diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,33759 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 4818,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00020755500207555002,
+      "grad_norm": 11.311776762787964,
+      "learning_rate": 3.9999996173438937e-07,
+      "loss": 1.8103,
+      "step": 1
+    },
+    {
+      "epoch": 0.00041511000415110004,
+      "grad_norm": 9.353463242142924,
+      "learning_rate": 3.9999984693757373e-07,
+      "loss": 1.8375,
+      "step": 2
+    },
+    {
+      "epoch": 0.0006226650062266501,
+      "grad_norm": 8.67336800381065,
+      "learning_rate": 3.999996556096019e-07,
+      "loss": 1.8509,
+      "step": 3
+    },
+    {
+      "epoch": 0.0008302200083022001,
+      "grad_norm": 8.863000043142469,
+      "learning_rate": 3.999993877505552e-07,
+      "loss": 1.8109,
+      "step": 4
+    },
+    {
+      "epoch": 0.0010377750103777502,
+      "grad_norm": 8.92142828494341,
+      "learning_rate": 3.9999904336054757e-07,
+      "loss": 1.8316,
+      "step": 5
+    },
+    {
+      "epoch": 0.0012453300124533001,
+      "grad_norm": 8.923191987375874,
+      "learning_rate": 3.999986224397254e-07,
+      "loss": 1.8512,
+      "step": 6
+    },
+    {
+      "epoch": 0.0014528850145288502,
+      "grad_norm": 10.559992136409935,
+      "learning_rate": 3.999981249882676e-07,
+      "loss": 1.8537,
+      "step": 7
+    },
+    {
+      "epoch": 0.0016604400166044002,
+      "grad_norm": 7.709226021472745,
+      "learning_rate": 3.999975510063859e-07,
+      "loss": 1.7617,
+      "step": 8
+    },
+    {
+      "epoch": 0.0018679950186799503,
+      "grad_norm": 7.602312931282997,
+      "learning_rate": 3.9999690049432405e-07,
+      "loss": 1.8194,
+      "step": 9
+    },
+    {
+      "epoch": 0.0020755500207555004,
+      "grad_norm": 7.326323018777289,
+      "learning_rate": 3.9999617345235876e-07,
+      "loss": 1.9101,
+      "step": 10
+    },
+    {
+      "epoch": 0.00228310502283105,
+      "grad_norm": 7.598220784049708,
+      "learning_rate": 3.9999536988079914e-07,
+      "loss": 1.8226,
+      "step": 11
+    },
+    {
+      "epoch": 0.0024906600249066002,
+      "grad_norm": 4.995333047136985,
+      "learning_rate": 3.9999448977998685e-07,
+      "loss": 1.7141,
+      "step": 12
+    },
+    {
+      "epoch": 0.0026982150269821504,
+      "grad_norm": 4.388908625648589,
+      "learning_rate": 3.999935331502961e-07,
+      "loss": 1.7453,
+      "step": 13
+    },
+    {
+      "epoch": 0.0029057700290577005,
+      "grad_norm": 3.655298270617133,
+      "learning_rate": 3.9999249999213364e-07,
+      "loss": 1.7598,
+      "step": 14
+    },
+    {
+      "epoch": 0.00311332503113325,
+      "grad_norm": 3.767616537060441,
+      "learning_rate": 3.999913903059387e-07,
+      "loss": 1.7484,
+      "step": 15
+    },
+    {
+      "epoch": 0.0033208800332088003,
+      "grad_norm": 4.390789350050781,
+      "learning_rate": 3.999902040921831e-07,
+      "loss": 1.7198,
+      "step": 16
+    },
+    {
+      "epoch": 0.0035284350352843504,
+      "grad_norm": 3.1143857817107343,
+      "learning_rate": 3.999889413513712e-07,
+      "loss": 1.6737,
+      "step": 17
+    },
+    {
+      "epoch": 0.0037359900373599006,
+      "grad_norm": 13.135689335510376,
+      "learning_rate": 3.999876020840398e-07,
+      "loss": 1.666,
+      "step": 18
+    },
+    {
+      "epoch": 0.00394354503943545,
+      "grad_norm": 2.7414767134828546,
+      "learning_rate": 3.9998618629075846e-07,
+      "loss": 1.6352,
+      "step": 19
+    },
+    {
+      "epoch": 0.004151100041511001,
+      "grad_norm": 2.7089328082011424,
+      "learning_rate": 3.9998469397212906e-07,
+      "loss": 1.714,
+      "step": 20
+    },
+    {
+      "epoch": 0.0043586550435865505,
+      "grad_norm": 4.299280097728705,
+      "learning_rate": 3.999831251287861e-07,
+      "loss": 1.6463,
+      "step": 21
+    },
+    {
+      "epoch": 0.0045662100456621,
+      "grad_norm": 2.9172203605477294,
+      "learning_rate": 3.999814797613966e-07,
+      "loss": 1.8079,
+      "step": 22
+    },
+    {
+      "epoch": 0.004773765047737651,
+      "grad_norm": 3.830109900005944,
+      "learning_rate": 3.999797578706602e-07,
+      "loss": 1.7054,
+      "step": 23
+    },
+    {
+      "epoch": 0.0049813200498132005,
+      "grad_norm": 6.031012034118779,
+      "learning_rate": 3.9997795945730887e-07,
+      "loss": 1.7954,
+      "step": 24
+    },
+    {
+      "epoch": 0.00518887505188875,
+      "grad_norm": 7.305625803127848,
+      "learning_rate": 3.9997608452210734e-07,
+      "loss": 1.6476,
+      "step": 25
+    },
+    {
+      "epoch": 0.005396430053964301,
+      "grad_norm": 3.199858149858899,
+      "learning_rate": 3.9997413306585275e-07,
+      "loss": 1.6505,
+      "step": 26
+    },
+    {
+      "epoch": 0.00560398505603985,
+      "grad_norm": 3.1464883611285788,
+      "learning_rate": 3.999721050893749e-07,
+      "loss": 1.6867,
+      "step": 27
+    },
+    {
+      "epoch": 0.005811540058115401,
+      "grad_norm": 3.2747257371112384,
+      "learning_rate": 3.9997000059353595e-07,
+      "loss": 1.7287,
+      "step": 28
+    },
+    {
+      "epoch": 0.006019095060190951,
+      "grad_norm": 2.970788897794001,
+      "learning_rate": 3.999678195792306e-07,
+      "loss": 1.7611,
+      "step": 29
+    },
+    {
+      "epoch": 0.0062266500622665,
+      "grad_norm": 5.449631086335021,
+      "learning_rate": 3.999655620473863e-07,
+      "loss": 1.6526,
+      "step": 30
+    },
+    {
+      "epoch": 0.006434205064342051,
+      "grad_norm": 1.6293671283027087,
+      "learning_rate": 3.999632279989628e-07,
+      "loss": 1.571,
+      "step": 31
+    },
+    {
+      "epoch": 0.006641760066417601,
+      "grad_norm": 1.6360904002070937,
+      "learning_rate": 3.9996081743495247e-07,
+      "loss": 1.6968,
+      "step": 32
+    },
+    {
+      "epoch": 0.00684931506849315,
+      "grad_norm": 1.5499493046506658,
+      "learning_rate": 3.9995833035638034e-07,
+      "loss": 1.6265,
+      "step": 33
+    },
+    {
+      "epoch": 0.007056870070568701,
+      "grad_norm": 1.627232202949144,
+      "learning_rate": 3.9995576676430375e-07,
+      "loss": 1.6673,
+      "step": 34
+    },
+    {
+      "epoch": 0.0072644250726442506,
+      "grad_norm": 1.1414425435382647,
+      "learning_rate": 3.999531266598126e-07,
+      "loss": 1.6336,
+      "step": 35
+    },
+    {
+      "epoch": 0.007471980074719801,
+      "grad_norm": 1.2406746982833923,
+      "learning_rate": 3.999504100440296e-07,
+      "loss": 1.6717,
+      "step": 36
+    },
+    {
+      "epoch": 0.007679535076795351,
+      "grad_norm": 1.1874081408656214,
+      "learning_rate": 3.9994761691810956e-07,
+      "loss": 1.6095,
+      "step": 37
+    },
+    {
+      "epoch": 0.0078870900788709,
+      "grad_norm": 1.5188163338014686,
+      "learning_rate": 3.999447472832402e-07,
+      "loss": 1.6461,
+      "step": 38
+    },
+    {
+      "epoch": 0.00809464508094645,
+      "grad_norm": 1.33634941500503,
+      "learning_rate": 3.999418011406415e-07,
+      "loss": 1.6274,
+      "step": 39
+    },
+    {
+      "epoch": 0.008302200083022002,
+      "grad_norm": 1.2824476876179023,
+      "learning_rate": 3.999387784915662e-07,
+      "loss": 1.6986,
+      "step": 40
+    },
+    {
+      "epoch": 0.008509755085097551,
+      "grad_norm": 1.1997067875355065,
+      "learning_rate": 3.9993567933729933e-07,
+      "loss": 1.6037,
+      "step": 41
+    },
+    {
+      "epoch": 0.008717310087173101,
+      "grad_norm": 1.371312130776846,
+      "learning_rate": 3.9993250367915873e-07,
+      "loss": 1.6133,
+      "step": 42
+    },
+    {
+      "epoch": 0.00892486508924865,
+      "grad_norm": 1.4148239664038342,
+      "learning_rate": 3.999292515184944e-07,
+      "loss": 1.6403,
+      "step": 43
+    },
+    {
+      "epoch": 0.0091324200913242,
+      "grad_norm": 1.0705459177594618,
+      "learning_rate": 3.9992592285668916e-07,
+      "loss": 1.5651,
+      "step": 44
+    },
+    {
+      "epoch": 0.00933997509339975,
+      "grad_norm": 1.141698744909979,
+      "learning_rate": 3.9992251769515837e-07,
+      "loss": 1.6246,
+      "step": 45
+    },
+    {
+      "epoch": 0.009547530095475302,
+      "grad_norm": 0.9921870431298467,
+      "learning_rate": 3.9991903603534964e-07,
+      "loss": 1.6022,
+      "step": 46
+    },
+    {
+      "epoch": 0.009755085097550851,
+      "grad_norm": 0.9010365335437199,
+      "learning_rate": 3.9991547787874343e-07,
+      "loss": 1.6159,
+      "step": 47
+    },
+    {
+      "epoch": 0.009962640099626401,
+      "grad_norm": 0.9693950271808062,
+      "learning_rate": 3.999118432268525e-07,
+      "loss": 1.6386,
+      "step": 48
+    },
+    {
+      "epoch": 0.01017019510170195,
+      "grad_norm": 1.0199944001952896,
+      "learning_rate": 3.9990813208122224e-07,
+      "loss": 1.6736,
+      "step": 49
+    },
+    {
+      "epoch": 0.0103777501037775,
+      "grad_norm": 1.2436120571389748,
+      "learning_rate": 3.999043444434305e-07,
+      "loss": 1.6482,
+      "step": 50
+    },
+    {
+      "epoch": 0.010585305105853052,
+      "grad_norm": 1.4153279715371405,
+      "learning_rate": 3.9990048031508765e-07,
+      "loss": 1.6776,
+      "step": 51
+    },
+    {
+      "epoch": 0.010792860107928601,
+      "grad_norm": 1.4925781849984703,
+      "learning_rate": 3.998965396978367e-07,
+      "loss": 1.6113,
+      "step": 52
+    },
+    {
+      "epoch": 0.011000415110004151,
+      "grad_norm": 1.0123982309527828,
+      "learning_rate": 3.99892522593353e-07,
+      "loss": 1.6865,
+      "step": 53
+    },
+    {
+      "epoch": 0.0112079701120797,
+      "grad_norm": 1.0459941412707985,
+      "learning_rate": 3.998884290033446e-07,
+      "loss": 1.6034,
+      "step": 54
+    },
+    {
+      "epoch": 0.01141552511415525,
+      "grad_norm": 0.9406488418358278,
+      "learning_rate": 3.99884258929552e-07,
+      "loss": 1.5439,
+      "step": 55
+    },
+    {
+      "epoch": 0.011623080116230802,
+      "grad_norm": 0.9041137465808828,
+      "learning_rate": 3.9988001237374804e-07,
+      "loss": 1.6299,
+      "step": 56
+    },
+    {
+      "epoch": 0.011830635118306352,
+      "grad_norm": 1.112009836123954,
+      "learning_rate": 3.9987568933773844e-07,
+      "loss": 1.7235,
+      "step": 57
+    },
+    {
+      "epoch": 0.012038190120381901,
+      "grad_norm": 1.047373763538019,
+      "learning_rate": 3.9987128982336114e-07,
+      "loss": 1.6787,
+      "step": 58
+    },
+    {
+      "epoch": 0.012245745122457451,
+      "grad_norm": 0.8835602431501268,
+      "learning_rate": 3.998668138324867e-07,
+      "loss": 1.6616,
+      "step": 59
+    },
+    {
+      "epoch": 0.012453300124533,
+      "grad_norm": 1.6835123820339188,
+      "learning_rate": 3.998622613670183e-07,
+      "loss": 1.6134,
+      "step": 60
+    },
+    {
+      "epoch": 0.01266085512660855,
+      "grad_norm": 0.9534183321700821,
+      "learning_rate": 3.998576324288914e-07,
+      "loss": 1.6385,
+      "step": 61
+    },
+    {
+      "epoch": 0.012868410128684102,
+      "grad_norm": 0.9104143262122223,
+      "learning_rate": 3.998529270200741e-07,
+      "loss": 1.6014,
+      "step": 62
+    },
+    {
+      "epoch": 0.013075965130759652,
+      "grad_norm": 1.530542587667927,
+      "learning_rate": 3.9984814514256715e-07,
+      "loss": 1.5427,
+      "step": 63
+    },
+    {
+      "epoch": 0.013283520132835201,
+      "grad_norm": 1.2104918255889494,
+      "learning_rate": 3.9984328679840343e-07,
+      "loss": 1.6357,
+      "step": 64
+    },
+    {
+      "epoch": 0.013491075134910751,
+      "grad_norm": 0.8108098711016526,
+      "learning_rate": 3.9983835198964885e-07,
+      "loss": 1.5577,
+      "step": 65
+    },
+    {
+      "epoch": 0.0136986301369863,
+      "grad_norm": 0.8894370000259616,
+      "learning_rate": 3.9983334071840135e-07,
+      "loss": 1.6112,
+      "step": 66
+    },
+    {
+      "epoch": 0.013906185139061852,
+      "grad_norm": 0.7398465159411312,
+      "learning_rate": 3.9982825298679176e-07,
+      "loss": 1.6371,
+      "step": 67
+    },
+    {
+      "epoch": 0.014113740141137402,
+      "grad_norm": 1.4236875339228963,
+      "learning_rate": 3.9982308879698317e-07,
+      "loss": 1.619,
+      "step": 68
+    },
+    {
+      "epoch": 0.014321295143212951,
+      "grad_norm": 3.265087745899655,
+      "learning_rate": 3.998178481511712e-07,
+      "loss": 1.6064,
+      "step": 69
+    },
+    {
+      "epoch": 0.014528850145288501,
+      "grad_norm": 0.7568948390200323,
+      "learning_rate": 3.998125310515841e-07,
+      "loss": 1.5552,
+      "step": 70
+    },
+    {
+      "epoch": 0.01473640514736405,
+      "grad_norm": 0.7756456280737289,
+      "learning_rate": 3.998071375004826e-07,
+      "loss": 1.6534,
+      "step": 71
+    },
+    {
+      "epoch": 0.014943960149439602,
+      "grad_norm": 3.5014949988448847,
+      "learning_rate": 3.9980166750015975e-07,
+      "loss": 1.5392,
+      "step": 72
+    },
+    {
+      "epoch": 0.015151515151515152,
+      "grad_norm": 0.8739187478417162,
+      "learning_rate": 3.9979612105294144e-07,
+      "loss": 1.5773,
+      "step": 73
+    },
+    {
+      "epoch": 0.015359070153590702,
+      "grad_norm": 1.3102361826816729,
+      "learning_rate": 3.997904981611857e-07,
+      "loss": 1.5842,
+      "step": 74
+    },
+    {
+      "epoch": 0.015566625155666251,
+      "grad_norm": 0.8691694903148511,
+      "learning_rate": 3.9978479882728335e-07,
+      "loss": 1.6076,
+      "step": 75
+    },
+    {
+      "epoch": 0.0157741801577418,
+      "grad_norm": 0.7842186698846596,
+      "learning_rate": 3.997790230536575e-07,
+      "loss": 1.5706,
+      "step": 76
+    },
+    {
+      "epoch": 0.01598173515981735,
+      "grad_norm": 0.8596725339526274,
+      "learning_rate": 3.99773170842764e-07,
+      "loss": 1.5754,
+      "step": 77
+    },
+    {
+      "epoch": 0.0161892901618929,
+      "grad_norm": 1.0555561024172766,
+      "learning_rate": 3.9976724219709095e-07,
+      "loss": 1.5816,
+      "step": 78
+    },
+    {
+      "epoch": 0.01639684516396845,
+      "grad_norm": 0.6843759261931969,
+      "learning_rate": 3.9976123711915897e-07,
+      "loss": 1.6009,
+      "step": 79
+    },
+    {
+      "epoch": 0.016604400166044003,
+      "grad_norm": 1.2404904928062643,
+      "learning_rate": 3.9975515561152145e-07,
+      "loss": 1.6722,
+      "step": 80
+    },
+    {
+      "epoch": 0.016811955168119553,
+      "grad_norm": 1.0796349325750587,
+      "learning_rate": 3.9974899767676395e-07,
+      "loss": 1.6735,
+      "step": 81
+    },
+    {
+      "epoch": 0.017019510170195103,
+      "grad_norm": 0.744744231983327,
+      "learning_rate": 3.997427633175047e-07,
+      "loss": 1.6355,
+      "step": 82
+    },
+    {
+      "epoch": 0.017227065172270652,
+      "grad_norm": 0.8136623677227462,
+      "learning_rate": 3.997364525363944e-07,
+      "loss": 1.5393,
+      "step": 83
+    },
+    {
+      "epoch": 0.017434620174346202,
+      "grad_norm": 0.7060002367144068,
+      "learning_rate": 3.997300653361162e-07,
+      "loss": 1.6214,
+      "step": 84
+    },
+    {
+      "epoch": 0.01764217517642175,
+      "grad_norm": 0.8792853840506358,
+      "learning_rate": 3.997236017193858e-07,
+      "loss": 1.5898,
+      "step": 85
+    },
+    {
+      "epoch": 0.0178497301784973,
+      "grad_norm": 0.8053031716414784,
+      "learning_rate": 3.9971706168895136e-07,
+      "loss": 1.5168,
+      "step": 86
+    },
+    {
+      "epoch": 0.01805728518057285,
+      "grad_norm": 0.8358084053196052,
+      "learning_rate": 3.9971044524759344e-07,
+      "loss": 1.597,
+      "step": 87
+    },
+    {
+      "epoch": 0.0182648401826484,
+      "grad_norm": 1.8858433136361286,
+      "learning_rate": 3.9970375239812525e-07,
+      "loss": 1.6603,
+      "step": 88
+    },
+    {
+      "epoch": 0.01847239518472395,
+      "grad_norm": 0.7772154947227216,
+      "learning_rate": 3.996969831433925e-07,
+      "loss": 1.6533,
+      "step": 89
+    },
+    {
+      "epoch": 0.0186799501867995,
+      "grad_norm": 0.7203965146552309,
+      "learning_rate": 3.996901374862731e-07,
+      "loss": 1.559,
+      "step": 90
+    },
+    {
+      "epoch": 0.018887505188875053,
+      "grad_norm": 0.9141428370702639,
+      "learning_rate": 3.996832154296778e-07,
+      "loss": 1.5923,
+      "step": 91
+    },
+    {
+      "epoch": 0.019095060190950603,
+      "grad_norm": 0.8730542345016723,
+      "learning_rate": 3.9967621697654955e-07,
+      "loss": 1.6517,
+      "step": 92
+    },
+    {
+      "epoch": 0.019302615193026153,
+      "grad_norm": 0.700994912781778,
+      "learning_rate": 3.996691421298641e-07,
+      "loss": 1.5687,
+      "step": 93
+    },
+    {
+      "epoch": 0.019510170195101702,
+      "grad_norm": 0.8546352866257564,
+      "learning_rate": 3.996619908926292e-07,
+      "loss": 1.635,
+      "step": 94
+    },
+    {
+      "epoch": 0.019717725197177252,
+      "grad_norm": 0.7118268331423449,
+      "learning_rate": 3.9965476326788563e-07,
+      "loss": 1.5886,
+      "step": 95
+    },
+    {
+      "epoch": 0.019925280199252802,
+      "grad_norm": 0.9067268803639301,
+      "learning_rate": 3.9964745925870626e-07,
+      "loss": 1.5618,
+      "step": 96
+    },
+    {
+      "epoch": 0.02013283520132835,
+      "grad_norm": 7.3842092978252305,
+      "learning_rate": 3.9964007886819656e-07,
+      "loss": 1.5876,
+      "step": 97
+    },
+    {
+      "epoch": 0.0203403902034039,
+      "grad_norm": 0.978924255031244,
+      "learning_rate": 3.996326220994945e-07,
+      "loss": 1.6045,
+      "step": 98
+    },
+    {
+      "epoch": 0.02054794520547945,
+      "grad_norm": 0.822612571727339,
+      "learning_rate": 3.996250889557706e-07,
+      "loss": 1.629,
+      "step": 99
+    },
+    {
+      "epoch": 0.020755500207555,
+      "grad_norm": 0.8769641944628137,
+      "learning_rate": 3.996174794402276e-07,
+      "loss": 1.6471,
+      "step": 100
+    },
+    {
+      "epoch": 0.020963055209630554,
+      "grad_norm": 0.8126776014287214,
+      "learning_rate": 3.9960979355610085e-07,
+      "loss": 1.6068,
+      "step": 101
+    },
+    {
+      "epoch": 0.021170610211706103,
+      "grad_norm": 0.7784589415169838,
+      "learning_rate": 3.9960203130665823e-07,
+      "loss": 1.6294,
+      "step": 102
+    },
+    {
+      "epoch": 0.021378165213781653,
+      "grad_norm": 1.199036492044139,
+      "learning_rate": 3.9959419269520013e-07,
+      "loss": 1.5832,
+      "step": 103
+    },
+    {
+      "epoch": 0.021585720215857203,
+      "grad_norm": 0.7984232635569228,
+      "learning_rate": 3.9958627772505924e-07,
+      "loss": 1.5183,
+      "step": 104
+    },
+    {
+      "epoch": 0.021793275217932753,
+      "grad_norm": 0.9139096875631566,
+      "learning_rate": 3.9957828639960083e-07,
+      "loss": 1.5612,
+      "step": 105
+    },
+    {
+      "epoch": 0.022000830220008302,
+      "grad_norm": 1.0650085114380075,
+      "learning_rate": 3.995702187222225e-07,
+      "loss": 1.5779,
+      "step": 106
+    },
+    {
+      "epoch": 0.022208385222083852,
+      "grad_norm": 0.8002619026048111,
+      "learning_rate": 3.9956207469635454e-07,
+      "loss": 1.5826,
+      "step": 107
+    },
+    {
+      "epoch": 0.0224159402241594,
+      "grad_norm": 1.0383859991179343,
+      "learning_rate": 3.995538543254595e-07,
+      "loss": 1.5817,
+      "step": 108
+    },
+    {
+      "epoch": 0.02262349522623495,
+      "grad_norm": 0.7189329429408791,
+      "learning_rate": 3.995455576130325e-07,
+      "loss": 1.5694,
+      "step": 109
+    },
+    {
+      "epoch": 0.0228310502283105,
+      "grad_norm": 0.7622695324143753,
+      "learning_rate": 3.9953718456260113e-07,
+      "loss": 1.6204,
+      "step": 110
+    },
+    {
+      "epoch": 0.02303860523038605,
+      "grad_norm": 0.7427754575751896,
+      "learning_rate": 3.9952873517772524e-07,
+      "loss": 1.5273,
+      "step": 111
+    },
+    {
+      "epoch": 0.023246160232461604,
+      "grad_norm": 1.3000202135831596,
+      "learning_rate": 3.995202094619974e-07,
+      "loss": 1.6408,
+      "step": 112
+    },
+    {
+      "epoch": 0.023453715234537154,
+      "grad_norm": 0.9593720967609558,
+      "learning_rate": 3.995116074190424e-07,
+      "loss": 1.6065,
+      "step": 113
+    },
+    {
+      "epoch": 0.023661270236612703,
+      "grad_norm": 0.6955278941718566,
+      "learning_rate": 3.995029290525178e-07,
+      "loss": 1.4961,
+      "step": 114
+    },
+    {
+      "epoch": 0.023868825238688253,
+      "grad_norm": 0.9759198828584535,
+      "learning_rate": 3.9949417436611325e-07,
+      "loss": 1.5576,
+      "step": 115
+    },
+    {
+      "epoch": 0.024076380240763803,
+      "grad_norm": 0.7554845221022716,
+      "learning_rate": 3.994853433635511e-07,
+      "loss": 1.5302,
+      "step": 116
+    },
+    {
+      "epoch": 0.024283935242839352,
+      "grad_norm": 0.7163127910598674,
+      "learning_rate": 3.99476436048586e-07,
+      "loss": 1.5539,
+      "step": 117
+    },
+    {
+      "epoch": 0.024491490244914902,
+      "grad_norm": 1.1803030168547048,
+      "learning_rate": 3.9946745242500507e-07,
+      "loss": 1.546,
+      "step": 118
+    },
+    {
+      "epoch": 0.02469904524699045,
+      "grad_norm": 1.107046921929907,
+      "learning_rate": 3.99458392496628e-07,
+      "loss": 1.5175,
+      "step": 119
+    },
+    {
+      "epoch": 0.024906600249066,
+      "grad_norm": 0.7044935713830988,
+      "learning_rate": 3.9944925626730676e-07,
+      "loss": 1.652,
+      "step": 120
+    },
+    {
+      "epoch": 0.02511415525114155,
+      "grad_norm": 1.4134203081620509,
+      "learning_rate": 3.994400437409259e-07,
+      "loss": 1.6474,
+      "step": 121
+    },
+    {
+      "epoch": 0.0253217102532171,
+      "grad_norm": 0.8450789118515795,
+      "learning_rate": 3.9943075492140234e-07,
+      "loss": 1.4921,
+      "step": 122
+    },
+    {
+      "epoch": 0.025529265255292654,
+      "grad_norm": 0.8784976315552246,
+      "learning_rate": 3.9942138981268536e-07,
+      "loss": 1.5586,
+      "step": 123
+    },
+    {
+      "epoch": 0.025736820257368204,
+      "grad_norm": 1.8578157384463314,
+      "learning_rate": 3.9941194841875676e-07,
+      "loss": 1.6001,
+      "step": 124
+    },
+    {
+      "epoch": 0.025944375259443753,
+      "grad_norm": 0.8477921362158418,
+      "learning_rate": 3.994024307436309e-07,
+      "loss": 1.5293,
+      "step": 125
+    },
+    {
+      "epoch": 0.026151930261519303,
+      "grad_norm": 0.8270019911234011,
+      "learning_rate": 3.993928367913543e-07,
+      "loss": 1.5722,
+      "step": 126
+    },
+    {
+      "epoch": 0.026359485263594853,
+      "grad_norm": 2.1323843964417595,
+      "learning_rate": 3.99383166566006e-07,
+      "loss": 1.5979,
+      "step": 127
+    },
+    {
+      "epoch": 0.026567040265670402,
+      "grad_norm": 0.8831376361302666,
+      "learning_rate": 3.9937342007169777e-07,
+      "loss": 1.5814,
+      "step": 128
+    },
+    {
+      "epoch": 0.026774595267745952,
+      "grad_norm": 0.7713679154410665,
+      "learning_rate": 3.993635973125734e-07,
+      "loss": 1.6059,
+      "step": 129
+    },
+    {
+      "epoch": 0.026982150269821502,
+      "grad_norm": 1.3485856894156136,
+      "learning_rate": 3.9935369829280924e-07,
+      "loss": 1.5848,
+      "step": 130
+    },
+    {
+      "epoch": 0.02718970527189705,
+      "grad_norm": 0.7679506123644734,
+      "learning_rate": 3.9934372301661416e-07,
+      "loss": 1.5869,
+      "step": 131
+    },
+    {
+      "epoch": 0.0273972602739726,
+      "grad_norm": 1.4099730664181218,
+      "learning_rate": 3.9933367148822936e-07,
+      "loss": 1.609,
+      "step": 132
+    },
+    {
+      "epoch": 0.027604815276048154,
+      "grad_norm": 0.8527046271639593,
+      "learning_rate": 3.993235437119285e-07,
+      "loss": 1.6518,
+      "step": 133
+    },
+    {
+      "epoch": 0.027812370278123704,
+      "grad_norm": 0.9876695238073084,
+      "learning_rate": 3.993133396920176e-07,
+      "loss": 1.5732,
+      "step": 134
+    },
+    {
+      "epoch": 0.028019925280199254,
+      "grad_norm": 0.8848818713241431,
+      "learning_rate": 3.993030594328352e-07,
+      "loss": 1.619,
+      "step": 135
+    },
+    {
+      "epoch": 0.028227480282274803,
+      "grad_norm": 0.8445329910933679,
+      "learning_rate": 3.9929270293875204e-07,
+      "loss": 1.6159,
+      "step": 136
+    },
+    {
+      "epoch": 0.028435035284350353,
+      "grad_norm": 0.7036153934743418,
+      "learning_rate": 3.992822702141717e-07,
+      "loss": 1.5072,
+      "step": 137
+    },
+    {
+      "epoch": 0.028642590286425903,
+      "grad_norm": 0.7875600395024726,
+      "learning_rate": 3.992717612635296e-07,
+      "loss": 1.684,
+      "step": 138
+    },
+    {
+      "epoch": 0.028850145288501453,
+      "grad_norm": 0.6610570426998603,
+      "learning_rate": 3.992611760912941e-07,
+      "loss": 1.5804,
+      "step": 139
+    },
+    {
+      "epoch": 0.029057700290577002,
+      "grad_norm": 1.0213832385611197,
+      "learning_rate": 3.992505147019656e-07,
+      "loss": 1.5811,
+      "step": 140
+    },
+    {
+      "epoch": 0.029265255292652552,
+      "grad_norm": 1.0008489967572651,
+      "learning_rate": 3.9923977710007705e-07,
+      "loss": 1.4963,
+      "step": 141
+    },
+    {
+      "epoch": 0.0294728102947281,
+      "grad_norm": 0.6649639875346893,
+      "learning_rate": 3.992289632901939e-07,
+      "loss": 1.5339,
+      "step": 142
+    },
+    {
+      "epoch": 0.02968036529680365,
+      "grad_norm": 2.2653111740366585,
+      "learning_rate": 3.9921807327691375e-07,
+      "loss": 1.6083,
+      "step": 143
+    },
+    {
+      "epoch": 0.029887920298879204,
+      "grad_norm": 1.1445291033894631,
+      "learning_rate": 3.992071070648668e-07,
+      "loss": 1.5568,
+      "step": 144
+    },
+    {
+      "epoch": 0.030095475300954754,
+      "grad_norm": 2.245761561167164,
+      "learning_rate": 3.9919606465871565e-07,
+      "loss": 1.5667,
+      "step": 145
+    },
+    {
+      "epoch": 0.030303030303030304,
+      "grad_norm": 0.9655237026435531,
+      "learning_rate": 3.991849460631552e-07,
+      "loss": 1.5883,
+      "step": 146
+    },
+    {
+      "epoch": 0.030510585305105854,
+      "grad_norm": 0.8122892559125585,
+      "learning_rate": 3.9917375128291276e-07,
+      "loss": 1.6374,
+      "step": 147
+    },
+    {
+      "epoch": 0.030718140307181403,
+      "grad_norm": 1.1617726676187454,
+      "learning_rate": 3.9916248032274807e-07,
+      "loss": 1.5375,
+      "step": 148
+    },
+    {
+      "epoch": 0.030925695309256953,
+      "grad_norm": 0.6676490919385982,
+      "learning_rate": 3.9915113318745327e-07,
+      "loss": 1.6109,
+      "step": 149
+    },
+    {
+      "epoch": 0.031133250311332503,
+      "grad_norm": 0.700265676155718,
+      "learning_rate": 3.9913970988185274e-07,
+      "loss": 1.5653,
+      "step": 150
+    },
+    {
+      "epoch": 0.031340805313408056,
+      "grad_norm": 0.6914303208551976,
+      "learning_rate": 3.9912821041080353e-07,
+      "loss": 1.5813,
+      "step": 151
+    },
+    {
+      "epoch": 0.0315483603154836,
+      "grad_norm": 1.5483990604207802,
+      "learning_rate": 3.9911663477919483e-07,
+      "loss": 1.6318,
+      "step": 152
+    },
+    {
+      "epoch": 0.031755915317559155,
+      "grad_norm": 2.426454657016846,
+      "learning_rate": 3.9910498299194825e-07,
+      "loss": 1.6519,
+      "step": 153
+    },
+    {
+      "epoch": 0.0319634703196347,
+      "grad_norm": 2.8984298572473453,
+      "learning_rate": 3.9909325505401795e-07,
+      "loss": 1.6033,
+      "step": 154
+    },
+    {
+      "epoch": 0.032171025321710255,
+      "grad_norm": 1.1062149773615026,
+      "learning_rate": 3.990814509703902e-07,
+      "loss": 1.5272,
+      "step": 155
+    },
+    {
+      "epoch": 0.0323785803237858,
+      "grad_norm": 1.2325479409853999,
+      "learning_rate": 3.9906957074608384e-07,
+      "loss": 1.5379,
+      "step": 156
+    },
+    {
+      "epoch": 0.032586135325861354,
+      "grad_norm": 0.6306153405028786,
+      "learning_rate": 3.9905761438615004e-07,
+      "loss": 1.5455,
+      "step": 157
+    },
+    {
+      "epoch": 0.0327936903279369,
+      "grad_norm": 0.8282588567870623,
+      "learning_rate": 3.990455818956723e-07,
+      "loss": 1.5969,
+      "step": 158
+    },
+    {
+      "epoch": 0.03300124533001245,
+      "grad_norm": 0.8312314136498741,
+      "learning_rate": 3.990334732797665e-07,
+      "loss": 1.5551,
+      "step": 159
+    },
+    {
+      "epoch": 0.033208800332088007,
+      "grad_norm": 0.8127972376519234,
+      "learning_rate": 3.99021288543581e-07,
+      "loss": 1.5505,
+      "step": 160
+    },
+    {
+      "epoch": 0.03341635533416355,
+      "grad_norm": 0.7182138887399596,
+      "learning_rate": 3.990090276922963e-07,
+      "loss": 1.5305,
+      "step": 161
+    },
+    {
+      "epoch": 0.033623910336239106,
+      "grad_norm": 0.8211628140749623,
+      "learning_rate": 3.9899669073112546e-07,
+      "loss": 1.5352,
+      "step": 162
+    },
+    {
+      "epoch": 0.03383146533831465,
+      "grad_norm": 1.1408798537766063,
+      "learning_rate": 3.9898427766531383e-07,
+      "loss": 1.6034,
+      "step": 163
+    },
+    {
+      "epoch": 0.034039020340390205,
+      "grad_norm": 0.8372043440866829,
+      "learning_rate": 3.9897178850013913e-07,
+      "loss": 1.6699,
+      "step": 164
+    },
+    {
+      "epoch": 0.03424657534246575,
+      "grad_norm": 0.700689461293022,
+      "learning_rate": 3.989592232409113e-07,
+      "loss": 1.5614,
+      "step": 165
+    },
+    {
+      "epoch": 0.034454130344541305,
+      "grad_norm": 0.6977281072027869,
+      "learning_rate": 3.9894658189297294e-07,
+      "loss": 1.6408,
+      "step": 166
+    },
+    {
+      "epoch": 0.03466168534661685,
+      "grad_norm": 1.094797373432667,
+      "learning_rate": 3.9893386446169863e-07,
+      "loss": 1.5868,
+      "step": 167
+    },
+    {
+      "epoch": 0.034869240348692404,
+      "grad_norm": 0.6908473634514661,
+      "learning_rate": 3.989210709524957e-07,
+      "loss": 1.5045,
+      "step": 168
+    },
+    {
+      "epoch": 0.03507679535076795,
+      "grad_norm": 1.1944367559596936,
+      "learning_rate": 3.9890820137080334e-07,
+      "loss": 1.5161,
+      "step": 169
+    },
+    {
+      "epoch": 0.0352843503528435,
+      "grad_norm": 0.773620165920913,
+      "learning_rate": 3.9889525572209363e-07,
+      "loss": 1.5781,
+      "step": 170
+    },
+    {
+      "epoch": 0.03549190535491906,
+      "grad_norm": 0.7063459328098319,
+      "learning_rate": 3.9888223401187047e-07,
+      "loss": 1.5963,
+      "step": 171
+    },
+    {
+      "epoch": 0.0356994603569946,
+      "grad_norm": 0.7325877920986794,
+      "learning_rate": 3.9886913624567054e-07,
+      "loss": 1.5681,
+      "step": 172
+    },
+    {
+      "epoch": 0.035907015359070156,
+      "grad_norm": 0.7487252631547822,
+      "learning_rate": 3.988559624290625e-07,
+      "loss": 1.6309,
+      "step": 173
+    },
+    {
+      "epoch": 0.0361145703611457,
+      "grad_norm": 0.7801430432423838,
+      "learning_rate": 3.988427125676477e-07,
+      "loss": 1.657,
+      "step": 174
+    },
+    {
+      "epoch": 0.036322125363221255,
+      "grad_norm": 1.0691624441098393,
+      "learning_rate": 3.988293866670595e-07,
+      "loss": 1.5523,
+      "step": 175
+    },
+    {
+      "epoch": 0.0365296803652968,
+      "grad_norm": 0.9498431902389803,
+      "learning_rate": 3.9881598473296367e-07,
+      "loss": 1.562,
+      "step": 176
+    },
+    {
+      "epoch": 0.036737235367372355,
+      "grad_norm": 0.6660373100674684,
+      "learning_rate": 3.9880250677105847e-07,
+      "loss": 1.6115,
+      "step": 177
+    },
+    {
+      "epoch": 0.0369447903694479,
+      "grad_norm": 0.7341769246356358,
+      "learning_rate": 3.987889527870743e-07,
+      "loss": 1.5515,
+      "step": 178
+    },
+    {
+      "epoch": 0.037152345371523454,
+      "grad_norm": 0.8412797012709253,
+      "learning_rate": 3.9877532278677396e-07,
+      "loss": 1.5441,
+      "step": 179
+    },
+    {
+      "epoch": 0.037359900373599,
+      "grad_norm": 1.6902112570845282,
+      "learning_rate": 3.9876161677595263e-07,
+      "loss": 1.4765,
+      "step": 180
+    },
+    {
+      "epoch": 0.037567455375674554,
+      "grad_norm": 0.8407033758598549,
+      "learning_rate": 3.987478347604377e-07,
+      "loss": 1.5988,
+      "step": 181
+    },
+    {
+      "epoch": 0.03777501037775011,
+      "grad_norm": 0.7351067608061718,
+      "learning_rate": 3.9873397674608895e-07,
+      "loss": 1.5781,
+      "step": 182
+    },
+    {
+      "epoch": 0.03798256537982565,
+      "grad_norm": 0.8740413274445975,
+      "learning_rate": 3.9872004273879834e-07,
+      "loss": 1.5492,
+      "step": 183
+    },
+    {
+      "epoch": 0.038190120381901206,
+      "grad_norm": 2.059436882263371,
+      "learning_rate": 3.987060327444904e-07,
+      "loss": 1.5662,
+      "step": 184
+    },
+    {
+      "epoch": 0.03839767538397675,
+      "grad_norm": 2.4231207058645037,
+      "learning_rate": 3.9869194676912164e-07,
+      "loss": 1.5922,
+      "step": 185
+    },
+    {
+      "epoch": 0.038605230386052306,
+      "grad_norm": 1.0766649990188597,
+      "learning_rate": 3.9867778481868114e-07,
+      "loss": 1.6493,
+      "step": 186
+    },
+    {
+      "epoch": 0.03881278538812785,
+      "grad_norm": 1.020653464015153,
+      "learning_rate": 3.986635468991901e-07,
+      "loss": 1.4228,
+      "step": 187
+    },
+    {
+      "epoch": 0.039020340390203405,
+      "grad_norm": 0.7514655130406892,
+      "learning_rate": 3.986492330167022e-07,
+      "loss": 1.5397,
+      "step": 188
+    },
+    {
+      "epoch": 0.03922789539227895,
+      "grad_norm": 0.6943121513169542,
+      "learning_rate": 3.986348431773033e-07,
+      "loss": 1.5768,
+      "step": 189
+    },
+    {
+      "epoch": 0.039435450394354504,
+      "grad_norm": 0.7628930984074667,
+      "learning_rate": 3.986203773871115e-07,
+      "loss": 1.5766,
+      "step": 190
+    },
+    {
+      "epoch": 0.03964300539643005,
+      "grad_norm": 1.0381762765052955,
+      "learning_rate": 3.9860583565227744e-07,
+      "loss": 1.5486,
+      "step": 191
+    },
+    {
+      "epoch": 0.039850560398505604,
+      "grad_norm": 0.7754563764940113,
+      "learning_rate": 3.985912179789838e-07,
+      "loss": 1.5418,
+      "step": 192
+    },
+    {
+      "epoch": 0.04005811540058116,
+      "grad_norm": 0.8452461811861927,
+      "learning_rate": 3.985765243734455e-07,
+      "loss": 1.5459,
+      "step": 193
+    },
+    {
+      "epoch": 0.0402656704026567,
+      "grad_norm": 0.756668308136434,
+      "learning_rate": 3.9856175484191004e-07,
+      "loss": 1.5337,
+      "step": 194
+    },
+    {
+      "epoch": 0.040473225404732256,
+      "grad_norm": 0.6655494883376677,
+      "learning_rate": 3.9854690939065693e-07,
+      "loss": 1.589,
+      "step": 195
+    },
+    {
+      "epoch": 0.0406807804068078,
+      "grad_norm": 1.311028732233012,
+      "learning_rate": 3.9853198802599806e-07,
+      "loss": 1.5563,
+      "step": 196
+    },
+    {
+      "epoch": 0.040888335408883356,
+      "grad_norm": 0.8465929555909902,
+      "learning_rate": 3.985169907542777e-07,
+      "loss": 1.6205,
+      "step": 197
+    },
+    {
+      "epoch": 0.0410958904109589,
+      "grad_norm": 2.1440242597231065,
+      "learning_rate": 3.9850191758187214e-07,
+      "loss": 1.5954,
+      "step": 198
+    },
+    {
+      "epoch": 0.041303445413034455,
+      "grad_norm": 0.7041663115618453,
+      "learning_rate": 3.984867685151903e-07,
+      "loss": 1.5534,
+      "step": 199
+    },
+    {
+      "epoch": 0.04151100041511,
+      "grad_norm": 0.8029009177790535,
+      "learning_rate": 3.98471543560673e-07,
+      "loss": 1.544,
+      "step": 200
+    },
+    {
+      "epoch": 0.041718555417185554,
+      "grad_norm": 0.8166296318967385,
+      "learning_rate": 3.984562427247935e-07,
+      "loss": 1.6173,
+      "step": 201
+    },
+    {
+      "epoch": 0.04192611041926111,
+      "grad_norm": 0.8011846342714176,
+      "learning_rate": 3.9844086601405734e-07,
+      "loss": 1.607,
+      "step": 202
+    },
+    {
+      "epoch": 0.042133665421336654,
+      "grad_norm": 0.6950052511381265,
+      "learning_rate": 3.9842541343500233e-07,
+      "loss": 1.5547,
+      "step": 203
+    },
+    {
+      "epoch": 0.04234122042341221,
+      "grad_norm": 0.9926840980550315,
+      "learning_rate": 3.9840988499419844e-07,
+      "loss": 1.5576,
+      "step": 204
+    },
+    {
+      "epoch": 0.04254877542548775,
+      "grad_norm": 1.0280956372225183,
+      "learning_rate": 3.9839428069824793e-07,
+      "loss": 1.518,
+      "step": 205
+    },
+    {
+      "epoch": 0.042756330427563306,
+      "grad_norm": 0.7912577642963848,
+      "learning_rate": 3.983786005537854e-07,
+      "loss": 1.535,
+      "step": 206
+    },
+    {
+      "epoch": 0.04296388542963885,
+      "grad_norm": 1.2668204255261601,
+      "learning_rate": 3.9836284456747753e-07,
+      "loss": 1.6266,
+      "step": 207
+    },
+    {
+      "epoch": 0.043171440431714406,
+      "grad_norm": 0.8746201750394597,
+      "learning_rate": 3.983470127460235e-07,
+      "loss": 1.5874,
+      "step": 208
+    },
+    {
+      "epoch": 0.04337899543378995,
+      "grad_norm": 1.3221081900743643,
+      "learning_rate": 3.9833110509615447e-07,
+      "loss": 1.5472,
+      "step": 209
+    },
+    {
+      "epoch": 0.043586550435865505,
+      "grad_norm": 0.7677515453124257,
+      "learning_rate": 3.9831512162463393e-07,
+      "loss": 1.5622,
+      "step": 210
+    },
+    {
+      "epoch": 0.04379410543794105,
+      "grad_norm": 0.7388518036238274,
+      "learning_rate": 3.982990623382577e-07,
+      "loss": 1.5878,
+      "step": 211
+    },
+    {
+      "epoch": 0.044001660440016604,
+      "grad_norm": 0.8349483578418847,
+      "learning_rate": 3.982829272438538e-07,
+      "loss": 1.6191,
+      "step": 212
+    },
+    {
+      "epoch": 0.04420921544209216,
+      "grad_norm": 0.7866287359865883,
+      "learning_rate": 3.982667163482823e-07,
+      "loss": 1.5913,
+      "step": 213
+    },
+    {
+      "epoch": 0.044416770444167704,
+      "grad_norm": 0.7337542585933059,
+      "learning_rate": 3.9825042965843574e-07,
+      "loss": 1.6263,
+      "step": 214
+    },
+    {
+      "epoch": 0.04462432544624326,
+      "grad_norm": 0.8087773209608627,
+      "learning_rate": 3.9823406718123876e-07,
+      "loss": 1.5302,
+      "step": 215
+    },
+    {
+      "epoch": 0.0448318804483188,
+      "grad_norm": 1.6289052960820165,
+      "learning_rate": 3.9821762892364824e-07,
+      "loss": 1.503,
+      "step": 216
+    },
+    {
+      "epoch": 0.045039435450394356,
+      "grad_norm": 0.7539242918102785,
+      "learning_rate": 3.9820111489265337e-07,
+      "loss": 1.4897,
+      "step": 217
+    },
+    {
+      "epoch": 0.0452469904524699,
+      "grad_norm": 0.7939260472145644,
+      "learning_rate": 3.981845250952754e-07,
+      "loss": 1.5782,
+      "step": 218
+    },
+    {
+      "epoch": 0.045454545454545456,
+      "grad_norm": 1.7727611690481937,
+      "learning_rate": 3.981678595385679e-07,
+      "loss": 1.5492,
+      "step": 219
+    },
+    {
+      "epoch": 0.045662100456621,
+      "grad_norm": 0.8533572039520807,
+      "learning_rate": 3.9815111822961653e-07,
+      "loss": 1.6221,
+      "step": 220
+    },
+    {
+      "epoch": 0.045869655458696555,
+      "grad_norm": 0.7203401381369927,
+      "learning_rate": 3.9813430117553944e-07,
+      "loss": 1.5412,
+      "step": 221
+    },
+    {
+      "epoch": 0.0460772104607721,
+      "grad_norm": 0.8959047237613945,
+      "learning_rate": 3.9811740838348664e-07,
+      "loss": 1.6073,
+      "step": 222
+    },
+    {
+      "epoch": 0.046284765462847655,
+      "grad_norm": 0.7759094459704792,
+      "learning_rate": 3.9810043986064053e-07,
+      "loss": 1.5732,
+      "step": 223
+    },
+    {
+      "epoch": 0.04649232046492321,
+      "grad_norm": 1.3792835863498725,
+      "learning_rate": 3.980833956142157e-07,
+      "loss": 1.6014,
+      "step": 224
+    },
+    {
+      "epoch": 0.046699875466998754,
+      "grad_norm": 1.1790585519053556,
+      "learning_rate": 3.9806627565145887e-07,
+      "loss": 1.5806,
+      "step": 225
+    },
+    {
+      "epoch": 0.04690743046907431,
+      "grad_norm": 0.8267158587237875,
+      "learning_rate": 3.9804907997964907e-07,
+      "loss": 1.5894,
+      "step": 226
+    },
+    {
+      "epoch": 0.04711498547114985,
+      "grad_norm": 1.4237674190785736,
+      "learning_rate": 3.9803180860609736e-07,
+      "loss": 1.5743,
+      "step": 227
+    },
+    {
+      "epoch": 0.047322540473225407,
+      "grad_norm": 0.7671641452665827,
+      "learning_rate": 3.980144615381472e-07,
+      "loss": 1.6446,
+      "step": 228
+    },
+    {
+      "epoch": 0.04753009547530095,
+      "grad_norm": 0.7846547235437324,
+      "learning_rate": 3.979970387831739e-07,
+      "loss": 1.6352,
+      "step": 229
+    },
+    {
+      "epoch": 0.047737650477376506,
+      "grad_norm": 1.051907504842811,
+      "learning_rate": 3.9797954034858534e-07,
+      "loss": 1.4974,
+      "step": 230
+    },
+    {
+      "epoch": 0.04794520547945205,
+      "grad_norm": 0.9643123621085294,
+      "learning_rate": 3.9796196624182127e-07,
+      "loss": 1.6384,
+      "step": 231
+    },
+    {
+      "epoch": 0.048152760481527605,
+      "grad_norm": 0.8182110474596903,
+      "learning_rate": 3.979443164703538e-07,
+      "loss": 1.5885,
+      "step": 232
+    },
+    {
+      "epoch": 0.04836031548360315,
+      "grad_norm": 1.8122984015829537,
+      "learning_rate": 3.9792659104168703e-07,
+      "loss": 1.5165,
+      "step": 233
+    },
+    {
+      "epoch": 0.048567870485678705,
+      "grad_norm": 0.9606591271592947,
+      "learning_rate": 3.9790878996335757e-07,
+      "loss": 1.5418,
+      "step": 234
+    },
+    {
+      "epoch": 0.04877542548775426,
+      "grad_norm": 1.097135309008659,
+      "learning_rate": 3.978909132429337e-07,
+      "loss": 1.5506,
+      "step": 235
+    },
+    {
+      "epoch": 0.048982980489829804,
+      "grad_norm": 1.2352878041003887,
+      "learning_rate": 3.9787296088801636e-07,
+      "loss": 1.5445,
+      "step": 236
+    },
+    {
+      "epoch": 0.04919053549190536,
+      "grad_norm": 1.5772240564416988,
+      "learning_rate": 3.9785493290623825e-07,
+      "loss": 1.5373,
+      "step": 237
+    },
+    {
+      "epoch": 0.0493980904939809,
+      "grad_norm": 0.7534201022763537,
+      "learning_rate": 3.9783682930526443e-07,
+      "loss": 1.6307,
+      "step": 238
+    },
+    {
+      "epoch": 0.04960564549605646,
+      "grad_norm": 0.7530367285207137,
+      "learning_rate": 3.9781865009279217e-07,
+      "loss": 1.5668,
+      "step": 239
+    },
+    {
+      "epoch": 0.049813200498132,
+      "grad_norm": 0.9391472494019696,
+      "learning_rate": 3.978003952765506e-07,
+      "loss": 1.585,
+      "step": 240
+    },
+    {
+      "epoch": 0.050020755500207556,
+      "grad_norm": 0.774852143433553,
+      "learning_rate": 3.977820648643014e-07,
+      "loss": 1.6155,
+      "step": 241
+    },
+    {
+      "epoch": 0.0502283105022831,
+      "grad_norm": 0.745677374247135,
+      "learning_rate": 3.97763658863838e-07,
+      "loss": 1.5293,
+      "step": 242
+    },
+    {
+      "epoch": 0.050435865504358655,
+      "grad_norm": 1.1837885959870564,
+      "learning_rate": 3.977451772829862e-07,
+      "loss": 1.5139,
+      "step": 243
+    },
+    {
+      "epoch": 0.0506434205064342,
+      "grad_norm": 0.7017706704601079,
+      "learning_rate": 3.977266201296039e-07,
+      "loss": 1.5863,
+      "step": 244
+    },
+    {
+      "epoch": 0.050850975508509755,
+      "grad_norm": 0.6737548118654552,
+      "learning_rate": 3.9770798741158113e-07,
+      "loss": 1.5699,
+      "step": 245
+    },
+    {
+      "epoch": 0.05105853051058531,
+      "grad_norm": 0.869460773178111,
+      "learning_rate": 3.976892791368399e-07,
+      "loss": 1.5654,
+      "step": 246
+    },
+    {
+      "epoch": 0.051266085512660854,
+      "grad_norm": 0.7162890058971442,
+      "learning_rate": 3.976704953133347e-07,
+      "loss": 1.5381,
+      "step": 247
+    },
+    {
+      "epoch": 0.05147364051473641,
+      "grad_norm": 0.8775845731013817,
+      "learning_rate": 3.976516359490517e-07,
+      "loss": 1.5739,
+      "step": 248
+    },
+    {
+      "epoch": 0.051681195516811954,
+      "grad_norm": 3.5312141103184507,
+      "learning_rate": 3.976327010520094e-07,
+      "loss": 1.5124,
+      "step": 249
+    },
+    {
+      "epoch": 0.05188875051888751,
+      "grad_norm": 0.734015116526283,
+      "learning_rate": 3.976136906302586e-07,
+      "loss": 1.6033,
+      "step": 250
+    },
+    {
+      "epoch": 0.05209630552096305,
+      "grad_norm": 0.8160873165015806,
+      "learning_rate": 3.975946046918819e-07,
+      "loss": 1.5691,
+      "step": 251
+    },
+    {
+      "epoch": 0.052303860523038606,
+      "grad_norm": 0.779206255074589,
+      "learning_rate": 3.9757544324499415e-07,
+      "loss": 1.5603,
+      "step": 252
+    },
+    {
+      "epoch": 0.05251141552511415,
+      "grad_norm": 0.7708855397759548,
+      "learning_rate": 3.9755620629774227e-07,
+      "loss": 1.5041,
+      "step": 253
+    },
+    {
+      "epoch": 0.052718970527189705,
+      "grad_norm": 0.7890497455375947,
+      "learning_rate": 3.9753689385830537e-07,
+      "loss": 1.5939,
+      "step": 254
+    },
+    {
+      "epoch": 0.05292652552926526,
+      "grad_norm": 0.9441825672726503,
+      "learning_rate": 3.975175059348945e-07,
+      "loss": 1.5454,
+      "step": 255
+    },
+    {
+      "epoch": 0.053134080531340805,
+      "grad_norm": 0.7639909825332569,
+      "learning_rate": 3.974980425357529e-07,
+      "loss": 1.6438,
+      "step": 256
+    },
+    {
+      "epoch": 0.05334163553341636,
+      "grad_norm": 1.692814812927853,
+      "learning_rate": 3.97478503669156e-07,
+      "loss": 1.5218,
+      "step": 257
+    },
+    {
+      "epoch": 0.053549190535491904,
+      "grad_norm": 0.7696333481095123,
+      "learning_rate": 3.9745888934341104e-07,
+      "loss": 1.5503,
+      "step": 258
+    },
+    {
+      "epoch": 0.05375674553756746,
+      "grad_norm": 0.9242829647779888,
+      "learning_rate": 3.9743919956685763e-07,
+      "loss": 1.5549,
+      "step": 259
+    },
+    {
+      "epoch": 0.053964300539643004,
+      "grad_norm": 0.8261988373601983,
+      "learning_rate": 3.974194343478673e-07,
+      "loss": 1.574,
+      "step": 260
+    },
+    {
+      "epoch": 0.05417185554171856,
+      "grad_norm": 0.8776349223600801,
+      "learning_rate": 3.9739959369484374e-07,
+      "loss": 1.5281,
+      "step": 261
+    },
+    {
+      "epoch": 0.0543794105437941,
+      "grad_norm": 0.7375384171292336,
+      "learning_rate": 3.973796776162226e-07,
+      "loss": 1.5369,
+      "step": 262
+    },
+    {
+      "epoch": 0.054586965545869656,
+      "grad_norm": 0.7404636618103695,
+      "learning_rate": 3.973596861204717e-07,
+      "loss": 1.59,
+      "step": 263
+    },
+    {
+      "epoch": 0.0547945205479452,
+      "grad_norm": 0.8365598934567327,
+      "learning_rate": 3.973396192160909e-07,
+      "loss": 1.5726,
+      "step": 264
+    },
+    {
+      "epoch": 0.055002075550020756,
+      "grad_norm": 0.9914600821937776,
+      "learning_rate": 3.9731947691161213e-07,
+      "loss": 1.5677,
+      "step": 265
+    },
+    {
+      "epoch": 0.05520963055209631,
+      "grad_norm": 0.8716730545567861,
+      "learning_rate": 3.972992592155993e-07,
+      "loss": 1.5474,
+      "step": 266
+    },
+    {
+      "epoch": 0.055417185554171855,
+      "grad_norm": 3.258720642013415,
+      "learning_rate": 3.972789661366485e-07,
+      "loss": 1.4957,
+      "step": 267
+    },
+    {
+      "epoch": 0.05562474055624741,
+      "grad_norm": 1.0773273816958533,
+      "learning_rate": 3.9725859768338776e-07,
+      "loss": 1.5619,
+      "step": 268
+    },
+    {
+      "epoch": 0.055832295558322954,
+      "grad_norm": 0.9024714694961539,
+      "learning_rate": 3.9723815386447727e-07,
+      "loss": 1.5378,
+      "step": 269
+    },
+    {
+      "epoch": 0.05603985056039851,
+      "grad_norm": 1.3831821553342059,
+      "learning_rate": 3.972176346886092e-07,
+      "loss": 1.5821,
+      "step": 270
+    },
+    {
+      "epoch": 0.056247405562474054,
+      "grad_norm": 0.952469940854049,
+      "learning_rate": 3.9719704016450766e-07,
+      "loss": 1.5554,
+      "step": 271
+    },
+    {
+      "epoch": 0.05645496056454961,
+      "grad_norm": 1.4725563826659172,
+      "learning_rate": 3.9717637030092897e-07,
+      "loss": 1.5343,
+      "step": 272
+    },
+    {
+      "epoch": 0.05666251556662515,
+      "grad_norm": 0.8363626557190338,
+      "learning_rate": 3.9715562510666136e-07,
+      "loss": 1.6163,
+      "step": 273
+    },
+    {
+      "epoch": 0.056870070568700706,
+      "grad_norm": 1.4087040072857615,
+      "learning_rate": 3.9713480459052524e-07,
+      "loss": 1.6336,
+      "step": 274
+    },
+    {
+      "epoch": 0.05707762557077625,
+      "grad_norm": 1.1491760517128764,
+      "learning_rate": 3.971139087613728e-07,
+      "loss": 1.5442,
+      "step": 275
+    },
+    {
+      "epoch": 0.057285180572851806,
+      "grad_norm": 0.688465467984676,
+      "learning_rate": 3.9709293762808846e-07,
+      "loss": 1.5556,
+      "step": 276
+    },
+    {
+      "epoch": 0.05749273557492736,
+      "grad_norm": 0.7734717471695348,
+      "learning_rate": 3.970718911995887e-07,
+      "loss": 1.5608,
+      "step": 277
+    },
+    {
+      "epoch": 0.057700290577002905,
+      "grad_norm": 0.9435481776922412,
+      "learning_rate": 3.970507694848217e-07,
+      "loss": 1.5769,
+      "step": 278
+    },
+    {
+      "epoch": 0.05790784557907846,
+      "grad_norm": 0.8203070725625297,
+      "learning_rate": 3.970295724927679e-07,
+      "loss": 1.5072,
+      "step": 279
+    },
+    {
+      "epoch": 0.058115400581154004,
+      "grad_norm": 0.7146359364444479,
+      "learning_rate": 3.970083002324399e-07,
+      "loss": 1.605,
+      "step": 280
+    },
+    {
+      "epoch": 0.05832295558322956,
+      "grad_norm": 1.3831084295560547,
+      "learning_rate": 3.9698695271288185e-07,
+      "loss": 1.5278,
+      "step": 281
+    },
+    {
+      "epoch": 0.058530510585305104,
+      "grad_norm": 0.9856642286692111,
+      "learning_rate": 3.9696552994317025e-07,
+      "loss": 1.6086,
+      "step": 282
+    },
+    {
+      "epoch": 0.05873806558738066,
+      "grad_norm": 0.7068077938134235,
+      "learning_rate": 3.9694403193241346e-07,
+      "loss": 1.6054,
+      "step": 283
+    },
+    {
+      "epoch": 0.0589456205894562,
+      "grad_norm": 1.627631425485338,
+      "learning_rate": 3.969224586897519e-07,
+      "loss": 1.5314,
+      "step": 284
+    },
+    {
+      "epoch": 0.059153175591531756,
+      "grad_norm": 0.7062886743411992,
+      "learning_rate": 3.9690081022435795e-07,
+      "loss": 1.5861,
+      "step": 285
+    },
+    {
+      "epoch": 0.0593607305936073,
+      "grad_norm": 0.789525351577236,
+      "learning_rate": 3.968790865454359e-07,
+      "loss": 1.5406,
+      "step": 286
+    },
+    {
+      "epoch": 0.059568285595682856,
+      "grad_norm": 0.7393779288397713,
+      "learning_rate": 3.968572876622222e-07,
+      "loss": 1.5273,
+      "step": 287
+    },
+    {
+      "epoch": 0.05977584059775841,
+      "grad_norm": 0.6496148935943981,
+      "learning_rate": 3.96835413583985e-07,
+      "loss": 1.486,
+      "step": 288
+    },
+    {
+      "epoch": 0.059983395599833955,
+      "grad_norm": 0.964319646479861,
+      "learning_rate": 3.968134643200247e-07,
+      "loss": 1.5263,
+      "step": 289
+    },
+    {
+      "epoch": 0.06019095060190951,
+      "grad_norm": 0.7009271072255719,
+      "learning_rate": 3.967914398796735e-07,
+      "loss": 1.6022,
+      "step": 290
+    },
+    {
+      "epoch": 0.060398505603985055,
+      "grad_norm": 0.9633499931880358,
+      "learning_rate": 3.9676934027229564e-07,
+      "loss": 1.6112,
+      "step": 291
+    },
+    {
+      "epoch": 0.06060606060606061,
+      "grad_norm": 1.131949634852106,
+      "learning_rate": 3.967471655072872e-07,
+      "loss": 1.5812,
+      "step": 292
+    },
+    {
+      "epoch": 0.060813615608136154,
+      "grad_norm": 0.7650738850666946,
+      "learning_rate": 3.9672491559407636e-07,
+      "loss": 1.6406,
+      "step": 293
+    },
+    {
+      "epoch": 0.06102117061021171,
+      "grad_norm": 0.8482872049035906,
+      "learning_rate": 3.967025905421232e-07,
+      "loss": 1.5406,
+      "step": 294
+    },
+    {
+      "epoch": 0.06122872561228725,
+      "grad_norm": 1.8054256569571165,
+      "learning_rate": 3.966801903609197e-07,
+      "loss": 1.5008,
+      "step": 295
+    },
+    {
+      "epoch": 0.061436280614362807,
+      "grad_norm": 0.7331796700490091,
+      "learning_rate": 3.966577150599899e-07,
+      "loss": 1.6281,
+      "step": 296
+    },
+    {
+      "epoch": 0.06164383561643835,
+      "grad_norm": 1.039524787278702,
+      "learning_rate": 3.966351646488896e-07,
+      "loss": 1.519,
+      "step": 297
+    },
+    {
+      "epoch": 0.061851390618513906,
+      "grad_norm": 0.9016996870861953,
+      "learning_rate": 3.9661253913720684e-07,
+      "loss": 1.525,
+      "step": 298
+    },
+    {
+      "epoch": 0.06205894562058946,
+      "grad_norm": 0.6794441374069633,
+      "learning_rate": 3.965898385345611e-07,
+      "loss": 1.6152,
+      "step": 299
+    },
+    {
+      "epoch": 0.062266500622665005,
+      "grad_norm": 1.3089635118220553,
+      "learning_rate": 3.965670628506042e-07,
+      "loss": 1.6183,
+      "step": 300
+    },
+    {
+      "epoch": 0.06247405562474056,
+      "grad_norm": 1.2221544240377789,
+      "learning_rate": 3.965442120950198e-07,
+      "loss": 1.6397,
+      "step": 301
+    },
+    {
+      "epoch": 0.06268161062681611,
+      "grad_norm": 0.6463045173637169,
+      "learning_rate": 3.9652128627752337e-07,
+      "loss": 1.5291,
+      "step": 302
+    },
+    {
+      "epoch": 0.06288916562889166,
+      "grad_norm": 0.9212512453584831,
+      "learning_rate": 3.9649828540786247e-07,
+      "loss": 1.5862,
+      "step": 303
+    },
+    {
+      "epoch": 0.0630967206309672,
+      "grad_norm": 0.9701094676987546,
+      "learning_rate": 3.964752094958163e-07,
+      "loss": 1.5255,
+      "step": 304
+    },
+    {
+      "epoch": 0.06330427563304275,
+      "grad_norm": 0.8050839790809957,
+      "learning_rate": 3.964520585511962e-07,
+      "loss": 1.566,
+      "step": 305
+    },
+    {
+      "epoch": 0.06351183063511831,
+      "grad_norm": 1.787638326017031,
+      "learning_rate": 3.964288325838454e-07,
+      "loss": 1.6107,
+      "step": 306
+    },
+    {
+      "epoch": 0.06371938563719386,
+      "grad_norm": 1.514367611958326,
+      "learning_rate": 3.964055316036388e-07,
+      "loss": 1.5261,
+      "step": 307
+    },
+    {
+      "epoch": 0.0639269406392694,
+      "grad_norm": 0.7560770150906774,
+      "learning_rate": 3.9638215562048355e-07,
+      "loss": 1.5291,
+      "step": 308
+    },
+    {
+      "epoch": 0.06413449564134496,
+      "grad_norm": 0.7726530494604988,
+      "learning_rate": 3.9635870464431837e-07,
+      "loss": 1.6066,
+      "step": 309
+    },
+    {
+      "epoch": 0.06434205064342051,
+      "grad_norm": 1.3009761711129673,
+      "learning_rate": 3.9633517868511407e-07,
+      "loss": 1.5895,
+      "step": 310
+    },
+    {
+      "epoch": 0.06454960564549606,
+      "grad_norm": 1.4546554518052999,
+      "learning_rate": 3.963115777528732e-07,
+      "loss": 1.5441,
+      "step": 311
+    },
+    {
+      "epoch": 0.0647571606475716,
+      "grad_norm": 1.5454337067593822,
+      "learning_rate": 3.962879018576303e-07,
+      "loss": 1.5033,
+      "step": 312
+    },
+    {
+      "epoch": 0.06496471564964716,
+      "grad_norm": 0.8111037531155808,
+      "learning_rate": 3.962641510094517e-07,
+      "loss": 1.5749,
+      "step": 313
+    },
+    {
+      "epoch": 0.06517227065172271,
+      "grad_norm": 2.096194509213919,
+      "learning_rate": 3.9624032521843563e-07,
+      "loss": 1.5651,
+      "step": 314
+    },
+    {
+      "epoch": 0.06537982565379825,
+      "grad_norm": 0.925326041461294,
+      "learning_rate": 3.962164244947122e-07,
+      "loss": 1.5303,
+      "step": 315
+    },
+    {
+      "epoch": 0.0655873806558738,
+      "grad_norm": 0.8005797991044269,
+      "learning_rate": 3.9619244884844335e-07,
+      "loss": 1.664,
+      "step": 316
+    },
+    {
+      "epoch": 0.06579493565794936,
+      "grad_norm": 1.1962393224185432,
+      "learning_rate": 3.9616839828982285e-07,
+      "loss": 1.6106,
+      "step": 317
+    },
+    {
+      "epoch": 0.0660024906600249,
+      "grad_norm": 0.7546956822237012,
+      "learning_rate": 3.9614427282907647e-07,
+      "loss": 1.5852,
+      "step": 318
+    },
+    {
+      "epoch": 0.06621004566210045,
+      "grad_norm": 0.8854439260997882,
+      "learning_rate": 3.961200724764616e-07,
+      "loss": 1.5192,
+      "step": 319
+    },
+    {
+      "epoch": 0.06641760066417601,
