
+    },
+    {
+      "epoch": 0.007317618362280932,
+      "grad_norm": 1.3809562921524048,
+      "learning_rate": 2.204704704704705e-05,
+      "loss": 1.5093,
+      "step": 5598
+    },
+    {
+      "epoch": 0.007318925546697203,
+      "grad_norm": 1.2399184703826904,
+      "learning_rate": 2.2042042042042045e-05,
+      "loss": 1.1607,
+      "step": 5599
+    },
+    {
+      "epoch": 0.007320232731113473,
+      "grad_norm": 1.0937659740447998,
+      "learning_rate": 2.2037037037037038e-05,
+      "loss": 0.7791,
+      "step": 5600
+    },
+    {
+      "epoch": 0.007321539915529743,
+      "grad_norm": 1.4154222011566162,
+      "learning_rate": 2.203203203203203e-05,
+      "loss": 1.7562,
+      "step": 5601
+    },
+    {
+      "epoch": 0.007322847099946014,
+      "grad_norm": 1.6267205476760864,
+      "learning_rate": 2.2027027027027027e-05,
+      "loss": 2.1229,
+      "step": 5602
+    },
+    {
+      "epoch": 0.007324154284362284,
+      "grad_norm": 1.299039602279663,
+      "learning_rate": 2.2022022022022024e-05,
+      "loss": 1.501,
+      "step": 5603
+    },
+    {
+      "epoch": 0.007325461468778553,
+      "grad_norm": 1.1801584959030151,
+      "learning_rate": 2.201701701701702e-05,
+      "loss": 0.7209,
+      "step": 5604
+    },
+    {
+      "epoch": 0.007326768653194824,
+      "grad_norm": 1.3467708826065063,
+      "learning_rate": 2.2012012012012013e-05,
+      "loss": 1.4847,
+      "step": 5605
+    },
+    {
+      "epoch": 0.007328075837611094,
+      "grad_norm": 1.9011162519454956,
+      "learning_rate": 2.200700700700701e-05,
+      "loss": 2.0973,
+      "step": 5606
+    },
+    {
+      "epoch": 0.007329383022027364,
+      "grad_norm": 1.1189218759536743,
+      "learning_rate": 2.2002002002002002e-05,
+      "loss": 0.8902,
+      "step": 5607
+    },
+    {
+      "epoch": 0.007330690206443635,
+      "grad_norm": 1.2982839345932007,
+      "learning_rate": 2.1996996996997e-05,
+      "loss": 1.6425,
+      "step": 5608
+    },
+    {
+      "epoch": 0.007331997390859905,
+      "grad_norm": 1.3885552883148193,
+      "learning_rate": 2.199199199199199e-05,
+      "loss": 1.4612,
+      "step": 5609
+    },
+    {
+      "epoch": 0.007333304575276175,
+      "grad_norm": 1.7174999713897705,
+      "learning_rate": 2.1986986986986988e-05,
+      "loss": 1.894,
+      "step": 5610
+    },
+    {
+      "epoch": 0.007334611759692446,
+      "grad_norm": 1.416738748550415,
+      "learning_rate": 2.1981981981981984e-05,
+      "loss": 2.1853,
+      "step": 5611
+    },
+    {
+      "epoch": 0.007335918944108716,
+      "grad_norm": 1.8563898801803589,
+      "learning_rate": 2.197697697697698e-05,
+      "loss": 2.121,
+      "step": 5612
+    },
+    {
+      "epoch": 0.007337226128524986,
+      "grad_norm": 1.4706101417541504,
+      "learning_rate": 2.197197197197197e-05,
+      "loss": 1.705,
+      "step": 5613
+    },
+    {
+      "epoch": 0.007338533312941257,
+      "grad_norm": 1.1593682765960693,
+      "learning_rate": 2.1966966966966967e-05,
+      "loss": 1.335,
+      "step": 5614
+    },
+    {
+      "epoch": 0.007339840497357527,
+      "grad_norm": 1.4033929109573364,
+      "learning_rate": 2.1961961961961963e-05,
+      "loss": 1.3484,
+      "step": 5615
+    },
+    {
+      "epoch": 0.007341147681773797,
+      "grad_norm": 1.4726946353912354,
+      "learning_rate": 2.195695695695696e-05,
+      "loss": 2.2119,
+      "step": 5616
+    },
+    {
+      "epoch": 0.007342454866190068,
+      "grad_norm": 1.4019861221313477,
+      "learning_rate": 2.1951951951951952e-05,
+      "loss": 1.2201,
+      "step": 5617
+    },
+    {
+      "epoch": 0.007343762050606338,
+      "grad_norm": 1.1956260204315186,
+      "learning_rate": 2.194694694694695e-05,
+      "loss": 1.3615,
+      "step": 5618
+    },
+    {
+      "epoch": 0.007345069235022607,
+      "grad_norm": 1.211209774017334,
+      "learning_rate": 2.194194194194194e-05,
+      "loss": 1.365,
+      "step": 5619
+    },
+    {
+      "epoch": 0.007346376419438878,
+      "grad_norm": 1.4523720741271973,
+      "learning_rate": 2.1936936936936938e-05,
+      "loss": 1.1292,
+      "step": 5620
+    },
+    {
+      "epoch": 0.007347683603855148,
+      "grad_norm": 1.3904895782470703,
+      "learning_rate": 2.193193193193193e-05,
+      "loss": 1.2571,
+      "step": 5621
+    },
+    {
+      "epoch": 0.007348990788271418,
+      "grad_norm": 1.260412573814392,
+      "learning_rate": 2.1926926926926927e-05,
+      "loss": 1.4779,
+      "step": 5622
+    },
+    {
+      "epoch": 0.007350297972687689,
+      "grad_norm": 1.3784288167953491,
+      "learning_rate": 2.1921921921921924e-05,
+      "loss": 1.9934,
+      "step": 5623
+    },
+    {
+      "epoch": 0.007351605157103959,
+      "grad_norm": 1.5240391492843628,
+      "learning_rate": 2.191691691691692e-05,
+      "loss": 1.5519,
+      "step": 5624
+    },
+    {
+      "epoch": 0.007352912341520229,
+      "grad_norm": 1.1035783290863037,
+      "learning_rate": 2.1911911911911913e-05,
+      "loss": 0.8692,
+      "step": 5625
+    },
+    {
+      "epoch": 0.0073542195259365,
+      "grad_norm": 1.416589617729187,
+      "learning_rate": 2.1906906906906906e-05,
+      "loss": 1.3776,
+      "step": 5626
+    },
+    {
+      "epoch": 0.00735552671035277,
+      "grad_norm": 0.8322731852531433,
+      "learning_rate": 2.1901901901901902e-05,
+      "loss": 0.5741,
+      "step": 5627
+    },
+    {
+      "epoch": 0.00735683389476904,
+      "grad_norm": 1.6225934028625488,
+      "learning_rate": 2.18968968968969e-05,
+      "loss": 2.223,
+      "step": 5628
+    },
+    {
+      "epoch": 0.0073581410791853106,
+      "grad_norm": 1.2560615539550781,
+      "learning_rate": 2.1891891891891895e-05,
+      "loss": 1.4002,
+      "step": 5629
+    },
+    {
+      "epoch": 0.007359448263601581,
+      "grad_norm": 3.4986822605133057,
+      "learning_rate": 2.1886886886886888e-05,
+      "loss": 1.3497,
+      "step": 5630
+    },
+    {
+      "epoch": 0.007360755448017851,
+      "grad_norm": 1.3134247064590454,
+      "learning_rate": 2.188188188188188e-05,
+      "loss": 1.9233,
+      "step": 5631
+    },
+    {
+      "epoch": 0.0073620626324341215,
+      "grad_norm": 0.6728571057319641,
+      "learning_rate": 2.1876876876876877e-05,
+      "loss": 0.5827,
+      "step": 5632
+    },
+    {
+      "epoch": 0.007363369816850392,
+      "grad_norm": 1.2923654317855835,
+      "learning_rate": 2.1871871871871874e-05,
+      "loss": 1.1979,
+      "step": 5633
+    },
+    {
+      "epoch": 0.007364677001266661,
+      "grad_norm": 1.6465610265731812,
+      "learning_rate": 2.1866866866866867e-05,
+      "loss": 1.442,
+      "step": 5634
+    },
+    {
+      "epoch": 0.007365984185682932,
+      "grad_norm": 1.5164844989776611,
+      "learning_rate": 2.1861861861861863e-05,
+      "loss": 1.4776,
+      "step": 5635
+    },
+    {
+      "epoch": 0.007367291370099202,
+      "grad_norm": 1.2900609970092773,
+      "learning_rate": 2.185685685685686e-05,
+      "loss": 1.7667,
+      "step": 5636
+    },
+    {
+      "epoch": 0.007368598554515472,
+      "grad_norm": 1.368440866470337,
+      "learning_rate": 2.1851851851851852e-05,
+      "loss": 1.4672,
+      "step": 5637
+    },
+    {
+      "epoch": 0.0073699057389317426,
+      "grad_norm": 1.3789539337158203,
+      "learning_rate": 2.1846846846846845e-05,
+      "loss": 0.9025,
+      "step": 5638
+    },
+    {
+      "epoch": 0.007371212923348013,
+      "grad_norm": 1.8318933248519897,
+      "learning_rate": 2.1841841841841842e-05,
+      "loss": 1.6256,
+      "step": 5639
+    },
+    {
+      "epoch": 0.007372520107764283,
+      "grad_norm": 2.4862210750579834,
+      "learning_rate": 2.1836836836836838e-05,
+      "loss": 1.6032,
+      "step": 5640
+    },
+    {
+      "epoch": 0.0073738272921805535,
+      "grad_norm": 1.63260018825531,
+      "learning_rate": 2.1831831831831835e-05,
+      "loss": 1.7734,
+      "step": 5641
+    },
+    {
+      "epoch": 0.007375134476596824,
+      "grad_norm": 1.4409980773925781,
+      "learning_rate": 2.1826826826826827e-05,
+      "loss": 1.7175,
+      "step": 5642
+    },
+    {
+      "epoch": 0.007376441661013094,
+      "grad_norm": 1.2662667036056519,
+      "learning_rate": 2.1821821821821824e-05,
+      "loss": 0.7762,
+      "step": 5643
+    },
+    {
+      "epoch": 0.0073777488454293645,
+      "grad_norm": 1.3209518194198608,
+      "learning_rate": 2.1816816816816817e-05,
+      "loss": 1.7429,
+      "step": 5644
+    },
+    {
+      "epoch": 0.007379056029845635,
+      "grad_norm": 1.326974868774414,
+      "learning_rate": 2.1811811811811813e-05,
+      "loss": 1.1498,
+      "step": 5645
+    },
+    {
+      "epoch": 0.007380363214261905,
+      "grad_norm": 1.2256845235824585,
+      "learning_rate": 2.1806806806806806e-05,
+      "loss": 1.0915,
+      "step": 5646
+    },
+    {
+      "epoch": 0.0073816703986781754,
+      "grad_norm": 0.9685318470001221,
+      "learning_rate": 2.1801801801801803e-05,
+      "loss": 0.6743,
+      "step": 5647
+    },
+    {
+      "epoch": 0.007382977583094446,
+      "grad_norm": 1.939993143081665,
+      "learning_rate": 2.17967967967968e-05,
+      "loss": 1.8006,
+      "step": 5648
+    },
+    {
+      "epoch": 0.007384284767510715,
+      "grad_norm": 1.2230247259140015,
+      "learning_rate": 2.1791791791791792e-05,
+      "loss": 1.4376,
+      "step": 5649
+    },
+    {
+      "epoch": 0.0073855919519269855,
+      "grad_norm": 1.464773416519165,
+      "learning_rate": 2.1786786786786788e-05,
+      "loss": 1.9453,
+      "step": 5650
+    },
+    {
+      "epoch": 0.007386899136343256,
+      "grad_norm": 1.1016473770141602,
+      "learning_rate": 2.178178178178178e-05,
+      "loss": 1.6022,
+      "step": 5651
+    },
+    {
+      "epoch": 0.007388206320759526,
+      "grad_norm": 1.108965277671814,
+      "learning_rate": 2.1776776776776778e-05,
+      "loss": 1.595,
+      "step": 5652
+    },
+    {
+      "epoch": 0.0073895135051757965,
+      "grad_norm": 1.2162917852401733,
+      "learning_rate": 2.1771771771771774e-05,
+      "loss": 1.1751,
+      "step": 5653
+    },
+    {
+      "epoch": 0.007390820689592067,
+      "grad_norm": 1.3809776306152344,
+      "learning_rate": 2.176676676676677e-05,
+      "loss": 1.805,
+      "step": 5654
+    },
+    {
+      "epoch": 0.007392127874008337,
+      "grad_norm": 1.526343584060669,
+      "learning_rate": 2.1761761761761763e-05,
+      "loss": 1.8609,
+      "step": 5655
+    },
+    {
+      "epoch": 0.0073934350584246075,
+      "grad_norm": 1.265377163887024,
+      "learning_rate": 2.1756756756756756e-05,
+      "loss": 2.0632,
+      "step": 5656
+    },
+    {
+      "epoch": 0.007394742242840878,
+      "grad_norm": 1.404964566230774,
+      "learning_rate": 2.1751751751751753e-05,
+      "loss": 1.4627,
+      "step": 5657
+    },
+    {
+      "epoch": 0.007396049427257148,
+      "grad_norm": 1.391205906867981,
+      "learning_rate": 2.174674674674675e-05,
+      "loss": 1.391,
+      "step": 5658
+    },
+    {
+      "epoch": 0.007397356611673418,
+      "grad_norm": 1.5252234935760498,
+      "learning_rate": 2.1741741741741742e-05,
+      "loss": 1.7638,
+      "step": 5659
+    },
+    {
+      "epoch": 0.007398663796089689,
+      "grad_norm": 1.1886703968048096,
+      "learning_rate": 2.1736736736736738e-05,
+      "loss": 1.2042,
+      "step": 5660
+    },
+    {
+      "epoch": 0.007399970980505959,
+      "grad_norm": 1.5854954719543457,
+      "learning_rate": 2.1731731731731735e-05,
+      "loss": 2.1043,
+      "step": 5661
+    },
+    {
+      "epoch": 0.007401278164922229,
+      "grad_norm": 1.5305730104446411,
+      "learning_rate": 2.1726726726726728e-05,
+      "loss": 1.5937,
+      "step": 5662
+    },
+    {
+      "epoch": 0.0074025853493385,
+      "grad_norm": 1.2588460445404053,
+      "learning_rate": 2.172172172172172e-05,
+      "loss": 1.6567,
+      "step": 5663
+    },
+    {
+      "epoch": 0.00740389253375477,
+      "grad_norm": 1.659976840019226,
+      "learning_rate": 2.1716716716716717e-05,
+      "loss": 1.0167,
+      "step": 5664
+    },
+    {
+      "epoch": 0.0074051997181710395,
+      "grad_norm": 0.861286997795105,
+      "learning_rate": 2.1711711711711713e-05,
+      "loss": 0.4712,
+      "step": 5665
+    },
+    {
+      "epoch": 0.00740650690258731,
+      "grad_norm": 0.8647375702857971,
+      "learning_rate": 2.170670670670671e-05,
+      "loss": 0.628,
+      "step": 5666
+    },
+    {
+      "epoch": 0.00740781408700358,
+      "grad_norm": 2.051731824874878,
+      "learning_rate": 2.1701701701701703e-05,
+      "loss": 0.7411,
+      "step": 5667
+    },
+    {
+      "epoch": 0.00740912127141985,
+      "grad_norm": 0.9711745381355286,
+      "learning_rate": 2.1696696696696696e-05,
+      "loss": 2.9603,
+      "step": 5668
+    },
+    {
+      "epoch": 0.007410428455836121,
+      "grad_norm": 1.2265836000442505,
+      "learning_rate": 2.1691691691691692e-05,
+      "loss": 1.9804,
+      "step": 5669
+    },
+    {
+      "epoch": 0.007411735640252391,
+      "grad_norm": 1.0439306497573853,
+      "learning_rate": 2.168668668668669e-05,
+      "loss": 2.2377,
+      "step": 5670
+    },
+    {
+      "epoch": 0.007413042824668661,
+      "grad_norm": 1.6202352046966553,
+      "learning_rate": 2.168168168168168e-05,
+      "loss": 2.3101,
+      "step": 5671
+    },
+    {
+      "epoch": 0.007414350009084932,
+      "grad_norm": 1.3469035625457764,
+      "learning_rate": 2.1676676676676678e-05,
+      "loss": 1.0529,
+      "step": 5672
+    },
+    {
+      "epoch": 0.007415657193501202,
+      "grad_norm": 1.6286064386367798,
+      "learning_rate": 2.1671671671671674e-05,
+      "loss": 1.8493,
+      "step": 5673
+    },
+    {
+      "epoch": 0.007416964377917472,
+      "grad_norm": 1.2346242666244507,
+      "learning_rate": 2.1666666666666667e-05,
+      "loss": 1.4128,
+      "step": 5674
+    },
+    {
+      "epoch": 0.007418271562333743,
+      "grad_norm": 1.384833812713623,
+      "learning_rate": 2.166166166166166e-05,
+      "loss": 1.5053,
+      "step": 5675
+    },
+    {
+      "epoch": 0.007419578746750013,
+      "grad_norm": 1.43271005153656,
+      "learning_rate": 2.1656656656656656e-05,
+      "loss": 1.7443,
+      "step": 5676
+    },
+    {
+      "epoch": 0.007420885931166283,
+      "grad_norm": 1.6075454950332642,
+      "learning_rate": 2.1651651651651653e-05,
+      "loss": 1.7723,
+      "step": 5677
+    },
+    {
+      "epoch": 0.007422193115582554,
+      "grad_norm": 1.9829641580581665,
+      "learning_rate": 2.164664664664665e-05,
+      "loss": 1.3489,
+      "step": 5678
+    },
+    {
+      "epoch": 0.007423500299998824,
+      "grad_norm": 1.3595476150512695,
+      "learning_rate": 2.1641641641641645e-05,
+      "loss": 2.1298,
+      "step": 5679
+    },
+    {
+      "epoch": 0.007424807484415093,
+      "grad_norm": 1.3271547555923462,
+      "learning_rate": 2.163663663663664e-05,
+      "loss": 2.1761,
+      "step": 5680
+    },
+    {
+      "epoch": 0.007426114668831364,
+      "grad_norm": 1.7198272943496704,
+      "learning_rate": 2.163163163163163e-05,
+      "loss": 2.0149,
+      "step": 5681
+    },
+    {
+      "epoch": 0.007427421853247634,
+      "grad_norm": 1.775626301765442,
+      "learning_rate": 2.1626626626626628e-05,
+      "loss": 2.3747,
+      "step": 5682
+    },
+    {
+      "epoch": 0.007428729037663904,
+      "grad_norm": 1.4073117971420288,
+      "learning_rate": 2.1621621621621624e-05,
+      "loss": 1.7698,
+      "step": 5683
+    },
+    {
+      "epoch": 0.007430036222080175,
+      "grad_norm": 1.1078728437423706,
+      "learning_rate": 2.1616616616616617e-05,
+      "loss": 0.6008,
+      "step": 5684
+    },
+    {
+      "epoch": 0.007431343406496445,
+      "grad_norm": 2.7076659202575684,
+      "learning_rate": 2.1611611611611613e-05,
+      "loss": 1.7512,
+      "step": 5685
+    },
+    {
+      "epoch": 0.007432650590912715,
+      "grad_norm": 1.9324427843093872,
+      "learning_rate": 2.1606606606606606e-05,
+      "loss": 1.5775,
+      "step": 5686
+    },
+    {
+      "epoch": 0.007433957775328986,
+      "grad_norm": 1.217018485069275,
+      "learning_rate": 2.1601601601601603e-05,
+      "loss": 1.2734,
+      "step": 5687
+    },
+    {
+      "epoch": 0.007435264959745256,
+      "grad_norm": 1.7399541139602661,
+      "learning_rate": 2.1596596596596596e-05,
+      "loss": 1.6661,
+      "step": 5688
+    },
+    {
+      "epoch": 0.007436572144161526,
+      "grad_norm": 1.4558336734771729,
+      "learning_rate": 2.1591591591591592e-05,
+      "loss": 1.8185,
+      "step": 5689
+    },
+    {
+      "epoch": 0.007437879328577797,
+      "grad_norm": 1.3901861906051636,
+      "learning_rate": 2.158658658658659e-05,
+      "loss": 1.4713,
+      "step": 5690
+    },
+    {
+      "epoch": 0.007439186512994067,
+      "grad_norm": 1.1614583730697632,
+      "learning_rate": 2.1581581581581585e-05,
+      "loss": 1.4305,
+      "step": 5691
+    },
+    {
+      "epoch": 0.007440493697410337,
+      "grad_norm": 1.2666375637054443,
+      "learning_rate": 2.1576576576576578e-05,
+      "loss": 1.3533,
+      "step": 5692
+    },
+    {
+      "epoch": 0.0074418008818266075,
+      "grad_norm": 1.3556989431381226,
+      "learning_rate": 2.157157157157157e-05,
+      "loss": 2.0314,
+      "step": 5693
+    },
+    {
+      "epoch": 0.007443108066242878,
+      "grad_norm": 1.7033751010894775,
+      "learning_rate": 2.1566566566566567e-05,
+      "loss": 0.6761,
+      "step": 5694
+    },
+    {
+      "epoch": 0.007444415250659147,
+      "grad_norm": 1.1292487382888794,
+      "learning_rate": 2.1561561561561564e-05,
+      "loss": 1.5777,
+      "step": 5695
+    },
+    {
+      "epoch": 0.007445722435075418,
+      "grad_norm": 1.018555998802185,
+      "learning_rate": 2.1556556556556557e-05,
+      "loss": 0.8393,
+      "step": 5696
+    },
+    {
+      "epoch": 0.007447029619491688,
+      "grad_norm": 1.274986982345581,
+      "learning_rate": 2.1551551551551553e-05,
+      "loss": 1.6291,
+      "step": 5697
+    },
+    {
+      "epoch": 0.007448336803907958,
+      "grad_norm": 1.2618142366409302,
+      "learning_rate": 2.154654654654655e-05,
+      "loss": 1.357,
+      "step": 5698
+    },
+    {
+      "epoch": 0.007449643988324229,
+      "grad_norm": 1.3670539855957031,
+      "learning_rate": 2.1541541541541542e-05,
+      "loss": 1.9064,
+      "step": 5699
+    },
+    {
+      "epoch": 0.007450951172740499,
+      "grad_norm": 1.2096188068389893,
+      "learning_rate": 2.1536536536536535e-05,
+      "loss": 1.648,
+      "step": 5700
+    },
+    {
+      "epoch": 0.007452258357156769,
+      "grad_norm": 1.2210229635238647,
+      "learning_rate": 2.153153153153153e-05,
+      "loss": 0.8387,
+      "step": 5701
+    },
+    {
+      "epoch": 0.0074535655415730396,
+      "grad_norm": 1.243396282196045,
+      "learning_rate": 2.1526526526526528e-05,
+      "loss": 1.6996,
+      "step": 5702
+    },
+    {
+      "epoch": 0.00745487272598931,
+      "grad_norm": 1.450654149055481,
+      "learning_rate": 2.1521521521521524e-05,
+      "loss": 1.8306,
+      "step": 5703
+    },
+    {
+      "epoch": 0.00745617991040558,
+      "grad_norm": 1.8644051551818848,
+      "learning_rate": 2.1516516516516517e-05,
+      "loss": 1.5848,
+      "step": 5704
+    },
+    {
+      "epoch": 0.0074574870948218505,
+      "grad_norm": 1.8080552816390991,
+      "learning_rate": 2.151151151151151e-05,
+      "loss": 1.2701,
+      "step": 5705
+    },
+    {
+      "epoch": 0.007458794279238121,
+      "grad_norm": 1.1469892263412476,
+      "learning_rate": 2.1506506506506507e-05,
+      "loss": 1.6415,
+      "step": 5706
+    },
+    {
+      "epoch": 0.007460101463654391,
+      "grad_norm": 1.3593077659606934,
+      "learning_rate": 2.1501501501501503e-05,
+      "loss": 1.3814,
+      "step": 5707
+    },
+    {
+      "epoch": 0.0074614086480706615,
+      "grad_norm": 1.7059824466705322,
+      "learning_rate": 2.14964964964965e-05,
+      "loss": 2.0793,
+      "step": 5708
+    },
+    {
+      "epoch": 0.007462715832486932,
+      "grad_norm": 1.2390475273132324,
+      "learning_rate": 2.1491491491491492e-05,
+      "loss": 1.2262,
+      "step": 5709
+    },
+    {
+      "epoch": 0.007464023016903202,
+      "grad_norm": 1.5648730993270874,
+      "learning_rate": 2.148648648648649e-05,
+      "loss": 1.4337,
+      "step": 5710
+    },
+    {
+      "epoch": 0.0074653302013194716,
+      "grad_norm": 1.431386947631836,
+      "learning_rate": 2.148148148148148e-05,
+      "loss": 1.2102,
+      "step": 5711
+    },
+    {
+      "epoch": 0.007466637385735742,
+      "grad_norm": 1.6882222890853882,
+      "learning_rate": 2.1476476476476478e-05,
+      "loss": 1.6641,
+      "step": 5712
+    },
+    {
+      "epoch": 0.007467944570152012,
+      "grad_norm": 1.698534607887268,
+      "learning_rate": 2.147147147147147e-05,
+      "loss": 1.8674,
+      "step": 5713
+    },
+    {
+      "epoch": 0.0074692517545682825,
+      "grad_norm": 0.9644426703453064,
+      "learning_rate": 2.1466466466466467e-05,
+      "loss": 1.0564,
+      "step": 5714
+    },
+    {
+      "epoch": 0.007470558938984553,
+      "grad_norm": 1.4472349882125854,
+      "learning_rate": 2.1461461461461464e-05,
+      "loss": 1.0047,
+      "step": 5715
+    },
+    {
+      "epoch": 0.007471866123400823,
+      "grad_norm": 1.6244828701019287,
+      "learning_rate": 2.145645645645646e-05,
+      "loss": 1.8004,
+      "step": 5716
+    },
+    {
+      "epoch": 0.0074731733078170935,
+      "grad_norm": 1.3620479106903076,
+      "learning_rate": 2.1451451451451453e-05,
+      "loss": 2.5856,
+      "step": 5717
+    },
+    {
+      "epoch": 0.007474480492233364,
+      "grad_norm": 0.8491674065589905,
+      "learning_rate": 2.1446446446446446e-05,
+      "loss": 0.5147,
+      "step": 5718
+    },
+    {
+      "epoch": 0.007475787676649634,
+      "grad_norm": 0.4322614073753357,
+      "learning_rate": 2.1441441441441442e-05,
+      "loss": 0.1547,
+      "step": 5719
+    },
+    {
+      "epoch": 0.0074770948610659044,
+      "grad_norm": 1.8007547855377197,
+      "learning_rate": 2.143643643643644e-05,
+      "loss": 2.5101,
+      "step": 5720
+    },
+    {
+      "epoch": 0.007478402045482175,
+      "grad_norm": 1.250522494316101,
+      "learning_rate": 2.143143143143143e-05,
+      "loss": 1.0405,
+      "step": 5721
+    },
+    {
+      "epoch": 0.007479709229898445,
+      "grad_norm": 1.3836826086044312,
+      "learning_rate": 2.1426426426426428e-05,
+      "loss": 1.943,
+      "step": 5722
+    },
+    {
+      "epoch": 0.007481016414314715,
+      "grad_norm": 1.3890635967254639,
+      "learning_rate": 2.142142142142142e-05,
+      "loss": 1.5822,
+      "step": 5723
+    },
+    {
+      "epoch": 0.007482323598730986,
+      "grad_norm": 1.0726567506790161,
+      "learning_rate": 2.1416416416416417e-05,
+      "loss": 1.0386,
+      "step": 5724
+    },
+    {
+      "epoch": 0.007483630783147256,
+      "grad_norm": 1.5514096021652222,
+      "learning_rate": 2.141141141141141e-05,
+      "loss": 2.2268,
+      "step": 5725
+    },
+    {
+      "epoch": 0.0074849379675635255,
+      "grad_norm": 1.1789122819900513,
+      "learning_rate": 2.1406406406406407e-05,
+      "loss": 1.3405,
+      "step": 5726
+    },
+    {
+      "epoch": 0.007486245151979796,
+      "grad_norm": 1.39958655834198,
+      "learning_rate": 2.1401401401401403e-05,
+      "loss": 2.4278,
+      "step": 5727
+    },
+    {
+      "epoch": 0.007487552336396066,
+      "grad_norm": 1.2138334512710571,
+      "learning_rate": 2.13963963963964e-05,
+      "loss": 1.0784,
+      "step": 5728
+    },
+    {
+      "epoch": 0.0074888595208123365,
+      "grad_norm": 1.554117202758789,
+      "learning_rate": 2.1391391391391392e-05,
+      "loss": 1.7823,
+      "step": 5729
+    },
+    {
+      "epoch": 0.007490166705228607,
+      "grad_norm": 1.5884853601455688,
+      "learning_rate": 2.1386386386386385e-05,
+      "loss": 1.8821,
+      "step": 5730
+    },
+    {
+      "epoch": 0.007491473889644877,
+      "grad_norm": 1.175134539604187,
+      "learning_rate": 2.1381381381381382e-05,
+      "loss": 1.401,
+      "step": 5731
+    },
+    {
+      "epoch": 0.007492781074061147,
+      "grad_norm": 1.914549708366394,
+      "learning_rate": 2.1376376376376378e-05,
+      "loss": 1.6984,
+      "step": 5732
+    },
+    {
+      "epoch": 0.007494088258477418,
+      "grad_norm": 1.2287685871124268,
+      "learning_rate": 2.1371371371371374e-05,
+      "loss": 0.7384,
+      "step": 5733
+    },
+    {
+      "epoch": 0.007495395442893688,
+      "grad_norm": 1.6021569967269897,
+      "learning_rate": 2.1366366366366367e-05,
+      "loss": 1.7023,
+      "step": 5734
+    },
+    {
+      "epoch": 0.007496702627309958,
+      "grad_norm": 1.1088522672653198,
+      "learning_rate": 2.1361361361361364e-05,
+      "loss": 1.2008,
+      "step": 5735
+    },
+    {
+      "epoch": 0.007498009811726229,
+      "grad_norm": 1.4118461608886719,
+      "learning_rate": 2.1356356356356357e-05,
+      "loss": 1.4965,
+      "step": 5736
+    },
+    {
+      "epoch": 0.007499316996142499,
+      "grad_norm": 2.4879817962646484,
+      "learning_rate": 2.1351351351351353e-05,
+      "loss": 1.1793,
+      "step": 5737
+    },
+    {
+      "epoch": 0.007500624180558769,
+      "grad_norm": 0.9464182257652283,
+      "learning_rate": 2.1346346346346346e-05,
+      "loss": 0.4328,
+      "step": 5738
+    },
+    {
+      "epoch": 0.00750193136497504,
+      "grad_norm": 1.0927802324295044,
+      "learning_rate": 2.1341341341341342e-05,
+      "loss": 1.0783,
+      "step": 5739
+    },
+    {
+      "epoch": 0.00750323854939131,
+      "grad_norm": 1.453410267829895,
+      "learning_rate": 2.133633633633634e-05,
+      "loss": 1.5539,
+      "step": 5740
+    },
+    {
+      "epoch": 0.007504545733807579,
+      "grad_norm": 1.3982367515563965,
+      "learning_rate": 2.1331331331331332e-05,
+      "loss": 1.4865,
+      "step": 5741
+    },
+    {
+      "epoch": 0.00750585291822385,
+      "grad_norm": 1.2394275665283203,
+      "learning_rate": 2.1326326326326325e-05,
+      "loss": 1.3242,
+      "step": 5742
+    },
+    {
+      "epoch": 0.00750716010264012,
+      "grad_norm": 1.6365493535995483,
+      "learning_rate": 2.132132132132132e-05,
+      "loss": 1.9171,
+      "step": 5743
+    },
+    {
+      "epoch": 0.00750846728705639,
+      "grad_norm": 1.185956597328186,
+      "learning_rate": 2.1316316316316318e-05,
+      "loss": 1.2883,
+      "step": 5744
+    },
+    {
+      "epoch": 0.007509774471472661,
+      "grad_norm": 1.4091086387634277,
+      "learning_rate": 2.1311311311311314e-05,
+      "loss": 1.5942,
+      "step": 5745
+    },
+    {
+      "epoch": 0.007511081655888931,
+      "grad_norm": 1.3884742259979248,
+      "learning_rate": 2.1306306306306307e-05,
+      "loss": 1.4265,
+      "step": 5746
+    },
+    {
+      "epoch": 0.007512388840305201,
+      "grad_norm": 1.3573551177978516,
+      "learning_rate": 2.1301301301301303e-05,
+      "loss": 1.9102,
+      "step": 5747
+    },
+    {
+      "epoch": 0.007513696024721472,
+      "grad_norm": 1.5227701663970947,
+      "learning_rate": 2.1296296296296296e-05,
+      "loss": 1.8577,
+      "step": 5748
+    },
+    {
+      "epoch": 0.007515003209137742,
+      "grad_norm": 2.507293939590454,
+      "learning_rate": 2.1291291291291293e-05,
+      "loss": 1.2625,
+      "step": 5749
+    },
+    {
+      "epoch": 0.007516310393554012,
+      "grad_norm": 1.0855906009674072,
+      "learning_rate": 2.1286286286286286e-05,
+      "loss": 0.4612,
+      "step": 5750
+    },
+    {
+      "epoch": 0.007517617577970283,
+      "grad_norm": 1.3184361457824707,
+      "learning_rate": 2.1281281281281282e-05,
+      "loss": 1.4781,
+      "step": 5751
+    },
+    {
+      "epoch": 0.007518924762386553,
+      "grad_norm": 1.3029979467391968,
+      "learning_rate": 2.1276276276276278e-05,
+      "loss": 1.1932,
+      "step": 5752
+    },
+    {
+      "epoch": 0.007520231946802823,
+      "grad_norm": 1.6538838148117065,
+      "learning_rate": 2.1271271271271275e-05,
+      "loss": 1.2211,
+      "step": 5753
+    },
+    {
+      "epoch": 0.007521539131219094,
+      "grad_norm": 1.1879535913467407,
+      "learning_rate": 2.1266266266266268e-05,
+      "loss": 1.0984,
+      "step": 5754
+    },
+    {
+      "epoch": 0.007522846315635364,
+      "grad_norm": 1.556908369064331,
+      "learning_rate": 2.126126126126126e-05,
+      "loss": 1.6639,
+      "step": 5755
+    },
+    {
+      "epoch": 0.007524153500051633,
+      "grad_norm": 1.564762830734253,
+      "learning_rate": 2.1256256256256257e-05,
+      "loss": 1.446,
+      "step": 5756
+    },
+    {
+      "epoch": 0.007525460684467904,
+      "grad_norm": 1.165137529373169,
+      "learning_rate": 2.1251251251251253e-05,
+      "loss": 1.1786,
+      "step": 5757
+    },
+    {
+      "epoch": 0.007526767868884174,
+      "grad_norm": 1.6033152341842651,
+      "learning_rate": 2.124624624624625e-05,
+      "loss": 1.5389,
+      "step": 5758
+    },
+    {
+      "epoch": 0.007528075053300444,
+      "grad_norm": 4.768499374389648,
+      "learning_rate": 2.1241241241241243e-05,
+      "loss": 1.9954,
+      "step": 5759
+    },
+    {
+      "epoch": 0.007529382237716715,
+      "grad_norm": 1.1106973886489868,
+      "learning_rate": 2.1236236236236236e-05,
+      "loss": 1.059,
+      "step": 5760
+    },
+    {
+      "epoch": 0.007530689422132985,
+      "grad_norm": 1.2443453073501587,
+      "learning_rate": 2.1231231231231232e-05,
+      "loss": 1.6414,
+      "step": 5761
+    },
+    {
+      "epoch": 0.007531996606549255,
+      "grad_norm": 1.2490054368972778,
+      "learning_rate": 2.122622622622623e-05,
+      "loss": 1.2028,
+      "step": 5762
+    },
+    {
+      "epoch": 0.007533303790965526,
+      "grad_norm": 1.4548256397247314,
+      "learning_rate": 2.122122122122122e-05,
+      "loss": 1.5645,
+      "step": 5763
+    },
+    {
+      "epoch": 0.007534610975381796,
+      "grad_norm": 2.4374847412109375,
+      "learning_rate": 2.1216216216216218e-05,
+      "loss": 1.3705,
+      "step": 5764
+    },
+    {
+      "epoch": 0.007535918159798066,
+      "grad_norm": 1.2590481042861938,
+      "learning_rate": 2.1211211211211214e-05,
+      "loss": 1.6402,
+      "step": 5765
+    },
+    {
+      "epoch": 0.0075372253442143365,
+      "grad_norm": 1.299000859260559,
+      "learning_rate": 2.1206206206206207e-05,
+      "loss": 1.659,
+      "step": 5766
+    },
+    {
+      "epoch": 0.007538532528630607,
+      "grad_norm": 1.5286226272583008,
+      "learning_rate": 2.12012012012012e-05,
+      "loss": 1.6324,
+      "step": 5767
+    },
+    {
+      "epoch": 0.007539839713046877,
+      "grad_norm": 1.2404968738555908,
+      "learning_rate": 2.1196196196196196e-05,
+      "loss": 1.464,
+      "step": 5768
+    },
+    {
+      "epoch": 0.0075411468974631475,
+      "grad_norm": 1.7361280918121338,
+      "learning_rate": 2.1191191191191193e-05,
+      "loss": 1.7504,
+      "step": 5769
+    },
+    {
+      "epoch": 0.007542454081879418,
+      "grad_norm": 1.0558686256408691,
+      "learning_rate": 2.118618618618619e-05,
+      "loss": 1.1485,
+      "step": 5770
+    },
+    {
+      "epoch": 0.007543761266295688,
+      "grad_norm": 1.3920823335647583,
+      "learning_rate": 2.1181181181181182e-05,
+      "loss": 1.8248,
+      "step": 5771
+    },
+    {
+      "epoch": 0.007545068450711958,
+      "grad_norm": 1.5967957973480225,
+      "learning_rate": 2.117617617617618e-05,
+      "loss": 1.9682,
+      "step": 5772
+    },
+    {
+      "epoch": 0.007546375635128228,