+      "grad_norm": 0.7476814369693596,
+      "learning_rate": 3.9609579724226763e-07,
+      "loss": 1.5597,
+      "step": 320
+    },
+    {
+      "epoch": 0.06662515566625156,
+      "grad_norm": 1.220053966221941,
+      "learning_rate": 3.960714471368158e-07,
+      "loss": 1.5834,
+      "step": 321
+    },
+    {
+      "epoch": 0.0668327106683271,
+      "grad_norm": 0.7171016406798384,
+      "learning_rate": 3.9604702217045903e-07,
+      "loss": 1.4922,
+      "step": 322
+    },
+    {
+      "epoch": 0.06704026567040265,
+      "grad_norm": 1.1864436162880359,
+      "learning_rate": 3.9602252235358227e-07,
+      "loss": 1.5939,
+      "step": 323
+    },
+    {
+      "epoch": 0.06724782067247821,
+      "grad_norm": 0.7851629580390745,
+      "learning_rate": 3.9599794769660214e-07,
+      "loss": 1.5338,
+      "step": 324
+    },
+    {
+      "epoch": 0.06745537567455376,
+      "grad_norm": 0.7731731941124776,
+      "learning_rate": 3.9597329820996704e-07,
+      "loss": 1.5185,
+      "step": 325
+    },
+    {
+      "epoch": 0.0676629306766293,
+      "grad_norm": 1.1481780697317034,
+      "learning_rate": 3.9594857390415744e-07,
+      "loss": 1.6273,
+      "step": 326
+    },
+    {
+      "epoch": 0.06787048567870485,
+      "grad_norm": 1.231706771563639,
+      "learning_rate": 3.9592377478968537e-07,
+      "loss": 1.5355,
+      "step": 327
+    },
+    {
+      "epoch": 0.06807804068078041,
+      "grad_norm": 0.791385321452644,
+      "learning_rate": 3.9589890087709475e-07,
+      "loss": 1.5941,
+      "step": 328
+    },
+    {
+      "epoch": 0.06828559568285596,
+      "grad_norm": 0.9820296859582639,
+      "learning_rate": 3.958739521769614e-07,
+      "loss": 1.4661,
+      "step": 329
+    },
+    {
+      "epoch": 0.0684931506849315,
+      "grad_norm": 0.9345534758361532,
+      "learning_rate": 3.958489286998927e-07,
+      "loss": 1.5782,
+      "step": 330
+    },
+    {
+      "epoch": 0.06870070568700706,
+      "grad_norm": 0.994239478069683,
+      "learning_rate": 3.958238304565281e-07,
+      "loss": 1.5355,
+      "step": 331
+    },
+    {
+      "epoch": 0.06890826068908261,
+      "grad_norm": 0.9422649938094205,
+      "learning_rate": 3.9579865745753854e-07,
+      "loss": 1.5736,
+      "step": 332
+    },
+    {
+      "epoch": 0.06911581569115816,
+      "grad_norm": 0.9246351674417703,
+      "learning_rate": 3.957734097136271e-07,
+      "loss": 1.4708,
+      "step": 333
+    },
+    {
+      "epoch": 0.0693233706932337,
+      "grad_norm": 2.569793437141846,
+      "learning_rate": 3.9574808723552834e-07,
+      "loss": 1.4702,
+      "step": 334
+    },
+    {
+      "epoch": 0.06953092569530926,
+      "grad_norm": 0.6471573546153861,
+      "learning_rate": 3.9572269003400876e-07,
+      "loss": 1.578,
+      "step": 335
+    },
+    {
+      "epoch": 0.06973848069738481,
+      "grad_norm": 0.7277938493644751,
+      "learning_rate": 3.9569721811986654e-07,
+      "loss": 1.5473,
+      "step": 336
+    },
+    {
+      "epoch": 0.06994603569946035,
+      "grad_norm": 0.783551313956181,
+      "learning_rate": 3.9567167150393163e-07,
+      "loss": 1.5488,
+      "step": 337
+    },
+    {
+      "epoch": 0.0701535907015359,
+      "grad_norm": 0.7109255168982866,
+      "learning_rate": 3.9564605019706586e-07,
+      "loss": 1.5575,
+      "step": 338
+    },
+    {
+      "epoch": 0.07036114570361146,
+      "grad_norm": 0.7915619295057836,
+      "learning_rate": 3.956203542101627e-07,
+      "loss": 1.6001,
+      "step": 339
+    },
+    {
+      "epoch": 0.070568700705687,
+      "grad_norm": 0.9391306518808272,
+      "learning_rate": 3.9559458355414734e-07,
+      "loss": 1.5346,
+      "step": 340
+    },
+    {
+      "epoch": 0.07077625570776255,
+      "grad_norm": 1.0878389625543508,
+      "learning_rate": 3.955687382399769e-07,
+      "loss": 1.5859,
+      "step": 341
+    },
+    {
+      "epoch": 0.07098381070983811,
+      "grad_norm": 2.529640080811518,
+      "learning_rate": 3.955428182786399e-07,
+      "loss": 1.5574,
+      "step": 342
+    },
+    {
+      "epoch": 0.07119136571191366,
+      "grad_norm": 0.7024340448196,
+      "learning_rate": 3.9551682368115706e-07,
+      "loss": 1.6081,
+      "step": 343
+    },
+    {
+      "epoch": 0.0713989207139892,
+      "grad_norm": 0.9818175237257888,
+      "learning_rate": 3.954907544585805e-07,
+      "loss": 1.57,
+      "step": 344
+    },
+    {
+      "epoch": 0.07160647571606475,
+      "grad_norm": 1.4630432972328837,
+      "learning_rate": 3.954646106219942e-07,
+      "loss": 1.5334,
+      "step": 345
+    },
+    {
+      "epoch": 0.07181403071814031,
+      "grad_norm": 1.3627602266113426,
+      "learning_rate": 3.9543839218251367e-07,
+      "loss": 1.4364,
+      "step": 346
+    },
+    {
+      "epoch": 0.07202158572021586,
+      "grad_norm": 0.7473269943068941,
+      "learning_rate": 3.954120991512865e-07,
+      "loss": 1.5877,
+      "step": 347
+    },
+    {
+      "epoch": 0.0722291407222914,
+      "grad_norm": 0.8405582475412623,
+      "learning_rate": 3.9538573153949166e-07,
+      "loss": 1.4467,
+      "step": 348
+    },
+    {
+      "epoch": 0.07243669572436695,
+      "grad_norm": 1.1594318553688343,
+      "learning_rate": 3.9535928935834e-07,
+      "loss": 1.6009,
+      "step": 349
+    },
+    {
+      "epoch": 0.07264425072644251,
+      "grad_norm": 1.2595260213243793,
+      "learning_rate": 3.9533277261907407e-07,
+      "loss": 1.5722,
+      "step": 350
+    },
+    {
+      "epoch": 0.07285180572851806,
+      "grad_norm": 1.0483692983589734,
+      "learning_rate": 3.9530618133296804e-07,
+      "loss": 1.6056,
+      "step": 351
+    },
+    {
+      "epoch": 0.0730593607305936,
+      "grad_norm": 1.878051520630954,
+      "learning_rate": 3.952795155113277e-07,
+      "loss": 1.5047,
+      "step": 352
+    },
+    {
+      "epoch": 0.07326691573266916,
+      "grad_norm": 0.8285069119449421,
+      "learning_rate": 3.9525277516549087e-07,
+      "loss": 1.5505,
+      "step": 353
+    },
+    {
+      "epoch": 0.07347447073474471,
+      "grad_norm": 1.0179567502745943,
+      "learning_rate": 3.952259603068267e-07,
+      "loss": 1.5815,
+      "step": 354
+    },
+    {
+      "epoch": 0.07368202573682026,
+      "grad_norm": 0.8437086108640751,
+      "learning_rate": 3.951990709467363e-07,
+      "loss": 1.6478,
+      "step": 355
+    },
+    {
+      "epoch": 0.0738895807388958,
+      "grad_norm": 0.7744403790360274,
+      "learning_rate": 3.951721070966521e-07,
+      "loss": 1.654,
+      "step": 356
+    },
+    {
+      "epoch": 0.07409713574097136,
+      "grad_norm": 0.9946033853448627,
+      "learning_rate": 3.9514506876803854e-07,
+      "loss": 1.5852,
+      "step": 357
+    },
+    {
+      "epoch": 0.07430469074304691,
+      "grad_norm": 1.2992465449230606,
+      "learning_rate": 3.9511795597239155e-07,
+      "loss": 1.6225,
+      "step": 358
+    },
+    {
+      "epoch": 0.07451224574512245,
+      "grad_norm": 0.6611632078314491,
+      "learning_rate": 3.9509076872123887e-07,
+      "loss": 1.5222,
+      "step": 359
+    },
+    {
+      "epoch": 0.074719800747198,
+      "grad_norm": 0.8554859218864674,
+      "learning_rate": 3.9506350702613966e-07,
+      "loss": 1.6026,
+      "step": 360
+    },
+    {
+      "epoch": 0.07492735574927356,
+      "grad_norm": 0.6806192227148197,
+      "learning_rate": 3.9503617089868496e-07,
+      "loss": 1.5226,
+      "step": 361
+    },
+    {
+      "epoch": 0.07513491075134911,
+      "grad_norm": 0.9228050332082446,
+      "learning_rate": 3.9500876035049735e-07,
+      "loss": 1.6104,
+      "step": 362
+    },
+    {
+      "epoch": 0.07534246575342465,
+      "grad_norm": 1.0163593089389478,
+      "learning_rate": 3.9498127539323105e-07,
+      "loss": 1.5641,
+      "step": 363
+    },
+    {
+      "epoch": 0.07555002075550021,
+      "grad_norm": 1.4912559905551543,
+      "learning_rate": 3.9495371603857193e-07,
+      "loss": 1.56,
+      "step": 364
+    },
+    {
+      "epoch": 0.07575757575757576,
+      "grad_norm": 0.8892116712870853,
+      "learning_rate": 3.9492608229823753e-07,
+      "loss": 1.5432,
+      "step": 365
+    },
+    {
+      "epoch": 0.0759651307596513,
+      "grad_norm": 0.9629533423498724,
+      "learning_rate": 3.9489837418397693e-07,
+      "loss": 1.5888,
+      "step": 366
+    },
+    {
+      "epoch": 0.07617268576172685,
+      "grad_norm": 0.655194629821129,
+      "learning_rate": 3.948705917075709e-07,
+      "loss": 1.5458,
+      "step": 367
+    },
+    {
+      "epoch": 0.07638024076380241,
+      "grad_norm": 1.7481375792049714,
+      "learning_rate": 3.9484273488083186e-07,
+      "loss": 1.5989,
+      "step": 368
+    },
+    {
+      "epoch": 0.07658779576587796,
+      "grad_norm": 0.695241734120242,
+      "learning_rate": 3.9481480371560375e-07,
+      "loss": 1.5241,
+      "step": 369
+    },
+    {
+      "epoch": 0.0767953507679535,
+      "grad_norm": 0.9343488497549051,
+      "learning_rate": 3.9478679822376216e-07,
+      "loss": 1.5785,
+      "step": 370
+    },
+    {
+      "epoch": 0.07700290577002905,
+      "grad_norm": 0.7423150462414291,
+      "learning_rate": 3.947587184172143e-07,
+      "loss": 1.6286,
+      "step": 371
+    },
+    {
+      "epoch": 0.07721046077210461,
+      "grad_norm": 0.7747788070171346,
+      "learning_rate": 3.9473056430789893e-07,
+      "loss": 1.5631,
+      "step": 372
+    },
+    {
+      "epoch": 0.07741801577418016,
+      "grad_norm": 0.7845084852813077,
+      "learning_rate": 3.947023359077865e-07,
+      "loss": 1.5857,
+      "step": 373
+    },
+    {
+      "epoch": 0.0776255707762557,
+      "grad_norm": 1.3942838974429645,
+      "learning_rate": 3.946740332288789e-07,
+      "loss": 1.5031,
+      "step": 374
+    },
+    {
+      "epoch": 0.07783312577833126,
+      "grad_norm": 0.6363998599574435,
+      "learning_rate": 3.9464565628320967e-07,
+      "loss": 1.5853,
+      "step": 375
+    },
+    {
+      "epoch": 0.07804068078040681,
+      "grad_norm": 0.7292506159159087,
+      "learning_rate": 3.94617205082844e-07,
+      "loss": 1.5485,
+      "step": 376
+    },
+    {
+      "epoch": 0.07824823578248236,
+      "grad_norm": 1.4592810508360627,
+      "learning_rate": 3.9458867963987856e-07,
+      "loss": 1.5113,
+      "step": 377
+    },
+    {
+      "epoch": 0.0784557907845579,
+      "grad_norm": 0.8599217550041893,
+      "learning_rate": 3.945600799664416e-07,
+      "loss": 1.5348,
+      "step": 378
+    },
+    {
+      "epoch": 0.07866334578663346,
+      "grad_norm": 0.6925878570206188,
+      "learning_rate": 3.94531406074693e-07,
+      "loss": 1.4851,
+      "step": 379
+    },
+    {
+      "epoch": 0.07887090078870901,
+      "grad_norm": 0.9552351209400144,
+      "learning_rate": 3.9450265797682396e-07,
+      "loss": 1.5127,
+      "step": 380
+    },
+    {
+      "epoch": 0.07907845579078455,
+      "grad_norm": 0.7312953405290957,
+      "learning_rate": 3.944738356850576e-07,
+      "loss": 1.5772,
+      "step": 381
+    },
+    {
+      "epoch": 0.0792860107928601,
+      "grad_norm": 0.686828088243894,
+      "learning_rate": 3.944449392116483e-07,
+      "loss": 1.4996,
+      "step": 382
+    },
+    {
+      "epoch": 0.07949356579493566,
+      "grad_norm": 1.002100881776077,
+      "learning_rate": 3.944159685688821e-07,
+      "loss": 1.4803,
+      "step": 383
+    },
+    {
+      "epoch": 0.07970112079701121,
+      "grad_norm": 1.274816320045783,
+      "learning_rate": 3.9438692376907657e-07,
+      "loss": 1.5238,
+      "step": 384
+    },
+    {
+      "epoch": 0.07990867579908675,
+      "grad_norm": 1.550214411197294,
+      "learning_rate": 3.943578048245807e-07,
+      "loss": 1.5407,
+      "step": 385
+    },
+    {
+      "epoch": 0.08011623080116231,
+      "grad_norm": 0.8221018746296057,
+      "learning_rate": 3.9432861174777525e-07,
+      "loss": 1.4838,
+      "step": 386
+    },
+    {
+      "epoch": 0.08032378580323786,
+      "grad_norm": 1.0610281147160094,
+      "learning_rate": 3.942993445510722e-07,
+      "loss": 1.5784,
+      "step": 387
+    },
+    {
+      "epoch": 0.0805313408053134,
+      "grad_norm": 1.6677082247902724,
+      "learning_rate": 3.9427000324691525e-07,
+      "loss": 1.548,
+      "step": 388
+    },
+    {
+      "epoch": 0.08073889580738895,
+      "grad_norm": 0.6911748828478762,
+      "learning_rate": 3.942405878477795e-07,
+      "loss": 1.538,
+      "step": 389
+    },
+    {
+      "epoch": 0.08094645080946451,
+      "grad_norm": 0.6959081935320677,
+      "learning_rate": 3.942110983661716e-07,
+      "loss": 1.6123,
+      "step": 390
+    },
+    {
+      "epoch": 0.08115400581154006,
+      "grad_norm": 4.6744001253512275,
+      "learning_rate": 3.9418153481462976e-07,
+      "loss": 1.5313,
+      "step": 391
+    },
+    {
+      "epoch": 0.0813615608136156,
+      "grad_norm": 0.8394496305680753,
+      "learning_rate": 3.941518972057235e-07,
+      "loss": 1.6077,
+      "step": 392
+    },
+    {
+      "epoch": 0.08156911581569116,
+      "grad_norm": 0.6411844023965299,
+      "learning_rate": 3.941221855520541e-07,
+      "loss": 1.5269,
+      "step": 393
+    },
+    {
+      "epoch": 0.08177667081776671,
+      "grad_norm": 4.091319459987459,
+      "learning_rate": 3.9409239986625405e-07,
+      "loss": 1.5721,
+      "step": 394
+    },
+    {
+      "epoch": 0.08198422581984226,
+      "grad_norm": 0.7884330335134523,
+      "learning_rate": 3.940625401609875e-07,
+      "loss": 1.5475,
+      "step": 395
+    },
+    {
+      "epoch": 0.0821917808219178,
+      "grad_norm": 0.78314552030894,
+      "learning_rate": 3.9403260644894993e-07,
+      "loss": 1.6138,
+      "step": 396
+    },
+    {
+      "epoch": 0.08239933582399336,
+      "grad_norm": 0.7622111269435049,
+      "learning_rate": 3.9400259874286844e-07,
+      "loss": 1.5111,
+      "step": 397
+    },
+    {
+      "epoch": 0.08260689082606891,
+      "grad_norm": 0.819239164663395,
+      "learning_rate": 3.9397251705550146e-07,
+      "loss": 1.5822,
+      "step": 398
+    },
+    {
+      "epoch": 0.08281444582814446,
+      "grad_norm": 0.7302769195935753,
+      "learning_rate": 3.9394236139963886e-07,
+      "loss": 1.5204,
+      "step": 399
+    },
+    {
+      "epoch": 0.08302200083022,
+      "grad_norm": 0.9658003009694968,
+      "learning_rate": 3.9391213178810223e-07,
+      "loss": 1.5423,
+      "step": 400
+    },
+    {
+      "epoch": 0.08322955583229556,
+      "grad_norm": 0.8609973872812837,
+      "learning_rate": 3.938818282337442e-07,
+      "loss": 1.6461,
+      "step": 401
+    },
+    {
+      "epoch": 0.08343711083437111,
+      "grad_norm": 0.9783977239516148,
+      "learning_rate": 3.938514507494491e-07,
+      "loss": 1.5774,
+      "step": 402
+    },
+    {
+      "epoch": 0.08364466583644665,
+      "grad_norm": 1.056699080196957,
+      "learning_rate": 3.9382099934813265e-07,
+      "loss": 1.5597,
+      "step": 403
+    },
+    {
+      "epoch": 0.08385222083852222,
+      "grad_norm": 0.7220006838934228,
+      "learning_rate": 3.937904740427419e-07,
+      "loss": 1.4923,
+      "step": 404
+    },
+    {
+      "epoch": 0.08405977584059776,
+      "grad_norm": 0.8430739116359129,
+      "learning_rate": 3.9375987484625555e-07,
+      "loss": 1.5426,
+      "step": 405
+    },
+    {
+      "epoch": 0.08426733084267331,
+      "grad_norm": 3.613166766223418,
+      "learning_rate": 3.937292017716834e-07,
+      "loss": 1.544,
+      "step": 406
+    },
+    {
+      "epoch": 0.08447488584474885,
+      "grad_norm": 0.9376576443914243,
+      "learning_rate": 3.93698454832067e-07,
+      "loss": 1.5918,
+      "step": 407
+    },
+    {
+      "epoch": 0.08468244084682441,
+      "grad_norm": 0.9094639893253319,
+      "learning_rate": 3.9366763404047896e-07,
+      "loss": 1.5672,
+      "step": 408
+    },
+    {
+      "epoch": 0.08488999584889996,
+      "grad_norm": 0.9663058776634983,
+      "learning_rate": 3.9363673941002366e-07,
+      "loss": 1.5924,
+      "step": 409
+    },
+    {
+      "epoch": 0.0850975508509755,
+      "grad_norm": 0.8002328143803354,
+      "learning_rate": 3.9360577095383644e-07,
+      "loss": 1.5445,
+      "step": 410
+    },
+    {
+      "epoch": 0.08530510585305105,
+      "grad_norm": 0.8623858127380615,
+      "learning_rate": 3.935747286850843e-07,
+      "loss": 1.5918,
+      "step": 411
+    },
+    {
+      "epoch": 0.08551266085512661,
+      "grad_norm": 0.7089643849188245,
+      "learning_rate": 3.935436126169658e-07,
+      "loss": 1.61,
+      "step": 412
+    },
+    {
+      "epoch": 0.08572021585720216,
+      "grad_norm": 0.7746493584397219,
+      "learning_rate": 3.935124227627105e-07,
+      "loss": 1.5202,
+      "step": 413
+    },
+    {
+      "epoch": 0.0859277708592777,
+      "grad_norm": 1.2816563699568104,
+      "learning_rate": 3.934811591355796e-07,
+      "loss": 1.5418,
+      "step": 414
+    },
+    {
+      "epoch": 0.08613532586135327,
+      "grad_norm": 0.9240654353762444,
+      "learning_rate": 3.934498217488654e-07,
+      "loss": 1.611,
+      "step": 415
+    },
+    {
+      "epoch": 0.08634288086342881,
+      "grad_norm": 2.1635633770292415,
+      "learning_rate": 3.934184106158919e-07,
+      "loss": 1.6338,
+      "step": 416
+    },
+    {
+      "epoch": 0.08655043586550436,
+      "grad_norm": 0.6485865496749437,
+      "learning_rate": 3.933869257500142e-07,
+      "loss": 1.5616,
+      "step": 417
+    },
+    {
+      "epoch": 0.0867579908675799,
+      "grad_norm": 0.6615085025280788,
+      "learning_rate": 3.933553671646188e-07,
+      "loss": 1.5267,
+      "step": 418
+    },
+    {
+      "epoch": 0.08696554586965546,
+      "grad_norm": 1.0369129853017083,
+      "learning_rate": 3.933237348731236e-07,
+      "loss": 1.5436,
+      "step": 419
+    },
+    {
+      "epoch": 0.08717310087173101,
+      "grad_norm": 1.1918789084504378,
+      "learning_rate": 3.932920288889778e-07,
+      "loss": 1.5543,
+      "step": 420
+    },
+    {
+      "epoch": 0.08738065587380656,
+      "grad_norm": 0.7387501874317923,
+      "learning_rate": 3.93260249225662e-07,
+      "loss": 1.5508,
+      "step": 421
+    },
+    {
+      "epoch": 0.0875882108758821,
+      "grad_norm": 0.8402063796071437,
+      "learning_rate": 3.93228395896688e-07,
+      "loss": 1.5808,
+      "step": 422
+    },
+    {
+      "epoch": 0.08779576587795766,
+      "grad_norm": 1.3213676966046553,
+      "learning_rate": 3.93196468915599e-07,
+      "loss": 1.6091,
+      "step": 423
+    },
+    {
+      "epoch": 0.08800332088003321,
+      "grad_norm": 0.8763352766282972,
+      "learning_rate": 3.931644682959696e-07,
+      "loss": 1.5402,
+      "step": 424
+    },
+    {
+      "epoch": 0.08821087588210876,
+      "grad_norm": 1.3891025379335946,
+      "learning_rate": 3.9313239405140545e-07,
+      "loss": 1.5404,
+      "step": 425
+    },
+    {
+      "epoch": 0.08841843088418432,
+      "grad_norm": 0.7251642927038664,
+      "learning_rate": 3.931002461955438e-07,
+      "loss": 1.4728,
+      "step": 426
+    },
+    {
+      "epoch": 0.08862598588625986,
+      "grad_norm": 0.8837030874737397,
+      "learning_rate": 3.9306802474205305e-07,
+      "loss": 1.5991,
+      "step": 427
+    },
+    {
+      "epoch": 0.08883354088833541,
+      "grad_norm": 0.6507271181113026,
+      "learning_rate": 3.9303572970463283e-07,
+      "loss": 1.5299,
+      "step": 428
+    },
+    {
+      "epoch": 0.08904109589041095,
+      "grad_norm": 0.6593824530272688,
+      "learning_rate": 3.930033610970141e-07,
+      "loss": 1.5542,
+      "step": 429
+    },
+    {
+      "epoch": 0.08924865089248651,
+      "grad_norm": 0.7472039099118118,
+      "learning_rate": 3.929709189329593e-07,
+      "loss": 1.5415,
+      "step": 430
+    },
+    {
+      "epoch": 0.08945620589456206,
+      "grad_norm": 1.7347531358597275,
+      "learning_rate": 3.929384032262619e-07,
+      "loss": 1.5233,
+      "step": 431
+    },
+    {
+      "epoch": 0.0896637608966376,
+      "grad_norm": 0.9845210129314712,
+      "learning_rate": 3.929058139907467e-07,
+      "loss": 1.5762,
+      "step": 432
+    },
+    {
+      "epoch": 0.08987131589871315,
+      "grad_norm": 0.820894568693371,
+      "learning_rate": 3.9287315124026973e-07,
+      "loss": 1.5065,
+      "step": 433
+    },
+    {
+      "epoch": 0.09007887090078871,
+      "grad_norm": 0.8277047313148947,
+      "learning_rate": 3.9284041498871835e-07,
+      "loss": 1.5968,
+      "step": 434
+    },
+    {
+      "epoch": 0.09028642590286426,
+      "grad_norm": 0.8556145328957938,
+      "learning_rate": 3.9280760525001123e-07,
+      "loss": 1.5912,
+      "step": 435
+    },
+    {
+      "epoch": 0.0904939809049398,
+      "grad_norm": 0.6697043902153069,
+      "learning_rate": 3.9277472203809813e-07,
+      "loss": 1.627,
+      "step": 436
+    },
+    {
+      "epoch": 0.09070153590701537,
+      "grad_norm": 0.9233328122184589,
+      "learning_rate": 3.927417653669601e-07,
+      "loss": 1.584,
+      "step": 437
+    },
+    {
+      "epoch": 0.09090909090909091,
+      "grad_norm": 0.9039799341906642,
+      "learning_rate": 3.9270873525060956e-07,
+      "loss": 1.6087,
+      "step": 438
+    },
+    {
+      "epoch": 0.09111664591116646,
+      "grad_norm": 0.7901578512951782,
+      "learning_rate": 3.9267563170308984e-07,
+      "loss": 1.4938,
+      "step": 439
+    },
+    {
+      "epoch": 0.091324200913242,
+      "grad_norm": 0.7466405510809797,
+      "learning_rate": 3.9264245473847584e-07,
+      "loss": 1.5476,
+      "step": 440
+    },
+    {
+      "epoch": 0.09153175591531756,
+      "grad_norm": 0.7505865861833354,
+      "learning_rate": 3.9260920437087347e-07,
+      "loss": 1.5833,
+      "step": 441
+    },
+    {
+      "epoch": 0.09173931091739311,
+      "grad_norm": 0.7746214690829071,
+      "learning_rate": 3.9257588061441993e-07,
+      "loss": 1.5374,
+      "step": 442
+    },
+    {
+      "epoch": 0.09194686591946866,
+      "grad_norm": 0.634379651499812,
+      "learning_rate": 3.925424834832835e-07,
+      "loss": 1.6653,
+      "step": 443
+    },
+    {
+      "epoch": 0.0921544209215442,
+      "grad_norm": 0.8289695484975603,
+      "learning_rate": 3.92509012991664e-07,
+      "loss": 1.523,
+      "step": 444
+    },
+    {
+      "epoch": 0.09236197592361976,
+      "grad_norm": 0.8986529863797156,
+      "learning_rate": 3.9247546915379186e-07,
+      "loss": 1.5213,
+      "step": 445
+    },
+    {
+      "epoch": 0.09256953092569531,
+      "grad_norm": 0.6947621755736146,
+      "learning_rate": 3.9244185198392933e-07,
+      "loss": 1.5785,
+      "step": 446
+    },
+    {
+      "epoch": 0.09277708592777086,
+      "grad_norm": 2.047831976460464,
+      "learning_rate": 3.9240816149636936e-07,
+      "loss": 1.5782,
+      "step": 447
+    },
+    {
+      "epoch": 0.09298464092984642,
+      "grad_norm": 0.6892096460532096,
+      "learning_rate": 3.923743977054363e-07,
+      "loss": 1.566,
+      "step": 448
+    },
+    {
+      "epoch": 0.09319219593192196,
+      "grad_norm": 3.6873811616487595,
+      "learning_rate": 3.923405606254856e-07,
+      "loss": 1.5002,
+      "step": 449
+    },
+    {
+      "epoch": 0.09339975093399751,
+      "grad_norm": 0.630593927946454,
+      "learning_rate": 3.92306650270904e-07,
+      "loss": 1.5924,
+      "step": 450
+    },
+    {
+      "epoch": 0.09360730593607305,
+      "grad_norm": 0.7655534901138857,
+      "learning_rate": 3.92272666656109e-07,
+      "loss": 1.5677,
+      "step": 451
+    },
+    {
+      "epoch": 0.09381486093814861,
+      "grad_norm": 1.9253770300963382,
+      "learning_rate": 3.9223860979554987e-07,
+      "loss": 1.5153,
+      "step": 452
+    },
+    {
+      "epoch": 0.09402241594022416,
+      "grad_norm": 1.444757510916554,
+      "learning_rate": 3.922044797037064e-07,
+      "loss": 1.56,
+      "step": 453
+    },
+    {
+      "epoch": 0.0942299709422997,
+      "grad_norm": 0.8291854720564956,
+      "learning_rate": 3.9217027639509004e-07,
+      "loss": 1.5342,
+      "step": 454
+    },
+    {
+      "epoch": 0.09443752594437525,
+      "grad_norm": 0.896395271499777,
+      "learning_rate": 3.92135999884243e-07,
+      "loss": 1.5403,
+      "step": 455
+    },
+    {
+      "epoch": 0.09464508094645081,
+      "grad_norm": 0.7046816160087109,
+      "learning_rate": 3.9210165018573874e-07,
+      "loss": 1.5961,
+      "step": 456
+    },
+    {
+      "epoch": 0.09485263594852636,
+      "grad_norm": 1.4111738556351614,
+      "learning_rate": 3.9206722731418187e-07,
+      "loss": 1.5996,
+      "step": 457
+    },
+    {
+      "epoch": 0.0950601909506019,
+      "grad_norm": 0.707850415479939,
+      "learning_rate": 3.9203273128420804e-07,
+      "loss": 1.6089,
+      "step": 458
+    },
+    {
+      "epoch": 0.09526774595267747,
+      "grad_norm": 0.737125279192193,
+      "learning_rate": 3.919981621104841e-07,
+      "loss": 1.6001,
+      "step": 459
+    },
+    {
+      "epoch": 0.09547530095475301,
+      "grad_norm": 1.0734406504459013,
+      "learning_rate": 3.9196351980770794e-07,
+      "loss": 1.5131,
+      "step": 460
+    },
+    {
+      "epoch": 0.09568285595682856,
+      "grad_norm": 1.2396986372388994,
+      "learning_rate": 3.9192880439060855e-07,
+      "loss": 1.5442,
+      "step": 461
+    },
+    {
+      "epoch": 0.0958904109589041,
+      "grad_norm": 0.769630394005304,
+      "learning_rate": 3.91894015873946e-07,
+      "loss": 1.5468,
+      "step": 462
+    },
+    {
+      "epoch": 0.09609796596097966,
+      "grad_norm": 0.6788045988237913,
+      "learning_rate": 3.9185915427251127e-07,
+      "loss": 1.5474,
+      "step": 463
+    },
+    {
+      "epoch": 0.09630552096305521,
+      "grad_norm": 0.7232298207770542,
+      "learning_rate": 3.9182421960112687e-07,
+      "loss": 1.5568,
+      "step": 464
+    },
+    {
+      "epoch": 0.09651307596513076,
+      "grad_norm": 1.0194928330350024,
+      "learning_rate": 3.91789211874646e-07,
+      "loss": 1.5612,
+      "step": 465
+    },
+    {
+      "epoch": 0.0967206309672063,
+      "grad_norm": 1.886389088624282,
+      "learning_rate": 3.917541311079529e-07,
+      "loss": 1.533,
+      "step": 466
+    },
+    {
+      "epoch": 0.09692818596928186,
+      "grad_norm": 0.72220049838547,
+      "learning_rate": 3.917189773159631e-07,
+      "loss": 1.5304,
+      "step": 467
+    },
+    {
+      "epoch": 0.09713574097135741,
+      "grad_norm": 0.8774491128392785,
+      "learning_rate": 3.91683750513623e-07,
+      "loss": 1.5579,
+      "step": 468
+    },
+    {
+      "epoch": 0.09734329597343296,
+      "grad_norm": 1.1621072478103602,
+      "learning_rate": 3.916484507159101e-07,
+      "loss": 1.6187,
+      "step": 469
+    },
+    {
+      "epoch": 0.09755085097550852,
+      "grad_norm": 1.26745053442655,
+      "learning_rate": 3.9161307793783307e-07,
+      "loss": 1.556,
+      "step": 470
+    },
+    {
+      "epoch": 0.09775840597758406,
+      "grad_norm": 0.8241008796622631,
+      "learning_rate": 3.9157763219443133e-07,
+      "loss": 1.535,
+      "step": 471
+    },
+    {
+      "epoch": 0.09796596097965961,
+      "grad_norm": 1.057297813454511,
+      "learning_rate": 3.9154211350077547e-07,
+      "loss": 1.5801,
+      "step": 472
+    },
+    {
+      "epoch": 0.09817351598173515,
+      "grad_norm": 1.1276494790143259,
+      "learning_rate": 3.915065218719672e-07,
+      "loss": 1.613,
+      "step": 473
+    },
+    {
+      "epoch": 0.09838107098381071,
+      "grad_norm": 1.0677648323961761,
+      "learning_rate": 3.9147085732313903e-07,
+      "loss": 1.538,
+      "step": 474
+    },
+    {
+      "epoch": 0.09858862598588626,
+      "grad_norm": 0.7344424296296667,
+      "learning_rate": 3.914351198694546e-07,
+      "loss": 1.6277,
+      "step": 475
+    },
+    {
+      "epoch": 0.0987961809879618,
+      "grad_norm": 0.7417892882452545,
+      "learning_rate": 3.9139930952610853e-07,
+      "loss": 1.5946,
+      "step": 476
+    },
+    {
+      "epoch": 0.09900373599003735,
+      "grad_norm": 0.7502122706201256,
+      "learning_rate": 3.9136342630832647e-07,
+      "loss": 1.5773,
+      "step": 477
+    },
+    {
+      "epoch": 0.09921129099211291,
+      "grad_norm": 0.8505035409534525,
+      "learning_rate": 3.9132747023136496e-07,
+      "loss": 1.5699,
+      "step": 478
+    },
+    {
+      "epoch": 0.09941884599418846,
+      "grad_norm": 0.7733069680646283,
+      "learning_rate": 3.9129144131051163e-07,
+      "loss": 1.5438,
+      "step": 479
+    },
+    {
+      "epoch": 0.099626400996264,
+      "grad_norm": 0.8433486036854327,
+      "learning_rate": 3.9125533956108495e-07,
+      "loss": 1.5143,
+      "step": 480
+    },
+    {
+      "epoch": 0.09983395599833957,
+      "grad_norm": 0.7807036090808576,
+      "learning_rate": 3.9121916499843454e-07,
+      "loss": 1.5665,
+      "step": 481
+    },
+    {
+      "epoch": 0.10004151100041511,
+      "grad_norm": 0.8926143161957559,
+      "learning_rate": 3.9118291763794067e-07,
+      "loss": 1.5545,
+      "step": 482
+    },
+    {
+      "epoch": 0.10024906600249066,
+      "grad_norm": 1.0697963944564333,
+      "learning_rate": 3.9114659749501494e-07,
+      "loss": 1.6027,
+      "step": 483
+    },
+    {
+      "epoch": 0.1004566210045662,
+      "grad_norm": 1.2286496973445413,
+      "learning_rate": 3.911102045850996e-07,
+      "loss": 1.5047,
+      "step": 484
+    },
+    {
+      "epoch": 0.10066417600664176,
+      "grad_norm": 0.7703375560644409,
+      "learning_rate": 3.91073738923668e-07,
+      "loss": 1.6369,
+      "step": 485
+    },
+    {
+      "epoch": 0.10087173100871731,
+      "grad_norm": 0.9181149106643345,
+      "learning_rate": 3.910372005262244e-07,
+      "loss": 1.6047,
+      "step": 486
+    },
+    {
+      "epoch": 0.10107928601079286,
+      "grad_norm": 0.8469189591961899,
+      "learning_rate": 3.910005894083039e-07,
+      "loss": 1.533,
+      "step": 487
+    },
+    {
+      "epoch": 0.1012868410128684,
+      "grad_norm": 0.7074828317641544,
+      "learning_rate": 3.9096390558547254e-07,
+      "loss": 1.5259,
+      "step": 488
+    },
+    {
+      "epoch": 0.10149439601494396,
+      "grad_norm": 0.6772308842654708,
+      "learning_rate": 3.9092714907332743e-07,
+      "loss": 1.5527,
+      "step": 489
+    },
+    {
+      "epoch": 0.10170195101701951,
+      "grad_norm": 0.9981354518878973,
+      "learning_rate": 3.9089031988749637e-07,
+      "loss": 1.6285,
+      "step": 490
+    },
+    {
+      "epoch": 0.10190950601909506,
+      "grad_norm": 1.4375112189702568,
+      "learning_rate": 3.908534180436381e-07,
+      "loss": 1.4481,
+      "step": 491
+    },
+    {
+      "epoch": 0.10211706102117062,
+      "grad_norm": 0.8634973329581775,
+      "learning_rate": 3.9081644355744246e-07,
+      "loss": 1.5023,
+      "step": 492
+    },
+    {
+      "epoch": 0.10232461602324616,
+      "grad_norm": 9.850171705636878,
+      "learning_rate": 3.907793964446299e-07,
+      "loss": 1.6363,
+      "step": 493
+    },
+    {
+      "epoch": 0.10253217102532171,
+      "grad_norm": 0.70073257959331,
+      "learning_rate": 3.9074227672095195e-07,
+      "loss": 1.4827,
+      "step": 494
+    },
+    {
+      "epoch": 0.10273972602739725,
+      "grad_norm": 1.2798585806140574,
+      "learning_rate": 3.907050844021909e-07,
+      "loss": 1.5361,
+      "step": 495
+    },
+    {
+      "epoch": 0.10294728102947281,
+      "grad_norm": 2.034903110718286,
+      "learning_rate": 3.9066781950415985e-07,
+      "loss": 1.556,
+      "step": 496
+    },
+    {
+      "epoch": 0.10315483603154836,
+      "grad_norm": 0.9206364005779825,
+      "learning_rate": 3.906304820427029e-07,
+      "loss": 1.5563,
+      "step": 497
+    },
+    {
+      "epoch": 0.10336239103362391,
+      "grad_norm": 0.7203478132170901,
+      "learning_rate": 3.905930720336951e-07,
+      "loss": 1.5716,
+      "step": 498
+    },
+    {
+      "epoch": 0.10356994603569947,
+      "grad_norm": 0.6886531134581791,
+      "learning_rate": 3.9055558949304196e-07,
+      "loss": 1.5734,
+      "step": 499
+    },
+    {
+      "epoch": 0.10377750103777501,
+      "grad_norm": 0.7754719826551352,
+      "learning_rate": 3.905180344366802e-07,
+      "loss": 1.5378,
+      "step": 500
+    },
+    {
+      "epoch": 0.10398505603985056,
+      "grad_norm": 0.7858072375779781,
+      "learning_rate": 3.904804068805772e-07,
+      "loss": 1.5048,
+      "step": 501
+    },
+    {
+      "epoch": 0.1041926110419261,
+      "grad_norm": 0.6312276157134135,
+      "learning_rate": 3.904427068407311e-07,
+      "loss": 1.5893,
+      "step": 502
+    },
+    {
+      "epoch": 0.10440016604400167,
+      "grad_norm": 0.8300735260334831,
+      "learning_rate": 3.9040493433317115e-07,
+      "loss": 1.5449,
+      "step": 503
+    },
+    {
+      "epoch": 0.10460772104607721,
+      "grad_norm": 1.6859111070979174,
+      "learning_rate": 3.9036708937395705e-07,
+      "loss": 1.5615,
+      "step": 504
+    },
+    {
+      "epoch": 0.10481527604815276,
+      "grad_norm": 0.8529431035924254,
+      "learning_rate": 3.903291719791796e-07,
+      "loss": 1.3919,
+      "step": 505
+    },
+    {
+      "epoch": 0.1050228310502283,
+      "grad_norm": 0.721944519553621,
+      "learning_rate": 3.902911821649602e-07,
+      "loss": 1.5649,
+      "step": 506
+    },
+    {
+      "epoch": 0.10523038605230386,
+      "grad_norm": 0.9336295365454339,
+      "learning_rate": 3.9025311994745106e-07,
+      "loss": 1.5755,
+      "step": 507
+    },
+    {
+      "epoch": 0.10543794105437941,
+      "grad_norm": 0.6434889668535757,
+      "learning_rate": 3.9021498534283534e-07,
+      "loss": 1.5289,
+      "step": 508
+    },
+    {
+      "epoch": 0.10564549605645496,
+      "grad_norm": 1.2063611498097493,
+      "learning_rate": 3.901767783673267e-07,
+      "loss": 1.5352,
+      "step": 509
+    },
+    {
+      "epoch": 0.10585305105853052,
+      "grad_norm": 0.9968073691449556,
+      "learning_rate": 3.9013849903716996e-07,
+      "loss": 1.5469,
+      "step": 510
+    },
+    {
+      "epoch": 0.10606060606060606,
+      "grad_norm": 0.6601769239968291,
+      "learning_rate": 3.9010014736864026e-07,
+      "loss": 1.5644,
+      "step": 511
+    },
+    {
+      "epoch": 0.10626816106268161,
+      "grad_norm": 0.7924401093548435,
+      "learning_rate": 3.9006172337804387e-07,
+      "loss": 1.5263,
+      "step": 512
+    },
+    {
+      "epoch": 0.10647571606475716,
+      "grad_norm": 0.7636002241439529,
+      "learning_rate": 3.900232270817176e-07,
+      "loss": 1.5028,
+      "step": 513
+    },
+    {
+      "epoch": 0.10668327106683272,
+      "grad_norm": 1.1518429581330218,
+      "learning_rate": 3.89984658496029e-07,
+      "loss": 1.5726,
+      "step": 514
+    },
+    {
+      "epoch": 0.10689082606890826,
+      "grad_norm": 0.9644898224754817,
+      "learning_rate": 3.8994601763737644e-07,
+      "loss": 1.5764,
+      "step": 515
+    },
+    {
+      "epoch": 0.10709838107098381,
+      "grad_norm": 0.9217493040293064,
+      "learning_rate": 3.8990730452218897e-07,
+      "loss": 1.5671,
+      "step": 516
+    },
+    {
+      "epoch": 0.10730593607305935,
+      "grad_norm": 0.840426624234865,
+      "learning_rate": 3.898685191669264e-07,
+      "loss": 1.5513,
+      "step": 517
+    },
+    {
+      "epoch": 0.10751349107513491,
+      "grad_norm": 0.9775317833774693,
+      "learning_rate": 3.8982966158807923e-07,
+      "loss": 1.5325,
+      "step": 518
+    },
+    {
+      "epoch": 0.10772104607721046,
+      "grad_norm": 0.8040808044825589,
+      "learning_rate": 3.897907318021687e-07,
+      "loss": 1.5659,
+      "step": 519
+    },
+    {
+      "epoch": 0.10792860107928601,
+      "grad_norm": 5.82499117295419,
+      "learning_rate": 3.897517298257467e-07,
+      "loss": 1.5402,
+      "step": 520
+    },
+    {
+      "epoch": 0.10813615608136157,
+      "grad_norm": 0.8871012219416111,
+      "learning_rate": 3.897126556753958e-07,
+      "loss": 1.5841,
+      "step": 521
+    },
+    {
+      "epoch": 0.10834371108343711,
+      "grad_norm": 0.9564522521750942,
+      "learning_rate": 3.8967350936772934e-07,
+      "loss": 1.6025,
+      "step": 522
+    },
+    {
+      "epoch": 0.10855126608551266,
+      "grad_norm": 1.0770848251096499,
+      "learning_rate": 3.8963429091939124e-07,
+      "loss": 1.6,
+      "step": 523
+    },
+    {
+      "epoch": 0.1087588210875882,
+      "grad_norm": 0.7609324386765703,
+      "learning_rate": 3.8959500034705625e-07,
+      "loss": 1.4665,
+      "step": 524
+    },
+    {
+      "epoch": 0.10896637608966377,
+      "grad_norm": 1.4349158794392791,
+      "learning_rate": 3.8955563766742957e-07,
+      "loss": 1.5503,
+      "step": 525
+    },
+    {
+      "epoch": 0.10917393109173931,
+      "grad_norm": 1.402721486633433,
+      "learning_rate": 3.895162028972472e-07,
+      "loss": 1.5648,
+      "step": 526
+    },
+    {
+      "epoch": 0.10938148609381486,
+      "grad_norm": 0.7625075276425013,
+      "learning_rate": 3.894766960532757e-07,
+      "loss": 1.5724,
+      "step": 527
+    },
+    {
+      "epoch": 0.1095890410958904,
+      "grad_norm": 0.7975975533982912,
+      "learning_rate": 3.8943711715231245e-07,
+      "loss": 1.5366,
+      "step": 528
+    },
+    {
+      "epoch": 0.10979659609796596,
+      "grad_norm": 0.7923064234091599,
+      "learning_rate": 3.8939746621118527e-07,
+      "loss": 1.5028,
+      "step": 529
+    },
+    {
+      "epoch": 0.11000415110004151,
+      "grad_norm": 0.6529154776680075,
+      "learning_rate": 3.893577432467527e-07,
+      "loss": 1.5727,
+      "step": 530
+    },
+    {
+      "epoch": 0.11021170610211706,
+      "grad_norm": 0.9128103636720795,
+      "learning_rate": 3.893179482759039e-07,
+      "loss": 1.5505,
+      "step": 531
+    },
+    {
+      "epoch": 0.11041926110419262,
+      "grad_norm": 0.6191678444772217,
+      "learning_rate": 3.892780813155586e-07,
+      "loss": 1.5546,
+      "step": 532
+    },
+    {
+      "epoch": 0.11062681610626816,
+      "grad_norm": 0.9440919177992679,
+      "learning_rate": 3.8923814238266724e-07,
+      "loss": 1.5667,
+      "step": 533
+    },
+    {
+      "epoch": 0.11083437110834371,
+      "grad_norm": 0.8064140974658476,
+      "learning_rate": 3.8919813149421076e-07,
+      "loss": 1.4646,
+      "step": 534
+    },
+    {
+      "epoch": 0.11104192611041926,
+      "grad_norm": 0.7003708073176622,
+      "learning_rate": 3.8915804866720074e-07,
+      "loss": 1.5063,
+      "step": 535
+    },
+    {
+      "epoch": 0.11124948111249482,
+      "grad_norm": 0.7781231267946289,
+      "learning_rate": 3.891178939186793e-07,
+      "loss": 1.4644,
+      "step": 536
+    },
+    {
+      "epoch": 0.11145703611457036,
+      "grad_norm": 0.7510775999916478,
+      "learning_rate": 3.8907766726571915e-07,
+      "loss": 1.5103,
+      "step": 537
+    },
+    {
+      "epoch": 0.11166459111664591,
+      "grad_norm": 0.7672721518525151,
+      "learning_rate": 3.8903736872542366e-07,
+      "loss": 1.508,
+      "step": 538
+    },
+    {
+      "epoch": 0.11187214611872145,
+      "grad_norm": 0.8585082254464806,
+      "learning_rate": 3.8899699831492676e-07,
+      "loss": 1.5373,
+      "step": 539
+    },
+    {
+      "epoch": 0.11207970112079702,
+      "grad_norm": 0.7975243202660994,
+      "learning_rate": 3.889565560513927e-07,
+      "loss": 1.5246,
+      "step": 540
+    },
+    {
+      "epoch": 0.11228725612287256,
+      "grad_norm": 0.7568660447624274,
+      "learning_rate": 3.8891604195201654e-07,
+      "loss": 1.5514,
+      "step": 541
+    },
+    {
+      "epoch": 0.11249481112494811,
+      "grad_norm": 1.0936425859018952,
+      "learning_rate": 3.888754560340238e-07,
+      "loss": 1.5369,
+      "step": 542
+    },
+    {
+      "epoch": 0.11270236612702367,
+      "grad_norm": 0.8112449139981895,
+      "learning_rate": 3.888347983146706e-07,
+      "loss": 1.5613,
+      "step": 543
+    },
+    {
+      "epoch": 0.11290992112909921,
+      "grad_norm": 0.802355469239535,
+      "learning_rate": 3.887940688112434e-07,
+      "loss": 1.5359,
+      "step": 544
+    },
+    {
+      "epoch": 0.11311747613117476,
+      "grad_norm": 0.6830065056849106,
+      "learning_rate": 3.8875326754105937e-07,
+      "loss": 1.5481,
+      "step": 545
+    },
+    {
+      "epoch": 0.1133250311332503,
+      "grad_norm": 1.4454059629737248,
+      "learning_rate": 3.887123945214662e-07,
+      "loss": 1.5787,
+      "step": 546
+    },
+    {
+      "epoch": 0.11353258613532587,
+      "grad_norm": 0.7262830908336716,
+      "learning_rate": 3.886714497698419e-07,
+      "loss": 1.5275,
+      "step": 547
+    },
+    {
+      "epoch": 0.11374014113740141,
+      "grad_norm": 0.9513931205545216,
+      "learning_rate": 3.886304333035951e-07,
+      "loss": 1.5453,
+      "step": 548
+    },
+    {
+      "epoch": 0.11394769613947696,
+      "grad_norm": 0.7397803180818738,
+      "learning_rate": 3.8858934514016497e-07,
+      "loss": 1.5529,
+      "step": 549
+    },
+    {
+      "epoch": 0.1141552511415525,
+      "grad_norm": 0.7839787107858897,
+      "learning_rate": 3.885481852970211e-07,
+      "loss": 1.6007,
+      "step": 550
+    },
+    {
+      "epoch": 0.11436280614362807,
+      "grad_norm": 2.311068445407042,
+      "learning_rate": 3.8850695379166356e-07,
+      "loss": 1.5355,
+      "step": 551
+    },
+    {
+      "epoch": 0.11457036114570361,
+      "grad_norm": 0.7580724649485342,
+      "learning_rate": 3.884656506416228e-07,
+      "loss": 1.5161,
+      "step": 552
+    },
+    {
+      "epoch": 0.11477791614777916,
+      "grad_norm": 1.0810735916216159,
+      "learning_rate": 3.8842427586445994e-07,
+      "loss": 1.5971,
+      "step": 553
+    },
+    {
+      "epoch": 0.11498547114985472,
+      "grad_norm": 0.7095799681671573,
+      "learning_rate": 3.883828294777664e-07,
+      "loss": 1.5519,
+      "step": 554
+    },
+    {
+      "epoch": 0.11519302615193026,
+      "grad_norm": 0.7776217604791965,
+      "learning_rate": 3.8834131149916407e-07,
+      "loss": 1.513,
+      "step": 555
+    },
+    {
+      "epoch": 0.11540058115400581,
+      "grad_norm": 1.5652313619207674,
+      "learning_rate": 3.882997219463053e-07,
+      "loss": 1.5832,
+      "step": 556
+    },
+    {
+      "epoch": 0.11560813615608136,
+      "grad_norm": 0.6584277949747235,
+      "learning_rate": 3.8825806083687285e-07,
+      "loss": 1.5643,
+      "step": 557
+    },
+    {
+      "epoch": 0.11581569115815692,
+      "grad_norm": 0.782165356640896,
+      "learning_rate": 3.882163281885799e-07,
+      "loss": 1.5433,
+      "step": 558
+    },
+    {
+      "epoch": 0.11602324616023246,
+      "grad_norm": 0.7278153007422951,
+      "learning_rate": 3.8817452401917017e-07,
+      "loss": 1.6027,
+      "step": 559
+    },
+    {
+      "epoch": 0.11623080116230801,
+      "grad_norm": 0.6554798336177031,
+      "learning_rate": 3.881326483464175e-07,
+      "loss": 1.5715,
+      "step": 560
+    },
+    {
+      "epoch": 0.11643835616438356,
+      "grad_norm": 0.6512248296103849,
+      "learning_rate": 3.8809070118812647e-07,
+      "loss": 1.5346,
+      "step": 561
+    },
+    {
+      "epoch": 0.11664591116645912,
+      "grad_norm": 0.7426102454894176,
+      "learning_rate": 3.880486825621319e-07,
+      "loss": 1.5214,
+      "step": 562
+    },
+    {
+      "epoch": 0.11685346616853466,
+      "grad_norm": 0.7986407710365486,
+      "learning_rate": 3.880065924862989e-07,
+      "loss": 1.574,
+      "step": 563
+    },
+    {
+      "epoch": 0.11706102117061021,
+      "grad_norm": 0.8621164944700053,
+      "learning_rate": 3.8796443097852313e-07,
+      "loss": 1.531,
+      "step": 564
+    },
+    {
+      "epoch": 0.11726857617268577,
+      "grad_norm": 0.9289635560898413,
+      "learning_rate": 3.8792219805673043e-07,
+      "loss": 1.5677,
+      "step": 565
+    },
+    {
+      "epoch": 0.11747613117476131,
+      "grad_norm": 0.7164626952110217,
+      "learning_rate": 3.878798937388773e-07,
+      "loss": 1.5604,
+      "step": 566
+    },
+    {
+      "epoch": 0.11768368617683686,
+      "grad_norm": 0.690439289874298,
+      "learning_rate": 3.8783751804295024e-07,
+      "loss": 1.5507,
+      "step": 567
+    },
+    {
+      "epoch": 0.1178912411789124,
+      "grad_norm": 1.4033573937712738,
+      "learning_rate": 3.8779507098696637e-07,
+      "loss": 1.5704,
+      "step": 568
+    },
+    {
+      "epoch": 0.11809879618098797,
+      "grad_norm": 0.9927587413657801,
+      "learning_rate": 3.8775255258897304e-07,
+      "loss": 1.5556,
+      "step": 569
+    },
+    {
+      "epoch": 0.11830635118306351,
+      "grad_norm": 0.7819344606449811,
+      "learning_rate": 3.877099628670479e-07,
+      "loss": 1.5152,
+      "step": 570
+    },
+    {
+      "epoch": 0.11851390618513906,
+      "grad_norm": 1.1938556215032117,
+      "learning_rate": 3.8766730183929893e-07,
+      "loss": 1.5554,
+      "step": 571
+    },
+    {
+      "epoch": 0.1187214611872146,
+      "grad_norm": 1.0039061048079752,
+      "learning_rate": 3.8762456952386466e-07,
+      "loss": 1.5625,
+      "step": 572
+    },
+    {
+      "epoch": 0.11892901618929017,
+      "grad_norm": 0.8867948281147221,
+      "learning_rate": 3.875817659389135e-07,
+      "loss": 1.5517,
+      "step": 573
+    },
+    {
+      "epoch": 0.11913657119136571,
+      "grad_norm": 0.6159689623344549,
+      "learning_rate": 3.8753889110264455e-07,
+      "loss": 1.5001,
+      "step": 574
+    },
+    {
+      "epoch": 0.11934412619344126,
+      "grad_norm": 0.7145769013620279,
+      "learning_rate": 3.87495945033287e-07,
+      "loss": 1.4948,
+      "step": 575
+    },
+    {
+      "epoch": 0.11955168119551682,
+      "grad_norm": 0.6740816373877853,
+      "learning_rate": 3.874529277491004e-07,
+      "loss": 1.6317,
+      "step": 576
+    },
+    {
+      "epoch": 0.11975923619759236,
+      "grad_norm": 0.7061900464119396,
+      "learning_rate": 3.8740983926837455e-07,
+      "loss": 1.5431,
+      "step": 577
+    },
+    {
+      "epoch": 0.11996679119966791,
+      "grad_norm": 0.8517078816663815,
+      "learning_rate": 3.873666796094295e-07,
+      "loss": 1.5775,
+      "step": 578
+    },
+    {
+      "epoch": 0.12017434620174346,
+      "grad_norm": 0.7228339470868413,
+      "learning_rate": 3.8732344879061565e-07,
+      "loss": 1.5779,
+      "step": 579
+    },
+    {
+      "epoch": 0.12038190120381902,
+      "grad_norm": 1.302469628343104,
+      "learning_rate": 3.8728014683031353e-07,
+      "loss": 1.6022,
+      "step": 580
+    },
+    {
+      "epoch": 0.12058945620589456,
+      "grad_norm": 0.8567454068966431,
+      "learning_rate": 3.87236773746934e-07,
+      "loss": 1.598,
+      "step": 581
+    },
+    {
+      "epoch": 0.12079701120797011,
+      "grad_norm": 0.8655054910951147,
+      "learning_rate": 3.8719332955891815e-07,
+      "loss": 1.5958,
+      "step": 582
+    },
+    {
+      "epoch": 0.12100456621004566,
+      "grad_norm": 0.9165375458943107,
+      "learning_rate": 3.8714981428473736e-07,
+      "loss": 1.6328,
+      "step": 583
+    },
+    {
+      "epoch": 0.12121212121212122,
+      "grad_norm": 0.7174289586456537,
+      "learning_rate": 3.8710622794289304e-07,
+      "loss": 1.4962,
+      "step": 584
+    },
+    {
+      "epoch": 0.12141967621419676,
+      "grad_norm": 0.8361028688471015,
+      "learning_rate": 3.8706257055191706e-07,
+      "loss": 1.5587,
+      "step": 585
+    },
+    {
+      "epoch": 0.12162723121627231,
+      "grad_norm": 0.816584592324343,
+      "learning_rate": 3.870188421303713e-07,
+      "loss": 1.49,
+      "step": 586
+    },
+    {
+      "epoch": 0.12183478621834787,
+      "grad_norm": 0.9940894117635181,
+      "learning_rate": 3.86975042696848e-07,
+      "loss": 1.597,
+      "step": 587
+    },
+    {
+      "epoch": 0.12204234122042341,
+      "grad_norm": 0.7345683751755205,
+      "learning_rate": 3.869311722699695e-07,
+      "loss": 1.5783,
+      "step": 588
+    },
+    {
+      "epoch": 0.12224989622249896,
+      "grad_norm": 1.7678038059645846,
+      "learning_rate": 3.868872308683883e-07,
+      "loss": 1.6367,
+      "step": 589
+    },
+    {
+      "epoch": 0.1224574512245745,
+      "grad_norm": 0.7465387206711173,
+      "learning_rate": 3.8684321851078714e-07,
+      "loss": 1.5989,
+      "step": 590
+    },
+    {
+      "epoch": 0.12266500622665007,
+      "grad_norm": 0.6844292182568991,
+      "learning_rate": 3.8679913521587893e-07,
+      "loss": 1.5181,
+      "step": 591
+    },
+    {
+      "epoch": 0.12287256122872561,
+      "grad_norm": 0.7188466303561866,
+      "learning_rate": 3.8675498100240664e-07,
+      "loss": 1.5175,
+      "step": 592
+    },
+    {
+      "epoch": 0.12308011623080116,
+      "grad_norm": 0.8255480499699501,
+      "learning_rate": 3.8671075588914355e-07,
+      "loss": 1.5483,
+      "step": 593
+    },
+    {
+      "epoch": 0.1232876712328767,
+      "grad_norm": 0.7309100774824837,
+      "learning_rate": 3.8666645989489293e-07,
+      "loss": 1.5319,
+      "step": 594
+    },
+    {
+      "epoch": 0.12349522623495227,
+      "grad_norm": 0.8771897348607793,
+      "learning_rate": 3.866220930384884e-07,
+      "loss": 1.5643,
+      "step": 595
+    },
+    {
+      "epoch": 0.12370278123702781,
+      "grad_norm": 0.7195101915185197,
+      "learning_rate": 3.865776553387934e-07,
+      "loss": 1.555,
+      "step": 596
+    },
+    {
+      "epoch": 0.12391033623910336,
+      "grad_norm": 0.896838102764189,
+      "learning_rate": 3.865331468147018e-07,
+      "loss": 1.5872,
+      "step": 597
+    },
+    {
+      "epoch": 0.12411789124117892,
+      "grad_norm": 0.8948986325445162,
+      "learning_rate": 3.864885674851372e-07,
+      "loss": 1.5039,
+      "step": 598
+    },
+    {
+      "epoch": 0.12432544624325446,
+      "grad_norm": 0.989534865109656,
+      "learning_rate": 3.8644391736905393e-07,
+      "loss": 1.5894,
+      "step": 599
+    },
+    {
+      "epoch": 0.12453300124533001,
+      "grad_norm": 1.1838407541445375,
+      "learning_rate": 3.8639919648543576e-07,
+      "loss": 1.565,
+      "step": 600
+    },
+    {
+      "epoch": 0.12474055624740556,
+      "grad_norm": 0.8347719714864444,
+      "learning_rate": 3.8635440485329686e-07,
+      "loss": 1.5311,
+      "step": 601
+    },
+    {
+      "epoch": 0.12494811124948112,
+      "grad_norm": 1.2274611309321541,
+      "learning_rate": 3.8630954249168156e-07,
+      "loss": 1.5233,
+      "step": 602
+    },
+    {
+      "epoch": 0.12515566625155666,
+      "grad_norm": 1.1781426891102558,
+      "learning_rate": 3.8626460941966397e-07,
+      "loss": 1.5471,
+      "step": 603
+    },
+    {
+      "epoch": 0.12536322125363222,
+      "grad_norm": 0.8713674847966427,
+      "learning_rate": 3.8621960565634854e-07,
+      "loss": 1.4641,
+      "step": 604
+    },
+    {
+      "epoch": 0.12557077625570776,
+      "grad_norm": 0.9842933850239068,
+      "learning_rate": 3.861745312208697e-07,
+      "loss": 1.5688,
+      "step": 605
+    },
+    {
+      "epoch": 0.12577833125778332,
+      "grad_norm": 0.6892514619832997,
+      "learning_rate": 3.861293861323919e-07,
+      "loss": 1.4371,
+      "step": 606
+    },
+    {
+      "epoch": 0.12598588625985888,
+      "grad_norm": 0.7223007839764674,
+      "learning_rate": 3.8608417041010954e-07,
+      "loss": 1.5333,
+      "step": 607
+    },
+    {
+      "epoch": 0.1261934412619344,
+      "grad_norm": 1.0990379125141843,
+      "learning_rate": 3.8603888407324724e-07,
+      "loss": 1.48,
+      "step": 608
+    },
+    {
+      "epoch": 0.12640099626400997,
+      "grad_norm": 1.4726701349794962,
+      "learning_rate": 3.859935271410595e-07,
+      "loss": 1.5996,
+      "step": 609
+    },
+    {
+      "epoch": 0.1266085512660855,
+      "grad_norm": 0.6901756012999231,
+      "learning_rate": 3.8594809963283083e-07,
+      "loss": 1.5191,
+      "step": 610
+    },
+    {
+      "epoch": 0.12681610626816106,
+      "grad_norm": 0.7790444189146841,
+      "learning_rate": 3.85902601567876e-07,
+      "loss": 1.5627,
+      "step": 611
+    },
+    {
+      "epoch": 0.12702366127023662,
+      "grad_norm": 1.300846328301186,
+      "learning_rate": 3.8585703296553934e-07,
+      "loss": 1.5636,
+      "step": 612
+    },
+    {
+      "epoch": 0.12723121627231215,
+      "grad_norm": 0.6181564106990775,
+      "learning_rate": 3.858113938451956e-07,
+      "loss": 1.5084,
+      "step": 613
+    },
+    {
+      "epoch": 0.1274387712743877,
+      "grad_norm": 0.7520014176919081,
+      "learning_rate": 3.857656842262492e-07,
+      "loss": 1.4436,
+      "step": 614
+    },
+    {
+      "epoch": 0.12764632627646327,
+      "grad_norm": 0.7423989341624727,
+      "learning_rate": 3.857199041281346e-07,
+      "loss": 1.4885,
+      "step": 615
+    },
+    {
+      "epoch": 0.1278538812785388,
+      "grad_norm": 0.8581992205715749,
+      "learning_rate": 3.8567405357031647e-07,
+      "loss": 1.4732,
+      "step": 616
+    },
+    {
+      "epoch": 0.12806143628061437,
+      "grad_norm": 0.79655873764379,
+      "learning_rate": 3.856281325722892e-07,
+      "loss": 1.5012,
+      "step": 617
+    },
+    {
+      "epoch": 0.12826899128268993,
+      "grad_norm": 0.8829311096154718,
+      "learning_rate": 3.8558214115357705e-07,
+      "loss": 1.4866,
+      "step": 618
+    },
+    {
+      "epoch": 0.12847654628476546,
+      "grad_norm": 0.9121143481722199,
+      "learning_rate": 3.855360793337345e-07,
+      "loss": 1.654,
+      "step": 619
+    },
+    {
+      "epoch": 0.12868410128684102,
+      "grad_norm": 0.644108671989746,
+      "learning_rate": 3.854899471323457e-07,
+      "loss": 1.5721,
+      "step": 620
+    },
+    {
+      "epoch": 0.12889165628891655,
+      "grad_norm": 0.8867091132403353,
+      "learning_rate": 3.85443744569025e-07,
+      "loss": 1.5161,
+      "step": 621
+    },
+    {
+      "epoch": 0.1290992112909921,
+      "grad_norm": 0.7994182237521943,
+      "learning_rate": 3.8539747166341625e-07,
+      "loss": 1.5188,
+      "step": 622
+    },
+    {
+      "epoch": 0.12930676629306767,
+      "grad_norm": 1.0066442837051297,
+      "learning_rate": 3.8535112843519373e-07,
+      "loss": 1.5315,
+      "step": 623
+    },
+    {
+      "epoch": 0.1295143212951432,
+      "grad_norm": 1.0004063717443135,
+      "learning_rate": 3.8530471490406107e-07,
+      "loss": 1.5701,
+      "step": 624
+    },
+    {
+      "epoch": 0.12972187629721876,
+      "grad_norm": 0.973800307539712,
+      "learning_rate": 3.8525823108975234e-07,
+      "loss": 1.6129,
+      "step": 625
+    },
+    {
+      "epoch": 0.12992943129929432,
+      "grad_norm": 0.8411830780758435,
+      "learning_rate": 3.8521167701203103e-07,
+      "loss": 1.5453,
+      "step": 626
+    },
+    {
+      "epoch": 0.13013698630136986,
+      "grad_norm": 0.9558907308226784,
+      "learning_rate": 3.8516505269069083e-07,
+      "loss": 1.5749,
+      "step": 627
+    },
+    {
+      "epoch": 0.13034454130344542,
+      "grad_norm": 1.0057699880050142,
+      "learning_rate": 3.851183581455551e-07,
+      "loss": 1.5444,
+      "step": 628
+    },
+    {
+      "epoch": 0.13055209630552098,
+      "grad_norm": 0.7043392810991221,
+      "learning_rate": 3.850715933964771e-07,
+      "loss": 1.4307,
+      "step": 629
+    },
+    {
+      "epoch": 0.1307596513075965,
+      "grad_norm": 1.0214604581379991,
+      "learning_rate": 3.850247584633401e-07,
+      "loss": 1.5608,
+      "step": 630
+    },
+    {
+      "epoch": 0.13096720630967207,
+      "grad_norm": 0.9167031619971703,
+      "learning_rate": 3.849778533660568e-07,
+      "loss": 1.4976,
+      "step": 631
+    },
+    {
+      "epoch": 0.1311747613117476,
+      "grad_norm": 1.1057247208033947,
+      "learning_rate": 3.849308781245703e-07,
+      "loss": 1.5619,
+      "step": 632
+    },
+    {
+      "epoch": 0.13138231631382316,
+      "grad_norm": 0.6765005367552381,
+      "learning_rate": 3.8488383275885297e-07,
+      "loss": 1.6289,
+      "step": 633
+    },
+    {
+      "epoch": 0.13158987131589872,
+      "grad_norm": 0.6786993235088649,
+      "learning_rate": 3.848367172889075e-07,
+      "loss": 1.6148,
+      "step": 634
+    },
+    {
+      "epoch": 0.13179742631797425,
+      "grad_norm": 0.7672669496048602,
+      "learning_rate": 3.847895317347659e-07,
+      "loss": 1.5458,
+      "step": 635
+    },
+    {
+      "epoch": 0.1320049813200498,
+      "grad_norm": 0.9748677926790578,
+      "learning_rate": 3.847422761164903e-07,
+      "loss": 1.4943,
+      "step": 636
+    },
+    {
+      "epoch": 0.13221253632212537,
+      "grad_norm": 0.7427900249427577,
+      "learning_rate": 3.8469495045417266e-07,
+      "loss": 1.5393,
+      "step": 637
+    },
+    {
+      "epoch": 0.1324200913242009,
+      "grad_norm": 1.1474297045666646,
+      "learning_rate": 3.8464755476793443e-07,
+      "loss": 1.538,
+      "step": 638
+    },
+    {
+      "epoch": 0.13262764632627647,
+      "grad_norm": 0.7909303657018172,
+      "learning_rate": 3.84600089077927e-07,
+      "loss": 1.5623,
+      "step": 639
+    },
+    {
+      "epoch": 0.13283520132835203,
+      "grad_norm": 0.9958495430750541,
+      "learning_rate": 3.8455255340433164e-07,