+      "grad_norm": 1.5194976329803467,
+      "learning_rate": 2.117117117117117e-05,
+      "loss": 1.7733,
+      "step": 5773
+    },
+    {
+      "epoch": 0.007547682819544498,
+      "grad_norm": 1.6498024463653564,
+      "learning_rate": 2.1166166166166168e-05,
+      "loss": 1.3608,
+      "step": 5774
+    },
+    {
+      "epoch": 0.0075489900039607686,
+      "grad_norm": 1.4018687009811401,
+      "learning_rate": 2.116116116116116e-05,
+      "loss": 1.9837,
+      "step": 5775
+    },
+    {
+      "epoch": 0.007550297188377039,
+      "grad_norm": 1.4245656728744507,
+      "learning_rate": 2.1156156156156157e-05,
+      "loss": 1.3044,
+      "step": 5776
+    },
+    {
+      "epoch": 0.007551604372793309,
+      "grad_norm": 2.234692096710205,
+      "learning_rate": 2.1151151151151153e-05,
+      "loss": 1.5845,
+      "step": 5777
+    },
+    {
+      "epoch": 0.0075529115572095795,
+      "grad_norm": 3.0807743072509766,
+      "learning_rate": 2.1146146146146146e-05,
+      "loss": 1.9823,
+      "step": 5778
+    },
+    {
+      "epoch": 0.00755421874162585,
+      "grad_norm": 1.6743425130844116,
+      "learning_rate": 2.114114114114114e-05,
+      "loss": 1.8108,
+      "step": 5779
+    },
+    {
+      "epoch": 0.00755552592604212,
+      "grad_norm": 1.328450083732605,
+      "learning_rate": 2.1136136136136136e-05,
+      "loss": 1.709,
+      "step": 5780
+    },
+    {
+      "epoch": 0.0075568331104583905,
+      "grad_norm": 1.3095370531082153,
+      "learning_rate": 2.1131131131131132e-05,
+      "loss": 1.822,
+      "step": 5781
+    },
+    {
+      "epoch": 0.007558140294874661,
+      "grad_norm": 1.4249248504638672,
+      "learning_rate": 2.112612612612613e-05,
+      "loss": 2.411,
+      "step": 5782
+    },
+    {
+      "epoch": 0.007559447479290931,
+      "grad_norm": 1.3458778858184814,
+      "learning_rate": 2.1121121121121125e-05,
+      "loss": 1.2031,
+      "step": 5783
+    },
+    {
+      "epoch": 0.0075607546637072014,
+      "grad_norm": 1.3746750354766846,
+      "learning_rate": 2.1116116116116118e-05,
+      "loss": 1.2425,
+      "step": 5784
+    },
+    {
+      "epoch": 0.007562061848123472,
+      "grad_norm": 1.2902106046676636,
+      "learning_rate": 2.111111111111111e-05,
+      "loss": 1.2538,
+      "step": 5785
+    },
+    {
+      "epoch": 0.007563369032539742,
+      "grad_norm": 1.9547595977783203,
+      "learning_rate": 2.1106106106106107e-05,
+      "loss": 2.1235,
+      "step": 5786
+    },
+    {
+      "epoch": 0.0075646762169560115,
+      "grad_norm": 1.4029871225357056,
+      "learning_rate": 2.1101101101101103e-05,
+      "loss": 1.8355,
+      "step": 5787
+    },
+    {
+      "epoch": 0.007565983401372282,
+      "grad_norm": 1.0005707740783691,
+      "learning_rate": 2.1096096096096096e-05,
+      "loss": 0.8164,
+      "step": 5788
+    },
+    {
+      "epoch": 0.007567290585788552,
+      "grad_norm": 1.3045367002487183,
+      "learning_rate": 2.1091091091091093e-05,
+      "loss": 1.6177,
+      "step": 5789
+    },
+    {
+      "epoch": 0.0075685977702048225,
+      "grad_norm": 1.5273183584213257,
+      "learning_rate": 2.108608608608609e-05,
+      "loss": 1.842,
+      "step": 5790
+    },
+    {
+      "epoch": 0.007569904954621093,
+      "grad_norm": 1.114370584487915,
+      "learning_rate": 2.1081081081081082e-05,
+      "loss": 0.8824,
+      "step": 5791
+    },
+    {
+      "epoch": 0.007571212139037363,
+      "grad_norm": 1.600193738937378,
+      "learning_rate": 2.1076076076076075e-05,
+      "loss": 1.7391,
+      "step": 5792
+    },
+    {
+      "epoch": 0.0075725193234536334,
+      "grad_norm": 1.4644272327423096,
+      "learning_rate": 2.107107107107107e-05,
+      "loss": 1.7231,
+      "step": 5793
+    },
+    {
+      "epoch": 0.007573826507869904,
+      "grad_norm": 1.1480265855789185,
+      "learning_rate": 2.1066066066066068e-05,
+      "loss": 1.2829,
+      "step": 5794
+    },
+    {
+      "epoch": 0.007575133692286174,
+      "grad_norm": 1.5388284921646118,
+      "learning_rate": 2.1061061061061064e-05,
+      "loss": 1.7354,
+      "step": 5795
+    },
+    {
+      "epoch": 0.007576440876702444,
+      "grad_norm": 1.7796659469604492,
+      "learning_rate": 2.1056056056056057e-05,
+      "loss": 2.6934,
+      "step": 5796
+    },
+    {
+      "epoch": 0.007577748061118715,
+      "grad_norm": 1.1102854013442993,
+      "learning_rate": 2.105105105105105e-05,
+      "loss": 0.9293,
+      "step": 5797
+    },
+    {
+      "epoch": 0.007579055245534985,
+      "grad_norm": 1.772068977355957,
+      "learning_rate": 2.1046046046046047e-05,
+      "loss": 0.9743,
+      "step": 5798
+    },
+    {
+      "epoch": 0.007580362429951255,
+      "grad_norm": 1.182788372039795,
+      "learning_rate": 2.1041041041041043e-05,
+      "loss": 0.982,
+      "step": 5799
+    },
+    {
+      "epoch": 0.007581669614367526,
+      "grad_norm": 1.2123584747314453,
+      "learning_rate": 2.1036036036036036e-05,
+      "loss": 1.3797,
+      "step": 5800
+    },
+    {
+      "epoch": 0.007582976798783796,
+      "grad_norm": 2.596842050552368,
+      "learning_rate": 2.1031031031031032e-05,
+      "loss": 1.5,
+      "step": 5801
+    },
+    {
+      "epoch": 0.0075842839832000655,
+      "grad_norm": 1.2168869972229004,
+      "learning_rate": 2.102602602602603e-05,
+      "loss": 1.7251,
+      "step": 5802
+    },
+    {
+      "epoch": 0.007585591167616336,
+      "grad_norm": 1.510818362236023,
+      "learning_rate": 2.102102102102102e-05,
+      "loss": 1.9646,
+      "step": 5803
+    },
+    {
+      "epoch": 0.007586898352032606,
+      "grad_norm": 1.404335856437683,
+      "learning_rate": 2.1016016016016015e-05,
+      "loss": 1.8862,
+      "step": 5804
+    },
+    {
+      "epoch": 0.007588205536448876,
+      "grad_norm": 1.2484445571899414,
+      "learning_rate": 2.101101101101101e-05,
+      "loss": 1.2546,
+      "step": 5805
+    },
+    {
+      "epoch": 0.007589512720865147,
+      "grad_norm": 1.3209325075149536,
+      "learning_rate": 2.1006006006006007e-05,
+      "loss": 1.5773,
+      "step": 5806
+    },
+    {
+      "epoch": 0.007590819905281417,
+      "grad_norm": 1.5267739295959473,
+      "learning_rate": 2.1001001001001004e-05,
+      "loss": 1.5859,
+      "step": 5807
+    },
+    {
+      "epoch": 0.007592127089697687,
+      "grad_norm": 1.4734594821929932,
+      "learning_rate": 2.0995995995995997e-05,
+      "loss": 2.2698,
+      "step": 5808
+    },
+    {
+      "epoch": 0.007593434274113958,
+      "grad_norm": 1.3915996551513672,
+      "learning_rate": 2.0990990990990993e-05,
+      "loss": 2.139,
+      "step": 5809
+    },
+    {
+      "epoch": 0.007594741458530228,
+      "grad_norm": 1.343042016029358,
+      "learning_rate": 2.0985985985985986e-05,
+      "loss": 1.237,
+      "step": 5810
+    },
+    {
+      "epoch": 0.007596048642946498,
+      "grad_norm": 1.7527186870574951,
+      "learning_rate": 2.0980980980980982e-05,
+      "loss": 2.122,
+      "step": 5811
+    },
+    {
+      "epoch": 0.007597355827362769,
+      "grad_norm": 1.5887272357940674,
+      "learning_rate": 2.097597597597598e-05,
+      "loss": 1.4465,
+      "step": 5812
+    },
+    {
+      "epoch": 0.007598663011779039,
+      "grad_norm": 1.6806362867355347,
+      "learning_rate": 2.097097097097097e-05,
+      "loss": 2.1884,
+      "step": 5813
+    },
+    {
+      "epoch": 0.007599970196195309,
+      "grad_norm": 1.2121689319610596,
+      "learning_rate": 2.0965965965965968e-05,
+      "loss": 1.4361,
+      "step": 5814
+    },
+    {
+      "epoch": 0.00760127738061158,
+      "grad_norm": 1.1269478797912598,
+      "learning_rate": 2.096096096096096e-05,
+      "loss": 1.1256,
+      "step": 5815
+    },
+    {
+      "epoch": 0.00760258456502785,
+      "grad_norm": 1.8055031299591064,
+      "learning_rate": 2.0955955955955957e-05,
+      "loss": 1.5753,
+      "step": 5816
+    },
+    {
+      "epoch": 0.00760389174944412,
+      "grad_norm": 1.3349933624267578,
+      "learning_rate": 2.095095095095095e-05,
+      "loss": 2.0039,
+      "step": 5817
+    },
+    {
+      "epoch": 0.00760519893386039,
+      "grad_norm": 1.598812222480774,
+      "learning_rate": 2.0945945945945947e-05,
+      "loss": 2.2174,
+      "step": 5818
+    },
+    {
+      "epoch": 0.00760650611827666,
+      "grad_norm": 1.516893982887268,
+      "learning_rate": 2.0940940940940943e-05,
+      "loss": 2.0841,
+      "step": 5819
+    },
+    {
+      "epoch": 0.00760781330269293,
+      "grad_norm": 1.4698141813278198,
+      "learning_rate": 2.093593593593594e-05,
+      "loss": 1.8211,
+      "step": 5820
+    },
+    {
+      "epoch": 0.007609120487109201,
+      "grad_norm": 1.5194389820098877,
+      "learning_rate": 2.0930930930930932e-05,
+      "loss": 2.0782,
+      "step": 5821
+    },
+    {
+      "epoch": 0.007610427671525471,
+      "grad_norm": 1.6036279201507568,
+      "learning_rate": 2.0925925925925925e-05,
+      "loss": 2.0742,
+      "step": 5822
+    },
+    {
+      "epoch": 0.007611734855941741,
+      "grad_norm": 1.402974009513855,
+      "learning_rate": 2.0920920920920922e-05,
+      "loss": 1.9,
+      "step": 5823
+    },
+    {
+      "epoch": 0.007613042040358012,
+      "grad_norm": 1.57710862159729,
+      "learning_rate": 2.0915915915915918e-05,
+      "loss": 1.5517,
+      "step": 5824
+    },
+    {
+      "epoch": 0.007614349224774282,
+      "grad_norm": 1.7418986558914185,
+      "learning_rate": 2.091091091091091e-05,
+      "loss": 1.8744,
+      "step": 5825
+    },
+    {
+      "epoch": 0.007615656409190552,
+      "grad_norm": 1.1683164834976196,
+      "learning_rate": 2.0905905905905907e-05,
+      "loss": 1.1415,
+      "step": 5826
+    },
+    {
+      "epoch": 0.007616963593606823,
+      "grad_norm": 1.3088507652282715,
+      "learning_rate": 2.0900900900900904e-05,
+      "loss": 1.248,
+      "step": 5827
+    },
+    {
+      "epoch": 0.007618270778023093,
+      "grad_norm": 1.216599702835083,
+      "learning_rate": 2.0895895895895897e-05,
+      "loss": 0.7736,
+      "step": 5828
+    },
+    {
+      "epoch": 0.007619577962439363,
+      "grad_norm": 0.9193097352981567,
+      "learning_rate": 2.089089089089089e-05,
+      "loss": 0.3731,
+      "step": 5829
+    },
+    {
+      "epoch": 0.0076208851468556335,
+      "grad_norm": 1.1291797161102295,
+      "learning_rate": 2.0885885885885886e-05,
+      "loss": 1.3656,
+      "step": 5830
+    },
+    {
+      "epoch": 0.007622192331271904,
+      "grad_norm": 1.203862190246582,
+      "learning_rate": 2.0880880880880882e-05,
+      "loss": 1.6655,
+      "step": 5831
+    },
+    {
+      "epoch": 0.007623499515688174,
+      "grad_norm": 1.8161746263504028,
+      "learning_rate": 2.087587587587588e-05,
+      "loss": 1.9144,
+      "step": 5832
+    },
+    {
+      "epoch": 0.007624806700104444,
+      "grad_norm": 1.3545180559158325,
+      "learning_rate": 2.0870870870870872e-05,
+      "loss": 1.1974,
+      "step": 5833
+    },
+    {
+      "epoch": 0.007626113884520714,
+      "grad_norm": 1.3327224254608154,
+      "learning_rate": 2.0865865865865865e-05,
+      "loss": 1.6853,
+      "step": 5834
+    },
+    {
+      "epoch": 0.007627421068936984,
+      "grad_norm": 1.4250637292861938,
+      "learning_rate": 2.086086086086086e-05,
+      "loss": 1.7332,
+      "step": 5835
+    },
+    {
+      "epoch": 0.007628728253353255,
+      "grad_norm": 1.4509693384170532,
+      "learning_rate": 2.0855855855855857e-05,
+      "loss": 0.891,
+      "step": 5836
+    },
+    {
+      "epoch": 0.007630035437769525,
+      "grad_norm": 1.818758487701416,
+      "learning_rate": 2.0850850850850854e-05,
+      "loss": 1.3235,
+      "step": 5837
+    },
+    {
+      "epoch": 0.007631342622185795,
+      "grad_norm": 1.5417382717132568,
+      "learning_rate": 2.0845845845845847e-05,
+      "loss": 1.8938,
+      "step": 5838
+    },
+    {
+      "epoch": 0.0076326498066020656,
+      "grad_norm": 1.4409698247909546,
+      "learning_rate": 2.0840840840840843e-05,
+      "loss": 2.097,
+      "step": 5839
+    },
+    {
+      "epoch": 0.007633956991018336,
+      "grad_norm": 1.1875897645950317,
+      "learning_rate": 2.0835835835835836e-05,
+      "loss": 0.9494,
+      "step": 5840
+    },
+    {
+      "epoch": 0.007635264175434606,
+      "grad_norm": 1.4848226308822632,
+      "learning_rate": 2.0830830830830833e-05,
+      "loss": 1.9963,
+      "step": 5841
+    },
+    {
+      "epoch": 0.0076365713598508765,
+      "grad_norm": 1.1044985055923462,
+      "learning_rate": 2.0825825825825825e-05,
+      "loss": 1.181,
+      "step": 5842
+    },
+    {
+      "epoch": 0.007637878544267147,
+      "grad_norm": 1.4135890007019043,
+      "learning_rate": 2.0820820820820822e-05,
+      "loss": 0.7471,
+      "step": 5843
+    },
+    {
+      "epoch": 0.007639185728683417,
+      "grad_norm": 1.0332424640655518,
+      "learning_rate": 2.0815815815815818e-05,
+      "loss": 0.7829,
+      "step": 5844
+    },
+    {
+      "epoch": 0.0076404929130996875,
+      "grad_norm": 1.2932864427566528,
+      "learning_rate": 2.0810810810810815e-05,
+      "loss": 1.1015,
+      "step": 5845
+    },
+    {
+      "epoch": 0.007641800097515958,
+      "grad_norm": 1.7552626132965088,
+      "learning_rate": 2.0805805805805804e-05,
+      "loss": 2.1881,
+      "step": 5846
+    },
+    {
+      "epoch": 0.007643107281932228,
+      "grad_norm": 1.4515517950057983,
+      "learning_rate": 2.08008008008008e-05,
+      "loss": 1.3821,
+      "step": 5847
+    },
+    {
+      "epoch": 0.0076444144663484976,
+      "grad_norm": 0.6383469700813293,
+      "learning_rate": 2.0795795795795797e-05,
+      "loss": 0.4452,
+      "step": 5848
+    },
+    {
+      "epoch": 0.007645721650764768,
+      "grad_norm": 1.8650742769241333,
+      "learning_rate": 2.0790790790790793e-05,
+      "loss": 1.9525,
+      "step": 5849
+    },
+    {
+      "epoch": 0.007647028835181038,
+      "grad_norm": 1.3957371711730957,
+      "learning_rate": 2.0785785785785786e-05,
+      "loss": 1.6717,
+      "step": 5850
+    },
+    {
+      "epoch": 0.0076483360195973085,
+      "grad_norm": 1.6116477251052856,
+      "learning_rate": 2.0780780780780783e-05,
+      "loss": 1.5052,
+      "step": 5851
+    },
+    {
+      "epoch": 0.007649643204013579,
+      "grad_norm": 1.468451738357544,
+      "learning_rate": 2.0775775775775776e-05,
+      "loss": 1.5055,
+      "step": 5852
+    },
+    {
+      "epoch": 0.007650950388429849,
+      "grad_norm": 1.6230586767196655,
+      "learning_rate": 2.0770770770770772e-05,
+      "loss": 1.4245,
+      "step": 5853
+    },
+    {
+      "epoch": 0.0076522575728461195,
+      "grad_norm": 1.3071726560592651,
+      "learning_rate": 2.0765765765765765e-05,
+      "loss": 1.0342,
+      "step": 5854
+    },
+    {
+      "epoch": 0.00765356475726239,
+      "grad_norm": 1.586289644241333,
+      "learning_rate": 2.076076076076076e-05,
+      "loss": 2.0332,
+      "step": 5855
+    },
+    {
+      "epoch": 0.00765487194167866,
+      "grad_norm": 1.7369320392608643,
+      "learning_rate": 2.0755755755755758e-05,
+      "loss": 1.7701,
+      "step": 5856
+    },
+    {
+      "epoch": 0.0076561791260949304,
+      "grad_norm": 1.2514981031417847,
+      "learning_rate": 2.0750750750750754e-05,
+      "loss": 1.3811,
+      "step": 5857
+    },
+    {
+      "epoch": 0.007657486310511201,
+      "grad_norm": 1.3001432418823242,
+      "learning_rate": 2.0745745745745747e-05,
+      "loss": 1.297,
+      "step": 5858
+    },
+    {
+      "epoch": 0.007658793494927471,
+      "grad_norm": 1.5417559146881104,
+      "learning_rate": 2.074074074074074e-05,
+      "loss": 1.505,
+      "step": 5859
+    },
+    {
+      "epoch": 0.007660100679343741,
+      "grad_norm": 1.4798685312271118,
+      "learning_rate": 2.0735735735735736e-05,
+      "loss": 2.1358,
+      "step": 5860
+    },
+    {
+      "epoch": 0.007661407863760012,
+      "grad_norm": 0.9986108541488647,
+      "learning_rate": 2.0730730730730733e-05,
+      "loss": 0.787,
+      "step": 5861
+    },
+    {
+      "epoch": 0.007662715048176282,
+      "grad_norm": 1.6853562593460083,
+      "learning_rate": 2.0725725725725726e-05,
+      "loss": 2.1207,
+      "step": 5862
+    },
+    {
+      "epoch": 0.0076640222325925515,
+      "grad_norm": 1.1812981367111206,
+      "learning_rate": 2.0720720720720722e-05,
+      "loss": 1.1519,
+      "step": 5863
+    },
+    {
+      "epoch": 0.007665329417008822,
+      "grad_norm": 1.488722324371338,
+      "learning_rate": 2.071571571571572e-05,
+      "loss": 1.1675,
+      "step": 5864
+    },
+    {
+      "epoch": 0.007666636601425092,
+      "grad_norm": 1.6156002283096313,
+      "learning_rate": 2.071071071071071e-05,
+      "loss": 2.4634,
+      "step": 5865
+    },
+    {
+      "epoch": 0.0076679437858413625,
+      "grad_norm": 1.6533606052398682,
+      "learning_rate": 2.0705705705705708e-05,
+      "loss": 2.1682,
+      "step": 5866
+    },
+    {
+      "epoch": 0.007669250970257633,
+      "grad_norm": 1.171040415763855,
+      "learning_rate": 2.07007007007007e-05,
+      "loss": 1.3012,
+      "step": 5867
+    },
+    {
+      "epoch": 0.007670558154673903,
+      "grad_norm": 1.5084662437438965,
+      "learning_rate": 2.0695695695695697e-05,
+      "loss": 1.8445,
+      "step": 5868
+    },
+    {
+      "epoch": 0.007671865339090173,
+      "grad_norm": 1.511272668838501,
+      "learning_rate": 2.0690690690690693e-05,
+      "loss": 2.1337,
+      "step": 5869
+    },
+    {
+      "epoch": 0.007673172523506444,
+      "grad_norm": 1.4853743314743042,
+      "learning_rate": 2.0685685685685686e-05,
+      "loss": 1.5454,
+      "step": 5870
+    },
+    {
+      "epoch": 0.007674479707922714,
+      "grad_norm": 1.5006532669067383,
+      "learning_rate": 2.068068068068068e-05,
+      "loss": 1.5095,
+      "step": 5871
+    },
+    {
+      "epoch": 0.007675786892338984,
+      "grad_norm": 1.5756572484970093,
+      "learning_rate": 2.0675675675675676e-05,
+      "loss": 1.8373,
+      "step": 5872
+    },
+    {
+      "epoch": 0.007677094076755255,
+      "grad_norm": 1.8633863925933838,
+      "learning_rate": 2.0670670670670672e-05,
+      "loss": 1.4269,
+      "step": 5873
+    },
+    {
+      "epoch": 0.007678401261171525,
+      "grad_norm": 1.4166085720062256,
+      "learning_rate": 2.066566566566567e-05,
+      "loss": 1.3078,
+      "step": 5874
+    },
+    {
+      "epoch": 0.007679708445587795,
+      "grad_norm": 1.2198978662490845,
+      "learning_rate": 2.066066066066066e-05,
+      "loss": 1.1745,
+      "step": 5875
+    },
+    {
+      "epoch": 0.007681015630004066,
+      "grad_norm": 1.670898675918579,
+      "learning_rate": 2.0655655655655658e-05,
+      "loss": 1.6263,
+      "step": 5876
+    },
+    {
+      "epoch": 0.007682322814420336,
+      "grad_norm": 1.5329192876815796,
+      "learning_rate": 2.065065065065065e-05,
+      "loss": 1.7911,
+      "step": 5877
+    },
+    {
+      "epoch": 0.007683629998836606,
+      "grad_norm": 1.356667399406433,
+      "learning_rate": 2.0645645645645647e-05,
+      "loss": 1.5292,
+      "step": 5878
+    },
+    {
+      "epoch": 0.007684937183252876,
+      "grad_norm": 1.645696997642517,
+      "learning_rate": 2.064064064064064e-05,
+      "loss": 2.068,
+      "step": 5879
+    },
+    {
+      "epoch": 0.007686244367669146,
+      "grad_norm": 1.6471024751663208,
+      "learning_rate": 2.0635635635635636e-05,
+      "loss": 1.8094,
+      "step": 5880
+    },
+    {
+      "epoch": 0.007687551552085416,
+      "grad_norm": 1.3703385591506958,
+      "learning_rate": 2.0630630630630633e-05,
+      "loss": 2.5214,
+      "step": 5881
+    },
+    {
+      "epoch": 0.007688858736501687,
+      "grad_norm": 0.964406430721283,
+      "learning_rate": 2.062562562562563e-05,
+      "loss": 0.8771,
+      "step": 5882
+    },
+    {
+      "epoch": 0.007690165920917957,
+      "grad_norm": 1.4578715562820435,
+      "learning_rate": 2.062062062062062e-05,
+      "loss": 1.8886,
+      "step": 5883
+    },
+    {
+      "epoch": 0.007691473105334227,
+      "grad_norm": 1.926645040512085,
+      "learning_rate": 2.0615615615615615e-05,
+      "loss": 1.77,
+      "step": 5884
+    },
+    {
+      "epoch": 0.007692780289750498,
+      "grad_norm": 1.1649967432022095,
+      "learning_rate": 2.061061061061061e-05,
+      "loss": 0.7829,
+      "step": 5885
+    },
+    {
+      "epoch": 0.007694087474166768,
+      "grad_norm": 1.5464801788330078,
+      "learning_rate": 2.0605605605605608e-05,
+      "loss": 1.6946,
+      "step": 5886
+    },
+    {
+      "epoch": 0.007695394658583038,
+      "grad_norm": 1.7677747011184692,
+      "learning_rate": 2.06006006006006e-05,
+      "loss": 1.7473,
+      "step": 5887
+    },
+    {
+      "epoch": 0.007696701842999309,
+      "grad_norm": 1.3729289770126343,
+      "learning_rate": 2.0595595595595597e-05,
+      "loss": 2.0715,
+      "step": 5888
+    },
+    {
+      "epoch": 0.007698009027415579,
+      "grad_norm": 1.2560036182403564,
+      "learning_rate": 2.059059059059059e-05,
+      "loss": 0.9763,
+      "step": 5889
+    },
+    {
+      "epoch": 0.007699316211831849,
+      "grad_norm": 1.2764647006988525,
+      "learning_rate": 2.0585585585585586e-05,
+      "loss": 1.9331,
+      "step": 5890
+    },
+    {
+      "epoch": 0.00770062339624812,
+      "grad_norm": 1.2574033737182617,
+      "learning_rate": 2.0580580580580583e-05,
+      "loss": 1.527,
+      "step": 5891
+    },
+    {
+      "epoch": 0.00770193058066439,
+      "grad_norm": 2.0608103275299072,
+      "learning_rate": 2.0575575575575576e-05,
+      "loss": 2.038,
+      "step": 5892
+    },
+    {
+      "epoch": 0.00770323776508066,
+      "grad_norm": 1.3610658645629883,
+      "learning_rate": 2.0570570570570572e-05,
+      "loss": 1.0683,
+      "step": 5893
+    },
+    {
+      "epoch": 0.00770454494949693,
+      "grad_norm": 1.5874361991882324,
+      "learning_rate": 2.056556556556557e-05,
+      "loss": 1.1812,
+      "step": 5894
+    },
+    {
+      "epoch": 0.0077058521339132,
+      "grad_norm": 1.3457640409469604,
+      "learning_rate": 2.056056056056056e-05,
+      "loss": 2.0206,
+      "step": 5895
+    },
+    {
+      "epoch": 0.00770715931832947,
+      "grad_norm": 1.4150817394256592,
+      "learning_rate": 2.0555555555555555e-05,
+      "loss": 1.4368,
+      "step": 5896
+    },
+    {
+      "epoch": 0.007708466502745741,
+      "grad_norm": 1.7326699495315552,
+      "learning_rate": 2.055055055055055e-05,
+      "loss": 1.1611,
+      "step": 5897
+    },
+    {
+      "epoch": 0.007709773687162011,
+      "grad_norm": 1.3326693773269653,
+      "learning_rate": 2.0545545545545547e-05,
+      "loss": 1.4469,
+      "step": 5898
+    },
+    {
+      "epoch": 0.007711080871578281,
+      "grad_norm": 1.7337961196899414,
+      "learning_rate": 2.0540540540540544e-05,
+      "loss": 2.2432,
+      "step": 5899
+    },
+    {
+      "epoch": 0.007712388055994552,
+      "grad_norm": 1.305687665939331,
+      "learning_rate": 2.0535535535535537e-05,
+      "loss": 0.8521,
+      "step": 5900
+    },
+    {
+      "epoch": 0.007713695240410822,
+      "grad_norm": 1.5702534914016724,
+      "learning_rate": 2.053053053053053e-05,
+      "loss": 1.5887,
+      "step": 5901
+    },
+    {
+      "epoch": 0.007715002424827092,
+      "grad_norm": 1.8270788192749023,
+      "learning_rate": 2.0525525525525526e-05,
+      "loss": 1.4097,
+      "step": 5902
+    },
+    {
+      "epoch": 0.0077163096092433625,
+      "grad_norm": 1.073147177696228,
+      "learning_rate": 2.0520520520520522e-05,
+      "loss": 1.2815,
+      "step": 5903
+    },
+    {
+      "epoch": 0.007717616793659633,
+      "grad_norm": 1.2971473932266235,
+      "learning_rate": 2.0515515515515515e-05,
+      "loss": 1.3388,
+      "step": 5904
+    },
+    {
+      "epoch": 0.007718923978075903,
+      "grad_norm": 1.1872632503509521,
+      "learning_rate": 2.051051051051051e-05,
+      "loss": 1.2421,
+      "step": 5905
+    },
+    {
+      "epoch": 0.0077202311624921735,
+      "grad_norm": 1.617693305015564,
+      "learning_rate": 2.0505505505505508e-05,
+      "loss": 1.876,
+      "step": 5906
+    },
+    {
+      "epoch": 0.007721538346908444,
+      "grad_norm": 2.002962589263916,
+      "learning_rate": 2.05005005005005e-05,
+      "loss": 1.6438,
+      "step": 5907
+    },
+    {
+      "epoch": 0.007722845531324714,
+      "grad_norm": 1.6477922201156616,
+      "learning_rate": 2.0495495495495494e-05,
+      "loss": 1.8,
+      "step": 5908
+    },
+    {
+      "epoch": 0.007724152715740984,
+      "grad_norm": 1.356479287147522,
+      "learning_rate": 2.049049049049049e-05,
+      "loss": 1.8907,
+      "step": 5909
+    },
+    {
+      "epoch": 0.007725459900157254,
+      "grad_norm": 1.2659120559692383,
+      "learning_rate": 2.0485485485485487e-05,
+      "loss": 1.7838,
+      "step": 5910
+    },
+    {
+      "epoch": 0.007726767084573524,
+      "grad_norm": 1.3491724729537964,
+      "learning_rate": 2.0480480480480483e-05,
+      "loss": 1.4802,
+      "step": 5911
+    },
+    {
+      "epoch": 0.0077280742689897946,
+      "grad_norm": 1.277526617050171,
+      "learning_rate": 2.0475475475475476e-05,
+      "loss": 1.1541,
+      "step": 5912
+    },
+    {
+      "epoch": 0.007729381453406065,
+      "grad_norm": 1.3715990781784058,
+      "learning_rate": 2.0470470470470472e-05,
+      "loss": 1.674,
+      "step": 5913
+    },
+    {
+      "epoch": 0.007730688637822335,
+      "grad_norm": 1.65455162525177,
+      "learning_rate": 2.0465465465465465e-05,
+      "loss": 2.109,
+      "step": 5914
+    },
+    {
+      "epoch": 0.0077319958222386055,
+      "grad_norm": 1.8216217756271362,
+      "learning_rate": 2.046046046046046e-05,
+      "loss": 2.0439,
+      "step": 5915
+    },
+    {
+      "epoch": 0.007733303006654876,
+      "grad_norm": 1.273499846458435,
+      "learning_rate": 2.0455455455455455e-05,
+      "loss": 1.3672,
+      "step": 5916
+    },
+    {
+      "epoch": 0.007734610191071146,
+      "grad_norm": 1.415332317352295,
+      "learning_rate": 2.045045045045045e-05,
+      "loss": 1.5926,
+      "step": 5917
+    },
+    {
+      "epoch": 0.0077359173754874165,
+      "grad_norm": 1.4518444538116455,
+      "learning_rate": 2.0445445445445447e-05,
+      "loss": 2.0138,
+      "step": 5918
+    },
+    {
+      "epoch": 0.007737224559903687,
+      "grad_norm": 1.5319194793701172,
+      "learning_rate": 2.0440440440440444e-05,
+      "loss": 1.7985,
+      "step": 5919
+    },
+    {
+      "epoch": 0.007738531744319957,
+      "grad_norm": 1.0620107650756836,
+      "learning_rate": 2.0435435435435437e-05,
+      "loss": 1.1277,
+      "step": 5920
+    },
+    {
+      "epoch": 0.007739838928736227,
+      "grad_norm": 1.1836704015731812,
+      "learning_rate": 2.043043043043043e-05,
+      "loss": 1.8036,
+      "step": 5921
+    },
+    {
+      "epoch": 0.007741146113152498,
+      "grad_norm": 1.5337775945663452,
+      "learning_rate": 2.0425425425425426e-05,
+      "loss": 1.7074,
+      "step": 5922
+    },
+    {
+      "epoch": 0.007742453297568768,
+      "grad_norm": 1.2984416484832764,
+      "learning_rate": 2.0420420420420422e-05,
+      "loss": 0.7639,
+      "step": 5923
+    },
+    {
+      "epoch": 0.007743760481985038,
+      "grad_norm": 1.2559980154037476,
+      "learning_rate": 2.041541541541542e-05,
+      "loss": 1.1147,
+      "step": 5924
+    },
+    {
+      "epoch": 0.007745067666401308,
+      "grad_norm": 1.2601885795593262,
+      "learning_rate": 2.0410410410410412e-05,
+      "loss": 1.6138,
+      "step": 5925
+    },
+    {
+      "epoch": 0.007746374850817578,
+      "grad_norm": 1.9089000225067139,
+      "learning_rate": 2.0405405405405405e-05,
+      "loss": 1.6425,
+      "step": 5926
+    },
+    {
+      "epoch": 0.0077476820352338485,
+      "grad_norm": 1.4575697183609009,
+      "learning_rate": 2.04004004004004e-05,
+      "loss": 1.8606,
+      "step": 5927
+    },
+    {
+      "epoch": 0.007748989219650119,
+      "grad_norm": 1.3318015336990356,
+      "learning_rate": 2.0395395395395397e-05,
+      "loss": 1.6193,
+      "step": 5928
+    },
+    {
+      "epoch": 0.007750296404066389,
+      "grad_norm": 1.6184309720993042,
+      "learning_rate": 2.039039039039039e-05,
+      "loss": 1.9512,
+      "step": 5929
+    },
+    {
+      "epoch": 0.0077516035884826594,
+      "grad_norm": 2.184877395629883,
+      "learning_rate": 2.0385385385385387e-05,
+      "loss": 1.8528,
+      "step": 5930
+    },
+    {
+      "epoch": 0.00775291077289893,
+      "grad_norm": 1.2110812664031982,
+      "learning_rate": 2.0380380380380383e-05,
+      "loss": 0.8749,
+      "step": 5931
+    },
+    {
+      "epoch": 0.0077542179573152,
+      "grad_norm": 1.4298036098480225,
+      "learning_rate": 2.0375375375375376e-05,
+      "loss": 1.203,
+      "step": 5932
+    },
+    {
+      "epoch": 0.00775552514173147,
+      "grad_norm": 1.2544312477111816,
+      "learning_rate": 2.037037037037037e-05,
+      "loss": 0.8771,
+      "step": 5933
+    },
+    {
+      "epoch": 0.007756832326147741,
+      "grad_norm": 1.7834994792938232,
+      "learning_rate": 2.0365365365365365e-05,
+      "loss": 1.4212,
+      "step": 5934
+    },
+    {
+      "epoch": 0.007758139510564011,
+      "grad_norm": 1.3167918920516968,
+      "learning_rate": 2.0360360360360362e-05,
+      "loss": 1.3058,
+      "step": 5935
+    },
+    {
+      "epoch": 0.007759446694980281,
+      "grad_norm": 1.7386082410812378,
+      "learning_rate": 2.0355355355355358e-05,
+      "loss": 1.1541,
+      "step": 5936
+    },
+    {
+      "epoch": 0.007760753879396552,
+      "grad_norm": 1.394890546798706,
+      "learning_rate": 2.035035035035035e-05,
+      "loss": 1.5012,
+      "step": 5937
+    },
+    {
+      "epoch": 0.007762061063812822,
+      "grad_norm": 1.5831032991409302,
+      "learning_rate": 2.0345345345345344e-05,
+      "loss": 1.311,
+      "step": 5938
+    },
+    {
+      "epoch": 0.007763368248229092,
+      "grad_norm": 1.8195135593414307,
+      "learning_rate": 2.034034034034034e-05,
+      "loss": 2.1188,
+      "step": 5939
+    },
+    {
+      "epoch": 0.007764675432645362,
+      "grad_norm": 1.475246548652649,
+      "learning_rate": 2.0335335335335337e-05,
+      "loss": 2.5548,
+      "step": 5940
+    },
+    {
+      "epoch": 0.007765982617061632,
+      "grad_norm": 1.9756669998168945,
+      "learning_rate": 2.033033033033033e-05,
+      "loss": 1.4687,
+      "step": 5941
+    },
+    {
+      "epoch": 0.007767289801477902,
+      "grad_norm": 1.5542175769805908,
+      "learning_rate": 2.0325325325325326e-05,
+      "loss": 1.4102,
+      "step": 5942
+    },
+    {
+      "epoch": 0.007768596985894173,
+      "grad_norm": 1.3413324356079102,
+      "learning_rate": 2.0320320320320323e-05,
+      "loss": 1.7525,
+      "step": 5943
+    },
+    {
+      "epoch": 0.007769904170310443,
+      "grad_norm": 2.1996026039123535,
+      "learning_rate": 2.0315315315315316e-05,
+      "loss": 1.7896,
+      "step": 5944
+    },
+    {
+      "epoch": 0.007771211354726713,
+      "grad_norm": 1.5729306936264038,
+      "learning_rate": 2.0310310310310312e-05,
+      "loss": 1.4429,
+      "step": 5945
+    },
+    {
+      "epoch": 0.007772518539142984,
+      "grad_norm": 1.5604287385940552,
+      "learning_rate": 2.0305305305305305e-05,
+      "loss": 1.9195,
+      "step": 5946
+    },
+    {
+      "epoch": 0.007773825723559254,
+      "grad_norm": 1.2712745666503906,
+      "learning_rate": 2.03003003003003e-05,
+      "loss": 1.4011,
+      "step": 5947
+    },
+    {
+      "epoch": 0.007775132907975524,
+      "grad_norm": 1.2403417825698853,
+      "learning_rate": 2.0295295295295298e-05,
+      "loss": 1.6751,
+      "step": 5948
+    },
+    {
+      "epoch": 0.007776440092391795,
+      "grad_norm": 1.79839026927948,
+      "learning_rate": 2.0290290290290294e-05,
+      "loss": 1.5419,
+      "step": 5949
+    },
+    {
+      "epoch": 0.007777747276808065,
+      "grad_norm": 1.5767366886138916,
+      "learning_rate": 2.0285285285285287e-05,
+      "loss": 1.0168,
+      "step": 5950
+    },
+    {
+      "epoch": 0.007779054461224335,
+      "grad_norm": 1.0506924390792847,
+      "learning_rate": 2.028028028028028e-05,
+      "loss": 0.7259,
+      "step": 5951
+    },
+    {
+      "epoch": 0.007780361645640606,
+      "grad_norm": 1.6685287952423096,
+      "learning_rate": 2.0275275275275276e-05,
+      "loss": 1.586,
+      "step": 5952
+    },
+    {
+      "epoch": 0.007781668830056876,
+      "grad_norm": 1.7932103872299194,
+      "learning_rate": 2.0270270270270273e-05,
+      "loss": 1.8082,
+      "step": 5953
+    },
+    {
+      "epoch": 0.007782976014473146,
+      "grad_norm": 1.0965818166732788,
+      "learning_rate": 2.0265265265265266e-05,
+      "loss": 0.8835,
+      "step": 5954
+    },
+    {
+      "epoch": 0.007784283198889416,
+      "grad_norm": 1.1170361042022705,
+      "learning_rate": 2.0260260260260262e-05,
+      "loss": 2.0131,
+      "step": 5955
+    },
+    {
+      "epoch": 0.007785590383305686,
+      "grad_norm": 1.4526325464248657,
+      "learning_rate": 2.0255255255255255e-05,
+      "loss": 2.5712,
+      "step": 5956
+    },
+    {
+      "epoch": 0.007786897567721956,
+      "grad_norm": 1.596399188041687,
+      "learning_rate": 2.025025025025025e-05,
+      "loss": 2.3647,
+      "step": 5957
+    },
+    {
+      "epoch": 0.007788204752138227,
+      "grad_norm": 1.585848093032837,
+      "learning_rate": 2.0245245245245244e-05,
+      "loss": 1.6248,
+      "step": 5958
+    },
+    {
+      "epoch": 0.007789511936554497,
+      "grad_norm": 1.7377886772155762,
+      "learning_rate": 2.024024024024024e-05,
+      "loss": 1.4336,
+      "step": 5959
+    },
+    {
+      "epoch": 0.007790819120970767,
+      "grad_norm": 1.9890599250793457,
+      "learning_rate": 2.0235235235235237e-05,
+      "loss": 1.7379,
+      "step": 5960
+    },
+    {
+      "epoch": 0.007792126305387038,
+      "grad_norm": 1.2768020629882812,
+      "learning_rate": 2.0230230230230233e-05,
+      "loss": 1.0929,
+      "step": 5961
+    },
+    {
+      "epoch": 0.007793433489803308,
+      "grad_norm": 1.9958888292312622,
+      "learning_rate": 2.0225225225225226e-05,
+      "loss": 1.5939,
+      "step": 5962
+    },
+    {
+      "epoch": 0.007794740674219578,
+      "grad_norm": 1.4434186220169067,
+      "learning_rate": 2.022022022022022e-05,
+      "loss": 1.3343,
+      "step": 5963
+    },
+    {
+      "epoch": 0.007796047858635849,
+      "grad_norm": 1.6255402565002441,
+      "learning_rate": 2.0215215215215216e-05,
+      "loss": 2.1608,
+      "step": 5964
+    },
+    {
+      "epoch": 0.007797355043052119,
+      "grad_norm": 1.486204743385315,
+      "learning_rate": 2.0210210210210212e-05,
+      "loss": 1.997,
+      "step": 5965
+    },
+    {
+      "epoch": 0.007798662227468389,
+      "grad_norm": 1.5443143844604492,
+      "learning_rate": 2.0205205205205205e-05,
+      "loss": 1.8581,
+      "step": 5966
+    },
+    {
+      "epoch": 0.0077999694118846595,
+      "grad_norm": 1.9042445421218872,
+      "learning_rate": 2.02002002002002e-05,
+      "loss": 1.4059,
+      "step": 5967
+    },
+    {
+      "epoch": 0.00780127659630093,
+      "grad_norm": 1.0450477600097656,
+      "learning_rate": 2.0195195195195198e-05,
+      "loss": 0.8473,
+      "step": 5968
+    },
+    {
+      "epoch": 0.0078025837807172,
+      "grad_norm": 1.4373540878295898,
+      "learning_rate": 2.019019019019019e-05,
+      "loss": 1.7343,
+      "step": 5969
+    },
+    {
+      "epoch": 0.0078038909651334705,
+      "grad_norm": 1.465095043182373,
+      "learning_rate": 2.0185185185185187e-05,
+      "loss": 1.7828,
+      "step": 5970
+    },
+    {
+      "epoch": 0.00780519814954974,
+      "grad_norm": 1.35258948802948,
+      "learning_rate": 2.018018018018018e-05,
+      "loss": 1.2322,
+      "step": 5971
+    },
+    {
+      "epoch": 0.00780650533396601,
+      "grad_norm": 1.1496264934539795,
+      "learning_rate": 2.0175175175175176e-05,
+      "loss": 1.4133,
+      "step": 5972
+    },
+    {
+      "epoch": 0.007807812518382281,
+      "grad_norm": 1.6152832508087158,
+      "learning_rate": 2.0170170170170173e-05,
+      "loss": 2.0889,
+      "step": 5973
+    },
+    {
+      "epoch": 0.007809119702798551,
+      "grad_norm": 1.3215690851211548,
+      "learning_rate": 2.0165165165165166e-05,
+      "loss": 1.1579,
+      "step": 5974
+    },
+    {
+      "epoch": 0.007810426887214821,
+      "grad_norm": 2.0656003952026367,
+      "learning_rate": 2.016016016016016e-05,
+      "loss": 1.7066,
+      "step": 5975
+    },
+    {
+      "epoch": 0.0078117340716310915,
+      "grad_norm": 1.535286545753479,
+      "learning_rate": 2.0155155155155155e-05,
+      "loss": 1.0631,
+      "step": 5976
+    },
+    {
+      "epoch": 0.007813041256047362,
+      "grad_norm": 1.3307989835739136,
+      "learning_rate": 2.015015015015015e-05,
+      "loss": 1.223,
+      "step": 5977
+    },
+    {
+      "epoch": 0.007814348440463631,
+      "grad_norm": 1.4120707511901855,
+      "learning_rate": 2.0145145145145148e-05,
+      "loss": 1.1753,
+      "step": 5978
+    },
+    {
+      "epoch": 0.007815655624879903,
+      "grad_norm": 1.3403708934783936,
+      "learning_rate": 2.014014014014014e-05,
+      "loss": 1.6877,
+      "step": 5979
+    },
+    {
+      "epoch": 0.007816962809296172,
+      "grad_norm": 1.3513883352279663,
+      "learning_rate": 2.0135135135135137e-05,
+      "loss": 1.2243,
+      "step": 5980
+    },
+    {
+      "epoch": 0.007818269993712443,
+      "grad_norm": 2.651571035385132,
+      "learning_rate": 2.013013013013013e-05,
+      "loss": 2.6167,
+      "step": 5981
+    },
+    {
+      "epoch": 0.007819577178128713,
+      "grad_norm": 1.2566003799438477,
+      "learning_rate": 2.0125125125125126e-05,
+      "loss": 0.6956,
+      "step": 5982
+    },
+    {
+      "epoch": 0.007820884362544984,
+      "grad_norm": 3.209019899368286,
+      "learning_rate": 2.012012012012012e-05,
+      "loss": 2.29,
+      "step": 5983
+    },
+    {
+      "epoch": 0.007822191546961253,
+      "grad_norm": 1.0665748119354248,
+      "learning_rate": 2.0115115115115116e-05,
+      "loss": 1.0423,
+      "step": 5984
+    },
+    {
+      "epoch": 0.007823498731377524,
+      "grad_norm": 1.3678921461105347,
+      "learning_rate": 2.0110110110110112e-05,
+      "loss": 1.9602,
+      "step": 5985
+    },
+    {
+      "epoch": 0.007824805915793794,
+      "grad_norm": 1.2592928409576416,
+      "learning_rate": 2.010510510510511e-05,
+      "loss": 1.5569,
+      "step": 5986
+    },
+    {
+      "epoch": 0.007826113100210065,
+      "grad_norm": 1.325223684310913,
+      "learning_rate": 2.01001001001001e-05,
+      "loss": 1.8173,
+      "step": 5987
+    },
+    {
+      "epoch": 0.007827420284626335,
+      "grad_norm": 1.1924699544906616,
+      "learning_rate": 2.0095095095095094e-05,
+      "loss": 1.5753,
+      "step": 5988
+    },
+    {
+      "epoch": 0.007828727469042606,
+      "grad_norm": 2.154745578765869,
+      "learning_rate": 2.009009009009009e-05,
+      "loss": 1.9518,
+      "step": 5989
+    },
+    {
+      "epoch": 0.007830034653458875,
+      "grad_norm": 1.569032907485962,
+      "learning_rate": 2.0085085085085087e-05,
+      "loss": 1.9224,
+      "step": 5990
+    },
+    {
+      "epoch": 0.007831341837875146,
+      "grad_norm": 1.4399555921554565,
+      "learning_rate": 2.008008008008008e-05,
+      "loss": 1.8288,
+      "step": 5991
+    },
+    {
+      "epoch": 0.007832649022291416,
+      "grad_norm": 1.975490689277649,
+      "learning_rate": 2.0075075075075077e-05,
+      "loss": 1.3211,
+      "step": 5992
+    },
+    {
+      "epoch": 0.007833956206707685,
+      "grad_norm": 0.1614423543214798,
+      "learning_rate": 2.007007007007007e-05,
+      "loss": 0.0053,
+      "step": 5993
+    },
+    {
+      "epoch": 0.007835263391123956,
+      "grad_norm": 1.3688040971755981,
+      "learning_rate": 2.0065065065065066e-05,
+      "loss": 1.9515,
+      "step": 5994
+    },
+    {
+      "epoch": 0.007836570575540226,
+      "grad_norm": 1.6069883108139038,
+      "learning_rate": 2.006006006006006e-05,
+      "loss": 2.0914,
+      "step": 5995
+    },
+    {
+      "epoch": 0.007837877759956497,
+      "grad_norm": 1.4972456693649292,
+      "learning_rate": 2.0055055055055055e-05,
+      "loss": 2.1382,
+      "step": 5996
+    },
+    {
+      "epoch": 0.007839184944372767,
+      "grad_norm": 1.1805561780929565,
+      "learning_rate": 2.005005005005005e-05,
+      "loss": 0.9551,
+      "step": 5997
+    },
+    {
+      "epoch": 0.007840492128789038,
+      "grad_norm": 1.723912000656128,
+      "learning_rate": 2.0045045045045048e-05,
+      "loss": 1.7553,
+      "step": 5998
+    },
+    {
+      "epoch": 0.007841799313205307,
+      "grad_norm": 1.6579593420028687,
+      "learning_rate": 2.004004004004004e-05,
+      "loss": 2.3849,
+      "step": 5999
+    },
+    {
+      "epoch": 0.007843106497621578,
+      "grad_norm": 1.6584858894348145,
+      "learning_rate": 2.0035035035035034e-05,
+      "loss": 1.6024,
+      "step": 6000
+    },
+    {
+      "epoch": 0.007844413682037848,
+      "grad_norm": 1.67653226852417,
+      "learning_rate": 2.003003003003003e-05,
+      "loss": 1.7395,
+      "step": 6001
+    },
+    {
+      "epoch": 0.007845720866454119,
+      "grad_norm": 1.8760404586791992,
+      "learning_rate": 2.0025025025025027e-05,
+      "loss": 1.3469,
+      "step": 6002
+    },
+    {
+      "epoch": 0.007847028050870388,
+      "grad_norm": 1.628304362297058,
+      "learning_rate": 2.0020020020020023e-05,
+      "loss": 1.6936,
+      "step": 6003
+    },
+    {
+      "epoch": 0.00784833523528666,
+      "grad_norm": 1.2725721597671509,
+      "learning_rate": 2.0015015015015016e-05,
+      "loss": 1.3506,
+      "step": 6004
+    },
+    {
+      "epoch": 0.007849642419702929,
+      "grad_norm": 1.5048253536224365,
+      "learning_rate": 2.0010010010010012e-05,
+      "loss": 2.4368,
+      "step": 6005
+    },
+    {
+      "epoch": 0.0078509496041192,
+      "grad_norm": 1.662551999092102,
+      "learning_rate": 2.0005005005005005e-05,
+      "loss": 1.6398,
+      "step": 6006
+    },
+    {
+      "epoch": 0.00785225678853547,
+      "grad_norm": 1.586049199104309,
+      "learning_rate": 2e-05,
+      "loss": 1.3284,
+      "step": 6007
+    },
+    {
+      "epoch": 0.007853563972951741,
+      "grad_norm": 1.6636046171188354,
+      "learning_rate": 1.9994994994994995e-05,
+      "loss": 1.998,
+      "step": 6008
+    },
+    {
+      "epoch": 0.00785487115736801,
+      "grad_norm": 1.3491284847259521,
+      "learning_rate": 1.998998998998999e-05,
+      "loss": 1.8333,
+      "step": 6009
+    },
+    {
+      "epoch": 0.00785617834178428,
+      "grad_norm": 1.8764374256134033,
+      "learning_rate": 1.9984984984984987e-05,
+      "loss": 1.2765,
+      "step": 6010
+    },
+    {
+      "epoch": 0.007857485526200551,
+      "grad_norm": 1.7102892398834229,
+      "learning_rate": 1.997997997997998e-05,
+      "loss": 2.0379,
+      "step": 6011
+    },
+    {
+      "epoch": 0.00785879271061682,
+      "grad_norm": 1.364081859588623,
+      "learning_rate": 1.9974974974974973e-05,
+      "loss": 1.5701,
+      "step": 6012
+    },
+    {
+      "epoch": 0.007860099895033092,
+      "grad_norm": 1.2410207986831665,
+      "learning_rate": 1.996996996996997e-05,
+      "loss": 1.6311,
+      "step": 6013
+    },
+    {
+      "epoch": 0.007861407079449361,
+      "grad_norm": 1.7439970970153809,
+      "learning_rate": 1.9964964964964966e-05,
+      "loss": 2.3954,
+      "step": 6014
+    },
+    {
+      "epoch": 0.007862714263865632,
+      "grad_norm": 1.352277398109436,
+      "learning_rate": 1.9959959959959962e-05,
+      "loss": 1.8576,
+      "step": 6015
+    },
+    {
+      "epoch": 0.007864021448281902,
+      "grad_norm": 1.1025581359863281,
+      "learning_rate": 1.9954954954954955e-05,
+      "loss": 1.0233,
+      "step": 6016
+    },
+    {
+      "epoch": 0.007865328632698173,
+      "grad_norm": 1.1484732627868652,
+      "learning_rate": 1.9949949949949952e-05,
+      "loss": 1.3906,
+      "step": 6017
+    },
+    {
+      "epoch": 0.007866635817114442,
+      "grad_norm": 1.3983436822891235,
+      "learning_rate": 1.9944944944944945e-05,
+      "loss": 2.1848,
+      "step": 6018
+    },
+    {
+      "epoch": 0.007867943001530714,
+      "grad_norm": 1.5435367822647095,
+      "learning_rate": 1.993993993993994e-05,
+      "loss": 2.2529,
+      "step": 6019
+    },
+    {
+      "epoch": 0.007869250185946983,
+      "grad_norm": 1.9888694286346436,
+      "learning_rate": 1.9934934934934934e-05,
+      "loss": 2.1902,
+      "step": 6020
+    },
+    {
+      "epoch": 0.007870557370363254,
+      "grad_norm": 0.6926677227020264,
+      "learning_rate": 1.992992992992993e-05,
+      "loss": 0.3374,
+      "step": 6021
+    },
+    {
+      "epoch": 0.007871864554779524,
+      "grad_norm": 1.6033910512924194,
+      "learning_rate": 1.9924924924924927e-05,
+      "loss": 1.439,
+      "step": 6022
+    },
+    {
+      "epoch": 0.007873171739195795,
+      "grad_norm": 1.2999788522720337,
+      "learning_rate": 1.9919919919919923e-05,
+      "loss": 1.443,
+      "step": 6023
+    },
+    {
+      "epoch": 0.007874478923612064,
+      "grad_norm": 1.4393408298492432,
+      "learning_rate": 1.9914914914914916e-05,
+      "loss": 1.5304,
+      "step": 6024
+    },
+    {
+      "epoch": 0.007875786108028334,
+      "grad_norm": 1.2706210613250732,
+      "learning_rate": 1.990990990990991e-05,
+      "loss": 1.5254,
+      "step": 6025
+    },
+    {
+      "epoch": 0.007877093292444605,
+      "grad_norm": 1.6079684495925903,
+      "learning_rate": 1.9904904904904905e-05,
+      "loss": 1.6994,
+      "step": 6026
+    },
+    {
+      "epoch": 0.007878400476860874,
+      "grad_norm": 1.5220767259597778,
+      "learning_rate": 1.9899899899899902e-05,
+      "loss": 1.8613,
+      "step": 6027
+    },
+    {
+      "epoch": 0.007879707661277146,
+      "grad_norm": 1.048128366470337,
+      "learning_rate": 1.9894894894894898e-05,
+      "loss": 0.6722,
+      "step": 6028
+    },
+    {
+      "epoch": 0.007881014845693415,
+      "grad_norm": 1.4976521730422974,
+      "learning_rate": 1.988988988988989e-05,
+      "loss": 1.8772,
+      "step": 6029
+    },
+    {
+      "epoch": 0.007882322030109686,
+      "grad_norm": 1.1012507677078247,
+      "learning_rate": 1.9884884884884884e-05,
+      "loss": 0.7949,
+      "step": 6030
+    },
+    {
+      "epoch": 0.007883629214525956,
+      "grad_norm": 1.327925443649292,
+      "learning_rate": 1.987987987987988e-05,
+      "loss": 2.1799,
+      "step": 6031
+    },
+    {
+      "epoch": 0.007884936398942227,
+      "grad_norm": 1.0930825471878052,
+      "learning_rate": 1.9874874874874877e-05,
+      "loss": 0.6699,
+      "step": 6032
+    },
+    {
+      "epoch": 0.007886243583358496,
+      "grad_norm": 1.4379768371582031,
+      "learning_rate": 1.986986986986987e-05,
+      "loss": 1.6278,
+      "step": 6033
+    },
+    {
+      "epoch": 0.007887550767774767,
+      "grad_norm": 1.0744563341140747,
+      "learning_rate": 1.9864864864864866e-05,
+      "loss": 0.7008,
+      "step": 6034
+    },
+    {
+      "epoch": 0.007888857952191037,
+      "grad_norm": 1.2937514781951904,
+      "learning_rate": 1.9859859859859863e-05,
+      "loss": 1.2607,
+      "step": 6035
+    },
+    {
+      "epoch": 0.007890165136607308,
+      "grad_norm": 1.2091472148895264,
+      "learning_rate": 1.9854854854854855e-05,
+      "loss": 1.0529,
+      "step": 6036
+    },
+    {
+      "epoch": 0.007891472321023578,
+      "grad_norm": 1.1875216960906982,
+      "learning_rate": 1.984984984984985e-05,
+      "loss": 0.9868,
+      "step": 6037
+    },
+    {
+      "epoch": 0.007892779505439849,
+      "grad_norm": 0.9461485743522644,
+      "learning_rate": 1.9844844844844845e-05,
+      "loss": 0.6395,
+      "step": 6038
+    },
+    {
+      "epoch": 0.007894086689856118,
+      "grad_norm": 1.7273821830749512,
+      "learning_rate": 1.983983983983984e-05,
+      "loss": 1.8022,
+      "step": 6039
+    },
+    {
+      "epoch": 0.007895393874272388,
+      "grad_norm": 1.5085952281951904,
+      "learning_rate": 1.9834834834834838e-05,
+      "loss": 1.4834,
+      "step": 6040
+    },
+    {
+      "epoch": 0.007896701058688659,
+      "grad_norm": 1.4202746152877808,
+      "learning_rate": 1.982982982982983e-05,
+      "loss": 1.9456,
+      "step": 6041
+    },
+    {
+      "epoch": 0.007898008243104928,
+      "grad_norm": 2.0495479106903076,
+      "learning_rate": 1.9824824824824827e-05,
+      "loss": 1.7853,
+      "step": 6042
+    },
+    {
+      "epoch": 0.0078993154275212,
+      "grad_norm": 1.4424266815185547,
+      "learning_rate": 1.981981981981982e-05,
+      "loss": 1.4473,
+      "step": 6043
+    },
+    {
+      "epoch": 0.007900622611937469,
+      "grad_norm": 0.9338734745979309,
+      "learning_rate": 1.9814814814814816e-05,
+      "loss": 0.5369,
+      "step": 6044
+    },
+    {
+      "epoch": 0.00790192979635374,
+      "grad_norm": 1.2991952896118164,
+      "learning_rate": 1.980980980980981e-05,
+      "loss": 1.1915,
+      "step": 6045
+    },
+    {
+      "epoch": 0.00790323698077001,
+      "grad_norm": 1.422650694847107,
+      "learning_rate": 1.9804804804804806e-05,
+      "loss": 1.5362,
+      "step": 6046
+    },
+    {
+      "epoch": 0.00790454416518628,
+      "grad_norm": 1.8383716344833374,
+      "learning_rate": 1.9799799799799802e-05,
+      "loss": 1.8643,
+      "step": 6047
+    },
+    {
+      "epoch": 0.00790585134960255,
+      "grad_norm": 1.4194101095199585,
+      "learning_rate": 1.9794794794794795e-05,
+      "loss": 2.3358,
+      "step": 6048
+    },
+    {
+      "epoch": 0.007907158534018821,
+      "grad_norm": 1.4250028133392334,
+      "learning_rate": 1.9789789789789788e-05,
+      "loss": 1.6624,
+      "step": 6049
+    },
+    {
+      "epoch": 0.007908465718435091,
+      "grad_norm": 1.3557292222976685,
+      "learning_rate": 1.9784784784784784e-05,
+      "loss": 1.7434,
+      "step": 6050
+    },
+    {
+      "epoch": 0.007909772902851362,
+      "grad_norm": 1.2083582878112793,
+      "learning_rate": 1.977977977977978e-05,
+      "loss": 0.9261,
+      "step": 6051
+    },
+    {
+      "epoch": 0.007911080087267632,
+      "grad_norm": 2.079345226287842,
+      "learning_rate": 1.9774774774774777e-05,
+      "loss": 2.0536,
+      "step": 6052
+    },
+    {
+      "epoch": 0.007912387271683903,
+      "grad_norm": 1.5318576097488403,
+      "learning_rate": 1.9769769769769773e-05,
+      "loss": 1.6722,
+      "step": 6053
+    },
+    {
+      "epoch": 0.007913694456100172,
+      "grad_norm": 1.7844712734222412,
+      "learning_rate": 1.9764764764764766e-05,
+      "loss": 2.1152,
+      "step": 6054
+    },
+    {
+      "epoch": 0.007915001640516442,
+      "grad_norm": 1.3009767532348633,
+      "learning_rate": 1.975975975975976e-05,
+      "loss": 1.542,
+      "step": 6055
+    },
+    {
+      "epoch": 0.007916308824932713,
+      "grad_norm": 1.485008716583252,
+      "learning_rate": 1.9754754754754756e-05,
+      "loss": 1.8358,
+      "step": 6056
+    },
+    {
+      "epoch": 0.007917616009348982,
+      "grad_norm": 1.435690999031067,
+      "learning_rate": 1.9749749749749752e-05,
+      "loss": 1.4922,
+      "step": 6057
+    },
+    {
+      "epoch": 0.007918923193765253,
+      "grad_norm": 1.1563990116119385,
+      "learning_rate": 1.9744744744744745e-05,
+      "loss": 2.2099,
+      "step": 6058
+    },
+    {
+      "epoch": 0.007920230378181523,
+      "grad_norm": 1.676208257675171,
+      "learning_rate": 1.973973973973974e-05,
+      "loss": 1.4974,
+      "step": 6059
+    },
+    {
+      "epoch": 0.007921537562597794,
+      "grad_norm": 1.5333201885223389,
+      "learning_rate": 1.9734734734734738e-05,
+      "loss": 1.0744,
+      "step": 6060
+    },
+    {
+      "epoch": 0.007922844747014064,
+      "grad_norm": 1.2672433853149414,
+      "learning_rate": 1.972972972972973e-05,
+      "loss": 1.2806,
+      "step": 6061
+    },
+    {
+      "epoch": 0.007924151931430335,
+      "grad_norm": 1.2498788833618164,
+      "learning_rate": 1.9724724724724724e-05,
+      "loss": 1.764,
+      "step": 6062
+    },
+    {
+      "epoch": 0.007925459115846604,
+      "grad_norm": 1.4103548526763916,
+      "learning_rate": 1.971971971971972e-05,
+      "loss": 1.3832,
+      "step": 6063
+    },
+    {
+      "epoch": 0.007926766300262875,
+      "grad_norm": 1.7981834411621094,
+      "learning_rate": 1.9714714714714716e-05,
+      "loss": 1.706,
+      "step": 6064
+    },
+    {
+      "epoch": 0.007928073484679145,
+      "grad_norm": 1.367862343788147,
+      "learning_rate": 1.9709709709709713e-05,
+      "loss": 1.9201,
+      "step": 6065
+    },
+    {
+      "epoch": 0.007929380669095416,
+      "grad_norm": 1.2502468824386597,
+      "learning_rate": 1.9704704704704706e-05,
+      "loss": 1.337,
+      "step": 6066
+    },
+    {
+      "epoch": 0.007930687853511685,
+      "grad_norm": 1.7942384481430054,
+      "learning_rate": 1.96996996996997e-05,
+      "loss": 1.6371,
+      "step": 6067
+    },
+    {
+      "epoch": 0.007931995037927957,
+      "grad_norm": 1.3712308406829834,
+      "learning_rate": 1.9694694694694695e-05,
+      "loss": 1.7206,
+      "step": 6068
+    },
+    {
+      "epoch": 0.007933302222344226,
+      "grad_norm": 1.642562985420227,
+      "learning_rate": 1.968968968968969e-05,
+      "loss": 1.6749,
+      "step": 6069
+    },
+    {
+      "epoch": 0.007934609406760496,
+      "grad_norm": 1.0280144214630127,
+      "learning_rate": 1.9684684684684684e-05,
+      "loss": 0.7152,
+      "step": 6070
+    },
+    {
+      "epoch": 0.007935916591176767,
+      "grad_norm": 1.942810297012329,
+      "learning_rate": 1.967967967967968e-05,
+      "loss": 1.7082,
+      "step": 6071
+    },
+    {
+      "epoch": 0.007937223775593036,
+      "grad_norm": 1.4168614149093628,
+      "learning_rate": 1.9674674674674677e-05,
+      "loss": 1.8884,
+      "step": 6072
+    },
+    {
+      "epoch": 0.007938530960009307,
+      "grad_norm": 1.3689569234848022,
+      "learning_rate": 1.966966966966967e-05,
+      "loss": 0.9782,
+      "step": 6073
+    },
+    {
+      "epoch": 0.007939838144425577,
+      "grad_norm": 0.8248261213302612,
+      "learning_rate": 1.9664664664664663e-05,
+      "loss": 0.295,
+      "step": 6074
+    },
+    {
+      "epoch": 0.007941145328841848,
+      "grad_norm": 1.6035082340240479,
+      "learning_rate": 1.965965965965966e-05,
+      "loss": 2.1499,
+      "step": 6075
+    },
+    {
+      "epoch": 0.007942452513258117,
+      "grad_norm": 1.8627866506576538,
+      "learning_rate": 1.9654654654654656e-05,
+      "loss": 1.5856,
+      "step": 6076
+    },
+    {
+      "epoch": 0.007943759697674389,
+      "grad_norm": 1.0543086528778076,
+      "learning_rate": 1.9649649649649652e-05,
+      "loss": 1.1674,
+      "step": 6077
+    },
+    {
+      "epoch": 0.007945066882090658,
+      "grad_norm": 1.2833918333053589,
+      "learning_rate": 1.964464464464465e-05,
+      "loss": 1.2559,
+      "step": 6078
+    },
+    {
+      "epoch": 0.00794637406650693,
+      "grad_norm": 1.6531795263290405,
+      "learning_rate": 1.963963963963964e-05,
+      "loss": 1.2288,
+      "step": 6079
+    },
+    {
+      "epoch": 0.007947681250923199,
+      "grad_norm": 1.384683609008789,
+      "learning_rate": 1.9634634634634634e-05,
+      "loss": 1.3754,
+      "step": 6080
+    },
+    {
+      "epoch": 0.00794898843533947,
+      "grad_norm": 1.5904109477996826,
+      "learning_rate": 1.962962962962963e-05,
+      "loss": 1.8205,
+      "step": 6081
+    },
+    {
+      "epoch": 0.00795029561975574,
+      "grad_norm": 2.374948740005493,
+      "learning_rate": 1.9624624624624627e-05,
+      "loss": 1.817,
+      "step": 6082
+    },
+    {
+      "epoch": 0.00795160280417201,
+      "grad_norm": 1.6804805994033813,
+      "learning_rate": 1.961961961961962e-05,
+      "loss": 1.0978,
+      "step": 6083
+    },
+    {
+      "epoch": 0.00795290998858828,
+      "grad_norm": 1.3803313970565796,
+      "learning_rate": 1.9614614614614616e-05,
+      "loss": 1.1608,
+      "step": 6084
+    },
+    {
+      "epoch": 0.00795421717300455,
+      "grad_norm": 1.1764214038848877,
+      "learning_rate": 1.960960960960961e-05,
+      "loss": 1.3941,
+      "step": 6085
+    },
+    {
+      "epoch": 0.00795552435742082,
+      "grad_norm": 1.5678223371505737,
+      "learning_rate": 1.9604604604604606e-05,
+      "loss": 1.5131,
+      "step": 6086
+    },
+    {
+      "epoch": 0.00795683154183709,
+      "grad_norm": 1.896040678024292,
+      "learning_rate": 1.95995995995996e-05,
+      "loss": 1.2205,
+      "step": 6087
+    },
+    {
+      "epoch": 0.007958138726253361,
+      "grad_norm": 0.46985968947410583,
+      "learning_rate": 1.9594594594594595e-05,
+      "loss": 0.1272,
+      "step": 6088
+    },
+    {
+      "epoch": 0.00795944591066963,
+      "grad_norm": 0.9546180367469788,
+      "learning_rate": 1.958958958958959e-05,
+      "loss": 0.8049,
+      "step": 6089
+    },
+    {
+      "epoch": 0.007960753095085902,
+      "grad_norm": 1.5061980485916138,
+      "learning_rate": 1.9584584584584588e-05,
+      "loss": 1.9616,
+      "step": 6090
+    },
+    {
+      "epoch": 0.007962060279502171,
+      "grad_norm": 1.126379132270813,
+      "learning_rate": 1.957957957957958e-05,
+      "loss": 0.861,
+      "step": 6091
+    },
+    {
+      "epoch": 0.007963367463918443,
+      "grad_norm": 1.5326406955718994,
+      "learning_rate": 1.9574574574574574e-05,
+      "loss": 1.9085,
+      "step": 6092
+    },
+    {
+      "epoch": 0.007964674648334712,
+      "grad_norm": 1.4049581289291382,
+      "learning_rate": 1.956956956956957e-05,
+      "loss": 2.0559,
+      "step": 6093
+    },
+    {
+      "epoch": 0.007965981832750983,
+      "grad_norm": 1.6101398468017578,
+      "learning_rate": 1.9564564564564567e-05,
+      "loss": 2.2862,
+      "step": 6094
+    },
+    {
+      "epoch": 0.007967289017167253,
+      "grad_norm": 1.624790072441101,
+      "learning_rate": 1.955955955955956e-05,
+      "loss": 2.0267,
+      "step": 6095
+    },
+    {
+      "epoch": 0.007968596201583524,
+      "grad_norm": 1.0847227573394775,
+      "learning_rate": 1.9554554554554556e-05,
+      "loss": 1.3289,
+      "step": 6096
+    },
+    {
+      "epoch": 0.007969903385999793,
+      "grad_norm": 1.2214972972869873,
+      "learning_rate": 1.9549549549549552e-05,
+      "loss": 1.7953,
+      "step": 6097
+    },
+    {
+      "epoch": 0.007971210570416064,
+      "grad_norm": 2.640829563140869,
+      "learning_rate": 1.9544544544544545e-05,
+      "loss": 2.5537,
+      "step": 6098
+    },
+    {
+      "epoch": 0.007972517754832334,
+      "grad_norm": 1.5014883279800415,
+      "learning_rate": 1.9539539539539538e-05,
+      "loss": 1.9184,
+      "step": 6099
+    },
+    {
+      "epoch": 0.007973824939248603,
+      "grad_norm": 1.6747337579727173,
+      "learning_rate": 1.9534534534534535e-05,
+      "loss": 2.1979,
+      "step": 6100
+    },
+    {
+      "epoch": 0.007975132123664875,
+      "grad_norm": 1.4607199430465698,
+      "learning_rate": 1.952952952952953e-05,
+      "loss": 1.614,
+      "step": 6101
+    },
+    {
+      "epoch": 0.007976439308081144,
+      "grad_norm": 1.759669542312622,
+      "learning_rate": 1.9524524524524527e-05,
+      "loss": 2.27,
+      "step": 6102
+    },
+    {
+      "epoch": 0.007977746492497415,
+      "grad_norm": 1.3246132135391235,
+      "learning_rate": 1.951951951951952e-05,
+      "loss": 1.8145,
+      "step": 6103
+    },
+    {
+      "epoch": 0.007979053676913685,
+      "grad_norm": 1.5541479587554932,
+      "learning_rate": 1.9514514514514513e-05,
+      "loss": 1.1111,
+      "step": 6104
+    },
+    {
+      "epoch": 0.007980360861329956,
+      "grad_norm": 1.443393349647522,
+      "learning_rate": 1.950950950950951e-05,
+      "loss": 2.1395,
+      "step": 6105
+    },
+    {
+      "epoch": 0.007981668045746225,
+      "grad_norm": 1.264482855796814,
+      "learning_rate": 1.9504504504504506e-05,
+      "loss": 1.7092,
+      "step": 6106
+    },
+    {
+      "epoch": 0.007982975230162496,
+      "grad_norm": 1.2724640369415283,
+      "learning_rate": 1.9499499499499502e-05,
+      "loss": 1.5889,
+      "step": 6107
+    },
+    {
+      "epoch": 0.007984282414578766,
+      "grad_norm": 1.5021992921829224,
+      "learning_rate": 1.9494494494494495e-05,
+      "loss": 2.5287,
+      "step": 6108
+    },
+    {
+      "epoch": 0.007985589598995037,
+      "grad_norm": 1.4258524179458618,
+      "learning_rate": 1.948948948948949e-05,
+      "loss": 1.9211,
+      "step": 6109
+    },
+    {
+      "epoch": 0.007986896783411307,
+      "grad_norm": 1.5243544578552246,
+      "learning_rate": 1.9484484484484485e-05,
+      "loss": 2.264,
+      "step": 6110
+    },
+    {
+      "epoch": 0.007988203967827578,
+      "grad_norm": 1.183401107788086,
+      "learning_rate": 1.947947947947948e-05,
+      "loss": 1.2424,
+      "step": 6111
+    },
+    {
+      "epoch": 0.007989511152243847,
+      "grad_norm": 1.423277497291565,
+      "learning_rate": 1.9474474474474474e-05,
+      "loss": 1.9422,
+      "step": 6112
+    },
+    {
+      "epoch": 0.007990818336660118,
+      "grad_norm": 1.259158968925476,
+      "learning_rate": 1.946946946946947e-05,
+      "loss": 1.4313,
+      "step": 6113
+    },
+    {
+      "epoch": 0.007992125521076388,
+      "grad_norm": 1.345898985862732,
+      "learning_rate": 1.9464464464464467e-05,
+      "loss": 1.6715,
+      "step": 6114
+    },
+    {
+      "epoch": 0.007993432705492659,
+      "grad_norm": 1.167940378189087,
+      "learning_rate": 1.9459459459459463e-05,
+      "loss": 0.6928,
+      "step": 6115
+    },
+    {
+      "epoch": 0.007994739889908929,
+      "grad_norm": 1.5921193361282349,
+      "learning_rate": 1.9454454454454456e-05,
+      "loss": 1.4147,
+      "step": 6116
+    },
+    {
+      "epoch": 0.007996047074325198,
+      "grad_norm": 1.2775211334228516,
+      "learning_rate": 1.944944944944945e-05,
+      "loss": 1.7135,
+      "step": 6117
+    },
+    {
+      "epoch": 0.00799735425874147,
+      "grad_norm": 1.5547246932983398,
+      "learning_rate": 1.9444444444444445e-05,
+      "loss": 1.9371,
+      "step": 6118
+    },
+    {
+      "epoch": 0.007998661443157739,
+      "grad_norm": 1.4164988994598389,
+      "learning_rate": 1.9439439439439442e-05,
+      "loss": 1.5117,
+      "step": 6119
+    },
+    {
+      "epoch": 0.00799996862757401,
+      "grad_norm": 1.5199204683303833,
+      "learning_rate": 1.9434434434434435e-05,
+      "loss": 2.2686,
+      "step": 6120
+    },
+    {
+      "epoch": 0.00800127581199028,
+      "grad_norm": 1.4213216304779053,
+      "learning_rate": 1.942942942942943e-05,
+      "loss": 1.7023,
+      "step": 6121
+    },
+    {
+      "epoch": 0.00800258299640655,
+      "grad_norm": 1.6998862028121948,
+      "learning_rate": 1.9424424424424424e-05,
+      "loss": 1.6497,
+      "step": 6122
+    },
+    {
+      "epoch": 0.00800389018082282,
+      "grad_norm": 1.7159817218780518,
+      "learning_rate": 1.941941941941942e-05,
+      "loss": 1.7818,
+      "step": 6123
+    },
+    {
+      "epoch": 0.008005197365239091,