+      "loss": 1.5238,
+      "step": 640
+    },
+    {
+      "epoch": 0.13304275633042756,
+      "grad_norm": 0.9154236743521144,
+      "learning_rate": 3.845049477673592e-07,
+      "loss": 1.519,
+      "step": 641
+    },
+    {
+      "epoch": 0.13325031133250312,
+      "grad_norm": 0.7310468539749906,
+      "learning_rate": 3.8445727218725034e-07,
+      "loss": 1.582,
+      "step": 642
+    },
+    {
+      "epoch": 0.13345786633457865,
+      "grad_norm": 0.943954823112964,
+      "learning_rate": 3.8440952668427537e-07,
+      "loss": 1.5671,
+      "step": 643
+    },
+    {
+      "epoch": 0.1336654213366542,
+      "grad_norm": 0.6532790701942882,
+      "learning_rate": 3.8436171127873456e-07,
+      "loss": 1.5546,
+      "step": 644
+    },
+    {
+      "epoch": 0.13387297633872977,
+      "grad_norm": 1.5859068736896227,
+      "learning_rate": 3.8431382599095765e-07,
+      "loss": 1.5196,
+      "step": 645
+    },
+    {
+      "epoch": 0.1340805313408053,
+      "grad_norm": 0.9087888700789262,
+      "learning_rate": 3.842658708413042e-07,
+      "loss": 1.5413,
+      "step": 646
+    },
+    {
+      "epoch": 0.13428808634288086,
+      "grad_norm": 0.7069536964471304,
+      "learning_rate": 3.842178458501634e-07,
+      "loss": 1.5058,
+      "step": 647
+    },
+    {
+      "epoch": 0.13449564134495642,
+      "grad_norm": 0.6928209414743897,
+      "learning_rate": 3.841697510379544e-07,
+      "loss": 1.5758,
+      "step": 648
+    },
+    {
+      "epoch": 0.13470319634703196,
+      "grad_norm": 0.8900614144594915,
+      "learning_rate": 3.841215864251257e-07,
+      "loss": 1.5876,
+      "step": 649
+    },
+    {
+      "epoch": 0.13491075134910752,
+      "grad_norm": 0.8163751453549446,
+      "learning_rate": 3.8407335203215555e-07,
+      "loss": 1.5545,
+      "step": 650
+    },
+    {
+      "epoch": 0.13511830635118308,
+      "grad_norm": 1.571663378176316,
+      "learning_rate": 3.8402504787955214e-07,
+      "loss": 1.5665,
+      "step": 651
+    },
+    {
+      "epoch": 0.1353258613532586,
+      "grad_norm": 0.8743853191616004,
+      "learning_rate": 3.839766739878529e-07,
+      "loss": 1.5691,
+      "step": 652
+    },
+    {
+      "epoch": 0.13553341635533417,
+      "grad_norm": 0.6463130888916893,
+      "learning_rate": 3.8392823037762524e-07,
+      "loss": 1.4404,
+      "step": 653
+    },
+    {
+      "epoch": 0.1357409713574097,
+      "grad_norm": 0.7358669078667202,
+      "learning_rate": 3.8387971706946607e-07,
+      "loss": 1.4908,
+      "step": 654
+    },
+    {
+      "epoch": 0.13594852635948526,
+      "grad_norm": 4.926003852656765,
+      "learning_rate": 3.8383113408400195e-07,
+      "loss": 1.5607,
+      "step": 655
+    },
+    {
+      "epoch": 0.13615608136156082,
+      "grad_norm": 0.7534878643020217,
+      "learning_rate": 3.8378248144188905e-07,
+      "loss": 1.5844,
+      "step": 656
+    },
+    {
+      "epoch": 0.13636363636363635,
+      "grad_norm": 1.452755487773869,
+      "learning_rate": 3.837337591638133e-07,
+      "loss": 1.5123,
+      "step": 657
+    },
+    {
+      "epoch": 0.1365711913657119,
+      "grad_norm": 1.206691011529855,
+      "learning_rate": 3.8368496727049e-07,
+      "loss": 1.5295,
+      "step": 658
+    },
+    {
+      "epoch": 0.13677874636778747,
+      "grad_norm": 0.8921177737430342,
+      "learning_rate": 3.8363610578266423e-07,
+      "loss": 1.602,
+      "step": 659
+    },
+    {
+      "epoch": 0.136986301369863,
+      "grad_norm": 0.8811356956302794,
+      "learning_rate": 3.835871747211105e-07,
+      "loss": 1.5455,
+      "step": 660
+    },
+    {
+      "epoch": 0.13719385637193857,
+      "grad_norm": 0.7778385237965392,
+      "learning_rate": 3.8353817410663314e-07,
+      "loss": 1.467,
+      "step": 661
+    },
+    {
+      "epoch": 0.13740141137401413,
+      "grad_norm": 1.138715814867982,
+      "learning_rate": 3.834891039600658e-07,
+      "loss": 1.4708,
+      "step": 662
+    },
+    {
+      "epoch": 0.13760896637608966,
+      "grad_norm": 0.6583494532720251,
+      "learning_rate": 3.834399643022719e-07,
+      "loss": 1.514,
+      "step": 663
+    },
+    {
+      "epoch": 0.13781652137816522,
+      "grad_norm": 0.7021534458239477,
+      "learning_rate": 3.833907551541442e-07,
+      "loss": 1.4935,
+      "step": 664
+    },
+    {
+      "epoch": 0.13802407638024075,
+      "grad_norm": 15.807717176804518,
+      "learning_rate": 3.833414765366052e-07,
+      "loss": 1.6038,
+      "step": 665
+    },
+    {
+      "epoch": 0.1382316313823163,
+      "grad_norm": 0.7486182420724226,
+      "learning_rate": 3.832921284706069e-07,
+      "loss": 1.4913,
+      "step": 666
+    },
+    {
+      "epoch": 0.13843918638439187,
+      "grad_norm": 0.6721947178541794,
+      "learning_rate": 3.8324271097713066e-07,
+      "loss": 1.5022,
+      "step": 667
+    },
+    {
+      "epoch": 0.1386467413864674,
+      "grad_norm": 0.8283243237841391,
+      "learning_rate": 3.831932240771876e-07,
+      "loss": 1.5541,
+      "step": 668
+    },
+    {
+      "epoch": 0.13885429638854296,
+      "grad_norm": 0.708608180990852,
+      "learning_rate": 3.831436677918182e-07,
+      "loss": 1.6247,
+      "step": 669
+    },
+    {
+      "epoch": 0.13906185139061852,
+      "grad_norm": 0.7473719286750843,
+      "learning_rate": 3.8309404214209245e-07,
+      "loss": 1.4948,
+      "step": 670
+    },
+    {
+      "epoch": 0.13926940639269406,
+      "grad_norm": 0.7672536569018873,
+      "learning_rate": 3.830443471491099e-07,
+      "loss": 1.5006,
+      "step": 671
+    },
+    {
+      "epoch": 0.13947696139476962,
+      "grad_norm": 0.7381461212903716,
+      "learning_rate": 3.8299458283399956e-07,
+      "loss": 1.5185,
+      "step": 672
+    },
+    {
+      "epoch": 0.13968451639684518,
+      "grad_norm": 1.164312657342,
+      "learning_rate": 3.8294474921791975e-07,
+      "loss": 1.5114,
+      "step": 673
+    },
+    {
+      "epoch": 0.1398920713989207,
+      "grad_norm": 0.8278958094453844,
+      "learning_rate": 3.8289484632205856e-07,
+      "loss": 1.6098,
+      "step": 674
+    },
+    {
+      "epoch": 0.14009962640099627,
+      "grad_norm": 0.6128504397269805,
+      "learning_rate": 3.828448741676334e-07,
+      "loss": 1.4965,
+      "step": 675
+    },
+    {
+      "epoch": 0.1403071814030718,
+      "grad_norm": 0.9878558357310917,
+      "learning_rate": 3.827948327758909e-07,
+      "loss": 1.5582,
+      "step": 676
+    },
+    {
+      "epoch": 0.14051473640514736,
+      "grad_norm": 0.81411650946353,
+      "learning_rate": 3.827447221681076e-07,
+      "loss": 1.5509,
+      "step": 677
+    },
+    {
+      "epoch": 0.14072229140722292,
+      "grad_norm": 1.0184776693565,
+      "learning_rate": 3.82694542365589e-07,
+      "loss": 1.531,
+      "step": 678
+    },
+    {
+      "epoch": 0.14092984640929845,
+      "grad_norm": 1.0808084582591457,
+      "learning_rate": 3.8264429338967025e-07,
+      "loss": 1.5492,
+      "step": 679
+    },
+    {
+      "epoch": 0.141137401411374,
+      "grad_norm": 0.9865246289865527,
+      "learning_rate": 3.8259397526171593e-07,
+      "loss": 1.5187,
+      "step": 680
+    },
+    {
+      "epoch": 0.14134495641344957,
+      "grad_norm": 0.7618113957239873,
+      "learning_rate": 3.8254358800311997e-07,
+      "loss": 1.4566,
+      "step": 681
+    },
+    {
+      "epoch": 0.1415525114155251,
+      "grad_norm": 0.8558260283975787,
+      "learning_rate": 3.824931316353057e-07,
+      "loss": 1.5643,
+      "step": 682
+    },
+    {
+      "epoch": 0.14176006641760067,
+      "grad_norm": 0.8080672825169757,
+      "learning_rate": 3.824426061797258e-07,
+      "loss": 1.6707,
+      "step": 683
+    },
+    {
+      "epoch": 0.14196762141967623,
+      "grad_norm": 0.8515026243443249,
+      "learning_rate": 3.823920116578623e-07,
+      "loss": 1.5116,
+      "step": 684
+    },
+    {
+      "epoch": 0.14217517642175176,
+      "grad_norm": 0.835529042057423,
+      "learning_rate": 3.823413480912267e-07,
+      "loss": 1.6197,
+      "step": 685
+    },
+    {
+      "epoch": 0.14238273142382732,
+      "grad_norm": 0.7493096217585754,
+      "learning_rate": 3.8229061550135994e-07,
+      "loss": 1.5813,
+      "step": 686
+    },
+    {
+      "epoch": 0.14259028642590285,
+      "grad_norm": 2.785738130234306,
+      "learning_rate": 3.822398139098319e-07,
+      "loss": 1.4898,
+      "step": 687
+    },
+    {
+      "epoch": 0.1427978414279784,
+      "grad_norm": 1.0164311430707942,
+      "learning_rate": 3.821889433382422e-07,
+      "loss": 1.5319,
+      "step": 688
+    },
+    {
+      "epoch": 0.14300539643005397,
+      "grad_norm": 0.6812387551679115,
+      "learning_rate": 3.8213800380821974e-07,
+      "loss": 1.5434,
+      "step": 689
+    },
+    {
+      "epoch": 0.1432129514321295,
+      "grad_norm": 0.6798379050196808,
+      "learning_rate": 3.820869953414226e-07,
+      "loss": 1.5579,
+      "step": 690
+    },
+    {
+      "epoch": 0.14342050643420506,
+      "grad_norm": 0.8432181785766621,
+      "learning_rate": 3.8203591795953815e-07,
+      "loss": 1.4414,
+      "step": 691
+    },
+    {
+      "epoch": 0.14362806143628062,
+      "grad_norm": 0.9751639850631328,
+      "learning_rate": 3.819847716842832e-07,
+      "loss": 1.5897,
+      "step": 692
+    },
+    {
+      "epoch": 0.14383561643835616,
+      "grad_norm": 0.7946200883797693,
+      "learning_rate": 3.819335565374038e-07,
+      "loss": 1.5444,
+      "step": 693
+    },
+    {
+      "epoch": 0.14404317144043172,
+      "grad_norm": 0.9780082218452316,
+      "learning_rate": 3.818822725406752e-07,
+      "loss": 1.6139,
+      "step": 694
+    },
+    {
+      "epoch": 0.14425072644250728,
+      "grad_norm": 0.6312627232542098,
+      "learning_rate": 3.818309197159021e-07,
+      "loss": 1.5396,
+      "step": 695
+    },
+    {
+      "epoch": 0.1444582814445828,
+      "grad_norm": 0.8541201372466952,
+      "learning_rate": 3.8177949808491834e-07,
+      "loss": 1.5175,
+      "step": 696
+    },
+    {
+      "epoch": 0.14466583644665837,
+      "grad_norm": 1.9111219014234557,
+      "learning_rate": 3.8172800766958694e-07,
+      "loss": 1.5128,
+      "step": 697
+    },
+    {
+      "epoch": 0.1448733914487339,
+      "grad_norm": 0.8248647559250584,
+      "learning_rate": 3.816764484918003e-07,
+      "loss": 1.591,
+      "step": 698
+    },
+    {
+      "epoch": 0.14508094645080946,
+      "grad_norm": 2.759536118128399,
+      "learning_rate": 3.8162482057348007e-07,
+      "loss": 1.5544,
+      "step": 699
+    },
+    {
+      "epoch": 0.14528850145288502,
+      "grad_norm": 1.552080997371948,
+      "learning_rate": 3.81573123936577e-07,
+      "loss": 1.5411,
+      "step": 700
+    },
+    {
+      "epoch": 0.14549605645496055,
+      "grad_norm": 0.7236085099946248,
+      "learning_rate": 3.815213586030711e-07,
+      "loss": 1.5529,
+      "step": 701
+    },
+    {
+      "epoch": 0.14570361145703611,
+      "grad_norm": 1.0174563013448228,
+      "learning_rate": 3.814695245949718e-07,
+      "loss": 1.5399,
+      "step": 702
+    },
+    {
+      "epoch": 0.14591116645911167,
+      "grad_norm": 0.654142014727374,
+      "learning_rate": 3.814176219343173e-07,
+      "loss": 1.5909,
+      "step": 703
+    },
+    {
+      "epoch": 0.1461187214611872,
+      "grad_norm": 1.1732915640822246,
+      "learning_rate": 3.813656506431754e-07,
+      "loss": 1.5265,
+      "step": 704
+    },
+    {
+      "epoch": 0.14632627646326277,
+      "grad_norm": 0.9579251669320937,
+      "learning_rate": 3.8131361074364287e-07,
+      "loss": 1.5227,
+      "step": 705
+    },
+    {
+      "epoch": 0.14653383146533833,
+      "grad_norm": 1.3655500747977294,
+      "learning_rate": 3.8126150225784563e-07,
+      "loss": 1.5429,
+      "step": 706
+    },
+    {
+      "epoch": 0.14674138646741386,
+      "grad_norm": 0.7642083193775192,
+      "learning_rate": 3.812093252079389e-07,
+      "loss": 1.5179,
+      "step": 707
+    },
+    {
+      "epoch": 0.14694894146948942,
+      "grad_norm": 0.6267765803780283,
+      "learning_rate": 3.81157079616107e-07,
+      "loss": 1.5097,
+      "step": 708
+    },
+    {
+      "epoch": 0.14715649647156495,
+      "grad_norm": 1.1608112381567908,
+      "learning_rate": 3.8110476550456325e-07,
+      "loss": 1.4683,
+      "step": 709
+    },
+    {
+      "epoch": 0.1473640514736405,
+      "grad_norm": 0.6530399334044678,
+      "learning_rate": 3.810523828955504e-07,
+      "loss": 1.5067,
+      "step": 710
+    },
+    {
+      "epoch": 0.14757160647571607,
+      "grad_norm": 2.295421073228988,
+      "learning_rate": 3.8099993181134e-07,
+      "loss": 1.5557,
+      "step": 711
+    },
+    {
+      "epoch": 0.1477791614777916,
+      "grad_norm": 1.3289193607487875,
+      "learning_rate": 3.8094741227423286e-07,
+      "loss": 1.5812,
+      "step": 712
+    },
+    {
+      "epoch": 0.14798671647986716,
+      "grad_norm": 0.721445697819485,
+      "learning_rate": 3.8089482430655895e-07,
+      "loss": 1.4894,
+      "step": 713
+    },
+    {
+      "epoch": 0.14819427148194272,
+      "grad_norm": 1.0615642218969679,
+      "learning_rate": 3.808421679306772e-07,
+      "loss": 1.5256,
+      "step": 714
+    },
+    {
+      "epoch": 0.14840182648401826,
+      "grad_norm": 0.701827676972807,
+      "learning_rate": 3.807894431689759e-07,
+      "loss": 1.5193,
+      "step": 715
+    },
+    {
+      "epoch": 0.14860938148609382,
+      "grad_norm": 1.0217882565219119,
+      "learning_rate": 3.8073665004387194e-07,
+      "loss": 1.535,
+      "step": 716
+    },
+    {
+      "epoch": 0.14881693648816938,
+      "grad_norm": 0.8035415744500214,
+      "learning_rate": 3.806837885778118e-07,
+      "loss": 1.5679,
+      "step": 717
+    },
+    {
+      "epoch": 0.1490244914902449,
+      "grad_norm": 1.652905192817568,
+      "learning_rate": 3.806308587932706e-07,
+      "loss": 1.5671,
+      "step": 718
+    },
+    {
+      "epoch": 0.14923204649232047,
+      "grad_norm": 0.7475609740898325,
+      "learning_rate": 3.805778607127528e-07,
+      "loss": 1.5687,
+      "step": 719
+    },
+    {
+      "epoch": 0.149439601494396,
+      "grad_norm": 0.6266407338147819,
+      "learning_rate": 3.805247943587917e-07,
+      "loss": 1.6176,
+      "step": 720
+    },
+    {
+      "epoch": 0.14964715649647156,
+      "grad_norm": 11.176807186817417,
+      "learning_rate": 3.8047165975394974e-07,
+      "loss": 1.6668,
+      "step": 721
+    },
+    {
+      "epoch": 0.14985471149854712,
+      "grad_norm": 0.8493468480389307,
+      "learning_rate": 3.8041845692081833e-07,
+      "loss": 1.5556,
+      "step": 722
+    },
+    {
+      "epoch": 0.15006226650062265,
+      "grad_norm": 0.6460753231064308,
+      "learning_rate": 3.803651858820179e-07,
+      "loss": 1.5545,
+      "step": 723
+    },
+    {
+      "epoch": 0.15026982150269821,
+      "grad_norm": 0.7042804767654001,
+      "learning_rate": 3.803118466601979e-07,
+      "loss": 1.5868,
+      "step": 724
+    },
+    {
+      "epoch": 0.15047737650477377,
+      "grad_norm": 0.7717898594920679,
+      "learning_rate": 3.802584392780367e-07,
+      "loss": 1.5569,
+      "step": 725
+    },
+    {
+      "epoch": 0.1506849315068493,
+      "grad_norm": 0.8045647664531013,
+      "learning_rate": 3.802049637582418e-07,
+      "loss": 1.6684,
+      "step": 726
+    },
+    {
+      "epoch": 0.15089248650892487,
+      "grad_norm": 0.7333857345999618,
+      "learning_rate": 3.801514201235495e-07,
+      "loss": 1.5355,
+      "step": 727
+    },
+    {
+      "epoch": 0.15110004151100043,
+      "grad_norm": 0.7489112675243124,
+      "learning_rate": 3.8009780839672504e-07,
+      "loss": 1.5582,
+      "step": 728
+    },
+    {
+      "epoch": 0.15130759651307596,
+      "grad_norm": 0.7184462961272369,
+      "learning_rate": 3.8004412860056293e-07,
+      "loss": 1.5919,
+      "step": 729
+    },
+    {
+      "epoch": 0.15151515151515152,
+      "grad_norm": 0.8550277459910752,
+      "learning_rate": 3.799903807578862e-07,
+      "loss": 1.5117,
+      "step": 730
+    },
+    {
+      "epoch": 0.15172270651722705,
+      "grad_norm": 1.672129718069143,
+      "learning_rate": 3.799365648915471e-07,
+      "loss": 1.5472,
+      "step": 731
+    },
+    {
+      "epoch": 0.1519302615193026,
+      "grad_norm": 1.095696233055675,
+      "learning_rate": 3.798826810244267e-07,
+      "loss": 1.5499,
+      "step": 732
+    },
+    {
+      "epoch": 0.15213781652137817,
+      "grad_norm": 0.878717413960594,
+      "learning_rate": 3.798287291794349e-07,
+      "loss": 1.5424,
+      "step": 733
+    },
+    {
+      "epoch": 0.1523453715234537,
+      "grad_norm": 0.7309357288980483,
+      "learning_rate": 3.7977470937951073e-07,
+      "loss": 1.5546,
+      "step": 734
+    },
+    {
+      "epoch": 0.15255292652552926,
+      "grad_norm": 0.7631169956712764,
+      "learning_rate": 3.797206216476219e-07,
+      "loss": 1.5668,
+      "step": 735
+    },
+    {
+      "epoch": 0.15276048152760482,
+      "grad_norm": 0.7792678761367231,
+      "learning_rate": 3.7966646600676515e-07,
+      "loss": 1.5557,
+      "step": 736
+    },
+    {
+      "epoch": 0.15296803652968036,
+      "grad_norm": 1.3692447146852718,
+      "learning_rate": 3.7961224247996585e-07,
+      "loss": 1.5175,
+      "step": 737
+    },
+    {
+      "epoch": 0.15317559153175592,
+      "grad_norm": 1.3137847200593238,
+      "learning_rate": 3.7955795109027854e-07,
+      "loss": 1.4981,
+      "step": 738
+    },
+    {
+      "epoch": 0.15338314653383148,
+      "grad_norm": 1.7798928876772429,
+      "learning_rate": 3.7950359186078647e-07,
+      "loss": 1.5234,
+      "step": 739
+    },
+    {
+      "epoch": 0.153590701535907,
+      "grad_norm": 0.9522088672416029,
+      "learning_rate": 3.794491648146017e-07,
+      "loss": 1.5268,
+      "step": 740
+    },
+    {
+      "epoch": 0.15379825653798257,
+      "grad_norm": 0.9478963550459683,
+      "learning_rate": 3.7939466997486516e-07,
+      "loss": 1.4965,
+      "step": 741
+    },
+    {
+      "epoch": 0.1540058115400581,
+      "grad_norm": 1.221112506801334,
+      "learning_rate": 3.793401073647467e-07,
+      "loss": 1.5211,
+      "step": 742
+    },
+    {
+      "epoch": 0.15421336654213366,
+      "grad_norm": 0.8030798001089229,
+      "learning_rate": 3.792854770074448e-07,
+      "loss": 1.5356,
+      "step": 743
+    },
+    {
+      "epoch": 0.15442092154420922,
+      "grad_norm": 0.7140004349172026,
+      "learning_rate": 3.7923077892618686e-07,
+      "loss": 1.6048,
+      "step": 744
+    },
+    {
+      "epoch": 0.15462847654628475,
+      "grad_norm": 1.0222560299712333,
+      "learning_rate": 3.791760131442291e-07,
+      "loss": 1.6742,
+      "step": 745
+    },
+    {
+      "epoch": 0.15483603154836031,
+      "grad_norm": 0.8331242423356838,
+      "learning_rate": 3.791211796848563e-07,
+      "loss": 1.5261,
+      "step": 746
+    },
+    {
+      "epoch": 0.15504358655043587,
+      "grad_norm": 0.7886865686912353,
+      "learning_rate": 3.7906627857138245e-07,
+      "loss": 1.4505,
+      "step": 747
+    },
+    {
+      "epoch": 0.1552511415525114,
+      "grad_norm": 10.464886450965508,
+      "learning_rate": 3.790113098271499e-07,
+      "loss": 1.5242,
+      "step": 748
+    },
+    {
+      "epoch": 0.15545869655458697,
+      "grad_norm": 0.6984608327745523,
+      "learning_rate": 3.7895627347552994e-07,
+      "loss": 1.4947,
+      "step": 749
+    },
+    {
+      "epoch": 0.15566625155666253,
+      "grad_norm": 0.6997890084690677,
+      "learning_rate": 3.7890116953992245e-07,
+      "loss": 1.54,
+      "step": 750
+    },
+    {
+      "epoch": 0.15587380655873806,
+      "grad_norm": 0.7052120132426379,
+      "learning_rate": 3.7884599804375637e-07,
+      "loss": 1.5285,
+      "step": 751
+    },
+    {
+      "epoch": 0.15608136156081362,
+      "grad_norm": 0.761894932884719,
+      "learning_rate": 3.78790759010489e-07,
+      "loss": 1.4949,
+      "step": 752
+    },
+    {
+      "epoch": 0.15628891656288915,
+      "grad_norm": 0.8524777648347153,
+      "learning_rate": 3.787354524636066e-07,
+      "loss": 1.5505,
+      "step": 753
+    },
+    {
+      "epoch": 0.1564964715649647,
+      "grad_norm": 0.7626434368528677,
+      "learning_rate": 3.7868007842662394e-07,
+      "loss": 1.5718,
+      "step": 754
+    },
+    {
+      "epoch": 0.15670402656704027,
+      "grad_norm": 0.9902730510266881,
+      "learning_rate": 3.786246369230846e-07,
+      "loss": 1.5988,
+      "step": 755
+    },
+    {
+      "epoch": 0.1569115815691158,
+      "grad_norm": 0.7973870159346816,
+      "learning_rate": 3.78569127976561e-07,
+      "loss": 1.6035,
+      "step": 756
+    },
+    {
+      "epoch": 0.15711913657119136,
+      "grad_norm": 0.890835409053048,
+      "learning_rate": 3.785135516106539e-07,
+      "loss": 1.4627,
+      "step": 757
+    },
+    {
+      "epoch": 0.15732669157326692,
+      "grad_norm": 0.8050582158496478,
+      "learning_rate": 3.784579078489929e-07,
+      "loss": 1.567,
+      "step": 758
+    },
+    {
+      "epoch": 0.15753424657534246,
+      "grad_norm": 0.7023748694599736,
+      "learning_rate": 3.784021967152364e-07,
+      "loss": 1.5677,
+      "step": 759
+    },
+    {
+      "epoch": 0.15774180157741802,
+      "grad_norm": 0.961776143990859,
+      "learning_rate": 3.7834641823307115e-07,
+      "loss": 1.5946,
+      "step": 760
+    },
+    {
+      "epoch": 0.15794935657949358,
+      "grad_norm": 0.7591178517703842,
+      "learning_rate": 3.782905724262127e-07,
+      "loss": 1.5049,
+      "step": 761
+    },
+    {
+      "epoch": 0.1581569115815691,
+      "grad_norm": 0.7277432229358922,
+      "learning_rate": 3.782346593184053e-07,
+      "loss": 1.5254,
+      "step": 762
+    },
+    {
+      "epoch": 0.15836446658364467,
+      "grad_norm": 0.6975379805271739,
+      "learning_rate": 3.781786789334216e-07,
+      "loss": 1.6107,
+      "step": 763
+    },
+    {
+      "epoch": 0.1585720215857202,
+      "grad_norm": 0.7293344042173459,
+      "learning_rate": 3.78122631295063e-07,
+      "loss": 1.5843,
+      "step": 764
+    },
+    {
+      "epoch": 0.15877957658779576,
+      "grad_norm": 1.0190360085064099,
+      "learning_rate": 3.780665164271595e-07,
+      "loss": 1.5502,
+      "step": 765
+    },
+    {
+      "epoch": 0.15898713158987132,
+      "grad_norm": 1.1213296450340349,
+      "learning_rate": 3.780103343535697e-07,
+      "loss": 1.4563,
+      "step": 766
+    },
+    {
+      "epoch": 0.15919468659194685,
+      "grad_norm": 2.250107999233427,
+      "learning_rate": 3.779540850981806e-07,
+      "loss": 1.6008,
+      "step": 767
+    },
+    {
+      "epoch": 0.15940224159402241,
+      "grad_norm": 0.6866332849690373,
+      "learning_rate": 3.77897768684908e-07,
+      "loss": 1.4932,
+      "step": 768
+    },
+    {
+      "epoch": 0.15960979659609797,
+      "grad_norm": 0.882567229268511,
+      "learning_rate": 3.778413851376961e-07,
+      "loss": 1.5265,
+      "step": 769
+    },
+    {
+      "epoch": 0.1598173515981735,
+      "grad_norm": 0.8981771172535955,
+      "learning_rate": 3.777849344805177e-07,
+      "loss": 1.4927,
+      "step": 770
+    },
+    {
+      "epoch": 0.16002490660024907,
+      "grad_norm": 1.8562640629195606,
+      "learning_rate": 3.7772841673737406e-07,
+      "loss": 1.5417,
+      "step": 771
+    },
+    {
+      "epoch": 0.16023246160232463,
+      "grad_norm": 0.7691663161504118,
+      "learning_rate": 3.776718319322951e-07,
+      "loss": 1.557,
+      "step": 772
+    },
+    {
+      "epoch": 0.16044001660440016,
+      "grad_norm": 0.7789547731555111,
+      "learning_rate": 3.776151800893392e-07,
+      "loss": 1.5567,
+      "step": 773
+    },
+    {
+      "epoch": 0.16064757160647572,
+      "grad_norm": 1.0760219210740214,
+      "learning_rate": 3.7755846123259316e-07,
+      "loss": 1.5842,
+      "step": 774
+    },
+    {
+      "epoch": 0.16085512660855128,
+      "grad_norm": 1.4460571422597417,
+      "learning_rate": 3.7750167538617225e-07,
+      "loss": 1.552,
+      "step": 775
+    },
+    {
+      "epoch": 0.1610626816106268,
+      "grad_norm": 0.7845528737582512,
+      "learning_rate": 3.7744482257422046e-07,
+      "loss": 1.4935,
+      "step": 776
+    },
+    {
+      "epoch": 0.16127023661270237,
+      "grad_norm": 1.2498517734264016,
+      "learning_rate": 3.7738790282091e-07,
+      "loss": 1.4891,
+      "step": 777
+    },
+    {
+      "epoch": 0.1614777916147779,
+      "grad_norm": 1.0218972126350483,
+      "learning_rate": 3.773309161504417e-07,
+      "loss": 1.5006,
+      "step": 778
+    },
+    {
+      "epoch": 0.16168534661685346,
+      "grad_norm": 0.819571163501824,
+      "learning_rate": 3.7727386258704484e-07,
+      "loss": 1.5268,
+      "step": 779
+    },
+    {
+      "epoch": 0.16189290161892902,
+      "grad_norm": 0.8005862565577075,
+      "learning_rate": 3.772167421549769e-07,
+      "loss": 1.5299,
+      "step": 780
+    },
+    {
+      "epoch": 0.16210045662100456,
+      "grad_norm": 0.9139233624251284,
+      "learning_rate": 3.7715955487852404e-07,
+      "loss": 1.4307,
+      "step": 781
+    },
+    {
+      "epoch": 0.16230801162308012,
+      "grad_norm": 0.7901949312198839,
+      "learning_rate": 3.7710230078200087e-07,
+      "loss": 1.5438,
+      "step": 782
+    },
+    {
+      "epoch": 0.16251556662515568,
+      "grad_norm": 0.7869595762301359,
+      "learning_rate": 3.770449798897502e-07,
+      "loss": 1.4998,
+      "step": 783
+    },
+    {
+      "epoch": 0.1627231216272312,
+      "grad_norm": 1.1239926204726076,
+      "learning_rate": 3.7698759222614333e-07,
+      "loss": 1.5444,
+      "step": 784
+    },
+    {
+      "epoch": 0.16293067662930677,
+      "grad_norm": 0.9839945942049779,
+      "learning_rate": 3.7693013781558007e-07,
+      "loss": 1.5096,
+      "step": 785
+    },
+    {
+      "epoch": 0.16313823163138233,
+      "grad_norm": 0.8446556706083901,
+      "learning_rate": 3.7687261668248846e-07,
+      "loss": 1.6214,
+      "step": 786
+    },
+    {
+      "epoch": 0.16334578663345786,
+      "grad_norm": 0.8215434477984572,
+      "learning_rate": 3.7681502885132505e-07,
+      "loss": 1.5724,
+      "step": 787
+    },
+    {
+      "epoch": 0.16355334163553342,
+      "grad_norm": 1.3136394967632907,
+      "learning_rate": 3.7675737434657443e-07,
+      "loss": 1.5573,
+      "step": 788
+    },
+    {
+      "epoch": 0.16376089663760895,
+      "grad_norm": 0.9960939864046063,
+      "learning_rate": 3.7669965319275007e-07,
+      "loss": 1.5317,
+      "step": 789
+    },
+    {
+      "epoch": 0.16396845163968451,
+      "grad_norm": 2.882541005839287,
+      "learning_rate": 3.766418654143932e-07,
+      "loss": 1.5517,
+      "step": 790
+    },
+    {
+      "epoch": 0.16417600664176007,
+      "grad_norm": 1.9054982923296206,
+      "learning_rate": 3.765840110360738e-07,
+      "loss": 1.6166,
+      "step": 791
+    },
+    {
+      "epoch": 0.1643835616438356,
+      "grad_norm": 0.7741643971008524,
+      "learning_rate": 3.7652609008238994e-07,
+      "loss": 1.613,
+      "step": 792
+    },
+    {
+      "epoch": 0.16459111664591117,
+      "grad_norm": 0.7858165882736603,
+      "learning_rate": 3.7646810257796815e-07,
+      "loss": 1.5427,
+      "step": 793
+    },
+    {
+      "epoch": 0.16479867164798673,
+      "grad_norm": 0.7609302281547298,
+      "learning_rate": 3.7641004854746316e-07,
+      "loss": 1.4877,
+      "step": 794
+    },
+    {
+      "epoch": 0.16500622665006226,
+      "grad_norm": 1.0767136850229462,
+      "learning_rate": 3.763519280155579e-07,
+      "loss": 1.4492,
+      "step": 795
+    },
+    {
+      "epoch": 0.16521378165213782,
+      "grad_norm": 1.0139039715169584,
+      "learning_rate": 3.762937410069638e-07,
+      "loss": 1.6387,
+      "step": 796
+    },
+    {
+      "epoch": 0.16542133665421338,
+      "grad_norm": 1.0029450432854055,
+      "learning_rate": 3.762354875464204e-07,
+      "loss": 1.5181,
+      "step": 797
+    },
+    {
+      "epoch": 0.1656288916562889,
+      "grad_norm": 0.6617654130628386,
+      "learning_rate": 3.761771676586955e-07,
+      "loss": 1.6173,
+      "step": 798
+    },
+    {
+      "epoch": 0.16583644665836447,
+      "grad_norm": 0.7694723985765943,
+      "learning_rate": 3.7611878136858515e-07,
+      "loss": 1.5429,
+      "step": 799
+    },
+    {
+      "epoch": 0.16604400166044,
+      "grad_norm": 1.4364158600848693,
+      "learning_rate": 3.7606032870091375e-07,
+      "loss": 1.5577,
+      "step": 800
+    },
+    {
+      "epoch": 0.16625155666251556,
+      "grad_norm": 0.9573899326543218,
+      "learning_rate": 3.7600180968053367e-07,
+      "loss": 1.5363,
+      "step": 801
+    },
+    {
+      "epoch": 0.16645911166459113,
+      "grad_norm": 0.7776387411836702,
+      "learning_rate": 3.7594322433232577e-07,
+      "loss": 1.4798,
+      "step": 802
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 0.7716791738728939,
+      "learning_rate": 3.7588457268119895e-07,
+      "loss": 1.4645,
+      "step": 803
+    },
+    {
+      "epoch": 0.16687422166874222,
+      "grad_norm": 0.739806369506702,
+      "learning_rate": 3.7582585475209034e-07,
+      "loss": 1.5515,
+      "step": 804
+    },
+    {
+      "epoch": 0.16708177667081778,
+      "grad_norm": 0.7183959212972962,
+      "learning_rate": 3.7576707056996513e-07,
+      "loss": 1.5006,
+      "step": 805
+    },
+    {
+      "epoch": 0.1672893316728933,
+      "grad_norm": 0.6887813563888008,
+      "learning_rate": 3.75708220159817e-07,
+      "loss": 1.5263,
+      "step": 806
+    },
+    {
+      "epoch": 0.16749688667496887,
+      "grad_norm": 1.299764998227441,
+      "learning_rate": 3.7564930354666746e-07,
+      "loss": 1.5331,
+      "step": 807
+    },
+    {
+      "epoch": 0.16770444167704443,
+      "grad_norm": 1.285255017078133,
+      "learning_rate": 3.755903207555663e-07,
+      "loss": 1.6044,
+      "step": 808
+    },
+    {
+      "epoch": 0.16791199667911996,
+      "grad_norm": 0.9214982513478615,
+      "learning_rate": 3.7553127181159143e-07,
+      "loss": 1.5554,
+      "step": 809
+    },
+    {
+      "epoch": 0.16811955168119552,
+      "grad_norm": 0.951877560860422,
+      "learning_rate": 3.7547215673984887e-07,
+      "loss": 1.4915,
+      "step": 810
+    },
+    {
+      "epoch": 0.16832710668327105,
+      "grad_norm": 0.7164806531172606,
+      "learning_rate": 3.754129755654729e-07,
+      "loss": 1.4705,
+      "step": 811
+    },
+    {
+      "epoch": 0.16853466168534662,
+      "grad_norm": 0.7600708642780232,
+      "learning_rate": 3.753537283136256e-07,
+      "loss": 1.5385,
+      "step": 812
+    },
+    {
+      "epoch": 0.16874221668742218,
+      "grad_norm": 1.3021846223681952,
+      "learning_rate": 3.7529441500949746e-07,
+      "loss": 1.5315,
+      "step": 813
+    },
+    {
+      "epoch": 0.1689497716894977,
+      "grad_norm": 1.0454533350964896,
+      "learning_rate": 3.7523503567830695e-07,
+      "loss": 1.5139,
+      "step": 814
+    },
+    {
+      "epoch": 0.16915732669157327,
+      "grad_norm": 1.063871912964034,
+      "learning_rate": 3.751755903453005e-07,
+      "loss": 1.4556,
+      "step": 815
+    },
+    {
+      "epoch": 0.16936488169364883,
+      "grad_norm": 0.7236547821306161,
+      "learning_rate": 3.751160790357527e-07,
+      "loss": 1.4297,
+      "step": 816
+    },
+    {
+      "epoch": 0.16957243669572436,
+      "grad_norm": 0.8154056622027921,
+      "learning_rate": 3.750565017749662e-07,
+      "loss": 1.5575,
+      "step": 817
+    },
+    {
+      "epoch": 0.16977999169779992,
+      "grad_norm": 0.7501780708777914,
+      "learning_rate": 3.7499685858827163e-07,
+      "loss": 1.5315,
+      "step": 818
+    },
+    {
+      "epoch": 0.16998754669987548,
+      "grad_norm": 1.5401868414692843,
+      "learning_rate": 3.7493714950102775e-07,
+      "loss": 1.4922,
+      "step": 819
+    },
+    {
+      "epoch": 0.170195101701951,
+      "grad_norm": 0.8836452541027164,
+      "learning_rate": 3.748773745386212e-07,
+      "loss": 1.6307,
+      "step": 820
+    },
+    {
+      "epoch": 0.17040265670402657,
+      "grad_norm": 0.9940818448233938,
+      "learning_rate": 3.748175337264669e-07,
+      "loss": 1.5827,
+      "step": 821
+    },
+    {
+      "epoch": 0.1706102117061021,
+      "grad_norm": 1.2346967925024543,
+      "learning_rate": 3.747576270900073e-07,
+      "loss": 1.5735,
+      "step": 822
+    },
+    {
+      "epoch": 0.17081776670817767,
+      "grad_norm": 0.7040502153082228,
+      "learning_rate": 3.746976546547132e-07,
+      "loss": 1.5422,
+      "step": 823
+    },
+    {
+      "epoch": 0.17102532171025323,
+      "grad_norm": 2.1395500261759843,
+      "learning_rate": 3.7463761644608345e-07,
+      "loss": 1.5326,
+      "step": 824
+    },
+    {
+      "epoch": 0.17123287671232876,
+      "grad_norm": 0.8031348156184384,
+      "learning_rate": 3.7457751248964453e-07,
+      "loss": 1.5139,
+      "step": 825
+    },
+    {
+      "epoch": 0.17144043171440432,
+      "grad_norm": 0.9867951561957972,
+      "learning_rate": 3.7451734281095113e-07,
+      "loss": 1.5467,
+      "step": 826
+    },
+    {
+      "epoch": 0.17164798671647988,
+      "grad_norm": 0.7330872726008544,
+      "learning_rate": 3.744571074355857e-07,
+      "loss": 1.4889,
+      "step": 827
+    },
+    {
+      "epoch": 0.1718555417185554,
+      "grad_norm": 0.6702570202411895,
+      "learning_rate": 3.7439680638915883e-07,
+      "loss": 1.5805,
+      "step": 828
+    },
+    {
+      "epoch": 0.17206309672063097,
+      "grad_norm": 1.7414225240176324,
+      "learning_rate": 3.743364396973089e-07,
+      "loss": 1.5067,
+      "step": 829
+    },
+    {
+      "epoch": 0.17227065172270653,
+      "grad_norm": 1.3788620966291183,
+      "learning_rate": 3.7427600738570223e-07,
+      "loss": 1.5045,
+      "step": 830
+    },
+    {
+      "epoch": 0.17247820672478206,
+      "grad_norm": 1.0441706124192782,
+      "learning_rate": 3.7421550948003293e-07,
+      "loss": 1.5275,
+      "step": 831
+    },
+    {
+      "epoch": 0.17268576172685762,
+      "grad_norm": 0.7625393754633494,
+      "learning_rate": 3.741549460060233e-07,
+      "loss": 1.4881,
+      "step": 832
+    },
+    {
+      "epoch": 0.17289331672893316,
+      "grad_norm": 1.0895229101877355,
+      "learning_rate": 3.740943169894232e-07,
+      "loss": 1.5459,
+      "step": 833
+    },
+    {
+      "epoch": 0.17310087173100872,
+      "grad_norm": 0.8040144948497726,
+      "learning_rate": 3.740336224560104e-07,
+      "loss": 1.4751,
+      "step": 834
+    },
+    {
+      "epoch": 0.17330842673308428,
+      "grad_norm": 0.7585509525417161,
+      "learning_rate": 3.739728624315907e-07,
+      "loss": 1.6045,
+      "step": 835
+    },
+    {
+      "epoch": 0.1735159817351598,
+      "grad_norm": 0.7499658473582391,
+      "learning_rate": 3.739120369419977e-07,
+      "loss": 1.5168,
+      "step": 836
+    },
+    {
+      "epoch": 0.17372353673723537,
+      "grad_norm": 0.6851865494435526,
+      "learning_rate": 3.738511460130927e-07,
+      "loss": 1.5575,
+      "step": 837
+    },
+    {
+      "epoch": 0.17393109173931093,
+      "grad_norm": 0.7406286892034353,
+      "learning_rate": 3.737901896707649e-07,
+      "loss": 1.5889,
+      "step": 838
+    },
+    {
+      "epoch": 0.17413864674138646,
+      "grad_norm": 1.5302447360162559,
+      "learning_rate": 3.737291679409314e-07,
+      "loss": 1.5361,
+      "step": 839
+    },
+    {
+      "epoch": 0.17434620174346202,
+      "grad_norm": 0.7674357037181953,
+      "learning_rate": 3.7366808084953694e-07,
+      "loss": 1.517,
+      "step": 840
+    },
+    {
+      "epoch": 0.17455375674553758,
+      "grad_norm": 0.8205955230958445,
+      "learning_rate": 3.736069284225542e-07,
+      "loss": 1.5043,
+      "step": 841
+    },
+    {
+      "epoch": 0.1747613117476131,
+      "grad_norm": 1.0117168377981043,
+      "learning_rate": 3.7354571068598346e-07,
+      "loss": 1.5449,
+      "step": 842
+    },
+    {
+      "epoch": 0.17496886674968867,
+      "grad_norm": 1.246631413984115,
+      "learning_rate": 3.7348442766585297e-07,
+      "loss": 1.4927,
+      "step": 843
+    },
+    {
+      "epoch": 0.1751764217517642,
+      "grad_norm": 0.9163661457828539,
+      "learning_rate": 3.734230793882186e-07,
+      "loss": 1.5516,
+      "step": 844
+    },
+    {
+      "epoch": 0.17538397675383977,
+      "grad_norm": 1.9286964648397114,
+      "learning_rate": 3.733616658791641e-07,
+      "loss": 1.5384,
+      "step": 845
+    },
+    {
+      "epoch": 0.17559153175591533,
+      "grad_norm": 0.7276743750278539,
+      "learning_rate": 3.733001871648007e-07,
+      "loss": 1.5774,
+      "step": 846
+    },
+    {
+      "epoch": 0.17579908675799086,
+      "grad_norm": 0.69341900550815,
+      "learning_rate": 3.732386432712677e-07,
+      "loss": 1.5085,
+      "step": 847
+    },
+    {
+      "epoch": 0.17600664176006642,
+      "grad_norm": 1.2247329914173868,
+      "learning_rate": 3.7317703422473176e-07,
+      "loss": 1.6026,
+      "step": 848
+    },
+    {
+      "epoch": 0.17621419676214198,
+      "grad_norm": 0.7201691933069883,
+      "learning_rate": 3.731153600513874e-07,
+      "loss": 1.53,
+      "step": 849
+    },
+    {
+      "epoch": 0.1764217517642175,
+      "grad_norm": 0.7935021363305134,
+      "learning_rate": 3.730536207774571e-07,
+      "loss": 1.5283,
+      "step": 850
+    },
+    {
+      "epoch": 0.17662930676629307,
+      "grad_norm": 0.8553407464966226,
+      "learning_rate": 3.729918164291905e-07,
+      "loss": 1.5535,
+      "step": 851
+    },
+    {
+      "epoch": 0.17683686176836863,
+      "grad_norm": 0.8521875048893446,
+      "learning_rate": 3.729299470328653e-07,
+      "loss": 1.5066,
+      "step": 852
+    },
+    {
+      "epoch": 0.17704441677044416,
+      "grad_norm": 0.6300596698085998,
+      "learning_rate": 3.728680126147867e-07,
+      "loss": 1.4873,
+      "step": 853
+    },
+    {
+      "epoch": 0.17725197177251972,
+      "grad_norm": 0.9716499542134922,
+      "learning_rate": 3.728060132012875e-07,
+      "loss": 1.4733,
+      "step": 854
+    },
+    {
+      "epoch": 0.17745952677459526,
+      "grad_norm": 0.9470577841313167,
+      "learning_rate": 3.7274394881872825e-07,
+      "loss": 1.5461,
+      "step": 855
+    },
+    {
+      "epoch": 0.17766708177667082,
+      "grad_norm": 0.6711753702436024,
+      "learning_rate": 3.7268181949349707e-07,
+      "loss": 1.4783,
+      "step": 856
+    },
+    {
+      "epoch": 0.17787463677874638,
+      "grad_norm": 0.7533868330444985,
+      "learning_rate": 3.7261962525200975e-07,
+      "loss": 1.5497,
+      "step": 857
+    },
+    {
+      "epoch": 0.1780821917808219,
+      "grad_norm": 0.771491046373101,
+      "learning_rate": 3.725573661207096e-07,
+      "loss": 1.6642,
+      "step": 858
+    },
+    {
+      "epoch": 0.17828974678289747,
+      "grad_norm": 2.0122488463077763,
+      "learning_rate": 3.724950421260675e-07,
+      "loss": 1.6609,
+      "step": 859
+    },
+    {
+      "epoch": 0.17849730178497303,
+      "grad_norm": 1.0075132812128422,
+      "learning_rate": 3.7243265329458207e-07,
+      "loss": 1.4208,
+      "step": 860
+    },
+    {
+      "epoch": 0.17870485678704856,
+      "grad_norm": 0.6737551666121391,
+      "learning_rate": 3.7237019965277925e-07,
+      "loss": 1.5304,
+      "step": 861
+    },
+    {
+      "epoch": 0.17891241178912412,
+      "grad_norm": 0.7536570759048754,
+      "learning_rate": 3.7230768122721276e-07,
+      "loss": 1.4994,
+      "step": 862
+    },
+    {
+      "epoch": 0.17911996679119968,
+      "grad_norm": 0.8170442508203793,
+      "learning_rate": 3.7224509804446374e-07,
+      "loss": 1.5288,
+      "step": 863
+    },
+    {
+      "epoch": 0.1793275217932752,
+      "grad_norm": 1.0398047490069944,
+      "learning_rate": 3.7218245013114096e-07,
+      "loss": 1.5256,
+      "step": 864
+    },
+    {
+      "epoch": 0.17953507679535077,
+      "grad_norm": 1.060674553509126,
+      "learning_rate": 3.721197375138805e-07,
+      "loss": 1.6188,
+      "step": 865
+    },
+    {
+      "epoch": 0.1797426317974263,
+      "grad_norm": 0.705775955581245,
+      "learning_rate": 3.720569602193463e-07,
+      "loss": 1.5185,
+      "step": 866
+    },
+    {
+      "epoch": 0.17995018679950187,
+      "grad_norm": 0.8728614881558762,
+      "learning_rate": 3.7199411827422945e-07,
+      "loss": 1.5611,
+      "step": 867
+    },
+    {
+      "epoch": 0.18015774180157743,
+      "grad_norm": 0.6688368020576857,
+      "learning_rate": 3.719312117052487e-07,
+      "loss": 1.5199,
+      "step": 868
+    },
+    {
+      "epoch": 0.18036529680365296,
+      "grad_norm": 0.7063100522250276,
+      "learning_rate": 3.7186824053915037e-07,
+      "loss": 1.5521,
+      "step": 869
+    },
+    {
+      "epoch": 0.18057285180572852,
+      "grad_norm": 0.8757094097801504,
+      "learning_rate": 3.7180520480270794e-07,
+      "loss": 1.5097,
+      "step": 870
+    },
+    {
+      "epoch": 0.18078040680780408,
+      "grad_norm": 0.6748031450837038,
+      "learning_rate": 3.7174210452272264e-07,
+      "loss": 1.6066,
+      "step": 871
+    },
+    {
+      "epoch": 0.1809879618098796,
+      "grad_norm": 0.8329110099647076,
+      "learning_rate": 3.716789397260231e-07,
+      "loss": 1.4707,
+      "step": 872
+    },
+    {
+      "epoch": 0.18119551681195517,
+      "grad_norm": 2.6263362634731564,
+      "learning_rate": 3.7161571043946514e-07,
+      "loss": 1.4995,
+      "step": 873
+    },
+    {
+      "epoch": 0.18140307181403073,
+      "grad_norm": 0.804298429773292,
+      "learning_rate": 3.715524166899323e-07,
+      "loss": 1.5935,
+      "step": 874
+    },
+    {
+      "epoch": 0.18161062681610626,
+      "grad_norm": 0.9471952862287576,
+      "learning_rate": 3.714890585043354e-07,
+      "loss": 1.4568,
+      "step": 875
+    },
+    {
+      "epoch": 0.18181818181818182,
+      "grad_norm": 0.6480174575770449,
+      "learning_rate": 3.7142563590961257e-07,
+      "loss": 1.5473,
+      "step": 876
+    },
+    {
+      "epoch": 0.18202573682025736,
+      "grad_norm": 0.6933639173981995,
+      "learning_rate": 3.7136214893272957e-07,
+      "loss": 1.4835,
+      "step": 877
+    },
+    {
+      "epoch": 0.18223329182233292,
+      "grad_norm": 0.877362937258417,
+      "learning_rate": 3.712985976006792e-07,
+      "loss": 1.4228,
+      "step": 878
+    },
+    {
+      "epoch": 0.18244084682440848,
+      "grad_norm": 0.9080778927031753,
+      "learning_rate": 3.712349819404819e-07,
+      "loss": 1.573,
+      "step": 879
+    },
+    {
+      "epoch": 0.182648401826484,
+      "grad_norm": 0.7452963501835226,
+      "learning_rate": 3.7117130197918535e-07,
+      "loss": 1.5271,
+      "step": 880
+    },
+    {
+      "epoch": 0.18285595682855957,
+      "grad_norm": 0.7400615040014613,
+      "learning_rate": 3.711075577438645e-07,
+      "loss": 1.6004,
+      "step": 881
+    },
+    {
+      "epoch": 0.18306351183063513,
+      "grad_norm": 0.6838797669559068,
+      "learning_rate": 3.7104374926162186e-07,
+      "loss": 1.4635,
+      "step": 882
+    },
+    {
+      "epoch": 0.18327106683271066,
+      "grad_norm": 0.7255774359287342,
+      "learning_rate": 3.70979876559587e-07,
+      "loss": 1.493,
+      "step": 883
+    },
+    {
+      "epoch": 0.18347862183478622,
+      "grad_norm": 0.6900781540426623,
+      "learning_rate": 3.709159396649169e-07,
+      "loss": 1.5631,
+      "step": 884
+    },
+    {
+      "epoch": 0.18368617683686178,
+      "grad_norm": 0.8364041827031164,
+      "learning_rate": 3.708519386047959e-07,
+      "loss": 1.5954,
+      "step": 885
+    },
+    {
+      "epoch": 0.1838937318389373,
+      "grad_norm": 0.8083933588543124,
+      "learning_rate": 3.707878734064354e-07,
+      "loss": 1.5223,
+      "step": 886
+    },
+    {
+      "epoch": 0.18410128684101287,
+      "grad_norm": 0.7386738334350924,
+      "learning_rate": 3.7072374409707437e-07,
+      "loss": 1.537,
+      "step": 887
+    },
+    {
+      "epoch": 0.1843088418430884,
+      "grad_norm": 1.380905427986444,
+      "learning_rate": 3.7065955070397884e-07,
+      "loss": 1.5444,
+      "step": 888
+    },
+    {
+      "epoch": 0.18451639684516397,
+      "grad_norm": 1.6275491039959697,
+      "learning_rate": 3.705952932544421e-07,
+      "loss": 1.4941,
+      "step": 889
+    },
+    {
+      "epoch": 0.18472395184723953,
+      "grad_norm": 1.5905168044510367,
+      "learning_rate": 3.7053097177578477e-07,
+      "loss": 1.5098,
+      "step": 890
+    },
+    {
+      "epoch": 0.18493150684931506,
+      "grad_norm": 0.8233384585158533,
+      "learning_rate": 3.7046658629535463e-07,
+      "loss": 1.5493,
+      "step": 891
+    },
+    {
+      "epoch": 0.18513906185139062,
+      "grad_norm": 0.6738388983752422,
+      "learning_rate": 3.704021368405266e-07,
+      "loss": 1.5084,
+      "step": 892
+    },
+    {
+      "epoch": 0.18534661685346618,
+      "grad_norm": 1.0263124288396517,
+      "learning_rate": 3.70337623438703e-07,
+      "loss": 1.5326,
+      "step": 893
+    },
+    {
+      "epoch": 0.1855541718555417,
+      "grad_norm": 0.653321010932783,
+      "learning_rate": 3.7027304611731314e-07,
+      "loss": 1.4953,
+      "step": 894
+    },
+    {
+      "epoch": 0.18576172685761727,
+      "grad_norm": 1.6973143792482301,
+      "learning_rate": 3.702084049038136e-07,
+      "loss": 1.5212,
+      "step": 895
+    },
+    {
+      "epoch": 0.18596928185969283,
+      "grad_norm": 0.6893987881742105,
+      "learning_rate": 3.7014369982568806e-07,
+      "loss": 1.5523,
+      "step": 896
+    },
+    {
+      "epoch": 0.18617683686176836,
+      "grad_norm": 0.8584832425803451,
+      "learning_rate": 3.700789309104475e-07,
+      "loss": 1.4919,
+      "step": 897
+    },
+    {
+      "epoch": 0.18638439186384392,
+      "grad_norm": 0.7513471717647343,
+      "learning_rate": 3.700140981856298e-07,
+      "loss": 1.5182,
+      "step": 898
+    },
+    {
+      "epoch": 0.18659194686591946,
+      "grad_norm": 0.7136741726048977,
+      "learning_rate": 3.699492016788003e-07,
+      "loss": 1.5409,
+      "step": 899
+    },
+    {
+      "epoch": 0.18679950186799502,
+      "grad_norm": 0.7957898079494814,
+      "learning_rate": 3.6988424141755104e-07,
+      "loss": 1.5401,
+      "step": 900
+    },
+    {
+      "epoch": 0.18700705687007058,
+      "grad_norm": 0.9799306871328185,
+      "learning_rate": 3.6981921742950164e-07,
+      "loss": 1.603,
+      "step": 901
+    },
+    {
+      "epoch": 0.1872146118721461,
+      "grad_norm": 1.5115008253198206,
+      "learning_rate": 3.6975412974229847e-07,
+      "loss": 1.5116,
+      "step": 902
+    },
+    {
+      "epoch": 0.18742216687422167,
+      "grad_norm": 0.6879554751128457,
+      "learning_rate": 3.6968897838361505e-07,
+      "loss": 1.5114,
+      "step": 903
+    },
+    {
+      "epoch": 0.18762972187629723,
+      "grad_norm": 0.9754190823941983,
+      "learning_rate": 3.69623763381152e-07,
+      "loss": 1.4938,
+      "step": 904
+    },
+    {
+      "epoch": 0.18783727687837276,
+      "grad_norm": 1.4617529588786387,
+      "learning_rate": 3.6955848476263706e-07,
+      "loss": 1.5417,
+      "step": 905
+    },
+    {
+      "epoch": 0.18804483188044832,
+      "grad_norm": 0.7202516177903175,
+      "learning_rate": 3.69493142555825e-07,
+      "loss": 1.5363,
+      "step": 906
+    },
+    {
+      "epoch": 0.18825238688252388,
+      "grad_norm": 0.8183756175881098,
+      "learning_rate": 3.694277367884975e-07,
+      "loss": 1.5479,
+      "step": 907
+    },
+    {
+      "epoch": 0.1884599418845994,
+      "grad_norm": 0.7059895066638223,
+      "learning_rate": 3.693622674884634e-07,
+      "loss": 1.5138,
+      "step": 908
+    },
+    {
+      "epoch": 0.18866749688667497,
+      "grad_norm": 0.810042677809564,
+      "learning_rate": 3.6929673468355846e-07,
+      "loss": 1.513,
+      "step": 909
+    },
+    {
+      "epoch": 0.1888750518887505,
+      "grad_norm": 0.6904054094558383,
+      "learning_rate": 3.692311384016457e-07,
+      "loss": 1.5107,
+      "step": 910
+    },
+    {
+      "epoch": 0.18908260689082607,
+      "grad_norm": 0.711361133621851,
+      "learning_rate": 3.691654786706146e-07,
+      "loss": 1.561,
+      "step": 911
+    },
+    {
+      "epoch": 0.18929016189290163,
+      "grad_norm": 0.9074535190825969,
+      "learning_rate": 3.6909975551838215e-07,
+      "loss": 1.5117,
+      "step": 912
+    },
+    {
+      "epoch": 0.18949771689497716,
+      "grad_norm": 1.0371280666571567,
+      "learning_rate": 3.69033968972892e-07,
+      "loss": 1.5839,
+      "step": 913
+    },
+    {
+      "epoch": 0.18970527189705272,
+      "grad_norm": 0.7474696174083512,
+      "learning_rate": 3.689681190621149e-07,
+      "loss": 1.484,
+      "step": 914
+    },
+    {
+      "epoch": 0.18991282689912828,
+      "grad_norm": 0.8819371001062164,
+      "learning_rate": 3.689022058140484e-07,
+      "loss": 1.4513,
+      "step": 915
+    },
+    {
+      "epoch": 0.1901203819012038,
+      "grad_norm": 0.7444399301468259,
+      "learning_rate": 3.6883622925671715e-07,
+      "loss": 1.6168,
+      "step": 916
+    },
+    {
+      "epoch": 0.19032793690327937,
+      "grad_norm": 0.9405246697918023,
+      "learning_rate": 3.687701894181726e-07,
+      "loss": 1.5099,
+      "step": 917
+    },
+    {
+      "epoch": 0.19053549190535493,
+      "grad_norm": 0.8131743488768268,
+      "learning_rate": 3.6870408632649315e-07,
+      "loss": 1.5917,
+      "step": 918
+    },
+    {
+      "epoch": 0.19074304690743046,
+      "grad_norm": 0.7051740423254481,
+      "learning_rate": 3.68637920009784e-07,
+      "loss": 1.6081,
+      "step": 919
+    },
+    {
+      "epoch": 0.19095060190950602,
+      "grad_norm": 0.66563976531375,
+      "learning_rate": 3.6857169049617746e-07,
+      "loss": 1.4995,
+      "step": 920
+    },
+    {
+      "epoch": 0.19115815691158156,
+      "grad_norm": 1.0106331116387341,
+      "learning_rate": 3.6850539781383237e-07,
+      "loss": 1.5841,
+      "step": 921
+    },
+    {
+      "epoch": 0.19136571191365712,
+      "grad_norm": 1.6137885675277175,
+      "learning_rate": 3.684390419909348e-07,
+      "loss": 1.4778,
+      "step": 922
+    },
+    {
+      "epoch": 0.19157326691573268,
+      "grad_norm": 0.9493816808872487,
+      "learning_rate": 3.6837262305569744e-07,
+      "loss": 1.5693,
+      "step": 923
+    },
+    {
+      "epoch": 0.1917808219178082,
+      "grad_norm": 1.2255426237804923,
+      "learning_rate": 3.6830614103635976e-07,
+      "loss": 1.46,
+      "step": 924
+    },
+    {
+      "epoch": 0.19198837691988377,
+      "grad_norm": 0.7348959172677256,
+      "learning_rate": 3.6823959596118825e-07,
+      "loss": 1.5317,
+      "step": 925
+    },
+    {
+      "epoch": 0.19219593192195933,
+      "grad_norm": 0.683884654379837,
+      "learning_rate": 3.68172987858476e-07,
+      "loss": 1.5956,
+      "step": 926
+    },
+    {
+      "epoch": 0.19240348692403486,
+      "grad_norm": 1.1680965814363948,
+      "learning_rate": 3.6810631675654316e-07,
+      "loss": 1.5995,
+      "step": 927
+    },
+    {
+      "epoch": 0.19261104192611042,
+      "grad_norm": 1.0226787982557772,
+      "learning_rate": 3.680395826837364e-07,
+      "loss": 1.5572,
+      "step": 928
+    },
+    {
+      "epoch": 0.19281859692818598,
+      "grad_norm": 0.7532513466103506,
+      "learning_rate": 3.6797278566842935e-07,
+      "loss": 1.5199,
+      "step": 929
+    },
+    {
+      "epoch": 0.1930261519302615,
+      "grad_norm": 1.281973832103749,
+      "learning_rate": 3.679059257390223e-07,
+      "loss": 1.4541,
+      "step": 930
+    },
+    {
+      "epoch": 0.19323370693233707,
+      "grad_norm": 0.7431152726788054,
+      "learning_rate": 3.678390029239422e-07,
+      "loss": 1.4473,
+      "step": 931
+    },
+    {
+      "epoch": 0.1934412619344126,
+      "grad_norm": 0.693932407814135,
+      "learning_rate": 3.6777201725164303e-07,
+      "loss": 1.5397,
+      "step": 932
+    },
+    {
+      "epoch": 0.19364881693648817,
+      "grad_norm": 0.7926036579526063,
+      "learning_rate": 3.6770496875060525e-07,
+      "loss": 1.5205,
+      "step": 933
+    },
+    {
+      "epoch": 0.19385637193856373,
+      "grad_norm": 0.77602026900396,
+      "learning_rate": 3.6763785744933614e-07,
+      "loss": 1.5172,
+      "step": 934
+    },
+    {
+      "epoch": 0.19406392694063926,
+      "grad_norm": 0.7133631793500962,
+      "learning_rate": 3.6757068337636955e-07,
+      "loss": 1.5946,
+      "step": 935
+    },
+    {
+      "epoch": 0.19427148194271482,
+      "grad_norm": 0.8588582311253812,
+      "learning_rate": 3.6750344656026617e-07,
+      "loss": 1.56,
+      "step": 936
+    },
+    {
+      "epoch": 0.19447903694479038,
+      "grad_norm": 1.525273574937671,
+      "learning_rate": 3.6743614702961334e-07,
+      "loss": 1.4686,
+      "step": 937
+    },
+    {
+      "epoch": 0.1946865919468659,
+      "grad_norm": 0.7366300680010816,
+      "learning_rate": 3.673687848130249e-07,
+      "loss": 1.5038,
+      "step": 938
+    },
+    {
+      "epoch": 0.19489414694894147,
+      "grad_norm": 2.1051036264314247,
+      "learning_rate": 3.673013599391417e-07,
+      "loss": 1.5599,
+      "step": 939
+    },
+    {
+      "epoch": 0.19510170195101703,
+      "grad_norm": 0.6794846997394501,
+      "learning_rate": 3.672338724366308e-07,
+      "loss": 1.4862,
+      "step": 940
+    },
+    {
+      "epoch": 0.19530925695309256,
+      "grad_norm": 0.8128017325234509,
+      "learning_rate": 3.6716632233418623e-07,
+      "loss": 1.5856,
+      "step": 941
+    },
+    {
+      "epoch": 0.19551681195516812,
+      "grad_norm": 0.7310885949394232,
+      "learning_rate": 3.6709870966052844e-07,
+      "loss": 1.4743,
+      "step": 942
+    },
+    {
+      "epoch": 0.19572436695724366,
+      "grad_norm": 0.8202315102887064,
+      "learning_rate": 3.6703103444440453e-07,
+      "loss": 1.4866,
+      "step": 943
+    },
+    {
+      "epoch": 0.19593192195931922,
+      "grad_norm": 0.8378121328849348,
+      "learning_rate": 3.6696329671458827e-07,
+      "loss": 1.5544,
+      "step": 944
+    },
+    {
+      "epoch": 0.19613947696139478,
+      "grad_norm": 2.4985488718347386,
+      "learning_rate": 3.6689549649987983e-07,
+      "loss": 1.5403,
+      "step": 945
+    },
+    {
+      "epoch": 0.1963470319634703,
+      "grad_norm": 0.751146122439667,
+      "learning_rate": 3.668276338291062e-07,
+      "loss": 1.5288,
+      "step": 946
+    },
+    {
+      "epoch": 0.19655458696554587,
+      "grad_norm": 0.7436847988145037,
+      "learning_rate": 3.6675970873112065e-07,
+      "loss": 1.5191,
+      "step": 947
+    },
+    {
+      "epoch": 0.19676214196762143,
+      "grad_norm": 0.7199080583641725,
+      "learning_rate": 3.6669172123480326e-07,
+      "loss": 1.5131,
+      "step": 948
+    },
+    {
+      "epoch": 0.19696969696969696,
+      "grad_norm": 0.7160545360517595,
+      "learning_rate": 3.666236713690604e-07,
+      "loss": 1.5916,
+      "step": 949
+    },
+    {
+      "epoch": 0.19717725197177252,
+      "grad_norm": 0.7501750418287525,
+      "learning_rate": 3.6655555916282515e-07,
+      "loss": 1.5683,
+      "step": 950
+    },
+    {
+      "epoch": 0.19738480697384808,
+      "grad_norm": 2.8370077003246053,
+      "learning_rate": 3.6648738464505697e-07,
+      "loss": 1.4981,
+      "step": 951
+    },
+    {
+      "epoch": 0.1975923619759236,
+      "grad_norm": 0.6644047637169577,
+      "learning_rate": 3.664191478447418e-07,
+      "loss": 1.4997,
+      "step": 952
+    },
+    {
+      "epoch": 0.19779991697799917,
+      "grad_norm": 0.7954094260949119,
+      "learning_rate": 3.6635084879089224e-07,
+      "loss": 1.5484,
+      "step": 953
+    },
+    {
+      "epoch": 0.1980074719800747,
+      "grad_norm": 0.7248770408981893,
+      "learning_rate": 3.662824875125471e-07,
+      "loss": 1.5204,
+      "step": 954
+    },
+    {
+      "epoch": 0.19821502698215027,
+      "grad_norm": 0.7250852204620137,
+      "learning_rate": 3.662140640387719e-07,
+      "loss": 1.4125,
+      "step": 955
+    },
+    {
+      "epoch": 0.19842258198422583,
+      "grad_norm": 3.1496163510577446,
+      "learning_rate": 3.661455783986584e-07,
+      "loss": 1.5167,
+      "step": 956
+    },
+    {
+      "epoch": 0.19863013698630136,
+      "grad_norm": 1.2661657661674592,
+      "learning_rate": 3.6607703062132496e-07,
+      "loss": 1.5511,
+      "step": 957
+    },
+    {
+      "epoch": 0.19883769198837692,
+      "grad_norm": 0.6500819347470116,
+      "learning_rate": 3.660084207359162e-07,
+      "loss": 1.5676,
+      "step": 958
+    },
+    {
+      "epoch": 0.19904524699045248,
+      "grad_norm": 1.3592590933844422,
+      "learning_rate": 3.659397487716032e-07,
+      "loss": 1.5252,
+      "step": 959
+    },
+    {
+      "epoch": 0.199252801992528,
+      "grad_norm": 2.669712555389213,
+      "learning_rate": 3.658710147575836e-07,
+      "loss": 1.5638,
+      "step": 960
+    },
+    {
+      "epoch": 0.19946035699460357,
+      "grad_norm": 1.3276804343738915,
+      "learning_rate": 3.6580221872308117e-07,
+      "loss": 1.4969,
+      "step": 961
+    },
+    {
+      "epoch": 0.19966791199667913,
+      "grad_norm": 1.332614242418297,
+      "learning_rate": 3.6573336069734607e-07,
+      "loss": 1.5383,
+      "step": 962
+    },
+    {
+      "epoch": 0.19987546699875466,
+      "grad_norm": 0.6914687241593803,
+      "learning_rate": 3.656644407096551e-07,
+      "loss": 1.6373,
+      "step": 963
+    },
+    {
+      "epoch": 0.20008302200083022,
+      "grad_norm": 0.7055726189531973,
+      "learning_rate": 3.65595458789311e-07,
+      "loss": 1.49,
+      "step": 964
+    },
+    {
+      "epoch": 0.20029057700290576,
+      "grad_norm": 1.0737501275797365,
+      "learning_rate": 3.655264149656432e-07,
+      "loss": 1.4932,
+      "step": 965
+    },
+    {
+      "epoch": 0.20049813200498132,
+      "grad_norm": 0.8143517884994579,
+      "learning_rate": 3.6545730926800734e-07,
+      "loss": 1.5204,
+      "step": 966
+    },
+    {
+      "epoch": 0.20070568700705688,
+      "grad_norm": 0.7068450228989227,
+      "learning_rate": 3.653881417257852e-07,
+      "loss": 1.4741,
+      "step": 967
+    },
+    {
+      "epoch": 0.2009132420091324,
+      "grad_norm": 0.8868365348832494,
+      "learning_rate": 3.65318912368385e-07,
+      "loss": 1.5691,
+      "step": 968
+    },
+    {
+      "epoch": 0.20112079701120797,
+      "grad_norm": 1.1549741126915505,
+      "learning_rate": 3.6524962122524133e-07,
+      "loss": 1.5953,
+      "step": 969
+    },
+    {
+      "epoch": 0.20132835201328353,
+      "grad_norm": 0.7458480493247286,
+      "learning_rate": 3.6518026832581483e-07,
+      "loss": 1.533,
+      "step": 970
+    },
+    {
+      "epoch": 0.20153590701535906,
+      "grad_norm": 0.7608734275214517,
+      "learning_rate": 3.6511085369959256e-07,
+      "loss": 1.4789,
+      "step": 971
+    },
+    {
+      "epoch": 0.20174346201743462,
+      "grad_norm": 0.7324264738249613,
+      "learning_rate": 3.650413773760878e-07,
+      "loss": 1.5215,
+      "step": 972
+    },
+    {
+      "epoch": 0.20195101701951018,
+      "grad_norm": 0.6115945082172204,
+      "learning_rate": 3.6497183938484e-07,
+      "loss": 1.4563,
+      "step": 973
+    },
+    {
+      "epoch": 0.20215857202158571,
+      "grad_norm": 0.9220845304511379,
+      "learning_rate": 3.6490223975541486e-07,
+      "loss": 1.5536,
+      "step": 974
+    },
+    {
+      "epoch": 0.20236612702366127,
+      "grad_norm": 0.6752012126198906,
+      "learning_rate": 3.648325785174043e-07,
+      "loss": 1.5684,
+      "step": 975
+    },
+    {
+      "epoch": 0.2025736820257368,
+      "grad_norm": 0.7300003639798002,
+      "learning_rate": 3.647628557004265e-07,
+      "loss": 1.4659,
+      "step": 976
+    },
+    {
+      "epoch": 0.20278123702781237,
+      "grad_norm": 1.1357719371195425,
+      "learning_rate": 3.6469307133412563e-07,
+      "loss": 1.4672,
+      "step": 977
+    },
+    {
+      "epoch": 0.20298879202988793,
+      "grad_norm": 0.7413194904020192,
+      "learning_rate": 3.646232254481722e-07,
+      "loss": 1.5692,
+      "step": 978
+    },
+    {
+      "epoch": 0.20319634703196346,
+      "grad_norm": 0.7805517462648436,
+      "learning_rate": 3.645533180722629e-07,
+      "loss": 1.6109,
+      "step": 979
+    },
+    {
+      "epoch": 0.20340390203403902,
+      "grad_norm": 0.7060715170778643,
+      "learning_rate": 3.644833492361204e-07,
+      "loss": 1.5008,
+      "step": 980
+    },
+    {
+      "epoch": 0.20361145703611458,
+      "grad_norm": 0.8046242189341722,
+      "learning_rate": 3.6441331896949357e-07,
+      "loss": 1.5687,
+      "step": 981
+    },
+    {
+      "epoch": 0.2038190120381901,
+      "grad_norm": 0.7023078197237743,
+      "learning_rate": 3.643432273021575e-07,
+      "loss": 1.6225,
+      "step": 982
+    },
+    {
+      "epoch": 0.20402656704026567,
+      "grad_norm": 0.9867549783489455,
+      "learning_rate": 3.6427307426391334e-07,
+      "loss": 1.5635,
+      "step": 983
+    },
+    {
+      "epoch": 0.20423412204234123,
+      "grad_norm": 1.1974338432887788,
+      "learning_rate": 3.642028598845882e-07,
+      "loss": 1.5652,
+      "step": 984
+    },
+    {
+      "epoch": 0.20444167704441676,
+      "grad_norm": 0.9192244124163903,
+      "learning_rate": 3.6413258419403536e-07,
+      "loss": 1.5241,
+      "step": 985
+    },
+    {
+      "epoch": 0.20464923204649232,
+      "grad_norm": 0.7081790678194302,
+      "learning_rate": 3.640622472221342e-07,
+      "loss": 1.5447,
+      "step": 986
+    },
+    {
+      "epoch": 0.20485678704856788,
+      "grad_norm": 1.1077921292682111,
+      "learning_rate": 3.6399184899879023e-07,
+      "loss": 1.5354,
+      "step": 987
+    },
+    {
+      "epoch": 0.20506434205064342,
+      "grad_norm": 1.4672286013569629,
+      "learning_rate": 3.639213895539349e-07,
+      "loss": 1.4619,
+      "step": 988
+    },
+    {
+      "epoch": 0.20527189705271898,
+      "grad_norm": 0.804495510022731,
+      "learning_rate": 3.6385086891752546e-07,
+      "loss": 1.498,
+      "step": 989
+    },
+    {
+      "epoch": 0.2054794520547945,
+      "grad_norm": 0.7088277940209514,
+      "learning_rate": 3.6378028711954565e-07,
+      "loss": 1.538,
+      "step": 990
+    },
+    {
+      "epoch": 0.20568700705687007,
+      "grad_norm": 0.9804258263528919,
+      "learning_rate": 3.637096441900049e-07,
+      "loss": 1.5722,
+      "step": 991
+    },
+    {
+      "epoch": 0.20589456205894563,
+      "grad_norm": 0.915570671843891,
+      "learning_rate": 3.6363894015893876e-07,
+      "loss": 1.507,
+      "step": 992
+    },
+    {
+      "epoch": 0.20610211706102116,
+      "grad_norm": 0.7401175250292767,
+      "learning_rate": 3.6356817505640865e-07,
+      "loss": 1.5251,
+      "step": 993
+    },
+    {
+      "epoch": 0.20630967206309672,
+      "grad_norm": 0.7839081841918215,
+      "learning_rate": 3.63497348912502e-07,
+      "loss": 1.5706,
+      "step": 994
+    },
+    {
+      "epoch": 0.20651722706517228,
+      "grad_norm": 0.923237808745583,
+      "learning_rate": 3.6342646175733226e-07,
+      "loss": 1.4727,
+      "step": 995
+    },
+    {
+      "epoch": 0.20672478206724781,
+      "grad_norm": 0.7647205333375906,
+      "learning_rate": 3.633555136210387e-07,
+      "loss": 1.5317,
+      "step": 996
+    },
+    {
+      "epoch": 0.20693233706932337,
+      "grad_norm": 0.7055991007802039,
+      "learning_rate": 3.632845045337866e-07,
+      "loss": 1.5603,
+      "step": 997
+    },
+    {
+      "epoch": 0.20713989207139893,
+      "grad_norm": 0.6390969879755061,
+      "learning_rate": 3.6321343452576716e-07,
+      "loss": 1.5589,
+      "step": 998
+    },
+    {
+      "epoch": 0.20734744707347447,
+      "grad_norm": 0.7282093784272053,
+      "learning_rate": 3.631423036271975e-07,
+      "loss": 1.6313,
+      "step": 999
+    },
+    {
+      "epoch": 0.20755500207555003,
+      "grad_norm": 1.475766775717144,
+      "learning_rate": 3.6307111186832057e-07,
+      "loss": 1.5253,
+      "step": 1000
+    },
+    {
+      "epoch": 0.20776255707762556,
+      "grad_norm": 0.9252510632177006,
+      "learning_rate": 3.6299985927940517e-07,
+      "loss": 1.4318,
+      "step": 1001
+    },
+    {
+      "epoch": 0.20797011207970112,
+      "grad_norm": 0.7737218048296447,
+      "learning_rate": 3.6292854589074604e-07,
+      "loss": 1.6498,
+      "step": 1002
+    },
+    {
+      "epoch": 0.20817766708177668,
+      "grad_norm": 1.0312417872879511,
+      "learning_rate": 3.6285717173266377e-07,
+      "loss": 1.5135,
+      "step": 1003
+    },
+    {
+      "epoch": 0.2083852220838522,
+      "grad_norm": 0.9303580669090593,
+      "learning_rate": 3.6278573683550464e-07,
+      "loss": 1.461,
+      "step": 1004
+    },
+    {
+      "epoch": 0.20859277708592777,
+      "grad_norm": 0.8987816836485442,
+      "learning_rate": 3.62714241229641e-07,
+      "loss": 1.5267,
+      "step": 1005
+    },
+    {
+      "epoch": 0.20880033208800333,
+      "grad_norm": 1.0573161719993172,
+      "learning_rate": 3.626426849454708e-07,
+      "loss": 1.4593,
+      "step": 1006
+    },
+    {
+      "epoch": 0.20900788709007886,
+      "grad_norm": 0.882982330954709,
+      "learning_rate": 3.6257106801341796e-07,
+      "loss": 1.5643,
+      "step": 1007
+    },
+    {
+      "epoch": 0.20921544209215442,
+      "grad_norm": 0.6274525117755215,
+      "learning_rate": 3.62499390463932e-07,
+      "loss": 1.6032,
+      "step": 1008
+    },
+    {
+      "epoch": 0.20942299709422998,
+      "grad_norm": 1.0591429128905,
+      "learning_rate": 3.6242765232748835e-07,
+      "loss": 1.5049,
+      "step": 1009
+    },
+    {
+      "epoch": 0.20963055209630552,
+      "grad_norm": 0.72923807888404,
+      "learning_rate": 3.6235585363458826e-07,
+      "loss": 1.4759,
+      "step": 1010
+    },
+    {
+      "epoch": 0.20983810709838108,
+      "grad_norm": 0.786796182824585,
+      "learning_rate": 3.6228399441575847e-07,
+      "loss": 1.4956,
+      "step": 1011
+    },
+    {
+      "epoch": 0.2100456621004566,
+      "grad_norm": 0.726427879589452,
+      "learning_rate": 3.622120747015517e-07,
+      "loss": 1.5475,
+      "step": 1012
+    },
+    {
+      "epoch": 0.21025321710253217,
+      "grad_norm": 0.692097382892561,
+      "learning_rate": 3.621400945225463e-07,
+      "loss": 1.5879,
+      "step": 1013
+    },
+    {
+      "epoch": 0.21046077210460773,
+      "grad_norm": 0.8368173235338542,
+      "learning_rate": 3.620680539093463e-07,
+      "loss": 1.5438,
+      "step": 1014
+    },
+    {
+      "epoch": 0.21066832710668326,
+      "grad_norm": 0.7655157283855119,
+      "learning_rate": 3.619959528925814e-07,
+      "loss": 1.5525,
+      "step": 1015
+    },
+    {
+      "epoch": 0.21087588210875882,
+      "grad_norm": 0.9930040953786426,
+      "learning_rate": 3.619237915029072e-07,
+      "loss": 1.5809,
+      "step": 1016
+    },
+    {
+      "epoch": 0.21108343711083438,
+      "grad_norm": 1.2270945411752896,
+      "learning_rate": 3.6185156977100465e-07,
+      "loss": 1.6398,
+      "step": 1017
+    },
+    {
+      "epoch": 0.21129099211290991,
+      "grad_norm": 0.7786878778901891,
+      "learning_rate": 3.617792877275806e-07,
+      "loss": 1.5092,
+      "step": 1018
+    },
+    {
+      "epoch": 0.21149854711498547,
+      "grad_norm": 0.760819941879524,
+      "learning_rate": 3.617069454033675e-07,
+      "loss": 1.5585,
+      "step": 1019
+    },
+    {
+      "epoch": 0.21170610211706103,
+      "grad_norm": 0.9630157097008851,
+      "learning_rate": 3.616345428291232e-07,
+      "loss": 1.6263,
+      "step": 1020
+    },
+    {
+      "epoch": 0.21191365711913657,
+      "grad_norm": 1.7057018584263368,
+      "learning_rate": 3.6156208003563154e-07,
+      "loss": 1.4996,
+      "step": 1021
+    },
+    {
+      "epoch": 0.21212121212121213,
+      "grad_norm": 0.7251056277062933,
+      "learning_rate": 3.614895570537017e-07,
+      "loss": 1.5644,
+      "step": 1022
+    },
+    {
+      "epoch": 0.21232876712328766,
+      "grad_norm": 1.1121933879502364,
+      "learning_rate": 3.6141697391416856e-07,
+      "loss": 1.4901,
+      "step": 1023
+    },
+    {
+      "epoch": 0.21253632212536322,
+      "grad_norm": 0.9028882682012563,
+      "learning_rate": 3.613443306478925e-07,
+      "loss": 1.5683,
+      "step": 1024
+    },
+    {
+      "epoch": 0.21274387712743878,
+      "grad_norm": 0.6827236143260521,
+      "learning_rate": 3.6127162728575957e-07,
+      "loss": 1.5061,
+      "step": 1025
+    },
+    {
+      "epoch": 0.2129514321295143,
+      "grad_norm": 0.8541047091180142,
+      "learning_rate": 3.6119886385868117e-07,
+      "loss": 1.4914,
+      "step": 1026
+    },
+    {
+      "epoch": 0.21315898713158987,
+      "grad_norm": 0.8808583449154139,
+      "learning_rate": 3.6112604039759455e-07,
+      "loss": 1.54,
+      "step": 1027
+    },
+    {
+      "epoch": 0.21336654213366543,
+      "grad_norm": 0.673300656333197,
+      "learning_rate": 3.610531569334622e-07,
+      "loss": 1.4505,
+      "step": 1028
+    },
+    {
+      "epoch": 0.21357409713574096,
+      "grad_norm": 0.7648703777499639,
+      "learning_rate": 3.6098021349727226e-07,
+      "loss": 1.5625,
+      "step": 1029
+    },
+    {
+      "epoch": 0.21378165213781652,
+      "grad_norm": 0.9545381270153108,
+      "learning_rate": 3.609072101200384e-07,
+      "loss": 1.5664,
+      "step": 1030
+    },
+    {
+      "epoch": 0.21398920713989208,
+      "grad_norm": 0.9043053771299132,
+      "learning_rate": 3.6083414683279966e-07,
+      "loss": 1.472,
+      "step": 1031
+    },
+    {
+      "epoch": 0.21419676214196762,
+      "grad_norm": 0.7919651328353591,
+      "learning_rate": 3.6076102366662056e-07,
+      "loss": 1.585,
+      "step": 1032
+    },
+    {
+      "epoch": 0.21440431714404318,
+      "grad_norm": 0.6758118295057567,
+      "learning_rate": 3.606878406525913e-07,
+      "loss": 1.4898,
+      "step": 1033
+    },
+    {
+      "epoch": 0.2146118721461187,
+      "grad_norm": 0.8554468899246301,
+      "learning_rate": 3.6061459782182715e-07,
+      "loss": 1.5818,
+      "step": 1034
+    },
+    {
+      "epoch": 0.21481942714819427,
+      "grad_norm": 1.1528963523575075,
+      "learning_rate": 3.6054129520546913e-07,
+      "loss": 1.6178,
+      "step": 1035
+    },
+    {
+      "epoch": 0.21502698215026983,
+      "grad_norm": 0.7750790035936163,
+      "learning_rate": 3.604679328346836e-07,
+      "loss": 1.5743,
+      "step": 1036
+    },
+    {
+      "epoch": 0.21523453715234536,
+      "grad_norm": 0.7969765912929939,
+      "learning_rate": 3.6039451074066214e-07,
+      "loss": 1.5569,
+      "step": 1037
+    },
+    {
+      "epoch": 0.21544209215442092,
+      "grad_norm": 2.2800106728768097,
+      "learning_rate": 3.60321028954622e-07,
+      "loss": 1.4997,
+      "step": 1038
+    },
+    {
+      "epoch": 0.21564964715649648,
+      "grad_norm": 0.6253355536777663,
+      "learning_rate": 3.602474875078058e-07,
+      "loss": 1.5601,
+      "step": 1039
+    },
+    {
+      "epoch": 0.21585720215857201,
+      "grad_norm": 0.7061438907606232,
+      "learning_rate": 3.6017388643148117e-07,
+      "loss": 1.5718,
+      "step": 1040
+    },
+    {
+      "epoch": 0.21606475716064757,
+      "grad_norm": 1.3897658753454325,
+      "learning_rate": 3.601002257569414e-07,
+      "loss": 1.5356,
+      "step": 1041
+    },
+    {
+      "epoch": 0.21627231216272313,
+      "grad_norm": 0.7199812896850963,
+      "learning_rate": 3.6002650551550515e-07,
+      "loss": 1.5841,
+      "step": 1042
+    },
+    {
+      "epoch": 0.21647986716479867,
+      "grad_norm": 0.6314436721875729,
+      "learning_rate": 3.599527257385162e-07,
+      "loss": 1.5291,
+      "step": 1043
+    },
+    {
+      "epoch": 0.21668742216687423,
+      "grad_norm": 0.8129927224059432,
+      "learning_rate": 3.5987888645734385e-07,
+      "loss": 1.6218,
+      "step": 1044
+    },
+    {
+      "epoch": 0.21689497716894976,
+      "grad_norm": 2.537115508337458,
+      "learning_rate": 3.598049877033825e-07,
+      "loss": 1.5163,
+      "step": 1045
+    },
+    {
+      "epoch": 0.21710253217102532,
+      "grad_norm": 1.1715340746389704,
+      "learning_rate": 3.59731029508052e-07,
+      "loss": 1.5736,
+      "step": 1046
+    },
+    {
+      "epoch": 0.21731008717310088,
+      "grad_norm": 0.8132969706930291,
+      "learning_rate": 3.596570119027974e-07,
+      "loss": 1.526,
+      "step": 1047
+    },
+    {
+      "epoch": 0.2175176421751764,
+      "grad_norm": 0.8127787669924381,
+      "learning_rate": 3.59582934919089e-07,
+      "loss": 1.6048,
+      "step": 1048
+    },
+    {
+      "epoch": 0.21772519717725197,
+      "grad_norm": 0.8264637428695621,
+      "learning_rate": 3.5950879858842246e-07,
+      "loss": 1.5814,
+      "step": 1049
+    },
+    {
+      "epoch": 0.21793275217932753,
+      "grad_norm": 0.7115903203561558,
+      "learning_rate": 3.594346029423184e-07,
+      "loss": 1.5222,
+      "step": 1050
+    },
+    {
+      "epoch": 0.21814030718140306,
+      "grad_norm": 0.8055282482632105,
+      "learning_rate": 3.5936034801232306e-07,
+      "loss": 1.4447,
+      "step": 1051
+    },
+    {
+      "epoch": 0.21834786218347862,
+      "grad_norm": 0.762652301141905,
+      "learning_rate": 3.592860338300075e-07,
+      "loss": 1.5596,
+      "step": 1052
+    },
+    {
+      "epoch": 0.21855541718555418,
+      "grad_norm": 0.8503416758302333,
+      "learning_rate": 3.592116604269682e-07,
+      "loss": 1.4316,
+      "step": 1053
+    },
+    {
+      "epoch": 0.21876297218762972,
+      "grad_norm": 0.9139415303597738,
+      "learning_rate": 3.5913722783482675e-07,
+      "loss": 1.5539,
+      "step": 1054
+    },
+    {
+      "epoch": 0.21897052718970528,
+      "grad_norm": 1.0700224764690445,
+      "learning_rate": 3.5906273608522984e-07,
+      "loss": 1.5497,
+      "step": 1055
+    },
+    {
+      "epoch": 0.2191780821917808,
+      "grad_norm": 0.8810454182499513,
+      "learning_rate": 3.5898818520984955e-07,
+      "loss": 1.5458,
+      "step": 1056
+    },
+    {
+      "epoch": 0.21938563719385637,
+      "grad_norm": 0.6512086563955461,
+      "learning_rate": 3.589135752403828e-07,
+      "loss": 1.4351,
+      "step": 1057
+    },
+    {
+      "epoch": 0.21959319219593193,
+      "grad_norm": 0.8475784004616901,
+      "learning_rate": 3.588389062085518e-07,
+      "loss": 1.6219,
+      "step": 1058
+    },
+    {
+      "epoch": 0.21980074719800746,
+      "grad_norm": 0.6729858419411026,
+      "learning_rate": 3.5876417814610385e-07,
+      "loss": 1.5021,
+      "step": 1059
+    },
+    {
+      "epoch": 0.22000830220008302,
+      "grad_norm": 1.0081611937448638,
+      "learning_rate": 3.5868939108481135e-07,
+      "loss": 1.5535,
+      "step": 1060
+    },
+    {
+      "epoch": 0.22021585720215858,
+      "grad_norm": 0.8878982715761973,
+      "learning_rate": 3.586145450564717e-07,
+      "loss": 1.5483,
+      "step": 1061
+    },
+    {
+      "epoch": 0.22042341220423411,
+      "grad_norm": 0.6796615513726271,
+      "learning_rate": 3.5853964009290755e-07,
+      "loss": 1.508,
+      "step": 1062
+    },
+    {
+      "epoch": 0.22063096720630967,
+      "grad_norm": 0.7574182326379953,
+      "learning_rate": 3.584646762259664e-07,
+      "loss": 1.6801,
+      "step": 1063
+    },
+    {
+      "epoch": 0.22083852220838524,
+      "grad_norm": 0.9577551078700357,
+      "learning_rate": 3.58389653487521e-07,
+      "loss": 1.5759,
+      "step": 1064
+    },
+    {
+      "epoch": 0.22104607721046077,
+      "grad_norm": 0.6856193490323519,
+      "learning_rate": 3.5831457190946896e-07,
+      "loss": 1.5297,
+      "step": 1065
+    },
+    {
+      "epoch": 0.22125363221253633,
+      "grad_norm": 0.9724950740515332,
+      "learning_rate": 3.582394315237329e-07,
+      "loss": 1.6117,
+      "step": 1066
+    },
+    {
+      "epoch": 0.22146118721461186,
+      "grad_norm": 1.0497306707331313,
+      "learning_rate": 3.581642323622607e-07,
+      "loss": 1.5273,
+      "step": 1067
+    },
+    {
+      "epoch": 0.22166874221668742,
+      "grad_norm": 0.6898577089872816,
+      "learning_rate": 3.58088974457025e-07,
+      "loss": 1.52,
+      "step": 1068
+    },
+    {
+      "epoch": 0.22187629721876298,
+      "grad_norm": 0.7527731255815604,
+      "learning_rate": 3.580136578400233e-07,
+      "loss": 1.5127,
+      "step": 1069
+    },
+    {
+      "epoch": 0.2220838522208385,
+      "grad_norm": 0.6751514202057007,
+      "learning_rate": 3.579382825432784e-07,
+      "loss": 1.5542,
+      "step": 1070
+    },
+    {
+      "epoch": 0.22229140722291407,
+      "grad_norm": 1.5318541751033339,
+      "learning_rate": 3.578628485988378e-07,
+      "loss": 1.6134,
+      "step": 1071
+    },
+    {
+      "epoch": 0.22249896222498963,
+      "grad_norm": 1.023045302434103,
+      "learning_rate": 3.57787356038774e-07,
+      "loss": 1.5041,
+      "step": 1072
+    },
+    {
+      "epoch": 0.22270651722706516,
+      "grad_norm": 1.082469823241705,
+      "learning_rate": 3.5771180489518457e-07,
+      "loss": 1.5004,
+      "step": 1073
+    },
+    {
+      "epoch": 0.22291407222914073,
+      "grad_norm": 0.8228670000560644,
+      "learning_rate": 3.576361952001916e-07,
+      "loss": 1.5062,
+      "step": 1074
+    },
+    {
+      "epoch": 0.22312162723121629,
+      "grad_norm": 1.0069016348680366,
+      "learning_rate": 3.575605269859425e-07,
+      "loss": 1.531,
+      "step": 1075
+    },
+    {
+      "epoch": 0.22332918223329182,
+      "grad_norm": 0.9620336005511578,
+      "learning_rate": 3.574848002846094e-07,
+      "loss": 1.5252,
+      "step": 1076
+    },
+    {
+      "epoch": 0.22353673723536738,
+      "grad_norm": 0.8815693906658952,
+      "learning_rate": 3.574090151283892e-07,
+      "loss": 1.4453,
+      "step": 1077
+    },
+    {
+      "epoch": 0.2237442922374429,
+      "grad_norm": 2.2965043054358794,
+      "learning_rate": 3.5733317154950384e-07,
+      "loss": 1.4871,
+      "step": 1078
+    },
+    {
+      "epoch": 0.22395184723951847,
+      "grad_norm": 0.7449502536445629,
+      "learning_rate": 3.5725726958019984e-07,
+      "loss": 1.606,
+      "step": 1079
+    },
+    {
+      "epoch": 0.22415940224159403,
+      "grad_norm": 1.0794989441493348,
+      "learning_rate": 3.5718130925274876e-07,
+      "loss": 1.5439,
+      "step": 1080
+    },
+    {
+      "epoch": 0.22436695724366956,
+      "grad_norm": 0.7358056517263604,
+      "learning_rate": 3.5710529059944703e-07,
+      "loss": 1.5326,
+      "step": 1081
+    },
+    {
+      "epoch": 0.22457451224574512,
+      "grad_norm": 3.387124782423694,
+      "learning_rate": 3.570292136526156e-07,
+      "loss": 1.4424,
+      "step": 1082
+    },
+    {
+      "epoch": 0.22478206724782068,
+      "grad_norm": 0.981958200213979,
+      "learning_rate": 3.5695307844460047e-07,
+      "loss": 1.5692,
+      "step": 1083
+    },
+    {
+      "epoch": 0.22498962224989622,
+      "grad_norm": 0.7052055577090767,
+      "learning_rate": 3.568768850077723e-07,
+      "loss": 1.5457,
+      "step": 1084
+    },
+    {
+      "epoch": 0.22519717725197178,
+      "grad_norm": 0.832611158573069,
+      "learning_rate": 3.5680063337452656e-07,
+      "loss": 1.5511,
+      "step": 1085
+    },
+    {
+      "epoch": 0.22540473225404734,
+      "grad_norm": 0.7052347028988847,
+      "learning_rate": 3.5672432357728323e-07,
+      "loss": 1.5436,
+      "step": 1086
+    },
+    {
+      "epoch": 0.22561228725612287,
+      "grad_norm": 0.6933559865988063,
+      "learning_rate": 3.5664795564848747e-07,
+      "loss": 1.5242,
+      "step": 1087
+    },
+    {
+      "epoch": 0.22581984225819843,
+      "grad_norm": 1.2194582774699534,
+      "learning_rate": 3.565715296206086e-07,
+      "loss": 1.4851,
+      "step": 1088
+    },
+    {
+      "epoch": 0.22602739726027396,
+      "grad_norm": 0.6719228601750747,
+      "learning_rate": 3.5649504552614126e-07,
+      "loss": 1.5395,
+      "step": 1089
+    },
+    {
+      "epoch": 0.22623495226234952,
+      "grad_norm": 0.7326097834920743,
+      "learning_rate": 3.5641850339760423e-07,
+      "loss": 1.5297,
+      "step": 1090
+    },
+    {
+      "epoch": 0.22644250726442508,
+      "grad_norm": 0.7912133563872042,
+      "learning_rate": 3.563419032675413e-07,
+      "loss": 1.4655,
+      "step": 1091
+    },
+    {
+      "epoch": 0.2266500622665006,
+      "grad_norm": 1.0059830551169904,
+      "learning_rate": 3.562652451685207e-07,
+      "loss": 1.5194,
+      "step": 1092
+    },
+    {
+      "epoch": 0.22685761726857617,
+      "grad_norm": 2.3626578532189755,
+      "learning_rate": 3.5618852913313555e-07,
+      "loss": 1.5309,
+      "step": 1093
+    },
+    {
+      "epoch": 0.22706517227065173,
+      "grad_norm": 0.7136902498245019,
+      "learning_rate": 3.5611175519400336e-07,
+      "loss": 1.5376,
+      "step": 1094
+    },
+    {
+      "epoch": 0.22727272727272727,
+      "grad_norm": 1.0857788038310587,
+      "learning_rate": 3.560349233837665e-07,
+      "loss": 1.5435,
+      "step": 1095
+    },
+    {
+      "epoch": 0.22748028227480283,
+      "grad_norm": 1.0249268271304817,
+      "learning_rate": 3.559580337350917e-07,
+      "loss": 1.5516,
+      "step": 1096
+    },
+    {
+      "epoch": 0.22768783727687839,
+      "grad_norm": 0.74967194236441,
+      "learning_rate": 3.558810862806704e-07,
+      "loss": 1.5891,
+      "step": 1097
+    },
+    {
+      "epoch": 0.22789539227895392,
+      "grad_norm": 0.6840225869912029,
+      "learning_rate": 3.558040810532187e-07,
+      "loss": 1.5944,
+      "step": 1098
+    },
+    {
+      "epoch": 0.22810294728102948,
+      "grad_norm": 1.186527241920316,
+      "learning_rate": 3.557270180854771e-07,
+      "loss": 1.4723,
+      "step": 1099
+    },
+    {
+      "epoch": 0.228310502283105,
+      "grad_norm": 3.081113774950376,
+      "learning_rate": 3.556498974102108e-07,
+      "loss": 1.5258,
+      "step": 1100
+    },
+    {
+      "epoch": 0.22851805728518057,
+      "grad_norm": 0.6837190802544153,
+      "learning_rate": 3.555727190602094e-07,
+      "loss": 1.5978,
+      "step": 1101
+    },
+    {
+      "epoch": 0.22872561228725613,
+      "grad_norm": 0.7006446109094233,
+      "learning_rate": 3.5549548306828705e-07,
+      "loss": 1.5107,
+      "step": 1102
+    },
+    {
+      "epoch": 0.22893316728933166,
+      "grad_norm": 1.142427625081276,
+      "learning_rate": 3.5541818946728254e-07,
+      "loss": 1.4936,
+      "step": 1103
+    },
+    {
+      "epoch": 0.22914072229140722,
+      "grad_norm": 0.9869752524805545,
+      "learning_rate": 3.553408382900591e-07,
+      "loss": 1.4953,
+      "step": 1104
+    },
+    {
+      "epoch": 0.22934827729348278,
+      "grad_norm": 0.8679464248422917,
+      "learning_rate": 3.552634295695042e-07,
+      "loss": 1.5196,
+      "step": 1105
+    },
+    {
+      "epoch": 0.22955583229555832,
+      "grad_norm": 0.8451704782978928,
+      "learning_rate": 3.551859633385301e-07,
+      "loss": 1.4815,
+      "step": 1106
+    },
+    {
+      "epoch": 0.22976338729763388,
+      "grad_norm": 0.819531145071641,
+      "learning_rate": 3.551084396300734e-07,
+      "loss": 1.4952,
+      "step": 1107
+    },
+    {
+      "epoch": 0.22997094229970944,
+      "grad_norm": 0.6780431844481092,
+      "learning_rate": 3.5503085847709513e-07,
+      "loss": 1.5472,
+      "step": 1108
+    },
+    {
+      "epoch": 0.23017849730178497,
+      "grad_norm": 0.8607147218439937,
+      "learning_rate": 3.5495321991258066e-07,
+      "loss": 1.6053,
+      "step": 1109
+    },
+    {
+      "epoch": 0.23038605230386053,
+      "grad_norm": 0.920641234673832,
+      "learning_rate": 3.548755239695399e-07,
+      "loss": 1.5012,
+      "step": 1110
+    },
+    {
+      "epoch": 0.23059360730593606,
+      "grad_norm": 1.1636507847521098,
+      "learning_rate": 3.547977706810071e-07,
+      "loss": 1.535,
+      "step": 1111
+    },
+    {
+      "epoch": 0.23080116230801162,
+      "grad_norm": 0.7499729548501789,
+      "learning_rate": 3.5471996008004086e-07,
+      "loss": 1.5131,
+      "step": 1112
+    },
+    {
+      "epoch": 0.23100871731008718,
+      "grad_norm": 0.8270941212869587,
+      "learning_rate": 3.5464209219972425e-07,
+      "loss": 1.5212,
+      "step": 1113
+    },
+    {
+      "epoch": 0.2312162723121627,
+      "grad_norm": 0.7444398963804137,
+      "learning_rate": 3.545641670731645e-07,
+      "loss": 1.5278,
+      "step": 1114
+    },
+    {
+      "epoch": 0.23142382731423827,
+      "grad_norm": 0.7729624006333331,
+      "learning_rate": 3.5448618473349344e-07,
+      "loss": 1.5791,
+      "step": 1115
+    },
+    {
+      "epoch": 0.23163138231631383,
+      "grad_norm": 0.7844818799700299,
+      "learning_rate": 3.5440814521386703e-07,
+      "loss": 1.5555,
+      "step": 1116
+    },
+    {
+      "epoch": 0.23183893731838937,
+      "grad_norm": 0.889674195599544,
+      "learning_rate": 3.543300485474656e-07,
+      "loss": 1.6032,
+      "step": 1117
+    },
+    {
+      "epoch": 0.23204649232046493,
+      "grad_norm": 0.9318115382968022,
+      "learning_rate": 3.542518947674938e-07,
+      "loss": 1.4618,
+      "step": 1118
+    },
+    {
+      "epoch": 0.23225404732254049,
+      "grad_norm": 0.788639798112483,
+      "learning_rate": 3.541736839071805e-07,
+      "loss": 1.5614,
+      "step": 1119
+    },
+    {
+      "epoch": 0.23246160232461602,
+      "grad_norm": 1.047236924867649,
+      "learning_rate": 3.5409541599977895e-07,
+      "loss": 1.5627,
+      "step": 1120
+    },
+    {
+      "epoch": 0.23266915732669158,
+      "grad_norm": 0.7247750234226287,
+      "learning_rate": 3.540170910785665e-07,
+      "loss": 1.5887,
+      "step": 1121
+    },
+    {
+      "epoch": 0.2328767123287671,
+      "grad_norm": 0.7439214209403497,
+      "learning_rate": 3.539387091768449e-07,
+      "loss": 1.4641,
+      "step": 1122
+    },
+    {
+      "epoch": 0.23308426733084267,
+      "grad_norm": 5.678155893947435,
+      "learning_rate": 3.538602703279401e-07,
+      "loss": 1.5128,
+      "step": 1123
+    },
+    {
+      "epoch": 0.23329182233291823,
+      "grad_norm": 1.1120363709533163,
+      "learning_rate": 3.53781774565202e-07,
+      "loss": 1.5181,
+      "step": 1124
+    },
+    {
+      "epoch": 0.23349937733499376,
+      "grad_norm": 0.7045229084175086,
+      "learning_rate": 3.5370322192200514e-07,
+      "loss": 1.5184,
+      "step": 1125
+    },
+    {
+      "epoch": 0.23370693233706932,
+      "grad_norm": 0.7956027699305732,
+      "learning_rate": 3.5362461243174795e-07,
+      "loss": 1.5407,
+      "step": 1126
+    },
+    {
+      "epoch": 0.23391448733914488,
+      "grad_norm": 0.7025158285313601,
+      "learning_rate": 3.5354594612785306e-07,
+      "loss": 1.4801,
+      "step": 1127
+    },
+    {
+      "epoch": 0.23412204234122042,
+      "grad_norm": 0.7680847312865177,
+      "learning_rate": 3.5346722304376734e-07,
+      "loss": 1.5415,
+      "step": 1128
+    },
+    {
+      "epoch": 0.23432959734329598,
+      "grad_norm": 0.9389118848809208,
+      "learning_rate": 3.5338844321296174e-07,
+      "loss": 1.5022,
+      "step": 1129
+    },
+    {
+      "epoch": 0.23453715234537154,
+      "grad_norm": 1.7567025870509037,
+      "learning_rate": 3.533096066689313e-07,
+      "loss": 1.5667,
+      "step": 1130
+    },
+    {
+      "epoch": 0.23474470734744707,
+      "grad_norm": 0.8307813913763255,
+      "learning_rate": 3.5323071344519526e-07,
+      "loss": 1.5235,
+      "step": 1131
+    },
+    {
+      "epoch": 0.23495226234952263,
+      "grad_norm": 0.6933498374453867,
+      "learning_rate": 3.5315176357529705e-07,
+      "loss": 1.6025,
+      "step": 1132
+    },
+    {
+      "epoch": 0.23515981735159816,
+      "grad_norm": 0.8281707935692568,
+      "learning_rate": 3.5307275709280386e-07,
+      "loss": 1.5123,
+      "step": 1133
+    },
+    {
+      "epoch": 0.23536737235367372,
+      "grad_norm": 0.7713002853216047,
+      "learning_rate": 3.529936940313073e-07,
+      "loss": 1.4931,
+      "step": 1134
+    },
+    {
+      "epoch": 0.23557492735574928,
+      "grad_norm": 0.8616488564890379,
+      "learning_rate": 3.529145744244227e-07,
+      "loss": 1.5537,
+      "step": 1135
+    },
+    {
+      "epoch": 0.2357824823578248,
+      "grad_norm": 2.080763493759813,
+      "learning_rate": 3.5283539830578986e-07,
+      "loss": 1.4982,
+      "step": 1136
+    },
+    {
+      "epoch": 0.23599003735990037,
+      "grad_norm": 1.9578650706350422,
+      "learning_rate": 3.527561657090722e-07,
+      "loss": 1.5173,
+      "step": 1137
+    },
+    {
+      "epoch": 0.23619759236197593,
+      "grad_norm": 1.834435107650694,
+      "learning_rate": 3.526768766679573e-07,
+      "loss": 1.5602,
+      "step": 1138
+    },
+    {
+      "epoch": 0.23640514736405147,
+      "grad_norm": 0.934201369962471,
+      "learning_rate": 3.525975312161569e-07,
+      "loss": 1.4968,
+      "step": 1139
+    },
+    {
+      "epoch": 0.23661270236612703,
+      "grad_norm": 1.1060332806062492,
+      "learning_rate": 3.525181293874064e-07,
+      "loss": 1.5265,
+      "step": 1140
+    },
+    {
+      "epoch": 0.23682025736820259,
+      "grad_norm": 0.7449954987756889,
+      "learning_rate": 3.5243867121546554e-07,
+      "loss": 1.5451,
+      "step": 1141
+    },
+    {
+      "epoch": 0.23702781237027812,
+      "grad_norm": 0.6175282902791747,
+      "learning_rate": 3.5235915673411765e-07,
+      "loss": 1.5583,
+      "step": 1142
+    },
+    {
+      "epoch": 0.23723536737235368,
+      "grad_norm": 0.7755460477275391,
+      "learning_rate": 3.522795859771703e-07,
+      "loss": 1.5564,
+      "step": 1143
+    },
+    {
+      "epoch": 0.2374429223744292,
+      "grad_norm": 0.6660308330359466,
+      "learning_rate": 3.521999589784548e-07,
+      "loss": 1.534,
+      "step": 1144
+    },
+    {
+      "epoch": 0.23765047737650477,
+      "grad_norm": 0.8255607264198397,
+      "learning_rate": 3.521202757718264e-07,
+      "loss": 1.4659,
+      "step": 1145
+    },
+    {
+      "epoch": 0.23785803237858033,
+      "grad_norm": 0.8266755481388101,
+      "learning_rate": 3.520405363911644e-07,
+      "loss": 1.4976,
+      "step": 1146
+    },
+    {
+      "epoch": 0.23806558738065586,
+      "grad_norm": 0.735387448014218,
+      "learning_rate": 3.5196074087037185e-07,
+      "loss": 1.4569,
+      "step": 1147
+    },
+    {
+      "epoch": 0.23827314238273142,
+      "grad_norm": 0.8202692357684406,
+      "learning_rate": 3.5188088924337554e-07,
+      "loss": 1.5411,
+      "step": 1148
+    },
+    {
+      "epoch": 0.23848069738480698,
+      "grad_norm": 1.1035662026438207,
+      "learning_rate": 3.518009815441264e-07,
+      "loss": 1.5346,
+      "step": 1149
+    },
+    {
+      "epoch": 0.23868825238688252,
+      "grad_norm": 0.6996633125514471,
+      "learning_rate": 3.51721017806599e-07,
+      "loss": 1.536,
+      "step": 1150
+    },
+    {
+      "epoch": 0.23889580738895808,
+      "grad_norm": 0.8105114216585455,
+      "learning_rate": 3.516409980647919e-07,
+      "loss": 1.5087,
+      "step": 1151
+    },
+    {
+      "epoch": 0.23910336239103364,
+      "grad_norm": 0.9357776618975657,
+      "learning_rate": 3.515609223527272e-07,
+      "loss": 1.512,
+      "step": 1152
+    },
+    {
+      "epoch": 0.23931091739310917,
+      "grad_norm": 0.7231486849647978,
+      "learning_rate": 3.514807907044511e-07,
+      "loss": 1.496,
+      "step": 1153
+    },
+    {
+      "epoch": 0.23951847239518473,
+      "grad_norm": 0.9511119162522497,
+      "learning_rate": 3.514006031540334e-07,
+      "loss": 1.5639,
+      "step": 1154
+    },
+    {
+      "epoch": 0.23972602739726026,
+      "grad_norm": 1.1355007292956965,
+      "learning_rate": 3.5132035973556773e-07,
+      "loss": 1.5735,
+      "step": 1155
+    },
+    {
+      "epoch": 0.23993358239933582,
+      "grad_norm": 0.7332457034201849,
+      "learning_rate": 3.512400604831715e-07,
+      "loss": 1.4312,
+      "step": 1156
+    },
+    {
+      "epoch": 0.24014113740141138,
+      "grad_norm": 0.7122281484165404,
+      "learning_rate": 3.511597054309857e-07,
+      "loss": 1.5995,
+      "step": 1157
+    },
+    {
+      "epoch": 0.2403486924034869,
+      "grad_norm": 0.8250980426862565,
+      "learning_rate": 3.510792946131753e-07,
+      "loss": 1.5275,
+      "step": 1158
+    },
+    {
+      "epoch": 0.24055624740556247,
+      "grad_norm": 0.9133774491873744,
+      "learning_rate": 3.5099882806392874e-07,
+      "loss": 1.5615,
+      "step": 1159
+    },
+    {
+      "epoch": 0.24076380240763803,
+      "grad_norm": 1.0882884865806335,
+      "learning_rate": 3.5091830581745833e-07,
+      "loss": 1.5537,
+      "step": 1160
+    },
+    {
+      "epoch": 0.24097135740971357,
+      "grad_norm": 0.8910267983356208,
+      "learning_rate": 3.50837727908e-07,
+      "loss": 1.482,
+      "step": 1161
+    },
+    {
+      "epoch": 0.24117891241178913,
+      "grad_norm": 0.8907082108587128,
+      "learning_rate": 3.5075709436981325e-07,
+      "loss": 1.5999,
+      "step": 1162
+    },
+    {
+      "epoch": 0.24138646741386469,
+      "grad_norm": 0.6677218222628633,
+      "learning_rate": 3.5067640523718145e-07,
+      "loss": 1.53,
+      "step": 1163
+    },
+    {
+      "epoch": 0.24159402241594022,
+      "grad_norm": 1.792291287577258,
+      "learning_rate": 3.5059566054441143e-07,
+      "loss": 1.5525,
+      "step": 1164
+    },
+    {
+      "epoch": 0.24180157741801578,
+      "grad_norm": 1.8506946777712572,
+      "learning_rate": 3.5051486032583354e-07,
+      "loss": 1.5414,
+      "step": 1165
+    },
+    {
+      "epoch": 0.2420091324200913,
+      "grad_norm": 0.9293223406808399,
+      "learning_rate": 3.504340046158021e-07,
+      "loss": 1.5237,
+      "step": 1166
+    },
+    {
+      "epoch": 0.24221668742216687,
+      "grad_norm": 0.7879397529892875,
+      "learning_rate": 3.503530934486947e-07,
+      "loss": 1.493,
+      "step": 1167
+    },
+    {
+      "epoch": 0.24242424242424243,
+      "grad_norm": 0.8808044646451905,
+      "learning_rate": 3.502721268589126e-07,
+      "loss": 1.5988,
+      "step": 1168
+    },
+    {
+      "epoch": 0.24263179742631796,
+      "grad_norm": 0.8643682076342636,
+      "learning_rate": 3.501911048808808e-07,
+      "loss": 1.5408,
+      "step": 1169
+    },
+    {
+      "epoch": 0.24283935242839352,
+      "grad_norm": 0.7270991475676218,
+      "learning_rate": 3.5011002754904744e-07,
+      "loss": 1.5179,
+      "step": 1170
+    },
+    {
+      "epoch": 0.24304690743046908,
+      "grad_norm": 3.0668902138950065,
+      "learning_rate": 3.5002889489788465e-07,
+      "loss": 1.554,
+      "step": 1171
+    },
+    {
+      "epoch": 0.24325446243254462,
+      "grad_norm": 0.7705644809816546,
+      "learning_rate": 3.499477069618879e-07,
+      "loss": 1.5145,
+      "step": 1172
+    },
+    {
+      "epoch": 0.24346201743462018,
+      "grad_norm": 0.7605964639763112,
+      "learning_rate": 3.4986646377557595e-07,
+      "loss": 1.514,
+      "step": 1173
+    },
+    {
+      "epoch": 0.24366957243669574,
+      "grad_norm": 0.9229541121161166,
+      "learning_rate": 3.4978516537349144e-07,
+      "loss": 1.4982,
+      "step": 1174
+    },
+    {
+      "epoch": 0.24387712743877127,
+      "grad_norm": 0.7433458249266687,
+      "learning_rate": 3.497038117902002e-07,
+      "loss": 1.3846,
+      "step": 1175
+    },
+    {
+      "epoch": 0.24408468244084683,
+      "grad_norm": 2.502884423063433,
+      "learning_rate": 3.4962240306029155e-07,
+      "loss": 1.4937,
+      "step": 1176
+    },
+    {
+      "epoch": 0.24429223744292236,
+      "grad_norm": 0.650115175962014,
+      "learning_rate": 3.4954093921837844e-07,
+      "loss": 1.5192,
+      "step": 1177
+    },
+    {
+      "epoch": 0.24449979244499792,
+      "grad_norm": 0.8937346709079073,
+      "learning_rate": 3.494594202990971e-07,
+      "loss": 1.5943,
+      "step": 1178
+    },
+    {
+      "epoch": 0.24470734744707348,
+      "grad_norm": 0.7765759918713224,
+      "learning_rate": 3.493778463371073e-07,
+      "loss": 1.5569,
+      "step": 1179
+    },
+    {
+      "epoch": 0.244914902449149,
+      "grad_norm": 0.7078389298835821,
+      "learning_rate": 3.492962173670919e-07,
+      "loss": 1.531,
+      "step": 1180
+    },
+    {
+      "epoch": 0.24512245745122457,
+      "grad_norm": 0.9731015621045409,
+      "learning_rate": 3.4921453342375753e-07,
+      "loss": 1.5447,
+      "step": 1181
+    },
+    {
+      "epoch": 0.24533001245330013,
+      "grad_norm": 0.9026889670118812,
+      "learning_rate": 3.4913279454183393e-07,
+      "loss": 1.5436,
+      "step": 1182
+    },
+    {
+      "epoch": 0.24553756745537567,
+      "grad_norm": 2.118697915346356,
+      "learning_rate": 3.490510007560745e-07,
+      "loss": 1.5048,
+      "step": 1183
+    },
+    {
+      "epoch": 0.24574512245745123,
+      "grad_norm": 0.7595503866912212,
+      "learning_rate": 3.4896915210125556e-07,
+      "loss": 1.5406,
+      "step": 1184
+    },
+    {
+      "epoch": 0.2459526774595268,
+      "grad_norm": 0.7273656000166633,
+      "learning_rate": 3.4888724861217715e-07,
+      "loss": 1.5175,
+      "step": 1185
+    },
+    {
+      "epoch": 0.24616023246160232,
+      "grad_norm": 1.087952926147746,
+      "learning_rate": 3.488052903236624e-07,
+      "loss": 1.5207,
+      "step": 1186
+    },
+    {
+      "epoch": 0.24636778746367788,
+      "grad_norm": 0.7850053086727021,
+      "learning_rate": 3.4872327727055767e-07,
+      "loss": 1.4612,
+      "step": 1187
+    },
+    {
+      "epoch": 0.2465753424657534,
+      "grad_norm": 0.672850110925625,
+      "learning_rate": 3.48641209487733e-07,
+      "loss": 1.5236,
+      "step": 1188
+    },
+    {
+      "epoch": 0.24678289746782897,
+      "grad_norm": 0.946695807598679,
+      "learning_rate": 3.485590870100812e-07,
+      "loss": 1.51,
+      "step": 1189
+    },
+    {
+      "epoch": 0.24699045246990453,
+      "grad_norm": 0.8087733984327321,
+      "learning_rate": 3.484769098725186e-07,
+      "loss": 1.4873,
+      "step": 1190
+    },
+    {
+      "epoch": 0.24719800747198006,
+      "grad_norm": 0.6202165210617465,
+      "learning_rate": 3.483946781099849e-07,
+      "loss": 1.6002,
+      "step": 1191
+    },
+    {
+      "epoch": 0.24740556247405562,
+      "grad_norm": 1.0469577936140428,
+      "learning_rate": 3.483123917574427e-07,
+      "loss": 1.506,
+      "step": 1192
+    },
+    {
+      "epoch": 0.24761311747613118,
+      "grad_norm": 0.6722479637831869,
+      "learning_rate": 3.482300508498781e-07,
+      "loss": 1.4863,
+      "step": 1193
+    },
+    {
+      "epoch": 0.24782067247820672,
+      "grad_norm": 0.6866664848811579,
+      "learning_rate": 3.4814765542230016e-07,
+      "loss": 1.5029,
+      "step": 1194
+    },
+    {
+      "epoch": 0.24802822748028228,
+      "grad_norm": 0.6951110690195187,
+      "learning_rate": 3.480652055097412e-07,
+      "loss": 1.4922,
+      "step": 1195
+    },
+    {
+      "epoch": 0.24823578248235784,
+      "grad_norm": 0.6210389677477952,
+      "learning_rate": 3.479827011472568e-07,
+      "loss": 1.5763,
+      "step": 1196
+    },
+    {
+      "epoch": 0.24844333748443337,
+      "grad_norm": 0.9507026327261713,
+      "learning_rate": 3.479001423699257e-07,
+      "loss": 1.5166,
+      "step": 1197
+    },
+    {
+      "epoch": 0.24865089248650893,
+      "grad_norm": 0.736680360946047,
+      "learning_rate": 3.4781752921284957e-07,
+      "loss": 1.4805,
+      "step": 1198
+    },
+    {
+      "epoch": 0.24885844748858446,
+      "grad_norm": 0.8055130492879062,
+      "learning_rate": 3.4773486171115336e-07,
+      "loss": 1.5249,
+      "step": 1199
+    },
+    {
+      "epoch": 0.24906600249066002,
+      "grad_norm": 0.7479198464303374,
+      "learning_rate": 3.476521398999851e-07,
+      "loss": 1.5196,
+      "step": 1200
+    },
+    {
+      "epoch": 0.24927355749273558,
+      "grad_norm": 0.864718739179395,
+      "learning_rate": 3.4756936381451604e-07,
+      "loss": 1.5517,
+      "step": 1201
+    },
+    {
+      "epoch": 0.2494811124948111,
+      "grad_norm": 2.2051962954573923,
+      "learning_rate": 3.4748653348994013e-07,
+      "loss": 1.4927,
+      "step": 1202
+    },
+    {
+      "epoch": 0.24968866749688667,
+      "grad_norm": 0.7097734943702743,
+      "learning_rate": 3.474036489614748e-07,
+      "loss": 1.5037,
+      "step": 1203
+    },
+    {
+      "epoch": 0.24989622249896223,
+      "grad_norm": 0.8835450203673174,
+      "learning_rate": 3.473207102643603e-07,
+      "loss": 1.5148,
+      "step": 1204
+    },
+    {
+      "epoch": 0.2501037775010378,
+      "grad_norm": 0.663424009458325,
+      "learning_rate": 3.4723771743386e-07,
+      "loss": 1.5525,
+      "step": 1205
+    },
+    {
+      "epoch": 0.2503113325031133,
+      "grad_norm": 0.8740416597945138,
+      "learning_rate": 3.471546705052602e-07,
+      "loss": 1.54,
+      "step": 1206
+    },
+    {
+      "epoch": 0.25051888750518886,
+      "grad_norm": 0.9838597095129007,
+      "learning_rate": 3.470715695138703e-07,
+      "loss": 1.4978,
+      "step": 1207
+    },
+    {
+      "epoch": 0.25072644250726445,
+      "grad_norm": 1.8699868039163179,
+      "learning_rate": 3.4698841449502255e-07,
+      "loss": 1.4988,
+      "step": 1208
+    },
+    {
+      "epoch": 0.25093399750934,
+      "grad_norm": 1.0253374426664066,
+      "learning_rate": 3.4690520548407234e-07,
+      "loss": 1.598,
+      "step": 1209
+    },
+    {
+      "epoch": 0.2511415525114155,
+      "grad_norm": 0.7399656560368909,
+      "learning_rate": 3.4682194251639785e-07,
+      "loss": 1.5229,
+      "step": 1210
+    },
+    {
+      "epoch": 0.2513491075134911,
+      "grad_norm": 1.0208535629189506,
+      "learning_rate": 3.467386256274004e-07,
+      "loss": 1.5697,
+      "step": 1211
+    },
+    {
+      "epoch": 0.25155666251556663,
+      "grad_norm": 1.4083379372509373,
+      "learning_rate": 3.46655254852504e-07,
+      "loss": 1.5417,
+      "step": 1212
+    },
+    {
+      "epoch": 0.25176421751764216,
+      "grad_norm": 0.7794650017371283,
+      "learning_rate": 3.465718302271558e-07,
+      "loss": 1.4597,
+      "step": 1213
+    },
+    {
+      "epoch": 0.25197177251971775,
+      "grad_norm": 0.9491926350594242,
+      "learning_rate": 3.464883517868256e-07,
+      "loss": 1.4628,
+      "step": 1214
+    },
+    {
+      "epoch": 0.2521793275217933,
+      "grad_norm": 0.8708290005181476,
+      "learning_rate": 3.4640481956700633e-07,
+      "loss": 1.4727,
+      "step": 1215
+    },
+    {
+      "epoch": 0.2523868825238688,
+      "grad_norm": 1.2105880211786428,
+      "learning_rate": 3.463212336032137e-07,
+      "loss": 1.5313,
+      "step": 1216
+    },
+    {
+      "epoch": 0.25259443752594435,
+      "grad_norm": 1.629160809077407,
+      "learning_rate": 3.462375939309861e-07,
+      "loss": 1.5039,
+      "step": 1217
+    },
+    {
+      "epoch": 0.25280199252801994,
+      "grad_norm": 0.9428744088974746,
+      "learning_rate": 3.46153900585885e-07,
+      "loss": 1.5279,
+      "step": 1218
+    },
+    {
+      "epoch": 0.25300954753009547,
+      "grad_norm": 1.0797364122207653,
+      "learning_rate": 3.4607015360349456e-07,
+      "loss": 1.4873,
+      "step": 1219
+    },
+    {
+      "epoch": 0.253217102532171,
+      "grad_norm": 0.6416712690429094,
+      "learning_rate": 3.4598635301942177e-07,
+      "loss": 1.5224,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2534246575342466,
+      "grad_norm": 0.747285449443041,
+      "learning_rate": 3.4590249886929647e-07,
+      "loss": 1.4692,
+      "step": 1221
+    },
+    {
+      "epoch": 0.2536322125363221,
+      "grad_norm": 0.910949929056786,
+      "learning_rate": 3.4581859118877117e-07,
+      "loss": 1.5673,
+      "step": 1222
+    },
+    {
+      "epoch": 0.25383976753839765,
+      "grad_norm": 0.9346303546125491,
+      "learning_rate": 3.4573463001352116e-07,
+      "loss": 1.5621,
+      "step": 1223
+    },
+    {
+      "epoch": 0.25404732254047324,
+      "grad_norm": 0.6971197936963709,
+      "learning_rate": 3.456506153792445e-07,
+      "loss": 1.51,
+      "step": 1224
+    },
+    {
+      "epoch": 0.2542548775425488,
+      "grad_norm": 2.8675278436766005,
+      "learning_rate": 3.4556654732166204e-07,
+      "loss": 1.526,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2544624325446243,
+      "grad_norm": 1.4822535966762262,
+      "learning_rate": 3.454824258765173e-07,
+      "loss": 1.5517,
+      "step": 1226
+    },
+    {
+      "epoch": 0.2546699875466999,
+      "grad_norm": 1.1853114078047695,
+      "learning_rate": 3.4539825107957643e-07,
+      "loss": 1.5357,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2548775425487754,
+      "grad_norm": 1.5281746994092364,
+      "learning_rate": 3.4531402296662827e-07,
+      "loss": 1.5171,
+      "step": 1228
+    },
+    {
+      "epoch": 0.25508509755085096,
+      "grad_norm": 0.6757285200932763,
+      "learning_rate": 3.4522974157348455e-07,
+      "loss": 1.4501,
+      "step": 1229
+    },
+    {
+      "epoch": 0.25529265255292655,
+      "grad_norm": 1.689320557195999,
+      "learning_rate": 3.4514540693597935e-07,
+      "loss": 1.56,
+      "step": 1230
+    },
+    {
+      "epoch": 0.2555002075550021,
+      "grad_norm": 1.2034464716258129,
+      "learning_rate": 3.450610190899695e-07,
+      "loss": 1.5487,
+      "step": 1231
+    },
+    {
+      "epoch": 0.2557077625570776,
+      "grad_norm": 0.8955944160494794,
+      "learning_rate": 3.449765780713345e-07,
+      "loss": 1.6101,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2559153175591532,
+      "grad_norm": 1.116097814053556,
+      "learning_rate": 3.4489208391597645e-07,
+      "loss": 1.4694,
+      "step": 1233
+    },
+    {
+      "epoch": 0.25612287256122873,
+      "grad_norm": 0.7482192292543534,
+      "learning_rate": 3.4480753665982007e-07,
+      "loss": 1.5272,
+      "step": 1234
+    },
+    {
+      "epoch": 0.25633042756330426,
+      "grad_norm": 0.6543260224943683,
+      "learning_rate": 3.4472293633881253e-07,
+      "loss": 1.5059,
+      "step": 1235
+    },
+    {
+      "epoch": 0.25653798256537985,
+      "grad_norm": 1.1024405593086009,
+      "learning_rate": 3.4463828298892363e-07,
+      "loss": 1.521,
+      "step": 1236
+    },
+    {
+      "epoch": 0.2567455375674554,
+      "grad_norm": 0.9239975172305424,
+      "learning_rate": 3.445535766461458e-07,
+      "loss": 1.5815,
+      "step": 1237
+    },
+    {
+      "epoch": 0.2569530925695309,
+      "grad_norm": 0.7747494025182335,
+      "learning_rate": 3.4446881734649387e-07,
+      "loss": 1.5642,
+      "step": 1238
+    },
+    {
+      "epoch": 0.25716064757160645,
+      "grad_norm": 0.8008164601774657,
+      "learning_rate": 3.443840051260053e-07,
+      "loss": 1.5377,
+      "step": 1239
+    },
+    {
+      "epoch": 0.25736820257368204,
+      "grad_norm": 0.7783037284441887,
+      "learning_rate": 3.442991400207399e-07,
+      "loss": 1.5056,
+      "step": 1240
+    },
+    {
+      "epoch": 0.25757575757575757,
+      "grad_norm": 0.8229201099728607,
+      "learning_rate": 3.442142220667802e-07,
+      "loss": 1.5631,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2577833125778331,
+      "grad_norm": 0.9064519814581001,
+      "learning_rate": 3.4412925130023086e-07,
+      "loss": 1.5287,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2579908675799087,
+      "grad_norm": 0.8732323880198436,
+      "learning_rate": 3.440442277572194e-07,
+      "loss": 1.5394,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2581984225819842,
+      "grad_norm": 8.232261890447024,
+      "learning_rate": 3.439591514738954e-07,
+      "loss": 1.5204,
+      "step": 1244
+    },
+    {
+      "epoch": 0.25840597758405975,
+      "grad_norm": 0.8898526844814135,
+      "learning_rate": 3.438740224864312e-07,
+      "loss": 1.5105,
+      "step": 1245
+    },
+    {
+      "epoch": 0.25861353258613534,
+      "grad_norm": 0.9339172497931623,
+      "learning_rate": 3.437888408310213e-07,
+      "loss": 1.514,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2588210875882109,
+      "grad_norm": 0.9333694988769258,
+      "learning_rate": 3.437036065438827e-07,
+      "loss": 1.6093,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2590286425902864,
+      "grad_norm": 0.6794141893096668,
+      "learning_rate": 3.4361831966125474e-07,
+      "loss": 1.446,
+      "step": 1248
+    },
+    {
+      "epoch": 0.259236197592362,
+      "grad_norm": 1.0455423324954585,
+      "learning_rate": 3.4353298021939907e-07,
+      "loss": 1.5688,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2594437525944375,
+      "grad_norm": 1.017536821245971,
+      "learning_rate": 3.434475882545999e-07,
+      "loss": 1.5796,
+      "step": 1250
+    },
+    {
+      "epoch": 0.25965130759651306,
+      "grad_norm": 1.1193592725916437,
+      "learning_rate": 3.433621438031635e-07,
+      "loss": 1.5464,
+      "step": 1251
+    },
+    {
+      "epoch": 0.25985886259858865,
+      "grad_norm": 1.766508056696705,
+      "learning_rate": 3.4327664690141865e-07,
+      "loss": 1.6188,
+      "step": 1252
+    },
+    {
+      "epoch": 0.2600664176006642,
+      "grad_norm": 0.7096153328756882,
+      "learning_rate": 3.4319109758571635e-07,
+      "loss": 1.5531,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2602739726027397,
+      "grad_norm": 0.7521727115271983,
+      "learning_rate": 3.431054958924299e-07,
+      "loss": 1.5499,
+      "step": 1254
+    },
+    {
+      "epoch": 0.2604815276048153,
+      "grad_norm": 1.0492497629841584,
+      "learning_rate": 3.4301984185795487e-07,
+      "loss": 1.4762,
+      "step": 1255
+    },
+    {
+      "epoch": 0.26068908260689083,
+      "grad_norm": 0.8157158938997188,
+      "learning_rate": 3.429341355187091e-07,
+      "loss": 1.5297,
+      "step": 1256
+    },
+    {
+      "epoch": 0.26089663760896636,
+      "grad_norm": 0.7112893149121823,
+      "learning_rate": 3.4284837691113255e-07,
+      "loss": 1.538,
+      "step": 1257
+    },
+    {
+      "epoch": 0.26110419261104195,
+      "grad_norm": 1.456030997449654,
+      "learning_rate": 3.427625660716876e-07,
+      "loss": 1.5762,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2613117476131175,
+      "grad_norm": 0.7537865709228221,
+      "learning_rate": 3.426767030368587e-07,
+      "loss": 1.5248,
+      "step": 1259
+    },
+    {
+      "epoch": 0.261519302615193,
+      "grad_norm": 0.9498334019222445,
+      "learning_rate": 3.425907878431526e-07,
+      "loss": 1.4351,
+      "step": 1260
+    },
+    {
+      "epoch": 0.26172685761726855,
+      "grad_norm": 0.6692248470470078,
+      "learning_rate": 3.425048205270981e-07,
+      "loss": 1.4829,
+      "step": 1261
+    },
+    {
+      "epoch": 0.26193441261934414,
+      "grad_norm": 1.0225832291961392,
+      "learning_rate": 3.424188011252462e-07,
+      "loss": 1.5155,
+      "step": 1262
+    },
+    {
+      "epoch": 0.26214196762141967,
+      "grad_norm": 1.3243499833191412,
+      "learning_rate": 3.4233272967417006e-07,
+      "loss": 1.561,
+      "step": 1263
+    },
+    {
+      "epoch": 0.2623495226234952,
+      "grad_norm": 0.6854015525731081,
+      "learning_rate": 3.422466062104651e-07,
+      "loss": 1.5237,
+      "step": 1264
+    },
+    {
+      "epoch": 0.2625570776255708,
+      "grad_norm": 0.658758538643818,
+      "learning_rate": 3.421604307707486e-07,
+      "loss": 1.5389,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2627646326276463,
+      "grad_norm": 0.6743833453979914,
+      "learning_rate": 3.420742033916601e-07,
+      "loss": 1.5162,
+      "step": 1266
+    },
+    {
+      "epoch": 0.26297218762972185,
+      "grad_norm": 0.94853333417178,
+      "learning_rate": 3.419879241098612e-07,
+      "loss": 1.5222,
+      "step": 1267
+    },
+    {
+      "epoch": 0.26317974263179744,
+      "grad_norm": 0.7766625914771926,
+      "learning_rate": 3.419015929620356e-07,
+      "loss": 1.4911,
+      "step": 1268
+    },
+    {
+      "epoch": 0.263387297633873,
+      "grad_norm": 0.6997670424334552,
+      "learning_rate": 3.4181520998488895e-07,
+      "loss": 1.56,
+      "step": 1269
+    },
+    {
+      "epoch": 0.2635948526359485,
+      "grad_norm": 0.77325213552909,
+      "learning_rate": 3.4172877521514905e-07,
+      "loss": 1.5923,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2638024076380241,
+      "grad_norm": 1.539055402435133,
+      "learning_rate": 3.4164228868956563e-07,
+      "loss": 1.574,
+      "step": 1271
+    },
+    {
+      "epoch": 0.2640099626400996,
+      "grad_norm": 1.2536444220547092,
+      "learning_rate": 3.415557504449105e-07,
+      "loss": 1.5091,
+      "step": 1272
+    },
+    {
+      "epoch": 0.26421751764217516,
+      "grad_norm": 0.7575551879670133,
+      "learning_rate": 3.4146916051797735e-07,
+      "loss": 1.5458,
+      "step": 1273
+    },
+    {
+      "epoch": 0.26442507264425075,
+      "grad_norm": 0.807457667790524,
+      "learning_rate": 3.41382518945582e-07,
+      "loss": 1.5439,
+      "step": 1274
+    },
+    {
+      "epoch": 0.2646326276463263,
+      "grad_norm": 0.7248201442239209,
+      "learning_rate": 3.4129582576456213e-07,
+      "loss": 1.5607,
+      "step": 1275
+    },
+    {
+      "epoch": 0.2648401826484018,
+      "grad_norm": 0.7703701356946464,
+      "learning_rate": 3.412090810117774e-07,
+      "loss": 1.5348,
+      "step": 1276
+    },
+    {
+      "epoch": 0.2650477376504774,
+      "grad_norm": 0.6790812074627258,
+      "learning_rate": 3.411222847241092e-07,
+      "loss": 1.605,
+      "step": 1277
+    },
+    {
+      "epoch": 0.26525529265255293,
+      "grad_norm": 0.8150481788979888,
+      "learning_rate": 3.4103543693846126e-07,
+      "loss": 1.5238,
+      "step": 1278
+    },
+    {
+      "epoch": 0.26546284765462846,
+      "grad_norm": 0.9352100500133862,
+      "learning_rate": 3.409485376917589e-07,
+      "loss": 1.5305,
+      "step": 1279
+    },
+    {
+      "epoch": 0.26567040265670405,
+      "grad_norm": 0.7780215728437782,
+      "learning_rate": 3.408615870209492e-07,
+      "loss": 1.5196,
+      "step": 1280