+      "grad_norm": 1.4591140747070312,
+      "learning_rate": 1.9414414414414413e-05,
+      "loss": 1.5965,
+      "step": 6124
+    },
+    {
+      "epoch": 0.00800650454965536,
+      "grad_norm": 1.8335078954696655,
+      "learning_rate": 1.940940940940941e-05,
+      "loss": 1.3548,
+      "step": 6125
+    },
+    {
+      "epoch": 0.008007811734071632,
+      "grad_norm": 1.4114692211151123,
+      "learning_rate": 1.9404404404404406e-05,
+      "loss": 1.9892,
+      "step": 6126
+    },
+    {
+      "epoch": 0.008009118918487901,
+      "grad_norm": 1.5428956747055054,
+      "learning_rate": 1.9399399399399402e-05,
+      "loss": 1.7548,
+      "step": 6127
+    },
+    {
+      "epoch": 0.008010426102904172,
+      "grad_norm": 1.2888081073760986,
+      "learning_rate": 1.9394394394394395e-05,
+      "loss": 1.0153,
+      "step": 6128
+    },
+    {
+      "epoch": 0.008011733287320442,
+      "grad_norm": 1.1248282194137573,
+      "learning_rate": 1.938938938938939e-05,
+      "loss": 0.9063,
+      "step": 6129
+    },
+    {
+      "epoch": 0.008013040471736713,
+      "grad_norm": 1.2655255794525146,
+      "learning_rate": 1.9384384384384385e-05,
+      "loss": 1.9616,
+      "step": 6130
+    },
+    {
+      "epoch": 0.008014347656152982,
+      "grad_norm": 1.486942172050476,
+      "learning_rate": 1.937937937937938e-05,
+      "loss": 1.9468,
+      "step": 6131
+    },
+    {
+      "epoch": 0.008015654840569252,
+      "grad_norm": 1.1762542724609375,
+      "learning_rate": 1.9374374374374378e-05,
+      "loss": 1.8287,
+      "step": 6132
+    },
+    {
+      "epoch": 0.008016962024985523,
+      "grad_norm": 1.5870975255966187,
+      "learning_rate": 1.936936936936937e-05,
+      "loss": 1.6538,
+      "step": 6133
+    },
+    {
+      "epoch": 0.008018269209401793,
+      "grad_norm": 1.537447452545166,
+      "learning_rate": 1.9364364364364367e-05,
+      "loss": 1.9986,
+      "step": 6134
+    },
+    {
+      "epoch": 0.008019576393818064,
+      "grad_norm": 1.3566901683807373,
+      "learning_rate": 1.935935935935936e-05,
+      "loss": 1.5805,
+      "step": 6135
+    },
+    {
+      "epoch": 0.008020883578234333,
+      "grad_norm": 1.6904418468475342,
+      "learning_rate": 1.9354354354354356e-05,
+      "loss": 1.2211,
+      "step": 6136
+    },
+    {
+      "epoch": 0.008022190762650604,
+      "grad_norm": 2.1861579418182373,
+      "learning_rate": 1.934934934934935e-05,
+      "loss": 2.5584,
+      "step": 6137
+    },
+    {
+      "epoch": 0.008023497947066874,
+      "grad_norm": 1.5565972328186035,
+      "learning_rate": 1.9344344344344346e-05,
+      "loss": 1.9144,
+      "step": 6138
+    },
+    {
+      "epoch": 0.008024805131483145,
+      "grad_norm": 1.5528043508529663,
+      "learning_rate": 1.9339339339339342e-05,
+      "loss": 2.6159,
+      "step": 6139
+    },
+    {
+      "epoch": 0.008026112315899414,
+      "grad_norm": 1.2439682483673096,
+      "learning_rate": 1.9334334334334335e-05,
+      "loss": 1.0868,
+      "step": 6140
+    },
+    {
+      "epoch": 0.008027419500315686,
+      "grad_norm": 1.5231190919876099,
+      "learning_rate": 1.9329329329329328e-05,
+      "loss": 1.3871,
+      "step": 6141
+    },
+    {
+      "epoch": 0.008028726684731955,
+      "grad_norm": 1.4658141136169434,
+      "learning_rate": 1.9324324324324324e-05,
+      "loss": 1.4852,
+      "step": 6142
+    },
+    {
+      "epoch": 0.008030033869148226,
+      "grad_norm": 1.416326880455017,
+      "learning_rate": 1.931931931931932e-05,
+      "loss": 1.2196,
+      "step": 6143
+    },
+    {
+      "epoch": 0.008031341053564496,
+      "grad_norm": 1.3268506526947021,
+      "learning_rate": 1.9314314314314317e-05,
+      "loss": 1.3436,
+      "step": 6144
+    },
+    {
+      "epoch": 0.008032648237980767,
+      "grad_norm": 1.2846823930740356,
+      "learning_rate": 1.930930930930931e-05,
+      "loss": 1.2839,
+      "step": 6145
+    },
+    {
+      "epoch": 0.008033955422397036,
+      "grad_norm": 1.5852793455123901,
+      "learning_rate": 1.9304304304304306e-05,
+      "loss": 1.8794,
+      "step": 6146
+    },
+    {
+      "epoch": 0.008035262606813306,
+      "grad_norm": 1.5016613006591797,
+      "learning_rate": 1.92992992992993e-05,
+      "loss": 2.0628,
+      "step": 6147
+    },
+    {
+      "epoch": 0.008036569791229577,
+      "grad_norm": 1.1888130903244019,
+      "learning_rate": 1.9294294294294296e-05,
+      "loss": 1.5251,
+      "step": 6148
+    },
+    {
+      "epoch": 0.008037876975645846,
+      "grad_norm": 1.2947707176208496,
+      "learning_rate": 1.928928928928929e-05,
+      "loss": 1.6487,
+      "step": 6149
+    },
+    {
+      "epoch": 0.008039184160062118,
+      "grad_norm": 1.3715989589691162,
+      "learning_rate": 1.9284284284284285e-05,
+      "loss": 1.9133,
+      "step": 6150
+    },
+    {
+      "epoch": 0.008040491344478387,
+      "grad_norm": 1.3147355318069458,
+      "learning_rate": 1.927927927927928e-05,
+      "loss": 1.621,
+      "step": 6151
+    },
+    {
+      "epoch": 0.008041798528894658,
+      "grad_norm": 2.3466567993164062,
+      "learning_rate": 1.9274274274274278e-05,
+      "loss": 1.7046,
+      "step": 6152
+    },
+    {
+      "epoch": 0.008043105713310928,
+      "grad_norm": 1.322914958000183,
+      "learning_rate": 1.9269269269269267e-05,
+      "loss": 1.7768,
+      "step": 6153
+    },
+    {
+      "epoch": 0.008044412897727199,
+      "grad_norm": 1.2623040676116943,
+      "learning_rate": 1.9264264264264264e-05,
+      "loss": 1.663,
+      "step": 6154
+    },
+    {
+      "epoch": 0.008045720082143468,
+      "grad_norm": 1.4054944515228271,
+      "learning_rate": 1.925925925925926e-05,
+      "loss": 1.2592,
+      "step": 6155
+    },
+    {
+      "epoch": 0.00804702726655974,
+      "grad_norm": 1.8334286212921143,
+      "learning_rate": 1.9254254254254256e-05,
+      "loss": 2.6598,
+      "step": 6156
+    },
+    {
+      "epoch": 0.008048334450976009,
+      "grad_norm": 1.4661685228347778,
+      "learning_rate": 1.9249249249249253e-05,
+      "loss": 2.0098,
+      "step": 6157
+    },
+    {
+      "epoch": 0.00804964163539228,
+      "grad_norm": 1.1246362924575806,
+      "learning_rate": 1.9244244244244246e-05,
+      "loss": 1.3912,
+      "step": 6158
+    },
+    {
+      "epoch": 0.00805094881980855,
+      "grad_norm": 1.323487639427185,
+      "learning_rate": 1.923923923923924e-05,
+      "loss": 1.0249,
+      "step": 6159
+    },
+    {
+      "epoch": 0.00805225600422482,
+      "grad_norm": 1.5399820804595947,
+      "learning_rate": 1.9234234234234235e-05,
+      "loss": 1.6433,
+      "step": 6160
+    },
+    {
+      "epoch": 0.00805356318864109,
+      "grad_norm": 1.564528465270996,
+      "learning_rate": 1.922922922922923e-05,
+      "loss": 1.8904,
+      "step": 6161
+    },
+    {
+      "epoch": 0.00805487037305736,
+      "grad_norm": 1.6742267608642578,
+      "learning_rate": 1.9224224224224224e-05,
+      "loss": 2.161,
+      "step": 6162
+    },
+    {
+      "epoch": 0.008056177557473631,
+      "grad_norm": 1.8857253789901733,
+      "learning_rate": 1.921921921921922e-05,
+      "loss": 2.2783,
+      "step": 6163
+    },
+    {
+      "epoch": 0.0080574847418899,
+      "grad_norm": 1.2992154359817505,
+      "learning_rate": 1.9214214214214217e-05,
+      "loss": 1.1323,
+      "step": 6164
+    },
+    {
+      "epoch": 0.008058791926306172,
+      "grad_norm": 1.6384350061416626,
+      "learning_rate": 1.920920920920921e-05,
+      "loss": 2.1522,
+      "step": 6165
+    },
+    {
+      "epoch": 0.008060099110722441,
+      "grad_norm": 1.3298938274383545,
+      "learning_rate": 1.9204204204204203e-05,
+      "loss": 1.2396,
+      "step": 6166
+    },
+    {
+      "epoch": 0.008061406295138712,
+      "grad_norm": 1.585729718208313,
+      "learning_rate": 1.91991991991992e-05,
+      "loss": 1.2672,
+      "step": 6167
+    },
+    {
+      "epoch": 0.008062713479554982,
+      "grad_norm": 1.2965375185012817,
+      "learning_rate": 1.9194194194194196e-05,
+      "loss": 2.0344,
+      "step": 6168
+    },
+    {
+      "epoch": 0.008064020663971253,
+      "grad_norm": 0.550769567489624,
+      "learning_rate": 1.9189189189189192e-05,
+      "loss": 0.1822,
+      "step": 6169
+    },
+    {
+      "epoch": 0.008065327848387522,
+      "grad_norm": 1.3698910474777222,
+      "learning_rate": 1.9184184184184185e-05,
+      "loss": 1.3619,
+      "step": 6170
+    },
+    {
+      "epoch": 0.008066635032803793,
+      "grad_norm": 1.2158056497573853,
+      "learning_rate": 1.9179179179179178e-05,
+      "loss": 1.7263,
+      "step": 6171
+    },
+    {
+      "epoch": 0.008067942217220063,
+      "grad_norm": 1.0512239933013916,
+      "learning_rate": 1.9174174174174174e-05,
+      "loss": 1.3478,
+      "step": 6172
+    },
+    {
+      "epoch": 0.008069249401636334,
+      "grad_norm": 1.4450831413269043,
+      "learning_rate": 1.916916916916917e-05,
+      "loss": 1.7654,
+      "step": 6173
+    },
+    {
+      "epoch": 0.008070556586052604,
+      "grad_norm": 0.9941189885139465,
+      "learning_rate": 1.9164164164164164e-05,
+      "loss": 0.6314,
+      "step": 6174
+    },
+    {
+      "epoch": 0.008071863770468875,
+      "grad_norm": 1.5783801078796387,
+      "learning_rate": 1.915915915915916e-05,
+      "loss": 1.0881,
+      "step": 6175
+    },
+    {
+      "epoch": 0.008073170954885144,
+      "grad_norm": 1.6749112606048584,
+      "learning_rate": 1.9154154154154156e-05,
+      "loss": 1.8352,
+      "step": 6176
+    },
+    {
+      "epoch": 0.008074478139301414,
+      "grad_norm": 1.4497984647750854,
+      "learning_rate": 1.914914914914915e-05,
+      "loss": 1.5423,
+      "step": 6177
+    },
+    {
+      "epoch": 0.008075785323717685,
+      "grad_norm": 1.3580571413040161,
+      "learning_rate": 1.9144144144144142e-05,
+      "loss": 1.483,
+      "step": 6178
+    },
+    {
+      "epoch": 0.008077092508133954,
+      "grad_norm": 1.3825535774230957,
+      "learning_rate": 1.913913913913914e-05,
+      "loss": 2.041,
+      "step": 6179
+    },
+    {
+      "epoch": 0.008078399692550225,
+      "grad_norm": 1.5147444009780884,
+      "learning_rate": 1.9134134134134135e-05,
+      "loss": 1.2938,
+      "step": 6180
+    },
+    {
+      "epoch": 0.008079706876966495,
+      "grad_norm": 1.19016695022583,
+      "learning_rate": 1.912912912912913e-05,
+      "loss": 1.3275,
+      "step": 6181
+    },
+    {
+      "epoch": 0.008081014061382766,
+      "grad_norm": 1.6266186237335205,
+      "learning_rate": 1.9124124124124124e-05,
+      "loss": 1.3929,
+      "step": 6182
+    },
+    {
+      "epoch": 0.008082321245799036,
+      "grad_norm": 1.451813817024231,
+      "learning_rate": 1.911911911911912e-05,
+      "loss": 1.5684,
+      "step": 6183
+    },
+    {
+      "epoch": 0.008083628430215307,
+      "grad_norm": 1.2928999662399292,
+      "learning_rate": 1.9114114114114114e-05,
+      "loss": 1.3881,
+      "step": 6184
+    },
+    {
+      "epoch": 0.008084935614631576,
+      "grad_norm": 0.980348527431488,
+      "learning_rate": 1.910910910910911e-05,
+      "loss": 0.9151,
+      "step": 6185
+    },
+    {
+      "epoch": 0.008086242799047847,
+      "grad_norm": 1.458312749862671,
+      "learning_rate": 1.9104104104104107e-05,
+      "loss": 1.2931,
+      "step": 6186
+    },
+    {
+      "epoch": 0.008087549983464117,
+      "grad_norm": 1.3756217956542969,
+      "learning_rate": 1.90990990990991e-05,
+      "loss": 1.5794,
+      "step": 6187
+    },
+    {
+      "epoch": 0.008088857167880388,
+      "grad_norm": 1.4012911319732666,
+      "learning_rate": 1.9094094094094096e-05,
+      "loss": 2.6644,
+      "step": 6188
+    },
+    {
+      "epoch": 0.008090164352296658,
+      "grad_norm": 1.6195783615112305,
+      "learning_rate": 1.9089089089089092e-05,
+      "loss": 2.3047,
+      "step": 6189
+    },
+    {
+      "epoch": 0.008091471536712929,
+      "grad_norm": 1.749226450920105,
+      "learning_rate": 1.9084084084084085e-05,
+      "loss": 2.4146,
+      "step": 6190
+    },
+    {
+      "epoch": 0.008092778721129198,
+      "grad_norm": 1.2322548627853394,
+      "learning_rate": 1.9079079079079078e-05,
+      "loss": 1.1,
+      "step": 6191
+    },
+    {
+      "epoch": 0.008094085905545468,
+      "grad_norm": 1.1624760627746582,
+      "learning_rate": 1.9074074074074075e-05,
+      "loss": 1.3543,
+      "step": 6192
+    },
+    {
+      "epoch": 0.008095393089961739,
+      "grad_norm": 1.5301541090011597,
+      "learning_rate": 1.906906906906907e-05,
+      "loss": 0.9232,
+      "step": 6193
+    },
+    {
+      "epoch": 0.008096700274378008,
+      "grad_norm": 1.4229650497436523,
+      "learning_rate": 1.9064064064064067e-05,
+      "loss": 1.3616,
+      "step": 6194
+    },
+    {
+      "epoch": 0.00809800745879428,
+      "grad_norm": 1.17621910572052,
+      "learning_rate": 1.905905905905906e-05,
+      "loss": 1.6236,
+      "step": 6195
+    },
+    {
+      "epoch": 0.008099314643210549,
+      "grad_norm": 1.4786063432693481,
+      "learning_rate": 1.9054054054054053e-05,
+      "loss": 1.3461,
+      "step": 6196
+    },
+    {
+      "epoch": 0.00810062182762682,
+      "grad_norm": 1.46193265914917,
+      "learning_rate": 1.904904904904905e-05,
+      "loss": 2.5997,
+      "step": 6197
+    },
+    {
+      "epoch": 0.00810192901204309,
+      "grad_norm": 1.765720009803772,
+      "learning_rate": 1.9044044044044046e-05,
+      "loss": 1.8416,
+      "step": 6198
+    },
+    {
+      "epoch": 0.00810323619645936,
+      "grad_norm": 1.6336590051651,
+      "learning_rate": 1.903903903903904e-05,
+      "loss": 1.1335,
+      "step": 6199
+    },
+    {
+      "epoch": 0.00810454338087563,
+      "grad_norm": 1.5237914323806763,
+      "learning_rate": 1.9034034034034035e-05,
+      "loss": 1.5523,
+      "step": 6200
+    },
+    {
+      "epoch": 0.008105850565291901,
+      "grad_norm": 1.3794652223587036,
+      "learning_rate": 1.902902902902903e-05,
+      "loss": 0.8291,
+      "step": 6201
+    },
+    {
+      "epoch": 0.00810715774970817,
+      "grad_norm": 1.3972578048706055,
+      "learning_rate": 1.9024024024024025e-05,
+      "loss": 2.3362,
+      "step": 6202
+    },
+    {
+      "epoch": 0.008108464934124442,
+      "grad_norm": 1.2730313539505005,
+      "learning_rate": 1.9019019019019018e-05,
+      "loss": 1.6792,
+      "step": 6203
+    },
+    {
+      "epoch": 0.008109772118540711,
+      "grad_norm": 1.398695468902588,
+      "learning_rate": 1.9014014014014014e-05,
+      "loss": 1.8479,
+      "step": 6204
+    },
+    {
+      "epoch": 0.008111079302956983,
+      "grad_norm": 1.6490472555160522,
+      "learning_rate": 1.900900900900901e-05,
+      "loss": 1.6974,
+      "step": 6205
+    },
+    {
+      "epoch": 0.008112386487373252,
+      "grad_norm": 1.6511844396591187,
+      "learning_rate": 1.9004004004004007e-05,
+      "loss": 1.3875,
+      "step": 6206
+    },
+    {
+      "epoch": 0.008113693671789522,
+      "grad_norm": 1.9420377016067505,
+      "learning_rate": 1.8998998998999e-05,
+      "loss": 2.4133,
+      "step": 6207
+    },
+    {
+      "epoch": 0.008115000856205793,
+      "grad_norm": 1.6544657945632935,
+      "learning_rate": 1.8993993993993993e-05,
+      "loss": 1.7459,
+      "step": 6208
+    },
+    {
+      "epoch": 0.008116308040622062,
+      "grad_norm": 1.6598191261291504,
+      "learning_rate": 1.898898898898899e-05,
+      "loss": 2.7337,
+      "step": 6209
+    },
+    {
+      "epoch": 0.008117615225038333,
+      "grad_norm": 0.9263902902603149,
+      "learning_rate": 1.8983983983983985e-05,
+      "loss": 0.7052,
+      "step": 6210
+    },
+    {
+      "epoch": 0.008118922409454603,
+      "grad_norm": 1.3565893173217773,
+      "learning_rate": 1.8978978978978982e-05,
+      "loss": 2.1465,
+      "step": 6211
+    },
+    {
+      "epoch": 0.008120229593870874,
+      "grad_norm": 0.5910101532936096,
+      "learning_rate": 1.8973973973973975e-05,
+      "loss": 0.2014,
+      "step": 6212
+    },
+    {
+      "epoch": 0.008121536778287143,
+      "grad_norm": 1.3576607704162598,
+      "learning_rate": 1.896896896896897e-05,
+      "loss": 1.9255,
+      "step": 6213
+    },
+    {
+      "epoch": 0.008122843962703415,
+      "grad_norm": 0.9894116520881653,
+      "learning_rate": 1.8963963963963964e-05,
+      "loss": 0.8712,
+      "step": 6214
+    },
+    {
+      "epoch": 0.008124151147119684,
+      "grad_norm": 1.3976962566375732,
+      "learning_rate": 1.895895895895896e-05,
+      "loss": 2.5015,
+      "step": 6215
+    },
+    {
+      "epoch": 0.008125458331535955,
+      "grad_norm": 1.5237274169921875,
+      "learning_rate": 1.8953953953953953e-05,
+      "loss": 1.2908,
+      "step": 6216
+    },
+    {
+      "epoch": 0.008126765515952225,
+      "grad_norm": 1.40516197681427,
+      "learning_rate": 1.894894894894895e-05,
+      "loss": 1.5094,
+      "step": 6217
+    },
+    {
+      "epoch": 0.008128072700368496,
+      "grad_norm": 1.4092708826065063,
+      "learning_rate": 1.8943943943943946e-05,
+      "loss": 1.6634,
+      "step": 6218
+    },
+    {
+      "epoch": 0.008129379884784765,
+      "grad_norm": 1.1871017217636108,
+      "learning_rate": 1.8938938938938942e-05,
+      "loss": 1.4813,
+      "step": 6219
+    },
+    {
+      "epoch": 0.008130687069201037,
+      "grad_norm": 1.4214684963226318,
+      "learning_rate": 1.8933933933933935e-05,
+      "loss": 1.7585,
+      "step": 6220
+    },
+    {
+      "epoch": 0.008131994253617306,
+      "grad_norm": 1.2758532762527466,
+      "learning_rate": 1.892892892892893e-05,
+      "loss": 1.3986,
+      "step": 6221
+    },
+    {
+      "epoch": 0.008133301438033577,
+      "grad_norm": 1.713358759880066,
+      "learning_rate": 1.8923923923923925e-05,
+      "loss": 2.0334,
+      "step": 6222
+    },
+    {
+      "epoch": 0.008134608622449847,
+      "grad_norm": 1.5374258756637573,
+      "learning_rate": 1.891891891891892e-05,
+      "loss": 1.7108,
+      "step": 6223
+    },
+    {
+      "epoch": 0.008135915806866116,
+      "grad_norm": 1.7605754137039185,
+      "learning_rate": 1.8913913913913914e-05,
+      "loss": 2.222,
+      "step": 6224
+    },
+    {
+      "epoch": 0.008137222991282387,
+      "grad_norm": 1.2366617918014526,
+      "learning_rate": 1.890890890890891e-05,
+      "loss": 1.8613,
+      "step": 6225
+    },
+    {
+      "epoch": 0.008138530175698657,
+      "grad_norm": 1.444818139076233,
+      "learning_rate": 1.8903903903903903e-05,
+      "loss": 1.5177,
+      "step": 6226
+    },
+    {
+      "epoch": 0.008139837360114928,
+      "grad_norm": 1.5056275129318237,
+      "learning_rate": 1.88988988988989e-05,
+      "loss": 1.5705,
+      "step": 6227
+    },
+    {
+      "epoch": 0.008141144544531197,
+      "grad_norm": 0.9855378270149231,
+      "learning_rate": 1.8893893893893893e-05,
+      "loss": 0.8797,
+      "step": 6228
+    },
+    {
+      "epoch": 0.008142451728947469,
+      "grad_norm": 2.097561836242676,
+      "learning_rate": 1.888888888888889e-05,
+      "loss": 2.0757,
+      "step": 6229
+    },
+    {
+      "epoch": 0.008143758913363738,
+      "grad_norm": 1.7507606744766235,
+      "learning_rate": 1.8883883883883885e-05,
+      "loss": 2.1304,
+      "step": 6230
+    },
+    {
+      "epoch": 0.00814506609778001,
+      "grad_norm": 1.4545581340789795,
+      "learning_rate": 1.8878878878878882e-05,
+      "loss": 1.7952,
+      "step": 6231
+    },
+    {
+      "epoch": 0.008146373282196279,
+      "grad_norm": 1.4387304782867432,
+      "learning_rate": 1.8873873873873875e-05,
+      "loss": 1.3216,
+      "step": 6232
+    },
+    {
+      "epoch": 0.00814768046661255,
+      "grad_norm": 1.1662107706069946,
+      "learning_rate": 1.8868868868868868e-05,
+      "loss": 1.0877,
+      "step": 6233
+    },
+    {
+      "epoch": 0.00814898765102882,
+      "grad_norm": 1.3542121648788452,
+      "learning_rate": 1.8863863863863864e-05,
+      "loss": 1.2024,
+      "step": 6234
+    },
+    {
+      "epoch": 0.00815029483544509,
+      "grad_norm": 1.3407913446426392,
+      "learning_rate": 1.885885885885886e-05,
+      "loss": 1.2123,
+      "step": 6235
+    },
+    {
+      "epoch": 0.00815160201986136,
+      "grad_norm": 1.3079156875610352,
+      "learning_rate": 1.8853853853853853e-05,
+      "loss": 1.8068,
+      "step": 6236
+    },
+    {
+      "epoch": 0.008152909204277631,
+      "grad_norm": 1.6254807710647583,
+      "learning_rate": 1.884884884884885e-05,
+      "loss": 2.0499,
+      "step": 6237
+    },
+    {
+      "epoch": 0.0081542163886939,
+      "grad_norm": 1.4348829984664917,
+      "learning_rate": 1.8843843843843846e-05,
+      "loss": 1.4099,
+      "step": 6238
+    },
+    {
+      "epoch": 0.00815552357311017,
+      "grad_norm": 1.4442334175109863,
+      "learning_rate": 1.883883883883884e-05,
+      "loss": 1.0081,
+      "step": 6239
+    },
+    {
+      "epoch": 0.008156830757526441,
+      "grad_norm": 1.6133842468261719,
+      "learning_rate": 1.8833833833833836e-05,
+      "loss": 2.1803,
+      "step": 6240
+    },
+    {
+      "epoch": 0.00815813794194271,
+      "grad_norm": 1.140149474143982,
+      "learning_rate": 1.882882882882883e-05,
+      "loss": 1.5925,
+      "step": 6241
+    },
+    {
+      "epoch": 0.008159445126358982,
+      "grad_norm": 1.4427179098129272,
+      "learning_rate": 1.8823823823823825e-05,
+      "loss": 1.6755,
+      "step": 6242
+    },
+    {
+      "epoch": 0.008160752310775251,
+      "grad_norm": 1.301640510559082,
+      "learning_rate": 1.881881881881882e-05,
+      "loss": 1.4266,
+      "step": 6243
+    },
+    {
+      "epoch": 0.008162059495191522,
+      "grad_norm": 1.171281099319458,
+      "learning_rate": 1.8813813813813814e-05,
+      "loss": 0.6642,
+      "step": 6244
+    },
+    {
+      "epoch": 0.008163366679607792,
+      "grad_norm": 1.2555326223373413,
+      "learning_rate": 1.8808808808808807e-05,
+      "loss": 0.9114,
+      "step": 6245
+    },
+    {
+      "epoch": 0.008164673864024063,
+      "grad_norm": 1.3973064422607422,
+      "learning_rate": 1.8803803803803804e-05,
+      "loss": 1.7672,
+      "step": 6246
+    },
+    {
+      "epoch": 0.008165981048440333,
+      "grad_norm": 1.4005424976348877,
+      "learning_rate": 1.87987987987988e-05,
+      "loss": 1.5372,
+      "step": 6247
+    },
+    {
+      "epoch": 0.008167288232856604,
+      "grad_norm": 1.2880964279174805,
+      "learning_rate": 1.8793793793793796e-05,
+      "loss": 1.8228,
+      "step": 6248
+    },
+    {
+      "epoch": 0.008168595417272873,
+      "grad_norm": 1.05202317237854,
+      "learning_rate": 1.878878878878879e-05,
+      "loss": 0.5889,
+      "step": 6249
+    },
+    {
+      "epoch": 0.008169902601689144,
+      "grad_norm": 1.3962854146957397,
+      "learning_rate": 1.8783783783783786e-05,
+      "loss": 1.6393,
+      "step": 6250
+    },
+    {
+      "epoch": 0.008171209786105414,
+      "grad_norm": 1.4954115152359009,
+      "learning_rate": 1.877877877877878e-05,
+      "loss": 1.5322,
+      "step": 6251
+    },
+    {
+      "epoch": 0.008172516970521685,
+      "grad_norm": 1.7493923902511597,
+      "learning_rate": 1.8773773773773775e-05,
+      "loss": 1.8181,
+      "step": 6252
+    },
+    {
+      "epoch": 0.008173824154937955,
+      "grad_norm": 1.1511130332946777,
+      "learning_rate": 1.8768768768768768e-05,
+      "loss": 0.7254,
+      "step": 6253
+    },
+    {
+      "epoch": 0.008175131339354224,
+      "grad_norm": 1.4653687477111816,
+      "learning_rate": 1.8763763763763764e-05,
+      "loss": 2.0783,
+      "step": 6254
+    },
+    {
+      "epoch": 0.008176438523770495,
+      "grad_norm": 1.4202955961227417,
+      "learning_rate": 1.875875875875876e-05,
+      "loss": 1.597,
+      "step": 6255
+    },
+    {
+      "epoch": 0.008177745708186765,
+      "grad_norm": 1.2466477155685425,
+      "learning_rate": 1.8753753753753757e-05,
+      "loss": 1.3347,
+      "step": 6256
+    },
+    {
+      "epoch": 0.008179052892603036,
+      "grad_norm": 1.5622001886367798,
+      "learning_rate": 1.874874874874875e-05,
+      "loss": 1.718,
+      "step": 6257
+    },
+    {
+      "epoch": 0.008180360077019305,
+      "grad_norm": 1.333413004875183,
+      "learning_rate": 1.8743743743743743e-05,
+      "loss": 1.1089,
+      "step": 6258
+    },
+    {
+      "epoch": 0.008181667261435576,
+      "grad_norm": 1.749738097190857,
+      "learning_rate": 1.873873873873874e-05,
+      "loss": 1.8046,
+      "step": 6259
+    },
+    {
+      "epoch": 0.008182974445851846,
+      "grad_norm": 1.2619025707244873,
+      "learning_rate": 1.8733733733733736e-05,
+      "loss": 1.6745,
+      "step": 6260
+    },
+    {
+      "epoch": 0.008184281630268117,
+      "grad_norm": 1.4279109239578247,
+      "learning_rate": 1.872872872872873e-05,
+      "loss": 2.3108,
+      "step": 6261
+    },
+    {
+      "epoch": 0.008185588814684387,
+      "grad_norm": 1.9877744913101196,
+      "learning_rate": 1.8723723723723725e-05,
+      "loss": 1.1352,
+      "step": 6262
+    },
+    {
+      "epoch": 0.008186895999100658,
+      "grad_norm": 1.3834372758865356,
+      "learning_rate": 1.8718718718718718e-05,
+      "loss": 2.3226,
+      "step": 6263
+    },
+    {
+      "epoch": 0.008188203183516927,
+      "grad_norm": 1.219789981842041,
+      "learning_rate": 1.8713713713713714e-05,
+      "loss": 1.7451,
+      "step": 6264
+    },
+    {
+      "epoch": 0.008189510367933198,
+      "grad_norm": 1.214350700378418,
+      "learning_rate": 1.870870870870871e-05,
+      "loss": 1.3878,
+      "step": 6265
+    },
+    {
+      "epoch": 0.008190817552349468,
+      "grad_norm": 1.481269359588623,
+      "learning_rate": 1.8703703703703704e-05,
+      "loss": 1.664,
+      "step": 6266
+    },
+    {
+      "epoch": 0.008192124736765739,
+      "grad_norm": 1.3241275548934937,
+      "learning_rate": 1.86986986986987e-05,
+      "loss": 1.4371,
+      "step": 6267
+    },
+    {
+      "epoch": 0.008193431921182008,
+      "grad_norm": 1.1102116107940674,
+      "learning_rate": 1.8693693693693696e-05,
+      "loss": 0.694,
+      "step": 6268
+    },
+    {
+      "epoch": 0.008194739105598278,
+      "grad_norm": 1.2464959621429443,
+      "learning_rate": 1.868868868868869e-05,
+      "loss": 1.4076,
+      "step": 6269
+    },
+    {
+      "epoch": 0.008196046290014549,
+      "grad_norm": 0.9579530358314514,
+      "learning_rate": 1.8683683683683682e-05,
+      "loss": 0.6931,
+      "step": 6270
+    },
+    {
+      "epoch": 0.008197353474430819,
+      "grad_norm": 1.8528544902801514,
+      "learning_rate": 1.867867867867868e-05,
+      "loss": 2.5014,
+      "step": 6271
+    },
+    {
+      "epoch": 0.00819866065884709,
+      "grad_norm": 1.4223881959915161,
+      "learning_rate": 1.8673673673673675e-05,
+      "loss": 1.1435,
+      "step": 6272
+    },
+    {
+      "epoch": 0.00819996784326336,
+      "grad_norm": 1.1673530340194702,
+      "learning_rate": 1.866866866866867e-05,
+      "loss": 1.3373,
+      "step": 6273
+    },
+    {
+      "epoch": 0.00820127502767963,
+      "grad_norm": 1.1808266639709473,
+      "learning_rate": 1.8663663663663664e-05,
+      "loss": 1.2678,
+      "step": 6274
+    },
+    {
+      "epoch": 0.0082025822120959,
+      "grad_norm": 1.4232062101364136,
+      "learning_rate": 1.865865865865866e-05,
+      "loss": 1.6266,
+      "step": 6275
+    },
+    {
+      "epoch": 0.008203889396512171,
+      "grad_norm": 1.3301951885223389,
+      "learning_rate": 1.8653653653653654e-05,
+      "loss": 1.6023,
+      "step": 6276
+    },
+    {
+      "epoch": 0.00820519658092844,
+      "grad_norm": 1.3295010328292847,
+      "learning_rate": 1.864864864864865e-05,
+      "loss": 1.4604,
+      "step": 6277
+    },
+    {
+      "epoch": 0.008206503765344712,
+      "grad_norm": 1.3246461153030396,
+      "learning_rate": 1.8643643643643643e-05,
+      "loss": 2.0727,
+      "step": 6278
+    },
+    {
+      "epoch": 0.008207810949760981,
+      "grad_norm": 1.0662879943847656,
+      "learning_rate": 1.863863863863864e-05,
+      "loss": 0.8844,
+      "step": 6279
+    },
+    {
+      "epoch": 0.008209118134177252,
+      "grad_norm": 1.3889623880386353,
+      "learning_rate": 1.8633633633633636e-05,
+      "loss": 1.8338,
+      "step": 6280
+    },
+    {
+      "epoch": 0.008210425318593522,
+      "grad_norm": 1.462066650390625,
+      "learning_rate": 1.862862862862863e-05,
+      "loss": 1.7239,
+      "step": 6281
+    },
+    {
+      "epoch": 0.008211732503009793,
+      "grad_norm": 1.605776071548462,
+      "learning_rate": 1.8623623623623622e-05,
+      "loss": 1.793,
+      "step": 6282
+    },
+    {
+      "epoch": 0.008213039687426062,
+      "grad_norm": 1.1192706823349,
+      "learning_rate": 1.8618618618618618e-05,
+      "loss": 1.4533,
+      "step": 6283
+    },
+    {
+      "epoch": 0.008214346871842332,
+      "grad_norm": 1.1550371646881104,
+      "learning_rate": 1.8613613613613614e-05,
+      "loss": 0.6085,
+      "step": 6284
+    },
+    {
+      "epoch": 0.008215654056258603,
+      "grad_norm": 1.189383864402771,
+      "learning_rate": 1.860860860860861e-05,
+      "loss": 1.1203,
+      "step": 6285
+    },
+    {
+      "epoch": 0.008216961240674872,
+      "grad_norm": 1.2319304943084717,
+      "learning_rate": 1.8603603603603604e-05,
+      "loss": 1.1626,
+      "step": 6286
+    },
+    {
+      "epoch": 0.008218268425091144,
+      "grad_norm": 1.1458969116210938,
+      "learning_rate": 1.85985985985986e-05,
+      "loss": 1.0781,
+      "step": 6287
+    },
+    {
+      "epoch": 0.008219575609507413,
+      "grad_norm": 1.3787761926651,
+      "learning_rate": 1.8593593593593593e-05,
+      "loss": 0.893,
+      "step": 6288
+    },
+    {
+      "epoch": 0.008220882793923684,
+      "grad_norm": 1.5353268384933472,
+      "learning_rate": 1.858858858858859e-05,
+      "loss": 1.6095,
+      "step": 6289
+    },
+    {
+      "epoch": 0.008222189978339954,
+      "grad_norm": 1.4656667709350586,
+      "learning_rate": 1.8583583583583583e-05,
+      "loss": 1.8741,
+      "step": 6290
+    },
+    {
+      "epoch": 0.008223497162756225,
+      "grad_norm": 1.7559832334518433,
+      "learning_rate": 1.857857857857858e-05,
+      "loss": 2.2803,
+      "step": 6291
+    },
+    {
+      "epoch": 0.008224804347172494,
+      "grad_norm": 1.6155341863632202,
+      "learning_rate": 1.8573573573573575e-05,
+      "loss": 1.1631,
+      "step": 6292
+    },
+    {
+      "epoch": 0.008226111531588766,
+      "grad_norm": 1.8866392374038696,
+      "learning_rate": 1.856856856856857e-05,
+      "loss": 2.3979,
+      "step": 6293
+    },
+    {
+      "epoch": 0.008227418716005035,
+      "grad_norm": 1.8857920169830322,
+      "learning_rate": 1.8563563563563565e-05,
+      "loss": 1.5498,
+      "step": 6294
+    },
+    {
+      "epoch": 0.008228725900421306,
+      "grad_norm": 1.008163571357727,
+      "learning_rate": 1.8558558558558558e-05,
+      "loss": 1.055,
+      "step": 6295
+    },
+    {
+      "epoch": 0.008230033084837576,
+      "grad_norm": 1.0751030445098877,
+      "learning_rate": 1.8553553553553554e-05,
+      "loss": 0.7987,
+      "step": 6296
+    },
+    {
+      "epoch": 0.008231340269253847,
+      "grad_norm": 1.3049677610397339,
+      "learning_rate": 1.854854854854855e-05,
+      "loss": 1.7454,
+      "step": 6297
+    },
+    {
+      "epoch": 0.008232647453670116,
+      "grad_norm": 1.3433533906936646,
+      "learning_rate": 1.8543543543543547e-05,
+      "loss": 1.3966,
+      "step": 6298
+    },
+    {
+      "epoch": 0.008233954638086386,
+      "grad_norm": 0.7153356671333313,
+      "learning_rate": 1.853853853853854e-05,
+      "loss": 0.4506,