+    },
+    {
+      "epoch": 0.2658779576587796,
+      "grad_norm": 0.6253958895837718,
+      "learning_rate": 3.4077458496300145e-07,
+      "loss": 1.5393,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2660855126608551,
+      "grad_norm": 0.9639590714174837,
+      "learning_rate": 3.406875315549066e-07,
+      "loss": 1.5241,
+      "step": 1282
+    },
+    {
+      "epoch": 0.26629306766293065,
+      "grad_norm": 0.6804335985210204,
+      "learning_rate": 3.406004268336773e-07,
+      "loss": 1.5337,
+      "step": 1283
+    },
+    {
+      "epoch": 0.26650062266500624,
+      "grad_norm": 0.9608046822867599,
+      "learning_rate": 3.405132708363483e-07,
+      "loss": 1.6063,
+      "step": 1284
+    },
+    {
+      "epoch": 0.26670817766708177,
+      "grad_norm": 0.8008725822506221,
+      "learning_rate": 3.40426063599976e-07,
+      "loss": 1.5602,
+      "step": 1285
+    },
+    {
+      "epoch": 0.2669157326691573,
+      "grad_norm": 0.9153818179597167,
+      "learning_rate": 3.4033880516163855e-07,
+      "loss": 1.5005,
+      "step": 1286
+    },
+    {
+      "epoch": 0.2671232876712329,
+      "grad_norm": 0.7747693875887444,
+      "learning_rate": 3.40251495558436e-07,
+      "loss": 1.5147,
+      "step": 1287
+    },
+    {
+      "epoch": 0.2673308426733084,
+      "grad_norm": 0.8738491672385178,
+      "learning_rate": 3.4016413482749006e-07,
+      "loss": 1.6194,
+      "step": 1288
+    },
+    {
+      "epoch": 0.26753839767538395,
+      "grad_norm": 7.597456785044188,
+      "learning_rate": 3.4007672300594414e-07,
+      "loss": 1.4556,
+      "step": 1289
+    },
+    {
+      "epoch": 0.26774595267745954,
+      "grad_norm": 0.6610778339741703,
+      "learning_rate": 3.3998926013096345e-07,
+      "loss": 1.5336,
+      "step": 1290
+    },
+    {
+      "epoch": 0.2679535076795351,
+      "grad_norm": 0.7975392858296627,
+      "learning_rate": 3.399017462397349e-07,
+      "loss": 1.6103,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2681610626816106,
+      "grad_norm": 0.870348270223364,
+      "learning_rate": 3.39814181369467e-07,
+      "loss": 1.6068,
+      "step": 1292
+    },
+    {
+      "epoch": 0.2683686176836862,
+      "grad_norm": 0.9142153106628116,
+      "learning_rate": 3.397265655573901e-07,
+      "loss": 1.5689,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2685761726857617,
+      "grad_norm": 0.831001094053461,
+      "learning_rate": 3.3963889884075613e-07,
+      "loss": 1.5699,
+      "step": 1294
+    },
+    {
+      "epoch": 0.26878372768783726,
+      "grad_norm": 1.0630847627657554,
+      "learning_rate": 3.395511812568386e-07,
+      "loss": 1.4947,
+      "step": 1295
+    },
+    {
+      "epoch": 0.26899128268991285,
+      "grad_norm": 1.0434866633138011,
+      "learning_rate": 3.394634128429326e-07,
+      "loss": 1.5765,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2691988376919884,
+      "grad_norm": 1.4759916928688286,
+      "learning_rate": 3.3937559363635517e-07,
+      "loss": 1.5064,
+      "step": 1297
+    },
+    {
+      "epoch": 0.2694063926940639,
+      "grad_norm": 0.8175486075063759,
+      "learning_rate": 3.392877236744445e-07,
+      "loss": 1.551,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2696139476961395,
+      "grad_norm": 1.2467956338180735,
+      "learning_rate": 3.391998029945606e-07,
+      "loss": 1.5079,
+      "step": 1299
+    },
+    {
+      "epoch": 0.26982150269821503,
+      "grad_norm": 0.7962141657570511,
+      "learning_rate": 3.391118316340851e-07,
+      "loss": 1.5688,
+      "step": 1300
+    },
+    {
+      "epoch": 0.27002905770029056,
+      "grad_norm": 0.8663931529115972,
+      "learning_rate": 3.3902380963042103e-07,
+      "loss": 1.4879,
+      "step": 1301
+    },
+    {
+      "epoch": 0.27023661270236615,
+      "grad_norm": 1.5227446569908352,
+      "learning_rate": 3.3893573702099295e-07,
+      "loss": 1.5449,
+      "step": 1302
+    },
+    {
+      "epoch": 0.2704441677044417,
+      "grad_norm": 1.5952769930727484,
+      "learning_rate": 3.388476138432471e-07,
+      "loss": 1.5795,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2706517227065172,
+      "grad_norm": 0.9920665270345874,
+      "learning_rate": 3.3875944013465106e-07,
+      "loss": 1.6059,
+      "step": 1304
+    },
+    {
+      "epoch": 0.27085927770859275,
+      "grad_norm": 0.7767794174497293,
+      "learning_rate": 3.38671215932694e-07,
+      "loss": 1.4674,
+      "step": 1305
+    },
+    {
+      "epoch": 0.27106683271066834,
+      "grad_norm": 0.7223885972104012,
+      "learning_rate": 3.3858294127488636e-07,
+      "loss": 1.5108,
+      "step": 1306
+    },
+    {
+      "epoch": 0.27127438771274387,
+      "grad_norm": 0.7703347315478367,
+      "learning_rate": 3.3849461619876033e-07,
+      "loss": 1.5391,
+      "step": 1307
+    },
+    {
+      "epoch": 0.2714819427148194,
+      "grad_norm": 1.083558951353024,
+      "learning_rate": 3.3840624074186946e-07,
+      "loss": 1.5693,
+      "step": 1308
+    },
+    {
+      "epoch": 0.271689497716895,
+      "grad_norm": 0.7960948613082768,
+      "learning_rate": 3.383178149417884e-07,
+      "loss": 1.6374,
+      "step": 1309
+    },
+    {
+      "epoch": 0.2718970527189705,
+      "grad_norm": 0.8181028809341938,
+      "learning_rate": 3.382293388361136e-07,
+      "loss": 1.5521,
+      "step": 1310
+    },
+    {
+      "epoch": 0.27210460772104605,
+      "grad_norm": 0.822942233450186,
+      "learning_rate": 3.381408124624628e-07,
+      "loss": 1.5792,
+      "step": 1311
+    },
+    {
+      "epoch": 0.27231216272312164,
+      "grad_norm": 1.0283885868319704,
+      "learning_rate": 3.3805223585847493e-07,
+      "loss": 1.5356,
+      "step": 1312
+    },
+    {
+      "epoch": 0.2725197177251972,
+      "grad_norm": 0.9953628936821117,
+      "learning_rate": 3.3796360906181054e-07,
+      "loss": 1.525,
+      "step": 1313
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 1.6313072242785849,
+      "learning_rate": 3.3787493211015133e-07,
+      "loss": 1.5858,
+      "step": 1314
+    },
+    {
+      "epoch": 0.2729348277293483,
+      "grad_norm": 0.7614107450589407,
+      "learning_rate": 3.377862050412003e-07,
+      "loss": 1.4772,
+      "step": 1315
+    },
+    {
+      "epoch": 0.2731423827314238,
+      "grad_norm": 0.7517826941209897,
+      "learning_rate": 3.376974278926821e-07,
+      "loss": 1.5588,
+      "step": 1316
+    },
+    {
+      "epoch": 0.27334993773349936,
+      "grad_norm": 0.9205421541542501,
+      "learning_rate": 3.376086007023421e-07,
+      "loss": 1.5645,
+      "step": 1317
+    },
+    {
+      "epoch": 0.27355749273557495,
+      "grad_norm": 0.644662343214532,
+      "learning_rate": 3.375197235079474e-07,
+      "loss": 1.5083,
+      "step": 1318
+    },
+    {
+      "epoch": 0.2737650477376505,
+      "grad_norm": 0.6802868003081509,
+      "learning_rate": 3.3743079634728637e-07,
+      "loss": 1.6246,
+      "step": 1319
+    },
+    {
+      "epoch": 0.273972602739726,
+      "grad_norm": 0.768816148544853,
+      "learning_rate": 3.3734181925816824e-07,
+      "loss": 1.5766,
+      "step": 1320
+    },
+    {
+      "epoch": 0.2741801577418016,
+      "grad_norm": 0.7713671834891636,
+      "learning_rate": 3.3725279227842385e-07,
+      "loss": 1.5415,
+      "step": 1321
+    },
+    {
+      "epoch": 0.27438771274387713,
+      "grad_norm": 0.6992736466420109,
+      "learning_rate": 3.371637154459051e-07,
+      "loss": 1.5487,
+      "step": 1322
+    },
+    {
+      "epoch": 0.27459526774595266,
+      "grad_norm": 1.1645797501442725,
+      "learning_rate": 3.37074588798485e-07,
+      "loss": 1.505,
+      "step": 1323
+    },
+    {
+      "epoch": 0.27480282274802825,
+      "grad_norm": 0.8082129119597378,
+      "learning_rate": 3.369854123740579e-07,
+      "loss": 1.4956,
+      "step": 1324
+    },
+    {
+      "epoch": 0.2750103777501038,
+      "grad_norm": 1.04479833148827,
+      "learning_rate": 3.3689618621053924e-07,
+      "loss": 1.5379,
+      "step": 1325
+    },
+    {
+      "epoch": 0.2752179327521793,
+      "grad_norm": 0.7179713369018662,
+      "learning_rate": 3.3680691034586565e-07,
+      "loss": 1.4695,
+      "step": 1326
+    },
+    {
+      "epoch": 0.2754254877542549,
+      "grad_norm": 0.7686131203548663,
+      "learning_rate": 3.3671758481799484e-07,
+      "loss": 1.4641,
+      "step": 1327
+    },
+    {
+      "epoch": 0.27563304275633044,
+      "grad_norm": 4.48297428836442,
+      "learning_rate": 3.366282096649056e-07,
+      "loss": 1.5001,
+      "step": 1328
+    },
+    {
+      "epoch": 0.27584059775840597,
+      "grad_norm": 0.6928145750710054,
+      "learning_rate": 3.3653878492459793e-07,
+      "loss": 1.5642,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2760481527604815,
+      "grad_norm": 0.9673980706438595,
+      "learning_rate": 3.3644931063509287e-07,
+      "loss": 1.6038,
+      "step": 1330
+    },
+    {
+      "epoch": 0.2762557077625571,
+      "grad_norm": 0.7952910298893379,
+      "learning_rate": 3.363597868344324e-07,
+      "loss": 1.5937,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2764632627646326,
+      "grad_norm": 1.4477721379840571,
+      "learning_rate": 3.3627021356067987e-07,
+      "loss": 1.5336,
+      "step": 1332
+    },
+    {
+      "epoch": 0.27667081776670815,
+      "grad_norm": 0.62242184799139,
+      "learning_rate": 3.361805908519192e-07,
+      "loss": 1.516,
+      "step": 1333
+    },
+    {
+      "epoch": 0.27687837276878374,
+      "grad_norm": 0.733347199404752,
+      "learning_rate": 3.360909187462558e-07,
+      "loss": 1.466,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2770859277708593,
+      "grad_norm": 0.7924465185545144,
+      "learning_rate": 3.360011972818158e-07,
+      "loss": 1.6244,
+      "step": 1335
+    },
+    {
+      "epoch": 0.2772934827729348,
+      "grad_norm": 1.1516186087648017,
+      "learning_rate": 3.359114264967463e-07,
+      "loss": 1.5155,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2775010377750104,
+      "grad_norm": 0.8312352958790239,
+      "learning_rate": 3.358216064292156e-07,
+      "loss": 1.5276,
+      "step": 1337
+    },
+    {
+      "epoch": 0.2777085927770859,
+      "grad_norm": 0.7314857669887004,
+      "learning_rate": 3.357317371174127e-07,
+      "loss": 1.5777,
+      "step": 1338
+    },
+    {
+      "epoch": 0.27791614777916146,
+      "grad_norm": 1.3839777919660108,
+      "learning_rate": 3.356418185995477e-07,
+      "loss": 1.571,
+      "step": 1339
+    },
+    {
+      "epoch": 0.27812370278123705,
+      "grad_norm": 0.8427359483829014,
+      "learning_rate": 3.355518509138515e-07,
+      "loss": 1.5695,
+      "step": 1340
+    },
+    {
+      "epoch": 0.2783312577833126,
+      "grad_norm": 0.8102609613745911,
+      "learning_rate": 3.3546183409857605e-07,
+      "loss": 1.5913,
+      "step": 1341
+    },
+    {
+      "epoch": 0.2785388127853881,
+      "grad_norm": 0.6309960468971255,
+      "learning_rate": 3.3537176819199407e-07,
+      "loss": 1.5117,
+      "step": 1342
+    },
+    {
+      "epoch": 0.2787463677874637,
+      "grad_norm": 0.655136803538487,
+      "learning_rate": 3.352816532323992e-07,
+      "loss": 1.5687,
+      "step": 1343
+    },
+    {
+      "epoch": 0.27895392278953923,
+      "grad_norm": 1.0736280622254268,
+      "learning_rate": 3.35191489258106e-07,
+      "loss": 1.526,
+      "step": 1344
+    },
+    {
+      "epoch": 0.27916147779161476,
+      "grad_norm": 2.5400433135755356,
+      "learning_rate": 3.351012763074496e-07,
+      "loss": 1.5219,
+      "step": 1345
+    },
+    {
+      "epoch": 0.27936903279369035,
+      "grad_norm": 0.8499342752672578,
+      "learning_rate": 3.350110144187864e-07,
+      "loss": 1.6216,
+      "step": 1346
+    },
+    {
+      "epoch": 0.2795765877957659,
+      "grad_norm": 0.9035787915782699,
+      "learning_rate": 3.3492070363049315e-07,
+      "loss": 1.6016,
+      "step": 1347
+    },
+    {
+      "epoch": 0.2797841427978414,
+      "grad_norm": 1.0699886207147347,
+      "learning_rate": 3.3483034398096777e-07,
+      "loss": 1.4731,
+      "step": 1348
+    },
+    {
+      "epoch": 0.279991697799917,
+      "grad_norm": 2.166681950803039,
+      "learning_rate": 3.347399355086286e-07,
+      "loss": 1.504,
+      "step": 1349
+    },
+    {
+      "epoch": 0.28019925280199254,
+      "grad_norm": 0.7351985886540563,
+      "learning_rate": 3.3464947825191507e-07,
+      "loss": 1.5447,
+      "step": 1350
+    },
+    {
+      "epoch": 0.28040680780406807,
+      "grad_norm": 0.8037359505487619,
+      "learning_rate": 3.3455897224928717e-07,
+      "loss": 1.5657,
+      "step": 1351
+    },
+    {
+      "epoch": 0.2806143628061436,
+      "grad_norm": 0.6751663426318365,
+      "learning_rate": 3.3446841753922565e-07,
+      "loss": 1.4916,
+      "step": 1352
+    },
+    {
+      "epoch": 0.2808219178082192,
+      "grad_norm": 1.0977158443448278,
+      "learning_rate": 3.343778141602319e-07,
+      "loss": 1.486,
+      "step": 1353
+    },
+    {
+      "epoch": 0.2810294728102947,
+      "grad_norm": 0.8736167191500952,
+      "learning_rate": 3.3428716215082823e-07,
+      "loss": 1.5328,
+      "step": 1354
+    },
+    {
+      "epoch": 0.28123702781237025,
+      "grad_norm": 0.8801826959208915,
+      "learning_rate": 3.341964615495573e-07,
+      "loss": 1.5398,
+      "step": 1355
+    },
+    {
+      "epoch": 0.28144458281444584,
+      "grad_norm": 0.8796282094919217,
+      "learning_rate": 3.3410571239498266e-07,
+      "loss": 1.5499,
+      "step": 1356
+    },
+    {
+      "epoch": 0.2816521378165214,
+      "grad_norm": 0.722629334914455,
+      "learning_rate": 3.3401491472568843e-07,
+      "loss": 1.4327,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2818596928185969,
+      "grad_norm": 1.0383693580401288,
+      "learning_rate": 3.339240685802794e-07,
+      "loss": 1.5016,
+      "step": 1358
+    },
+    {
+      "epoch": 0.2820672478206725,
+      "grad_norm": 1.0356899312072005,
+      "learning_rate": 3.338331739973809e-07,
+      "loss": 1.5734,
+      "step": 1359
+    },
+    {
+      "epoch": 0.282274802822748,
+      "grad_norm": 0.7177652800960455,
+      "learning_rate": 3.3374223101563894e-07,
+      "loss": 1.578,
+      "step": 1360
+    },
+    {
+      "epoch": 0.28248235782482356,
+      "grad_norm": 0.7229667525162314,
+      "learning_rate": 3.3365123967372e-07,
+      "loss": 1.5139,
+      "step": 1361
+    },
+    {
+      "epoch": 0.28268991282689915,
+      "grad_norm": 1.4787085351149962,
+      "learning_rate": 3.3356020001031126e-07,
+      "loss": 1.5397,
+      "step": 1362
+    },
+    {
+      "epoch": 0.2828974678289747,
+      "grad_norm": 0.6660004355114754,
+      "learning_rate": 3.3346911206412033e-07,
+      "loss": 1.5141,
+      "step": 1363
+    },
+    {
+      "epoch": 0.2831050228310502,
+      "grad_norm": 0.6433280477219206,
+      "learning_rate": 3.333779758738754e-07,
+      "loss": 1.5792,
+      "step": 1364
+    },
+    {
+      "epoch": 0.2833125778331258,
+      "grad_norm": 0.7489890878045019,
+      "learning_rate": 3.3328679147832516e-07,
+      "loss": 1.5808,
+      "step": 1365
+    },
+    {
+      "epoch": 0.28352013283520133,
+      "grad_norm": 0.8010819810767609,
+      "learning_rate": 3.3319555891623864e-07,
+      "loss": 1.5154,
+      "step": 1366
+    },
+    {
+      "epoch": 0.28372768783727687,
+      "grad_norm": 0.9181979113229911,
+      "learning_rate": 3.331042782264058e-07,
+      "loss": 1.497,
+      "step": 1367
+    },
+    {
+      "epoch": 0.28393524283935245,
+      "grad_norm": 0.7165485941022961,
+      "learning_rate": 3.3301294944763647e-07,
+      "loss": 1.5121,
+      "step": 1368
+    },
+    {
+      "epoch": 0.284142797841428,
+      "grad_norm": 0.6830942915816728,
+      "learning_rate": 3.3292157261876146e-07,
+      "loss": 1.482,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2843503528435035,
+      "grad_norm": 0.8371948148412527,
+      "learning_rate": 3.3283014777863165e-07,
+      "loss": 1.5007,
+      "step": 1370
+    },
+    {
+      "epoch": 0.2845579078455791,
+      "grad_norm": 0.762599666183577,
+      "learning_rate": 3.327386749661185e-07,
+      "loss": 1.4943,
+      "step": 1371
+    },
+    {
+      "epoch": 0.28476546284765464,
+      "grad_norm": 0.8275233197463118,
+      "learning_rate": 3.326471542201137e-07,
+      "loss": 1.5951,
+      "step": 1372
+    },
+    {
+      "epoch": 0.28497301784973017,
+      "grad_norm": 0.7922012907311398,
+      "learning_rate": 3.3255558557952965e-07,
+      "loss": 1.4832,
+      "step": 1373
+    },
+    {
+      "epoch": 0.2851805728518057,
+      "grad_norm": 0.6497828187569625,
+      "learning_rate": 3.324639690832987e-07,
+      "loss": 1.6181,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2853881278538813,
+      "grad_norm": 0.8024742814319767,
+      "learning_rate": 3.3237230477037387e-07,
+      "loss": 1.5728,
+      "step": 1375
+    },
+    {
+      "epoch": 0.2855956828559568,
+      "grad_norm": 1.2464912098915424,
+      "learning_rate": 3.322805926797284e-07,
+      "loss": 1.5018,
+      "step": 1376
+    },
+    {
+      "epoch": 0.28580323785803236,
+      "grad_norm": 1.2118377303976169,
+      "learning_rate": 3.321888328503558e-07,
+      "loss": 1.488,
+      "step": 1377
+    },
+    {
+      "epoch": 0.28601079286010794,
+      "grad_norm": 0.8677127039471785,
+      "learning_rate": 3.320970253212699e-07,
+      "loss": 1.52,
+      "step": 1378
+    },
+    {
+      "epoch": 0.2862183478621835,
+      "grad_norm": 0.7946896236611976,
+      "learning_rate": 3.3200517013150485e-07,
+      "loss": 1.6039,
+      "step": 1379
+    },
+    {
+      "epoch": 0.286425902864259,
+      "grad_norm": 1.0209073010802001,
+      "learning_rate": 3.31913267320115e-07,
+      "loss": 1.51,
+      "step": 1380
+    },
+    {
+      "epoch": 0.2866334578663346,
+      "grad_norm": 1.25456686165671,
+      "learning_rate": 3.3182131692617497e-07,
+      "loss": 1.557,
+      "step": 1381
+    },
+    {
+      "epoch": 0.28684101286841013,
+      "grad_norm": 1.0615213634109164,
+      "learning_rate": 3.3172931898877976e-07,
+      "loss": 1.5744,
+      "step": 1382
+    },
+    {
+      "epoch": 0.28704856787048566,
+      "grad_norm": 0.6858341477285249,
+      "learning_rate": 3.316372735470444e-07,
+      "loss": 1.4688,
+      "step": 1383
+    },
+    {
+      "epoch": 0.28725612287256125,
+      "grad_norm": 0.694940603241787,
+      "learning_rate": 3.3154518064010403e-07,
+      "loss": 1.5382,
+      "step": 1384
+    },
+    {
+      "epoch": 0.2874636778746368,
+      "grad_norm": 4.924309431518236,
+      "learning_rate": 3.314530403071142e-07,
+      "loss": 1.4993,
+      "step": 1385
+    },
+    {
+      "epoch": 0.2876712328767123,
+      "grad_norm": 0.878152732898713,
+      "learning_rate": 3.313608525872506e-07,
+      "loss": 1.5711,
+      "step": 1386
+    },
+    {
+      "epoch": 0.2878787878787879,
+      "grad_norm": 0.8255705074112311,
+      "learning_rate": 3.312686175197089e-07,
+      "loss": 1.5321,
+      "step": 1387
+    },
+    {
+      "epoch": 0.28808634288086343,
+      "grad_norm": 0.670398028968472,
+      "learning_rate": 3.311763351437051e-07,
+      "loss": 1.529,
+      "step": 1388
+    },
+    {
+      "epoch": 0.28829389788293897,
+      "grad_norm": 1.141528302956471,
+      "learning_rate": 3.310840054984751e-07,
+      "loss": 1.4962,
+      "step": 1389
+    },
+    {
+      "epoch": 0.28850145288501455,
+      "grad_norm": 0.980374076707978,
+      "learning_rate": 3.3099162862327517e-07,
+      "loss": 1.571,
+      "step": 1390
+    },
+    {
+      "epoch": 0.2887090078870901,
+      "grad_norm": 0.7968563122606515,
+      "learning_rate": 3.3089920455738135e-07,
+      "loss": 1.4797,
+      "step": 1391
+    },
+    {
+      "epoch": 0.2889165628891656,
+      "grad_norm": 0.7778048330000662,
+      "learning_rate": 3.308067333400899e-07,
+      "loss": 1.5637,
+      "step": 1392
+    },
+    {
+      "epoch": 0.2891241178912412,
+      "grad_norm": 0.6469492821123097,
+      "learning_rate": 3.307142150107172e-07,
+      "loss": 1.5405,
+      "step": 1393
+    },
+    {
+      "epoch": 0.28933167289331674,
+      "grad_norm": 0.8904714328368389,
+      "learning_rate": 3.306216496085996e-07,
+      "loss": 1.5123,
+      "step": 1394
+    },
+    {
+      "epoch": 0.28953922789539227,
+      "grad_norm": 0.841095944065256,
+      "learning_rate": 3.305290371730935e-07,
+      "loss": 1.5405,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2897467828974678,
+      "grad_norm": 0.9836085117911635,
+      "learning_rate": 3.304363777435751e-07,
+      "loss": 1.5187,
+      "step": 1396
+    },
+    {
+      "epoch": 0.2899543378995434,
+      "grad_norm": 0.6847796469797756,
+      "learning_rate": 3.3034367135944077e-07,
+      "loss": 1.5553,
+      "step": 1397
+    },
+    {
+      "epoch": 0.2901618929016189,
+      "grad_norm": 1.3767365081878549,
+      "learning_rate": 3.302509180601069e-07,
+      "loss": 1.4842,
+      "step": 1398
+    },
+    {
+      "epoch": 0.29036944790369446,
+      "grad_norm": 1.7179775329560176,
+      "learning_rate": 3.3015811788500965e-07,
+      "loss": 1.5059,
+      "step": 1399
+    },
+    {
+      "epoch": 0.29057700290577004,
+      "grad_norm": 0.8127238892282448,
+      "learning_rate": 3.300652708736052e-07,
+      "loss": 1.4776,
+      "step": 1400
+    },
+    {
+      "epoch": 0.2907845579078456,
+      "grad_norm": 0.8727007834826351,
+      "learning_rate": 3.299723770653696e-07,
+      "loss": 1.5215,
+      "step": 1401
+    },
+    {
+      "epoch": 0.2909921129099211,
+      "grad_norm": 0.6508045803468447,
+      "learning_rate": 3.2987943649979894e-07,
+      "loss": 1.5088,
+      "step": 1402
+    },
+    {
+      "epoch": 0.2911996679119967,
+      "grad_norm": 0.8812833851828636,
+      "learning_rate": 3.29786449216409e-07,
+      "loss": 1.5782,
+      "step": 1403
+    },
+    {
+      "epoch": 0.29140722291407223,
+      "grad_norm": 1.1967984831572973,
+      "learning_rate": 3.2969341525473545e-07,
+      "loss": 1.5297,
+      "step": 1404
+    },
+    {
+      "epoch": 0.29161477791614776,
+      "grad_norm": 2.2821788784399586,
+      "learning_rate": 3.2960033465433404e-07,
+      "loss": 1.5594,
+      "step": 1405
+    },
+    {
+      "epoch": 0.29182233291822335,
+      "grad_norm": 0.9436529448984353,
+      "learning_rate": 3.2950720745477995e-07,
+      "loss": 1.5813,
+      "step": 1406
+    },
+    {
+      "epoch": 0.2920298879202989,
+      "grad_norm": 0.6671248652370418,
+      "learning_rate": 3.294140336956686e-07,
+      "loss": 1.6244,
+      "step": 1407
+    },
+    {
+      "epoch": 0.2922374429223744,
+      "grad_norm": 0.7773983954069197,
+      "learning_rate": 3.293208134166148e-07,
+      "loss": 1.5582,
+      "step": 1408
+    },
+    {
+      "epoch": 0.29244499792445,
+      "grad_norm": 1.306306347959827,
+      "learning_rate": 3.292275466572535e-07,
+      "loss": 1.5607,
+      "step": 1409
+    },
+    {
+      "epoch": 0.29265255292652553,
+      "grad_norm": 0.7771192019268828,
+      "learning_rate": 3.291342334572392e-07,
+      "loss": 1.5766,
+      "step": 1410
+    },
+    {
+      "epoch": 0.29286010792860107,
+      "grad_norm": 1.053851512481581,
+      "learning_rate": 3.290408738562462e-07,
+      "loss": 1.5459,
+      "step": 1411
+    },
+    {
+      "epoch": 0.29306766293067665,
+      "grad_norm": 0.7854016789739342,
+      "learning_rate": 3.2894746789396843e-07,
+      "loss": 1.5953,
+      "step": 1412
+    },
+    {
+      "epoch": 0.2932752179327522,
+      "grad_norm": 0.8811751599708441,
+      "learning_rate": 3.288540156101197e-07,
+      "loss": 1.5447,
+      "step": 1413
+    },
+    {
+      "epoch": 0.2934827729348277,
+      "grad_norm": 0.8532443790437596,
+      "learning_rate": 3.2876051704443356e-07,
+      "loss": 1.504,
+      "step": 1414
+    },
+    {
+      "epoch": 0.2936903279369033,
+      "grad_norm": 0.7224694881158574,
+      "learning_rate": 3.286669722366628e-07,
+      "loss": 1.5505,
+      "step": 1415
+    },
+    {
+      "epoch": 0.29389788293897884,
+      "grad_norm": 0.754753634337428,
+      "learning_rate": 3.2857338122658054e-07,
+      "loss": 1.4468,
+      "step": 1416
+    },
+    {
+      "epoch": 0.29410543794105437,
+      "grad_norm": 1.2362141982583785,
+      "learning_rate": 3.2847974405397904e-07,
+      "loss": 1.4769,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2943129929431299,
+      "grad_norm": 0.9164007262195232,
+      "learning_rate": 3.283860607586703e-07,
+      "loss": 1.4946,
+      "step": 1418
+    },
+    {
+      "epoch": 0.2945205479452055,
+      "grad_norm": 1.1049340030253059,
+      "learning_rate": 3.28292331380486e-07,
+      "loss": 1.5109,
+      "step": 1419
+    },
+    {
+      "epoch": 0.294728102947281,
+      "grad_norm": 1.1513378433777173,
+      "learning_rate": 3.281985559592775e-07,
+      "loss": 1.4919,
+      "step": 1420
+    },
+    {
+      "epoch": 0.29493565794935656,
+      "grad_norm": 1.3770131462140454,
+      "learning_rate": 3.281047345349154e-07,
+      "loss": 1.6033,
+      "step": 1421
+    },
+    {
+      "epoch": 0.29514321295143214,
+      "grad_norm": 0.7988219249609612,
+      "learning_rate": 3.280108671472902e-07,
+      "loss": 1.524,
+      "step": 1422
+    },
+    {
+      "epoch": 0.2953507679535077,
+      "grad_norm": 1.8136097396798223,
+      "learning_rate": 3.279169538363119e-07,
+      "loss": 1.5501,
+      "step": 1423
+    },
+    {
+      "epoch": 0.2955583229555832,
+      "grad_norm": 1.0985061222247552,
+      "learning_rate": 3.2782299464190977e-07,
+      "loss": 1.5024,
+      "step": 1424
+    },
+    {
+      "epoch": 0.2957658779576588,
+      "grad_norm": 2.274843550556237,
+      "learning_rate": 3.277289896040329e-07,
+      "loss": 1.4161,
+      "step": 1425
+    },
+    {
+      "epoch": 0.29597343295973433,
+      "grad_norm": 0.7016579066219234,
+      "learning_rate": 3.276349387626497e-07,
+      "loss": 1.5348,
+      "step": 1426
+    },
+    {
+      "epoch": 0.29618098796180986,
+      "grad_norm": 3.9305990413286143,
+      "learning_rate": 3.2754084215774805e-07,
+      "loss": 1.5662,
+      "step": 1427
+    },
+    {
+      "epoch": 0.29638854296388545,
+      "grad_norm": 1.3508454303397472,
+      "learning_rate": 3.2744669982933533e-07,
+      "loss": 1.4868,
+      "step": 1428
+    },
+    {
+      "epoch": 0.296596097965961,
+      "grad_norm": 0.8133253397220569,
+      "learning_rate": 3.273525118174385e-07,
+      "loss": 1.5166,
+      "step": 1429
+    },
+    {
+      "epoch": 0.2968036529680365,
+      "grad_norm": 0.7292492744462185,
+      "learning_rate": 3.272582781621036e-07,
+      "loss": 1.4296,
+      "step": 1430
+    },
+    {
+      "epoch": 0.2970112079701121,
+      "grad_norm": 1.1744269924600776,
+      "learning_rate": 3.271639989033964e-07,
+      "loss": 1.5518,
+      "step": 1431
+    },
+    {
+      "epoch": 0.29721876297218763,
+      "grad_norm": 0.8617304688474138,
+      "learning_rate": 3.270696740814019e-07,
+      "loss": 1.5373,
+      "step": 1432
+    },
+    {
+      "epoch": 0.29742631797426317,
+      "grad_norm": 0.7748498227845443,
+      "learning_rate": 3.2697530373622456e-07,
+      "loss": 1.5572,
+      "step": 1433
+    },
+    {
+      "epoch": 0.29763387297633875,
+      "grad_norm": 0.996435065936527,
+      "learning_rate": 3.2688088790798805e-07,
+      "loss": 1.5103,
+      "step": 1434
+    },
+    {
+      "epoch": 0.2978414279784143,
+      "grad_norm": 0.7514739154389984,
+      "learning_rate": 3.267864266368356e-07,
+      "loss": 1.5786,
+      "step": 1435
+    },
+    {
+      "epoch": 0.2980489829804898,
+      "grad_norm": 0.7802674892361465,
+      "learning_rate": 3.266919199629295e-07,
+      "loss": 1.5494,
+      "step": 1436
+    },
+    {
+      "epoch": 0.2982565379825654,
+      "grad_norm": 2.4547405386977643,
+      "learning_rate": 3.265973679264515e-07,
+      "loss": 1.523,
+      "step": 1437
+    },
+    {
+      "epoch": 0.29846409298464094,
+      "grad_norm": 1.8008797986363987,
+      "learning_rate": 3.2650277056760277e-07,
+      "loss": 1.5008,
+      "step": 1438
+    },
+    {
+      "epoch": 0.29867164798671647,
+      "grad_norm": 0.7646547976048881,
+      "learning_rate": 3.264081279266034e-07,
+      "loss": 1.4195,
+      "step": 1439
+    },
+    {
+      "epoch": 0.298879202988792,
+      "grad_norm": 0.8377092749577859,
+      "learning_rate": 3.2631344004369303e-07,
+      "loss": 1.6047,
+      "step": 1440
+    },
+    {
+      "epoch": 0.2990867579908676,
+      "grad_norm": 0.7624046233112225,
+      "learning_rate": 3.262187069591304e-07,
+      "loss": 1.5179,
+      "step": 1441
+    },
+    {
+      "epoch": 0.2992943129929431,
+      "grad_norm": 1.5539460774173561,
+      "learning_rate": 3.2612392871319356e-07,
+      "loss": 1.5303,
+      "step": 1442
+    },
+    {
+      "epoch": 0.29950186799501866,
+      "grad_norm": 1.8339698584285773,
+      "learning_rate": 3.2602910534617966e-07,
+      "loss": 1.5572,
+      "step": 1443
+    },
+    {
+      "epoch": 0.29970942299709424,
+      "grad_norm": 0.7229225258599659,
+      "learning_rate": 3.2593423689840504e-07,
+      "loss": 1.5798,
+      "step": 1444
+    },
+    {
+      "epoch": 0.2999169779991698,
+      "grad_norm": 1.1951479312285485,
+      "learning_rate": 3.2583932341020524e-07,
+      "loss": 1.4973,
+      "step": 1445
+    },
+    {
+      "epoch": 0.3001245330012453,
+      "grad_norm": 0.6510380730429101,
+      "learning_rate": 3.2574436492193507e-07,
+      "loss": 1.4627,
+      "step": 1446
+    },
+    {
+      "epoch": 0.3003320880033209,
+      "grad_norm": 8.980845521490364,
+      "learning_rate": 3.2564936147396826e-07,
+      "loss": 1.5012,
+      "step": 1447
+    },
+    {
+      "epoch": 0.30053964300539643,
+      "grad_norm": 0.649235346083898,
+      "learning_rate": 3.2555431310669786e-07,
+      "loss": 1.5352,
+      "step": 1448
+    },
+    {
+      "epoch": 0.30074719800747196,
+      "grad_norm": 0.6584661656135817,
+      "learning_rate": 3.2545921986053574e-07,
+      "loss": 1.5303,
+      "step": 1449
+    },
+    {
+      "epoch": 0.30095475300954755,
+      "grad_norm": 0.8579868121235932,
+      "learning_rate": 3.253640817759132e-07,
+      "loss": 1.4733,
+      "step": 1450
+    },
+    {
+      "epoch": 0.3011623080116231,
+      "grad_norm": 0.874759434511237,
+      "learning_rate": 3.252688988932803e-07,
+      "loss": 1.5217,
+      "step": 1451
+    },
+    {
+      "epoch": 0.3013698630136986,
+      "grad_norm": 1.8249093374354721,
+      "learning_rate": 3.2517367125310633e-07,
+      "loss": 1.5272,
+      "step": 1452
+    },
+    {
+      "epoch": 0.3015774180157742,
+      "grad_norm": 0.9111732378935296,
+      "learning_rate": 3.250783988958795e-07,
+      "loss": 1.5509,
+      "step": 1453
+    },
+    {
+      "epoch": 0.30178497301784973,
+      "grad_norm": 6.763506849431699,
+      "learning_rate": 3.2498308186210716e-07,
+      "loss": 1.545,
+      "step": 1454
+    },
+    {
+      "epoch": 0.30199252801992527,
+      "grad_norm": 0.6229391503821824,
+      "learning_rate": 3.248877201923156e-07,
+      "loss": 1.4267,
+      "step": 1455
+    },
+    {
+      "epoch": 0.30220008302200085,
+      "grad_norm": 0.7965010096817351,
+      "learning_rate": 3.2479231392704994e-07,
+      "loss": 1.5477,
+      "step": 1456
+    },
+    {
+      "epoch": 0.3024076380240764,
+      "grad_norm": 0.7047090854295499,
+      "learning_rate": 3.2469686310687453e-07,
+      "loss": 1.5514,
+      "step": 1457
+    },
+    {
+      "epoch": 0.3026151930261519,
+      "grad_norm": 1.2636396073348313,
+      "learning_rate": 3.246013677723725e-07,
+      "loss": 1.552,
+      "step": 1458
+    },
+    {
+      "epoch": 0.3028227480282275,
+      "grad_norm": 0.6724373666837078,
+      "learning_rate": 3.2450582796414583e-07,
+      "loss": 1.6598,
+      "step": 1459
+    },
+    {
+      "epoch": 0.30303030303030304,
+      "grad_norm": 0.8141770331673945,
+      "learning_rate": 3.244102437228157e-07,
+      "loss": 1.5257,
+      "step": 1460
+    },
+    {
+      "epoch": 0.30323785803237857,
+      "grad_norm": 0.6351867484208295,
+      "learning_rate": 3.2431461508902177e-07,
+      "loss": 1.4958,
+      "step": 1461
+    },
+    {
+      "epoch": 0.3034454130344541,
+      "grad_norm": 1.1014433562112305,
+      "learning_rate": 3.2421894210342294e-07,
+      "loss": 1.5116,
+      "step": 1462
+    },
+    {
+      "epoch": 0.3036529680365297,
+      "grad_norm": 0.9439205676959231,
+      "learning_rate": 3.241232248066967e-07,
+      "loss": 1.6225,
+      "step": 1463
+    },
+    {
+      "epoch": 0.3038605230386052,
+      "grad_norm": 1.2747688983522443,
+      "learning_rate": 3.2402746323953973e-07,
+      "loss": 1.5608,
+      "step": 1464
+    },
+    {
+      "epoch": 0.30406807804068076,
+      "grad_norm": 1.2117643087248915,
+      "learning_rate": 3.239316574426671e-07,
+      "loss": 1.5473,
+      "step": 1465
+    },
+    {
+      "epoch": 0.30427563304275634,
+      "grad_norm": 1.4327599611642121,
+      "learning_rate": 3.2383580745681287e-07,
+      "loss": 1.5785,
+      "step": 1466
+    },
+    {
+      "epoch": 0.3044831880448319,
+      "grad_norm": 0.7804375176536272,
+      "learning_rate": 3.2373991332273005e-07,
+      "loss": 1.5125,
+      "step": 1467
+    },
+    {
+      "epoch": 0.3046907430469074,
+      "grad_norm": 0.8744877261273809,
+      "learning_rate": 3.2364397508119025e-07,
+      "loss": 1.5426,
+      "step": 1468
+    },
+    {
+      "epoch": 0.304898298048983,
+      "grad_norm": 0.652419939295772,
+      "learning_rate": 3.235479927729838e-07,
+      "loss": 1.5356,
+      "step": 1469
+    },
+    {
+      "epoch": 0.30510585305105853,
+      "grad_norm": 0.7963126907186158,
+      "learning_rate": 3.2345196643891997e-07,
+      "loss": 1.575,
+      "step": 1470
+    },
+    {
+      "epoch": 0.30531340805313406,
+      "grad_norm": 0.7956472285325714,
+      "learning_rate": 3.233558961198264e-07,
+      "loss": 1.4802,
+      "step": 1471
+    },
+    {
+      "epoch": 0.30552096305520965,
+      "grad_norm": 0.7809079460953622,
+      "learning_rate": 3.2325978185654973e-07,
+      "loss": 1.4406,
+      "step": 1472
+    },
+    {
+      "epoch": 0.3057285180572852,
+      "grad_norm": 0.8946802679198177,
+      "learning_rate": 3.2316362368995524e-07,
+      "loss": 1.5148,
+      "step": 1473
+    },
+    {
+      "epoch": 0.3059360730593607,
+      "grad_norm": 0.9007462742163707,
+      "learning_rate": 3.230674216609268e-07,
+      "loss": 1.5207,
+      "step": 1474
+    },
+    {
+      "epoch": 0.3061436280614363,
+      "grad_norm": 1.1915168831473029,
+      "learning_rate": 3.2297117581036697e-07,
+      "loss": 1.4669,
+      "step": 1475
+    },
+    {
+      "epoch": 0.30635118306351183,
+      "grad_norm": 0.7114571722835049,
+      "learning_rate": 3.2287488617919693e-07,
+      "loss": 1.5498,
+      "step": 1476
+    },
+    {
+      "epoch": 0.30655873806558737,
+      "grad_norm": 0.6974210022531402,
+      "learning_rate": 3.227785528083564e-07,
+      "loss": 1.5246,
+      "step": 1477
+    },
+    {
+      "epoch": 0.30676629306766295,
+      "grad_norm": 0.8418907685530905,
+      "learning_rate": 3.2268217573880387e-07,
+      "loss": 1.508,
+      "step": 1478
+    },
+    {
+      "epoch": 0.3069738480697385,
+      "grad_norm": 0.6965700820659818,
+      "learning_rate": 3.225857550115162e-07,
+      "loss": 1.5171,
+      "step": 1479
+    },
+    {
+      "epoch": 0.307181403071814,
+      "grad_norm": 0.7194809493418677,
+      "learning_rate": 3.2248929066748906e-07,
+      "loss": 1.6061,
+      "step": 1480
+    },
+    {
+      "epoch": 0.3073889580738896,
+      "grad_norm": 0.7097689446817526,
+      "learning_rate": 3.2239278274773644e-07,
+      "loss": 1.4629,
+      "step": 1481
+    },
+    {
+      "epoch": 0.30759651307596514,
+      "grad_norm": 0.8917484868027727,
+      "learning_rate": 3.2229623129329104e-07,
+      "loss": 1.552,
+      "step": 1482
+    },
+    {
+      "epoch": 0.30780406807804067,
+      "grad_norm": 0.7241034861127932,
+      "learning_rate": 3.2219963634520385e-07,
+      "loss": 1.5119,
+      "step": 1483
+    },
+    {
+      "epoch": 0.3080116230801162,
+      "grad_norm": 0.7168320259652754,
+      "learning_rate": 3.221029979445445e-07,
+      "loss": 1.5064,
+      "step": 1484
+    },
+    {
+      "epoch": 0.3082191780821918,
+      "grad_norm": 0.7963308947496888,
+      "learning_rate": 3.2200631613240114e-07,
+      "loss": 1.5255,
+      "step": 1485
+    },
+    {
+      "epoch": 0.3084267330842673,
+      "grad_norm": 0.6938457222888774,
+      "learning_rate": 3.219095909498803e-07,
+      "loss": 1.5677,
+      "step": 1486
+    },
+    {
+      "epoch": 0.30863428808634286,
+      "grad_norm": 1.2338821644468418,
+      "learning_rate": 3.218128224381069e-07,
+      "loss": 1.5321,
+      "step": 1487
+    },
+    {
+      "epoch": 0.30884184308841844,
+      "grad_norm": 0.6823592784556315,
+      "learning_rate": 3.217160106382244e-07,
+      "loss": 1.5544,
+      "step": 1488
+    },
+    {
+      "epoch": 0.309049398090494,
+      "grad_norm": 0.8444088039991224,
+      "learning_rate": 3.216191555913946e-07,
+      "loss": 1.4365,
+      "step": 1489
+    },
+    {
+      "epoch": 0.3092569530925695,
+      "grad_norm": 1.1429283479613144,
+      "learning_rate": 3.215222573387976e-07,
+      "loss": 1.5496,
+      "step": 1490
+    },
+    {
+      "epoch": 0.3094645080946451,
+      "grad_norm": 1.0426805694076446,
+      "learning_rate": 3.214253159216321e-07,
+      "loss": 1.5383,
+      "step": 1491
+    },
+    {
+      "epoch": 0.30967206309672063,
+      "grad_norm": 0.9045859135706596,
+      "learning_rate": 3.213283313811149e-07,
+      "loss": 1.5721,
+      "step": 1492
+    },
+    {
+      "epoch": 0.30987961809879616,
+      "grad_norm": 0.6478698967547891,
+      "learning_rate": 3.2123130375848136e-07,
+      "loss": 1.5669,
+      "step": 1493
+    },
+    {
+      "epoch": 0.31008717310087175,
+      "grad_norm": 1.2844247912688618,
+      "learning_rate": 3.21134233094985e-07,
+      "loss": 1.5101,
+      "step": 1494
+    },
+    {
+      "epoch": 0.3102947281029473,
+      "grad_norm": 1.26847834702015,
+      "learning_rate": 3.210371194318977e-07,
+      "loss": 1.5249,
+      "step": 1495
+    },
+    {
+      "epoch": 0.3105022831050228,
+      "grad_norm": 0.8212468144770589,
+      "learning_rate": 3.2093996281050956e-07,
+      "loss": 1.5571,
+      "step": 1496
+    },
+    {
+      "epoch": 0.3107098381070984,
+      "grad_norm": 0.9878428196111082,
+      "learning_rate": 3.2084276327212905e-07,
+      "loss": 1.5576,
+      "step": 1497
+    },
+    {
+      "epoch": 0.31091739310917393,
+      "grad_norm": 0.8026416846886552,
+      "learning_rate": 3.207455208580828e-07,
+      "loss": 1.5075,
+      "step": 1498
+    },
+    {
+      "epoch": 0.31112494811124947,
+      "grad_norm": 0.6570688907094375,
+      "learning_rate": 3.2064823560971587e-07,
+      "loss": 1.4921,
+      "step": 1499
+    },
+    {
+      "epoch": 0.31133250311332505,
+      "grad_norm": 0.6857331667772839,
+      "learning_rate": 3.2055090756839103e-07,
+      "loss": 1.5331,
+      "step": 1500
+    },
+    {
+      "epoch": 0.3115400581154006,
+      "grad_norm": 0.9833241251790387,
+      "learning_rate": 3.204535367754899e-07,
+      "loss": 1.4873,
+      "step": 1501
+    },
+    {
+      "epoch": 0.3117476131174761,
+      "grad_norm": 0.8292374552917077,
+      "learning_rate": 3.203561232724118e-07,
+      "loss": 1.5424,
+      "step": 1502
+    },
+    {
+      "epoch": 0.3119551681195517,
+      "grad_norm": 0.8069331256009402,
+      "learning_rate": 3.202586671005743e-07,
+      "loss": 1.5921,
+      "step": 1503
+    },
+    {
+      "epoch": 0.31216272312162724,
+      "grad_norm": 0.6086234165403009,
+      "learning_rate": 3.201611683014133e-07,
+      "loss": 1.5202,
+      "step": 1504
+    },
+    {
+      "epoch": 0.31237027812370277,
+      "grad_norm": 0.8160269642831041,
+      "learning_rate": 3.200636269163827e-07,
+      "loss": 1.5802,
+      "step": 1505
+    },
+    {
+      "epoch": 0.3125778331257783,
+      "grad_norm": 0.7076949765826489,
+      "learning_rate": 3.1996604298695444e-07,
+      "loss": 1.5407,
+      "step": 1506
+    },
+    {
+      "epoch": 0.3127853881278539,
+      "grad_norm": 0.7648446115921089,
+      "learning_rate": 3.198684165546187e-07,
+      "loss": 1.4971,
+      "step": 1507
+    },
+    {
+      "epoch": 0.3129929431299294,
+      "grad_norm": 1.723302421532576,
+      "learning_rate": 3.1977074766088355e-07,
+      "loss": 1.5415,
+      "step": 1508
+    },
+    {
+      "epoch": 0.31320049813200496,
+      "grad_norm": 0.8775475948486927,
+      "learning_rate": 3.1967303634727525e-07,
+      "loss": 1.5314,
+      "step": 1509
+    },
+    {
+      "epoch": 0.31340805313408054,
+      "grad_norm": 0.8967103397146254,
+      "learning_rate": 3.195752826553381e-07,
+      "loss": 1.5252,
+      "step": 1510
+    },
+    {
+      "epoch": 0.3136156081361561,
+      "grad_norm": 0.6926574208637071,
+      "learning_rate": 3.194774866266343e-07,
+      "loss": 1.6335,
+      "step": 1511
+    },
+    {
+      "epoch": 0.3138231631382316,
+      "grad_norm": 0.6361002593920386,
+      "learning_rate": 3.193796483027442e-07,
+      "loss": 1.5409,
+      "step": 1512
+    },
+    {
+      "epoch": 0.3140307181403072,
+      "grad_norm": 0.6470691998681614,
+      "learning_rate": 3.1928176772526597e-07,
+      "loss": 1.5559,
+      "step": 1513
+    },
+    {
+      "epoch": 0.31423827314238273,
+      "grad_norm": 0.9979400111895484,
+      "learning_rate": 3.1918384493581603e-07,
+      "loss": 1.5492,
+      "step": 1514
+    },
+    {
+      "epoch": 0.31444582814445826,
+      "grad_norm": 0.7579857391826024,
+      "learning_rate": 3.1908587997602824e-07,
+      "loss": 1.5712,
+      "step": 1515
+    },
+    {
+      "epoch": 0.31465338314653385,
+      "grad_norm": 1.2558704077900733,
+      "learning_rate": 3.189878728875549e-07,
+      "loss": 1.624,
+      "step": 1516
+    },
+    {
+      "epoch": 0.3148609381486094,
+      "grad_norm": 0.6426516950425566,
+      "learning_rate": 3.1888982371206604e-07,
+      "loss": 1.5013,
+      "step": 1517
+    },
+    {
+      "epoch": 0.3150684931506849,
+      "grad_norm": 0.6837986299827858,
+      "learning_rate": 3.187917324912494e-07,
+      "loss": 1.5526,
+      "step": 1518
+    },
+    {
+      "epoch": 0.3152760481527605,
+      "grad_norm": 0.7238030627995922,
+      "learning_rate": 3.1869359926681097e-07,
+      "loss": 1.5368,
+      "step": 1519
+    },
+    {
+      "epoch": 0.31548360315483603,
+      "grad_norm": 0.7227540923543799,
+      "learning_rate": 3.1859542408047435e-07,
+      "loss": 1.5626,
+      "step": 1520
+    },
+    {
+      "epoch": 0.31569115815691157,
+      "grad_norm": 0.7174582378218057,
+      "learning_rate": 3.1849720697398093e-07,
+      "loss": 1.5067,
+      "step": 1521
+    },
+    {
+      "epoch": 0.31589871315898715,
+      "grad_norm": 0.8053684446427255,
+      "learning_rate": 3.1839894798909e-07,
+      "loss": 1.6452,
+      "step": 1522
+    },
+    {
+      "epoch": 0.3161062681610627,
+      "grad_norm": 1.25638468654692,
+      "learning_rate": 3.1830064716757875e-07,
+      "loss": 1.529,
+      "step": 1523
+    },
+    {
+      "epoch": 0.3163138231631382,
+      "grad_norm": 0.7546442315082738,
+      "learning_rate": 3.182023045512421e-07,
+      "loss": 1.5167,
+      "step": 1524
+    },
+    {
+      "epoch": 0.3165213781652138,
+      "grad_norm": 0.7048190227954557,
+      "learning_rate": 3.181039201818926e-07,
+      "loss": 1.5397,
+      "step": 1525
+    },
+    {
+      "epoch": 0.31672893316728934,
+      "grad_norm": 0.7912242124900313,
+      "learning_rate": 3.180054941013608e-07,
+      "loss": 1.5004,
+      "step": 1526
+    },
+    {
+      "epoch": 0.31693648816936487,
+      "grad_norm": 0.7566673249850133,
+      "learning_rate": 3.1790702635149483e-07,
+      "loss": 1.4853,
+      "step": 1527
+    },
+    {
+      "epoch": 0.3171440431714404,
+      "grad_norm": 1.0302422292310887,
+      "learning_rate": 3.178085169741605e-07,
+      "loss": 1.5521,
+      "step": 1528
+    },
+    {
+      "epoch": 0.317351598173516,
+      "grad_norm": 1.5335784294377346,
+      "learning_rate": 3.177099660112414e-07,
+      "loss": 1.5057,
+      "step": 1529
+    },
+    {
+      "epoch": 0.3175591531755915,
+      "grad_norm": 0.7161477842631548,
+      "learning_rate": 3.1761137350463883e-07,
+      "loss": 1.4984,
+      "step": 1530
+    },
+    {
+      "epoch": 0.31776670817766706,
+      "grad_norm": 1.9737898514115628,
+      "learning_rate": 3.175127394962717e-07,
+      "loss": 1.5586,
+      "step": 1531
+    },
+    {
+      "epoch": 0.31797426317974264,
+      "grad_norm": 0.8116458110557672,
+      "learning_rate": 3.1741406402807655e-07,
+      "loss": 1.4885,
+      "step": 1532
+    },
+    {
+      "epoch": 0.3181818181818182,
+      "grad_norm": 1.0863419919213408,
+      "learning_rate": 3.173153471420076e-07,
+      "loss": 1.515,
+      "step": 1533
+    },
+    {
+      "epoch": 0.3183893731838937,
+      "grad_norm": 0.7121551013623241,
+      "learning_rate": 3.172165888800365e-07,
+      "loss": 1.5156,
+      "step": 1534
+    },
+    {
+      "epoch": 0.3185969281859693,
+      "grad_norm": 0.6970222090298633,
+      "learning_rate": 3.171177892841528e-07,
+      "loss": 1.5274,
+      "step": 1535
+    },
+    {
+      "epoch": 0.31880448318804483,
+      "grad_norm": 0.7122464096822435,
+      "learning_rate": 3.170189483963635e-07,
+      "loss": 1.6166,
+      "step": 1536
+    },
+    {
+      "epoch": 0.31901203819012036,
+      "grad_norm": 0.6782791764309692,
+      "learning_rate": 3.169200662586931e-07,
+      "loss": 1.4854,
+      "step": 1537
+    },
+    {
+      "epoch": 0.31921959319219595,
+      "grad_norm": 0.7654414523252254,
+      "learning_rate": 3.168211429131835e-07,
+      "loss": 1.5325,
+      "step": 1538
+    },
+    {
+      "epoch": 0.3194271481942715,
+      "grad_norm": 0.6722686560973263,
+      "learning_rate": 3.1672217840189443e-07,
+      "loss": 1.4301,
+      "step": 1539
+    },
+    {
+      "epoch": 0.319634703196347,
+      "grad_norm": 0.8152407686796453,
+      "learning_rate": 3.166231727669029e-07,
+      "loss": 1.5538,
+      "step": 1540
+    },
+    {
+      "epoch": 0.3198422581984226,
+      "grad_norm": 0.7463891481452275,
+      "learning_rate": 3.165241260503035e-07,
+      "loss": 1.536,
+      "step": 1541
+    },
+    {
+      "epoch": 0.32004981320049813,
+      "grad_norm": 0.7667119659520718,
+      "learning_rate": 3.164250382942083e-07,
+      "loss": 1.5675,
+      "step": 1542
+    },
+    {
+      "epoch": 0.32025736820257367,
+      "grad_norm": 0.8600403298533835,
+      "learning_rate": 3.163259095407468e-07,
+      "loss": 1.5936,
+      "step": 1543
+    },
+    {
+      "epoch": 0.32046492320464925,
+      "grad_norm": 0.843974281229249,
+      "learning_rate": 3.1622673983206577e-07,
+      "loss": 1.5806,
+      "step": 1544
+    },
+    {
+      "epoch": 0.3206724782067248,
+      "grad_norm": 0.8409252655849362,
+      "learning_rate": 3.161275292103297e-07,
+      "loss": 1.5206,
+      "step": 1545
+    },
+    {
+      "epoch": 0.3208800332088003,
+      "grad_norm": 1.3683881215793297,
+      "learning_rate": 3.160282777177203e-07,
+      "loss": 1.5176,
+      "step": 1546
+    },
+    {
+      "epoch": 0.3210875882108759,
+      "grad_norm": 0.8530317832654084,
+      "learning_rate": 3.1592898539643653e-07,
+      "loss": 1.546,
+      "step": 1547
+    },
+    {
+      "epoch": 0.32129514321295144,
+      "grad_norm": 1.3046243530778758,
+      "learning_rate": 3.15829652288695e-07,
+      "loss": 1.5187,
+      "step": 1548
+    },
+    {
+      "epoch": 0.32150269821502697,
+      "grad_norm": 0.6626824338689945,
+      "learning_rate": 3.157302784367294e-07,
+      "loss": 1.5824,
+      "step": 1549
+    },
+    {
+      "epoch": 0.32171025321710256,
+      "grad_norm": 0.7150204682433556,
+      "learning_rate": 3.156308638827909e-07,
+      "loss": 1.5273,
+      "step": 1550
+    },
+    {
+      "epoch": 0.3219178082191781,
+      "grad_norm": 0.7082587380156864,
+      "learning_rate": 3.15531408669148e-07,
+      "loss": 1.5714,
+      "step": 1551
+    },
+    {
+      "epoch": 0.3221253632212536,
+      "grad_norm": 0.8129077176326319,
+      "learning_rate": 3.1543191283808633e-07,
+      "loss": 1.5416,
+      "step": 1552
+    },
+    {
+      "epoch": 0.32233291822332916,
+      "grad_norm": 0.6643762418379271,
+      "learning_rate": 3.153323764319088e-07,
+      "loss": 1.5461,
+      "step": 1553
+    },
+    {
+      "epoch": 0.32254047322540474,
+      "grad_norm": 2.719052829084681,
+      "learning_rate": 3.1523279949293584e-07,
+      "loss": 1.5452,
+      "step": 1554
+    },
+    {
+      "epoch": 0.3227480282274803,
+      "grad_norm": 0.7982673759207699,
+      "learning_rate": 3.151331820635048e-07,
+      "loss": 1.4871,
+      "step": 1555
+    },
+    {
+      "epoch": 0.3229555832295558,
+      "grad_norm": 1.4882366877931574,
+      "learning_rate": 3.1503352418597043e-07,
+      "loss": 1.5315,
+      "step": 1556
+    },
+    {
+      "epoch": 0.3231631382316314,
+      "grad_norm": 0.7444536583231158,
+      "learning_rate": 3.149338259027045e-07,
+      "loss": 1.5269,
+      "step": 1557
+    },
+    {
+      "epoch": 0.32337069323370693,
+      "grad_norm": 1.538861324119853,
+      "learning_rate": 3.1483408725609615e-07,
+      "loss": 1.4852,
+      "step": 1558
+    },
+    {
+      "epoch": 0.32357824823578246,
+      "grad_norm": 0.6568849312673052,
+      "learning_rate": 3.1473430828855164e-07,
+      "loss": 1.5304,
+      "step": 1559
+    },
+    {
+      "epoch": 0.32378580323785805,
+      "grad_norm": 1.1182090435908014,
+      "learning_rate": 3.146344890424943e-07,
+      "loss": 1.5594,
+      "step": 1560
+    },
+    {
+      "epoch": 0.3239933582399336,
+      "grad_norm": 0.7766036895438667,
+      "learning_rate": 3.145346295603646e-07,
+      "loss": 1.4441,
+      "step": 1561
+    },
+    {
+      "epoch": 0.3242009132420091,
+      "grad_norm": 0.9127264212547783,
+      "learning_rate": 3.144347298846202e-07,
+      "loss": 1.5319,
+      "step": 1562
+    },
+    {
+      "epoch": 0.3244084682440847,
+      "grad_norm": 0.6593738763657883,
+      "learning_rate": 3.1433479005773567e-07,
+      "loss": 1.5741,
+      "step": 1563
+    },
+    {
+      "epoch": 0.32461602324616023,
+      "grad_norm": 0.8911329221554665,
+      "learning_rate": 3.142348101222029e-07,
+      "loss": 1.4774,
+      "step": 1564
+    },
+    {
+      "epoch": 0.32482357824823577,
+      "grad_norm": 1.1799757042561831,
+      "learning_rate": 3.1413479012053065e-07,
+      "loss": 1.5811,
+      "step": 1565
+    },
+    {
+      "epoch": 0.32503113325031135,
+      "grad_norm": 30.82941877357409,
+      "learning_rate": 3.140347300952448e-07,
+      "loss": 1.6183,
+      "step": 1566
+    },
+    {
+      "epoch": 0.3252386882523869,
+      "grad_norm": 0.8686183829245135,
+      "learning_rate": 3.139346300888882e-07,
+      "loss": 1.5173,
+      "step": 1567
+    },
+    {
+      "epoch": 0.3254462432544624,
+      "grad_norm": 0.934343222546769,
+      "learning_rate": 3.138344901440207e-07,
+      "loss": 1.4974,
+      "step": 1568
+    },
+    {
+      "epoch": 0.325653798256538,
+      "grad_norm": 0.8288251586647396,
+      "learning_rate": 3.137343103032191e-07,
+      "loss": 1.4969,
+      "step": 1569
+    },
+    {
+      "epoch": 0.32586135325861354,
+      "grad_norm": 1.1756031358058125,
+      "learning_rate": 3.1363409060907735e-07,
+      "loss": 1.5287,
+      "step": 1570
+    },
+    {
+      "epoch": 0.32606890826068907,
+      "grad_norm": 0.6872982056877541,
+      "learning_rate": 3.1353383110420607e-07,
+      "loss": 1.5368,
+      "step": 1571
+    },
+    {
+      "epoch": 0.32627646326276466,
+      "grad_norm": 0.7173370722988562,
+      "learning_rate": 3.13433531831233e-07,
+      "loss": 1.518,
+      "step": 1572
+    },
+    {
+      "epoch": 0.3264840182648402,
+      "grad_norm": 0.7077151170311493,
+      "learning_rate": 3.1333319283280274e-07,
+      "loss": 1.6073,
+      "step": 1573
+    },
+    {
+      "epoch": 0.3266915732669157,
+      "grad_norm": 0.8533943399033741,
+      "learning_rate": 3.1323281415157665e-07,
+      "loss": 1.4611,
+      "step": 1574
+    },
+    {
+      "epoch": 0.32689912826899126,
+      "grad_norm": 0.7216180828932892,
+      "learning_rate": 3.1313239583023327e-07,
+      "loss": 1.5433,
+      "step": 1575
+    },
+    {
+      "epoch": 0.32710668327106684,
+      "grad_norm": 0.7657188170898415,
+      "learning_rate": 3.1303193791146767e-07,
+      "loss": 1.5461,
+      "step": 1576
+    },
+    {
+      "epoch": 0.3273142382731424,
+      "grad_norm": 0.7547586499023344,
+      "learning_rate": 3.129314404379919e-07,
+      "loss": 1.5458,
+      "step": 1577
+    },
+    {
+      "epoch": 0.3275217932752179,
+      "grad_norm": 0.6492722320495253,
+      "learning_rate": 3.1283090345253494e-07,
+      "loss": 1.449,
+      "step": 1578
+    },
+    {
+      "epoch": 0.3277293482772935,
+      "grad_norm": 0.694932859316317,
+      "learning_rate": 3.1273032699784223e-07,
+      "loss": 1.5038,
+      "step": 1579
+    },
+    {
+      "epoch": 0.32793690327936903,
+      "grad_norm": 0.760421015961153,
+      "learning_rate": 3.1262971111667643e-07,
+      "loss": 1.4807,
+      "step": 1580
+    },
+    {
+      "epoch": 0.32814445828144456,
+      "grad_norm": 1.0312798410817425,
+      "learning_rate": 3.125290558518166e-07,
+      "loss": 1.5945,
+      "step": 1581
+    },
+    {
+      "epoch": 0.32835201328352015,
+      "grad_norm": 1.7143441712325553,
+      "learning_rate": 3.1242836124605866e-07,
+      "loss": 1.4721,
+      "step": 1582
+    },
+    {
+      "epoch": 0.3285595682855957,
+      "grad_norm": 0.6655217244718277,
+      "learning_rate": 3.123276273422155e-07,
+      "loss": 1.4986,
+      "step": 1583
+    },
+    {
+      "epoch": 0.3287671232876712,
+      "grad_norm": 0.8346966955789498,
+      "learning_rate": 3.1222685418311625e-07,
+      "loss": 1.5709,
+      "step": 1584
+    },
+    {
+      "epoch": 0.3289746782897468,
+      "grad_norm": 0.6331326576478944,
+      "learning_rate": 3.121260418116071e-07,
+      "loss": 1.5167,
+      "step": 1585
+    },
+    {
+      "epoch": 0.32918223329182233,
+      "grad_norm": 0.7412456391611238,
+      "learning_rate": 3.120251902705508e-07,
+      "loss": 1.5368,
+      "step": 1586
+    },
+    {
+      "epoch": 0.32938978829389787,
+      "grad_norm": 0.6753891380104349,
+      "learning_rate": 3.1192429960282666e-07,
+      "loss": 1.5062,
+      "step": 1587
+    },
+    {
+      "epoch": 0.32959734329597346,
+      "grad_norm": 0.6640327679259239,
+      "learning_rate": 3.1182336985133083e-07,
+      "loss": 1.5371,
+      "step": 1588
+    },
+    {
+      "epoch": 0.329804898298049,
+      "grad_norm": 1.1952829517448376,
+      "learning_rate": 3.1172240105897596e-07,
+      "loss": 1.5394,
+      "step": 1589
+    },
+    {
+      "epoch": 0.3300124533001245,
+      "grad_norm": 0.9634946214688638,
+      "learning_rate": 3.116213932686912e-07,
+      "loss": 1.6199,
+      "step": 1590
+    },
+    {
+      "epoch": 0.3302200083022001,
+      "grad_norm": 0.8110358709993737,
+      "learning_rate": 3.1152034652342243e-07,
+      "loss": 1.5702,
+      "step": 1591
+    },
+    {
+      "epoch": 0.33042756330427564,
+      "grad_norm": 0.6553060442490959,
+      "learning_rate": 3.114192608661321e-07,
+      "loss": 1.5057,
+      "step": 1592
+    },
+    {
+      "epoch": 0.3306351183063512,
+      "grad_norm": 0.624511976827147,
+      "learning_rate": 3.1131813633979905e-07,
+      "loss": 1.5156,
+      "step": 1593
+    },
+    {
+      "epoch": 0.33084267330842676,
+      "grad_norm": 3.9508105574571273,
+      "learning_rate": 3.1121697298741874e-07,
+      "loss": 1.5963,
+      "step": 1594
+    },
+    {
+      "epoch": 0.3310502283105023,
+      "grad_norm": 0.6582180790077864,
+      "learning_rate": 3.1111577085200323e-07,
+      "loss": 1.4956,
+      "step": 1595
+    },
+    {
+      "epoch": 0.3312577833125778,
+      "grad_norm": 1.522087814498903,
+      "learning_rate": 3.1101452997658097e-07,
+      "loss": 1.5421,
+      "step": 1596
+    },
+    {
+      "epoch": 0.33146533831465336,
+      "grad_norm": 0.7153593782532924,
+      "learning_rate": 3.109132504041968e-07,
+      "loss": 1.4977,
+      "step": 1597
+    },
+    {
+      "epoch": 0.33167289331672895,
+      "grad_norm": 0.79503716445836,
+      "learning_rate": 3.1081193217791226e-07,
+      "loss": 1.5354,
+      "step": 1598
+    },
+    {
+      "epoch": 0.3318804483188045,
+      "grad_norm": 1.2461131906558136,
+      "learning_rate": 3.10710575340805e-07,
+      "loss": 1.4358,
+      "step": 1599
+    },
+    {
+      "epoch": 0.33208800332088,
+      "grad_norm": 0.718938286134833,
+      "learning_rate": 3.1060917993596933e-07,
+      "loss": 1.4774,
+      "step": 1600
+    },
+    {
+      "epoch": 0.3322955583229556,
+      "grad_norm": 0.68334836747786,
+      "learning_rate": 3.105077460065159e-07,
+      "loss": 1.5268,
+      "step": 1601
+    },
+    {
+      "epoch": 0.33250311332503113,
+      "grad_norm": 2.343952333780335,
+      "learning_rate": 3.1040627359557175e-07,
+      "loss": 1.4907,
+      "step": 1602
+    },
+    {
+      "epoch": 0.33271066832710666,
+      "grad_norm": 0.7640612952723437,
+      "learning_rate": 3.103047627462802e-07,
+      "loss": 1.4855,
+      "step": 1603
+    },
+    {
+      "epoch": 0.33291822332918225,
+      "grad_norm": 0.7532144797063448,
+      "learning_rate": 3.102032135018009e-07,
+      "loss": 1.5227,
+      "step": 1604
+    },
+    {
+      "epoch": 0.3331257783312578,
+      "grad_norm": 1.386706642060875,
+      "learning_rate": 3.101016259053101e-07,
+      "loss": 1.4513,
+      "step": 1605
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.7132200211075789,
+      "learning_rate": 3.1e-07,
+      "loss": 1.4837,
+      "step": 1606
+    },
+    {
+      "epoch": 0.3335408883354089,
+      "grad_norm": 0.7624633968700086,
+      "learning_rate": 3.098983358290792e-07,
+      "loss": 1.4883,
+      "step": 1607
+    },
+    {
+      "epoch": 0.33374844333748444,
+      "grad_norm": 1.6305508823080885,
+      "learning_rate": 3.0979663343577277e-07,
+      "loss": 1.5252,
+      "step": 1608
+    },
+    {
+      "epoch": 0.33395599833955997,
+      "grad_norm": 1.8227199718129374,
+      "learning_rate": 3.0969489286332174e-07,
+      "loss": 1.5444,
+      "step": 1609
+    },
+    {
+      "epoch": 0.33416355334163556,
+      "grad_norm": 0.7052644520628132,
+      "learning_rate": 3.0959311415498345e-07,
+      "loss": 1.4436,
+      "step": 1610
+    },
+    {
+      "epoch": 0.3343711083437111,
+      "grad_norm": 0.9998415197954762,
+      "learning_rate": 3.0949129735403165e-07,
+      "loss": 1.5216,
+      "step": 1611
+    },
+    {
+      "epoch": 0.3345786633457866,
+      "grad_norm": 1.9674711545113885,
+      "learning_rate": 3.093894425037561e-07,
+      "loss": 1.4963,
+      "step": 1612
+    },
+    {
+      "epoch": 0.3347862183478622,
+      "grad_norm": 1.113441702031003,
+      "learning_rate": 3.092875496474627e-07,
+      "loss": 1.4454,
+      "step": 1613
+    },
+    {
+      "epoch": 0.33499377334993774,
+      "grad_norm": 0.8016579227198851,
+      "learning_rate": 3.091856188284736e-07,
+      "loss": 1.5209,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3352013283520133,
+      "grad_norm": 0.6890212291227804,
+      "learning_rate": 3.090836500901272e-07,
+      "loss": 1.5137,
+      "step": 1615
+    },
+    {
+      "epoch": 0.33540888335408886,
+      "grad_norm": 0.8209015007259222,
+      "learning_rate": 3.0898164347577775e-07,
+      "loss": 1.4832,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3356164383561644,
+      "grad_norm": 0.7018157065024992,
+      "learning_rate": 3.0887959902879586e-07,
+      "loss": 1.5736,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3358239933582399,
+      "grad_norm": 1.4473239754467748,
+      "learning_rate": 3.087775167925681e-07,
+      "loss": 1.6125,
+      "step": 1618
+    },
+    {
+      "epoch": 0.33603154836031546,
+      "grad_norm": 0.7490796166398493,
+      "learning_rate": 3.086753968104971e-07,
+      "loss": 1.4745,
+      "step": 1619
+    },
+    {
+      "epoch": 0.33623910336239105,
+      "grad_norm": 1.2519146778893293,
+      "learning_rate": 3.085732391260016e-07,
+      "loss": 1.4962,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3364466583644666,
+      "grad_norm": 2.8196203071619137,
+      "learning_rate": 3.0847104378251623e-07,
+      "loss": 1.5235,
+      "step": 1621
+    },
+    {
+      "epoch": 0.3366542133665421,
+      "grad_norm": 0.815405779431192,
+      "learning_rate": 3.083688108234919e-07,
+      "loss": 1.6065,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3368617683686177,
+      "grad_norm": 0.9621634342903317,
+      "learning_rate": 3.082665402923952e-07,
+      "loss": 1.5325,
+      "step": 1623
+    },
+    {
+      "epoch": 0.33706932337069323,
+      "grad_norm": 0.8621212106888736,
+      "learning_rate": 3.08164232232709e-07,
+      "loss": 1.4955,
+      "step": 1624
+    },
+    {
+      "epoch": 0.33727687837276876,
+      "grad_norm": 0.6612875148662144,
+      "learning_rate": 3.0806188668793176e-07,
+      "loss": 1.5019,
+      "step": 1625
+    },
+    {
+      "epoch": 0.33748443337484435,
+      "grad_norm": 0.663738492465499,
+      "learning_rate": 3.079595037015783e-07,
+      "loss": 1.5535,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3376919883769199,
+      "grad_norm": 0.9422035390470533,
+      "learning_rate": 3.07857083317179e-07,
+      "loss": 1.4713,
+      "step": 1627
+    },
+    {
+      "epoch": 0.3378995433789954,
+      "grad_norm": 0.7883623781541931,
+      "learning_rate": 3.077546255782804e-07,
+      "loss": 1.5072,
+      "step": 1628
+    },
+    {
+      "epoch": 0.338107098381071,
+      "grad_norm": 0.7574632719595523,
+      "learning_rate": 3.076521305284447e-07,
+      "loss": 1.5881,
+      "step": 1629
+    },
+    {
+      "epoch": 0.33831465338314654,
+      "grad_norm": 0.7536334871385633,
+      "learning_rate": 3.0754959821125017e-07,
+      "loss": 1.496,
+      "step": 1630
+    },
+    {
+      "epoch": 0.33852220838522207,
+      "grad_norm": 0.7909478959005793,
+      "learning_rate": 3.074470286702908e-07,
+      "loss": 1.5074,
+      "step": 1631
+    },
+    {
+      "epoch": 0.33872976338729766,
+      "grad_norm": 0.8249333290394759,
+      "learning_rate": 3.073444219491764e-07,
+      "loss": 1.5348,
+      "step": 1632
+    },
+    {
+      "epoch": 0.3389373183893732,
+      "grad_norm": 1.023051553917948,
+      "learning_rate": 3.072417780915327e-07,
+      "loss": 1.5532,
+      "step": 1633
+    },
+    {
+      "epoch": 0.3391448733914487,
+      "grad_norm": 0.8990458627368273,
+      "learning_rate": 3.07139097141001e-07,
+      "loss": 1.4847,
+      "step": 1634
+    },
+    {
+      "epoch": 0.3393524283935243,
+      "grad_norm": 0.7114981462728718,
+      "learning_rate": 3.0703637914123864e-07,
+      "loss": 1.4415,
+      "step": 1635
+    },
+    {
+      "epoch": 0.33955998339559984,
+      "grad_norm": 0.9838345121038751,
+      "learning_rate": 3.069336241359186e-07,
+      "loss": 1.5449,
+      "step": 1636
+    },
+    {
+      "epoch": 0.3397675383976754,
+      "grad_norm": 0.7081896978793626,
+      "learning_rate": 3.068308321687296e-07,
+      "loss": 1.5356,
+      "step": 1637
+    },
+    {
+      "epoch": 0.33997509339975096,
+      "grad_norm": 0.6901392950455878,
+      "learning_rate": 3.0672800328337583e-07,
+      "loss": 1.5497,
+      "step": 1638
+    },
+    {
+      "epoch": 0.3401826484018265,
+      "grad_norm": 0.728927656555971,
+      "learning_rate": 3.0662513752357767e-07,
+      "loss": 1.5151,
+      "step": 1639
+    },
+    {
+      "epoch": 0.340390203403902,
+      "grad_norm": 0.8221644777235805,
+      "learning_rate": 3.0652223493307066e-07,
+      "loss": 1.51,
+      "step": 1640
+    },
+    {
+      "epoch": 0.34059775840597756,
+      "grad_norm": 0.6619593327455161,
+      "learning_rate": 3.064192955556066e-07,
+      "loss": 1.5578,
+      "step": 1641
+    },
+    {
+      "epoch": 0.34080531340805315,
+      "grad_norm": 0.6711668860031297,
+      "learning_rate": 3.063163194349522e-07,
+      "loss": 1.5519,
+      "step": 1642
+    },
+    {
+      "epoch": 0.3410128684101287,
+      "grad_norm": 1.0665924311144357,
+      "learning_rate": 3.062133066148904e-07,
+      "loss": 1.6023,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3412204234122042,
+      "grad_norm": 1.0919493056165805,
+      "learning_rate": 3.061102571392195e-07,
+      "loss": 1.6064,
+      "step": 1644
+    },
+    {
+      "epoch": 0.3414279784142798,
+      "grad_norm": 0.8579946574964484,
+      "learning_rate": 3.0600717105175327e-07,
+      "loss": 1.5201,
+      "step": 1645
+    },
+    {
+      "epoch": 0.34163553341635533,
+      "grad_norm": 0.8572452091318757,
+      "learning_rate": 3.059040483963214e-07,
+      "loss": 1.5426,
+      "step": 1646
+    },
+    {
+      "epoch": 0.34184308841843086,
+      "grad_norm": 0.6299908917917835,
+      "learning_rate": 3.058008892167687e-07,
+      "loss": 1.5534,
+      "step": 1647
+    },
+    {
+      "epoch": 0.34205064342050645,
+      "grad_norm": 1.6094766890804149,
+      "learning_rate": 3.0569769355695575e-07,
+      "loss": 1.4546,
+      "step": 1648
+    },
+    {
+      "epoch": 0.342258198422582,
+      "grad_norm": 0.94109579604955,
+      "learning_rate": 3.055944614607587e-07,
+      "loss": 1.5961,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3424657534246575,
+      "grad_norm": 0.714434693264507,
+      "learning_rate": 3.054911929720691e-07,
+      "loss": 1.5651,
+      "step": 1650
+    },
+    {
+      "epoch": 0.3426733084267331,
+      "grad_norm": 0.824520636839947,
+      "learning_rate": 3.053878881347938e-07,
+      "loss": 1.5642,
+      "step": 1651
+    },
+    {
+      "epoch": 0.34288086342880864,
+      "grad_norm": 0.7509265291697873,
+      "learning_rate": 3.052845469928554e-07,
+      "loss": 1.4685,
+      "step": 1652
+    },
+    {
+      "epoch": 0.34308841843088417,
+      "grad_norm": 1.0078662464984964,
+      "learning_rate": 3.051811695901918e-07,
+      "loss": 1.444,
+      "step": 1653
+    },
+    {
+      "epoch": 0.34329597343295976,
+      "grad_norm": 2.3313540365874497,
+      "learning_rate": 3.0507775597075634e-07,
+      "loss": 1.4941,
+      "step": 1654
+    },
+    {
+      "epoch": 0.3435035284350353,
+      "grad_norm": 0.813050234345806,
+      "learning_rate": 3.049743061785177e-07,
+      "loss": 1.5321,
+      "step": 1655
+    },
+    {
+      "epoch": 0.3437110834371108,
+      "grad_norm": 0.6732430021714549,
+      "learning_rate": 3.0487082025746007e-07,
+      "loss": 1.4882,
+      "step": 1656
+    },
+    {
+      "epoch": 0.3439186384391864,
+      "grad_norm": 0.8195534802277931,
+      "learning_rate": 3.047672982515828e-07,
+      "loss": 1.6275,
+      "step": 1657
+    },
+    {
+      "epoch": 0.34412619344126194,
+      "grad_norm": 0.8052939454467235,
+      "learning_rate": 3.046637402049008e-07,
+      "loss": 1.5414,
+      "step": 1658
+    },
+    {
+      "epoch": 0.3443337484433375,
+      "grad_norm": 0.9324483000140465,
+      "learning_rate": 3.045601461614442e-07,
+      "loss": 1.5398,
+      "step": 1659
+    },
+    {
+      "epoch": 0.34454130344541306,
+      "grad_norm": 0.77879291620623,
+      "learning_rate": 3.044565161652583e-07,
+      "loss": 1.5063,
+      "step": 1660
+    },
+    {
+      "epoch": 0.3447488584474886,
+      "grad_norm": 0.8697844337300528,
+      "learning_rate": 3.0435285026040393e-07,
+      "loss": 1.4811,
+      "step": 1661
+    },
+    {
+      "epoch": 0.3449564134495641,
+      "grad_norm": 0.7403197810690658,
+      "learning_rate": 3.0424914849095715e-07,
+      "loss": 1.4977,
+      "step": 1662
+    },
+    {
+      "epoch": 0.34516396845163966,
+      "grad_norm": 1.1406082873430088,
+      "learning_rate": 3.0414541090100907e-07,
+      "loss": 1.5005,
+      "step": 1663
+    },
+    {
+      "epoch": 0.34537152345371525,
+      "grad_norm": 0.8772203904166385,
+      "learning_rate": 3.040416375346662e-07,
+      "loss": 1.5712,
+      "step": 1664
+    },
+    {
+      "epoch": 0.3455790784557908,
+      "grad_norm": 1.0097804432803577,
+      "learning_rate": 3.0393782843605025e-07,
+      "loss": 1.4485,
+      "step": 1665
+    },
+    {
+      "epoch": 0.3457866334578663,
+      "grad_norm": 0.8945710588706391,
+      "learning_rate": 3.0383398364929807e-07,
+      "loss": 1.4628,
+      "step": 1666
+    },
+    {
+      "epoch": 0.3459941884599419,
+      "grad_norm": 1.2166364844098123,
+      "learning_rate": 3.0373010321856164e-07,
+      "loss": 1.554,
+      "step": 1667
+    },
+    {
+      "epoch": 0.34620174346201743,
+      "grad_norm": 0.8278050827604245,
+      "learning_rate": 3.0362618718800834e-07,
+      "loss": 1.4815,
+      "step": 1668
+    },
+    {
+      "epoch": 0.34640929846409296,
+      "grad_norm": 0.6830986360349536,
+      "learning_rate": 3.035222356018203e-07,
+      "loss": 1.5209,
+      "step": 1669
+    },
+    {
+      "epoch": 0.34661685346616855,
+      "grad_norm": 0.8092350944352389,
+      "learning_rate": 3.034182485041951e-07,
+      "loss": 1.5253,
+      "step": 1670
+    },
+    {
+      "epoch": 0.3468244084682441,
+      "grad_norm": 0.8330013459319859,
+      "learning_rate": 3.033142259393453e-07,
+      "loss": 1.5411,
+      "step": 1671
+    },
+    {
+      "epoch": 0.3470319634703196,
+      "grad_norm": 0.752721215090041,
+      "learning_rate": 3.0321016795149847e-07,
+      "loss": 1.5835,
+      "step": 1672
+    },
+    {
+      "epoch": 0.3472395184723952,
+      "grad_norm": 0.6675975075083221,
+      "learning_rate": 3.0310607458489734e-07,
+      "loss": 1.4351,
+      "step": 1673
+    },
+    {
+      "epoch": 0.34744707347447074,
+      "grad_norm": 0.622488184844189,
+      "learning_rate": 3.0300194588379964e-07,
+      "loss": 1.537,
+      "step": 1674
+    },
+    {
+      "epoch": 0.34765462847654627,
+      "grad_norm": 0.8522550700196722,
+      "learning_rate": 3.0289778189247816e-07,
+      "loss": 1.5081,
+      "step": 1675
+    },
+    {
+      "epoch": 0.34786218347862186,
+      "grad_norm": 0.69898024471361,
+      "learning_rate": 3.0279358265522053e-07,
+      "loss": 1.4991,
+      "step": 1676
+    },
+    {
+      "epoch": 0.3480697384806974,
+      "grad_norm": 0.7433784810098976,
+      "learning_rate": 3.026893482163297e-07,
+      "loss": 1.5236,
+      "step": 1677
+    },
+    {
+      "epoch": 0.3482772934827729,
+      "grad_norm": 1.0199962591268212,
+      "learning_rate": 3.025850786201232e-07,
+      "loss": 1.5772,
+      "step": 1678
+    },
+    {
+      "epoch": 0.3484848484848485,
+      "grad_norm": 0.6550305302333714,
+      "learning_rate": 3.0248077391093384e-07,
+      "loss": 1.5698,
+      "step": 1679
+    },
+    {
+      "epoch": 0.34869240348692404,
+      "grad_norm": 0.7776577145060054,
+      "learning_rate": 3.023764341331092e-07,
+      "loss": 1.5546,
+      "step": 1680
+    },
+    {
+      "epoch": 0.3488999584889996,
+      "grad_norm": 1.2816873166807066,
+      "learning_rate": 3.0227205933101166e-07,
+      "loss": 1.5292,
+      "step": 1681
+    },
+    {
+      "epoch": 0.34910751349107516,
+      "grad_norm": 1.261557960183847,
+      "learning_rate": 3.0216764954901865e-07,
+      "loss": 1.4805,
+      "step": 1682
+    },
+    {
+      "epoch": 0.3493150684931507,
+      "grad_norm": 0.6774303188709085,
+      "learning_rate": 3.020632048315226e-07,
+      "loss": 1.5187,
+      "step": 1683
+    },
+    {
+      "epoch": 0.3495226234952262,
+      "grad_norm": 2.035685522152664,
+      "learning_rate": 3.019587252229304e-07,
+      "loss": 1.5804,
+      "step": 1684
+    },
+    {
+      "epoch": 0.34973017849730176,
+      "grad_norm": 1.1254319215830477,
+      "learning_rate": 3.018542107676642e-07,
+      "loss": 1.5053,
+      "step": 1685
+    },
+    {
+      "epoch": 0.34993773349937735,
+      "grad_norm": 0.8275366226382337,
+      "learning_rate": 3.0174966151016064e-07,
+      "loss": 1.5517,
+      "step": 1686
+    },
+    {
+      "epoch": 0.3501452885014529,
+      "grad_norm": 0.6859446592128088,
+      "learning_rate": 3.016450774948713e-07,
+      "loss": 1.5276,
+      "step": 1687
+    },
+    {
+      "epoch": 0.3503528435035284,
+      "grad_norm": 0.8553843053007176,
+      "learning_rate": 3.0154045876626264e-07,
+      "loss": 1.5983,
+      "step": 1688
+    },
+    {
+      "epoch": 0.350560398505604,
+      "grad_norm": 0.869081237501683,
+      "learning_rate": 3.014358053688157e-07,
+      "loss": 1.6072,
+      "step": 1689
+    },
+    {
+      "epoch": 0.35076795350767953,
+      "grad_norm": 0.7909907178079894,
+      "learning_rate": 3.013311173470262e-07,
+      "loss": 1.5368,
+      "step": 1690
+    },
+    {
+      "epoch": 0.35097550850975506,
+      "grad_norm": 0.6394119879036785,
+      "learning_rate": 3.0122639474540493e-07,
+      "loss": 1.4718,
+      "step": 1691
+    },
+    {
+      "epoch": 0.35118306351183065,
+      "grad_norm": 0.9759307754666919,
+      "learning_rate": 3.01121637608477e-07,
+      "loss": 1.5045,
+      "step": 1692
+    },
+    {
+      "epoch": 0.3513906185139062,
+      "grad_norm": 0.8695526425880886,
+      "learning_rate": 3.0101684598078244e-07,
+      "loss": 1.5284,
+      "step": 1693
+    },
+    {
+      "epoch": 0.3515981735159817,
+      "grad_norm": 0.7888488891482031,
+      "learning_rate": 3.0091201990687586e-07,
+      "loss": 1.5402,
+      "step": 1694
+    },
+    {
+      "epoch": 0.3518057285180573,
+      "grad_norm": 0.6291405641894751,
+      "learning_rate": 3.0080715943132646e-07,
+      "loss": 1.5051,
+      "step": 1695
+    },
+    {
+      "epoch": 0.35201328352013284,
+      "grad_norm": 0.7103108192927109,
+      "learning_rate": 3.007022645987182e-07,
+      "loss": 1.5028,
+      "step": 1696
+    },
+    {
+      "epoch": 0.35222083852220837,
+      "grad_norm": 0.7627372914788422,
+      "learning_rate": 3.005973354536496e-07,
+      "loss": 1.5835,
+      "step": 1697
+    },
+    {
+      "epoch": 0.35242839352428396,
+      "grad_norm": 0.725042565721643,
+      "learning_rate": 3.004923720407336e-07,
+      "loss": 1.5028,
+      "step": 1698
+    },
+    {
+      "epoch": 0.3526359485263595,
+      "grad_norm": 0.8297371250754101,
+      "learning_rate": 3.00387374404598e-07,
+      "loss": 1.5794,
+      "step": 1699
+    },
+    {
+      "epoch": 0.352843503528435,
+      "grad_norm": 1.247375235901272,
+      "learning_rate": 3.0028234258988503e-07,
+      "loss": 1.501,
+      "step": 1700
+    },
+    {
+      "epoch": 0.3530510585305106,
+      "grad_norm": 0.7979091860000276,
+      "learning_rate": 3.001772766412513e-07,
+      "loss": 1.5222,
+      "step": 1701
+    },
+    {
+      "epoch": 0.35325861353258614,
+      "grad_norm": 0.7936083419914423,
+      "learning_rate": 3.0007217660336816e-07,
+      "loss": 1.5614,
+      "step": 1702
+    },
+    {
+      "epoch": 0.3534661685346617,
+      "grad_norm": 0.8009676053899694,
+      "learning_rate": 2.9996704252092137e-07,
+      "loss": 1.5254,
+      "step": 1703
+    },
+    {
+      "epoch": 0.35367372353673726,
+      "grad_norm": 1.043574240180602,
+      "learning_rate": 2.9986187443861103e-07,
+      "loss": 1.5266,
+      "step": 1704
+    },
+    {
+      "epoch": 0.3538812785388128,
+      "grad_norm": 0.7379523630474946,
+      "learning_rate": 2.997566724011519e-07,
+      "loss": 1.5135,
+      "step": 1705
+    },
+    {
+      "epoch": 0.3540888335408883,
+      "grad_norm": 0.6707427472142247,
+      "learning_rate": 2.996514364532731e-07,
+      "loss": 1.598,
+      "step": 1706
+    },
+    {
+      "epoch": 0.35429638854296386,
+      "grad_norm": 0.7105516763227429,
+      "learning_rate": 2.995461666397181e-07,
+      "loss": 1.472,
+      "step": 1707
+    },
+    {
+      "epoch": 0.35450394354503945,
+      "grad_norm": 0.9825429304315128,
+      "learning_rate": 2.9944086300524493e-07,
+      "loss": 1.5137,
+      "step": 1708
+    },
+    {
+      "epoch": 0.354711498547115,
+      "grad_norm": 0.8266271207201454,
+      "learning_rate": 2.9933552559462586e-07,
+      "loss": 1.5352,
+      "step": 1709
+    },
+    {
+      "epoch": 0.3549190535491905,
+      "grad_norm": 0.8420143656260886,
+      "learning_rate": 2.9923015445264746e-07,
+      "loss": 1.5245,
+      "step": 1710
+    },
+    {
+      "epoch": 0.3551266085512661,
+      "grad_norm": 0.9319206561434165,
+      "learning_rate": 2.991247496241109e-07,
+      "loss": 1.5242,
+      "step": 1711
+    },
+    {
+      "epoch": 0.35533416355334163,
+      "grad_norm": 0.7284856736795287,
+      "learning_rate": 2.9901931115383143e-07,
+      "loss": 1.5128,
+      "step": 1712
+    },
+    {
+      "epoch": 0.35554171855541716,
+      "grad_norm": 1.1061317297106878,
+      "learning_rate": 2.989138390866388e-07,
+      "loss": 1.5403,
+      "step": 1713
+    },
+    {
+      "epoch": 0.35574927355749275,
+      "grad_norm": 0.8702660793802259,
+      "learning_rate": 2.9880833346737664e-07,
+      "loss": 1.4661,
+      "step": 1714
+    },
+    {
+      "epoch": 0.3559568285595683,
+      "grad_norm": 3.1766649706898957,
+      "learning_rate": 2.9870279434090346e-07,
+      "loss": 1.5005,
+      "step": 1715
+    },
+    {
+      "epoch": 0.3561643835616438,
+      "grad_norm": 0.7205636259189994,
+      "learning_rate": 2.9859722175209153e-07,
+      "loss": 1.5109,
+      "step": 1716
+    },
+    {
+      "epoch": 0.3563719385637194,
+      "grad_norm": 0.6283346875682314,
+      "learning_rate": 2.984916157458275e-07,
+      "loss": 1.4877,
+      "step": 1717
+    },
+    {
+      "epoch": 0.35657949356579494,
+      "grad_norm": 1.6563760549803141,
+      "learning_rate": 2.983859763670123e-07,
+      "loss": 1.5204,
+      "step": 1718
+    },
+    {
+      "epoch": 0.35678704856787047,
+      "grad_norm": 1.1189558043,
+      "learning_rate": 2.9828030366056106e-07,
+      "loss": 1.5492,
+      "step": 1719
+    },
+    {
+      "epoch": 0.35699460356994606,
+      "grad_norm": 0.7216092395063066,
+      "learning_rate": 2.9817459767140286e-07,
+      "loss": 1.5167,
+      "step": 1720
+    },
+    {
+      "epoch": 0.3572021585720216,
+      "grad_norm": 0.7441271563187284,
+      "learning_rate": 2.980688584444812e-07,
+      "loss": 1.4984,
+      "step": 1721
+    },
+    {
+      "epoch": 0.3574097135740971,
+      "grad_norm": 0.6889872910853054,
+      "learning_rate": 2.979630860247535e-07,
+      "loss": 1.5651,
+      "step": 1722
+    },
+    {
+      "epoch": 0.3576172685761727,
+      "grad_norm": 0.6704762098524681,
+      "learning_rate": 2.978572804571914e-07,
+      "loss": 1.4289,
+      "step": 1723
+    },
+    {
+      "epoch": 0.35782482357824824,
+      "grad_norm": 0.958615776488211,
+      "learning_rate": 2.977514417867807e-07,
+      "loss": 1.5443,
+      "step": 1724
+    },
+    {
+      "epoch": 0.3580323785803238,
+      "grad_norm": 0.7406615626568862,
+      "learning_rate": 2.9764557005852113e-07,
+      "loss": 1.4956,
+      "step": 1725
+    },
+    {
+      "epoch": 0.35823993358239936,
+      "grad_norm": 0.8771154807976586,
+      "learning_rate": 2.9753966531742645e-07,
+      "loss": 1.4966,
+      "step": 1726
+    },
+    {
+      "epoch": 0.3584474885844749,
+      "grad_norm": 0.7483783007618476,
+      "learning_rate": 2.974337276085248e-07,
+      "loss": 1.5266,
+      "step": 1727
+    },
+    {
+      "epoch": 0.3586550435865504,
+      "grad_norm": 0.754628660201579,
+      "learning_rate": 2.973277569768578e-07,
+      "loss": 1.5666,
+      "step": 1728
+    },
+    {
+      "epoch": 0.35886259858862596,
+      "grad_norm": 0.8747559113838734,
+      "learning_rate": 2.972217534674815e-07,
+      "loss": 1.6055,
+      "step": 1729
+    },
+    {
+      "epoch": 0.35907015359070155,
+      "grad_norm": 1.116113725806758,
+      "learning_rate": 2.971157171254658e-07,
+      "loss": 1.5103,
+      "step": 1730
+    },
+    {
+      "epoch": 0.3592777085927771,
+      "grad_norm": 0.7861015005845284,
+      "learning_rate": 2.970096479958944e-07,
+      "loss": 1.6098,
+      "step": 1731
+    },
+    {
+      "epoch": 0.3594852635948526,
+      "grad_norm": 0.9617823634146951,
+      "learning_rate": 2.969035461238652e-07,
+      "loss": 1.5086,
+      "step": 1732
+    },
+    {
+      "epoch": 0.3596928185969282,
+      "grad_norm": 0.961114820817053,
+      "learning_rate": 2.9679741155448983e-07,
+      "loss": 1.4787,
+      "step": 1733
+    },
+    {
+      "epoch": 0.35990037359900373,
+      "grad_norm": 0.855547225438654,
+      "learning_rate": 2.9669124433289396e-07,
+      "loss": 1.5384,
+      "step": 1734
+    },
+    {
+      "epoch": 0.36010792860107926,
+      "grad_norm": 0.7805276087870763,
+      "learning_rate": 2.96585044504217e-07,
+      "loss": 1.5389,
+      "step": 1735
+    },
+    {
+      "epoch": 0.36031548360315485,
+      "grad_norm": 0.8136550380958589,
+      "learning_rate": 2.9647881211361237e-07,
+      "loss": 1.5576,
+      "step": 1736
+    },
+    {
+      "epoch": 0.3605230386052304,
+      "grad_norm": 0.66497298643468,
+      "learning_rate": 2.963725472062472e-07,
+      "loss": 1.454,
+      "step": 1737
+    },
+    {
+      "epoch": 0.3607305936073059,
+      "grad_norm": 0.6276381689953632,
+      "learning_rate": 2.962662498273026e-07,
+      "loss": 1.4555,
+      "step": 1738
+    },
+    {
+      "epoch": 0.3609381486093815,
+      "grad_norm": 0.6614671787854753,
+      "learning_rate": 2.9615992002197325e-07,
+      "loss": 1.5379,
+      "step": 1739
+    },
+    {
+      "epoch": 0.36114570361145704,
+      "grad_norm": 0.8525112543558966,
+      "learning_rate": 2.9605355783546787e-07,
+      "loss": 1.5815,
+      "step": 1740
+    },
+    {
+      "epoch": 0.36135325861353257,
+      "grad_norm": 0.7129837198729112,
+      "learning_rate": 2.959471633130088e-07,
+      "loss": 1.4539,
+      "step": 1741
+    },
+    {
+      "epoch": 0.36156081361560816,
+      "grad_norm": 0.9130059709848856,
+      "learning_rate": 2.958407364998322e-07,
+      "loss": 1.4706,
+      "step": 1742
+    },
+    {
+      "epoch": 0.3617683686176837,
+      "grad_norm": 0.8027343517265195,
+      "learning_rate": 2.957342774411878e-07,
+      "loss": 1.5139,
+      "step": 1743
+    },
+    {
+      "epoch": 0.3619759236197592,
+      "grad_norm": 0.7397254986736014,
+      "learning_rate": 2.956277861823394e-07,
+      "loss": 1.5687,
+      "step": 1744
+    },
+    {
+      "epoch": 0.3621834786218348,
+      "grad_norm": 0.9829383375014513,
+      "learning_rate": 2.9552126276856404e-07,
+      "loss": 1.5216,
+      "step": 1745
+    },
+    {
+      "epoch": 0.36239103362391034,
+      "grad_norm": 0.6952958906345301,
+      "learning_rate": 2.954147072451527e-07,
+      "loss": 1.4976,
+      "step": 1746
+    },
+    {
+      "epoch": 0.3625985886259859,
+      "grad_norm": 0.6378002366491281,
+      "learning_rate": 2.9530811965741003e-07,
+      "loss": 1.577,
+      "step": 1747
+    },
+    {
+      "epoch": 0.36280614362806146,
+      "grad_norm": 0.6219270941064697,
+      "learning_rate": 2.9520150005065414e-07,
+      "loss": 1.4592,
+      "step": 1748
+    },
+    {
+      "epoch": 0.363013698630137,
+      "grad_norm": 0.8451233477653258,
+      "learning_rate": 2.9509484847021704e-07,
+      "loss": 1.5592,
+      "step": 1749
+    },
+    {
+      "epoch": 0.3632212536322125,
+      "grad_norm": 1.2275080116382218,
+      "learning_rate": 2.9498816496144394e-07,
+      "loss": 1.5534,
+      "step": 1750
+    },
+    {
+      "epoch": 0.36342880863428806,
+      "grad_norm": 1.1182092905517573,
+      "learning_rate": 2.9488144956969394e-07,
+      "loss": 1.4897,
+      "step": 1751
+    },
+    {
+      "epoch": 0.36363636363636365,
+      "grad_norm": 0.7721664941862921,
+      "learning_rate": 2.947747023403396e-07,
+      "loss": 1.5315,
+      "step": 1752
+    },
+    {
+      "epoch": 0.3638439186384392,
+      "grad_norm": 0.7818740065443556,
+      "learning_rate": 2.946679233187669e-07,
+      "loss": 1.5258,
+      "step": 1753
+    },
+    {
+      "epoch": 0.3640514736405147,
+      "grad_norm": 0.7051419859263878,
+      "learning_rate": 2.9456111255037556e-07,
+      "loss": 1.5029,
+      "step": 1754
+    },
+    {
+      "epoch": 0.3642590286425903,
+      "grad_norm": 0.747064840265756,
+      "learning_rate": 2.944542700805787e-07,
+      "loss": 1.4954,
+      "step": 1755
+    },
+    {
+      "epoch": 0.36446658364466583,
+      "grad_norm": 0.707961041524801,
+      "learning_rate": 2.943473959548028e-07,
+      "loss": 1.5186,
+      "step": 1756
+    },
+    {
+      "epoch": 0.36467413864674136,
+      "grad_norm": 0.884401846257606,
+      "learning_rate": 2.942404902184879e-07,
+      "loss": 1.4792,
+      "step": 1757
+    },
+    {
+      "epoch": 0.36488169364881695,
+      "grad_norm": 0.7312295811536884,
+      "learning_rate": 2.941335529170876e-07,
+      "loss": 1.4889,
+      "step": 1758
+    },
+    {
+      "epoch": 0.3650892486508925,
+      "grad_norm": 1.0756486533974505,
+      "learning_rate": 2.940265840960687e-07,
+      "loss": 1.5116,
+      "step": 1759
+    },
+    {
+      "epoch": 0.365296803652968,
+      "grad_norm": 3.969806928744894,
+      "learning_rate": 2.939195838009116e-07,
+      "loss": 1.5481,
+      "step": 1760
+    },
+    {
+      "epoch": 0.3655043586550436,
+      "grad_norm": 7.209525118202673,
+      "learning_rate": 2.938125520771099e-07,
+      "loss": 1.5164,
+      "step": 1761
+    },
+    {
+      "epoch": 0.36571191365711914,
+      "grad_norm": 0.7153874088280192,
+      "learning_rate": 2.937054889701706e-07,
+      "loss": 1.4693,
+      "step": 1762
+    },
+    {
+      "epoch": 0.36591946865919467,
+      "grad_norm": 0.850598206314216,
+      "learning_rate": 2.9359839452561426e-07,
+      "loss": 1.5628,
+      "step": 1763
+    },
+    {
+      "epoch": 0.36612702366127026,
+      "grad_norm": 0.8585994069962113,
+      "learning_rate": 2.934912687889744e-07,
+      "loss": 1.4284,
+      "step": 1764
+    },
+    {
+      "epoch": 0.3663345786633458,
+      "grad_norm": 1.0627420158289687,
+      "learning_rate": 2.933841118057982e-07,
+      "loss": 1.5067,
+      "step": 1765
+    },
+    {
+      "epoch": 0.3665421336654213,
+      "grad_norm": 1.1408889396014805,
+      "learning_rate": 2.932769236216459e-07,
+      "loss": 1.5725,
+      "step": 1766
+    },
+    {
+      "epoch": 0.3667496886674969,
+      "grad_norm": 0.7987928242629375,
+      "learning_rate": 2.9316970428209104e-07,
+      "loss": 1.5193,
+      "step": 1767
+    },
+    {
+      "epoch": 0.36695724366957244,
+      "grad_norm": 0.7225197506249644,
+      "learning_rate": 2.930624538327205e-07,
+      "loss": 1.5201,
+      "step": 1768
+    },
+    {
+      "epoch": 0.367164798671648,
+      "grad_norm": 0.8497929654463315,
+      "learning_rate": 2.9295517231913423e-07,
+      "loss": 1.5435,
+      "step": 1769
+    },
+    {
+      "epoch": 0.36737235367372356,
+      "grad_norm": 0.6124953585769698,
+      "learning_rate": 2.928478597869456e-07,
+      "loss": 1.5074,
+      "step": 1770
+    },
+    {
+      "epoch": 0.3675799086757991,
+      "grad_norm": 0.8525349756883807,
+      "learning_rate": 2.927405162817809e-07,
+      "loss": 1.5803,
+      "step": 1771
+    },
+    {
+      "epoch": 0.3677874636778746,
+      "grad_norm": 3.031277381744521,
+      "learning_rate": 2.9263314184927987e-07,
+      "loss": 1.559,
+      "step": 1772
+    },
+    {
+      "epoch": 0.3679950186799502,
+      "grad_norm": 0.6602677669129073,
+      "learning_rate": 2.925257365350952e-07,
+      "loss": 1.4774,
+      "step": 1773
+    },
+    {
+      "epoch": 0.36820257368202575,
+      "grad_norm": 0.6551522227737647,
+      "learning_rate": 2.9241830038489293e-07,
+      "loss": 1.5337,
+      "step": 1774
+    },
+    {
+      "epoch": 0.3684101286841013,
+      "grad_norm": 0.8985469487129819,
+      "learning_rate": 2.9231083344435185e-07,
+      "loss": 1.4968,
+      "step": 1775
+    },
+    {
+      "epoch": 0.3686176836861768,
+      "grad_norm": 0.6463507985368129,
+      "learning_rate": 2.9220333575916414e-07,
+      "loss": 1.5401,
+      "step": 1776
+    },
+    {
+      "epoch": 0.3688252386882524,
+      "grad_norm": 1.0489510793874743,
+      "learning_rate": 2.920958073750349e-07,
+      "loss": 1.4496,
+      "step": 1777
+    },
+    {
+      "epoch": 0.36903279369032793,
+      "grad_norm": 0.9971658252145071,
+      "learning_rate": 2.9198824833768245e-07,
+      "loss": 1.4645,
+      "step": 1778
+    },
+    {
+      "epoch": 0.36924034869240346,
+      "grad_norm": 0.6773521974366477,
+      "learning_rate": 2.91880658692838e-07,
+      "loss": 1.4861,
+      "step": 1779
+    },
+    {
+      "epoch": 0.36944790369447905,
+      "grad_norm": 0.797381790314639,
+      "learning_rate": 2.9177303848624576e-07,
+      "loss": 1.4955,
+      "step": 1780
+    },
+    {
+      "epoch": 0.3696554586965546,
+      "grad_norm": 0.7243669833038442,
+      "learning_rate": 2.9166538776366305e-07,
+      "loss": 1.5523,
+      "step": 1781
+    },
+    {
+      "epoch": 0.3698630136986301,
+      "grad_norm": 0.8073564724124318,
+      "learning_rate": 2.915577065708601e-07,
+      "loss": 1.4681,
+      "step": 1782
+    },
+    {
+      "epoch": 0.3700705687007057,
+      "grad_norm": 0.7486457762552564,
+      "learning_rate": 2.9144999495361993e-07,
+      "loss": 1.492,
+      "step": 1783
+    },
+    {
+      "epoch": 0.37027812370278124,
+      "grad_norm": 0.73969202222816,
+      "learning_rate": 2.913422529577389e-07,
+      "loss": 1.5026,
+      "step": 1784
+    },
+    {
+      "epoch": 0.37048567870485677,
+      "grad_norm": 0.7982331611979465,
+      "learning_rate": 2.91234480629026e-07,
+      "loss": 1.4837,
+      "step": 1785
+    },
+    {
+      "epoch": 0.37069323370693236,
+      "grad_norm": 1.433331735456798,
+      "learning_rate": 2.91126678013303e-07,
+      "loss": 1.5643,
+      "step": 1786
+    },
+    {
+      "epoch": 0.3709007887090079,
+      "grad_norm": 0.6966227458426625,
+      "learning_rate": 2.9101884515640486e-07,
+      "loss": 1.5435,
+      "step": 1787
+    },
+    {
+      "epoch": 0.3711083437110834,
+      "grad_norm": 1.1889136660738293,
+      "learning_rate": 2.909109821041792e-07,
+      "loss": 1.48,
+      "step": 1788
+    },
+    {
+      "epoch": 0.371315898713159,
+      "grad_norm": 1.2852233645627547,
+      "learning_rate": 2.9080308890248646e-07,
+      "loss": 1.4536,
+      "step": 1789
+    },
+    {
+      "epoch": 0.37152345371523454,
+      "grad_norm": 0.7116942681505585,
+      "learning_rate": 2.906951655972001e-07,
+      "loss": 1.449,
+      "step": 1790
+    },
+    {
+      "epoch": 0.3717310087173101,
+      "grad_norm": 0.7306489325876637,
+      "learning_rate": 2.905872122342062e-07,
+      "loss": 1.5153,
+      "step": 1791
+    },
+    {
+      "epoch": 0.37193856371938566,
+      "grad_norm": 0.6744167899478379,
+      "learning_rate": 2.904792288594036e-07,
+      "loss": 1.5627,
+      "step": 1792
+    },
+    {
+      "epoch": 0.3721461187214612,
+      "grad_norm": 0.9241052132026019,
+      "learning_rate": 2.9037121551870406e-07,
+      "loss": 1.4503,
+      "step": 1793
+    },
+    {
+      "epoch": 0.3723536737235367,
+      "grad_norm": 0.6830456692624526,
+      "learning_rate": 2.902631722580319e-07,
+      "loss": 1.5196,
+      "step": 1794
+    },
+    {
+      "epoch": 0.3725612287256123,
+      "grad_norm": 0.7699863146828911,
+      "learning_rate": 2.9015509912332425e-07,
+      "loss": 1.4719,
+      "step": 1795
+    },
+    {
+      "epoch": 0.37276878372768785,
+      "grad_norm": 0.9407019799499728,
+      "learning_rate": 2.9004699616053094e-07,
+      "loss": 1.5128,
+      "step": 1796
+    },
+    {
+      "epoch": 0.3729763387297634,
+      "grad_norm": 0.7331956559267344,
+      "learning_rate": 2.899388634156146e-07,
+      "loss": 1.5367,
+      "step": 1797
+    },
+    {
+      "epoch": 0.3731838937318389,
+      "grad_norm": 1.5790529025933957,
+      "learning_rate": 2.8983070093455024e-07,
+      "loss": 1.475,
+      "step": 1798
+    },
+    {
+      "epoch": 0.3733914487339145,
+      "grad_norm": 0.9006149220411652,
+      "learning_rate": 2.8972250876332573e-07,
+      "loss": 1.5965,
+      "step": 1799
+    },
+    {
+      "epoch": 0.37359900373599003,
+      "grad_norm": 1.3755555683212195,
+      "learning_rate": 2.8961428694794156e-07,
+      "loss": 1.535,
+      "step": 1800
+    },
+    {
+      "epoch": 0.37380655873806556,
+      "grad_norm": 0.7353782961824376,
+      "learning_rate": 2.8950603553441073e-07,
+      "loss": 1.417,
+      "step": 1801
+    },
+    {
+      "epoch": 0.37401411374014115,
+      "grad_norm": 1.0600416247613942,
+      "learning_rate": 2.893977545687589e-07,
+      "loss": 1.54,
+      "step": 1802
+    },
+    {
+      "epoch": 0.3742216687422167,
+      "grad_norm": 0.8104960875457892,
+      "learning_rate": 2.8928944409702414e-07,
+      "loss": 1.5143,
+      "step": 1803
+    },
+    {
+      "epoch": 0.3744292237442922,
+      "grad_norm": 0.747597815078873,
+      "learning_rate": 2.891811041652574e-07,
+      "loss": 1.5215,
+      "step": 1804
+    },
+    {
+      "epoch": 0.3746367787463678,
+      "grad_norm": 2.5472919710932693,
+      "learning_rate": 2.890727348195217e-07,
+      "loss": 1.5684,
+      "step": 1805
+    },
+    {
+      "epoch": 0.37484433374844334,
+      "grad_norm": 0.6653594327993576,
+      "learning_rate": 2.88964336105893e-07,
+      "loss": 1.4674,
+      "step": 1806
+    },
+    {
+      "epoch": 0.37505188875051887,
+      "grad_norm": 0.7682038926295779,
+      "learning_rate": 2.888559080704595e-07,
+      "loss": 1.5117,
+      "step": 1807
+    },
+    {
+      "epoch": 0.37525944375259446,
+      "grad_norm": 0.7781683832905401,
+      "learning_rate": 2.8874745075932184e-07,
+      "loss": 1.5887,
+      "step": 1808
+    },
+    {
+      "epoch": 0.37546699875467,
+      "grad_norm": 0.7733843964903611,
+      "learning_rate": 2.886389642185934e-07,
+      "loss": 1.6289,
+      "step": 1809
+    },
+    {
+      "epoch": 0.3756745537567455,
+      "grad_norm": 0.7425122137061511,
+      "learning_rate": 2.8853044849439946e-07,
+      "loss": 1.546,
+      "step": 1810
+    },
+    {
+      "epoch": 0.3758821087588211,
+      "grad_norm": 0.7630938308509255,
+      "learning_rate": 2.884219036328783e-07,
+      "loss": 1.5082,
+      "step": 1811
+    },
+    {
+      "epoch": 0.37608966376089664,
+      "grad_norm": 0.7544870434777171,
+      "learning_rate": 2.883133296801802e-07,
+      "loss": 1.51,
+      "step": 1812
+    },
+    {
+      "epoch": 0.3762972187629722,
+      "grad_norm": 0.701721632843437,
+      "learning_rate": 2.8820472668246794e-07,
+      "loss": 1.5119,
+      "step": 1813
+    },
+    {
+      "epoch": 0.37650477376504776,
+      "grad_norm": 0.8288652246980378,
+      "learning_rate": 2.880960946859166e-07,
+      "loss": 1.4584,
+      "step": 1814
+    },
+    {
+      "epoch": 0.3767123287671233,
+      "grad_norm": 0.6831096352226285,
+      "learning_rate": 2.8798743373671366e-07,
+      "loss": 1.5326,
+      "step": 1815
+    },
+    {
+      "epoch": 0.3769198837691988,
+      "grad_norm": 0.7037553431320874,
+      "learning_rate": 2.878787438810589e-07,
+      "loss": 1.5215,
+      "step": 1816
+    },
+    {
+      "epoch": 0.3771274387712744,
+      "grad_norm": 0.9937775764151584,
+      "learning_rate": 2.8777002516516425e-07,
+      "loss": 1.4809,
+      "step": 1817
+    },
+    {
+      "epoch": 0.37733499377334995,
+      "grad_norm": 0.8753637791110365,
+      "learning_rate": 2.8766127763525417e-07,
+      "loss": 1.4619,
+      "step": 1818
+    },
+    {
+      "epoch": 0.3775425487754255,
+      "grad_norm": 1.381331686046734,
+      "learning_rate": 2.875525013375651e-07,
+      "loss": 1.4532,
+      "step": 1819
+    },
+    {
+      "epoch": 0.377750103777501,
+      "grad_norm": 5.828872507809322,
+      "learning_rate": 2.874436963183459e-07,
+      "loss": 1.4929,
+      "step": 1820
+    },
+    {
+      "epoch": 0.3779576587795766,
+      "grad_norm": 0.8000148440892989,
+      "learning_rate": 2.873348626238575e-07,
+      "loss": 1.5261,
+      "step": 1821
+    },
+    {
+      "epoch": 0.37816521378165213,
+      "grad_norm": 1.4545838082783766,
+      "learning_rate": 2.8722600030037314e-07,
+      "loss": 1.5992,
+      "step": 1822
+    },
+    {
+      "epoch": 0.37837276878372766,
+      "grad_norm": 0.7454331605890411,
+      "learning_rate": 2.871171093941782e-07,
+      "loss": 1.4785,
+      "step": 1823
+    },
+    {
+      "epoch": 0.37858032378580325,
+      "grad_norm": 0.8913187537978086,
+      "learning_rate": 2.870081899515703e-07,
+      "loss": 1.5635,
+      "step": 1824
+    },
+    {
+      "epoch": 0.3787878787878788,
+      "grad_norm": 2.1580098516173813,
+      "learning_rate": 2.8689924201885894e-07,
+      "loss": 1.4963,
+      "step": 1825
+    },
+    {
+      "epoch": 0.3789954337899543,
+      "grad_norm": 1.2800966783503454,
+      "learning_rate": 2.8679026564236596e-07,
+      "loss": 1.5661,
+      "step": 1826
+    },
+    {
+      "epoch": 0.3792029887920299,
+      "grad_norm": 0.6756002911824516,
+      "learning_rate": 2.8668126086842523e-07,
+      "loss": 1.4652,
+      "step": 1827
+    },
+    {
+      "epoch": 0.37941054379410544,
+      "grad_norm": 1.3902061654048554,
+      "learning_rate": 2.865722277433826e-07,
+      "loss": 1.552,
+      "step": 1828
+    },
+    {
+      "epoch": 0.37961809879618097,
+      "grad_norm": 3.58459254832797,
+      "learning_rate": 2.864631663135962e-07,
+      "loss": 1.5641,
+      "step": 1829
+    },
+    {
+      "epoch": 0.37982565379825656,
+      "grad_norm": 0.6543936366061491,
+      "learning_rate": 2.863540766254359e-07,
+      "loss": 1.5607,
+      "step": 1830
+    },
+    {
+      "epoch": 0.3800332088003321,
+      "grad_norm": 0.6606988767087205,
+      "learning_rate": 2.862449587252839e-07,
+      "loss": 1.5429,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3802407638024076,
+      "grad_norm": 1.3028334828837254,
+      "learning_rate": 2.861358126595341e-07,
+      "loss": 1.5575,
+      "step": 1832
+    },
+    {
+      "epoch": 0.3804483188044832,
+      "grad_norm": 0.878861104681702,
+      "learning_rate": 2.860266384745925e-07,
+      "loss": 1.5603,
+      "step": 1833
+    },
+    {
+      "epoch": 0.38065587380655874,
+      "grad_norm": 0.7172936096231933,
+      "learning_rate": 2.859174362168773e-07,
+      "loss": 1.5112,
+      "step": 1834
+    },
+    {
+      "epoch": 0.3808634288086343,
+      "grad_norm": 0.672063637043049,
+      "learning_rate": 2.8580820593281816e-07,
+      "loss": 1.5849,
+      "step": 1835
+    },
+    {
+      "epoch": 0.38107098381070986,
+      "grad_norm": 0.7124377997873157,
+      "learning_rate": 2.8569894766885694e-07,
+      "loss": 1.5265,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3812785388127854,
+      "grad_norm": 0.7413567235389821,
+      "learning_rate": 2.8558966147144736e-07,
+      "loss": 1.5913,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3814860938148609,
+      "grad_norm": 0.9272380461197702,
+      "learning_rate": 2.8548034738705507e-07,
+      "loss": 1.5424,
+      "step": 1838
+    },
+    {
+      "epoch": 0.3816936488169365,
+      "grad_norm": 1.8804828007492838,
+      "learning_rate": 2.853710054621574e-07,
+      "loss": 1.5533,
+      "step": 1839
+    },
+    {
+      "epoch": 0.38190120381901205,
+      "grad_norm": 0.7636573500978951,
+      "learning_rate": 2.852616357432438e-07,
+      "loss": 1.5569,
+      "step": 1840
+    },
+    {
+      "epoch": 0.3821087588210876,
+      "grad_norm": 1.1772095457271807,
+      "learning_rate": 2.851522382768153e-07,
+      "loss": 1.592,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3823163138231631,
+      "grad_norm": 0.7493612440366139,
+      "learning_rate": 2.8504281310938467e-07,
+      "loss": 1.5642,
+      "step": 1842
+    },
+    {
+      "epoch": 0.3825238688252387,
+      "grad_norm": 0.9451664195050361,
+      "learning_rate": 2.849333602874768e-07,
+      "loss": 1.5107,
+      "step": 1843
+    },
+    {
+      "epoch": 0.38273142382731423,
+      "grad_norm": 0.6392717217497093,
+      "learning_rate": 2.848238798576279e-07,
+      "loss": 1.5097,
+      "step": 1844
+    },
+    {
+      "epoch": 0.38293897882938976,
+      "grad_norm": 1.1872601313715576,
+      "learning_rate": 2.8471437186638637e-07,
+      "loss": 1.4252,
+      "step": 1845
+    },
+    {
+      "epoch": 0.38314653383146535,
+      "grad_norm": 0.6530782862274989,
+      "learning_rate": 2.846048363603119e-07,
+      "loss": 1.4733,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3833540888335409,
+      "grad_norm": 0.8818869721313423,
+      "learning_rate": 2.844952733859763e-07,
+      "loss": 1.6101,
+      "step": 1847
+    },
+    {
+      "epoch": 0.3835616438356164,
+      "grad_norm": 1.9324674517805251,
+      "learning_rate": 2.8438568298996264e-07,
+      "loss": 1.5288,
+      "step": 1848
+    },
+    {
+      "epoch": 0.383769198837692,
+      "grad_norm": 0.6803426081958092,
+      "learning_rate": 2.842760652188658e-07,
+      "loss": 1.5297,
+      "step": 1849
+    },
+    {
+      "epoch": 0.38397675383976754,
+      "grad_norm": 0.7028778133935468,
+      "learning_rate": 2.841664201192926e-07,
+      "loss": 1.5611,
+      "step": 1850
+    },
+    {
+      "epoch": 0.38418430884184307,
+      "grad_norm": 0.6692229665364089,
+      "learning_rate": 2.84056747737861e-07,
+      "loss": 1.4909,
+      "step": 1851
+    },
+    {
+      "epoch": 0.38439186384391866,
+      "grad_norm": 0.83404468757173,
+      "learning_rate": 2.83947048121201e-07,
+      "loss": 1.5241,
+      "step": 1852
+    },
+    {
+      "epoch": 0.3845994188459942,
+      "grad_norm": 0.7449454132251969,
+      "learning_rate": 2.838373213159537e-07,
+      "loss": 1.5242,
+      "step": 1853
+    },
+    {
+      "epoch": 0.3848069738480697,
+      "grad_norm": 0.7572259514250277,
+      "learning_rate": 2.8372756736877223e-07,
+      "loss": 1.577,
+      "step": 1854
+    },
+    {
+      "epoch": 0.3850145288501453,
+      "grad_norm": 0.7814337178458187,
+      "learning_rate": 2.83617786326321e-07,
+      "loss": 1.5791,
+      "step": 1855
+    },
+    {
+      "epoch": 0.38522208385222084,
+      "grad_norm": 0.9747877083420584,
+      "learning_rate": 2.8350797823527595e-07,
+      "loss": 1.5588,
+      "step": 1856
+    },
+    {
+      "epoch": 0.3854296388542964,
+      "grad_norm": 1.0560047301941708,
+      "learning_rate": 2.8339814314232467e-07,
+      "loss": 1.4953,
+      "step": 1857
+    },
+    {
+      "epoch": 0.38563719385637196,
+      "grad_norm": 1.0473786774544744,
+      "learning_rate": 2.832882810941659e-07,
+      "loss": 1.5532,
+      "step": 1858
+    },
+    {
+      "epoch": 0.3858447488584475,
+      "grad_norm": 0.967706897548344,
+      "learning_rate": 2.8317839213751036e-07,
+      "loss": 1.559,
+      "step": 1859
+    },
+    {
+      "epoch": 0.386052303860523,
+      "grad_norm": 1.1212019870201413,
+      "learning_rate": 2.830684763190797e-07,
+      "loss": 1.4703,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3862598588625986,
+      "grad_norm": 0.8705750913419449,
+      "learning_rate": 2.829585336856073e-07,
+      "loss": 1.5377,
+      "step": 1861
+    },
+    {
+      "epoch": 0.38646741386467415,
+      "grad_norm": 0.8890857236355498,
+      "learning_rate": 2.8284856428383783e-07,
+      "loss": 1.4978,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3866749688667497,
+      "grad_norm": 0.7373074560343306,
+      "learning_rate": 2.827385681605273e-07,
+      "loss": 1.5216,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3868825238688252,
+      "grad_norm": 0.7450661492329643,
+      "learning_rate": 2.8262854536244333e-07,
+      "loss": 1.4973,
+      "step": 1864
+    },
+    {
+      "epoch": 0.3870900788709008,
+      "grad_norm": 0.6472947892747584,
+      "learning_rate": 2.8251849593636444e-07,
+      "loss": 1.5811,
+      "step": 1865
+    },
+    {
+      "epoch": 0.38729763387297633,
+      "grad_norm": 0.8515492354658755,
+      "learning_rate": 2.8240841992908093e-07,
+      "loss": 1.5378,
+      "step": 1866
+    },
+    {
+      "epoch": 0.38750518887505186,
+      "grad_norm": 1.3033141826532952,
+      "learning_rate": 2.822983173873941e-07,
+      "loss": 1.5013,
+      "step": 1867
+    },
+    {
+      "epoch": 0.38771274387712745,
+      "grad_norm": 0.9766285733196798,
+      "learning_rate": 2.8218818835811664e-07,
+      "loss": 1.6529,
+      "step": 1868
+    },
+    {
+      "epoch": 0.387920298879203,
+      "grad_norm": 0.6403998153684116,
+      "learning_rate": 2.820780328880725e-07,
+      "loss": 1.583,
+      "step": 1869
+    },
+    {
+      "epoch": 0.3881278538812785,
+      "grad_norm": 0.773413876447274,
+      "learning_rate": 2.8196785102409683e-07,
+      "loss": 1.4987,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3883354088833541,
+      "grad_norm": 1.7284876751288045,
+      "learning_rate": 2.81857642813036e-07,
+      "loss": 1.5257,
+      "step": 1871
+    },
+    {
+      "epoch": 0.38854296388542964,
+      "grad_norm": 0.7333806711353849,
+      "learning_rate": 2.8174740830174777e-07,
+      "loss": 1.5563,
+      "step": 1872
+    },
+    {
+      "epoch": 0.38875051888750517,
+      "grad_norm": 1.001070885952592,
+      "learning_rate": 2.8163714753710084e-07,
+      "loss": 1.5177,
+      "step": 1873
+    },
+    {
+      "epoch": 0.38895807388958076,
+      "grad_norm": 0.8328437223603238,
+      "learning_rate": 2.815268605659751e-07,
+      "loss": 1.5247,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3891656288916563,
+      "grad_norm": 0.7583358914287197,
+      "learning_rate": 2.814165474352617e-07,
+      "loss": 1.4546,
+      "step": 1875
+    },
+    {
+      "epoch": 0.3893731838937318,
+      "grad_norm": 0.7124746928026011,
+      "learning_rate": 2.8130620819186284e-07,
+      "loss": 1.507,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3895807388958074,
+      "grad_norm": 0.9292279294119694,
+      "learning_rate": 2.811958428826918e-07,
+      "loss": 1.492,
+      "step": 1877
+    },
+    {
+      "epoch": 0.38978829389788294,
+      "grad_norm": 0.665360962298018,
+      "learning_rate": 2.810854515546731e-07,
+      "loss": 1.5523,
+      "step": 1878
+    },
+    {
+      "epoch": 0.3899958488999585,
+      "grad_norm": 0.9000411119834659,
+      "learning_rate": 2.8097503425474215e-07,
+      "loss": 1.5552,
+      "step": 1879
+    },
+    {
+      "epoch": 0.39020340390203406,
+      "grad_norm": 0.8156011417190381,
+      "learning_rate": 2.808645910298454e-07,
+      "loss": 1.5059,
+      "step": 1880
+    },
+    {
+      "epoch": 0.3904109589041096,
+      "grad_norm": 0.6787718400614382,
+      "learning_rate": 2.807541219269404e-07,
+      "loss": 1.4733,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3906185139061851,
+      "grad_norm": 0.7672358766353656,
+      "learning_rate": 2.8064362699299565e-07,
+      "loss": 1.5063,
+      "step": 1882
+    },
+    {
+      "epoch": 0.3908260689082607,
+      "grad_norm": 0.6275728138528871,
+      "learning_rate": 2.805331062749907e-07,
+      "loss": 1.5565,
+      "step": 1883
+    },
+    {
+      "epoch": 0.39103362391033625,
+      "grad_norm": 0.9776960823012801,
+      "learning_rate": 2.8042255981991607e-07,
+      "loss": 1.5652,
+      "step": 1884
+    },
+    {
+      "epoch": 0.3912411789124118,
+      "grad_norm": 0.6852027388058879,
+      "learning_rate": 2.8031198767477314e-07,
+      "loss": 1.5819,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3914487339144873,
+      "grad_norm": 0.9193455622299851,
+      "learning_rate": 2.8020138988657424e-07,
+      "loss": 1.5018,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3916562889165629,
+      "grad_norm": 0.777764201613916,
+      "learning_rate": 2.800907665023426e-07,
+      "loss": 1.5728,
+      "step": 1887
+    },
+    {
+      "epoch": 0.39186384391863843,
+      "grad_norm": 0.7520658019294972,
+      "learning_rate": 2.7998011756911233e-07,
+      "loss": 1.5694,
+      "step": 1888
+    },
+    {
+      "epoch": 0.39207139892071396,
+      "grad_norm": 0.6706414072368084,
+      "learning_rate": 2.798694431339285e-07,
+      "loss": 1.5159,
+      "step": 1889
+    },
+    {
+      "epoch": 0.39227895392278955,
+      "grad_norm": 0.9576753099157093,
+      "learning_rate": 2.797587432438468e-07,
+      "loss": 1.5203,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3924865089248651,
+      "grad_norm": 0.7720234725496633,
+      "learning_rate": 2.796480179459341e-07,
+      "loss": 1.4727,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3926940639269406,
+      "grad_norm": 1.0040258258328383,
+      "learning_rate": 2.795372672872677e-07,
+      "loss": 1.5072,
+      "step": 1892
+    },
+    {
+      "epoch": 0.3929016189290162,
+      "grad_norm": 0.8197918622449006,
+      "learning_rate": 2.7942649131493583e-07,
+      "loss": 1.5386,
+      "step": 1893
+    },
+    {
+      "epoch": 0.39310917393109174,
+      "grad_norm": 0.8466677779127744,
+      "learning_rate": 2.793156900760376e-07,
+      "loss": 1.5824,
+      "step": 1894
+    },
+    {
+      "epoch": 0.39331672893316727,
+      "grad_norm": 0.9189457018412223,
+      "learning_rate": 2.7920486361768265e-07,
+      "loss": 1.4849,
+      "step": 1895
+    },
+    {
+      "epoch": 0.39352428393524286,
+      "grad_norm": 1.8072961179005242,
+      "learning_rate": 2.7909401198699147e-07,
+      "loss": 1.5832,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3937318389373184,
+      "grad_norm": 1.3560238159837612,
+      "learning_rate": 2.789831352310953e-07,
+      "loss": 1.5138,
+      "step": 1897
+    },
+    {
+      "epoch": 0.3939393939393939,
+      "grad_norm": 1.2251198407633839,
+      "learning_rate": 2.788722333971359e-07,
+      "loss": 1.5635,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3941469489414695,
+      "grad_norm": 0.7103893180844721,
+      "learning_rate": 2.7876130653226583e-07,
+      "loss": 1.5849,
+      "step": 1899
+    },
+    {
+      "epoch": 0.39435450394354504,
+      "grad_norm": 0.6425399457519182,
+      "learning_rate": 2.786503546836482e-07,
+      "loss": 1.5406,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3945620589456206,
+      "grad_norm": 0.8424368335860033,
+      "learning_rate": 2.7853937789845703e-07,
+      "loss": 1.5178,
+      "step": 1901
+    },
+    {
+      "epoch": 0.39476961394769616,
+      "grad_norm": 0.7443914766213149,
+      "learning_rate": 2.7842837622387634e-07,
+      "loss": 1.559,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3949771689497717,
+      "grad_norm": 0.9345212891903224,
+      "learning_rate": 2.7831734970710124e-07,
+      "loss": 1.5163,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3951847239518472,
+      "grad_norm": 1.101698639321631,
+      "learning_rate": 2.7820629839533735e-07,
+      "loss": 1.5418,
+      "step": 1904
+    },
+    {
+      "epoch": 0.3953922789539228,
+      "grad_norm": 0.7369936160309475,
+      "learning_rate": 2.7809522233580067e-07,
+      "loss": 1.5049,
+      "step": 1905
+    },
+    {
+      "epoch": 0.39559983395599835,
+      "grad_norm": 0.82063854054892,
+      "learning_rate": 2.779841215757178e-07,
+      "loss": 1.476,
+      "step": 1906
+    },
+    {
+      "epoch": 0.3958073889580739,
+      "grad_norm": 1.2804337591301285,
+      "learning_rate": 2.7787299616232587e-07,
+      "loss": 1.5276,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3960149439601494,
+      "grad_norm": 0.7294762129030232,
+      "learning_rate": 2.777618461428723e-07,
+      "loss": 1.5479,
+      "step": 1908
+    },
+    {
+      "epoch": 0.396222498962225,
+      "grad_norm": 0.6339456556839771,
+      "learning_rate": 2.7765067156461533e-07,
+      "loss": 1.5121,
+      "step": 1909
+    },
+    {
+      "epoch": 0.39643005396430053,
+      "grad_norm": 0.9371184268948276,
+      "learning_rate": 2.775394724748233e-07,
+      "loss": 1.568,
+      "step": 1910
+    },
+    {
+      "epoch": 0.39663760896637607,
+      "grad_norm": 1.1074910091988812,
+      "learning_rate": 2.774282489207752e-07,
+      "loss": 1.5305,
+      "step": 1911
+    },
+    {
+      "epoch": 0.39684516396845165,
+      "grad_norm": 1.3944087746190492,
+      "learning_rate": 2.773170009497602e-07,
+      "loss": 1.498,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3970527189705272,
+      "grad_norm": 0.7427476163230735,
+      "learning_rate": 2.7720572860907825e-07,
+      "loss": 1.546,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3972602739726027,
+      "grad_norm": 0.7726285525939837,
+      "learning_rate": 2.770944319460391e-07,
+      "loss": 1.5231,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3974678289746783,
+      "grad_norm": 0.798868942757301,
+      "learning_rate": 2.769831110079632e-07,
+      "loss": 1.5249,
+      "step": 1915
+    },
+    {
+      "epoch": 0.39767538397675384,
+      "grad_norm": 1.0416876829616981,
+      "learning_rate": 2.7687176584218137e-07,
+      "loss": 1.507,
+      "step": 1916
+    },
+    {
+      "epoch": 0.39788293897882937,
+      "grad_norm": 0.9686546628175399,
+      "learning_rate": 2.767603964960346e-07,
+      "loss": 1.503,
+      "step": 1917
+    },
+    {
+      "epoch": 0.39809049398090496,
+      "grad_norm": 0.8320277908681334,
+      "learning_rate": 2.766490030168742e-07,
+      "loss": 1.5501,
+      "step": 1918
+    },
+    {
+      "epoch": 0.3982980489829805,
+      "grad_norm": 3.0362071168980322,
+      "learning_rate": 2.765375854520616e-07,
+      "loss": 1.5429,
+      "step": 1919
+    },
+    {
+      "epoch": 0.398505603985056,
+      "grad_norm": 2.383344141689591,
+      "learning_rate": 2.764261438489686e-07,