+      "step": 6299
+    },
+    {
+      "epoch": 0.008235261822502657,
+      "grad_norm": 1.6498936414718628,
+      "learning_rate": 1.8533533533533533e-05,
+      "loss": 2.0854,
+      "step": 6300
+    },
+    {
+      "epoch": 0.008236569006918926,
+      "grad_norm": 1.3126577138900757,
+      "learning_rate": 1.852852852852853e-05,
+      "loss": 1.3798,
+      "step": 6301
+    },
+    {
+      "epoch": 0.008237876191335198,
+      "grad_norm": 1.4931023120880127,
+      "learning_rate": 1.8523523523523525e-05,
+      "loss": 1.1853,
+      "step": 6302
+    },
+    {
+      "epoch": 0.008239183375751467,
+      "grad_norm": 1.582301378250122,
+      "learning_rate": 1.8518518518518518e-05,
+      "loss": 1.6279,
+      "step": 6303
+    },
+    {
+      "epoch": 0.008240490560167738,
+      "grad_norm": 1.801631212234497,
+      "learning_rate": 1.8513513513513515e-05,
+      "loss": 1.4826,
+      "step": 6304
+    },
+    {
+      "epoch": 0.008241797744584008,
+      "grad_norm": 1.3345496654510498,
+      "learning_rate": 1.850850850850851e-05,
+      "loss": 1.3847,
+      "step": 6305
+    },
+    {
+      "epoch": 0.008243104929000279,
+      "grad_norm": 1.5879416465759277,
+      "learning_rate": 1.8503503503503504e-05,
+      "loss": 0.6189,
+      "step": 6306
+    },
+    {
+      "epoch": 0.008244412113416548,
+      "grad_norm": 1.6946147680282593,
+      "learning_rate": 1.8498498498498497e-05,
+      "loss": 1.7637,
+      "step": 6307
+    },
+    {
+      "epoch": 0.00824571929783282,
+      "grad_norm": 1.213861346244812,
+      "learning_rate": 1.8493493493493493e-05,
+      "loss": 1.1015,
+      "step": 6308
+    },
+    {
+      "epoch": 0.008247026482249089,
+      "grad_norm": 1.6994218826293945,
+      "learning_rate": 1.848848848848849e-05,
+      "loss": 2.3188,
+      "step": 6309
+    },
+    {
+      "epoch": 0.00824833366666536,
+      "grad_norm": 1.149584174156189,
+      "learning_rate": 1.8483483483483486e-05,
+      "loss": 1.1,
+      "step": 6310
+    },
+    {
+      "epoch": 0.00824964085108163,
+      "grad_norm": 1.8066807985305786,
+      "learning_rate": 1.847847847847848e-05,
+      "loss": 1.5961,
+      "step": 6311
+    },
+    {
+      "epoch": 0.0082509480354979,
+      "grad_norm": 1.4775680303573608,
+      "learning_rate": 1.8473473473473475e-05,
+      "loss": 1.6228,
+      "step": 6312
+    },
+    {
+      "epoch": 0.00825225521991417,
+      "grad_norm": 1.5239574909210205,
+      "learning_rate": 1.846846846846847e-05,
+      "loss": 2.2769,
+      "step": 6313
+    },
+    {
+      "epoch": 0.00825356240433044,
+      "grad_norm": 1.618314504623413,
+      "learning_rate": 1.8463463463463465e-05,
+      "loss": 1.2877,
+      "step": 6314
+    },
+    {
+      "epoch": 0.00825486958874671,
+      "grad_norm": 1.3970030546188354,
+      "learning_rate": 1.8458458458458458e-05,
+      "loss": 1.4042,
+      "step": 6315
+    },
+    {
+      "epoch": 0.00825617677316298,
+      "grad_norm": 1.426716923713684,
+      "learning_rate": 1.8453453453453454e-05,
+      "loss": 1.9637,
+      "step": 6316
+    },
+    {
+      "epoch": 0.008257483957579251,
+      "grad_norm": 1.1363551616668701,
+      "learning_rate": 1.844844844844845e-05,
+      "loss": 1.0654,
+      "step": 6317
+    },
+    {
+      "epoch": 0.008258791141995521,
+      "grad_norm": 1.5384882688522339,
+      "learning_rate": 1.8443443443443443e-05,
+      "loss": 1.7851,
+      "step": 6318
+    },
+    {
+      "epoch": 0.008260098326411792,
+      "grad_norm": 1.9253578186035156,
+      "learning_rate": 1.843843843843844e-05,
+      "loss": 1.7167,
+      "step": 6319
+    },
+    {
+      "epoch": 0.008261405510828062,
+      "grad_norm": 1.6930334568023682,
+      "learning_rate": 1.8433433433433433e-05,
+      "loss": 1.6603,
+      "step": 6320
+    },
+    {
+      "epoch": 0.008262712695244333,
+      "grad_norm": 1.0556018352508545,
+      "learning_rate": 1.842842842842843e-05,
+      "loss": 0.6634,
+      "step": 6321
+    },
+    {
+      "epoch": 0.008264019879660602,
+      "grad_norm": 1.6448642015457153,
+      "learning_rate": 1.8423423423423425e-05,
+      "loss": 2.5531,
+      "step": 6322
+    },
+    {
+      "epoch": 0.008265327064076873,
+      "grad_norm": 1.3424268960952759,
+      "learning_rate": 1.8418418418418422e-05,
+      "loss": 1.6367,
+      "step": 6323
+    },
+    {
+      "epoch": 0.008266634248493143,
+      "grad_norm": 1.3995227813720703,
+      "learning_rate": 1.8413413413413415e-05,
+      "loss": 1.3657,
+      "step": 6324
+    },
+    {
+      "epoch": 0.008267941432909414,
+      "grad_norm": 1.3035091161727905,
+      "learning_rate": 1.8408408408408408e-05,
+      "loss": 1.2252,
+      "step": 6325
+    },
+    {
+      "epoch": 0.008269248617325684,
+      "grad_norm": 1.762397050857544,
+      "learning_rate": 1.8403403403403404e-05,
+      "loss": 1.2747,
+      "step": 6326
+    },
+    {
+      "epoch": 0.008270555801741955,
+      "grad_norm": 1.4347635507583618,
+      "learning_rate": 1.83983983983984e-05,
+      "loss": 1.8403,
+      "step": 6327
+    },
+    {
+      "epoch": 0.008271862986158224,
+      "grad_norm": 1.7272084951400757,
+      "learning_rate": 1.8393393393393393e-05,
+      "loss": 1.8139,
+      "step": 6328
+    },
+    {
+      "epoch": 0.008273170170574495,
+      "grad_norm": 1.4466111660003662,
+      "learning_rate": 1.838838838838839e-05,
+      "loss": 1.4819,
+      "step": 6329
+    },
+    {
+      "epoch": 0.008274477354990765,
+      "grad_norm": 1.3984220027923584,
+      "learning_rate": 1.8383383383383386e-05,
+      "loss": 1.4717,
+      "step": 6330
+    },
+    {
+      "epoch": 0.008275784539407034,
+      "grad_norm": 1.4767919778823853,
+      "learning_rate": 1.837837837837838e-05,
+      "loss": 1.6155,
+      "step": 6331
+    },
+    {
+      "epoch": 0.008277091723823305,
+      "grad_norm": 1.6241779327392578,
+      "learning_rate": 1.8373373373373372e-05,
+      "loss": 1.7396,
+      "step": 6332
+    },
+    {
+      "epoch": 0.008278398908239575,
+      "grad_norm": 1.9534870386123657,
+      "learning_rate": 1.836836836836837e-05,
+      "loss": 1.8818,
+      "step": 6333
+    },
+    {
+      "epoch": 0.008279706092655846,
+      "grad_norm": 1.5548218488693237,
+      "learning_rate": 1.8363363363363365e-05,
+      "loss": 1.3125,
+      "step": 6334
+    },
+    {
+      "epoch": 0.008281013277072116,
+      "grad_norm": 1.5471489429473877,
+      "learning_rate": 1.835835835835836e-05,
+      "loss": 2.1028,
+      "step": 6335
+    },
+    {
+      "epoch": 0.008282320461488387,
+      "grad_norm": 1.229709506034851,
+      "learning_rate": 1.8353353353353354e-05,
+      "loss": 1.488,
+      "step": 6336
+    },
+    {
+      "epoch": 0.008283627645904656,
+      "grad_norm": 1.4629777669906616,
+      "learning_rate": 1.8348348348348347e-05,
+      "loss": 1.3484,
+      "step": 6337
+    },
+    {
+      "epoch": 0.008284934830320927,
+      "grad_norm": 1.2428723573684692,
+      "learning_rate": 1.8343343343343344e-05,
+      "loss": 0.9425,
+      "step": 6338
+    },
+    {
+      "epoch": 0.008286242014737197,
+      "grad_norm": 1.267740249633789,
+      "learning_rate": 1.833833833833834e-05,
+      "loss": 1.0423,
+      "step": 6339
+    },
+    {
+      "epoch": 0.008287549199153468,
+      "grad_norm": 1.304635763168335,
+      "learning_rate": 1.8333333333333333e-05,
+      "loss": 1.7492,
+      "step": 6340
+    },
+    {
+      "epoch": 0.008288856383569737,
+      "grad_norm": 1.6973445415496826,
+      "learning_rate": 1.832832832832833e-05,
+      "loss": 1.8007,
+      "step": 6341
+    },
+    {
+      "epoch": 0.008290163567986009,
+      "grad_norm": 1.1339062452316284,
+      "learning_rate": 1.8323323323323326e-05,
+      "loss": 0.7629,
+      "step": 6342
+    },
+    {
+      "epoch": 0.008291470752402278,
+      "grad_norm": 1.5441731214523315,
+      "learning_rate": 1.831831831831832e-05,
+      "loss": 1.2301,
+      "step": 6343
+    },
+    {
+      "epoch": 0.00829277793681855,
+      "grad_norm": 1.712719202041626,
+      "learning_rate": 1.8313313313313315e-05,
+      "loss": 1.9256,
+      "step": 6344
+    },
+    {
+      "epoch": 0.008294085121234819,
+      "grad_norm": 1.4463411569595337,
+      "learning_rate": 1.8308308308308308e-05,
+      "loss": 1.6648,
+      "step": 6345
+    },
+    {
+      "epoch": 0.008295392305651088,
+      "grad_norm": 1.4114418029785156,
+      "learning_rate": 1.8303303303303304e-05,
+      "loss": 1.2888,
+      "step": 6346
+    },
+    {
+      "epoch": 0.00829669949006736,
+      "grad_norm": 1.9361258745193481,
+      "learning_rate": 1.82982982982983e-05,
+      "loss": 3.0828,
+      "step": 6347
+    },
+    {
+      "epoch": 0.008298006674483629,
+      "grad_norm": 1.2454068660736084,
+      "learning_rate": 1.8293293293293297e-05,
+      "loss": 1.1768,
+      "step": 6348
+    },
+    {
+      "epoch": 0.0082993138588999,
+      "grad_norm": 1.3398100137710571,
+      "learning_rate": 1.828828828828829e-05,
+      "loss": 1.3226,
+      "step": 6349
+    },
+    {
+      "epoch": 0.00830062104331617,
+      "grad_norm": 1.3560271263122559,
+      "learning_rate": 1.8283283283283283e-05,
+      "loss": 1.45,
+      "step": 6350
+    },
+    {
+      "epoch": 0.00830192822773244,
+      "grad_norm": 2.653871774673462,
+      "learning_rate": 1.827827827827828e-05,
+      "loss": 2.2168,
+      "step": 6351
+    },
+    {
+      "epoch": 0.00830323541214871,
+      "grad_norm": 1.1013672351837158,
+      "learning_rate": 1.8273273273273276e-05,
+      "loss": 0.9591,
+      "step": 6352
+    },
+    {
+      "epoch": 0.008304542596564981,
+      "grad_norm": 2.082571268081665,
+      "learning_rate": 1.826826826826827e-05,
+      "loss": 2.3062,
+      "step": 6353
+    },
+    {
+      "epoch": 0.00830584978098125,
+      "grad_norm": 1.34363853931427,
+      "learning_rate": 1.8263263263263265e-05,
+      "loss": 1.7463,
+      "step": 6354
+    },
+    {
+      "epoch": 0.008307156965397522,
+      "grad_norm": 1.3702508211135864,
+      "learning_rate": 1.8258258258258258e-05,
+      "loss": 1.7331,
+      "step": 6355
+    },
+    {
+      "epoch": 0.008308464149813791,
+      "grad_norm": 1.5913704633712769,
+      "learning_rate": 1.8253253253253254e-05,
+      "loss": 1.5927,
+      "step": 6356
+    },
+    {
+      "epoch": 0.008309771334230063,
+      "grad_norm": 1.3661609888076782,
+      "learning_rate": 1.8248248248248247e-05,
+      "loss": 1.6021,
+      "step": 6357
+    },
+    {
+      "epoch": 0.008311078518646332,
+      "grad_norm": 1.7113715410232544,
+      "learning_rate": 1.8243243243243244e-05,
+      "loss": 2.3604,
+      "step": 6358
+    },
+    {
+      "epoch": 0.008312385703062603,
+      "grad_norm": 1.3951841592788696,
+      "learning_rate": 1.823823823823824e-05,
+      "loss": 2.0689,
+      "step": 6359
+    },
+    {
+      "epoch": 0.008313692887478873,
+      "grad_norm": 1.7038748264312744,
+      "learning_rate": 1.8233233233233236e-05,
+      "loss": 2.2428,
+      "step": 6360
+    },
+    {
+      "epoch": 0.008315000071895142,
+      "grad_norm": 1.5391063690185547,
+      "learning_rate": 1.822822822822823e-05,
+      "loss": 1.2031,
+      "step": 6361
+    },
+    {
+      "epoch": 0.008316307256311413,
+      "grad_norm": 1.475633978843689,
+      "learning_rate": 1.8223223223223222e-05,
+      "loss": 2.0882,
+      "step": 6362
+    },
+    {
+      "epoch": 0.008317614440727683,
+      "grad_norm": 1.4056137800216675,
+      "learning_rate": 1.821821821821822e-05,
+      "loss": 1.5992,
+      "step": 6363
+    },
+    {
+      "epoch": 0.008318921625143954,
+      "grad_norm": 1.2477256059646606,
+      "learning_rate": 1.8213213213213215e-05,
+      "loss": 1.2993,
+      "step": 6364
+    },
+    {
+      "epoch": 0.008320228809560223,
+      "grad_norm": 1.2482731342315674,
+      "learning_rate": 1.8208208208208208e-05,
+      "loss": 1.3723,
+      "step": 6365
+    },
+    {
+      "epoch": 0.008321535993976495,
+      "grad_norm": 1.4055628776550293,
+      "learning_rate": 1.8203203203203204e-05,
+      "loss": 1.5188,
+      "step": 6366
+    },
+    {
+      "epoch": 0.008322843178392764,
+      "grad_norm": 0.8476602435112,
+      "learning_rate": 1.81981981981982e-05,
+      "loss": 0.6717,
+      "step": 6367
+    },
+    {
+      "epoch": 0.008324150362809035,
+      "grad_norm": 1.6473441123962402,
+      "learning_rate": 1.8193193193193194e-05,
+      "loss": 1.6159,
+      "step": 6368
+    },
+    {
+      "epoch": 0.008325457547225305,
+      "grad_norm": 1.4665833711624146,
+      "learning_rate": 1.8188188188188187e-05,
+      "loss": 1.8091,
+      "step": 6369
+    },
+    {
+      "epoch": 0.008326764731641576,
+      "grad_norm": 1.1065492630004883,
+      "learning_rate": 1.8183183183183183e-05,
+      "loss": 1.0595,
+      "step": 6370
+    },
+    {
+      "epoch": 0.008328071916057845,
+      "grad_norm": 1.9728772640228271,
+      "learning_rate": 1.817817817817818e-05,
+      "loss": 1.7912,
+      "step": 6371
+    },
+    {
+      "epoch": 0.008329379100474116,
+      "grad_norm": 1.5492225885391235,
+      "learning_rate": 1.8173173173173176e-05,
+      "loss": 2.159,
+      "step": 6372
+    },
+    {
+      "epoch": 0.008330686284890386,
+      "grad_norm": 1.8420453071594238,
+      "learning_rate": 1.816816816816817e-05,
+      "loss": 1.2859,
+      "step": 6373
+    },
+    {
+      "epoch": 0.008331993469306657,
+      "grad_norm": 1.2815078496932983,
+      "learning_rate": 1.8163163163163162e-05,
+      "loss": 1.0286,
+      "step": 6374
+    },
+    {
+      "epoch": 0.008333300653722927,
+      "grad_norm": 1.8655595779418945,
+      "learning_rate": 1.8158158158158158e-05,
+      "loss": 1.5758,
+      "step": 6375
+    },
+    {
+      "epoch": 0.008334607838139196,
+      "grad_norm": 1.1927909851074219,
+      "learning_rate": 1.8153153153153154e-05,
+      "loss": 1.2772,
+      "step": 6376
+    },
+    {
+      "epoch": 0.008335915022555467,
+      "grad_norm": 1.5637260675430298,
+      "learning_rate": 1.814814814814815e-05,
+      "loss": 1.2175,
+      "step": 6377
+    },
+    {
+      "epoch": 0.008337222206971737,
+      "grad_norm": 1.381056547164917,
+      "learning_rate": 1.8143143143143144e-05,
+      "loss": 0.5844,
+      "step": 6378
+    },
+    {
+      "epoch": 0.008338529391388008,
+      "grad_norm": 1.125281810760498,
+      "learning_rate": 1.813813813813814e-05,
+      "loss": 1.1795,
+      "step": 6379
+    },
+    {
+      "epoch": 0.008339836575804277,
+      "grad_norm": 1.3138717412948608,
+      "learning_rate": 1.8133133133133133e-05,
+      "loss": 1.1225,
+      "step": 6380
+    },
+    {
+      "epoch": 0.008341143760220548,
+      "grad_norm": 0.9434112906455994,
+      "learning_rate": 1.812812812812813e-05,
+      "loss": 0.6241,
+      "step": 6381
+    },
+    {
+      "epoch": 0.008342450944636818,
+      "grad_norm": 1.3067516088485718,
+      "learning_rate": 1.8123123123123122e-05,
+      "loss": 1.9066,
+      "step": 6382
+    },
+    {
+      "epoch": 0.008343758129053089,
+      "grad_norm": 1.5010817050933838,
+      "learning_rate": 1.811811811811812e-05,
+      "loss": 1.7757,
+      "step": 6383
+    },
+    {
+      "epoch": 0.008345065313469359,
+      "grad_norm": 1.376915693283081,
+      "learning_rate": 1.8113113113113115e-05,
+      "loss": 1.2789,
+      "step": 6384
+    },
+    {
+      "epoch": 0.00834637249788563,
+      "grad_norm": 1.3108426332473755,
+      "learning_rate": 1.810810810810811e-05,
+      "loss": 1.884,
+      "step": 6385
+    },
+    {
+      "epoch": 0.0083476796823019,
+      "grad_norm": 1.8863136768341064,
+      "learning_rate": 1.8103103103103105e-05,
+      "loss": 2.1042,
+      "step": 6386
+    },
+    {
+      "epoch": 0.00834898686671817,
+      "grad_norm": 1.793798565864563,
+      "learning_rate": 1.8098098098098098e-05,
+      "loss": 1.8287,
+      "step": 6387
+    },
+    {
+      "epoch": 0.00835029405113444,
+      "grad_norm": 1.4103461503982544,
+      "learning_rate": 1.8093093093093094e-05,
+      "loss": 0.9995,
+      "step": 6388
+    },
+    {
+      "epoch": 0.008351601235550711,
+      "grad_norm": 1.4399807453155518,
+      "learning_rate": 1.808808808808809e-05,
+      "loss": 1.7221,
+      "step": 6389
+    },
+    {
+      "epoch": 0.00835290841996698,
+      "grad_norm": 1.2273454666137695,
+      "learning_rate": 1.8083083083083083e-05,
+      "loss": 0.9865,
+      "step": 6390
+    },
+    {
+      "epoch": 0.00835421560438325,
+      "grad_norm": 1.85224449634552,
+      "learning_rate": 1.807807807807808e-05,
+      "loss": 1.5183,
+      "step": 6391
+    },
+    {
+      "epoch": 0.008355522788799521,
+      "grad_norm": 1.0086297988891602,
+      "learning_rate": 1.8073073073073073e-05,
+      "loss": 0.6789,
+      "step": 6392
+    },
+    {
+      "epoch": 0.00835682997321579,
+      "grad_norm": 1.381419062614441,
+      "learning_rate": 1.806806806806807e-05,
+      "loss": 2.3223,
+      "step": 6393
+    },
+    {
+      "epoch": 0.008358137157632062,
+      "grad_norm": 0.3620203733444214,
+      "learning_rate": 1.8063063063063062e-05,
+      "loss": 0.0998,
+      "step": 6394
+    },
+    {
+      "epoch": 0.008359444342048331,
+      "grad_norm": 1.8040612936019897,
+      "learning_rate": 1.8058058058058058e-05,
+      "loss": 1.935,
+      "step": 6395
+    },
+    {
+      "epoch": 0.008360751526464602,
+      "grad_norm": 1.4697967767715454,
+      "learning_rate": 1.8053053053053055e-05,
+      "loss": 1.2112,
+      "step": 6396
+    },
+    {
+      "epoch": 0.008362058710880872,
+      "grad_norm": 1.5619410276412964,
+      "learning_rate": 1.804804804804805e-05,
+      "loss": 1.6079,
+      "step": 6397
+    },
+    {
+      "epoch": 0.008363365895297143,
+      "grad_norm": 2.029658794403076,
+      "learning_rate": 1.8043043043043044e-05,
+      "loss": 2.7304,
+      "step": 6398
+    },
+    {
+      "epoch": 0.008364673079713413,
+      "grad_norm": 1.7922471761703491,
+      "learning_rate": 1.8038038038038037e-05,
+      "loss": 1.6871,
+      "step": 6399
+    },
+    {
+      "epoch": 0.008365980264129684,
+      "grad_norm": 1.7215683460235596,
+      "learning_rate": 1.8033033033033033e-05,
+      "loss": 1.83,
+      "step": 6400
+    },
+    {
+      "epoch": 0.008367287448545953,
+      "grad_norm": 1.6571084260940552,
+      "learning_rate": 1.802802802802803e-05,
+      "loss": 1.0172,
+      "step": 6401
+    },
+    {
+      "epoch": 0.008368594632962224,
+      "grad_norm": 1.2091617584228516,
+      "learning_rate": 1.8023023023023026e-05,
+      "loss": 1.5676,
+      "step": 6402
+    },
+    {
+      "epoch": 0.008369901817378494,
+      "grad_norm": 1.3481963872909546,
+      "learning_rate": 1.801801801801802e-05,
+      "loss": 1.1443,
+      "step": 6403
+    },
+    {
+      "epoch": 0.008371209001794765,
+      "grad_norm": 1.5685815811157227,
+      "learning_rate": 1.8013013013013015e-05,
+      "loss": 1.5887,
+      "step": 6404
+    },
+    {
+      "epoch": 0.008372516186211034,
+      "grad_norm": 1.5026166439056396,
+      "learning_rate": 1.800800800800801e-05,
+      "loss": 1.9294,
+      "step": 6405
+    },
+    {
+      "epoch": 0.008373823370627304,
+      "grad_norm": 1.9851096868515015,
+      "learning_rate": 1.8003003003003005e-05,
+      "loss": 2.2378,
+      "step": 6406
+    },
+    {
+      "epoch": 0.008375130555043575,
+      "grad_norm": 1.644208550453186,
+      "learning_rate": 1.7997997997997998e-05,
+      "loss": 1.9241,
+      "step": 6407
+    },
+    {
+      "epoch": 0.008376437739459845,
+      "grad_norm": 1.3623074293136597,
+      "learning_rate": 1.7992992992992994e-05,
+      "loss": 1.0759,
+      "step": 6408
+    },
+    {
+      "epoch": 0.008377744923876116,
+      "grad_norm": 1.545291543006897,
+      "learning_rate": 1.798798798798799e-05,
+      "loss": 1.9086,
+      "step": 6409
+    },
+    {
+      "epoch": 0.008379052108292385,
+      "grad_norm": 1.4542279243469238,
+      "learning_rate": 1.7982982982982983e-05,
+      "loss": 1.5444,
+      "step": 6410
+    },
+    {
+      "epoch": 0.008380359292708656,
+      "grad_norm": 1.663552165031433,
+      "learning_rate": 1.7977977977977976e-05,
+      "loss": 1.9807,
+      "step": 6411
+    },
+    {
+      "epoch": 0.008381666477124926,
+      "grad_norm": 1.6835460662841797,
+      "learning_rate": 1.7972972972972973e-05,
+      "loss": 1.7484,
+      "step": 6412
+    },
+    {
+      "epoch": 0.008382973661541197,
+      "grad_norm": 1.6011698246002197,
+      "learning_rate": 1.796796796796797e-05,
+      "loss": 2.2448,
+      "step": 6413
+    },
+    {
+      "epoch": 0.008384280845957466,
+      "grad_norm": 1.3422085046768188,
+      "learning_rate": 1.7962962962962965e-05,
+      "loss": 1.3597,
+      "step": 6414
+    },
+    {
+      "epoch": 0.008385588030373738,
+      "grad_norm": 1.274091362953186,
+      "learning_rate": 1.795795795795796e-05,
+      "loss": 1.2738,
+      "step": 6415
+    },
+    {
+      "epoch": 0.008386895214790007,
+      "grad_norm": 0.9373536109924316,
+      "learning_rate": 1.7952952952952955e-05,
+      "loss": 0.9455,
+      "step": 6416
+    },
+    {
+      "epoch": 0.008388202399206278,
+      "grad_norm": 1.372843623161316,
+      "learning_rate": 1.7947947947947948e-05,
+      "loss": 1.0614,
+      "step": 6417
+    },
+    {
+      "epoch": 0.008389509583622548,
+      "grad_norm": 1.5960743427276611,
+      "learning_rate": 1.7942942942942944e-05,
+      "loss": 1.3783,
+      "step": 6418
+    },
+    {
+      "epoch": 0.008390816768038819,
+      "grad_norm": 1.1954265832901,
+      "learning_rate": 1.7937937937937937e-05,
+      "loss": 1.4332,
+      "step": 6419
+    },
+    {
+      "epoch": 0.008392123952455088,
+      "grad_norm": 1.3324917554855347,
+      "learning_rate": 1.7932932932932933e-05,
+      "loss": 1.353,
+      "step": 6420
+    },
+    {
+      "epoch": 0.00839343113687136,
+      "grad_norm": 0.9907099008560181,
+      "learning_rate": 1.792792792792793e-05,
+      "loss": 0.5637,
+      "step": 6421
+    },
+    {
+      "epoch": 0.008394738321287629,
+      "grad_norm": 1.499237298965454,
+      "learning_rate": 1.7922922922922926e-05,
+      "loss": 1.7512,
+      "step": 6422
+    },
+    {
+      "epoch": 0.008396045505703898,
+      "grad_norm": 1.5187965631484985,
+      "learning_rate": 1.7917917917917916e-05,
+      "loss": 1.7253,
+      "step": 6423
+    },
+    {
+      "epoch": 0.00839735269012017,
+      "grad_norm": 1.1108216047286987,
+      "learning_rate": 1.7912912912912912e-05,
+      "loss": 1.127,
+      "step": 6424
+    },
+    {
+      "epoch": 0.008398659874536439,
+      "grad_norm": 1.5034611225128174,
+      "learning_rate": 1.790790790790791e-05,
+      "loss": 2.1544,
+      "step": 6425
+    },
+    {
+      "epoch": 0.00839996705895271,
+      "grad_norm": 1.3642117977142334,
+      "learning_rate": 1.7902902902902905e-05,
+      "loss": 1.4368,
+      "step": 6426
+    },
+    {
+      "epoch": 0.00840127424336898,
+      "grad_norm": 1.0408982038497925,
+      "learning_rate": 1.78978978978979e-05,
+      "loss": 0.772,
+      "step": 6427
+    },
+    {
+      "epoch": 0.008402581427785251,
+      "grad_norm": 1.763648509979248,
+      "learning_rate": 1.7892892892892894e-05,
+      "loss": 1.4301,
+      "step": 6428
+    },
+    {
+      "epoch": 0.00840388861220152,
+      "grad_norm": 1.808457851409912,
+      "learning_rate": 1.7887887887887887e-05,
+      "loss": 1.2217,
+      "step": 6429
+    },
+    {
+      "epoch": 0.008405195796617792,
+      "grad_norm": 1.5414410829544067,
+      "learning_rate": 1.7882882882882883e-05,
+      "loss": 0.9495,
+      "step": 6430
+    },
+    {
+      "epoch": 0.008406502981034061,
+      "grad_norm": 1.770066261291504,
+      "learning_rate": 1.787787787787788e-05,
+      "loss": 1.4655,
+      "step": 6431
+    },
+    {
+      "epoch": 0.008407810165450332,
+      "grad_norm": 1.6303621530532837,
+      "learning_rate": 1.7872872872872873e-05,
+      "loss": 0.8314,
+      "step": 6432
+    },
+    {
+      "epoch": 0.008409117349866602,
+      "grad_norm": 1.6146042346954346,
+      "learning_rate": 1.786786786786787e-05,
+      "loss": 1.5959,
+      "step": 6433
+    },
+    {
+      "epoch": 0.008410424534282873,
+      "grad_norm": 1.2927511930465698,
+      "learning_rate": 1.7862862862862866e-05,
+      "loss": 1.3714,
+      "step": 6434
+    },
+    {
+      "epoch": 0.008411731718699142,
+      "grad_norm": 0.6642452478408813,
+      "learning_rate": 1.785785785785786e-05,
+      "loss": 0.4274,
+      "step": 6435
+    },
+    {
+      "epoch": 0.008413038903115413,
+      "grad_norm": 1.7351791858673096,
+      "learning_rate": 1.785285285285285e-05,
+      "loss": 1.3165,
+      "step": 6436
+    },
+    {
+      "epoch": 0.008414346087531683,
+      "grad_norm": 1.4387867450714111,
+      "learning_rate": 1.7847847847847848e-05,
+      "loss": 1.6183,
+      "step": 6437
+    },
+    {
+      "epoch": 0.008415653271947952,
+      "grad_norm": 1.9416606426239014,
+      "learning_rate": 1.7842842842842844e-05,
+      "loss": 1.6773,
+      "step": 6438
+    },
+    {
+      "epoch": 0.008416960456364224,
+      "grad_norm": 1.7922090291976929,
+      "learning_rate": 1.783783783783784e-05,
+      "loss": 1.7976,
+      "step": 6439
+    },
+    {
+      "epoch": 0.008418267640780493,
+      "grad_norm": 1.7467809915542603,
+      "learning_rate": 1.7832832832832834e-05,
+      "loss": 1.2855,
+      "step": 6440
+    },
+    {
+      "epoch": 0.008419574825196764,
+      "grad_norm": 1.7581554651260376,
+      "learning_rate": 1.782782782782783e-05,
+      "loss": 1.7203,
+      "step": 6441
+    },
+    {
+      "epoch": 0.008420882009613034,
+      "grad_norm": 1.305393934249878,
+      "learning_rate": 1.7822822822822823e-05,
+      "loss": 1.6315,
+      "step": 6442
+    },
+    {
+      "epoch": 0.008422189194029305,
+      "grad_norm": 1.31243896484375,
+      "learning_rate": 1.781781781781782e-05,
+      "loss": 1.3868,
+      "step": 6443
+    },
+    {
+      "epoch": 0.008423496378445574,
+      "grad_norm": 2.0405962467193604,
+      "learning_rate": 1.7812812812812812e-05,
+      "loss": 2.2773,
+      "step": 6444
+    },
+    {
+      "epoch": 0.008424803562861845,
+      "grad_norm": 1.431647539138794,
+      "learning_rate": 1.780780780780781e-05,
+      "loss": 1.7939,
+      "step": 6445
+    },
+    {
+      "epoch": 0.008426110747278115,
+      "grad_norm": 1.4288309812545776,
+      "learning_rate": 1.7802802802802805e-05,
+      "loss": 1.3846,
+      "step": 6446
+    },
+    {
+      "epoch": 0.008427417931694386,
+      "grad_norm": 1.4952192306518555,
+      "learning_rate": 1.7797797797797798e-05,
+      "loss": 1.86,
+      "step": 6447
+    },
+    {
+      "epoch": 0.008428725116110656,
+      "grad_norm": 1.4935742616653442,
+      "learning_rate": 1.779279279279279e-05,
+      "loss": 1.4099,
+      "step": 6448
+    },
+    {
+      "epoch": 0.008430032300526927,
+      "grad_norm": 1.6427861452102661,
+      "learning_rate": 1.7787787787787787e-05,
+      "loss": 1.8425,
+      "step": 6449
+    },
+    {
+      "epoch": 0.008431339484943196,
+      "grad_norm": 1.7751133441925049,
+      "learning_rate": 1.7782782782782784e-05,
+      "loss": 2.2275,
+      "step": 6450
+    },
+    {
+      "epoch": 0.008432646669359467,
+      "grad_norm": 1.3370062112808228,
+      "learning_rate": 1.777777777777778e-05,
+      "loss": 0.8674,
+      "step": 6451
+    },
+    {
+      "epoch": 0.008433953853775737,
+      "grad_norm": 1.1876823902130127,
+      "learning_rate": 1.7772772772772776e-05,
+      "loss": 1.0973,
+      "step": 6452
+    },
+    {
+      "epoch": 0.008435261038192006,
+      "grad_norm": 1.9688209295272827,
+      "learning_rate": 1.776776776776777e-05,
+      "loss": 2.4128,
+      "step": 6453
+    },
+    {
+      "epoch": 0.008436568222608277,
+      "grad_norm": 1.5301251411437988,
+      "learning_rate": 1.7762762762762762e-05,
+      "loss": 1.8838,
+      "step": 6454
+    },
+    {
+      "epoch": 0.008437875407024547,
+      "grad_norm": 1.1433639526367188,
+      "learning_rate": 1.775775775775776e-05,
+      "loss": 0.9207,
+      "step": 6455
+    },
+    {
+      "epoch": 0.008439182591440818,
+      "grad_norm": 1.385521650314331,
+      "learning_rate": 1.7752752752752755e-05,
+      "loss": 1.3602,
+      "step": 6456
+    },
+    {
+      "epoch": 0.008440489775857088,
+      "grad_norm": 1.597213864326477,
+      "learning_rate": 1.7747747747747748e-05,
+      "loss": 0.9828,
+      "step": 6457
+    },
+    {
+      "epoch": 0.008441796960273359,
+      "grad_norm": 1.0501298904418945,
+      "learning_rate": 1.7742742742742744e-05,
+      "loss": 0.9437,
+      "step": 6458
+    },
+    {
+      "epoch": 0.008443104144689628,
+      "grad_norm": 1.3188316822052002,
+      "learning_rate": 1.773773773773774e-05,
+      "loss": 1.3965,
+      "step": 6459
+    },
+    {
+      "epoch": 0.0084444113291059,
+      "grad_norm": 1.8004968166351318,
+      "learning_rate": 1.7732732732732734e-05,
+      "loss": 2.0219,
+      "step": 6460
+    },
+    {
+      "epoch": 0.008445718513522169,
+      "grad_norm": 1.3028119802474976,
+      "learning_rate": 1.7727727727727727e-05,
+      "loss": 1.1174,
+      "step": 6461
+    },
+    {
+      "epoch": 0.00844702569793844,
+      "grad_norm": 1.701041340827942,
+      "learning_rate": 1.7722722722722723e-05,
+      "loss": 1.4784,
+      "step": 6462
+    },
+    {
+      "epoch": 0.00844833288235471,
+      "grad_norm": 1.3994437456130981,
+      "learning_rate": 1.771771771771772e-05,
+      "loss": 1.1301,
+      "step": 6463
+    },
+    {
+      "epoch": 0.00844964006677098,
+      "grad_norm": 2.2067294120788574,
+      "learning_rate": 1.7712712712712716e-05,
+      "loss": 1.9888,
+      "step": 6464
+    },
+    {
+      "epoch": 0.00845094725118725,
+      "grad_norm": 1.6221777200698853,
+      "learning_rate": 1.770770770770771e-05,
+      "loss": 1.0022,
+      "step": 6465
+    },
+    {
+      "epoch": 0.008452254435603521,
+      "grad_norm": 1.6722205877304077,
+      "learning_rate": 1.7702702702702702e-05,
+      "loss": 1.2904,
+      "step": 6466
+    },
+    {
+      "epoch": 0.00845356162001979,
+      "grad_norm": 1.3530011177062988,
+      "learning_rate": 1.7697697697697698e-05,
+      "loss": 1.5638,
+      "step": 6467
+    },
+    {
+      "epoch": 0.00845486880443606,
+      "grad_norm": 1.1884881258010864,
+      "learning_rate": 1.7692692692692694e-05,
+      "loss": 1.023,
+      "step": 6468
+    },
+    {
+      "epoch": 0.008456175988852331,
+      "grad_norm": 1.3773472309112549,
+      "learning_rate": 1.7687687687687687e-05,
+      "loss": 1.6967,
+      "step": 6469
+    },
+    {
+      "epoch": 0.008457483173268601,
+      "grad_norm": 1.378190279006958,
+      "learning_rate": 1.7682682682682684e-05,
+      "loss": 1.3926,
+      "step": 6470
+    },
+    {
+      "epoch": 0.008458790357684872,
+      "grad_norm": 1.4093787670135498,
+      "learning_rate": 1.767767767767768e-05,
+      "loss": 1.5418,
+      "step": 6471
+    },
+    {
+      "epoch": 0.008460097542101142,
+      "grad_norm": 3.2040350437164307,
+      "learning_rate": 1.7672672672672673e-05,
+      "loss": 2.0918,
+      "step": 6472
+    },
+    {
+      "epoch": 0.008461404726517413,
+      "grad_norm": 1.380606770515442,
+      "learning_rate": 1.7667667667667666e-05,
+      "loss": 1.0406,
+      "step": 6473
+    },
+    {
+      "epoch": 0.008462711910933682,
+      "grad_norm": 1.1675785779953003,
+      "learning_rate": 1.7662662662662662e-05,
+      "loss": 1.1966,
+      "step": 6474
+    },
+    {
+      "epoch": 0.008464019095349953,
+      "grad_norm": 1.3273314237594604,
+      "learning_rate": 1.765765765765766e-05,
+      "loss": 2.0449,
+      "step": 6475
+    },
+    {
+      "epoch": 0.008465326279766223,
+      "grad_norm": 2.0390400886535645,
+      "learning_rate": 1.7652652652652655e-05,
+      "loss": 1.8168,
+      "step": 6476
+    },
+    {
+      "epoch": 0.008466633464182494,
+      "grad_norm": 1.384561538696289,
+      "learning_rate": 1.7647647647647648e-05,
+      "loss": 2.0556,
+      "step": 6477
+    },
+    {
+      "epoch": 0.008467940648598763,
+      "grad_norm": 1.4235466718673706,
+      "learning_rate": 1.764264264264264e-05,
+      "loss": 1.6557,
+      "step": 6478
+    },
+    {
+      "epoch": 0.008469247833015035,
+      "grad_norm": 1.2112685441970825,
+      "learning_rate": 1.7637637637637637e-05,
+      "loss": 0.2909,
+      "step": 6479
+    },
+    {
+      "epoch": 0.008470555017431304,
+      "grad_norm": 1.7012403011322021,
+      "learning_rate": 1.7632632632632634e-05,
+      "loss": 1.3097,
+      "step": 6480
+    },
+    {
+      "epoch": 0.008471862201847575,
+      "grad_norm": 1.5302737951278687,
+      "learning_rate": 1.762762762762763e-05,
+      "loss": 1.9678,
+      "step": 6481
+    },
+    {
+      "epoch": 0.008473169386263845,
+      "grad_norm": 1.401724934577942,
+      "learning_rate": 1.7622622622622623e-05,
+      "loss": 2.1321,
+      "step": 6482
+    },
+    {
+      "epoch": 0.008474476570680114,
+      "grad_norm": 1.3824788331985474,
+      "learning_rate": 1.761761761761762e-05,
+      "loss": 1.4376,
+      "step": 6483
+    },
+    {
+      "epoch": 0.008475783755096385,
+      "grad_norm": 1.812900424003601,
+      "learning_rate": 1.7612612612612613e-05,
+      "loss": 1.1036,
+      "step": 6484
+    },
+    {
+      "epoch": 0.008477090939512655,
+      "grad_norm": 1.9244412183761597,
+      "learning_rate": 1.760760760760761e-05,
+      "loss": 1.5524,
+      "step": 6485
+    },
+    {
+      "epoch": 0.008478398123928926,
+      "grad_norm": 1.3349802494049072,
+      "learning_rate": 1.7602602602602602e-05,
+      "loss": 1.5142,
+      "step": 6486
+    },
+    {
+      "epoch": 0.008479705308345195,
+      "grad_norm": 1.439102053642273,
+      "learning_rate": 1.7597597597597598e-05,
+      "loss": 1.4854,
+      "step": 6487
+    },
+    {
+      "epoch": 0.008481012492761467,
+      "grad_norm": 1.8436557054519653,
+      "learning_rate": 1.7592592592592595e-05,
+      "loss": 2.3627,
+      "step": 6488
+    },
+    {
+      "epoch": 0.008482319677177736,
+      "grad_norm": 1.8799740076065063,
+      "learning_rate": 1.758758758758759e-05,
+      "loss": 1.6371,
+      "step": 6489
+    },
+    {
+      "epoch": 0.008483626861594007,
+      "grad_norm": 1.4273706674575806,
+      "learning_rate": 1.7582582582582584e-05,
+      "loss": 1.6816,
+      "step": 6490
+    },
+    {
+      "epoch": 0.008484934046010277,
+      "grad_norm": 1.5686638355255127,
+      "learning_rate": 1.7577577577577577e-05,
+      "loss": 1.7414,
+      "step": 6491
+    },
+    {
+      "epoch": 0.008486241230426548,
+      "grad_norm": 1.574822187423706,
+      "learning_rate": 1.7572572572572573e-05,
+      "loss": 1.8009,
+      "step": 6492
+    },
+    {
+      "epoch": 0.008487548414842817,
+      "grad_norm": 1.4001760482788086,
+      "learning_rate": 1.756756756756757e-05,
+      "loss": 2.0088,
+      "step": 6493
+    },
+    {
+      "epoch": 0.008488855599259089,
+      "grad_norm": 1.402129054069519,
+      "learning_rate": 1.7562562562562563e-05,
+      "loss": 1.5692,
+      "step": 6494
+    },
+    {
+      "epoch": 0.008490162783675358,
+      "grad_norm": 1.0241062641143799,
+      "learning_rate": 1.755755755755756e-05,
+      "loss": 0.7343,
+      "step": 6495
+    },
+    {
+      "epoch": 0.00849146996809163,
+      "grad_norm": 1.3004467487335205,
+      "learning_rate": 1.7552552552552552e-05,
+      "loss": 1.2547,
+      "step": 6496
+    },
+    {
+      "epoch": 0.008492777152507899,
+      "grad_norm": 0.7227709889411926,
+      "learning_rate": 1.7547547547547548e-05,
+      "loss": 0.3429,
+      "step": 6497
+    },
+    {
+      "epoch": 0.008494084336924168,
+      "grad_norm": 1.3914262056350708,
+      "learning_rate": 1.754254254254254e-05,
+      "loss": 1.4783,
+      "step": 6498
+    },
+    {
+      "epoch": 0.00849539152134044,
+      "grad_norm": 1.2414250373840332,
+      "learning_rate": 1.7537537537537538e-05,
+      "loss": 1.4212,
+      "step": 6499
+    },
+    {
+      "epoch": 0.008496698705756709,
+      "grad_norm": 1.5546578168869019,
+      "learning_rate": 1.7532532532532534e-05,
+      "loss": 2.0179,
+      "step": 6500
+    },
+    {
+      "epoch": 0.00849800589017298,
+      "grad_norm": 1.2971240282058716,
+      "learning_rate": 1.752752752752753e-05,
+      "loss": 1.034,
+      "step": 6501
+    },
+    {
+      "epoch": 0.00849931307458925,
+      "grad_norm": 1.5127595663070679,
+      "learning_rate": 1.7522522522522523e-05,
+      "loss": 2.2717,
+      "step": 6502
+    },
+    {
+      "epoch": 0.00850062025900552,
+      "grad_norm": 1.2601970434188843,
+      "learning_rate": 1.7517517517517516e-05,
+      "loss": 1.3494,
+      "step": 6503
+    },
+    {
+      "epoch": 0.00850192744342179,
+      "grad_norm": 1.3744269609451294,
+      "learning_rate": 1.7512512512512513e-05,
+      "loss": 1.6326,
+      "step": 6504
+    },
+    {
+      "epoch": 0.008503234627838061,
+      "grad_norm": 1.2352550029754639,
+      "learning_rate": 1.750750750750751e-05,
+      "loss": 1.1131,
+      "step": 6505
+    },
+    {
+      "epoch": 0.00850454181225433,
+      "grad_norm": 1.5802282094955444,
+      "learning_rate": 1.7502502502502505e-05,
+      "loss": 1.5548,
+      "step": 6506
+    },
+    {
+      "epoch": 0.008505848996670602,
+      "grad_norm": 1.4878593683242798,
+      "learning_rate": 1.74974974974975e-05,
+      "loss": 2.1493,
+      "step": 6507
+    },
+    {
+      "epoch": 0.008507156181086871,
+      "grad_norm": 0.967311680316925,
+      "learning_rate": 1.7492492492492495e-05,
+      "loss": 0.3658,
+      "step": 6508
+    },
+    {
+      "epoch": 0.008508463365503142,
+      "grad_norm": 1.0908420085906982,
+      "learning_rate": 1.7487487487487488e-05,
+      "loss": 0.3056,
+      "step": 6509
+    },
+    {
+      "epoch": 0.008509770549919412,
+      "grad_norm": 0.8348388671875,
+      "learning_rate": 1.7482482482482484e-05,
+      "loss": 0.1244,
+      "step": 6510
+    },
+    {
+      "epoch": 0.008511077734335683,
+      "grad_norm": 1.284062147140503,
+      "learning_rate": 1.7477477477477477e-05,
+      "loss": 1.1007,
+      "step": 6511
+    },
+    {
+      "epoch": 0.008512384918751953,
+      "grad_norm": 1.2084236145019531,
+      "learning_rate": 1.7472472472472473e-05,
+      "loss": 1.7049,
+      "step": 6512
+    },
+    {
+      "epoch": 0.008513692103168222,
+      "grad_norm": 0.20860786736011505,
+      "learning_rate": 1.746746746746747e-05,
+      "loss": 0.007,
+      "step": 6513
+    },
+    {
+      "epoch": 0.008514999287584493,
+      "grad_norm": 1.3644169569015503,
+      "learning_rate": 1.7462462462462466e-05,
+      "loss": 1.0257,
+      "step": 6514
+    },
+    {
+      "epoch": 0.008516306472000763,
+      "grad_norm": 1.3817188739776611,
+      "learning_rate": 1.7457457457457456e-05,
+      "loss": 1.7166,
+      "step": 6515
+    },
+    {
+      "epoch": 0.008517613656417034,
+      "grad_norm": 1.763692021369934,
+      "learning_rate": 1.7452452452452452e-05,
+      "loss": 2.2895,
+      "step": 6516
+    },
+    {
+      "epoch": 0.008518920840833303,
+      "grad_norm": 1.045519232749939,
+      "learning_rate": 1.744744744744745e-05,
+      "loss": 1.1352,
+      "step": 6517
+    },
+    {
+      "epoch": 0.008520228025249574,
+      "grad_norm": 1.3474986553192139,
+      "learning_rate": 1.7442442442442445e-05,
+      "loss": 2.3246,
+      "step": 6518
+    },
+    {
+      "epoch": 0.008521535209665844,
+      "grad_norm": 1.400022268295288,
+      "learning_rate": 1.7437437437437438e-05,
+      "loss": 1.851,
+      "step": 6519
+    },
+    {
+      "epoch": 0.008522842394082115,
+      "grad_norm": 1.402436375617981,
+      "learning_rate": 1.7432432432432434e-05,
+      "loss": 2.033,
+      "step": 6520
+    },
+    {
+      "epoch": 0.008524149578498385,
+      "grad_norm": 1.590602993965149,
+      "learning_rate": 1.7427427427427427e-05,
+      "loss": 1.323,
+      "step": 6521
+    },
+    {
+      "epoch": 0.008525456762914656,
+      "grad_norm": 1.4256433248519897,
+      "learning_rate": 1.7422422422422423e-05,
+      "loss": 1.665,
+      "step": 6522
+    },
+    {
+      "epoch": 0.008526763947330925,
+      "grad_norm": 2.033761501312256,
+      "learning_rate": 1.7417417417417416e-05,
+      "loss": 1.5988,
+      "step": 6523
+    },
+    {
+      "epoch": 0.008528071131747196,
+      "grad_norm": 1.3563417196273804,
+      "learning_rate": 1.7412412412412413e-05,
+      "loss": 1.4314,
+      "step": 6524
+    },
+    {
+      "epoch": 0.008529378316163466,
+      "grad_norm": 2.1943519115448,
+      "learning_rate": 1.740740740740741e-05,
+      "loss": 1.2434,
+      "step": 6525
+    },
+    {
+      "epoch": 0.008530685500579737,
+      "grad_norm": 1.6851998567581177,
+      "learning_rate": 1.7402402402402406e-05,
+      "loss": 1.7754,
+      "step": 6526
+    },
+    {
+      "epoch": 0.008531992684996006,
+      "grad_norm": 1.4579871892929077,
+      "learning_rate": 1.73973973973974e-05,
+      "loss": 1.5607,
+      "step": 6527
+    },
+    {
+      "epoch": 0.008533299869412278,
+      "grad_norm": 1.2198681831359863,
+      "learning_rate": 1.739239239239239e-05,
+      "loss": 1.2831,
+      "step": 6528
+    },
+    {
+      "epoch": 0.008534607053828547,
+      "grad_norm": 1.5666279792785645,
+      "learning_rate": 1.7387387387387388e-05,
+      "loss": 2.1905,
+      "step": 6529
+    },
+    {
+      "epoch": 0.008535914238244817,
+      "grad_norm": 1.718334674835205,
+      "learning_rate": 1.7382382382382384e-05,
+      "loss": 1.8148,
+      "step": 6530
+    },
+    {
+      "epoch": 0.008537221422661088,
+      "grad_norm": 1.49354887008667,
+      "learning_rate": 1.737737737737738e-05,
+      "loss": 1.2469,
+      "step": 6531
+    },
+    {
+      "epoch": 0.008538528607077357,
+      "grad_norm": 2.258219003677368,
+      "learning_rate": 1.7372372372372374e-05,
+      "loss": 1.9716,
+      "step": 6532
+    },
+    {
+      "epoch": 0.008539835791493628,
+      "grad_norm": 1.4020365476608276,
+      "learning_rate": 1.7367367367367366e-05,
+      "loss": 1.6311,
+      "step": 6533
+    },
+    {
+      "epoch": 0.008541142975909898,
+      "grad_norm": 1.3617963790893555,
+      "learning_rate": 1.7362362362362363e-05,
+      "loss": 1.4916,
+      "step": 6534
+    },
+    {
+      "epoch": 0.008542450160326169,
+      "grad_norm": 1.3482071161270142,
+      "learning_rate": 1.735735735735736e-05,
+      "loss": 2.0623,
+      "step": 6535
+    },
+    {
+      "epoch": 0.008543757344742439,
+      "grad_norm": 1.8056416511535645,
+      "learning_rate": 1.7352352352352352e-05,
+      "loss": 1.9799,
+      "step": 6536
+    },
+    {
+      "epoch": 0.00854506452915871,
+      "grad_norm": 1.064400553703308,
+      "learning_rate": 1.734734734734735e-05,
+      "loss": 1.0723,
+      "step": 6537
+    },
+    {
+      "epoch": 0.00854637171357498,
+      "grad_norm": 2.705655097961426,
+      "learning_rate": 1.7342342342342345e-05,
+      "loss": 2.0333,
+      "step": 6538
+    },
+    {
+      "epoch": 0.00854767889799125,
+      "grad_norm": 1.5135964155197144,
+      "learning_rate": 1.7337337337337338e-05,
+      "loss": 1.7103,
+      "step": 6539
+    },
+    {
+      "epoch": 0.00854898608240752,
+      "grad_norm": 1.3175292015075684,
+      "learning_rate": 1.733233233233233e-05,
+      "loss": 0.9414,
+      "step": 6540
+    },
+    {
+      "epoch": 0.008550293266823791,
+      "grad_norm": 1.4497907161712646,
+      "learning_rate": 1.7327327327327327e-05,
+      "loss": 1.1394,
+      "step": 6541
+    },
+    {
+      "epoch": 0.00855160045124006,
+      "grad_norm": 1.4928697347640991,
+      "learning_rate": 1.7322322322322324e-05,
+      "loss": 1.8079,
+      "step": 6542
+    },
+    {
+      "epoch": 0.008552907635656332,
+      "grad_norm": 1.1362396478652954,
+      "learning_rate": 1.731731731731732e-05,
+      "loss": 1.6614,
+      "step": 6543
+    },
+    {
+      "epoch": 0.008554214820072601,
+      "grad_norm": 1.2200061082839966,
+      "learning_rate": 1.7312312312312313e-05,
+      "loss": 0.9663,
+      "step": 6544
+    },
+    {
+      "epoch": 0.00855552200448887,
+      "grad_norm": 1.8217788934707642,
+      "learning_rate": 1.730730730730731e-05,
+      "loss": 1.983,
+      "step": 6545
+    },
+    {
+      "epoch": 0.008556829188905142,
+      "grad_norm": 1.8891481161117554,
+      "learning_rate": 1.7302302302302302e-05,
+      "loss": 2.0544,
+      "step": 6546
+    },
+    {
+      "epoch": 0.008558136373321411,
+      "grad_norm": 1.3598425388336182,
+      "learning_rate": 1.72972972972973e-05,
+      "loss": 0.9247,
+      "step": 6547
+    },
+    {
+      "epoch": 0.008559443557737682,
+      "grad_norm": 1.160760760307312,
+      "learning_rate": 1.729229229229229e-05,
+      "loss": 1.0348,
+      "step": 6548
+    },
+    {
+      "epoch": 0.008560750742153952,
+      "grad_norm": 1.4121906757354736,
+      "learning_rate": 1.7287287287287288e-05,
+      "loss": 1.7619,
+      "step": 6549
+    },
+    {
+      "epoch": 0.008562057926570223,
+      "grad_norm": 1.1793842315673828,
+      "learning_rate": 1.7282282282282284e-05,
+      "loss": 1.2606,
+      "step": 6550
+    },
+    {
+      "epoch": 0.008563365110986492,
+      "grad_norm": 1.5831375122070312,
+      "learning_rate": 1.7277277277277277e-05,
+      "loss": 1.815,
+      "step": 6551
+    },
+    {
+      "epoch": 0.008564672295402764,
+      "grad_norm": 1.347679615020752,
+      "learning_rate": 1.727227227227227e-05,
+      "loss": 1.5132,
+      "step": 6552
+    },
+    {
+      "epoch": 0.008565979479819033,
+      "grad_norm": 1.1889508962631226,
+      "learning_rate": 1.7267267267267267e-05,
+      "loss": 0.6992,
+      "step": 6553
+    },
+    {
+      "epoch": 0.008567286664235304,
+      "grad_norm": 1.2128857374191284,
+      "learning_rate": 1.7262262262262263e-05,
+      "loss": 0.9918,
+      "step": 6554
+    },
+    {
+      "epoch": 0.008568593848651574,
+      "grad_norm": 1.586735725402832,
+      "learning_rate": 1.725725725725726e-05,
+      "loss": 1.7155,
+      "step": 6555
+    },
+    {
+      "epoch": 0.008569901033067845,
+      "grad_norm": 1.8313721418380737,
+      "learning_rate": 1.7252252252252252e-05,
+      "loss": 1.0405,
+      "step": 6556
+    },
+    {
+      "epoch": 0.008571208217484114,
+      "grad_norm": 1.451818585395813,
+      "learning_rate": 1.724724724724725e-05,
+      "loss": 1.5772,
+      "step": 6557
+    },
+    {
+      "epoch": 0.008572515401900386,
+      "grad_norm": 1.1553593873977661,
+      "learning_rate": 1.724224224224224e-05,
+      "loss": 1.0992,
+      "step": 6558
+    },
+    {
+      "epoch": 0.008573822586316655,
+      "grad_norm": 1.1481719017028809,
+      "learning_rate": 1.7237237237237238e-05,
+      "loss": 1.5204,
+      "step": 6559
+    },
+    {
+      "epoch": 0.008575129770732924,
+      "grad_norm": 1.1963696479797363,
+      "learning_rate": 1.7232232232232234e-05,
+      "loss": 1.1353,
+      "step": 6560
+    },
+    {
+      "epoch": 0.008576436955149196,
+      "grad_norm": 1.3946319818496704,
+      "learning_rate": 1.7227227227227227e-05,
+      "loss": 1.9457,
+      "step": 6561
+    },
+    {
+      "epoch": 0.008577744139565465,
+      "grad_norm": 1.3328797817230225,
+      "learning_rate": 1.7222222222222224e-05,
+      "loss": 1.7649,
+      "step": 6562
+    },
+    {
+      "epoch": 0.008579051323981736,
+      "grad_norm": 1.255276083946228,
+      "learning_rate": 1.721721721721722e-05,
+      "loss": 1.2084,
+      "step": 6563
+    },
+    {
+      "epoch": 0.008580358508398006,
+      "grad_norm": 1.37405264377594,
+      "learning_rate": 1.7212212212212213e-05,
+      "loss": 1.8827,
+      "step": 6564
+    },
+    {
+      "epoch": 0.008581665692814277,
+      "grad_norm": 1.6172566413879395,
+      "learning_rate": 1.7207207207207206e-05,
+      "loss": 1.9237,
+      "step": 6565
+    },
+    {
+      "epoch": 0.008582972877230546,
+      "grad_norm": 1.3998901844024658,
+      "learning_rate": 1.7202202202202202e-05,
+      "loss": 1.4488,
+      "step": 6566
+    },
+    {
+      "epoch": 0.008584280061646818,
+      "grad_norm": 1.5911133289337158,
+      "learning_rate": 1.71971971971972e-05,
+      "loss": 1.6058,
+      "step": 6567
+    },
+    {
+      "epoch": 0.008585587246063087,
+      "grad_norm": 1.3326234817504883,
+      "learning_rate": 1.7192192192192195e-05,
+      "loss": 1.6302,
+      "step": 6568
+    },
+    {
+      "epoch": 0.008586894430479358,
+      "grad_norm": 1.4015599489212036,
+      "learning_rate": 1.7187187187187188e-05,
+      "loss": 1.2984,
+      "step": 6569
+    },
+    {
+      "epoch": 0.008588201614895628,
+      "grad_norm": 1.728670358657837,
+      "learning_rate": 1.718218218218218e-05,
+      "loss": 1.9412,
+      "step": 6570
+    },
+    {
+      "epoch": 0.008589508799311899,
+      "grad_norm": 1.6335312128067017,
+      "learning_rate": 1.7177177177177177e-05,
+      "loss": 1.5344,
+      "step": 6571
+    },
+    {
+      "epoch": 0.008590815983728168,
+      "grad_norm": 1.1830601692199707,
+      "learning_rate": 1.7172172172172174e-05,
+      "loss": 1.3134,
+      "step": 6572
+    },
+    {
+      "epoch": 0.00859212316814444,
+      "grad_norm": 1.1250618696212769,
+      "learning_rate": 1.7167167167167167e-05,
+      "loss": 0.8421,
+      "step": 6573
+    },
+    {
+      "epoch": 0.008593430352560709,
+      "grad_norm": 1.318769097328186,
+      "learning_rate": 1.7162162162162163e-05,
+      "loss": 1.6701,
+      "step": 6574
+    },
+    {
+      "epoch": 0.008594737536976978,
+      "grad_norm": 1.5590603351593018,
+      "learning_rate": 1.715715715715716e-05,
+      "loss": 2.0025,
+      "step": 6575
+    },
+    {
+      "epoch": 0.00859604472139325,
+      "grad_norm": 1.593848705291748,
+      "learning_rate": 1.7152152152152152e-05,
+      "loss": 1.6084,
+      "step": 6576
+    },
+    {
+      "epoch": 0.008597351905809519,
+      "grad_norm": 1.6471682786941528,
+      "learning_rate": 1.7147147147147145e-05,
+      "loss": 1.7055,
+      "step": 6577
+    },
+    {
+      "epoch": 0.00859865909022579,
+      "grad_norm": 1.5125977993011475,
+      "learning_rate": 1.7142142142142142e-05,
+      "loss": 1.5621,
+      "step": 6578
+    },
+    {
+      "epoch": 0.00859996627464206,
+      "grad_norm": 1.030529499053955,
+      "learning_rate": 1.7137137137137138e-05,
+      "loss": 0.6916,
+      "step": 6579
+    },
+    {
+      "epoch": 0.00860127345905833,
+      "grad_norm": 2.116105079650879,
+      "learning_rate": 1.7132132132132135e-05,
+      "loss": 1.5578,
+      "step": 6580
+    },
+    {
+      "epoch": 0.0086025806434746,
+      "grad_norm": 1.4792028665542603,
+      "learning_rate": 1.7127127127127128e-05,
+      "loss": 1.7183,
+      "step": 6581
+    },
+    {
+      "epoch": 0.008603887827890871,
+      "grad_norm": 0.826803982257843,
+      "learning_rate": 1.7122122122122124e-05,
+      "loss": 0.5427,
+      "step": 6582
+    },
+    {
+      "epoch": 0.008605195012307141,
+      "grad_norm": 1.1315858364105225,
+      "learning_rate": 1.7117117117117117e-05,
+      "loss": 0.8865,
+      "step": 6583
+    },
+    {
+      "epoch": 0.008606502196723412,
+      "grad_norm": 1.3966597318649292,
+      "learning_rate": 1.7112112112112113e-05,
+      "loss": 1.9451,
+      "step": 6584
+    },
+    {
+      "epoch": 0.008607809381139682,
+      "grad_norm": 1.4716246128082275,
+      "learning_rate": 1.710710710710711e-05,
+      "loss": 1.7144,
+      "step": 6585
+    },
+    {
+      "epoch": 0.008609116565555953,
+      "grad_norm": 1.4603126049041748,
+      "learning_rate": 1.7102102102102103e-05,
+      "loss": 1.5236,
+      "step": 6586
+    },
+    {
+      "epoch": 0.008610423749972222,
+      "grad_norm": 1.593819260597229,
+      "learning_rate": 1.70970970970971e-05,
+      "loss": 1.0068,
+      "step": 6587
+    },
+    {
+      "epoch": 0.008611730934388493,
+      "grad_norm": 1.702612280845642,
+      "learning_rate": 1.7092092092092092e-05,
+      "loss": 1.9818,
+      "step": 6588
+    },
+    {
+      "epoch": 0.008613038118804763,
+      "grad_norm": 1.197610855102539,
+      "learning_rate": 1.7087087087087088e-05,
+      "loss": 1.2493,
+      "step": 6589
+    },
+    {
+      "epoch": 0.008614345303221032,
+      "grad_norm": 1.5751311779022217,
+      "learning_rate": 1.708208208208208e-05,
+      "loss": 2.0452,
+      "step": 6590
+    },
+    {
+      "epoch": 0.008615652487637303,
+      "grad_norm": 1.7066060304641724,
+      "learning_rate": 1.7077077077077078e-05,
+      "loss": 1.6424,
+      "step": 6591
+    },
+    {
+      "epoch": 0.008616959672053573,
+      "grad_norm": 1.321193814277649,
+      "learning_rate": 1.7072072072072074e-05,
+      "loss": 1.8189,
+      "step": 6592
+    },
+    {
+      "epoch": 0.008618266856469844,
+      "grad_norm": 1.4580059051513672,
+      "learning_rate": 1.706706706706707e-05,
+      "loss": 1.2819,
+      "step": 6593
+    },
+    {
+      "epoch": 0.008619574040886114,
+      "grad_norm": 1.3514399528503418,
+      "learning_rate": 1.7062062062062063e-05,
+      "loss": 1.6471,
+      "step": 6594
+    },
+    {
+      "epoch": 0.008620881225302385,
+      "grad_norm": 1.5401005744934082,
+      "learning_rate": 1.7057057057057056e-05,
+      "loss": 1.7042,
+      "step": 6595
+    },
+    {
+      "epoch": 0.008622188409718654,
+      "grad_norm": 0.9740469455718994,
+      "learning_rate": 1.7052052052052053e-05,
+      "loss": 0.843,
+      "step": 6596
+    },
+    {
+      "epoch": 0.008623495594134925,
+      "grad_norm": 1.1000760793685913,
+      "learning_rate": 1.704704704704705e-05,
+      "loss": 0.7566,
+      "step": 6597
+    },
+    {
+      "epoch": 0.008624802778551195,
+      "grad_norm": 1.2391518354415894,
+      "learning_rate": 1.7042042042042042e-05,
+      "loss": 1.9883,
+      "step": 6598
+    },
+    {
+      "epoch": 0.008626109962967466,
+      "grad_norm": 1.2511746883392334,
+      "learning_rate": 1.7037037037037038e-05,
+      "loss": 1.8903,
+      "step": 6599
+    },
+    {
+      "epoch": 0.008627417147383735,
+      "grad_norm": 1.201683521270752,
+      "learning_rate": 1.7032032032032035e-05,
+      "loss": 0.9776,
+      "step": 6600
+    },
+    {
+      "epoch": 0.008628724331800007,
+      "grad_norm": 1.0939501523971558,
+      "learning_rate": 1.7027027027027028e-05,
+      "loss": 0.8766,
+      "step": 6601
+    },
+    {
+      "epoch": 0.008630031516216276,
+      "grad_norm": 1.218754529953003,
+      "learning_rate": 1.702202202202202e-05,
+      "loss": 1.407,
+      "step": 6602
+    },
+    {
+      "epoch": 0.008631338700632547,
+      "grad_norm": 1.3480890989303589,
+      "learning_rate": 1.7017017017017017e-05,
+      "loss": 1.3025,
+      "step": 6603
+    },
+    {
+      "epoch": 0.008632645885048817,
+      "grad_norm": 1.4150480031967163,
+      "learning_rate": 1.7012012012012013e-05,
+      "loss": 1.4546,
+      "step": 6604
+    },
+    {
+      "epoch": 0.008633953069465086,
+      "grad_norm": 1.150962471961975,
+      "learning_rate": 1.700700700700701e-05,
+      "loss": 0.9852,
+      "step": 6605
+    },
+    {
+      "epoch": 0.008635260253881357,
+      "grad_norm": 1.6067228317260742,
+      "learning_rate": 1.7002002002002003e-05,
+      "loss": 1.5349,
+      "step": 6606
+    },
+    {
+      "epoch": 0.008636567438297627,
+      "grad_norm": 1.663934350013733,
+      "learning_rate": 1.6996996996996996e-05,
+      "loss": 1.7651,
+      "step": 6607
+    },
+    {
+      "epoch": 0.008637874622713898,
+      "grad_norm": 1.673395037651062,
+      "learning_rate": 1.6991991991991992e-05,
+      "loss": 1.5095,
+      "step": 6608
+    },
+    {
+      "epoch": 0.008639181807130168,
+      "grad_norm": 1.4116712808609009,
+      "learning_rate": 1.698698698698699e-05,
+      "loss": 1.7316,
+      "step": 6609
+    },
+    {
+      "epoch": 0.008640488991546439,
+      "grad_norm": 1.577516794204712,
+      "learning_rate": 1.698198198198198e-05,
+      "loss": 1.0777,
+      "step": 6610
+    },
+    {
+      "epoch": 0.008641796175962708,
+      "grad_norm": 1.2335835695266724,
+      "learning_rate": 1.6976976976976978e-05,
+      "loss": 0.9839,
+      "step": 6611
+    },
+    {
+      "epoch": 0.00864310336037898,
+      "grad_norm": 1.1776411533355713,
+      "learning_rate": 1.6971971971971974e-05,
+      "loss": 1.5281,
+      "step": 6612
+    },
+    {
+      "epoch": 0.008644410544795249,
+      "grad_norm": 1.2412837743759155,
+      "learning_rate": 1.6966966966966967e-05,
+      "loss": 1.2284,
+      "step": 6613
+    },
+    {
+      "epoch": 0.00864571772921152,
+      "grad_norm": 1.170417308807373,
+      "learning_rate": 1.6961961961961963e-05,
+      "loss": 1.8174,
+      "step": 6614
+    },
+    {
+      "epoch": 0.00864702491362779,
+      "grad_norm": 1.3451905250549316,
+      "learning_rate": 1.6956956956956956e-05,
+      "loss": 1.4775,
+      "step": 6615
+    },
+    {
+      "epoch": 0.00864833209804406,
+      "grad_norm": 1.8078742027282715,
+      "learning_rate": 1.6951951951951953e-05,
+      "loss": 2.2976,
+      "step": 6616
+    },
+    {
+      "epoch": 0.00864963928246033,
+      "grad_norm": 1.7708909511566162,
+      "learning_rate": 1.694694694694695e-05,
+      "loss": 1.6548,
+      "step": 6617
+    },
+    {
+      "epoch": 0.008650946466876601,
+      "grad_norm": 1.6014842987060547,
+      "learning_rate": 1.6941941941941945e-05,
+      "loss": 1.1232,
+      "step": 6618
+    },
+    {
+      "epoch": 0.00865225365129287,
+      "grad_norm": 1.3794059753417969,
+      "learning_rate": 1.693693693693694e-05,
+      "loss": 1.6097,
+      "step": 6619
+    },
+    {
+      "epoch": 0.00865356083570914,
+      "grad_norm": 1.2640392780303955,
+      "learning_rate": 1.693193193193193e-05,
+      "loss": 1.6451,
+      "step": 6620
+    },
+    {
+      "epoch": 0.008654868020125411,
+      "grad_norm": 1.7823740243911743,
+      "learning_rate": 1.6926926926926928e-05,
+      "loss": 1.3505,
+      "step": 6621
+    },
+    {
+      "epoch": 0.00865617520454168,
+      "grad_norm": 1.6241542100906372,
+      "learning_rate": 1.6921921921921924e-05,
+      "loss": 1.5987,
+      "step": 6622
+    },
+    {
+      "epoch": 0.008657482388957952,
+      "grad_norm": 1.58807373046875,
+      "learning_rate": 1.6916916916916917e-05,
+      "loss": 2.0377,
+      "step": 6623
+    },
+    {
+      "epoch": 0.008658789573374221,
+      "grad_norm": 1.4395304918289185,
+      "learning_rate": 1.6911911911911913e-05,
+      "loss": 1.7344,
+      "step": 6624
+    },
+    {
+      "epoch": 0.008660096757790493,
+      "grad_norm": 1.2542924880981445,
+      "learning_rate": 1.6906906906906906e-05,
+      "loss": 2.1325,
+      "step": 6625
+    },
+    {
+      "epoch": 0.008661403942206762,
+      "grad_norm": 1.6839350461959839,
+      "learning_rate": 1.6901901901901903e-05,
+      "loss": 1.5916,
+      "step": 6626
+    },
+    {
+      "epoch": 0.008662711126623033,
+      "grad_norm": 1.512624979019165,
+      "learning_rate": 1.6896896896896896e-05,
+      "loss": 1.4077,
+      "step": 6627
+    },
+    {
+      "epoch": 0.008664018311039303,
+      "grad_norm": 1.2833216190338135,
+      "learning_rate": 1.6891891891891892e-05,
+      "loss": 1.5494,
+      "step": 6628
+    },
+    {
+      "epoch": 0.008665325495455574,
+      "grad_norm": 1.1440625190734863,
+      "learning_rate": 1.688688688688689e-05,
+      "loss": 1.1416,
+      "step": 6629
+    },
+    {
+      "epoch": 0.008666632679871843,
+      "grad_norm": 1.2954131364822388,
+      "learning_rate": 1.6881881881881885e-05,
+      "loss": 1.4985,
+      "step": 6630
+    },
+    {
+      "epoch": 0.008667939864288115,
+      "grad_norm": 1.4625028371810913,
+      "learning_rate": 1.6876876876876878e-05,
+      "loss": 1.4325,
+      "step": 6631
+    },
+    {
+      "epoch": 0.008669247048704384,
+      "grad_norm": 1.69624662399292,
+      "learning_rate": 1.687187187187187e-05,
+      "loss": 1.3833,
+      "step": 6632
+    },
+    {
+      "epoch": 0.008670554233120655,
+      "grad_norm": 1.4296391010284424,
+      "learning_rate": 1.6866866866866867e-05,
+      "loss": 1.0626,
+      "step": 6633
+    },
+    {
+      "epoch": 0.008671861417536925,
+      "grad_norm": 1.3262748718261719,
+      "learning_rate": 1.6861861861861864e-05,
+      "loss": 1.7004,
+      "step": 6634
+    },
+    {
+      "epoch": 0.008673168601953196,
+      "grad_norm": 1.4077486991882324,
+      "learning_rate": 1.6856856856856857e-05,
+      "loss": 1.3993,
+      "step": 6635
+    },
+    {
+      "epoch": 0.008674475786369465,
+      "grad_norm": 1.5457004308700562,
+      "learning_rate": 1.6851851851851853e-05,
+      "loss": 1.4788,
+      "step": 6636
+    },
+    {
+      "epoch": 0.008675782970785735,
+      "grad_norm": 1.024810791015625,
+      "learning_rate": 1.684684684684685e-05,
+      "loss": 0.9697,
+      "step": 6637
+    },
+    {
+      "epoch": 0.008677090155202006,
+      "grad_norm": 1.4806605577468872,
+      "learning_rate": 1.6841841841841842e-05,
+      "loss": 1.3888,
+      "step": 6638
+    },
+    {
+      "epoch": 0.008678397339618275,
+      "grad_norm": 1.3484513759613037,
+      "learning_rate": 1.683683683683684e-05,
+      "loss": 1.2942,
+      "step": 6639
+    },
+    {
+      "epoch": 0.008679704524034547,
+      "grad_norm": 1.818913221359253,
+      "learning_rate": 1.683183183183183e-05,
+      "loss": 2.1203,
+      "step": 6640
+    },
+    {
+      "epoch": 0.008681011708450816,
+      "grad_norm": 1.3280267715454102,
+      "learning_rate": 1.6826826826826828e-05,
+      "loss": 2.4279,
+      "step": 6641
+    },
+    {
+      "epoch": 0.008682318892867087,
+      "grad_norm": 1.2506970167160034,
+      "learning_rate": 1.6821821821821824e-05,
+      "loss": 2.0616,
+      "step": 6642
+    },
+    {
+      "epoch": 0.008683626077283357,
+      "grad_norm": 1.877939224243164,
+      "learning_rate": 1.6816816816816817e-05,
+      "loss": 1.544,
+      "step": 6643
+    },
+    {
+      "epoch": 0.008684933261699628,
+      "grad_norm": 1.1956266164779663,
+      "learning_rate": 1.681181181181181e-05,
+      "loss": 1.4852,
+      "step": 6644
+    },
+    {
+      "epoch": 0.008686240446115897,
+      "grad_norm": 1.0721189975738525,
+      "learning_rate": 1.6806806806806807e-05,
+      "loss": 0.9691,
+      "step": 6645
+    },
+    {
+      "epoch": 0.008687547630532168,
+      "grad_norm": 1.2936511039733887,
+      "learning_rate": 1.6801801801801803e-05,
+      "loss": 1.7968,
+      "step": 6646
+    },
+    {
+      "epoch": 0.008688854814948438,
+      "grad_norm": 0.3741351366043091,
+      "learning_rate": 1.67967967967968e-05,
+      "loss": 0.11,
+      "step": 6647
+    },
+    {
+      "epoch": 0.008690161999364709,
+      "grad_norm": 1.451588749885559,
+      "learning_rate": 1.6791791791791792e-05,
+      "loss": 1.154,
+      "step": 6648
+    },
+    {
+      "epoch": 0.008691469183780979,
+      "grad_norm": 1.45819091796875,
+      "learning_rate": 1.678678678678679e-05,
+      "loss": 1.4949,
+      "step": 6649
+    },
+    {
+      "epoch": 0.00869277636819725,
+      "grad_norm": 1.1751042604446411,
+      "learning_rate": 1.678178178178178e-05,
+      "loss": 1.6756,
+      "step": 6650