+      "loss": 1.4693,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3987131589871316,
+      "grad_norm": 0.9330224023721052,
+      "learning_rate": 2.763146782549773e-07,
+      "loss": 1.5399,
+      "step": 1921
+    },
+    {
+      "epoch": 0.39892071398920714,
+      "grad_norm": 0.6333921145130824,
+      "learning_rate": 2.7620318871747986e-07,
+      "loss": 1.4676,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3991282689912827,
+      "grad_norm": 0.7247784637507091,
+      "learning_rate": 2.7609167528387877e-07,
+      "loss": 1.6343,
+      "step": 1923
+    },
+    {
+      "epoch": 0.39933582399335826,
+      "grad_norm": 0.7223408769449733,
+      "learning_rate": 2.7598013800158637e-07,
+      "loss": 1.531,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3995433789954338,
+      "grad_norm": 0.9010314709398364,
+      "learning_rate": 2.758685769180256e-07,
+      "loss": 1.599,
+      "step": 1925
+    },
+    {
+      "epoch": 0.39975093399750933,
+      "grad_norm": 0.6944061389856688,
+      "learning_rate": 2.7575699208062914e-07,
+      "loss": 1.4845,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3999584889995849,
+      "grad_norm": 0.8364905591527232,
+      "learning_rate": 2.7564538353683984e-07,
+      "loss": 1.553,
+      "step": 1927
+    },
+    {
+      "epoch": 0.40016604400166045,
+      "grad_norm": 0.7529657059424397,
+      "learning_rate": 2.7553375133411075e-07,
+      "loss": 1.5283,
+      "step": 1928
+    },
+    {
+      "epoch": 0.400373599003736,
+      "grad_norm": 0.7190180928510772,
+      "learning_rate": 2.7542209551990495e-07,
+      "loss": 1.5077,
+      "step": 1929
+    },
+    {
+      "epoch": 0.4005811540058115,
+      "grad_norm": 0.6925289951992526,
+      "learning_rate": 2.7531041614169556e-07,
+      "loss": 1.5486,
+      "step": 1930
+    },
+    {
+      "epoch": 0.4007887090078871,
+      "grad_norm": 0.7915887904090121,
+      "learning_rate": 2.751987132469656e-07,
+      "loss": 1.5124,
+      "step": 1931
+    },
+    {
+      "epoch": 0.40099626400996263,
+      "grad_norm": 0.7627898228572834,
+      "learning_rate": 2.750869868832082e-07,
+      "loss": 1.5642,
+      "step": 1932
+    },
+    {
+      "epoch": 0.40120381901203817,
+      "grad_norm": 2.3036117094911464,
+      "learning_rate": 2.7497523709792656e-07,
+      "loss": 1.5635,
+      "step": 1933
+    },
+    {
+      "epoch": 0.40141137401411375,
+      "grad_norm": 1.1773529869650565,
+      "learning_rate": 2.7486346393863345e-07,
+      "loss": 1.5154,
+      "step": 1934
+    },
+    {
+      "epoch": 0.4016189290161893,
+      "grad_norm": 0.9991890160932483,
+      "learning_rate": 2.747516674528521e-07,
+      "loss": 1.5523,
+      "step": 1935
+    },
+    {
+      "epoch": 0.4018264840182648,
+      "grad_norm": 0.7512122792523664,
+      "learning_rate": 2.7463984768811533e-07,
+      "loss": 1.5325,
+      "step": 1936
+    },
+    {
+      "epoch": 0.4020340390203404,
+      "grad_norm": 0.8008766142155629,
+      "learning_rate": 2.745280046919659e-07,
+      "loss": 1.5534,
+      "step": 1937
+    },
+    {
+      "epoch": 0.40224159402241594,
+      "grad_norm": 0.939625216040225,
+      "learning_rate": 2.7441613851195657e-07,
+      "loss": 1.5254,
+      "step": 1938
+    },
+    {
+      "epoch": 0.40244914902449147,
+      "grad_norm": 0.9794917016143329,
+      "learning_rate": 2.7430424919564976e-07,
+      "loss": 1.4973,
+      "step": 1939
+    },
+    {
+      "epoch": 0.40265670402656706,
+      "grad_norm": 0.9013638156949378,
+      "learning_rate": 2.7419233679061785e-07,
+      "loss": 1.475,
+      "step": 1940
+    },
+    {
+      "epoch": 0.4028642590286426,
+      "grad_norm": 0.7953984772173054,
+      "learning_rate": 2.740804013444431e-07,
+      "loss": 1.4647,
+      "step": 1941
+    },
+    {
+      "epoch": 0.4030718140307181,
+      "grad_norm": 0.7251998686131645,
+      "learning_rate": 2.7396844290471745e-07,
+      "loss": 1.5457,
+      "step": 1942
+    },
+    {
+      "epoch": 0.4032793690327937,
+      "grad_norm": 0.6345492526265151,
+      "learning_rate": 2.7385646151904264e-07,
+      "loss": 1.4983,
+      "step": 1943
+    },
+    {
+      "epoch": 0.40348692403486924,
+      "grad_norm": 1.1620281048635444,
+      "learning_rate": 2.7374445723503024e-07,
+      "loss": 1.5516,
+      "step": 1944
+    },
+    {
+      "epoch": 0.4036944790369448,
+      "grad_norm": 0.8362171382145378,
+      "learning_rate": 2.7363243010030143e-07,
+      "loss": 1.5405,
+      "step": 1945
+    },
+    {
+      "epoch": 0.40390203403902036,
+      "grad_norm": 1.0620848685033106,
+      "learning_rate": 2.735203801624872e-07,
+      "loss": 1.5088,
+      "step": 1946
+    },
+    {
+      "epoch": 0.4041095890410959,
+      "grad_norm": 0.8707960728684162,
+      "learning_rate": 2.7340830746922826e-07,
+      "loss": 1.5262,
+      "step": 1947
+    },
+    {
+      "epoch": 0.40431714404317143,
+      "grad_norm": 1.071232075744944,
+      "learning_rate": 2.7329621206817484e-07,
+      "loss": 1.4723,
+      "step": 1948
+    },
+    {
+      "epoch": 0.404524699045247,
+      "grad_norm": 0.7257287686710381,
+      "learning_rate": 2.7318409400698695e-07,
+      "loss": 1.5015,
+      "step": 1949
+    },
+    {
+      "epoch": 0.40473225404732255,
+      "grad_norm": 0.8668421279176032,
+      "learning_rate": 2.7307195333333434e-07,
+      "loss": 1.5108,
+      "step": 1950
+    },
+    {
+      "epoch": 0.4049398090493981,
+      "grad_norm": 0.7627042084590963,
+      "learning_rate": 2.7295979009489613e-07,
+      "loss": 1.5181,
+      "step": 1951
+    },
+    {
+      "epoch": 0.4051473640514736,
+      "grad_norm": 0.8092281435814975,
+      "learning_rate": 2.7284760433936116e-07,
+      "loss": 1.5521,
+      "step": 1952
+    },
+    {
+      "epoch": 0.4053549190535492,
+      "grad_norm": 1.0549612633542718,
+      "learning_rate": 2.727353961144278e-07,
+      "loss": 1.4943,
+      "step": 1953
+    },
+    {
+      "epoch": 0.40556247405562473,
+      "grad_norm": 0.8105043827481816,
+      "learning_rate": 2.726231654678041e-07,
+      "loss": 1.5111,
+      "step": 1954
+    },
+    {
+      "epoch": 0.40577002905770027,
+      "grad_norm": 0.7174589377144522,
+      "learning_rate": 2.725109124472075e-07,
+      "loss": 1.4832,
+      "step": 1955
+    },
+    {
+      "epoch": 0.40597758405977585,
+      "grad_norm": 0.9245630002901994,
+      "learning_rate": 2.72398637100365e-07,
+      "loss": 1.5428,
+      "step": 1956
+    },
+    {
+      "epoch": 0.4061851390618514,
+      "grad_norm": 0.9163491355752184,
+      "learning_rate": 2.7228633947501313e-07,
+      "loss": 1.5254,
+      "step": 1957
+    },
+    {
+      "epoch": 0.4063926940639269,
+      "grad_norm": 1.108975707723247,
+      "learning_rate": 2.721740196188978e-07,
+      "loss": 1.5512,
+      "step": 1958
+    },
+    {
+      "epoch": 0.4066002490660025,
+      "grad_norm": 0.7944624451383683,
+      "learning_rate": 2.7206167757977453e-07,
+      "loss": 1.5106,
+      "step": 1959
+    },
+    {
+      "epoch": 0.40680780406807804,
+      "grad_norm": 5.609741949134925,
+      "learning_rate": 2.719493134054081e-07,
+      "loss": 1.5096,
+      "step": 1960
+    },
+    {
+      "epoch": 0.40701535907015357,
+      "grad_norm": 1.1187070928880356,
+      "learning_rate": 2.718369271435728e-07,
+      "loss": 1.5925,
+      "step": 1961
+    },
+    {
+      "epoch": 0.40722291407222916,
+      "grad_norm": 1.0546010614843362,
+      "learning_rate": 2.7172451884205216e-07,
+      "loss": 1.5623,
+      "step": 1962
+    },
+    {
+      "epoch": 0.4074304690743047,
+      "grad_norm": 0.9062508446385186,
+      "learning_rate": 2.716120885486395e-07,
+      "loss": 1.5115,
+      "step": 1963
+    },
+    {
+      "epoch": 0.4076380240763802,
+      "grad_norm": 0.6314903926596144,
+      "learning_rate": 2.714996363111369e-07,
+      "loss": 1.557,
+      "step": 1964
+    },
+    {
+      "epoch": 0.4078455790784558,
+      "grad_norm": 0.795655228469788,
+      "learning_rate": 2.713871621773562e-07,
+      "loss": 1.5546,
+      "step": 1965
+    },
+    {
+      "epoch": 0.40805313408053134,
+      "grad_norm": 1.370501547675247,
+      "learning_rate": 2.712746661951184e-07,
+      "loss": 1.5506,
+      "step": 1966
+    },
+    {
+      "epoch": 0.4082606890826069,
+      "grad_norm": 0.7452116526347807,
+      "learning_rate": 2.7116214841225375e-07,
+      "loss": 1.4974,
+      "step": 1967
+    },
+    {
+      "epoch": 0.40846824408468246,
+      "grad_norm": 0.7842228040721118,
+      "learning_rate": 2.710496088766019e-07,
+      "loss": 1.5392,
+      "step": 1968
+    },
+    {
+      "epoch": 0.408675799086758,
+      "grad_norm": 0.8855829100805783,
+      "learning_rate": 2.709370476360116e-07,
+      "loss": 1.5472,
+      "step": 1969
+    },
+    {
+      "epoch": 0.40888335408883353,
+      "grad_norm": 0.7929509687283204,
+      "learning_rate": 2.70824464738341e-07,
+      "loss": 1.4534,
+      "step": 1970
+    },
+    {
+      "epoch": 0.4090909090909091,
+      "grad_norm": 0.69688109019725,
+      "learning_rate": 2.7071186023145736e-07,
+      "loss": 1.6032,
+      "step": 1971
+    },
+    {
+      "epoch": 0.40929846409298465,
+      "grad_norm": 0.8672389131683208,
+      "learning_rate": 2.7059923416323694e-07,
+      "loss": 1.5137,
+      "step": 1972
+    },
+    {
+      "epoch": 0.4095060190950602,
+      "grad_norm": 2.161778956729785,
+      "learning_rate": 2.704865865815656e-07,
+      "loss": 1.5691,
+      "step": 1973
+    },
+    {
+      "epoch": 0.40971357409713577,
+      "grad_norm": 0.9402531109283854,
+      "learning_rate": 2.70373917534338e-07,
+      "loss": 1.5073,
+      "step": 1974
+    },
+    {
+      "epoch": 0.4099211290992113,
+      "grad_norm": 0.7410762861775534,
+      "learning_rate": 2.7026122706945796e-07,
+      "loss": 1.5349,
+      "step": 1975
+    },
+    {
+      "epoch": 0.41012868410128683,
+      "grad_norm": 0.8531186208854797,
+      "learning_rate": 2.7014851523483854e-07,
+      "loss": 1.478,
+      "step": 1976
+    },
+    {
+      "epoch": 0.41033623910336237,
+      "grad_norm": 1.0683833121668473,
+      "learning_rate": 2.7003578207840185e-07,
+      "loss": 1.553,
+      "step": 1977
+    },
+    {
+      "epoch": 0.41054379410543795,
+      "grad_norm": 0.7647988716687542,
+      "learning_rate": 2.699230276480789e-07,
+      "loss": 1.4995,
+      "step": 1978
+    },
+    {
+      "epoch": 0.4107513491075135,
+      "grad_norm": 0.7178294451724859,
+      "learning_rate": 2.6981025199181e-07,
+      "loss": 1.5437,
+      "step": 1979
+    },
+    {
+      "epoch": 0.410958904109589,
+      "grad_norm": 0.7650489216282447,
+      "learning_rate": 2.6969745515754444e-07,
+      "loss": 1.4938,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4111664591116646,
+      "grad_norm": 0.8511815549654166,
+      "learning_rate": 2.695846371932402e-07,
+      "loss": 1.4929,
+      "step": 1981
+    },
+    {
+      "epoch": 0.41137401411374014,
+      "grad_norm": 0.789166193611733,
+      "learning_rate": 2.694717981468647e-07,
+      "loss": 1.5835,
+      "step": 1982
+    },
+    {
+      "epoch": 0.41158156911581567,
+      "grad_norm": 0.9913750998034185,
+      "learning_rate": 2.69358938066394e-07,
+      "loss": 1.5683,
+      "step": 1983
+    },
+    {
+      "epoch": 0.41178912411789126,
+      "grad_norm": 0.7305718241286214,
+      "learning_rate": 2.692460569998133e-07,
+      "loss": 1.574,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4119966791199668,
+      "grad_norm": 0.7436803249449166,
+      "learning_rate": 2.6913315499511647e-07,
+      "loss": 1.5329,
+      "step": 1985
+    },
+    {
+      "epoch": 0.4122042341220423,
+      "grad_norm": 1.0178942563673272,
+      "learning_rate": 2.690202321003067e-07,
+      "loss": 1.4834,
+      "step": 1986
+    },
+    {
+      "epoch": 0.4124117891241179,
+      "grad_norm": 0.7376959817155728,
+      "learning_rate": 2.6890728836339545e-07,
+      "loss": 1.5433,
+      "step": 1987
+    },
+    {
+      "epoch": 0.41261934412619344,
+      "grad_norm": 0.9555009382855892,
+      "learning_rate": 2.6879432383240376e-07,
+      "loss": 1.5354,
+      "step": 1988
+    },
+    {
+      "epoch": 0.412826899128269,
+      "grad_norm": 0.6662939897421363,
+      "learning_rate": 2.68681338555361e-07,
+      "loss": 1.4966,
+      "step": 1989
+    },
+    {
+      "epoch": 0.41303445413034456,
+      "grad_norm": 0.843215642797735,
+      "learning_rate": 2.6856833258030536e-07,
+      "loss": 1.4844,
+      "step": 1990
+    },
+    {
+      "epoch": 0.4132420091324201,
+      "grad_norm": 0.9991433835316956,
+      "learning_rate": 2.6845530595528426e-07,
+      "loss": 1.5481,
+      "step": 1991
+    },
+    {
+      "epoch": 0.41344956413449563,
+      "grad_norm": 0.8435969999898885,
+      "learning_rate": 2.6834225872835343e-07,
+      "loss": 1.4435,
+      "step": 1992
+    },
+    {
+      "epoch": 0.4136571191365712,
+      "grad_norm": 0.6791292461181895,
+      "learning_rate": 2.682291909475776e-07,
+      "loss": 1.516,
+      "step": 1993
+    },
+    {
+      "epoch": 0.41386467413864675,
+      "grad_norm": 0.9605644876741145,
+      "learning_rate": 2.6811610266103027e-07,
+      "loss": 1.4849,
+      "step": 1994
+    },
+    {
+      "epoch": 0.4140722291407223,
+      "grad_norm": 1.0086035047298292,
+      "learning_rate": 2.680029939167934e-07,
+      "loss": 1.4928,
+      "step": 1995
+    },
+    {
+      "epoch": 0.41427978414279787,
+      "grad_norm": 0.8368771415626136,
+      "learning_rate": 2.678898647629579e-07,
+      "loss": 1.5075,
+      "step": 1996
+    },
+    {
+      "epoch": 0.4144873391448734,
+      "grad_norm": 0.8690980104636004,
+      "learning_rate": 2.6777671524762333e-07,
+      "loss": 1.5928,
+      "step": 1997
+    },
+    {
+      "epoch": 0.41469489414694893,
+      "grad_norm": 0.720130372239292,
+      "learning_rate": 2.6766354541889787e-07,
+      "loss": 1.4572,
+      "step": 1998
+    },
+    {
+      "epoch": 0.41490244914902447,
+      "grad_norm": 0.7549670726723419,
+      "learning_rate": 2.6755035532489833e-07,
+      "loss": 1.5779,
+      "step": 1999
+    },
+    {
+      "epoch": 0.41511000415110005,
+      "grad_norm": 0.9098930657658568,
+      "learning_rate": 2.6743714501375003e-07,
+      "loss": 1.5239,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4153175591531756,
+      "grad_norm": 1.2050494538479672,
+      "learning_rate": 2.6732391453358713e-07,
+      "loss": 1.462,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4155251141552511,
+      "grad_norm": 0.8173804395085473,
+      "learning_rate": 2.672106639325521e-07,
+      "loss": 1.5721,
+      "step": 2002
+    },
+    {
+      "epoch": 0.4157326691573267,
+      "grad_norm": 0.7557662067098537,
+      "learning_rate": 2.670973932587961e-07,
+      "loss": 1.5223,
+      "step": 2003
+    },
+    {
+      "epoch": 0.41594022415940224,
+      "grad_norm": 0.8747910090242246,
+      "learning_rate": 2.669841025604789e-07,
+      "loss": 1.5398,
+      "step": 2004
+    },
+    {
+      "epoch": 0.41614777916147777,
+      "grad_norm": 1.1885443368681692,
+      "learning_rate": 2.668707918857687e-07,
+      "loss": 1.4637,
+      "step": 2005
+    },
+    {
+      "epoch": 0.41635533416355336,
+      "grad_norm": 0.8769914767750531,
+      "learning_rate": 2.66757461282842e-07,
+      "loss": 1.5319,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4165628891656289,
+      "grad_norm": 0.789150243313903,
+      "learning_rate": 2.666441107998842e-07,
+      "loss": 1.5372,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4167704441677044,
+      "grad_norm": 0.9155065626204896,
+      "learning_rate": 2.665307404850887e-07,
+      "loss": 1.5586,
+      "step": 2008
+    },
+    {
+      "epoch": 0.41697799916978,
+      "grad_norm": 0.6258097564708011,
+      "learning_rate": 2.664173503866578e-07,
+      "loss": 1.5735,
+      "step": 2009
+    },
+    {
+      "epoch": 0.41718555417185554,
+      "grad_norm": 0.8174737773932771,
+      "learning_rate": 2.6630394055280175e-07,
+      "loss": 1.5316,
+      "step": 2010
+    },
+    {
+      "epoch": 0.4173931091739311,
+      "grad_norm": 0.7038536993377364,
+      "learning_rate": 2.6619051103173957e-07,
+      "loss": 1.4509,
+      "step": 2011
+    },
+    {
+      "epoch": 0.41760066417600666,
+      "grad_norm": 0.7542523517708427,
+      "learning_rate": 2.660770618716983e-07,
+      "loss": 1.5315,
+      "step": 2012
+    },
+    {
+      "epoch": 0.4178082191780822,
+      "grad_norm": 0.8299481166178969,
+      "learning_rate": 2.659635931209137e-07,
+      "loss": 1.5386,
+      "step": 2013
+    },
+    {
+      "epoch": 0.41801577418015773,
+      "grad_norm": 0.6961370689734985,
+      "learning_rate": 2.6585010482762946e-07,
+      "loss": 1.4669,
+      "step": 2014
+    },
+    {
+      "epoch": 0.4182233291822333,
+      "grad_norm": 0.7680902273366674,
+      "learning_rate": 2.6573659704009794e-07,
+      "loss": 1.5235,
+      "step": 2015
+    },
+    {
+      "epoch": 0.41843088418430885,
+      "grad_norm": 0.8476145513507719,
+      "learning_rate": 2.656230698065796e-07,
+      "loss": 1.5182,
+      "step": 2016
+    },
+    {
+      "epoch": 0.4186384391863844,
+      "grad_norm": 0.7733746156412612,
+      "learning_rate": 2.655095231753432e-07,
+      "loss": 1.543,
+      "step": 2017
+    },
+    {
+      "epoch": 0.41884599418845997,
+      "grad_norm": 1.7757191044600509,
+      "learning_rate": 2.653959571946657e-07,
+      "loss": 1.5023,
+      "step": 2018
+    },
+    {
+      "epoch": 0.4190535491905355,
+      "grad_norm": 0.6452839633580992,
+      "learning_rate": 2.652823719128325e-07,
+      "loss": 1.4772,
+      "step": 2019
+    },
+    {
+      "epoch": 0.41926110419261103,
+      "grad_norm": 0.7799227453446673,
+      "learning_rate": 2.6516876737813685e-07,
+      "loss": 1.475,
+      "step": 2020
+    },
+    {
+      "epoch": 0.41946865919468657,
+      "grad_norm": 0.8006026534273608,
+      "learning_rate": 2.6505514363888056e-07,
+      "loss": 1.5116,
+      "step": 2021
+    },
+    {
+      "epoch": 0.41967621419676215,
+      "grad_norm": 1.5066166955921338,
+      "learning_rate": 2.6494150074337324e-07,
+      "loss": 1.5478,
+      "step": 2022
+    },
+    {
+      "epoch": 0.4198837691988377,
+      "grad_norm": 0.7222997155125618,
+      "learning_rate": 2.64827838739933e-07,
+      "loss": 1.4958,
+      "step": 2023
+    },
+    {
+      "epoch": 0.4200913242009132,
+      "grad_norm": 0.8223025333169117,
+      "learning_rate": 2.647141576768858e-07,
+      "loss": 1.494,
+      "step": 2024
+    },
+    {
+      "epoch": 0.4202988792029888,
+      "grad_norm": 0.6656864767856296,
+      "learning_rate": 2.646004576025659e-07,
+      "loss": 1.546,
+      "step": 2025
+    },
+    {
+      "epoch": 0.42050643420506434,
+      "grad_norm": 0.7805902033504355,
+      "learning_rate": 2.6448673856531543e-07,
+      "loss": 1.552,
+      "step": 2026
+    },
+    {
+      "epoch": 0.42071398920713987,
+      "grad_norm": 1.272531855540796,
+      "learning_rate": 2.6437300061348485e-07,
+      "loss": 1.6448,
+      "step": 2027
+    },
+    {
+      "epoch": 0.42092154420921546,
+      "grad_norm": 0.6355549472224202,
+      "learning_rate": 2.642592437954324e-07,
+      "loss": 1.5084,
+      "step": 2028
+    },
+    {
+      "epoch": 0.421129099211291,
+      "grad_norm": 1.3291418867193618,
+      "learning_rate": 2.641454681595246e-07,
+      "loss": 1.5703,
+      "step": 2029
+    },
+    {
+      "epoch": 0.4213366542133665,
+      "grad_norm": 0.789434902757764,
+      "learning_rate": 2.640316737541356e-07,
+      "loss": 1.4814,
+      "step": 2030
+    },
+    {
+      "epoch": 0.4215442092154421,
+      "grad_norm": 1.2506765962226911,
+      "learning_rate": 2.6391786062764794e-07,
+      "loss": 1.4988,
+      "step": 2031
+    },
+    {
+      "epoch": 0.42175176421751764,
+      "grad_norm": 1.1450936888881258,
+      "learning_rate": 2.638040288284519e-07,
+      "loss": 1.6556,
+      "step": 2032
+    },
+    {
+      "epoch": 0.4219593192195932,
+      "grad_norm": 3.6865029122549933,
+      "learning_rate": 2.6369017840494576e-07,
+      "loss": 1.511,
+      "step": 2033
+    },
+    {
+      "epoch": 0.42216687422166876,
+      "grad_norm": 0.8230006227676449,
+      "learning_rate": 2.6357630940553564e-07,
+      "loss": 1.5834,
+      "step": 2034
+    },
+    {
+      "epoch": 0.4223744292237443,
+      "grad_norm": 0.9182976130910437,
+      "learning_rate": 2.634624218786356e-07,
+      "loss": 1.4871,
+      "step": 2035
+    },
+    {
+      "epoch": 0.42258198422581983,
+      "grad_norm": 0.7335896230065206,
+      "learning_rate": 2.633485158726677e-07,
+      "loss": 1.4475,
+      "step": 2036
+    },
+    {
+      "epoch": 0.4227895392278954,
+      "grad_norm": 1.4666424136570828,
+      "learning_rate": 2.632345914360617e-07,
+      "loss": 1.5274,
+      "step": 2037
+    },
+    {
+      "epoch": 0.42299709422997095,
+      "grad_norm": 2.4023424201431913,
+      "learning_rate": 2.6312064861725526e-07,
+      "loss": 1.4725,
+      "step": 2038
+    },
+    {
+      "epoch": 0.4232046492320465,
+      "grad_norm": 0.7697042861056618,
+      "learning_rate": 2.630066874646938e-07,
+      "loss": 1.5226,
+      "step": 2039
+    },
+    {
+      "epoch": 0.42341220423412207,
+      "grad_norm": 0.6845749982480749,
+      "learning_rate": 2.628927080268305e-07,
+      "loss": 1.55,
+      "step": 2040
+    },
+    {
+      "epoch": 0.4236197592361976,
+      "grad_norm": 0.7471716141410968,
+      "learning_rate": 2.6277871035212653e-07,
+      "loss": 1.5378,
+      "step": 2041
+    },
+    {
+      "epoch": 0.42382731423827313,
+      "grad_norm": 0.7208714626204153,
+      "learning_rate": 2.626646944890507e-07,
+      "loss": 1.4736,
+      "step": 2042
+    },
+    {
+      "epoch": 0.42403486924034867,
+      "grad_norm": 3.9840396195222674,
+      "learning_rate": 2.625506604860794e-07,
+      "loss": 1.4753,
+      "step": 2043
+    },
+    {
+      "epoch": 0.42424242424242425,
+      "grad_norm": 1.0022284636716088,
+      "learning_rate": 2.624366083916969e-07,
+      "loss": 1.5241,
+      "step": 2044
+    },
+    {
+      "epoch": 0.4244499792444998,
+      "grad_norm": 0.706265138873214,
+      "learning_rate": 2.6232253825439515e-07,
+      "loss": 1.4645,
+      "step": 2045
+    },
+    {
+      "epoch": 0.4246575342465753,
+      "grad_norm": 0.6291291912454011,
+      "learning_rate": 2.622084501226737e-07,
+      "loss": 1.4578,
+      "step": 2046
+    },
+    {
+      "epoch": 0.4248650892486509,
+      "grad_norm": 0.876011593692041,
+      "learning_rate": 2.6209434404503994e-07,
+      "loss": 1.5483,
+      "step": 2047
+    },
+    {
+      "epoch": 0.42507264425072644,
+      "grad_norm": 1.0912952966883218,
+      "learning_rate": 2.619802200700085e-07,
+      "loss": 1.493,
+      "step": 2048
+    },
+    {
+      "epoch": 0.42528019925280197,
+      "grad_norm": 0.6980180837308734,
+      "learning_rate": 2.618660782461021e-07,
+      "loss": 1.5547,
+      "step": 2049
+    },
+    {
+      "epoch": 0.42548775425487756,
+      "grad_norm": 0.9037102877983608,
+      "learning_rate": 2.6175191862185066e-07,
+      "loss": 1.483,
+      "step": 2050
+    },
+    {
+      "epoch": 0.4256953092569531,
+      "grad_norm": 0.7792476610232986,
+      "learning_rate": 2.616377412457919e-07,
+      "loss": 1.5847,
+      "step": 2051
+    },
+    {
+      "epoch": 0.4259028642590286,
+      "grad_norm": 1.0935208527142122,
+      "learning_rate": 2.61523546166471e-07,
+      "loss": 1.534,
+      "step": 2052
+    },
+    {
+      "epoch": 0.4261104192611042,
+      "grad_norm": 0.7221317998016549,
+      "learning_rate": 2.6140933343244057e-07,
+      "loss": 1.4952,
+      "step": 2053
+    },
+    {
+      "epoch": 0.42631797426317974,
+      "grad_norm": 0.8560585391970406,
+      "learning_rate": 2.61295103092261e-07,
+      "loss": 1.5385,
+      "step": 2054
+    },
+    {
+      "epoch": 0.4265255292652553,
+      "grad_norm": 0.839086834158676,
+      "learning_rate": 2.6118085519449993e-07,
+      "loss": 1.4509,
+      "step": 2055
+    },
+    {
+      "epoch": 0.42673308426733086,
+      "grad_norm": 0.6664727499837579,
+      "learning_rate": 2.6106658978773244e-07,
+      "loss": 1.5664,
+      "step": 2056
+    },
+    {
+      "epoch": 0.4269406392694064,
+      "grad_norm": 0.6553495126242809,
+      "learning_rate": 2.609523069205413e-07,
+      "loss": 1.4837,
+      "step": 2057
+    },
+    {
+      "epoch": 0.42714819427148193,
+      "grad_norm": 0.64964437906978,
+      "learning_rate": 2.608380066415164e-07,
+      "loss": 1.5497,
+      "step": 2058
+    },
+    {
+      "epoch": 0.4273557492735575,
+      "grad_norm": 0.6071201962002736,
+      "learning_rate": 2.6072368899925536e-07,
+      "loss": 1.5691,
+      "step": 2059
+    },
+    {
+      "epoch": 0.42756330427563305,
+      "grad_norm": 1.5713414629083455,
+      "learning_rate": 2.6060935404236286e-07,
+      "loss": 1.4795,
+      "step": 2060
+    },
+    {
+      "epoch": 0.4277708592777086,
+      "grad_norm": 1.1262097489189138,
+      "learning_rate": 2.6049500181945113e-07,
+      "loss": 1.576,
+      "step": 2061
+    },
+    {
+      "epoch": 0.42797841427978417,
+      "grad_norm": 0.79002612241337,
+      "learning_rate": 2.603806323791397e-07,
+      "loss": 1.5142,
+      "step": 2062
+    },
+    {
+      "epoch": 0.4281859692818597,
+      "grad_norm": 1.4422448117375584,
+      "learning_rate": 2.6026624577005546e-07,
+      "loss": 1.5416,
+      "step": 2063
+    },
+    {
+      "epoch": 0.42839352428393523,
+      "grad_norm": 0.7503413597555127,
+      "learning_rate": 2.601518420408325e-07,
+      "loss": 1.5221,
+      "step": 2064
+    },
+    {
+      "epoch": 0.42860107928601077,
+      "grad_norm": 0.7857775711499824,
+      "learning_rate": 2.600374212401123e-07,
+      "loss": 1.55,
+      "step": 2065
+    },
+    {
+      "epoch": 0.42880863428808635,
+      "grad_norm": 0.6273231122356161,
+      "learning_rate": 2.599229834165436e-07,
+      "loss": 1.5003,
+      "step": 2066
+    },
+    {
+      "epoch": 0.4290161892901619,
+      "grad_norm": 1.1792941125455756,
+      "learning_rate": 2.5980852861878213e-07,
+      "loss": 1.4907,
+      "step": 2067
+    },
+    {
+      "epoch": 0.4292237442922374,
+      "grad_norm": 0.7812811257183906,
+      "learning_rate": 2.5969405689549113e-07,
+      "loss": 1.4668,
+      "step": 2068
+    },
+    {
+      "epoch": 0.429431299294313,
+      "grad_norm": 0.7665258415497694,
+      "learning_rate": 2.59579568295341e-07,
+      "loss": 1.5086,
+      "step": 2069
+    },
+    {
+      "epoch": 0.42963885429638854,
+      "grad_norm": 0.6803406900341635,
+      "learning_rate": 2.594650628670092e-07,
+      "loss": 1.5059,
+      "step": 2070
+    },
+    {
+      "epoch": 0.42984640929846407,
+      "grad_norm": 0.9726321494992176,
+      "learning_rate": 2.5935054065918047e-07,
+      "loss": 1.4858,
+      "step": 2071
+    },
+    {
+      "epoch": 0.43005396430053966,
+      "grad_norm": 1.1539561214168268,
+      "learning_rate": 2.5923600172054645e-07,
+      "loss": 1.5302,
+      "step": 2072
+    },
+    {
+      "epoch": 0.4302615193026152,
+      "grad_norm": 0.7474077553400984,
+      "learning_rate": 2.591214460998062e-07,
+      "loss": 1.4986,
+      "step": 2073
+    },
+    {
+      "epoch": 0.4304690743046907,
+      "grad_norm": 0.921469828434731,
+      "learning_rate": 2.5900687384566565e-07,
+      "loss": 1.5212,
+      "step": 2074
+    },
+    {
+      "epoch": 0.4306766293067663,
+      "grad_norm": 1.1254870648032753,
+      "learning_rate": 2.588922850068379e-07,
+      "loss": 1.5289,
+      "step": 2075
+    },
+    {
+      "epoch": 0.43088418430884184,
+      "grad_norm": 2.8821663495470937,
+      "learning_rate": 2.587776796320432e-07,
+      "loss": 1.5083,
+      "step": 2076
+    },
+    {
+      "epoch": 0.4310917393109174,
+      "grad_norm": 0.7817431035285098,
+      "learning_rate": 2.586630577700086e-07,
+      "loss": 1.4893,
+      "step": 2077
+    },
+    {
+      "epoch": 0.43129929431299296,
+      "grad_norm": 1.3322400918914807,
+      "learning_rate": 2.585484194694682e-07,
+      "loss": 1.5307,
+      "step": 2078
+    },
+    {
+      "epoch": 0.4315068493150685,
+      "grad_norm": 0.8341806325564548,
+      "learning_rate": 2.584337647791633e-07,
+      "loss": 1.5426,
+      "step": 2079
+    },
+    {
+      "epoch": 0.43171440431714403,
+      "grad_norm": 0.692326746479079,
+      "learning_rate": 2.5831909374784194e-07,
+      "loss": 1.4693,
+      "step": 2080
+    },
+    {
+      "epoch": 0.4319219593192196,
+      "grad_norm": 0.6304855163810402,
+      "learning_rate": 2.5820440642425923e-07,
+      "loss": 1.5104,
+      "step": 2081
+    },
+    {
+      "epoch": 0.43212951432129515,
+      "grad_norm": 0.6651898945885343,
+      "learning_rate": 2.580897028571772e-07,
+      "loss": 1.5929,
+      "step": 2082
+    },
+    {
+      "epoch": 0.4323370693233707,
+      "grad_norm": 0.9950781603408949,
+      "learning_rate": 2.5797498309536474e-07,
+      "loss": 1.5657,
+      "step": 2083
+    },
+    {
+      "epoch": 0.43254462432544627,
+      "grad_norm": 1.0425445371864661,
+      "learning_rate": 2.5786024718759763e-07,
+      "loss": 1.5491,
+      "step": 2084
+    },
+    {
+      "epoch": 0.4327521793275218,
+      "grad_norm": 0.6945210803612328,
+      "learning_rate": 2.577454951826586e-07,
+      "loss": 1.4589,
+      "step": 2085
+    },
+    {
+      "epoch": 0.43295973432959733,
+      "grad_norm": 1.1443682126659054,
+      "learning_rate": 2.5763072712933706e-07,
+      "loss": 1.4681,
+      "step": 2086
+    },
+    {
+      "epoch": 0.43316728933167287,
+      "grad_norm": 0.6723891645248855,
+      "learning_rate": 2.575159430764293e-07,
+      "loss": 1.537,
+      "step": 2087
+    },
+    {
+      "epoch": 0.43337484433374845,
+      "grad_norm": 0.7076545854764452,
+      "learning_rate": 2.5740114307273867e-07,
+      "loss": 1.5585,
+      "step": 2088
+    },
+    {
+      "epoch": 0.433582399335824,
+      "grad_norm": 0.8414062525776694,
+      "learning_rate": 2.5728632716707493e-07,
+      "loss": 1.5487,
+      "step": 2089
+    },
+    {
+      "epoch": 0.4337899543378995,
+      "grad_norm": 0.6796298457061739,
+      "learning_rate": 2.5717149540825473e-07,
+      "loss": 1.5151,
+      "step": 2090
+    },
+    {
+      "epoch": 0.4339975093399751,
+      "grad_norm": 0.6982565888292679,
+      "learning_rate": 2.570566478451015e-07,
+      "loss": 1.5584,
+      "step": 2091
+    },
+    {
+      "epoch": 0.43420506434205064,
+      "grad_norm": 2.2419501741257237,
+      "learning_rate": 2.5694178452644547e-07,
+      "loss": 1.4493,
+      "step": 2092
+    },
+    {
+      "epoch": 0.43441261934412617,
+      "grad_norm": 1.2497028443646439,
+      "learning_rate": 2.5682690550112343e-07,
+      "loss": 1.5084,
+      "step": 2093
+    },
+    {
+      "epoch": 0.43462017434620176,
+      "grad_norm": 0.6264273778094389,
+      "learning_rate": 2.5671201081797883e-07,
+      "loss": 1.535,
+      "step": 2094
+    },
+    {
+      "epoch": 0.4348277293482773,
+      "grad_norm": 0.7809303293601078,
+      "learning_rate": 2.5659710052586185e-07,
+      "loss": 1.5293,
+      "step": 2095
+    },
+    {
+      "epoch": 0.4350352843503528,
+      "grad_norm": 0.9091501470396954,
+      "learning_rate": 2.564821746736294e-07,
+      "loss": 1.5058,
+      "step": 2096
+    },
+    {
+      "epoch": 0.4352428393524284,
+      "grad_norm": 0.6747206005797378,
+      "learning_rate": 2.563672333101447e-07,
+      "loss": 1.4499,
+      "step": 2097
+    },
+    {
+      "epoch": 0.43545039435450394,
+      "grad_norm": 1.2416595502264431,
+      "learning_rate": 2.5625227648427794e-07,
+      "loss": 1.581,
+      "step": 2098
+    },
+    {
+      "epoch": 0.4356579493565795,
+      "grad_norm": 0.6952150313047731,
+      "learning_rate": 2.5613730424490574e-07,
+      "loss": 1.5475,
+      "step": 2099
+    },
+    {
+      "epoch": 0.43586550435865506,
+      "grad_norm": 0.7732241194368042,
+      "learning_rate": 2.560223166409111e-07,
+      "loss": 1.6035,
+      "step": 2100
+    },
+    {
+      "epoch": 0.4360730593607306,
+      "grad_norm": 0.6977680960106218,
+      "learning_rate": 2.5590731372118377e-07,
+      "loss": 1.5819,
+      "step": 2101
+    },
+    {
+      "epoch": 0.43628061436280613,
+      "grad_norm": 0.942322246097702,
+      "learning_rate": 2.5579229553461994e-07,
+      "loss": 1.4804,
+      "step": 2102
+    },
+    {
+      "epoch": 0.4364881693648817,
+      "grad_norm": 0.9454414137347728,
+      "learning_rate": 2.556772621301223e-07,
+      "loss": 1.4412,
+      "step": 2103
+    },
+    {
+      "epoch": 0.43669572436695725,
+      "grad_norm": 0.6823896129596078,
+      "learning_rate": 2.555622135566e-07,
+      "loss": 1.5715,
+      "step": 2104
+    },
+    {
+      "epoch": 0.4369032793690328,
+      "grad_norm": 2.5678452853475626,
+      "learning_rate": 2.554471498629685e-07,
+      "loss": 1.5717,
+      "step": 2105
+    },
+    {
+      "epoch": 0.43711083437110837,
+      "grad_norm": 1.0413133255808267,
+      "learning_rate": 2.5533207109815004e-07,
+      "loss": 1.5249,
+      "step": 2106
+    },
+    {
+      "epoch": 0.4373183893731839,
+      "grad_norm": 0.873315196618782,
+      "learning_rate": 2.55216977311073e-07,
+      "loss": 1.5106,
+      "step": 2107
+    },
+    {
+      "epoch": 0.43752594437525943,
+      "grad_norm": 1.2472732961964537,
+      "learning_rate": 2.5510186855067205e-07,
+      "loss": 1.5647,
+      "step": 2108
+    },
+    {
+      "epoch": 0.43773349937733497,
+      "grad_norm": 1.1743021102139204,
+      "learning_rate": 2.5498674486588857e-07,
+      "loss": 1.5248,
+      "step": 2109
+    },
+    {
+      "epoch": 0.43794105437941055,
+      "grad_norm": 0.6864692264212933,
+      "learning_rate": 2.5487160630567e-07,
+      "loss": 1.4887,
+      "step": 2110
+    },
+    {
+      "epoch": 0.4381486093814861,
+      "grad_norm": 1.7098830341219857,
+      "learning_rate": 2.547564529189702e-07,
+      "loss": 1.4521,
+      "step": 2111
+    },
+    {
+      "epoch": 0.4383561643835616,
+      "grad_norm": 2.1376306290838745,
+      "learning_rate": 2.5464128475474935e-07,
+      "loss": 1.5057,
+      "step": 2112
+    },
+    {
+      "epoch": 0.4385637193856372,
+      "grad_norm": 0.7218449571633895,
+      "learning_rate": 2.54526101861974e-07,
+      "loss": 1.4417,
+      "step": 2113
+    },
+    {
+      "epoch": 0.43877127438771274,
+      "grad_norm": 0.7117824999076906,
+      "learning_rate": 2.544109042896166e-07,
+      "loss": 1.5356,
+      "step": 2114
+    },
+    {
+      "epoch": 0.43897882938978827,
+      "grad_norm": 0.9469277854892874,
+      "learning_rate": 2.542956920866564e-07,
+      "loss": 1.5012,
+      "step": 2115
+    },
+    {
+      "epoch": 0.43918638439186386,
+      "grad_norm": 1.49389575265253,
+      "learning_rate": 2.5418046530207827e-07,
+      "loss": 1.4665,
+      "step": 2116
+    },
+    {
+      "epoch": 0.4393939393939394,
+      "grad_norm": 1.178001804103002,
+      "learning_rate": 2.5406522398487383e-07,
+      "loss": 1.5106,
+      "step": 2117
+    },
+    {
+      "epoch": 0.4396014943960149,
+      "grad_norm": 0.7205061218601415,
+      "learning_rate": 2.539499681840405e-07,
+      "loss": 1.4869,
+      "step": 2118
+    },
+    {
+      "epoch": 0.4398090493980905,
+      "grad_norm": 0.7064580384542527,
+      "learning_rate": 2.53834697948582e-07,
+      "loss": 1.5286,
+      "step": 2119
+    },
+    {
+      "epoch": 0.44001660440016604,
+      "grad_norm": 1.057498935743862,
+      "learning_rate": 2.5371941332750827e-07,
+      "loss": 1.4963,
+      "step": 2120
+    },
+    {
+      "epoch": 0.4402241594022416,
+      "grad_norm": 0.7914071785461088,
+      "learning_rate": 2.536041143698351e-07,
+      "loss": 1.4887,
+      "step": 2121
+    },
+    {
+      "epoch": 0.44043171440431717,
+      "grad_norm": 0.8098510614416544,
+      "learning_rate": 2.5348880112458475e-07,
+      "loss": 1.4643,
+      "step": 2122
+    },
+    {
+      "epoch": 0.4406392694063927,
+      "grad_norm": 0.8920768556497111,
+      "learning_rate": 2.5337347364078513e-07,
+      "loss": 1.5333,
+      "step": 2123
+    },
+    {
+      "epoch": 0.44084682440846823,
+      "grad_norm": 0.8260414924562738,
+      "learning_rate": 2.5325813196747063e-07,
+      "loss": 1.5388,
+      "step": 2124
+    },
+    {
+      "epoch": 0.4410543794105438,
+      "grad_norm": 1.2216266672317042,
+      "learning_rate": 2.531427761536813e-07,
+      "loss": 1.5691,
+      "step": 2125
+    },
+    {
+      "epoch": 0.44126193441261935,
+      "grad_norm": 1.019132578632894,
+      "learning_rate": 2.5302740624846345e-07,
+      "loss": 1.4846,
+      "step": 2126
+    },
+    {
+      "epoch": 0.4414694894146949,
+      "grad_norm": 0.6730609713596192,
+      "learning_rate": 2.529120223008693e-07,
+      "loss": 1.6148,
+      "step": 2127
+    },
+    {
+      "epoch": 0.44167704441677047,
+      "grad_norm": 0.7364599063826658,
+      "learning_rate": 2.52796624359957e-07,
+      "loss": 1.4845,
+      "step": 2128
+    },
+    {
+      "epoch": 0.441884599418846,
+      "grad_norm": 1.4282277236522039,
+      "learning_rate": 2.526812124747907e-07,
+      "loss": 1.5591,
+      "step": 2129
+    },
+    {
+      "epoch": 0.44209215442092153,
+      "grad_norm": 0.8735489865557484,
+      "learning_rate": 2.525657866944406e-07,
+      "loss": 1.5901,
+      "step": 2130
+    },
+    {
+      "epoch": 0.44229970942299707,
+      "grad_norm": 1.314174621489636,
+      "learning_rate": 2.5245034706798255e-07,
+      "loss": 1.4845,
+      "step": 2131
+    },
+    {
+      "epoch": 0.44250726442507266,
+      "grad_norm": 0.7844126704488753,
+      "learning_rate": 2.523348936444984e-07,
+      "loss": 1.4985,
+      "step": 2132
+    },
+    {
+      "epoch": 0.4427148194271482,
+      "grad_norm": 1.29598726479746,
+      "learning_rate": 2.5221942647307595e-07,
+      "loss": 1.4843,
+      "step": 2133
+    },
+    {
+      "epoch": 0.4429223744292237,
+      "grad_norm": 0.881226393157972,
+      "learning_rate": 2.521039456028087e-07,
+      "loss": 1.5349,
+      "step": 2134
+    },
+    {
+      "epoch": 0.4431299294312993,
+      "grad_norm": 1.2013944878876999,
+      "learning_rate": 2.5198845108279606e-07,
+      "loss": 1.4395,
+      "step": 2135
+    },
+    {
+      "epoch": 0.44333748443337484,
+      "grad_norm": 1.792177002966969,
+      "learning_rate": 2.518729429621433e-07,
+      "loss": 1.4914,
+      "step": 2136
+    },
+    {
+      "epoch": 0.4435450394354504,
+      "grad_norm": 0.7623098932584398,
+      "learning_rate": 2.517574212899614e-07,
+      "loss": 1.5324,
+      "step": 2137
+    },
+    {
+      "epoch": 0.44375259443752596,
+      "grad_norm": 1.3308952326930836,
+      "learning_rate": 2.51641886115367e-07,
+      "loss": 1.4793,
+      "step": 2138
+    },
+    {
+      "epoch": 0.4439601494396015,
+      "grad_norm": 1.0284826176503037,
+      "learning_rate": 2.5152633748748274e-07,
+      "loss": 1.5472,
+      "step": 2139
+    },
+    {
+      "epoch": 0.444167704441677,
+      "grad_norm": 0.6638555943882263,
+      "learning_rate": 2.5141077545543676e-07,
+      "loss": 1.4954,
+      "step": 2140
+    },
+    {
+      "epoch": 0.4443752594437526,
+      "grad_norm": 0.7581208311184571,
+      "learning_rate": 2.512952000683629e-07,
+      "loss": 1.4955,
+      "step": 2141
+    },
+    {
+      "epoch": 0.44458281444582815,
+      "grad_norm": 0.772749048068988,
+      "learning_rate": 2.511796113754009e-07,
+      "loss": 1.5363,
+      "step": 2142
+    },
+    {
+      "epoch": 0.4447903694479037,
+      "grad_norm": 0.8878423028150323,
+      "learning_rate": 2.5106400942569585e-07,
+      "loss": 1.5375,
+      "step": 2143
+    },
+    {
+      "epoch": 0.44499792444997927,
+      "grad_norm": 0.8595138628111212,
+      "learning_rate": 2.509483942683987e-07,
+      "loss": 1.5544,
+      "step": 2144
+    },
+    {
+      "epoch": 0.4452054794520548,
+      "grad_norm": 1.1120578591520924,
+      "learning_rate": 2.5083276595266595e-07,
+      "loss": 1.588,
+      "step": 2145
+    },
+    {
+      "epoch": 0.44541303445413033,
+      "grad_norm": 0.9408018476804988,
+      "learning_rate": 2.507171245276597e-07,
+      "loss": 1.4777,
+      "step": 2146
+    },
+    {
+      "epoch": 0.4456205894562059,
+      "grad_norm": 1.0914886847384382,
+      "learning_rate": 2.5060147004254754e-07,
+      "loss": 1.536,
+      "step": 2147
+    },
+    {
+      "epoch": 0.44582814445828145,
+      "grad_norm": 1.0940914759700866,
+      "learning_rate": 2.504858025465028e-07,
+      "loss": 1.5379,
+      "step": 2148
+    },
+    {
+      "epoch": 0.446035699460357,
+      "grad_norm": 1.0911655521561265,
+      "learning_rate": 2.503701220887042e-07,
+      "loss": 1.5028,
+      "step": 2149
+    },
+    {
+      "epoch": 0.44624325446243257,
+      "grad_norm": 0.6978686356143317,
+      "learning_rate": 2.5025442871833585e-07,
+      "loss": 1.4942,
+      "step": 2150
+    },
+    {
+      "epoch": 0.4464508094645081,
+      "grad_norm": 0.9637473364927642,
+      "learning_rate": 2.5013872248458777e-07,
+      "loss": 1.5235,
+      "step": 2151
+    },
+    {
+      "epoch": 0.44665836446658364,
+      "grad_norm": 0.73052714759084,
+      "learning_rate": 2.5002300343665485e-07,
+      "loss": 1.6125,
+      "step": 2152
+    },
+    {
+      "epoch": 0.44686591946865917,
+      "grad_norm": 0.9468713510678555,
+      "learning_rate": 2.4990727162373806e-07,
+      "loss": 1.5183,
+      "step": 2153
+    },
+    {
+      "epoch": 0.44707347447073476,
+      "grad_norm": 0.6504356193109583,
+      "learning_rate": 2.4979152709504334e-07,
+      "loss": 1.4517,
+      "step": 2154
+    },
+    {
+      "epoch": 0.4472810294728103,
+      "grad_norm": 0.7420101022035055,
+      "learning_rate": 2.496757698997822e-07,
+      "loss": 1.4707,
+      "step": 2155
+    },
+    {
+      "epoch": 0.4474885844748858,
+      "grad_norm": 0.6549410960572217,
+      "learning_rate": 2.495600000871716e-07,
+      "loss": 1.5719,
+      "step": 2156
+    },
+    {
+      "epoch": 0.4476961394769614,
+      "grad_norm": 2.359159900418399,
+      "learning_rate": 2.494442177064336e-07,
+      "loss": 1.4731,
+      "step": 2157
+    },
+    {
+      "epoch": 0.44790369447903694,
+      "grad_norm": 0.7331777173551153,
+      "learning_rate": 2.493284228067961e-07,
+      "loss": 1.4802,
+      "step": 2158
+    },
+    {
+      "epoch": 0.4481112494811125,
+      "grad_norm": 1.2440970935529934,
+      "learning_rate": 2.492126154374917e-07,
+      "loss": 1.5337,
+      "step": 2159
+    },
+    {
+      "epoch": 0.44831880448318806,
+      "grad_norm": 0.7913351001443824,
+      "learning_rate": 2.490967956477589e-07,
+      "loss": 1.4839,
+      "step": 2160
+    },
+    {
+      "epoch": 0.4485263594852636,
+      "grad_norm": 0.7738195761239326,
+      "learning_rate": 2.48980963486841e-07,
+      "loss": 1.5918,
+      "step": 2161
+    },
+    {
+      "epoch": 0.4487339144873391,
+      "grad_norm": 0.7822292770972306,
+      "learning_rate": 2.4886511900398683e-07,
+      "loss": 1.432,
+      "step": 2162
+    },
+    {
+      "epoch": 0.4489414694894147,
+      "grad_norm": 0.9965774970451594,
+      "learning_rate": 2.487492622484504e-07,
+      "loss": 1.5116,
+      "step": 2163
+    },
+    {
+      "epoch": 0.44914902449149025,
+      "grad_norm": 0.8766209811602343,
+      "learning_rate": 2.4863339326949094e-07,
+      "loss": 1.441,
+      "step": 2164
+    },
+    {
+      "epoch": 0.4493565794935658,
+      "grad_norm": 0.652786520871881,
+      "learning_rate": 2.485175121163727e-07,
+      "loss": 1.5011,
+      "step": 2165
+    },
+    {
+      "epoch": 0.44956413449564137,
+      "grad_norm": 0.7154615701241811,
+      "learning_rate": 2.4840161883836554e-07,
+      "loss": 1.583,
+      "step": 2166
+    },
+    {
+      "epoch": 0.4497716894977169,
+      "grad_norm": 0.7150291241211217,
+      "learning_rate": 2.48285713484744e-07,
+      "loss": 1.5129,
+      "step": 2167
+    },
+    {
+      "epoch": 0.44997924449979243,
+      "grad_norm": 0.7413374877217682,
+      "learning_rate": 2.481697961047881e-07,
+      "loss": 1.4973,
+      "step": 2168
+    },
+    {
+      "epoch": 0.450186799501868,
+      "grad_norm": 0.936433186319151,
+      "learning_rate": 2.480538667477827e-07,
+      "loss": 1.5095,
+      "step": 2169
+    },
+    {
+      "epoch": 0.45039435450394355,
+      "grad_norm": 0.9069472615583587,
+      "learning_rate": 2.4793792546301804e-07,
+      "loss": 1.5027,
+      "step": 2170
+    },
+    {
+      "epoch": 0.4506019095060191,
+      "grad_norm": 1.1355643095112342,
+      "learning_rate": 2.478219722997891e-07,
+      "loss": 1.5675,
+      "step": 2171
+    },
+    {
+      "epoch": 0.45080946450809467,
+      "grad_norm": 0.7132307173851706,
+      "learning_rate": 2.4770600730739615e-07,
+      "loss": 1.5252,
+      "step": 2172
+    },
+    {
+      "epoch": 0.4510170195101702,
+      "grad_norm": 0.8272218787985647,
+      "learning_rate": 2.475900305351445e-07,
+      "loss": 1.5477,
+      "step": 2173
+    },
+    {
+      "epoch": 0.45122457451224574,
+      "grad_norm": 0.7242767194465567,
+      "learning_rate": 2.474740420323443e-07,
+      "loss": 1.5261,
+      "step": 2174
+    },
+    {
+      "epoch": 0.45143212951432127,
+      "grad_norm": 0.8973888682735535,
+      "learning_rate": 2.4735804184831086e-07,
+      "loss": 1.5265,
+      "step": 2175
+    },
+    {
+      "epoch": 0.45163968451639686,
+      "grad_norm": 0.6911003042600978,
+      "learning_rate": 2.472420300323643e-07,
+      "loss": 1.4984,
+      "step": 2176
+    },
+    {
+      "epoch": 0.4518472395184724,
+      "grad_norm": 0.671625565554024,
+      "learning_rate": 2.471260066338298e-07,
+      "loss": 1.5356,
+      "step": 2177
+    },
+    {
+      "epoch": 0.4520547945205479,
+      "grad_norm": 0.7630655099457171,
+      "learning_rate": 2.4700997170203746e-07,
+      "loss": 1.496,
+      "step": 2178
+    },
+    {
+      "epoch": 0.4522623495226235,
+      "grad_norm": 0.7700659474829851,
+      "learning_rate": 2.468939252863223e-07,
+      "loss": 1.6678,
+      "step": 2179
+    },
+    {
+      "epoch": 0.45246990452469904,
+      "grad_norm": 1.4565397812827277,
+      "learning_rate": 2.4677786743602396e-07,
+      "loss": 1.4823,
+      "step": 2180
+    },
+    {
+      "epoch": 0.4526774595267746,
+      "grad_norm": 1.0367986862582488,
+      "learning_rate": 2.466617982004874e-07,
+      "loss": 1.489,
+      "step": 2181
+    },
+    {
+      "epoch": 0.45288501452885016,
+      "grad_norm": 0.651934882544263,
+      "learning_rate": 2.46545717629062e-07,
+      "loss": 1.4984,
+      "step": 2182
+    },
+    {
+      "epoch": 0.4530925695309257,
+      "grad_norm": 1.1317650152368461,
+      "learning_rate": 2.4642962577110225e-07,
+      "loss": 1.578,
+      "step": 2183
+    },
+    {
+      "epoch": 0.4533001245330012,
+      "grad_norm": 0.8274764600224538,
+      "learning_rate": 2.4631352267596734e-07,
+      "loss": 1.4889,
+      "step": 2184
+    },
+    {
+      "epoch": 0.4535076795350768,
+      "grad_norm": 0.7508653095657366,
+      "learning_rate": 2.4619740839302105e-07,
+      "loss": 1.4549,
+      "step": 2185
+    },
+    {
+      "epoch": 0.45371523453715235,
+      "grad_norm": 0.7464091447727699,
+      "learning_rate": 2.460812829716323e-07,
+      "loss": 1.5054,
+      "step": 2186
+    },
+    {
+      "epoch": 0.4539227895392279,
+      "grad_norm": 3.2129031051020234,
+      "learning_rate": 2.459651464611745e-07,
+      "loss": 1.5716,
+      "step": 2187
+    },
+    {
+      "epoch": 0.45413034454130347,
+      "grad_norm": 2.439099305498632,
+      "learning_rate": 2.4584899891102566e-07,
+      "loss": 1.5479,
+      "step": 2188
+    },
+    {
+      "epoch": 0.454337899543379,
+      "grad_norm": 0.8031207600909883,
+      "learning_rate": 2.4573284037056876e-07,
+      "loss": 1.4635,
+      "step": 2189
+    },
+    {
+      "epoch": 0.45454545454545453,
+      "grad_norm": 1.0006031458842293,
+      "learning_rate": 2.4561667088919135e-07,
+      "loss": 1.6125,
+      "step": 2190
+    },
+    {
+      "epoch": 0.4547530095475301,
+      "grad_norm": 0.810242147883144,
+      "learning_rate": 2.455004905162855e-07,
+      "loss": 1.4435,
+      "step": 2191
+    },
+    {
+      "epoch": 0.45496056454960565,
+      "grad_norm": 2.0254887730141404,
+      "learning_rate": 2.4538429930124814e-07,
+      "loss": 1.5417,
+      "step": 2192
+    },
+    {
+      "epoch": 0.4551681195516812,
+      "grad_norm": 0.7985861811236729,
+      "learning_rate": 2.4526809729348056e-07,
+      "loss": 1.4958,
+      "step": 2193
+    },
+    {
+      "epoch": 0.45537567455375677,
+      "grad_norm": 0.9310305620406012,
+      "learning_rate": 2.4515188454238887e-07,
+      "loss": 1.5145,
+      "step": 2194
+    },
+    {
+      "epoch": 0.4555832295558323,
+      "grad_norm": 0.8109195266052031,
+      "learning_rate": 2.450356610973836e-07,
+      "loss": 1.5166,
+      "step": 2195
+    },
+    {
+      "epoch": 0.45579078455790784,
+      "grad_norm": 0.9150030933398595,
+      "learning_rate": 2.4491942700787993e-07,
+      "loss": 1.5094,
+      "step": 2196
+    },
+    {
+      "epoch": 0.4559983395599834,
+      "grad_norm": 0.951889116884637,
+      "learning_rate": 2.4480318232329746e-07,
+      "loss": 1.5348,
+      "step": 2197
+    },
+    {
+      "epoch": 0.45620589456205896,
+      "grad_norm": 1.1850439055436415,
+      "learning_rate": 2.4468692709306036e-07,
+      "loss": 1.4985,
+      "step": 2198
+    },
+    {
+      "epoch": 0.4564134495641345,
+      "grad_norm": 0.8749768594294257,
+      "learning_rate": 2.445706613665973e-07,
+      "loss": 1.4924,
+      "step": 2199
+    },
+    {
+      "epoch": 0.45662100456621,
+      "grad_norm": 0.9163160438438379,
+      "learning_rate": 2.4445438519334127e-07,
+      "loss": 1.4742,
+      "step": 2200
+    },
+    {
+      "epoch": 0.4568285595682856,
+      "grad_norm": 0.643252125770128,
+      "learning_rate": 2.443380986227299e-07,
+      "loss": 1.4366,
+      "step": 2201
+    },
+    {
+      "epoch": 0.45703611457036114,
+      "grad_norm": 0.7997659922866655,
+      "learning_rate": 2.442218017042051e-07,
+      "loss": 1.4971,
+      "step": 2202
+    },
+    {
+      "epoch": 0.4572436695724367,
+      "grad_norm": 1.0279960693569334,
+      "learning_rate": 2.4410549448721334e-07,
+      "loss": 1.5281,
+      "step": 2203
+    },
+    {
+      "epoch": 0.45745122457451226,
+      "grad_norm": 0.8381311521165757,
+      "learning_rate": 2.439891770212053e-07,
+      "loss": 1.574,
+      "step": 2204
+    },
+    {
+      "epoch": 0.4576587795765878,
+      "grad_norm": 0.9408991150849904,
+      "learning_rate": 2.438728493556359e-07,
+      "loss": 1.471,
+      "step": 2205
+    },
+    {
+      "epoch": 0.4578663345786633,
+      "grad_norm": 0.9170328175042292,
+      "learning_rate": 2.437565115399649e-07,
+      "loss": 1.4747,
+      "step": 2206
+    },
+    {
+      "epoch": 0.4580738895807389,
+      "grad_norm": 0.6427472013442638,
+      "learning_rate": 2.436401636236559e-07,
+      "loss": 1.4991,
+      "step": 2207
+    },
+    {
+      "epoch": 0.45828144458281445,
+      "grad_norm": 0.7063233564871986,
+      "learning_rate": 2.435238056561769e-07,
+      "loss": 1.5374,
+      "step": 2208
+    },
+    {
+      "epoch": 0.45848899958489,
+      "grad_norm": 0.8076142970763248,
+      "learning_rate": 2.4340743768700026e-07,
+      "loss": 1.4997,
+      "step": 2209
+    },
+    {
+      "epoch": 0.45869655458696557,
+      "grad_norm": 1.3361724639534662,
+      "learning_rate": 2.432910597656025e-07,
+      "loss": 1.5836,
+      "step": 2210
+    },
+    {
+      "epoch": 0.4589041095890411,
+      "grad_norm": 0.7501759904041626,
+      "learning_rate": 2.4317467194146455e-07,
+      "loss": 1.5164,
+      "step": 2211
+    },
+    {
+      "epoch": 0.45911166459111663,
+      "grad_norm": 1.0543792063166513,
+      "learning_rate": 2.430582742640713e-07,
+      "loss": 1.545,
+      "step": 2212
+    },
+    {
+      "epoch": 0.4593192195931922,
+      "grad_norm": 0.8380079336488809,
+      "learning_rate": 2.4294186678291194e-07,
+      "loss": 1.5199,
+      "step": 2213
+    },
+    {
+      "epoch": 0.45952677459526775,
+      "grad_norm": 0.9021296785095388,
+      "learning_rate": 2.4282544954748003e-07,
+      "loss": 1.5034,
+      "step": 2214
+    },
+    {
+      "epoch": 0.4597343295973433,
+      "grad_norm": 0.7521408333219058,
+      "learning_rate": 2.4270902260727284e-07,
+      "loss": 1.5231,
+      "step": 2215
+    },
+    {
+      "epoch": 0.45994188459941887,
+      "grad_norm": 0.844388611353209,
+      "learning_rate": 2.4259258601179214e-07,
+      "loss": 1.5478,
+      "step": 2216
+    },
+    {
+      "epoch": 0.4601494396014944,
+      "grad_norm": 0.9820550374511889,
+      "learning_rate": 2.4247613981054373e-07,
+      "loss": 1.5211,
+      "step": 2217
+    },
+    {
+      "epoch": 0.46035699460356994,
+      "grad_norm": 0.7235320752272337,
+      "learning_rate": 2.423596840530373e-07,
+      "loss": 1.5455,
+      "step": 2218
+    },
+    {
+      "epoch": 0.4605645496056455,
+      "grad_norm": 0.8324983002567111,
+      "learning_rate": 2.4224321878878694e-07,
+      "loss": 1.5362,
+      "step": 2219
+    },
+    {
+      "epoch": 0.46077210460772106,
+      "grad_norm": 1.0399830447329832,
+      "learning_rate": 2.421267440673103e-07,
+      "loss": 1.5326,
+      "step": 2220
+    },
+    {
+      "epoch": 0.4609796596097966,
+      "grad_norm": 1.0575362914644075,
+      "learning_rate": 2.4201025993812967e-07,
+      "loss": 1.5321,
+      "step": 2221
+    },
+    {
+      "epoch": 0.4611872146118721,
+      "grad_norm": 0.6324880735125413,
+      "learning_rate": 2.4189376645077077e-07,
+      "loss": 1.4759,
+      "step": 2222
+    },
+    {
+      "epoch": 0.4613947696139477,
+      "grad_norm": 9.613455962360714,
+      "learning_rate": 2.417772636547637e-07,
+      "loss": 1.6108,
+      "step": 2223
+    },
+    {
+      "epoch": 0.46160232461602324,
+      "grad_norm": 0.796027678126697,
+      "learning_rate": 2.4166075159964224e-07,
+      "loss": 1.5208,
+      "step": 2224
+    },
+    {
+      "epoch": 0.4618098796180988,
+      "grad_norm": 1.340168327814824,
+      "learning_rate": 2.4154423033494427e-07,
+      "loss": 1.5384,
+      "step": 2225
+    },
+    {
+      "epoch": 0.46201743462017436,
+      "grad_norm": 0.743504323645428,
+      "learning_rate": 2.4142769991021147e-07,
+      "loss": 1.5876,
+      "step": 2226
+    },
+    {
+      "epoch": 0.4622249896222499,
+      "grad_norm": 0.7251114209535581,
+      "learning_rate": 2.413111603749896e-07,
+      "loss": 1.5301,
+      "step": 2227
+    },
+    {
+      "epoch": 0.4624325446243254,
+      "grad_norm": 0.8329547937009193,
+      "learning_rate": 2.4119461177882816e-07,
+      "loss": 1.5098,
+      "step": 2228
+    },
+    {
+      "epoch": 0.462640099626401,
+      "grad_norm": 0.7595128985859629,
+      "learning_rate": 2.4107805417128035e-07,
+      "loss": 1.535,
+      "step": 2229
+    },
+    {
+      "epoch": 0.46284765462847655,
+      "grad_norm": 1.0289447997454348,
+      "learning_rate": 2.409614876019036e-07,
+      "loss": 1.5338,
+      "step": 2230
+    },
+    {
+      "epoch": 0.4630552096305521,
+      "grad_norm": 0.876545684650963,
+      "learning_rate": 2.4084491212025873e-07,
+      "loss": 1.5588,
+      "step": 2231
+    },
+    {
+      "epoch": 0.46326276463262767,
+      "grad_norm": 0.8742852720639952,
+      "learning_rate": 2.4072832777591075e-07,
+      "loss": 1.4348,
+      "step": 2232
+    },
+    {
+      "epoch": 0.4634703196347032,
+      "grad_norm": 0.7026562412805656,
+      "learning_rate": 2.40611734618428e-07,
+      "loss": 1.457,
+      "step": 2233
+    },
+    {
+      "epoch": 0.46367787463677873,
+      "grad_norm": 1.1270635462841694,
+      "learning_rate": 2.404951326973829e-07,
+      "loss": 1.551,
+      "step": 2234
+    },
+    {
+      "epoch": 0.4638854296388543,
+      "grad_norm": 0.8683412207048049,
+      "learning_rate": 2.403785220623515e-07,
+      "loss": 1.4483,
+      "step": 2235
+    },
+    {
+      "epoch": 0.46409298464092985,
+      "grad_norm": 0.7555098610212173,
+      "learning_rate": 2.402619027629136e-07,
+      "loss": 1.5824,
+      "step": 2236
+    },
+    {
+      "epoch": 0.4643005396430054,
+      "grad_norm": 0.7981558658921464,
+      "learning_rate": 2.401452748486525e-07,
+      "loss": 1.4537,
+      "step": 2237
+    },
+    {
+      "epoch": 0.46450809464508097,
+      "grad_norm": 1.3371005366076087,
+      "learning_rate": 2.400286383691554e-07,
+      "loss": 1.5096,
+      "step": 2238
+    },
+    {
+      "epoch": 0.4647156496471565,
+      "grad_norm": 0.8455773124255425,
+      "learning_rate": 2.39911993374013e-07,
+      "loss": 1.5191,
+      "step": 2239
+    },
+    {
+      "epoch": 0.46492320464923204,