+    },
+    {
+      "epoch": 0.00869408355261352,
+      "grad_norm": 1.771199345588684,
+      "learning_rate": 1.6776776776776778e-05,
+      "loss": 1.3897,
+      "step": 6651
+    },
+    {
+      "epoch": 0.008695390737029789,
+      "grad_norm": 1.5398545265197754,
+      "learning_rate": 1.677177177177177e-05,
+      "loss": 1.8836,
+      "step": 6652
+    },
+    {
+      "epoch": 0.00869669792144606,
+      "grad_norm": 1.6017918586730957,
+      "learning_rate": 1.6766766766766767e-05,
+      "loss": 2.3208,
+      "step": 6653
+    },
+    {
+      "epoch": 0.00869800510586233,
+      "grad_norm": 1.4289231300354004,
+      "learning_rate": 1.6761761761761764e-05,
+      "loss": 1.0626,
+      "step": 6654
+    },
+    {
+      "epoch": 0.0086993122902786,
+      "grad_norm": 1.9925352334976196,
+      "learning_rate": 1.675675675675676e-05,
+      "loss": 2.4033,
+      "step": 6655
+    },
+    {
+      "epoch": 0.00870061947469487,
+      "grad_norm": 1.610689401626587,
+      "learning_rate": 1.6751751751751753e-05,
+      "loss": 1.5502,
+      "step": 6656
+    },
+    {
+      "epoch": 0.008701926659111141,
+      "grad_norm": 1.3631068468093872,
+      "learning_rate": 1.6746746746746746e-05,
+      "loss": 1.8406,
+      "step": 6657
+    },
+    {
+      "epoch": 0.00870323384352741,
+      "grad_norm": 1.0560575723648071,
+      "learning_rate": 1.6741741741741742e-05,
+      "loss": 0.3681,
+      "step": 6658
+    },
+    {
+      "epoch": 0.008704541027943682,
+      "grad_norm": 1.7092682123184204,
+      "learning_rate": 1.673673673673674e-05,
+      "loss": 2.0459,
+      "step": 6659
+    },
+    {
+      "epoch": 0.008705848212359951,
+      "grad_norm": 1.4070724248886108,
+      "learning_rate": 1.6731731731731732e-05,
+      "loss": 1.6725,
+      "step": 6660
+    },
+    {
+      "epoch": 0.008707155396776222,
+      "grad_norm": 1.1875488758087158,
+      "learning_rate": 1.6726726726726728e-05,
+      "loss": 0.9471,
+      "step": 6661
+    },
+    {
+      "epoch": 0.008708462581192492,
+      "grad_norm": 1.5060644149780273,
+      "learning_rate": 1.672172172172172e-05,
+      "loss": 1.9012,
+      "step": 6662
+    },
+    {
+      "epoch": 0.008709769765608763,
+      "grad_norm": 1.11302649974823,
+      "learning_rate": 1.6716716716716717e-05,
+      "loss": 0.9949,
+      "step": 6663
+    },
+    {
+      "epoch": 0.008711076950025032,
+      "grad_norm": 1.5182650089263916,
+      "learning_rate": 1.671171171171171e-05,
+      "loss": 1.3229,
+      "step": 6664
+    },
+    {
+      "epoch": 0.008712384134441304,
+      "grad_norm": 1.7640209197998047,
+      "learning_rate": 1.6706706706706707e-05,
+      "loss": 1.814,
+      "step": 6665
+    },
+    {
+      "epoch": 0.008713691318857573,
+      "grad_norm": 1.6542978286743164,
+      "learning_rate": 1.6701701701701703e-05,
+      "loss": 1.3456,
+      "step": 6666
+    },
+    {
+      "epoch": 0.008714998503273843,
+      "grad_norm": 1.4910675287246704,
+      "learning_rate": 1.66966966966967e-05,
+      "loss": 1.3531,
+      "step": 6667
+    },
+    {
+      "epoch": 0.008716305687690114,
+      "grad_norm": 1.0565179586410522,
+      "learning_rate": 1.6691691691691692e-05,
+      "loss": 0.8779,
+      "step": 6668
+    },
+    {
+      "epoch": 0.008717612872106383,
+      "grad_norm": 1.373948574066162,
+      "learning_rate": 1.6686686686686685e-05,
+      "loss": 1.275,
+      "step": 6669
+    },
+    {
+      "epoch": 0.008718920056522654,
+      "grad_norm": 1.680833101272583,
+      "learning_rate": 1.6681681681681682e-05,
+      "loss": 1.4576,
+      "step": 6670
+    },
+    {
+      "epoch": 0.008720227240938924,
+      "grad_norm": 1.1494618654251099,
+      "learning_rate": 1.6676676676676678e-05,
+      "loss": 1.626,
+      "step": 6671
+    },
+    {
+      "epoch": 0.008721534425355195,
+      "grad_norm": 1.4084579944610596,
+      "learning_rate": 1.6671671671671674e-05,
+      "loss": 2.0067,
+      "step": 6672
+    },
+    {
+      "epoch": 0.008722841609771464,
+      "grad_norm": 1.3984785079956055,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.2278,
+      "step": 6673
+    },
+    {
+      "epoch": 0.008724148794187736,
+      "grad_norm": 1.1531147956848145,
+      "learning_rate": 1.6661661661661664e-05,
+      "loss": 1.3879,
+      "step": 6674
+    },
+    {
+      "epoch": 0.008725455978604005,
+      "grad_norm": 1.657248616218567,
+      "learning_rate": 1.6656656656656657e-05,
+      "loss": 1.6249,
+      "step": 6675
+    },
+    {
+      "epoch": 0.008726763163020276,
+      "grad_norm": 1.4212586879730225,
+      "learning_rate": 1.6651651651651653e-05,
+      "loss": 1.563,
+      "step": 6676
+    },
+    {
+      "epoch": 0.008728070347436546,
+      "grad_norm": 1.3403770923614502,
+      "learning_rate": 1.6646646646646646e-05,
+      "loss": 1.4933,
+      "step": 6677
+    },
+    {
+      "epoch": 0.008729377531852817,
+      "grad_norm": 1.5957987308502197,
+      "learning_rate": 1.6641641641641643e-05,
+      "loss": 0.6249,
+      "step": 6678
+    },
+    {
+      "epoch": 0.008730684716269086,
+      "grad_norm": 1.7927981615066528,
+      "learning_rate": 1.663663663663664e-05,
+      "loss": 1.8584,
+      "step": 6679
+    },
+    {
+      "epoch": 0.008731991900685358,
+      "grad_norm": 1.3457629680633545,
+      "learning_rate": 1.6631631631631632e-05,
+      "loss": 1.2505,
+      "step": 6680
+    },
+    {
+      "epoch": 0.008733299085101627,
+      "grad_norm": 2.766538381576538,
+      "learning_rate": 1.6626626626626625e-05,
+      "loss": 1.7272,
+      "step": 6681
+    },
+    {
+      "epoch": 0.008734606269517897,
+      "grad_norm": 1.6684551239013672,
+      "learning_rate": 1.662162162162162e-05,
+      "loss": 1.5301,
+      "step": 6682
+    },
+    {
+      "epoch": 0.008735913453934168,
+      "grad_norm": 2.165250062942505,
+      "learning_rate": 1.6616616616616618e-05,
+      "loss": 2.4551,
+      "step": 6683
+    },
+    {
+      "epoch": 0.008737220638350437,
+      "grad_norm": 1.409316062927246,
+      "learning_rate": 1.6611611611611614e-05,
+      "loss": 2.4289,
+      "step": 6684
+    },
+    {
+      "epoch": 0.008738527822766708,
+      "grad_norm": 0.9290755987167358,
+      "learning_rate": 1.6606606606606607e-05,
+      "loss": 0.5438,
+      "step": 6685
+    },
+    {
+      "epoch": 0.008739835007182978,
+      "grad_norm": 1.3902826309204102,
+      "learning_rate": 1.6601601601601603e-05,
+      "loss": 1.3658,
+      "step": 6686
+    },
+    {
+      "epoch": 0.008741142191599249,
+      "grad_norm": 1.3156968355178833,
+      "learning_rate": 1.6596596596596596e-05,
+      "loss": 1.5241,
+      "step": 6687
+    },
+    {
+      "epoch": 0.008742449376015518,
+      "grad_norm": 1.6371721029281616,
+      "learning_rate": 1.6591591591591593e-05,
+      "loss": 1.6078,
+      "step": 6688
+    },
+    {
+      "epoch": 0.00874375656043179,
+      "grad_norm": 1.689353108406067,
+      "learning_rate": 1.6586586586586586e-05,
+      "loss": 1.3198,
+      "step": 6689
+    },
+    {
+      "epoch": 0.008745063744848059,
+      "grad_norm": 1.3208091259002686,
+      "learning_rate": 1.6581581581581582e-05,
+      "loss": 1.6786,
+      "step": 6690
+    },
+    {
+      "epoch": 0.00874637092926433,
+      "grad_norm": 1.4769662618637085,
+      "learning_rate": 1.6576576576576578e-05,
+      "loss": 2.4405,
+      "step": 6691
+    },
+    {
+      "epoch": 0.0087476781136806,
+      "grad_norm": 1.2001208066940308,
+      "learning_rate": 1.6571571571571575e-05,
+      "loss": 1.1936,
+      "step": 6692
+    },
+    {
+      "epoch": 0.008748985298096871,
+      "grad_norm": 1.4363027811050415,
+      "learning_rate": 1.6566566566566568e-05,
+      "loss": 1.6636,
+      "step": 6693
+    },
+    {
+      "epoch": 0.00875029248251314,
+      "grad_norm": 1.558960199356079,
+      "learning_rate": 1.656156156156156e-05,
+      "loss": 1.5684,
+      "step": 6694
+    },
+    {
+      "epoch": 0.008751599666929412,
+      "grad_norm": 1.1881210803985596,
+      "learning_rate": 1.6556556556556557e-05,
+      "loss": 0.5094,
+      "step": 6695
+    },
+    {
+      "epoch": 0.008752906851345681,
+      "grad_norm": 1.3455973863601685,
+      "learning_rate": 1.6551551551551553e-05,
+      "loss": 1.2791,
+      "step": 6696
+    },
+    {
+      "epoch": 0.00875421403576195,
+      "grad_norm": 1.4890570640563965,
+      "learning_rate": 1.654654654654655e-05,
+      "loss": 2.0642,
+      "step": 6697
+    },
+    {
+      "epoch": 0.008755521220178222,
+      "grad_norm": 1.5027759075164795,
+      "learning_rate": 1.6541541541541543e-05,
+      "loss": 1.8288,
+      "step": 6698
+    },
+    {
+      "epoch": 0.008756828404594491,
+      "grad_norm": 1.3593223094940186,
+      "learning_rate": 1.6536536536536536e-05,
+      "loss": 1.6389,
+      "step": 6699
+    },
+    {
+      "epoch": 0.008758135589010762,
+      "grad_norm": 1.2759027481079102,
+      "learning_rate": 1.6531531531531532e-05,
+      "loss": 1.6106,
+      "step": 6700
+    },
+    {
+      "epoch": 0.008759442773427032,
+      "grad_norm": 1.3860821723937988,
+      "learning_rate": 1.652652652652653e-05,
+      "loss": 1.6389,
+      "step": 6701
+    },
+    {
+      "epoch": 0.008760749957843303,
+      "grad_norm": 1.5477522611618042,
+      "learning_rate": 1.652152152152152e-05,
+      "loss": 1.5838,
+      "step": 6702
+    },
+    {
+      "epoch": 0.008762057142259572,
+      "grad_norm": 1.136335849761963,
+      "learning_rate": 1.6516516516516518e-05,
+      "loss": 0.8625,
+      "step": 6703
+    },
+    {
+      "epoch": 0.008763364326675844,
+      "grad_norm": 1.3826769590377808,
+      "learning_rate": 1.6511511511511514e-05,
+      "loss": 1.7782,
+      "step": 6704
+    },
+    {
+      "epoch": 0.008764671511092113,
+      "grad_norm": 1.5120704174041748,
+      "learning_rate": 1.6506506506506507e-05,
+      "loss": 1.699,
+      "step": 6705
+    },
+    {
+      "epoch": 0.008765978695508384,
+      "grad_norm": 1.471538782119751,
+      "learning_rate": 1.65015015015015e-05,
+      "loss": 2.4128,
+      "step": 6706
+    },
+    {
+      "epoch": 0.008767285879924654,
+      "grad_norm": 1.512444257736206,
+      "learning_rate": 1.6496496496496496e-05,
+      "loss": 1.528,
+      "step": 6707
+    },
+    {
+      "epoch": 0.008768593064340925,
+      "grad_norm": 1.5936181545257568,
+      "learning_rate": 1.6491491491491493e-05,
+      "loss": 2.1387,
+      "step": 6708
+    },
+    {
+      "epoch": 0.008769900248757194,
+      "grad_norm": 1.578963279724121,
+      "learning_rate": 1.648648648648649e-05,
+      "loss": 1.7241,
+      "step": 6709
+    },
+    {
+      "epoch": 0.008771207433173465,
+      "grad_norm": 1.5635260343551636,
+      "learning_rate": 1.6481481481481482e-05,
+      "loss": 1.3969,
+      "step": 6710
+    },
+    {
+      "epoch": 0.008772514617589735,
+      "grad_norm": 1.343226432800293,
+      "learning_rate": 1.647647647647648e-05,
+      "loss": 1.6564,
+      "step": 6711
+    },
+    {
+      "epoch": 0.008773821802006004,
+      "grad_norm": 1.596472144126892,
+      "learning_rate": 1.647147147147147e-05,
+      "loss": 1.8604,
+      "step": 6712
+    },
+    {
+      "epoch": 0.008775128986422276,
+      "grad_norm": 1.598349690437317,
+      "learning_rate": 1.6466466466466468e-05,
+      "loss": 2.5113,
+      "step": 6713
+    },
+    {
+      "epoch": 0.008776436170838545,
+      "grad_norm": 1.1724307537078857,
+      "learning_rate": 1.646146146146146e-05,
+      "loss": 1.5235,
+      "step": 6714
+    },
+    {
+      "epoch": 0.008777743355254816,
+      "grad_norm": 1.3333185911178589,
+      "learning_rate": 1.6456456456456457e-05,
+      "loss": 1.6064,
+      "step": 6715
+    },
+    {
+      "epoch": 0.008779050539671086,
+      "grad_norm": 1.637776255607605,
+      "learning_rate": 1.6451451451451453e-05,
+      "loss": 2.0591,
+      "step": 6716
+    },
+    {
+      "epoch": 0.008780357724087357,
+      "grad_norm": 1.155055046081543,
+      "learning_rate": 1.6446446446446446e-05,
+      "loss": 1.1154,
+      "step": 6717
+    },
+    {
+      "epoch": 0.008781664908503626,
+      "grad_norm": 1.450541615486145,
+      "learning_rate": 1.6441441441441443e-05,
+      "loss": 1.9267,
+      "step": 6718
+    },
+    {
+      "epoch": 0.008782972092919897,
+      "grad_norm": 1.3940937519073486,
+      "learning_rate": 1.6436436436436436e-05,
+      "loss": 1.3136,
+      "step": 6719
+    },
+    {
+      "epoch": 0.008784279277336167,
+      "grad_norm": 1.0742074251174927,
+      "learning_rate": 1.6431431431431432e-05,
+      "loss": 1.1326,
+      "step": 6720
+    },
+    {
+      "epoch": 0.008785586461752438,
+      "grad_norm": 1.015051007270813,
+      "learning_rate": 1.642642642642643e-05,
+      "loss": 0.8592,
+      "step": 6721
+    },
+    {
+      "epoch": 0.008786893646168708,
+      "grad_norm": 1.3341408967971802,
+      "learning_rate": 1.6421421421421425e-05,
+      "loss": 1.5341,
+      "step": 6722
+    },
+    {
+      "epoch": 0.008788200830584979,
+      "grad_norm": 1.0386451482772827,
+      "learning_rate": 1.6416416416416418e-05,
+      "loss": 0.7812,
+      "step": 6723
+    },
+    {
+      "epoch": 0.008789508015001248,
+      "grad_norm": 1.3599796295166016,
+      "learning_rate": 1.641141141141141e-05,
+      "loss": 1.708,
+      "step": 6724
+    },
+    {
+      "epoch": 0.00879081519941752,
+      "grad_norm": 1.7491966485977173,
+      "learning_rate": 1.6406406406406407e-05,
+      "loss": 1.2599,
+      "step": 6725
+    },
+    {
+      "epoch": 0.008792122383833789,
+      "grad_norm": 2.1537868976593018,
+      "learning_rate": 1.6401401401401404e-05,
+      "loss": 2.0214,
+      "step": 6726
+    },
+    {
+      "epoch": 0.008793429568250058,
+      "grad_norm": 1.391038179397583,
+      "learning_rate": 1.6396396396396396e-05,
+      "loss": 1.3988,
+      "step": 6727
+    },
+    {
+      "epoch": 0.00879473675266633,
+      "grad_norm": 1.6241068840026855,
+      "learning_rate": 1.6391391391391393e-05,
+      "loss": 1.3332,
+      "step": 6728
+    },
+    {
+      "epoch": 0.008796043937082599,
+      "grad_norm": 1.3256312608718872,
+      "learning_rate": 1.638638638638639e-05,
+      "loss": 1.1243,
+      "step": 6729
+    },
+    {
+      "epoch": 0.00879735112149887,
+      "grad_norm": 1.5751688480377197,
+      "learning_rate": 1.6381381381381382e-05,
+      "loss": 1.9014,
+      "step": 6730
+    },
+    {
+      "epoch": 0.00879865830591514,
+      "grad_norm": 2.0089762210845947,
+      "learning_rate": 1.6376376376376375e-05,
+      "loss": 1.7016,
+      "step": 6731
+    },
+    {
+      "epoch": 0.00879996549033141,
+      "grad_norm": 1.2128419876098633,
+      "learning_rate": 1.637137137137137e-05,
+      "loss": 1.009,
+      "step": 6732
+    },
+    {
+      "epoch": 0.00880127267474768,
+      "grad_norm": 1.7220306396484375,
+      "learning_rate": 1.6366366366366368e-05,
+      "loss": 1.0463,
+      "step": 6733
+    },
+    {
+      "epoch": 0.008802579859163951,
+      "grad_norm": 1.7378060817718506,
+      "learning_rate": 1.6361361361361364e-05,
+      "loss": 1.3789,
+      "step": 6734
+    },
+    {
+      "epoch": 0.00880388704358022,
+      "grad_norm": 1.444074273109436,
+      "learning_rate": 1.6356356356356357e-05,
+      "loss": 0.8064,
+      "step": 6735
+    },
+    {
+      "epoch": 0.008805194227996492,
+      "grad_norm": 1.3891587257385254,
+      "learning_rate": 1.635135135135135e-05,
+      "loss": 2.0638,
+      "step": 6736
+    },
+    {
+      "epoch": 0.008806501412412761,
+      "grad_norm": 1.6312363147735596,
+      "learning_rate": 1.6346346346346347e-05,
+      "loss": 2.1394,
+      "step": 6737
+    },
+    {
+      "epoch": 0.008807808596829033,
+      "grad_norm": 1.5263774394989014,
+      "learning_rate": 1.6341341341341343e-05,
+      "loss": 1.7512,
+      "step": 6738
+    },
+    {
+      "epoch": 0.008809115781245302,
+      "grad_norm": 1.2537130117416382,
+      "learning_rate": 1.6336336336336336e-05,
+      "loss": 1.341,
+      "step": 6739
+    },
+    {
+      "epoch": 0.008810422965661573,
+      "grad_norm": 1.273511290550232,
+      "learning_rate": 1.6331331331331332e-05,
+      "loss": 1.2823,
+      "step": 6740
+    },
+    {
+      "epoch": 0.008811730150077843,
+      "grad_norm": 1.5487428903579712,
+      "learning_rate": 1.632632632632633e-05,
+      "loss": 1.5054,
+      "step": 6741
+    },
+    {
+      "epoch": 0.008813037334494114,
+      "grad_norm": 1.1159732341766357,
+      "learning_rate": 1.632132132132132e-05,
+      "loss": 0.9848,
+      "step": 6742
+    },
+    {
+      "epoch": 0.008814344518910383,
+      "grad_norm": 1.186806082725525,
+      "learning_rate": 1.6316316316316315e-05,
+      "loss": 1.2406,
+      "step": 6743
+    },
+    {
+      "epoch": 0.008815651703326653,
+      "grad_norm": 1.957993745803833,
+      "learning_rate": 1.631131131131131e-05,
+      "loss": 2.192,
+      "step": 6744
+    },
+    {
+      "epoch": 0.008816958887742924,
+      "grad_norm": 1.721553087234497,
+      "learning_rate": 1.6306306306306307e-05,
+      "loss": 1.3408,
+      "step": 6745
+    },
+    {
+      "epoch": 0.008818266072159194,
+      "grad_norm": 1.1703215837478638,
+      "learning_rate": 1.6301301301301304e-05,
+      "loss": 0.8967,
+      "step": 6746
+    },
+    {
+      "epoch": 0.008819573256575465,
+      "grad_norm": 1.4136805534362793,
+      "learning_rate": 1.62962962962963e-05,
+      "loss": 2.1449,
+      "step": 6747
+    },
+    {
+      "epoch": 0.008820880440991734,
+      "grad_norm": 1.8290131092071533,
+      "learning_rate": 1.629129129129129e-05,
+      "loss": 2.4193,
+      "step": 6748
+    },
+    {
+      "epoch": 0.008822187625408005,
+      "grad_norm": 1.2937935590744019,
+      "learning_rate": 1.6286286286286286e-05,
+      "loss": 0.8397,
+      "step": 6749
+    },
+    {
+      "epoch": 0.008823494809824275,
+      "grad_norm": 1.3545846939086914,
+      "learning_rate": 1.6281281281281282e-05,
+      "loss": 1.8232,
+      "step": 6750
+    },
+    {
+      "epoch": 0.008824801994240546,
+      "grad_norm": 1.227016568183899,
+      "learning_rate": 1.627627627627628e-05,
+      "loss": 1.5596,
+      "step": 6751
+    },
+    {
+      "epoch": 0.008826109178656815,
+      "grad_norm": 1.274713158607483,
+      "learning_rate": 1.627127127127127e-05,
+      "loss": 2.1362,
+      "step": 6752
+    },
+    {
+      "epoch": 0.008827416363073087,
+      "grad_norm": 1.534730315208435,
+      "learning_rate": 1.6266266266266268e-05,
+      "loss": 1.2429,
+      "step": 6753
+    },
+    {
+      "epoch": 0.008828723547489356,
+      "grad_norm": 1.3164405822753906,
+      "learning_rate": 1.626126126126126e-05,
+      "loss": 1.6983,
+      "step": 6754
+    },
+    {
+      "epoch": 0.008830030731905627,
+      "grad_norm": 1.237406849861145,
+      "learning_rate": 1.6256256256256257e-05,
+      "loss": 1.2241,
+      "step": 6755
+    },
+    {
+      "epoch": 0.008831337916321897,
+      "grad_norm": 1.409958839416504,
+      "learning_rate": 1.625125125125125e-05,
+      "loss": 1.2877,
+      "step": 6756
+    },
+    {
+      "epoch": 0.008832645100738168,
+      "grad_norm": 2.4400012493133545,
+      "learning_rate": 1.6246246246246247e-05,
+      "loss": 1.5326,
+      "step": 6757
+    },
+    {
+      "epoch": 0.008833952285154437,
+      "grad_norm": 1.5411534309387207,
+      "learning_rate": 1.6241241241241243e-05,
+      "loss": 2.1962,
+      "step": 6758
+    },
+    {
+      "epoch": 0.008835259469570707,
+      "grad_norm": 1.4882636070251465,
+      "learning_rate": 1.623623623623624e-05,
+      "loss": 1.3149,
+      "step": 6759
+    },
+    {
+      "epoch": 0.008836566653986978,
+      "grad_norm": 2.057269811630249,
+      "learning_rate": 1.6231231231231232e-05,
+      "loss": 1.9937,
+      "step": 6760
+    },
+    {
+      "epoch": 0.008837873838403247,
+      "grad_norm": 1.5623546838760376,
+      "learning_rate": 1.6226226226226225e-05,
+      "loss": 1.4824,
+      "step": 6761
+    },
+    {
+      "epoch": 0.008839181022819519,
+      "grad_norm": 1.3371094465255737,
+      "learning_rate": 1.6221221221221222e-05,
+      "loss": 1.0364,
+      "step": 6762
+    },
+    {
+      "epoch": 0.008840488207235788,
+      "grad_norm": 1.8581725358963013,
+      "learning_rate": 1.6216216216216218e-05,
+      "loss": 1.4137,
+      "step": 6763
+    },
+    {
+      "epoch": 0.00884179539165206,
+      "grad_norm": 1.7339259386062622,
+      "learning_rate": 1.621121121121121e-05,
+      "loss": 1.4599,
+      "step": 6764
+    },
+    {
+      "epoch": 0.008843102576068329,
+      "grad_norm": 1.5113145112991333,
+      "learning_rate": 1.6206206206206207e-05,
+      "loss": 2.1794,
+      "step": 6765
+    },
+    {
+      "epoch": 0.0088444097604846,
+      "grad_norm": 1.753060221672058,
+      "learning_rate": 1.6201201201201204e-05,
+      "loss": 0.3111,
+      "step": 6766
+    },
+    {
+      "epoch": 0.00884571694490087,
+      "grad_norm": 1.4010261297225952,
+      "learning_rate": 1.6196196196196197e-05,
+      "loss": 1.7954,
+      "step": 6767
+    },
+    {
+      "epoch": 0.00884702412931714,
+      "grad_norm": 2.079833507537842,
+      "learning_rate": 1.619119119119119e-05,
+      "loss": 1.3976,
+      "step": 6768
+    },
+    {
+      "epoch": 0.00884833131373341,
+      "grad_norm": 1.6059303283691406,
+      "learning_rate": 1.6186186186186186e-05,
+      "loss": 1.787,
+      "step": 6769
+    },
+    {
+      "epoch": 0.008849638498149681,
+      "grad_norm": 1.094563364982605,
+      "learning_rate": 1.6181181181181182e-05,
+      "loss": 1.0139,
+      "step": 6770
+    },
+    {
+      "epoch": 0.00885094568256595,
+      "grad_norm": 0.4922714829444885,
+      "learning_rate": 1.617617617617618e-05,
+      "loss": 0.1579,
+      "step": 6771
+    },
+    {
+      "epoch": 0.008852252866982222,
+      "grad_norm": 1.3835803270339966,
+      "learning_rate": 1.6171171171171172e-05,
+      "loss": 1.4666,
+      "step": 6772
+    },
+    {
+      "epoch": 0.008853560051398491,
+      "grad_norm": 1.3250312805175781,
+      "learning_rate": 1.6166166166166165e-05,
+      "loss": 1.5637,
+      "step": 6773
+    },
+    {
+      "epoch": 0.00885486723581476,
+      "grad_norm": 1.5183013677597046,
+      "learning_rate": 1.616116116116116e-05,
+      "loss": 1.2916,
+      "step": 6774
+    },
+    {
+      "epoch": 0.008856174420231032,
+      "grad_norm": 1.4419571161270142,
+      "learning_rate": 1.6156156156156157e-05,
+      "loss": 1.6664,
+      "step": 6775
+    },
+    {
+      "epoch": 0.008857481604647301,
+      "grad_norm": 1.2849586009979248,
+      "learning_rate": 1.6151151151151154e-05,
+      "loss": 1.3082,
+      "step": 6776
+    },
+    {
+      "epoch": 0.008858788789063573,
+      "grad_norm": 1.4720687866210938,
+      "learning_rate": 1.6146146146146147e-05,
+      "loss": 1.9621,
+      "step": 6777
+    },
+    {
+      "epoch": 0.008860095973479842,
+      "grad_norm": 1.5500644445419312,
+      "learning_rate": 1.6141141141141143e-05,
+      "loss": 1.2869,
+      "step": 6778
+    },
+    {
+      "epoch": 0.008861403157896113,
+      "grad_norm": 1.3936514854431152,
+      "learning_rate": 1.6136136136136136e-05,
+      "loss": 2.0669,
+      "step": 6779
+    },
+    {
+      "epoch": 0.008862710342312383,
+      "grad_norm": 0.9226558804512024,
+      "learning_rate": 1.6131131131131133e-05,
+      "loss": 0.8783,
+      "step": 6780
+    },
+    {
+      "epoch": 0.008864017526728654,
+      "grad_norm": 1.2178924083709717,
+      "learning_rate": 1.6126126126126126e-05,
+      "loss": 1.0051,
+      "step": 6781
+    },
+    {
+      "epoch": 0.008865324711144923,
+      "grad_norm": 1.7843505144119263,
+      "learning_rate": 1.6121121121121122e-05,
+      "loss": 1.2526,
+      "step": 6782
+    },
+    {
+      "epoch": 0.008866631895561194,
+      "grad_norm": 2.7131521701812744,
+      "learning_rate": 1.6116116116116118e-05,
+      "loss": 2.5131,
+      "step": 6783
+    },
+    {
+      "epoch": 0.008867939079977464,
+      "grad_norm": 1.4323841333389282,
+      "learning_rate": 1.6111111111111115e-05,
+      "loss": 2.3086,
+      "step": 6784
+    },
+    {
+      "epoch": 0.008869246264393735,
+      "grad_norm": 1.7348802089691162,
+      "learning_rate": 1.6106106106106104e-05,
+      "loss": 1.6892,
+      "step": 6785
+    },
+    {
+      "epoch": 0.008870553448810005,
+      "grad_norm": 1.5218486785888672,
+      "learning_rate": 1.61011011011011e-05,
+      "loss": 0.9978,
+      "step": 6786
+    },
+    {
+      "epoch": 0.008871860633226276,
+      "grad_norm": 1.3018760681152344,
+      "learning_rate": 1.6096096096096097e-05,
+      "loss": 1.0234,
+      "step": 6787
+    },
+    {
+      "epoch": 0.008873167817642545,
+      "grad_norm": 1.6150978803634644,
+      "learning_rate": 1.6091091091091093e-05,
+      "loss": 1.6972,
+      "step": 6788
+    },
+    {
+      "epoch": 0.008874475002058815,
+      "grad_norm": 1.4947410821914673,
+      "learning_rate": 1.6086086086086086e-05,
+      "loss": 1.6362,
+      "step": 6789
+    },
+    {
+      "epoch": 0.008875782186475086,
+      "grad_norm": 1.5297735929489136,
+      "learning_rate": 1.6081081081081083e-05,
+      "loss": 1.3672,
+      "step": 6790
+    },
+    {
+      "epoch": 0.008877089370891355,
+      "grad_norm": 1.1784470081329346,
+      "learning_rate": 1.6076076076076076e-05,
+      "loss": 1.3688,
+      "step": 6791
+    },
+    {
+      "epoch": 0.008878396555307626,
+      "grad_norm": 1.5736511945724487,
+      "learning_rate": 1.6071071071071072e-05,
+      "loss": 1.8816,
+      "step": 6792
+    },
+    {
+      "epoch": 0.008879703739723896,
+      "grad_norm": 1.4543061256408691,
+      "learning_rate": 1.6066066066066065e-05,
+      "loss": 1.6561,
+      "step": 6793
+    },
+    {
+      "epoch": 0.008881010924140167,
+      "grad_norm": 1.3913172483444214,
+      "learning_rate": 1.606106106106106e-05,
+      "loss": 1.8473,
+      "step": 6794
+    },
+    {
+      "epoch": 0.008882318108556437,
+      "grad_norm": 1.7435532808303833,
+      "learning_rate": 1.6056056056056058e-05,
+      "loss": 1.1287,
+      "step": 6795
+    },
+    {
+      "epoch": 0.008883625292972708,
+      "grad_norm": 0.9122709035873413,
+      "learning_rate": 1.6051051051051054e-05,
+      "loss": 0.7665,
+      "step": 6796
+    },
+    {
+      "epoch": 0.008884932477388977,
+      "grad_norm": 1.1911466121673584,
+      "learning_rate": 1.6046046046046047e-05,
+      "loss": 1.3298,
+      "step": 6797
+    },
+    {
+      "epoch": 0.008886239661805248,
+      "grad_norm": 1.077325463294983,
+      "learning_rate": 1.604104104104104e-05,
+      "loss": 0.5402,
+      "step": 6798
+    },
+    {
+      "epoch": 0.008887546846221518,
+      "grad_norm": 1.4121025800704956,
+      "learning_rate": 1.6036036036036036e-05,
+      "loss": 1.1836,
+      "step": 6799
+    },
+    {
+      "epoch": 0.008888854030637789,
+      "grad_norm": 1.3093801736831665,
+      "learning_rate": 1.6031031031031033e-05,
+      "loss": 1.3398,
+      "step": 6800
+    },
+    {
+      "epoch": 0.008890161215054058,
+      "grad_norm": 1.4665488004684448,
+      "learning_rate": 1.602602602602603e-05,
+      "loss": 1.6369,
+      "step": 6801
+    },
+    {
+      "epoch": 0.00889146839947033,
+      "grad_norm": 1.2597053050994873,
+      "learning_rate": 1.6021021021021022e-05,
+      "loss": 1.2842,
+      "step": 6802
+    },
+    {
+      "epoch": 0.008892775583886599,
+      "grad_norm": 1.4754912853240967,
+      "learning_rate": 1.6016016016016015e-05,
+      "loss": 1.8592,
+      "step": 6803
+    },
+    {
+      "epoch": 0.008894082768302869,
+      "grad_norm": 1.3846949338912964,
+      "learning_rate": 1.601101101101101e-05,
+      "loss": 1.1707,
+      "step": 6804
+    },
+    {
+      "epoch": 0.00889538995271914,
+      "grad_norm": 1.4617149829864502,
+      "learning_rate": 1.6006006006006008e-05,
+      "loss": 1.4369,
+      "step": 6805
+    },
+    {
+      "epoch": 0.00889669713713541,
+      "grad_norm": 1.3111882209777832,
+      "learning_rate": 1.6001001001001e-05,
+      "loss": 1.5242,
+      "step": 6806
+    },
+    {
+      "epoch": 0.00889800432155168,
+      "grad_norm": 1.281277060508728,
+      "learning_rate": 1.5995995995995997e-05,
+      "loss": 0.9764,
+      "step": 6807
+    },
+    {
+      "epoch": 0.00889931150596795,
+      "grad_norm": 1.646628499031067,
+      "learning_rate": 1.5990990990990993e-05,
+      "loss": 1.7197,
+      "step": 6808
+    },
+    {
+      "epoch": 0.008900618690384221,
+      "grad_norm": 1.3368743658065796,
+      "learning_rate": 1.5985985985985986e-05,
+      "loss": 2.1341,
+      "step": 6809
+    },
+    {
+      "epoch": 0.00890192587480049,
+      "grad_norm": 1.3595813512802124,
+      "learning_rate": 1.598098098098098e-05,
+      "loss": 1.686,
+      "step": 6810
+    },
+    {
+      "epoch": 0.008903233059216762,
+      "grad_norm": 1.5778623819351196,
+      "learning_rate": 1.5975975975975976e-05,
+      "loss": 1.6714,
+      "step": 6811
+    },
+    {
+      "epoch": 0.008904540243633031,
+      "grad_norm": 2.0751140117645264,
+      "learning_rate": 1.5970970970970972e-05,
+      "loss": 0.8999,
+      "step": 6812
+    },
+    {
+      "epoch": 0.008905847428049302,
+      "grad_norm": 1.3845704793930054,
+      "learning_rate": 1.596596596596597e-05,
+      "loss": 1.1203,
+      "step": 6813
+    },
+    {
+      "epoch": 0.008907154612465572,
+      "grad_norm": 1.5723121166229248,
+      "learning_rate": 1.596096096096096e-05,
+      "loss": 2.1886,
+      "step": 6814
+    },
+    {
+      "epoch": 0.008908461796881843,
+      "grad_norm": 1.3139495849609375,
+      "learning_rate": 1.5955955955955958e-05,
+      "loss": 1.4962,
+      "step": 6815
+    },
+    {
+      "epoch": 0.008909768981298112,
+      "grad_norm": 1.647049903869629,
+      "learning_rate": 1.595095095095095e-05,
+      "loss": 1.544,
+      "step": 6816
+    },
+    {
+      "epoch": 0.008911076165714384,
+      "grad_norm": 1.5831794738769531,
+      "learning_rate": 1.5945945945945947e-05,
+      "loss": 1.3804,
+      "step": 6817
+    },
+    {
+      "epoch": 0.008912383350130653,
+      "grad_norm": 1.6204311847686768,
+      "learning_rate": 1.594094094094094e-05,
+      "loss": 1.9956,
+      "step": 6818
+    },
+    {
+      "epoch": 0.008913690534546923,
+      "grad_norm": 1.4886915683746338,
+      "learning_rate": 1.5935935935935936e-05,
+      "loss": 2.8316,
+      "step": 6819
+    },
+    {
+      "epoch": 0.008914997718963194,
+      "grad_norm": 1.863429069519043,
+      "learning_rate": 1.5930930930930933e-05,
+      "loss": 1.5641,
+      "step": 6820
+    },
+    {
+      "epoch": 0.008916304903379463,
+      "grad_norm": 1.6092323064804077,
+      "learning_rate": 1.5925925925925926e-05,
+      "loss": 1.4477,
+      "step": 6821
+    },
+    {
+      "epoch": 0.008917612087795734,
+      "grad_norm": 1.6681382656097412,
+      "learning_rate": 1.592092092092092e-05,
+      "loss": 2.5486,
+      "step": 6822
+    },
+    {
+      "epoch": 0.008918919272212004,
+      "grad_norm": 1.1999106407165527,
+      "learning_rate": 1.5915915915915915e-05,
+      "loss": 1.0148,
+      "step": 6823
+    },
+    {
+      "epoch": 0.008920226456628275,
+      "grad_norm": 1.6669408082962036,
+      "learning_rate": 1.591091091091091e-05,
+      "loss": 2.1586,
+      "step": 6824
+    },
+    {
+      "epoch": 0.008921533641044544,
+      "grad_norm": 1.1745102405548096,
+      "learning_rate": 1.5905905905905908e-05,
+      "loss": 1.2769,
+      "step": 6825
+    },
+    {
+      "epoch": 0.008922840825460816,
+      "grad_norm": 1.666249394416809,
+      "learning_rate": 1.5900900900900904e-05,
+      "loss": 2.2248,
+      "step": 6826
+    },
+    {
+      "epoch": 0.008924148009877085,
+      "grad_norm": 1.5616997480392456,
+      "learning_rate": 1.5895895895895897e-05,
+      "loss": 1.7144,
+      "step": 6827
+    },
+    {
+      "epoch": 0.008925455194293356,
+      "grad_norm": 1.3135626316070557,
+      "learning_rate": 1.589089089089089e-05,
+      "loss": 1.7722,
+      "step": 6828
+    },
+    {
+      "epoch": 0.008926762378709626,
+      "grad_norm": 1.7937337160110474,
+      "learning_rate": 1.5885885885885887e-05,
+      "loss": 1.4502,
+      "step": 6829
+    },
+    {
+      "epoch": 0.008928069563125897,
+      "grad_norm": 1.2927981615066528,
+      "learning_rate": 1.5880880880880883e-05,
+      "loss": 1.278,
+      "step": 6830
+    },
+    {
+      "epoch": 0.008929376747542166,
+      "grad_norm": 1.6234650611877441,
+      "learning_rate": 1.5875875875875876e-05,
+      "loss": 1.639,
+      "step": 6831
+    },
+    {
+      "epoch": 0.008930683931958438,
+      "grad_norm": 1.525168776512146,
+      "learning_rate": 1.5870870870870872e-05,
+      "loss": 2.1852,
+      "step": 6832
+    },
+    {
+      "epoch": 0.008931991116374707,
+      "grad_norm": 1.447853922843933,
+      "learning_rate": 1.586586586586587e-05,
+      "loss": 0.9717,
+      "step": 6833
+    },
+    {
+      "epoch": 0.008933298300790976,
+      "grad_norm": 1.4563162326812744,
+      "learning_rate": 1.586086086086086e-05,
+      "loss": 1.2754,
+      "step": 6834
+    },
+    {
+      "epoch": 0.008934605485207248,
+      "grad_norm": 1.8562885522842407,
+      "learning_rate": 1.5855855855855855e-05,
+      "loss": 1.8731,
+      "step": 6835
+    },
+    {
+      "epoch": 0.008935912669623517,
+      "grad_norm": 1.507948398590088,
+      "learning_rate": 1.585085085085085e-05,
+      "loss": 3.4955,
+      "step": 6836
+    },
+    {
+      "epoch": 0.008937219854039788,
+      "grad_norm": 1.3423963785171509,
+      "learning_rate": 1.5845845845845847e-05,
+      "loss": 1.9035,
+      "step": 6837
+    },
+    {
+      "epoch": 0.008938527038456058,
+      "grad_norm": 1.2618061304092407,
+      "learning_rate": 1.5840840840840844e-05,
+      "loss": 1.9507,
+      "step": 6838
+    },
+    {
+      "epoch": 0.008939834222872329,
+      "grad_norm": 1.1388509273529053,
+      "learning_rate": 1.5835835835835837e-05,
+      "loss": 0.8583,
+      "step": 6839
+    },
+    {
+      "epoch": 0.008941141407288598,
+      "grad_norm": 1.1081013679504395,
+      "learning_rate": 1.583083083083083e-05,
+      "loss": 0.9437,
+      "step": 6840
+    },
+    {
+      "epoch": 0.00894244859170487,
+      "grad_norm": 2.1595849990844727,
+      "learning_rate": 1.5825825825825826e-05,
+      "loss": 1.5249,
+      "step": 6841
+    },
+    {
+      "epoch": 0.008943755776121139,
+      "grad_norm": 1.2565821409225464,
+      "learning_rate": 1.5820820820820822e-05,
+      "loss": 1.1429,
+      "step": 6842
+    },
+    {
+      "epoch": 0.00894506296053741,
+      "grad_norm": 1.2837661504745483,
+      "learning_rate": 1.5815815815815815e-05,
+      "loss": 1.4289,
+      "step": 6843
+    },
+    {
+      "epoch": 0.00894637014495368,
+      "grad_norm": 1.4385298490524292,
+      "learning_rate": 1.581081081081081e-05,
+      "loss": 2.4675,
+      "step": 6844
+    },
+    {
+      "epoch": 0.00894767732936995,
+      "grad_norm": 1.4806745052337646,
+      "learning_rate": 1.5805805805805808e-05,
+      "loss": 1.7846,
+      "step": 6845
+    },
+    {
+      "epoch": 0.00894898451378622,
+      "grad_norm": 1.7229747772216797,
+      "learning_rate": 1.58008008008008e-05,
+      "loss": 0.8443,
+      "step": 6846
+    },
+    {
+      "epoch": 0.008950291698202491,
+      "grad_norm": 1.4250985383987427,
+      "learning_rate": 1.5795795795795794e-05,
+      "loss": 1.5066,
+      "step": 6847
+    },
+    {
+      "epoch": 0.008951598882618761,
+      "grad_norm": 1.9392746686935425,
+      "learning_rate": 1.579079079079079e-05,
+      "loss": 1.9983,
+      "step": 6848
+    },
+    {
+      "epoch": 0.008952906067035032,
+      "grad_norm": 1.6741257905960083,
+      "learning_rate": 1.5785785785785787e-05,
+      "loss": 1.1066,
+      "step": 6849
+    },
+    {
+      "epoch": 0.008954213251451302,
+      "grad_norm": 1.3093879222869873,
+      "learning_rate": 1.5780780780780783e-05,
+      "loss": 1.848,
+      "step": 6850
+    },
+    {
+      "epoch": 0.008955520435867571,
+      "grad_norm": 1.6284639835357666,
+      "learning_rate": 1.5775775775775776e-05,
+      "loss": 1.5521,
+      "step": 6851
+    },
+    {
+      "epoch": 0.008956827620283842,
+      "grad_norm": 1.5501229763031006,
+      "learning_rate": 1.5770770770770772e-05,
+      "loss": 2.3222,
+      "step": 6852
+    },
+    {
+      "epoch": 0.008958134804700112,
+      "grad_norm": 1.1320677995681763,
+      "learning_rate": 1.5765765765765765e-05,
+      "loss": 0.8332,
+      "step": 6853
+    },
+    {
+      "epoch": 0.008959441989116383,
+      "grad_norm": 1.807198166847229,
+      "learning_rate": 1.576076076076076e-05,
+      "loss": 1.9239,
+      "step": 6854
+    },
+    {
+      "epoch": 0.008960749173532652,
+      "grad_norm": 1.3642265796661377,
+      "learning_rate": 1.5755755755755758e-05,
+      "loss": 1.511,
+      "step": 6855
+    },
+    {
+      "epoch": 0.008962056357948923,
+      "grad_norm": 1.428087592124939,
+      "learning_rate": 1.575075075075075e-05,
+      "loss": 1.4012,
+      "step": 6856
+    },
+    {
+      "epoch": 0.008963363542365193,
+      "grad_norm": 1.0498007535934448,
+      "learning_rate": 1.5745745745745747e-05,
+      "loss": 0.8144,
+      "step": 6857
+    },
+    {
+      "epoch": 0.008964670726781464,
+      "grad_norm": 1.2841564416885376,
+      "learning_rate": 1.574074074074074e-05,
+      "loss": 1.5529,
+      "step": 6858
+    },
+    {
+      "epoch": 0.008965977911197734,
+      "grad_norm": 1.516652226448059,
+      "learning_rate": 1.5735735735735737e-05,
+      "loss": 1.7345,
+      "step": 6859
+    },
+    {
+      "epoch": 0.008967285095614005,
+      "grad_norm": 1.1196974515914917,
+      "learning_rate": 1.573073073073073e-05,
+      "loss": 0.8445,
+      "step": 6860
+    },
+    {
+      "epoch": 0.008968592280030274,
+      "grad_norm": 1.2797346115112305,
+      "learning_rate": 1.5725725725725726e-05,
+      "loss": 1.7762,
+      "step": 6861
+    },
+    {
+      "epoch": 0.008969899464446545,
+      "grad_norm": 1.6984885931015015,
+      "learning_rate": 1.5720720720720722e-05,
+      "loss": 1.76,
+      "step": 6862
+    },
+    {
+      "epoch": 0.008971206648862815,
+      "grad_norm": 2.0073726177215576,
+      "learning_rate": 1.571571571571572e-05,
+      "loss": 2.1262,
+      "step": 6863
+    },
+    {
+      "epoch": 0.008972513833279086,
+      "grad_norm": 1.2387480735778809,
+      "learning_rate": 1.5710710710710712e-05,
+      "loss": 1.2458,
+      "step": 6864
+    },
+    {
+      "epoch": 0.008973821017695355,
+      "grad_norm": 1.4712715148925781,
+      "learning_rate": 1.5705705705705705e-05,
+      "loss": 2.2798,
+      "step": 6865
+    },
+    {
+      "epoch": 0.008975128202111625,
+      "grad_norm": 1.5491924285888672,
+      "learning_rate": 1.57007007007007e-05,
+      "loss": 2.5254,
+      "step": 6866
+    },
+    {
+      "epoch": 0.008976435386527896,
+      "grad_norm": 1.248387098312378,
+      "learning_rate": 1.5695695695695697e-05,
+      "loss": 1.4697,
+      "step": 6867
+    },
+    {
+      "epoch": 0.008977742570944166,
+      "grad_norm": 1.3036999702453613,
+      "learning_rate": 1.569069069069069e-05,
+      "loss": 1.0879,
+      "step": 6868
+    },
+    {
+      "epoch": 0.008979049755360437,
+      "grad_norm": 1.355079174041748,
+      "learning_rate": 1.5685685685685687e-05,
+      "loss": 1.4243,
+      "step": 6869
+    },
+    {
+      "epoch": 0.008980356939776706,
+      "grad_norm": 1.730860710144043,
+      "learning_rate": 1.5680680680680683e-05,
+      "loss": 1.4357,
+      "step": 6870
+    },
+    {
+      "epoch": 0.008981664124192977,
+      "grad_norm": 1.3638508319854736,
+      "learning_rate": 1.5675675675675676e-05,
+      "loss": 1.7296,
+      "step": 6871
+    },
+    {
+      "epoch": 0.008982971308609247,
+      "grad_norm": 1.2024645805358887,
+      "learning_rate": 1.567067067067067e-05,
+      "loss": 1.0552,
+      "step": 6872
+    },
+    {
+      "epoch": 0.008984278493025518,
+      "grad_norm": 1.6213204860687256,
+      "learning_rate": 1.5665665665665665e-05,
+      "loss": 1.3151,
+      "step": 6873
+    },
+    {
+      "epoch": 0.008985585677441787,
+      "grad_norm": 1.460486650466919,
+      "learning_rate": 1.5660660660660662e-05,
+      "loss": 2.0677,
+      "step": 6874
+    },
+    {
+      "epoch": 0.008986892861858059,
+      "grad_norm": 1.6433194875717163,
+      "learning_rate": 1.5655655655655658e-05,
+      "loss": 2.2554,
+      "step": 6875
+    },
+    {
+      "epoch": 0.008988200046274328,
+      "grad_norm": 1.028313398361206,
+      "learning_rate": 1.565065065065065e-05,
+      "loss": 0.6352,
+      "step": 6876
+    },
+    {
+      "epoch": 0.0089895072306906,
+      "grad_norm": 1.3749562501907349,
+      "learning_rate": 1.5645645645645644e-05,
+      "loss": 1.4357,
+      "step": 6877
+    },
+    {
+      "epoch": 0.008990814415106869,
+      "grad_norm": 0.9878309369087219,
+      "learning_rate": 1.564064064064064e-05,
+      "loss": 1.0413,
+      "step": 6878
+    },
+    {
+      "epoch": 0.00899212159952314,
+      "grad_norm": 1.4349502325057983,
+      "learning_rate": 1.5635635635635637e-05,
+      "loss": 1.0327,
+      "step": 6879
+    },
+    {
+      "epoch": 0.00899342878393941,
+      "grad_norm": 1.5023289918899536,
+      "learning_rate": 1.5630630630630633e-05,
+      "loss": 1.7011,
+      "step": 6880
+    },
+    {
+      "epoch": 0.008994735968355679,
+      "grad_norm": 1.1189044713974,
+      "learning_rate": 1.5625625625625626e-05,
+      "loss": 1.4741,
+      "step": 6881
+    },
+    {
+      "epoch": 0.00899604315277195,
+      "grad_norm": 1.6854981184005737,
+      "learning_rate": 1.5620620620620623e-05,
+      "loss": 2.0651,
+      "step": 6882
+    },
+    {
+      "epoch": 0.00899735033718822,
+      "grad_norm": 1.198167324066162,
+      "learning_rate": 1.5615615615615616e-05,
+      "loss": 0.9898,
+      "step": 6883
+    },
+    {
+      "epoch": 0.00899865752160449,
+      "grad_norm": 1.192941665649414,
+      "learning_rate": 1.5610610610610612e-05,
+      "loss": 1.3972,
+      "step": 6884
+    },
+    {
+      "epoch": 0.00899996470602076,
+      "grad_norm": 2.111043691635132,
+      "learning_rate": 1.5605605605605605e-05,
+      "loss": 1.8409,
+      "step": 6885
+    },
+    {
+      "epoch": 0.009001271890437031,
+      "grad_norm": 1.6118987798690796,
+      "learning_rate": 1.56006006006006e-05,
+      "loss": 1.4939,
+      "step": 6886
+    },
+    {
+      "epoch": 0.0090025790748533,
+      "grad_norm": 1.0801957845687866,
+      "learning_rate": 1.5595595595595598e-05,
+      "loss": 0.8558,
+      "step": 6887
+    },
+    {
+      "epoch": 0.009003886259269572,
+      "grad_norm": 1.729120135307312,
+      "learning_rate": 1.5590590590590594e-05,
+      "loss": 2.2291,
+      "step": 6888
+    },
+    {
+      "epoch": 0.009005193443685841,
+      "grad_norm": 1.4175816774368286,
+      "learning_rate": 1.5585585585585587e-05,
+      "loss": 0.9877,
+      "step": 6889
+    },
+    {
+      "epoch": 0.009006500628102113,
+      "grad_norm": 1.558062195777893,
+      "learning_rate": 1.558058058058058e-05,
+      "loss": 1.4814,
+      "step": 6890
+    },
+    {
+      "epoch": 0.009007807812518382,
+      "grad_norm": 1.7434206008911133,
+      "learning_rate": 1.5575575575575576e-05,
+      "loss": 1.768,
+      "step": 6891
+    },
+    {
+      "epoch": 0.009009114996934653,
+      "grad_norm": 1.461486577987671,
+      "learning_rate": 1.5570570570570573e-05,
+      "loss": 1.3372,
+      "step": 6892
+    },
+    {
+      "epoch": 0.009010422181350923,
+      "grad_norm": 1.2614681720733643,
+      "learning_rate": 1.5565565565565566e-05,
+      "loss": 2.0732,
+      "step": 6893
+    },
+    {
+      "epoch": 0.009011729365767194,
+      "grad_norm": 1.6604398488998413,
+      "learning_rate": 1.5560560560560562e-05,
+      "loss": 2.4065,
+      "step": 6894
+    },
+    {
+      "epoch": 0.009013036550183463,
+      "grad_norm": 1.6958496570587158,
+      "learning_rate": 1.5555555555555555e-05,
+      "loss": 1.9597,
+      "step": 6895
+    },
+    {
+      "epoch": 0.009014343734599733,
+      "grad_norm": 1.3080248832702637,
+      "learning_rate": 1.555055055055055e-05,
+      "loss": 2.3835,
+      "step": 6896
+    },
+    {
+      "epoch": 0.009015650919016004,
+      "grad_norm": 1.3903577327728271,
+      "learning_rate": 1.5545545545545544e-05,
+      "loss": 1.016,
+      "step": 6897
+    },
+    {
+      "epoch": 0.009016958103432273,
+      "grad_norm": 1.2525666952133179,
+      "learning_rate": 1.554054054054054e-05,
+      "loss": 0.8794,
+      "step": 6898
+    },
+    {
+      "epoch": 0.009018265287848545,
+      "grad_norm": 1.6158853769302368,
+      "learning_rate": 1.5535535535535537e-05,
+      "loss": 1.3077,
+      "step": 6899
+    },
+    {
+      "epoch": 0.009019572472264814,
+      "grad_norm": 1.5927802324295044,
+      "learning_rate": 1.5530530530530533e-05,
+      "loss": 1.8207,
+      "step": 6900
+    },
+    {
+      "epoch": 0.009020879656681085,
+      "grad_norm": 1.418837308883667,
+      "learning_rate": 1.5525525525525526e-05,
+      "loss": 1.1868,
+      "step": 6901
+    },
+    {
+      "epoch": 0.009022186841097355,
+      "grad_norm": 1.3699965476989746,
+      "learning_rate": 1.552052052052052e-05,
+      "loss": 1.5448,
+      "step": 6902
+    },
+    {
+      "epoch": 0.009023494025513626,
+      "grad_norm": 1.273872971534729,
+      "learning_rate": 1.5515515515515516e-05,
+      "loss": 1.0333,
+      "step": 6903
+    },
+    {
+      "epoch": 0.009024801209929895,
+      "grad_norm": 1.4030042886734009,
+      "learning_rate": 1.5510510510510512e-05,
+      "loss": 1.4767,
+      "step": 6904
+    },
+    {
+      "epoch": 0.009026108394346167,
+      "grad_norm": 1.2879765033721924,
+      "learning_rate": 1.550550550550551e-05,
+      "loss": 1.2128,
+      "step": 6905
+    },
+    {
+      "epoch": 0.009027415578762436,
+      "grad_norm": 1.2809842824935913,
+      "learning_rate": 1.55005005005005e-05,
+      "loss": 0.9326,
+      "step": 6906
+    },
+    {
+      "epoch": 0.009028722763178707,
+      "grad_norm": 1.3342094421386719,
+      "learning_rate": 1.5495495495495498e-05,
+      "loss": 1.8946,
+      "step": 6907
+    },
+    {
+      "epoch": 0.009030029947594977,
+      "grad_norm": 1.425167202949524,
+      "learning_rate": 1.549049049049049e-05,
+      "loss": 1.9836,
+      "step": 6908
+    },
+    {
+      "epoch": 0.009031337132011248,
+      "grad_norm": 1.3269357681274414,
+      "learning_rate": 1.5485485485485487e-05,
+      "loss": 0.863,
+      "step": 6909
+    },
+    {
+      "epoch": 0.009032644316427517,
+      "grad_norm": 1.4251599311828613,
+      "learning_rate": 1.548048048048048e-05,
+      "loss": 1.5567,
+      "step": 6910
+    },
+    {
+      "epoch": 0.009033951500843787,
+      "grad_norm": 1.2442957162857056,
+      "learning_rate": 1.5475475475475476e-05,
+      "loss": 1.5867,
+      "step": 6911
+    },
+    {
+      "epoch": 0.009035258685260058,
+      "grad_norm": 1.3447675704956055,
+      "learning_rate": 1.5470470470470473e-05,
+      "loss": 0.3868,
+      "step": 6912
+    },
+    {
+      "epoch": 0.009036565869676327,
+      "grad_norm": 1.1141862869262695,
+      "learning_rate": 1.5465465465465466e-05,
+      "loss": 0.945,
+      "step": 6913
+    },
+    {
+      "epoch": 0.009037873054092599,
+      "grad_norm": 1.5237133502960205,
+      "learning_rate": 1.546046046046046e-05,
+      "loss": 1.7505,
+      "step": 6914
+    },
+    {
+      "epoch": 0.009039180238508868,
+      "grad_norm": 1.601372480392456,
+      "learning_rate": 1.5455455455455455e-05,
+      "loss": 1.7019,
+      "step": 6915
+    },
+    {
+      "epoch": 0.00904048742292514,
+      "grad_norm": 1.303246021270752,
+      "learning_rate": 1.545045045045045e-05,
+      "loss": 1.2043,
+      "step": 6916
+    },
+    {
+      "epoch": 0.009041794607341409,
+      "grad_norm": 2.0998764038085938,
+      "learning_rate": 1.5445445445445448e-05,
+      "loss": 1.4148,
+      "step": 6917
+    },
+    {
+      "epoch": 0.00904310179175768,
+      "grad_norm": 1.6047621965408325,
+      "learning_rate": 1.544044044044044e-05,
+      "loss": 1.4668,
+      "step": 6918
+    },
+    {
+      "epoch": 0.00904440897617395,
+      "grad_norm": 1.1133348941802979,
+      "learning_rate": 1.5435435435435437e-05,
+      "loss": 0.5764,
+      "step": 6919
+    },
+    {
+      "epoch": 0.00904571616059022,
+      "grad_norm": 1.4619442224502563,
+      "learning_rate": 1.543043043043043e-05,
+      "loss": 1.3573,
+      "step": 6920
+    },
+    {
+      "epoch": 0.00904702334500649,
+      "grad_norm": 1.5784854888916016,
+      "learning_rate": 1.5425425425425426e-05,
+      "loss": 1.7361,
+      "step": 6921
+    },
+    {
+      "epoch": 0.009048330529422761,
+      "grad_norm": 1.4800266027450562,
+      "learning_rate": 1.542042042042042e-05,
+      "loss": 0.7636,
+      "step": 6922
+    },
+    {
+      "epoch": 0.00904963771383903,
+      "grad_norm": 1.5391526222229004,
+      "learning_rate": 1.5415415415415416e-05,
+      "loss": 2.421,
+      "step": 6923
+    },
+    {
+      "epoch": 0.009050944898255302,
+      "grad_norm": 1.3793609142303467,
+      "learning_rate": 1.5410410410410412e-05,
+      "loss": 1.8622,
+      "step": 6924
+    },
+    {
+      "epoch": 0.009052252082671571,
+      "grad_norm": 1.2066377401351929,
+      "learning_rate": 1.540540540540541e-05,
+      "loss": 0.6673,
+      "step": 6925
+    },
+    {
+      "epoch": 0.00905355926708784,
+      "grad_norm": 1.436437964439392,
+      "learning_rate": 1.54004004004004e-05,
+      "loss": 1.7219,
+      "step": 6926
+    },
+    {
+      "epoch": 0.009054866451504112,
+      "grad_norm": 1.615725040435791,
+      "learning_rate": 1.5395395395395394e-05,
+      "loss": 1.8501,
+      "step": 6927
+    },
+    {
+      "epoch": 0.009056173635920381,
+      "grad_norm": 1.660314917564392,
+      "learning_rate": 1.539039039039039e-05,
+      "loss": 1.7123,
+      "step": 6928
+    },
+    {
+      "epoch": 0.009057480820336652,
+      "grad_norm": 1.421047568321228,
+      "learning_rate": 1.5385385385385387e-05,
+      "loss": 2.0296,
+      "step": 6929
+    },
+    {
+      "epoch": 0.009058788004752922,
+      "grad_norm": 1.3448107242584229,
+      "learning_rate": 1.538038038038038e-05,
+      "loss": 2.1962,
+      "step": 6930
+    },
+    {
+      "epoch": 0.009060095189169193,
+      "grad_norm": 1.6655653715133667,
+      "learning_rate": 1.5375375375375377e-05,
+      "loss": 1.0826,
+      "step": 6931
+    },
+    {
+      "epoch": 0.009061402373585463,
+      "grad_norm": 1.1854863166809082,
+      "learning_rate": 1.537037037037037e-05,
+      "loss": 1.0707,
+      "step": 6932
+    },
+    {
+      "epoch": 0.009062709558001734,
+      "grad_norm": 1.8032697439193726,
+      "learning_rate": 1.5365365365365366e-05,
+      "loss": 1.5472,
+      "step": 6933
+    },
+    {
+      "epoch": 0.009064016742418003,
+      "grad_norm": 1.7158112525939941,
+      "learning_rate": 1.5360360360360362e-05,
+      "loss": 1.9742,
+      "step": 6934
+    },
+    {
+      "epoch": 0.009065323926834274,
+      "grad_norm": 1.370894432067871,
+      "learning_rate": 1.5355355355355355e-05,
+      "loss": 1.9454,
+      "step": 6935
+    },
+    {
+      "epoch": 0.009066631111250544,
+      "grad_norm": 1.3220961093902588,
+      "learning_rate": 1.535035035035035e-05,
+      "loss": 1.3994,
+      "step": 6936
+    },
+    {
+      "epoch": 0.009067938295666815,
+      "grad_norm": 1.7119014263153076,
+      "learning_rate": 1.5345345345345348e-05,
+      "loss": 1.8111,
+      "step": 6937
+    },
+    {
+      "epoch": 0.009069245480083084,
+      "grad_norm": 1.4147604703903198,
+      "learning_rate": 1.534034034034034e-05,
+      "loss": 1.09,
+      "step": 6938
+    },
+    {
+      "epoch": 0.009070552664499356,
+      "grad_norm": 1.3791738748550415,
+      "learning_rate": 1.5335335335335334e-05,
+      "loss": 1.5507,
+      "step": 6939
+    },
+    {
+      "epoch": 0.009071859848915625,
+      "grad_norm": 1.1774777173995972,
+      "learning_rate": 1.533033033033033e-05,
+      "loss": 1.5583,
+      "step": 6940
+    },
+    {
+      "epoch": 0.009073167033331895,
+      "grad_norm": 1.4745421409606934,
+      "learning_rate": 1.5325325325325327e-05,
+      "loss": 1.4288,
+      "step": 6941
+    },
+    {
+      "epoch": 0.009074474217748166,
+      "grad_norm": 1.8062962293624878,
+      "learning_rate": 1.5320320320320323e-05,
+      "loss": 1.6825,
+      "step": 6942
+    },
+    {
+      "epoch": 0.009075781402164435,
+      "grad_norm": 1.8964476585388184,
+      "learning_rate": 1.5315315315315316e-05,
+      "loss": 1.8646,
+      "step": 6943
+    },
+    {
+      "epoch": 0.009077088586580706,
+      "grad_norm": 1.5000170469284058,
+      "learning_rate": 1.5310310310310312e-05,
+      "loss": 1.8571,
+      "step": 6944
+    },
+    {
+      "epoch": 0.009078395770996976,
+      "grad_norm": 1.8876135349273682,
+      "learning_rate": 1.5305305305305305e-05,
+      "loss": 2.2926,
+      "step": 6945
+    },
+    {
+      "epoch": 0.009079702955413247,
+      "grad_norm": 1.2169692516326904,
+      "learning_rate": 1.53003003003003e-05,
+      "loss": 1.469,
+      "step": 6946
+    },
+    {
+      "epoch": 0.009081010139829516,
+      "grad_norm": 0.7966275215148926,
+      "learning_rate": 1.5295295295295295e-05,
+      "loss": 0.419,
+      "step": 6947
+    },
+    {
+      "epoch": 0.009082317324245788,
+      "grad_norm": 1.473838210105896,
+      "learning_rate": 1.529029029029029e-05,
+      "loss": 1.5583,
+      "step": 6948
+    },
+    {
+      "epoch": 0.009083624508662057,
+      "grad_norm": 2.4029386043548584,
+      "learning_rate": 1.5285285285285287e-05,
+      "loss": 2.8811,
+      "step": 6949
+    },
+    {
+      "epoch": 0.009084931693078328,
+      "grad_norm": 1.5876059532165527,
+      "learning_rate": 1.528028028028028e-05,
+      "loss": 1.8321,
+      "step": 6950
+    },
+    {
+      "epoch": 0.009086238877494598,
+      "grad_norm": 1.2058485746383667,
+      "learning_rate": 1.5275275275275273e-05,
+      "loss": 0.6278,
+      "step": 6951
+    },
+    {
+      "epoch": 0.009087546061910869,
+      "grad_norm": 1.4080345630645752,
+      "learning_rate": 1.527027027027027e-05,
+      "loss": 1.3944,
+      "step": 6952
+    },
+    {
+      "epoch": 0.009088853246327138,
+      "grad_norm": 1.5900695323944092,
+      "learning_rate": 1.5265265265265266e-05,
+      "loss": 1.7895,
+      "step": 6953
+    },
+    {
+      "epoch": 0.00909016043074341,
+      "grad_norm": 1.6412582397460938,
+      "learning_rate": 1.5260260260260262e-05,
+      "loss": 2.3519,
+      "step": 6954
+    },
+    {
+      "epoch": 0.009091467615159679,
+      "grad_norm": 1.66777503490448,
+      "learning_rate": 1.5255255255255255e-05,
+      "loss": 2.1288,
+      "step": 6955
+    },
+    {
+      "epoch": 0.00909277479957595,
+      "grad_norm": 1.5893572568893433,
+      "learning_rate": 1.525025025025025e-05,
+      "loss": 1.7549,
+      "step": 6956
+    },
+    {
+      "epoch": 0.00909408198399222,
+      "grad_norm": 2.513761520385742,
+      "learning_rate": 1.5245245245245246e-05,
+      "loss": 1.9058,
+      "step": 6957
+    },
+    {
+      "epoch": 0.00909538916840849,
+      "grad_norm": 1.4900075197219849,
+      "learning_rate": 1.5240240240240241e-05,
+      "loss": 1.9915,
+      "step": 6958
+    },
+    {
+      "epoch": 0.00909669635282476,
+      "grad_norm": 1.1877477169036865,
+      "learning_rate": 1.5235235235235237e-05,
+      "loss": 0.6804,
+      "step": 6959
+    },
+    {
+      "epoch": 0.00909800353724103,
+      "grad_norm": 1.3640377521514893,
+      "learning_rate": 1.523023023023023e-05,
+      "loss": 1.6654,
+      "step": 6960
+    },
+    {
+      "epoch": 0.009099310721657301,
+      "grad_norm": 1.516669750213623,
+      "learning_rate": 1.5225225225225225e-05,
+      "loss": 1.6182,
+      "step": 6961
+    },
+    {
+      "epoch": 0.00910061790607357,
+      "grad_norm": 1.5814027786254883,
+      "learning_rate": 1.5220220220220221e-05,
+      "loss": 1.7734,
+      "step": 6962
+    },
+    {
+      "epoch": 0.009101925090489842,
+      "grad_norm": 1.1578916311264038,
+      "learning_rate": 1.5215215215215218e-05,
+      "loss": 0.7944,
+      "step": 6963
+    },
+    {
+      "epoch": 0.009103232274906111,
+      "grad_norm": 2.382713794708252,
+      "learning_rate": 1.521021021021021e-05,
+      "loss": 1.9216,
+      "step": 6964
+    },
+    {
+      "epoch": 0.009104539459322382,
+      "grad_norm": 1.546707272529602,
+      "learning_rate": 1.5205205205205205e-05,
+      "loss": 2.1648,
+      "step": 6965
+    },
+    {
+      "epoch": 0.009105846643738652,
+      "grad_norm": 1.3666800260543823,
+      "learning_rate": 1.5200200200200202e-05,
+      "loss": 1.0959,
+      "step": 6966
+    },
+    {
+      "epoch": 0.009107153828154923,
+      "grad_norm": 1.5071203708648682,
+      "learning_rate": 1.5195195195195196e-05,
+      "loss": 1.5054,
+      "step": 6967
+    },
+    {
+      "epoch": 0.009108461012571192,
+      "grad_norm": 1.2945618629455566,
+      "learning_rate": 1.519019019019019e-05,
+      "loss": 1.3637,
+      "step": 6968
+    },
+    {
+      "epoch": 0.009109768196987464,
+      "grad_norm": 1.5005626678466797,
+      "learning_rate": 1.5185185185185186e-05,
+      "loss": 1.3716,
+      "step": 6969
+    },
+    {
+      "epoch": 0.009111075381403733,
+      "grad_norm": 1.2758702039718628,
+      "learning_rate": 1.518018018018018e-05,
+      "loss": 1.4908,
+      "step": 6970
+    },
+    {
+      "epoch": 0.009112382565820004,
+      "grad_norm": 1.4560043811798096,
+      "learning_rate": 1.5175175175175177e-05,
+      "loss": 1.7507,
+      "step": 6971
+    },
+    {
+      "epoch": 0.009113689750236274,
+      "grad_norm": 1.5260823965072632,
+      "learning_rate": 1.517017017017017e-05,
+      "loss": 1.5409,
+      "step": 6972
+    },
+    {
+      "epoch": 0.009114996934652543,
+      "grad_norm": 1.6010547876358032,
+      "learning_rate": 1.5165165165165166e-05,
+      "loss": 1.3669,
+      "step": 6973
+    },
+    {
+      "epoch": 0.009116304119068814,
+      "grad_norm": 1.4660228490829468,
+      "learning_rate": 1.516016016016016e-05,
+      "loss": 1.2339,
+      "step": 6974
+    },
+    {
+      "epoch": 0.009117611303485084,
+      "grad_norm": 1.4592586755752563,
+      "learning_rate": 1.5155155155155157e-05,
+      "loss": 2.0019,
+      "step": 6975
+    },
+    {
+      "epoch": 0.009118918487901355,
+      "grad_norm": 1.0084089040756226,
+      "learning_rate": 1.515015015015015e-05,
+      "loss": 0.5121,
+      "step": 6976
+    },
+    {
+      "epoch": 0.009120225672317624,
+      "grad_norm": 1.4256819486618042,
+      "learning_rate": 1.5145145145145145e-05,
+      "loss": 1.6603,
+      "step": 6977
+    },
+    {
+      "epoch": 0.009121532856733896,
+      "grad_norm": 1.4420849084854126,
+      "learning_rate": 1.5140140140140141e-05,
+      "loss": 2.1427,
+      "step": 6978
+    },
+    {
+      "epoch": 0.009122840041150165,
+      "grad_norm": 1.4300018548965454,
+      "learning_rate": 1.5135135135135136e-05,
+      "loss": 1.8687,
+      "step": 6979
+    },
+    {
+      "epoch": 0.009124147225566436,
+      "grad_norm": 1.230709195137024,
+      "learning_rate": 1.5130130130130129e-05,
+      "loss": 1.4588,
+      "step": 6980
+    },
+    {
+      "epoch": 0.009125454409982706,
+      "grad_norm": 1.2282710075378418,
+      "learning_rate": 1.5125125125125125e-05,
+      "loss": 1.2663,
+      "step": 6981
+    },
+    {
+      "epoch": 0.009126761594398977,
+      "grad_norm": 1.1365900039672852,
+      "learning_rate": 1.5120120120120122e-05,
+      "loss": 1.2248,
+      "step": 6982
+    },
+    {
+      "epoch": 0.009128068778815246,
+      "grad_norm": 1.3917306661605835,
+      "learning_rate": 1.5115115115115116e-05,
+      "loss": 1.3723,
+      "step": 6983
+    },
+    {
+      "epoch": 0.009129375963231517,
+      "grad_norm": 1.3935201168060303,
+      "learning_rate": 1.511011011011011e-05,
+      "loss": 0.9319,
+      "step": 6984
+    },
+    {
+      "epoch": 0.009130683147647787,
+      "grad_norm": 1.520301342010498,
+      "learning_rate": 1.5105105105105106e-05,
+      "loss": 0.9619,
+      "step": 6985
+    },
+    {
+      "epoch": 0.009131990332064058,
+      "grad_norm": 0.16384270787239075,
+      "learning_rate": 1.51001001001001e-05,
+      "loss": 0.0054,
+      "step": 6986
+    },
+    {
+      "epoch": 0.009133297516480328,
+      "grad_norm": 1.8263185024261475,
+      "learning_rate": 1.5095095095095097e-05,
+      "loss": 0.8117,
+      "step": 6987
+    },
+    {
+      "epoch": 0.009134604700896597,
+      "grad_norm": 1.1673376560211182,
+      "learning_rate": 1.5090090090090091e-05,
+      "loss": 0.8369,
+      "step": 6988
+    },
+    {
+      "epoch": 0.009135911885312868,
+      "grad_norm": 1.4714758396148682,
+      "learning_rate": 1.5085085085085084e-05,
+      "loss": 0.943,
+      "step": 6989
+    },
+    {
+      "epoch": 0.009137219069729138,
+      "grad_norm": 1.403088927268982,
+      "learning_rate": 1.508008008008008e-05,
+      "loss": 1.7492,
+      "step": 6990
+    },
+    {
+      "epoch": 0.009138526254145409,
+      "grad_norm": 1.5092284679412842,
+      "learning_rate": 1.5075075075075077e-05,
+      "loss": 1.8211,
+      "step": 6991
+    },
+    {
+      "epoch": 0.009139833438561678,
+      "grad_norm": 0.9949844479560852,
+      "learning_rate": 1.5070070070070072e-05,
+      "loss": 0.7308,
+      "step": 6992
+    },
+    {
+      "epoch": 0.00914114062297795,
+      "grad_norm": 1.3226675987243652,
+      "learning_rate": 1.5065065065065065e-05,
+      "loss": 1.0688,
+      "step": 6993
+    },
+    {
+      "epoch": 0.009142447807394219,
+      "grad_norm": 1.7290098667144775,
+      "learning_rate": 1.5060060060060061e-05,
+      "loss": 2.3471,
+      "step": 6994
+    },
+    {
+      "epoch": 0.00914375499181049,
+      "grad_norm": 2.1745071411132812,
+      "learning_rate": 1.5055055055055056e-05,
+      "loss": 2.1034,
+      "step": 6995
+    },
+    {
+      "epoch": 0.00914506217622676,
+      "grad_norm": 1.3611079454421997,
+      "learning_rate": 1.5050050050050052e-05,
+      "loss": 0.877,
+      "step": 6996
+    },
+    {
+      "epoch": 0.00914636936064303,
+      "grad_norm": 1.4538486003875732,
+      "learning_rate": 1.5045045045045045e-05,
+      "loss": 1.4764,
+      "step": 6997
+    },
+    {
+      "epoch": 0.0091476765450593,
+      "grad_norm": 1.391932725906372,
+      "learning_rate": 1.504004004004004e-05,
+      "loss": 1.8162,
+      "step": 6998
+    },
+    {
+      "epoch": 0.009148983729475571,
+      "grad_norm": 1.7986490726470947,
+      "learning_rate": 1.5035035035035036e-05,
+      "loss": 0.8955,
+      "step": 6999
+    },
+    {
+      "epoch": 0.00915029091389184,
+      "grad_norm": 1.566508173942566,
+      "learning_rate": 1.5030030030030032e-05,
+      "loss": 1.4927,
+      "step": 7000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 10000,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 5995556366106624.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}