+      "grad_norm": 0.7691116102035135,
+      "learning_rate": 2.3979533991281966e-07,
+      "loss": 1.5462,
+      "step": 2240
+    },
+    {
+      "epoch": 0.4651307596513076,
+      "grad_norm": 0.6431765198247197,
+      "learning_rate": 2.396786780351733e-07,
+      "loss": 1.4937,
+      "step": 2241
+    },
+    {
+      "epoch": 0.46533831465338316,
+      "grad_norm": 0.7657358944398197,
+      "learning_rate": 2.395620077906755e-07,
+      "loss": 1.4717,
+      "step": 2242
+    },
+    {
+      "epoch": 0.4655458696554587,
+      "grad_norm": 0.9294180369349442,
+      "learning_rate": 2.394453292289313e-07,
+      "loss": 1.4809,
+      "step": 2243
+    },
+    {
+      "epoch": 0.4657534246575342,
+      "grad_norm": 11.50667086537391,
+      "learning_rate": 2.3932864239954937e-07,
+      "loss": 1.6035,
+      "step": 2244
+    },
+    {
+      "epoch": 0.4659609796596098,
+      "grad_norm": 1.1673548490255723,
+      "learning_rate": 2.3921194735214183e-07,
+      "loss": 1.5091,
+      "step": 2245
+    },
+    {
+      "epoch": 0.46616853466168534,
+      "grad_norm": 0.9266484129151237,
+      "learning_rate": 2.390952441363243e-07,
+      "loss": 1.5089,
+      "step": 2246
+    },
+    {
+      "epoch": 0.4663760896637609,
+      "grad_norm": 2.0752055352963095,
+      "learning_rate": 2.38978532801716e-07,
+      "loss": 1.5774,
+      "step": 2247
+    },
+    {
+      "epoch": 0.46658364466583646,
+      "grad_norm": 0.7924802957558427,
+      "learning_rate": 2.388618133979393e-07,
+      "loss": 1.5107,
+      "step": 2248
+    },
+    {
+      "epoch": 0.466791199667912,
+      "grad_norm": 0.6988179244314432,
+      "learning_rate": 2.3874508597462036e-07,
+      "loss": 1.4901,
+      "step": 2249
+    },
+    {
+      "epoch": 0.4669987546699875,
+      "grad_norm": 0.7851342111636496,
+      "learning_rate": 2.386283505813885e-07,
+      "loss": 1.5156,
+      "step": 2250
+    },
+    {
+      "epoch": 0.4672063096720631,
+      "grad_norm": 0.7180013615537107,
+      "learning_rate": 2.385116072678765e-07,
+      "loss": 1.5447,
+      "step": 2251
+    },
+    {
+      "epoch": 0.46741386467413865,
+      "grad_norm": 0.6499516346801097,
+      "learning_rate": 2.383948560837206e-07,
+      "loss": 1.5142,
+      "step": 2252
+    },
+    {
+      "epoch": 0.4676214196762142,
+      "grad_norm": 0.6610478488921054,
+      "learning_rate": 2.3827809707856023e-07,
+      "loss": 1.482,
+      "step": 2253
+    },
+    {
+      "epoch": 0.46782897467828977,
+      "grad_norm": 1.2337671553158112,
+      "learning_rate": 2.3816133030203818e-07,
+      "loss": 1.4226,
+      "step": 2254
+    },
+    {
+      "epoch": 0.4680365296803653,
+      "grad_norm": 0.6947826865818518,
+      "learning_rate": 2.3804455580380072e-07,
+      "loss": 1.5019,
+      "step": 2255
+    },
+    {
+      "epoch": 0.46824408468244083,
+      "grad_norm": 0.6997069130088122,
+      "learning_rate": 2.3792777363349722e-07,
+      "loss": 1.5007,
+      "step": 2256
+    },
+    {
+      "epoch": 0.4684516396845164,
+      "grad_norm": 0.7088211810414526,
+      "learning_rate": 2.378109838407802e-07,
+      "loss": 1.5571,
+      "step": 2257
+    },
+    {
+      "epoch": 0.46865919468659195,
+      "grad_norm": 0.9213998911953192,
+      "learning_rate": 2.3769418647530576e-07,
+      "loss": 1.5353,
+      "step": 2258
+    },
+    {
+      "epoch": 0.4688667496886675,
+      "grad_norm": 4.383534052373395,
+      "learning_rate": 2.3757738158673302e-07,
+      "loss": 1.5843,
+      "step": 2259
+    },
+    {
+      "epoch": 0.46907430469074307,
+      "grad_norm": 0.9149189232816582,
+      "learning_rate": 2.3746056922472423e-07,
+      "loss": 1.4917,
+      "step": 2260
+    },
+    {
+      "epoch": 0.4692818596928186,
+      "grad_norm": 0.8751886951499699,
+      "learning_rate": 2.3734374943894502e-07,
+      "loss": 1.4453,
+      "step": 2261
+    },
+    {
+      "epoch": 0.46948941469489414,
+      "grad_norm": 0.8068067117599571,
+      "learning_rate": 2.3722692227906394e-07,
+      "loss": 1.4656,
+      "step": 2262
+    },
+    {
+      "epoch": 0.4696969696969697,
+      "grad_norm": 0.7183821599326902,
+      "learning_rate": 2.3711008779475286e-07,
+      "loss": 1.5502,
+      "step": 2263
+    },
+    {
+      "epoch": 0.46990452469904526,
+      "grad_norm": 1.4283606380429645,
+      "learning_rate": 2.3699324603568674e-07,
+      "loss": 1.5851,
+      "step": 2264
+    },
+    {
+      "epoch": 0.4701120797011208,
+      "grad_norm": 0.9085930663017272,
+      "learning_rate": 2.3687639705154354e-07,
+      "loss": 1.4868,
+      "step": 2265
+    },
+    {
+      "epoch": 0.4703196347031963,
+      "grad_norm": 0.9242844557414177,
+      "learning_rate": 2.3675954089200437e-07,
+      "loss": 1.4919,
+      "step": 2266
+    },
+    {
+      "epoch": 0.4705271897052719,
+      "grad_norm": 0.721031895461535,
+      "learning_rate": 2.366426776067533e-07,
+      "loss": 1.488,
+      "step": 2267
+    },
+    {
+      "epoch": 0.47073474470734744,
+      "grad_norm": 0.9094884078174456,
+      "learning_rate": 2.3652580724547758e-07,
+      "loss": 1.5261,
+      "step": 2268
+    },
+    {
+      "epoch": 0.470942299709423,
+      "grad_norm": 0.8241169253099095,
+      "learning_rate": 2.3640892985786736e-07,
+      "loss": 1.4752,
+      "step": 2269
+    },
+    {
+      "epoch": 0.47114985471149856,
+      "grad_norm": 0.8928604733558316,
+      "learning_rate": 2.362920454936158e-07,
+      "loss": 1.455,
+      "step": 2270
+    },
+    {
+      "epoch": 0.4713574097135741,
+      "grad_norm": 0.7805080742315295,
+      "learning_rate": 2.3617515420241897e-07,
+      "loss": 1.4613,
+      "step": 2271
+    },
+    {
+      "epoch": 0.4715649647156496,
+      "grad_norm": 0.6460143432090598,
+      "learning_rate": 2.3605825603397606e-07,
+      "loss": 1.4943,
+      "step": 2272
+    },
+    {
+      "epoch": 0.4717725197177252,
+      "grad_norm": 1.1613912271518472,
+      "learning_rate": 2.3594135103798894e-07,
+      "loss": 1.5032,
+      "step": 2273
+    },
+    {
+      "epoch": 0.47198007471980075,
+      "grad_norm": 1.3145593790529866,
+      "learning_rate": 2.3582443926416267e-07,
+      "loss": 1.5537,
+      "step": 2274
+    },
+    {
+      "epoch": 0.4721876297218763,
+      "grad_norm": 0.9746060269879496,
+      "learning_rate": 2.357075207622049e-07,
+      "loss": 1.5431,
+      "step": 2275
+    },
+    {
+      "epoch": 0.47239518472395187,
+      "grad_norm": 0.7274668284374193,
+      "learning_rate": 2.3559059558182626e-07,
+      "loss": 1.5456,
+      "step": 2276
+    },
+    {
+      "epoch": 0.4726027397260274,
+      "grad_norm": 0.7978097033622888,
+      "learning_rate": 2.3547366377274035e-07,
+      "loss": 1.509,
+      "step": 2277
+    },
+    {
+      "epoch": 0.47281029472810293,
+      "grad_norm": 0.6792619886445909,
+      "learning_rate": 2.3535672538466343e-07,
+      "loss": 1.5635,
+      "step": 2278
+    },
+    {
+      "epoch": 0.4730178497301785,
+      "grad_norm": 0.7510328569408615,
+      "learning_rate": 2.352397804673145e-07,
+      "loss": 1.5883,
+      "step": 2279
+    },
+    {
+      "epoch": 0.47322540473225405,
+      "grad_norm": 0.8196013965152724,
+      "learning_rate": 2.3512282907041557e-07,
+      "loss": 1.4882,
+      "step": 2280
+    },
+    {
+      "epoch": 0.4734329597343296,
+      "grad_norm": 0.9124355227306552,
+      "learning_rate": 2.3500587124369124e-07,
+      "loss": 1.4656,
+      "step": 2281
+    },
+    {
+      "epoch": 0.47364051473640517,
+      "grad_norm": 0.7035988412176885,
+      "learning_rate": 2.348889070368687e-07,
+      "loss": 1.5247,
+      "step": 2282
+    },
+    {
+      "epoch": 0.4738480697384807,
+      "grad_norm": 0.8640509539676204,
+      "learning_rate": 2.347719364996783e-07,
+      "loss": 1.5071,
+      "step": 2283
+    },
+    {
+      "epoch": 0.47405562474055624,
+      "grad_norm": 0.6564767860950007,
+      "learning_rate": 2.3465495968185257e-07,
+      "loss": 1.5044,
+      "step": 2284
+    },
+    {
+      "epoch": 0.4742631797426318,
+      "grad_norm": 1.1517678340453643,
+      "learning_rate": 2.345379766331271e-07,
+      "loss": 1.507,
+      "step": 2285
+    },
+    {
+      "epoch": 0.47447073474470736,
+      "grad_norm": 0.6889288333671003,
+      "learning_rate": 2.344209874032399e-07,
+      "loss": 1.5451,
+      "step": 2286
+    },
+    {
+      "epoch": 0.4746782897467829,
+      "grad_norm": 0.6398958340841971,
+      "learning_rate": 2.3430399204193176e-07,
+      "loss": 1.5083,
+      "step": 2287
+    },
+    {
+      "epoch": 0.4748858447488584,
+      "grad_norm": 0.9696278280102735,
+      "learning_rate": 2.3418699059894596e-07,
+      "loss": 1.5542,
+      "step": 2288
+    },
+    {
+      "epoch": 0.475093399750934,
+      "grad_norm": 0.7131244976565894,
+      "learning_rate": 2.3406998312402836e-07,
+      "loss": 1.5225,
+      "step": 2289
+    },
+    {
+      "epoch": 0.47530095475300954,
+      "grad_norm": 0.8488901136050451,
+      "learning_rate": 2.3395296966692753e-07,
+      "loss": 1.5077,
+      "step": 2290
+    },
+    {
+      "epoch": 0.4755085097550851,
+      "grad_norm": 0.7116597973479669,
+      "learning_rate": 2.3383595027739438e-07,
+      "loss": 1.5291,
+      "step": 2291
+    },
+    {
+      "epoch": 0.47571606475716066,
+      "grad_norm": 0.7912824666543066,
+      "learning_rate": 2.337189250051825e-07,
+      "loss": 1.4718,
+      "step": 2292
+    },
+    {
+      "epoch": 0.4759236197592362,
+      "grad_norm": 0.7762120187147161,
+      "learning_rate": 2.3360189390004788e-07,
+      "loss": 1.5113,
+      "step": 2293
+    },
+    {
+      "epoch": 0.4761311747613117,
+      "grad_norm": 1.029205861466796,
+      "learning_rate": 2.3348485701174918e-07,
+      "loss": 1.5491,
+      "step": 2294
+    },
+    {
+      "epoch": 0.4763387297633873,
+      "grad_norm": 5.880820379340486,
+      "learning_rate": 2.3336781439004717e-07,
+      "loss": 1.4188,
+      "step": 2295
+    },
+    {
+      "epoch": 0.47654628476546285,
+      "grad_norm": 0.8120818762675087,
+      "learning_rate": 2.3325076608470544e-07,
+      "loss": 1.4372,
+      "step": 2296
+    },
+    {
+      "epoch": 0.4767538397675384,
+      "grad_norm": 0.7106779144237448,
+      "learning_rate": 2.3313371214548976e-07,
+      "loss": 1.5736,
+      "step": 2297
+    },
+    {
+      "epoch": 0.47696139476961397,
+      "grad_norm": 1.2711830984132426,
+      "learning_rate": 2.3301665262216837e-07,
+      "loss": 1.5762,
+      "step": 2298
+    },
+    {
+      "epoch": 0.4771689497716895,
+      "grad_norm": 0.7380329117026444,
+      "learning_rate": 2.3289958756451176e-07,
+      "loss": 1.5254,
+      "step": 2299
+    },
+    {
+      "epoch": 0.47737650477376503,
+      "grad_norm": 0.6689751624502145,
+      "learning_rate": 2.327825170222931e-07,
+      "loss": 1.5081,
+      "step": 2300
+    },
+    {
+      "epoch": 0.4775840597758406,
+      "grad_norm": 0.8735488480645404,
+      "learning_rate": 2.3266544104528747e-07,
+      "loss": 1.5229,
+      "step": 2301
+    },
+    {
+      "epoch": 0.47779161477791615,
+      "grad_norm": 7.491064486529877,
+      "learning_rate": 2.3254835968327263e-07,
+      "loss": 1.5328,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4779991697799917,
+      "grad_norm": 0.9445451691796192,
+      "learning_rate": 2.324312729860284e-07,
+      "loss": 1.5257,
+      "step": 2303
+    },
+    {
+      "epoch": 0.47820672478206727,
+      "grad_norm": 0.8216093833592433,
+      "learning_rate": 2.3231418100333688e-07,
+      "loss": 1.5427,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4784142797841428,
+      "grad_norm": 0.8910714939683786,
+      "learning_rate": 2.3219708378498258e-07,
+      "loss": 1.4807,
+      "step": 2305
+    },
+    {
+      "epoch": 0.47862183478621834,
+      "grad_norm": 0.7606860366842499,
+      "learning_rate": 2.32079981380752e-07,
+      "loss": 1.502,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4788293897882939,
+      "grad_norm": 0.9012764651918588,
+      "learning_rate": 2.3196287384043404e-07,
+      "loss": 1.5503,
+      "step": 2307
+    },
+    {
+      "epoch": 0.47903694479036946,
+      "grad_norm": 0.9579651908088471,
+      "learning_rate": 2.3184576121381973e-07,
+      "loss": 1.5675,
+      "step": 2308
+    },
+    {
+      "epoch": 0.479244499792445,
+      "grad_norm": 0.7836072654754713,
+      "learning_rate": 2.317286435507023e-07,
+      "loss": 1.4791,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4794520547945205,
+      "grad_norm": 0.8674005482187821,
+      "learning_rate": 2.3161152090087695e-07,
+      "loss": 1.5291,
+      "step": 2310
+    },
+    {
+      "epoch": 0.4796596097965961,
+      "grad_norm": 0.7839208017615653,
+      "learning_rate": 2.3149439331414116e-07,
+      "loss": 1.5115,
+      "step": 2311
+    },
+    {
+      "epoch": 0.47986716479867164,
+      "grad_norm": 0.8487706287298489,
+      "learning_rate": 2.3137726084029455e-07,
+      "loss": 1.5373,
+      "step": 2312
+    },
+    {
+      "epoch": 0.4800747198007472,
+      "grad_norm": 0.6871791952873016,
+      "learning_rate": 2.3126012352913867e-07,
+      "loss": 1.5507,
+      "step": 2313
+    },
+    {
+      "epoch": 0.48028227480282276,
+      "grad_norm": 0.8373394184859119,
+      "learning_rate": 2.3114298143047718e-07,
+      "loss": 1.5851,
+      "step": 2314
+    },
+    {
+      "epoch": 0.4804898298048983,
+      "grad_norm": 1.515144902140093,
+      "learning_rate": 2.310258345941158e-07,
+      "loss": 1.5506,
+      "step": 2315
+    },
+    {
+      "epoch": 0.4806973848069738,
+      "grad_norm": 1.6261595708990992,
+      "learning_rate": 2.3090868306986233e-07,
+      "loss": 1.5354,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4809049398090494,
+      "grad_norm": 1.4336662702773295,
+      "learning_rate": 2.3079152690752637e-07,
+      "loss": 1.514,
+      "step": 2317
+    },
+    {
+      "epoch": 0.48111249481112495,
+      "grad_norm": 0.6615473570534988,
+      "learning_rate": 2.306743661569197e-07,
+      "loss": 1.4657,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4813200498132005,
+      "grad_norm": 0.7773467156877616,
+      "learning_rate": 2.3055720086785598e-07,
+      "loss": 1.4369,
+      "step": 2319
+    },
+    {
+      "epoch": 0.48152760481527607,
+      "grad_norm": 1.194637578620655,
+      "learning_rate": 2.304400310901506e-07,
+      "loss": 1.5464,
+      "step": 2320
+    },
+    {
+      "epoch": 0.4817351598173516,
+      "grad_norm": 0.6918974180078192,
+      "learning_rate": 2.3032285687362126e-07,
+      "loss": 1.5278,
+      "step": 2321
+    },
+    {
+      "epoch": 0.48194271481942713,
+      "grad_norm": 1.142409739512506,
+      "learning_rate": 2.3020567826808724e-07,
+      "loss": 1.5155,
+      "step": 2322
+    },
+    {
+      "epoch": 0.4821502698215027,
+      "grad_norm": 0.9249476922835445,
+      "learning_rate": 2.3008849532336971e-07,
+      "loss": 1.4607,
+      "step": 2323
+    },
+    {
+      "epoch": 0.48235782482357825,
+      "grad_norm": 0.7017221801410658,
+      "learning_rate": 2.2997130808929183e-07,
+      "loss": 1.4927,
+      "step": 2324
+    },
+    {
+      "epoch": 0.4825653798256538,
+      "grad_norm": 2.9660027003228544,
+      "learning_rate": 2.2985411661567843e-07,
+      "loss": 1.5249,
+      "step": 2325
+    },
+    {
+      "epoch": 0.48277293482772937,
+      "grad_norm": 0.6642418128238853,
+      "learning_rate": 2.297369209523563e-07,
+      "loss": 1.5503,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4829804898298049,
+      "grad_norm": 0.9172176372091863,
+      "learning_rate": 2.296197211491539e-07,
+      "loss": 1.5265,
+      "step": 2327
+    },
+    {
+      "epoch": 0.48318804483188044,
+      "grad_norm": 0.938666710961786,
+      "learning_rate": 2.2950251725590145e-07,
+      "loss": 1.4752,
+      "step": 2328
+    },
+    {
+      "epoch": 0.483395599833956,
+      "grad_norm": 0.6923747271119006,
+      "learning_rate": 2.293853093224309e-07,
+      "loss": 1.4951,
+      "step": 2329
+    },
+    {
+      "epoch": 0.48360315483603156,
+      "grad_norm": 3.7339901418898993,
+      "learning_rate": 2.2926809739857606e-07,
+      "loss": 1.5496,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4838107098381071,
+      "grad_norm": 0.9441949551559098,
+      "learning_rate": 2.291508815341722e-07,
+      "loss": 1.5485,
+      "step": 2331
+    },
+    {
+      "epoch": 0.4840182648401826,
+      "grad_norm": 0.7341260522535987,
+      "learning_rate": 2.290336617790565e-07,
+      "loss": 1.5495,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4842258198422582,
+      "grad_norm": 0.643687181295844,
+      "learning_rate": 2.2891643818306757e-07,
+      "loss": 1.541,
+      "step": 2333
+    },
+    {
+      "epoch": 0.48443337484433374,
+      "grad_norm": 0.7378296106714559,
+      "learning_rate": 2.287992107960459e-07,
+      "loss": 1.5333,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4846409298464093,
+      "grad_norm": 0.6845583694851329,
+      "learning_rate": 2.2868197966783343e-07,
+      "loss": 1.4853,
+      "step": 2335
+    },
+    {
+      "epoch": 0.48484848484848486,
+      "grad_norm": 0.7279902969198789,
+      "learning_rate": 2.2856474484827367e-07,
+      "loss": 1.5799,
+      "step": 2336
+    },
+    {
+      "epoch": 0.4850560398505604,
+      "grad_norm": 0.7684932353486686,
+      "learning_rate": 2.2844750638721168e-07,
+      "loss": 1.5329,
+      "step": 2337
+    },
+    {
+      "epoch": 0.4852635948526359,
+      "grad_norm": 0.6606907016769956,
+      "learning_rate": 2.2833026433449438e-07,
+      "loss": 1.5158,
+      "step": 2338
+    },
+    {
+      "epoch": 0.4854711498547115,
+      "grad_norm": 0.6035452251027119,
+      "learning_rate": 2.2821301873996978e-07,
+      "loss": 1.5473,
+      "step": 2339
+    },
+    {
+      "epoch": 0.48567870485678705,
+      "grad_norm": 1.649840124667926,
+      "learning_rate": 2.280957696534877e-07,
+      "loss": 1.499,
+      "step": 2340
+    },
+    {
+      "epoch": 0.4858862598588626,
+      "grad_norm": 0.6911183197901467,
+      "learning_rate": 2.2797851712489927e-07,
+      "loss": 1.5669,
+      "step": 2341
+    },
+    {
+      "epoch": 0.48609381486093817,
+      "grad_norm": 0.8624375163771072,
+      "learning_rate": 2.2786126120405714e-07,
+      "loss": 1.548,
+      "step": 2342
+    },
+    {
+      "epoch": 0.4863013698630137,
+      "grad_norm": 0.6485684695885501,
+      "learning_rate": 2.277440019408155e-07,
+      "loss": 1.5588,
+      "step": 2343
+    },
+    {
+      "epoch": 0.48650892486508923,
+      "grad_norm": 1.0583012242280798,
+      "learning_rate": 2.276267393850298e-07,
+      "loss": 1.5101,
+      "step": 2344
+    },
+    {
+      "epoch": 0.4867164798671648,
+      "grad_norm": 0.7303335413368193,
+      "learning_rate": 2.27509473586557e-07,
+      "loss": 1.4889,
+      "step": 2345
+    },
+    {
+      "epoch": 0.48692403486924035,
+      "grad_norm": 0.6507484640374903,
+      "learning_rate": 2.273922045952554e-07,
+      "loss": 1.5165,
+      "step": 2346
+    },
+    {
+      "epoch": 0.4871315898713159,
+      "grad_norm": 0.7355554998343299,
+      "learning_rate": 2.2727493246098465e-07,
+      "loss": 1.5499,
+      "step": 2347
+    },
+    {
+      "epoch": 0.48733914487339147,
+      "grad_norm": 0.8334077385243107,
+      "learning_rate": 2.2715765723360576e-07,
+      "loss": 1.5599,
+      "step": 2348
+    },
+    {
+      "epoch": 0.487546699875467,
+      "grad_norm": 0.6835985906661994,
+      "learning_rate": 2.2704037896298106e-07,
+      "loss": 1.5884,
+      "step": 2349
+    },
+    {
+      "epoch": 0.48775425487754254,
+      "grad_norm": 0.7642188923655947,
+      "learning_rate": 2.2692309769897408e-07,
+      "loss": 1.4738,
+      "step": 2350
+    },
+    {
+      "epoch": 0.4879618098796181,
+      "grad_norm": 0.7961778772622933,
+      "learning_rate": 2.268058134914498e-07,
+      "loss": 1.548,
+      "step": 2351
+    },
+    {
+      "epoch": 0.48816936488169366,
+      "grad_norm": 0.7717140197464556,
+      "learning_rate": 2.266885263902743e-07,
+      "loss": 1.4434,
+      "step": 2352
+    },
+    {
+      "epoch": 0.4883769198837692,
+      "grad_norm": 1.1344999282159138,
+      "learning_rate": 2.2657123644531495e-07,
+      "loss": 1.5362,
+      "step": 2353
+    },
+    {
+      "epoch": 0.4885844748858447,
+      "grad_norm": 0.8086824019107909,
+      "learning_rate": 2.2645394370644033e-07,
+      "loss": 1.5124,
+      "step": 2354
+    },
+    {
+      "epoch": 0.4887920298879203,
+      "grad_norm": 0.6734997115762309,
+      "learning_rate": 2.2633664822352015e-07,
+      "loss": 1.4235,
+      "step": 2355
+    },
+    {
+      "epoch": 0.48899958488999584,
+      "grad_norm": 0.926579261439982,
+      "learning_rate": 2.2621935004642542e-07,
+      "loss": 1.542,
+      "step": 2356
+    },
+    {
+      "epoch": 0.4892071398920714,
+      "grad_norm": 1.2807860653760692,
+      "learning_rate": 2.2610204922502816e-07,
+      "loss": 1.4462,
+      "step": 2357
+    },
+    {
+      "epoch": 0.48941469489414696,
+      "grad_norm": 0.6648117251299703,
+      "learning_rate": 2.2598474580920154e-07,
+      "loss": 1.4751,
+      "step": 2358
+    },
+    {
+      "epoch": 0.4896222498962225,
+      "grad_norm": 0.7426686724046041,
+      "learning_rate": 2.2586743984881992e-07,
+      "loss": 1.5033,
+      "step": 2359
+    },
+    {
+      "epoch": 0.489829804898298,
+      "grad_norm": 0.9369948294317266,
+      "learning_rate": 2.2575013139375864e-07,
+      "loss": 1.5495,
+      "step": 2360
+    },
+    {
+      "epoch": 0.4900373599003736,
+      "grad_norm": 0.7104454748457598,
+      "learning_rate": 2.2563282049389411e-07,
+      "loss": 1.5369,
+      "step": 2361
+    },
+    {
+      "epoch": 0.49024491490244915,
+      "grad_norm": 0.666662957904917,
+      "learning_rate": 2.255155071991039e-07,
+      "loss": 1.5705,
+      "step": 2362
+    },
+    {
+      "epoch": 0.4904524699045247,
+      "grad_norm": 0.76992241718709,
+      "learning_rate": 2.253981915592664e-07,
+      "loss": 1.5185,
+      "step": 2363
+    },
+    {
+      "epoch": 0.49066002490660027,
+      "grad_norm": 1.3641406282741546,
+      "learning_rate": 2.252808736242612e-07,
+      "loss": 1.4954,
+      "step": 2364
+    },
+    {
+      "epoch": 0.4908675799086758,
+      "grad_norm": 0.6534250671595311,
+      "learning_rate": 2.2516355344396873e-07,
+      "loss": 1.5424,
+      "step": 2365
+    },
+    {
+      "epoch": 0.49107513491075133,
+      "grad_norm": 1.3755655313249537,
+      "learning_rate": 2.2504623106827046e-07,
+      "loss": 1.5184,
+      "step": 2366
+    },
+    {
+      "epoch": 0.4912826899128269,
+      "grad_norm": 0.8016967943967964,
+      "learning_rate": 2.2492890654704862e-07,
+      "loss": 1.5138,
+      "step": 2367
+    },
+    {
+      "epoch": 0.49149024491490245,
+      "grad_norm": 0.7283545108228767,
+      "learning_rate": 2.2481157993018667e-07,
+      "loss": 1.4976,
+      "step": 2368
+    },
+    {
+      "epoch": 0.491697799916978,
+      "grad_norm": 0.7184022988714208,
+      "learning_rate": 2.2469425126756865e-07,
+      "loss": 1.5948,
+      "step": 2369
+    },
+    {
+      "epoch": 0.4919053549190536,
+      "grad_norm": 0.7466552043569856,
+      "learning_rate": 2.2457692060907952e-07,
+      "loss": 1.4851,
+      "step": 2370
+    },
+    {
+      "epoch": 0.4921129099211291,
+      "grad_norm": 0.7263271990122764,
+      "learning_rate": 2.2445958800460537e-07,
+      "loss": 1.4852,
+      "step": 2371
+    },
+    {
+      "epoch": 0.49232046492320464,
+      "grad_norm": 1.0766862289288277,
+      "learning_rate": 2.2434225350403278e-07,
+      "loss": 1.5401,
+      "step": 2372
+    },
+    {
+      "epoch": 0.4925280199252802,
+      "grad_norm": 0.9796289910443803,
+      "learning_rate": 2.2422491715724923e-07,
+      "loss": 1.4356,
+      "step": 2373
+    },
+    {
+      "epoch": 0.49273557492735576,
+      "grad_norm": 0.810236605123269,
+      "learning_rate": 2.241075790141431e-07,
+      "loss": 1.5169,
+      "step": 2374
+    },
+    {
+      "epoch": 0.4929431299294313,
+      "grad_norm": 1.0078173013194331,
+      "learning_rate": 2.2399023912460345e-07,
+      "loss": 1.509,
+      "step": 2375
+    },
+    {
+      "epoch": 0.4931506849315068,
+      "grad_norm": 1.2735091932440763,
+      "learning_rate": 2.2387289753852002e-07,
+      "loss": 1.5409,
+      "step": 2376
+    },
+    {
+      "epoch": 0.4933582399335824,
+      "grad_norm": 0.7829239059246687,
+      "learning_rate": 2.2375555430578332e-07,
+      "loss": 1.5498,
+      "step": 2377
+    },
+    {
+      "epoch": 0.49356579493565794,
+      "grad_norm": 0.7209757964077856,
+      "learning_rate": 2.2363820947628472e-07,
+      "loss": 1.4812,
+      "step": 2378
+    },
+    {
+      "epoch": 0.4937733499377335,
+      "grad_norm": 1.0953402268958026,
+      "learning_rate": 2.2352086309991605e-07,
+      "loss": 1.5725,
+      "step": 2379
+    },
+    {
+      "epoch": 0.49398090493980906,
+      "grad_norm": 0.7140442658739737,
+      "learning_rate": 2.2340351522656982e-07,
+      "loss": 1.5611,
+      "step": 2380
+    },
+    {
+      "epoch": 0.4941884599418846,
+      "grad_norm": 0.6439189073376079,
+      "learning_rate": 2.2328616590613927e-07,
+      "loss": 1.5731,
+      "step": 2381
+    },
+    {
+      "epoch": 0.4943960149439601,
+      "grad_norm": 0.8369451397426825,
+      "learning_rate": 2.2316881518851827e-07,
+      "loss": 1.5482,
+      "step": 2382
+    },
+    {
+      "epoch": 0.4946035699460357,
+      "grad_norm": 1.3822314719291078,
+      "learning_rate": 2.2305146312360113e-07,
+      "loss": 1.5507,
+      "step": 2383
+    },
+    {
+      "epoch": 0.49481112494811125,
+      "grad_norm": 0.7388750690291305,
+      "learning_rate": 2.229341097612829e-07,
+      "loss": 1.5374,
+      "step": 2384
+    },
+    {
+      "epoch": 0.4950186799501868,
+      "grad_norm": 0.7308634363153971,
+      "learning_rate": 2.2281675515145911e-07,
+      "loss": 1.4615,
+      "step": 2385
+    },
+    {
+      "epoch": 0.49522623495226237,
+      "grad_norm": 0.6561389768958427,
+      "learning_rate": 2.2269939934402579e-07,
+      "loss": 1.55,
+      "step": 2386
+    },
+    {
+      "epoch": 0.4954337899543379,
+      "grad_norm": 0.6491519946471916,
+      "learning_rate": 2.2258204238887952e-07,
+      "loss": 1.4425,
+      "step": 2387
+    },
+    {
+      "epoch": 0.49564134495641343,
+      "grad_norm": 0.7899172122781211,
+      "learning_rate": 2.2246468433591738e-07,
+      "loss": 1.4685,
+      "step": 2388
+    },
+    {
+      "epoch": 0.495848899958489,
+      "grad_norm": 0.9312882067013943,
+      "learning_rate": 2.223473252350369e-07,
+      "loss": 1.5024,
+      "step": 2389
+    },
+    {
+      "epoch": 0.49605645496056455,
+      "grad_norm": 0.92402736165806,
+      "learning_rate": 2.2222996513613607e-07,
+      "loss": 1.6208,
+      "step": 2390
+    },
+    {
+      "epoch": 0.4962640099626401,
+      "grad_norm": 0.8201952361478746,
+      "learning_rate": 2.2211260408911326e-07,
+      "loss": 1.5908,
+      "step": 2391
+    },
+    {
+      "epoch": 0.4964715649647157,
+      "grad_norm": 0.7475155499084326,
+      "learning_rate": 2.2199524214386725e-07,
+      "loss": 1.5815,
+      "step": 2392
+    },
+    {
+      "epoch": 0.4966791199667912,
+      "grad_norm": 1.888323899346627,
+      "learning_rate": 2.2187787935029729e-07,
+      "loss": 1.4394,
+      "step": 2393
+    },
+    {
+      "epoch": 0.49688667496886674,
+      "grad_norm": 0.6832350446707858,
+      "learning_rate": 2.2176051575830287e-07,
+      "loss": 1.4805,
+      "step": 2394
+    },
+    {
+      "epoch": 0.4970942299709423,
+      "grad_norm": 0.8165171691449675,
+      "learning_rate": 2.2164315141778385e-07,
+      "loss": 1.4602,
+      "step": 2395
+    },
+    {
+      "epoch": 0.49730178497301786,
+      "grad_norm": 1.0322843220373226,
+      "learning_rate": 2.2152578637864052e-07,
+      "loss": 1.5236,
+      "step": 2396
+    },
+    {
+      "epoch": 0.4975093399750934,
+      "grad_norm": 0.7792230662731928,
+      "learning_rate": 2.2140842069077332e-07,
+      "loss": 1.5274,
+      "step": 2397
+    },
+    {
+      "epoch": 0.4977168949771689,
+      "grad_norm": 0.7198823116536779,
+      "learning_rate": 2.2129105440408305e-07,
+      "loss": 1.4906,
+      "step": 2398
+    },
+    {
+      "epoch": 0.4979244499792445,
+      "grad_norm": 0.6890951297207466,
+      "learning_rate": 2.211736875684707e-07,
+      "loss": 1.5051,
+      "step": 2399
+    },
+    {
+      "epoch": 0.49813200498132004,
+      "grad_norm": 0.6984522177679134,
+      "learning_rate": 2.2105632023383754e-07,
+      "loss": 1.5105,
+      "step": 2400
+    },
+    {
+      "epoch": 0.4983395599833956,
+      "grad_norm": 0.6524769170303281,
+      "learning_rate": 2.209389524500851e-07,
+      "loss": 1.5476,
+      "step": 2401
+    },
+    {
+      "epoch": 0.49854711498547116,
+      "grad_norm": 0.851430811558582,
+      "learning_rate": 2.2082158426711498e-07,
+      "loss": 1.5251,
+      "step": 2402
+    },
+    {
+      "epoch": 0.4987546699875467,
+      "grad_norm": 1.8419335179908904,
+      "learning_rate": 2.2070421573482898e-07,
+      "loss": 1.4877,
+      "step": 2403
+    },
+    {
+      "epoch": 0.4989622249896222,
+      "grad_norm": 0.8383365556589939,
+      "learning_rate": 2.205868469031292e-07,
+      "loss": 1.4606,
+      "step": 2404
+    },
+    {
+      "epoch": 0.4991697799916978,
+      "grad_norm": 1.0272126896521991,
+      "learning_rate": 2.204694778219177e-07,
+      "loss": 1.4775,
+      "step": 2405
+    },
+    {
+      "epoch": 0.49937733499377335,
+      "grad_norm": 1.0330333576597883,
+      "learning_rate": 2.2035210854109672e-07,
+      "loss": 1.5555,
+      "step": 2406
+    },
+    {
+      "epoch": 0.4995848899958489,
+      "grad_norm": 0.7326413275343511,
+      "learning_rate": 2.2023473911056852e-07,
+      "loss": 1.536,
+      "step": 2407
+    },
+    {
+      "epoch": 0.49979244499792447,
+      "grad_norm": 0.8270757016848693,
+      "learning_rate": 2.2011736958023546e-07,
+      "loss": 1.5758,
+      "step": 2408
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.9499466976856001,
+      "learning_rate": 2.2e-07,
+      "loss": 1.5102,
+      "step": 2409
+    },
+    {
+      "epoch": 0.5002075550020756,
+      "grad_norm": 0.7388147924857219,
+      "learning_rate": 2.198826304197645e-07,
+      "loss": 1.5057,
+      "step": 2410
+    },
+    {
+      "epoch": 0.5004151100041511,
+      "grad_norm": 1.8802440139792262,
+      "learning_rate": 2.197652608894315e-07,
+      "loss": 1.5136,
+      "step": 2411
+    },
+    {
+      "epoch": 0.5006226650062267,
+      "grad_norm": 0.744049750632982,
+      "learning_rate": 2.1964789145890327e-07,
+      "loss": 1.5254,
+      "step": 2412
+    },
+    {
+      "epoch": 0.5008302200083022,
+      "grad_norm": 0.9103896865032503,
+      "learning_rate": 2.1953052217808227e-07,
+      "loss": 1.6083,
+      "step": 2413
+    },
+    {
+      "epoch": 0.5010377750103777,
+      "grad_norm": 15.482267004015949,
+      "learning_rate": 2.194131530968708e-07,
+      "loss": 1.5511,
+      "step": 2414
+    },
+    {
+      "epoch": 0.5012453300124533,
+      "grad_norm": 0.6691221392789174,
+      "learning_rate": 2.1929578426517104e-07,
+      "loss": 1.4803,
+      "step": 2415
+    },
+    {
+      "epoch": 0.5014528850145289,
+      "grad_norm": 0.8649620252717285,
+      "learning_rate": 2.1917841573288504e-07,
+      "loss": 1.5206,
+      "step": 2416
+    },
+    {
+      "epoch": 0.5016604400166044,
+      "grad_norm": 0.7945827938797361,
+      "learning_rate": 2.1906104754991496e-07,
+      "loss": 1.5048,
+      "step": 2417
+    },
+    {
+      "epoch": 0.50186799501868,
+      "grad_norm": 0.8105239033777475,
+      "learning_rate": 2.1894367976616248e-07,
+      "loss": 1.4898,
+      "step": 2418
+    },
+    {
+      "epoch": 0.5020755500207555,
+      "grad_norm": 0.7066074602117844,
+      "learning_rate": 2.1882631243152932e-07,
+      "loss": 1.4823,
+      "step": 2419
+    },
+    {
+      "epoch": 0.502283105022831,
+      "grad_norm": 1.6046386942377815,
+      "learning_rate": 2.1870894559591702e-07,
+      "loss": 1.5445,
+      "step": 2420
+    },
+    {
+      "epoch": 0.5024906600249066,
+      "grad_norm": 0.9681527331244452,
+      "learning_rate": 2.185915793092267e-07,
+      "loss": 1.5275,
+      "step": 2421
+    },
+    {
+      "epoch": 0.5026982150269822,
+      "grad_norm": 0.708434096699083,
+      "learning_rate": 2.1847421362135945e-07,
+      "loss": 1.5532,
+      "step": 2422
+    },
+    {
+      "epoch": 0.5029057700290577,
+      "grad_norm": 0.713965479785203,
+      "learning_rate": 2.1835684858221618e-07,
+      "loss": 1.4534,
+      "step": 2423
+    },
+    {
+      "epoch": 0.5031133250311333,
+      "grad_norm": 1.2962602570015131,
+      "learning_rate": 2.1823948424169715e-07,
+      "loss": 1.5067,
+      "step": 2424
+    },
+    {
+      "epoch": 0.5033208800332089,
+      "grad_norm": 0.7339499594838154,
+      "learning_rate": 2.181221206497027e-07,
+      "loss": 1.667,
+      "step": 2425
+    },
+    {
+      "epoch": 0.5035284350352843,
+      "grad_norm": 0.6634648430427076,
+      "learning_rate": 2.1800475785613277e-07,
+      "loss": 1.5251,
+      "step": 2426
+    },
+    {
+      "epoch": 0.5037359900373599,
+      "grad_norm": 3.0871159735918665,
+      "learning_rate": 2.1788739591088677e-07,
+      "loss": 1.5387,
+      "step": 2427
+    },
+    {
+      "epoch": 0.5039435450394355,
+      "grad_norm": 0.7638438643105517,
+      "learning_rate": 2.177700348638639e-07,
+      "loss": 1.5176,
+      "step": 2428
+    },
+    {
+      "epoch": 0.504151100041511,
+      "grad_norm": 0.8473858309718701,
+      "learning_rate": 2.1765267476496308e-07,
+      "loss": 1.5254,
+      "step": 2429
+    },
+    {
+      "epoch": 0.5043586550435866,
+      "grad_norm": 0.8177431453202664,
+      "learning_rate": 2.175353156640826e-07,
+      "loss": 1.5109,
+      "step": 2430
+    },
+    {
+      "epoch": 0.5045662100456622,
+      "grad_norm": 1.0904511203969505,
+      "learning_rate": 2.174179576111205e-07,
+      "loss": 1.4932,
+      "step": 2431
+    },
+    {
+      "epoch": 0.5047737650477376,
+      "grad_norm": 0.6510980267204981,
+      "learning_rate": 2.1730060065597424e-07,
+      "loss": 1.4741,
+      "step": 2432
+    },
+    {
+      "epoch": 0.5049813200498132,
+      "grad_norm": 0.9533777191581878,
+      "learning_rate": 2.1718324484854088e-07,
+      "loss": 1.5348,
+      "step": 2433
+    },
+    {
+      "epoch": 0.5051888750518887,
+      "grad_norm": 0.6783813190698431,
+      "learning_rate": 2.1706589023871714e-07,
+      "loss": 1.5372,
+      "step": 2434
+    },
+    {
+      "epoch": 0.5053964300539643,
+      "grad_norm": 0.7938763627829857,
+      "learning_rate": 2.169485368763989e-07,
+      "loss": 1.5557,
+      "step": 2435
+    },
+    {
+      "epoch": 0.5056039850560399,
+      "grad_norm": 0.7996241166136091,
+      "learning_rate": 2.1683118481148175e-07,
+      "loss": 1.5017,
+      "step": 2436
+    },
+    {
+      "epoch": 0.5058115400581153,
+      "grad_norm": 0.884268792283763,
+      "learning_rate": 2.1671383409386075e-07,
+      "loss": 1.5164,
+      "step": 2437
+    },
+    {
+      "epoch": 0.5060190950601909,
+      "grad_norm": 0.6716713247883358,
+      "learning_rate": 2.165964847734302e-07,
+      "loss": 1.4735,
+      "step": 2438
+    },
+    {
+      "epoch": 0.5062266500622665,
+      "grad_norm": 1.1598691760507256,
+      "learning_rate": 2.1647913690008398e-07,
+      "loss": 1.4989,
+      "step": 2439
+    },
+    {
+      "epoch": 0.506434205064342,
+      "grad_norm": 1.2055544529619706,
+      "learning_rate": 2.1636179052371525e-07,
+      "loss": 1.4991,
+      "step": 2440
+    },
+    {
+      "epoch": 0.5066417600664176,
+      "grad_norm": 0.7595865867619096,
+      "learning_rate": 2.1624444569421665e-07,
+      "loss": 1.5393,
+      "step": 2441
+    },
+    {
+      "epoch": 0.5068493150684932,
+      "grad_norm": 0.6702215983733609,
+      "learning_rate": 2.1612710246148e-07,
+      "loss": 1.5017,
+      "step": 2442
+    },
+    {
+      "epoch": 0.5070568700705687,
+      "grad_norm": 0.9219748956820397,
+      "learning_rate": 2.1600976087539663e-07,
+      "loss": 1.5682,
+      "step": 2443
+    },
+    {
+      "epoch": 0.5072644250726442,
+      "grad_norm": 0.6697674762341783,
+      "learning_rate": 2.1589242098585688e-07,
+      "loss": 1.5021,
+      "step": 2444
+    },
+    {
+      "epoch": 0.5074719800747198,
+      "grad_norm": 0.6356975413297162,
+      "learning_rate": 2.1577508284275074e-07,
+      "loss": 1.4923,
+      "step": 2445
+    },
+    {
+      "epoch": 0.5076795350767953,
+      "grad_norm": 1.473554865655388,
+      "learning_rate": 2.156577464959673e-07,
+      "loss": 1.4707,
+      "step": 2446
+    },
+    {
+      "epoch": 0.5078870900788709,
+      "grad_norm": 0.8449252727151781,
+      "learning_rate": 2.1554041199539465e-07,
+      "loss": 1.5072,
+      "step": 2447
+    },
+    {
+      "epoch": 0.5080946450809465,
+      "grad_norm": 0.9176107996429508,
+      "learning_rate": 2.1542307939092043e-07,
+      "loss": 1.4434,
+      "step": 2448
+    },
+    {
+      "epoch": 0.508302200083022,
+      "grad_norm": 0.9618820285083549,
+      "learning_rate": 2.1530574873243142e-07,
+      "loss": 1.4842,
+      "step": 2449
+    },
+    {
+      "epoch": 0.5085097550850975,
+      "grad_norm": 0.9252681176960654,
+      "learning_rate": 2.1518842006981335e-07,
+      "loss": 1.5339,
+      "step": 2450
+    },
+    {
+      "epoch": 0.5087173100871731,
+      "grad_norm": 0.7061334241468691,
+      "learning_rate": 2.1507109345295135e-07,
+      "loss": 1.5524,
+      "step": 2451
+    },
+    {
+      "epoch": 0.5089248650892486,
+      "grad_norm": 0.7019032867443913,
+      "learning_rate": 2.149537689317296e-07,
+      "loss": 1.4591,
+      "step": 2452
+    },
+    {
+      "epoch": 0.5091324200913242,
+      "grad_norm": 0.8982633216221373,
+      "learning_rate": 2.148364465560313e-07,
+      "loss": 1.5385,
+      "step": 2453
+    },
+    {
+      "epoch": 0.5093399750933998,
+      "grad_norm": 0.766863641714324,
+      "learning_rate": 2.1471912637573877e-07,
+      "loss": 1.5031,
+      "step": 2454
+    },
+    {
+      "epoch": 0.5095475300954753,
+      "grad_norm": 0.7062281624244494,
+      "learning_rate": 2.1460180844073358e-07,
+      "loss": 1.4823,
+      "step": 2455
+    },
+    {
+      "epoch": 0.5097550850975509,
+      "grad_norm": 0.8723446422793907,
+      "learning_rate": 2.144844928008961e-07,
+      "loss": 1.5528,
+      "step": 2456
+    },
+    {
+      "epoch": 0.5099626400996264,
+      "grad_norm": 0.7422745245421455,
+      "learning_rate": 2.143671795061059e-07,
+      "loss": 1.5478,
+      "step": 2457
+    },
+    {
+      "epoch": 0.5101701951017019,
+      "grad_norm": 0.7799161085115599,
+      "learning_rate": 2.142498686062414e-07,
+      "loss": 1.5311,
+      "step": 2458
+    },
+    {
+      "epoch": 0.5103777501037775,
+      "grad_norm": 0.6945089826557154,
+      "learning_rate": 2.1413256015118008e-07,
+      "loss": 1.5037,
+      "step": 2459
+    },
+    {
+      "epoch": 0.5105853051058531,
+      "grad_norm": 0.7138520715435961,
+      "learning_rate": 2.1401525419079846e-07,
+      "loss": 1.491,
+      "step": 2460
+    },
+    {
+      "epoch": 0.5107928601079286,
+      "grad_norm": 0.6366437637946678,
+      "learning_rate": 2.1389795077497192e-07,
+      "loss": 1.4952,
+      "step": 2461
+    },
+    {
+      "epoch": 0.5110004151100042,
+      "grad_norm": 0.7726024325988109,
+      "learning_rate": 2.1378064995357458e-07,
+      "loss": 1.4565,
+      "step": 2462
+    },
+    {
+      "epoch": 0.5112079701120797,
+      "grad_norm": 0.8521489148994181,
+      "learning_rate": 2.1366335177647982e-07,
+      "loss": 1.5389,
+      "step": 2463
+    },
+    {
+      "epoch": 0.5114155251141552,
+      "grad_norm": 1.21942086179602,
+      "learning_rate": 2.1354605629355972e-07,
+      "loss": 1.5497,
+      "step": 2464
+    },
+    {
+      "epoch": 0.5116230801162308,
+      "grad_norm": 0.8609674445921635,
+      "learning_rate": 2.1342876355468507e-07,
+      "loss": 1.5593,
+      "step": 2465
+    },
+    {
+      "epoch": 0.5118306351183064,
+      "grad_norm": 3.2921579618116916,
+      "learning_rate": 2.1331147360972567e-07,
+      "loss": 1.4479,
+      "step": 2466
+    },
+    {
+      "epoch": 0.5120381901203819,
+      "grad_norm": 0.7598708051121843,
+      "learning_rate": 2.131941865085502e-07,
+      "loss": 1.5294,
+      "step": 2467
+    },
+    {
+      "epoch": 0.5122457451224575,
+      "grad_norm": 0.6713978721464637,
+      "learning_rate": 2.1307690230102594e-07,
+      "loss": 1.5361,
+      "step": 2468
+    },
+    {
+      "epoch": 0.512453300124533,
+      "grad_norm": 0.8762878443709896,
+      "learning_rate": 2.1295962103701894e-07,
+      "loss": 1.497,
+      "step": 2469
+    },
+    {
+      "epoch": 0.5126608551266085,
+      "grad_norm": 0.9048152481656528,
+      "learning_rate": 2.1284234276639426e-07,
+      "loss": 1.4793,
+      "step": 2470
+    },
+    {
+      "epoch": 0.5128684101286841,
+      "grad_norm": 0.7156511024565448,
+      "learning_rate": 2.1272506753901534e-07,
+      "loss": 1.5092,
+      "step": 2471
+    },
+    {
+      "epoch": 0.5130759651307597,
+      "grad_norm": 0.696532616525458,
+      "learning_rate": 2.1260779540474457e-07,
+      "loss": 1.5417,
+      "step": 2472
+    },
+    {
+      "epoch": 0.5132835201328352,
+      "grad_norm": 0.6927337066286205,
+      "learning_rate": 2.1249052641344302e-07,
+      "loss": 1.5053,
+      "step": 2473
+    },
+    {
+      "epoch": 0.5134910751349108,
+      "grad_norm": 0.7442089820585011,
+      "learning_rate": 2.1237326061497017e-07,
+      "loss": 1.4757,
+      "step": 2474
+    },
+    {
+      "epoch": 0.5136986301369864,
+      "grad_norm": 0.8246407513578571,
+      "learning_rate": 2.1225599805918448e-07,
+      "loss": 1.5193,
+      "step": 2475
+    },
+    {
+      "epoch": 0.5139061851390618,
+      "grad_norm": 0.8641013857303477,
+      "learning_rate": 2.1213873879594288e-07,
+      "loss": 1.5032,
+      "step": 2476
+    },
+    {
+      "epoch": 0.5141137401411374,
+      "grad_norm": 2.247438914337527,
+      "learning_rate": 2.1202148287510075e-07,
+      "loss": 1.4839,
+      "step": 2477
+    },
+    {
+      "epoch": 0.5143212951432129,
+      "grad_norm": 0.6932341896726947,
+      "learning_rate": 2.119042303465124e-07,
+      "loss": 1.5523,
+      "step": 2478
+    },
+    {
+      "epoch": 0.5145288501452885,
+      "grad_norm": 0.7732484587753624,
+      "learning_rate": 2.1178698126003027e-07,
+      "loss": 1.4939,
+      "step": 2479
+    },
+    {
+      "epoch": 0.5147364051473641,
+      "grad_norm": 0.9041529393067483,
+      "learning_rate": 2.1166973566550564e-07,
+      "loss": 1.5704,
+      "step": 2480
+    },
+    {
+      "epoch": 0.5149439601494396,
+      "grad_norm": 0.8410247703754223,
+      "learning_rate": 2.1155249361278832e-07,
+      "loss": 1.5702,
+      "step": 2481
+    },
+    {
+      "epoch": 0.5151515151515151,
+      "grad_norm": 0.8103445323108253,
+      "learning_rate": 2.114352551517264e-07,
+      "loss": 1.5048,
+      "step": 2482
+    },
+    {
+      "epoch": 0.5153590701535907,
+      "grad_norm": 1.1187607543104976,
+      "learning_rate": 2.113180203321666e-07,
+      "loss": 1.5225,
+      "step": 2483
+    },
+    {
+      "epoch": 0.5155666251556662,
+      "grad_norm": 0.7353532302101807,
+      "learning_rate": 2.1120078920395414e-07,
+      "loss": 1.4607,
+      "step": 2484
+    },
+    {
+      "epoch": 0.5157741801577418,
+      "grad_norm": 1.0769957396925915,
+      "learning_rate": 2.1108356181693242e-07,
+      "loss": 1.4906,
+      "step": 2485
+    },
+    {
+      "epoch": 0.5159817351598174,
+      "grad_norm": 1.527339486433276,
+      "learning_rate": 2.109663382209435e-07,
+      "loss": 1.5283,
+      "step": 2486
+    },
+    {
+      "epoch": 0.5161892901618929,
+      "grad_norm": 0.7213794450479653,
+      "learning_rate": 2.1084911846582782e-07,
+      "loss": 1.581,
+      "step": 2487
+    },
+    {
+      "epoch": 0.5163968451639684,
+      "grad_norm": 1.005543951952498,
+      "learning_rate": 2.10731902601424e-07,
+      "loss": 1.5208,
+      "step": 2488
+    },
+    {
+      "epoch": 0.516604400166044,
+      "grad_norm": 0.826196582424875,
+      "learning_rate": 2.1061469067756907e-07,
+      "loss": 1.4791,
+      "step": 2489
+    },
+    {
+      "epoch": 0.5168119551681195,
+      "grad_norm": 1.0287051244940715,
+      "learning_rate": 2.1049748274409863e-07,
+      "loss": 1.5513,
+      "step": 2490
+    },
+    {
+      "epoch": 0.5170195101701951,
+      "grad_norm": 0.9585164554639976,
+      "learning_rate": 2.1038027885084612e-07,
+      "loss": 1.5184,
+      "step": 2491
+    },
+    {
+      "epoch": 0.5172270651722707,
+      "grad_norm": 0.7865200612734766,
+      "learning_rate": 2.1026307904764367e-07,
+      "loss": 1.4974,
+      "step": 2492
+    },
+    {
+      "epoch": 0.5174346201743462,
+      "grad_norm": 1.3845069727907573,
+      "learning_rate": 2.1014588338432157e-07,
+      "loss": 1.4595,
+      "step": 2493
+    },
+    {
+      "epoch": 0.5176421751764217,
+      "grad_norm": 0.690391057181294,
+      "learning_rate": 2.1002869191070825e-07,
+      "loss": 1.5095,
+      "step": 2494
+    },
+    {
+      "epoch": 0.5178497301784973,
+      "grad_norm": 0.758964411680427,
+      "learning_rate": 2.099115046766303e-07,
+      "loss": 1.4986,
+      "step": 2495
+    },
+    {
+      "epoch": 0.5180572851805728,
+      "grad_norm": 0.6995193011205069,
+      "learning_rate": 2.0979432173191284e-07,
+      "loss": 1.5073,
+      "step": 2496
+    },
+    {
+      "epoch": 0.5182648401826484,
+      "grad_norm": 0.6745500359558347,
+      "learning_rate": 2.0967714312637877e-07,
+      "loss": 1.5596,
+      "step": 2497
+    },
+    {
+      "epoch": 0.518472395184724,
+      "grad_norm": 8.379139661600988,
+      "learning_rate": 2.0955996890984938e-07,
+      "loss": 1.4595,
+      "step": 2498
+    },
+    {
+      "epoch": 0.5186799501867995,
+      "grad_norm": 0.7309652404488921,
+      "learning_rate": 2.0944279913214414e-07,
+      "loss": 1.5826,
+      "step": 2499
+    },
+    {
+      "epoch": 0.518887505188875,
+      "grad_norm": 0.7936397903089487,
+      "learning_rate": 2.0932563384308032e-07,
+      "loss": 1.5901,
+      "step": 2500
+    },
+    {
+      "epoch": 0.5190950601909506,
+      "grad_norm": 1.2203690484503538,
+      "learning_rate": 2.092084730924736e-07,
+      "loss": 1.4968,
+      "step": 2501
+    },
+    {
+      "epoch": 0.5193026151930261,
+      "grad_norm": 0.8466712403175027,
+      "learning_rate": 2.0909131693013772e-07,
+      "loss": 1.5396,
+      "step": 2502
+    },
+    {
+      "epoch": 0.5195101701951017,
+      "grad_norm": 0.6963177397067928,
+      "learning_rate": 2.0897416540588418e-07,
+      "loss": 1.4834,
+      "step": 2503
+    },
+    {
+      "epoch": 0.5197177251971773,
+      "grad_norm": 0.6624086497063163,
+      "learning_rate": 2.088570185695228e-07,
+      "loss": 1.5081,
+      "step": 2504
+    },
+    {
+      "epoch": 0.5199252801992528,
+      "grad_norm": 0.6258416757370212,
+      "learning_rate": 2.087398764708614e-07,
+      "loss": 1.4706,
+      "step": 2505
+    },
+    {
+      "epoch": 0.5201328352013284,
+      "grad_norm": 0.9068187223454897,
+      "learning_rate": 2.0862273915970548e-07,
+      "loss": 1.5718,
+      "step": 2506
+    },
+    {
+      "epoch": 0.520340390203404,
+      "grad_norm": 0.7166672898104127,
+      "learning_rate": 2.085056066858588e-07,
+      "loss": 1.558,
+      "step": 2507
+    },
+    {
+      "epoch": 0.5205479452054794,
+      "grad_norm": 0.9092022844852634,
+      "learning_rate": 2.083884790991231e-07,
+      "loss": 1.588,
+      "step": 2508
+    },
+    {
+      "epoch": 0.520755500207555,
+      "grad_norm": 0.7282167983136697,
+      "learning_rate": 2.0827135644929771e-07,
+      "loss": 1.4866,
+      "step": 2509
+    },
+    {
+      "epoch": 0.5209630552096306,
+      "grad_norm": 1.4645709270681502,
+      "learning_rate": 2.0815423878618024e-07,
+      "loss": 1.4826,
+      "step": 2510
+    },
+    {
+      "epoch": 0.5211706102117061,
+      "grad_norm": 1.1634417026671537,
+      "learning_rate": 2.0803712615956598e-07,
+      "loss": 1.5624,
+      "step": 2511
+    },
+    {
+      "epoch": 0.5213781652137817,
+      "grad_norm": 0.5994248350347131,
+      "learning_rate": 2.07920018619248e-07,
+      "loss": 1.4907,
+      "step": 2512
+    },
+    {
+      "epoch": 0.5215857202158573,
+      "grad_norm": 0.854130339608346,
+      "learning_rate": 2.0780291621501745e-07,
+      "loss": 1.4797,
+      "step": 2513
+    },
+    {
+      "epoch": 0.5217932752179327,
+      "grad_norm": 1.4762152580693124,
+      "learning_rate": 2.0768581899666314e-07,
+      "loss": 1.4978,
+      "step": 2514
+    },
+    {
+      "epoch": 0.5220008302200083,
+      "grad_norm": 0.70458294709752,
+      "learning_rate": 2.075687270139716e-07,
+      "loss": 1.5031,
+      "step": 2515
+    },
+    {
+      "epoch": 0.5222083852220839,
+      "grad_norm": 0.6863497182521193,
+      "learning_rate": 2.0745164031672734e-07,
+      "loss": 1.5239,
+      "step": 2516
+    },
+    {
+      "epoch": 0.5224159402241594,
+      "grad_norm": 1.2072481621083504,
+      "learning_rate": 2.073345589547125e-07,
+      "loss": 1.5273,
+      "step": 2517
+    },
+    {
+      "epoch": 0.522623495226235,
+      "grad_norm": 0.778691246550858,
+      "learning_rate": 2.0721748297770691e-07,
+      "loss": 1.6001,
+      "step": 2518
+    },
+    {
+      "epoch": 0.5228310502283106,
+      "grad_norm": 1.1136332407815033,
+      "learning_rate": 2.0710041243548818e-07,
+      "loss": 1.5537,
+      "step": 2519
+    },
+    {
+      "epoch": 0.523038605230386,
+      "grad_norm": 0.8649465687966319,
+      "learning_rate": 2.0698334737783166e-07,
+      "loss": 1.4557,
+      "step": 2520
+    },
+    {
+      "epoch": 0.5232461602324616,
+      "grad_norm": 1.1643855257735907,
+      "learning_rate": 2.0686628785451027e-07,
+      "loss": 1.5895,
+      "step": 2521
+    },
+    {
+      "epoch": 0.5234537152345371,
+      "grad_norm": 1.1166779145001688,
+      "learning_rate": 2.0674923391529458e-07,
+      "loss": 1.5445,
+      "step": 2522
+    },
+    {
+      "epoch": 0.5236612702366127,
+      "grad_norm": 1.478756010933752,
+      "learning_rate": 2.0663218560995285e-07,
+      "loss": 1.5654,
+      "step": 2523
+    },
+    {
+      "epoch": 0.5238688252386883,
+      "grad_norm": 0.6516633123967659,
+      "learning_rate": 2.0651514298825087e-07,
+      "loss": 1.5366,
+      "step": 2524
+    },
+    {
+      "epoch": 0.5240763802407638,
+      "grad_norm": 0.7161622602381621,
+      "learning_rate": 2.0639810609995214e-07,
+      "loss": 1.5313,
+      "step": 2525
+    },
+    {
+      "epoch": 0.5242839352428393,
+      "grad_norm": 0.7078844579987807,
+      "learning_rate": 2.0628107499481756e-07,
+      "loss": 1.5615,
+      "step": 2526
+    },
+    {
+      "epoch": 0.5244914902449149,
+      "grad_norm": 0.6908546622783227,
+      "learning_rate": 2.0616404972260565e-07,
+      "loss": 1.5223,
+      "step": 2527
+    },
+    {
+      "epoch": 0.5246990452469904,
+      "grad_norm": 0.9980560645697012,
+      "learning_rate": 2.0604703033307257e-07,
+      "loss": 1.5172,
+      "step": 2528
+    },
+    {
+      "epoch": 0.524906600249066,
+      "grad_norm": 0.7512472252202701,
+      "learning_rate": 2.0593001687597167e-07,
+      "loss": 1.5654,
+      "step": 2529
+    },
+    {
+      "epoch": 0.5251141552511416,
+      "grad_norm": 1.579620471724761,
+      "learning_rate": 2.0581300940105403e-07,
+      "loss": 1.4265,
+      "step": 2530
+    },
+    {
+      "epoch": 0.5253217102532171,
+      "grad_norm": 0.7751885985002903,
+      "learning_rate": 2.056960079580683e-07,
+      "loss": 1.4854,
+      "step": 2531
+    },
+    {
+      "epoch": 0.5255292652552926,
+      "grad_norm": 0.655931305698934,
+      "learning_rate": 2.055790125967601e-07,
+      "loss": 1.5715,
+      "step": 2532
+    },
+    {
+      "epoch": 0.5257368202573682,
+      "grad_norm": 1.783292058223657,
+      "learning_rate": 2.0546202336687291e-07,
+      "loss": 1.5745,
+      "step": 2533
+    },
+    {
+      "epoch": 0.5259443752594437,
+      "grad_norm": 0.787009686242263,
+      "learning_rate": 2.0534504031814746e-07,
+      "loss": 1.52,
+      "step": 2534
+    },
+    {
+      "epoch": 0.5261519302615193,
+      "grad_norm": 0.959048562325366,
+      "learning_rate": 2.0522806350032175e-07,
+      "loss": 1.5391,
+      "step": 2535
+    },
+    {
+      "epoch": 0.5263594852635949,
+      "grad_norm": 0.8527148164380743,
+      "learning_rate": 2.0511109296313126e-07,
+      "loss": 1.6084,
+      "step": 2536
+    },
+    {
+      "epoch": 0.5265670402656704,
+      "grad_norm": 0.7107121194407611,
+      "learning_rate": 2.049941287563089e-07,
+      "loss": 1.5214,
+      "step": 2537
+    },
+    {
+      "epoch": 0.526774595267746,
+      "grad_norm": 1.2482997698995304,
+      "learning_rate": 2.0487717092958446e-07,
+      "loss": 1.4906,
+      "step": 2538
+    },
+    {
+      "epoch": 0.5269821502698215,
+      "grad_norm": 0.7072247277798667,
+      "learning_rate": 2.0476021953268546e-07,
+      "loss": 1.5156,
+      "step": 2539
+    },
+    {
+      "epoch": 0.527189705271897,
+      "grad_norm": 0.7728478784544087,
+      "learning_rate": 2.0464327461533664e-07,
+      "loss": 1.558,
+      "step": 2540
+    },
+    {
+      "epoch": 0.5273972602739726,
+      "grad_norm": 0.8284025800637722,
+      "learning_rate": 2.0452633622725964e-07,
+      "loss": 1.4653,
+      "step": 2541
+    },
+    {
+      "epoch": 0.5276048152760482,
+      "grad_norm": 0.7864178193578892,
+      "learning_rate": 2.0440940441817368e-07,
+      "loss": 1.5375,
+      "step": 2542
+    },
+    {
+      "epoch": 0.5278123702781237,
+      "grad_norm": 0.983310189126021,
+      "learning_rate": 2.0429247923779513e-07,
+      "loss": 1.4949,
+      "step": 2543
+    },
+    {
+      "epoch": 0.5280199252801993,
+      "grad_norm": 0.7467909625989806,
+      "learning_rate": 2.0417556073583735e-07,
+      "loss": 1.5651,
+      "step": 2544
+    },
+    {
+      "epoch": 0.5282274802822748,
+      "grad_norm": 0.9296866953037852,
+      "learning_rate": 2.0405864896201103e-07,
+      "loss": 1.4697,
+      "step": 2545
+    },
+    {
+      "epoch": 0.5284350352843503,
+      "grad_norm": 1.22643417988263,
+      "learning_rate": 2.0394174396602398e-07,
+      "loss": 1.5398,
+      "step": 2546
+    },
+    {
+      "epoch": 0.5286425902864259,
+      "grad_norm": 0.8074935583982986,
+      "learning_rate": 2.0382484579758103e-07,
+      "loss": 1.5069,
+      "step": 2547
+    },
+    {
+      "epoch": 0.5288501452885015,
+      "grad_norm": 0.7339167135498497,
+      "learning_rate": 2.0370795450638423e-07,
+      "loss": 1.4936,
+      "step": 2548
+    },
+    {
+      "epoch": 0.529057700290577,
+      "grad_norm": 0.8308385068673543,
+      "learning_rate": 2.035910701421327e-07,
+      "loss": 1.5502,
+      "step": 2549
+    },
+    {
+      "epoch": 0.5292652552926526,
+      "grad_norm": 0.9291711743124204,
+      "learning_rate": 2.0347419275452244e-07,
+      "loss": 1.5193,
+      "step": 2550
+    },
+    {
+      "epoch": 0.5294728102947281,
+      "grad_norm": 0.7043226671689163,
+      "learning_rate": 2.0335732239324668e-07,
+      "loss": 1.5815,
+      "step": 2551
+    },
+    {
+      "epoch": 0.5296803652968036,
+      "grad_norm": 0.6734361721890403,
+      "learning_rate": 2.032404591079957e-07,
+      "loss": 1.6102,
+      "step": 2552
+    },
+    {
+      "epoch": 0.5298879202988792,
+      "grad_norm": 0.6880087259342141,
+      "learning_rate": 2.0312360294845649e-07,
+      "loss": 1.5458,
+      "step": 2553
+    },
+    {
+      "epoch": 0.5300954753009548,
+      "grad_norm": 1.2872455794096174,
+      "learning_rate": 2.0300675396431325e-07,
+      "loss": 1.5311,
+      "step": 2554
+    },
+    {
+      "epoch": 0.5303030303030303,
+      "grad_norm": 0.7272249077581653,
+      "learning_rate": 2.0288991220524716e-07,
+      "loss": 1.5008,
+      "step": 2555
+    },
+    {
+      "epoch": 0.5305105853051059,
+      "grad_norm": 0.7762425445406153,
+      "learning_rate": 2.0277307772093608e-07,
+      "loss": 1.4858,
+      "step": 2556
+    },
+    {
+      "epoch": 0.5307181403071815,
+      "grad_norm": 0.713950185719675,
+      "learning_rate": 2.02656250561055e-07,
+      "loss": 1.5028,
+      "step": 2557
+    },
+    {
+      "epoch": 0.5309256953092569,
+      "grad_norm": 0.7962183171099543,
+      "learning_rate": 2.0253943077527582e-07,
+      "loss": 1.5346,
+      "step": 2558
+    },
+    {
+      "epoch": 0.5311332503113325,
+      "grad_norm": 0.6455917051944241,
+      "learning_rate": 2.02422618413267e-07,
+      "loss": 1.5303,
+      "step": 2559
+    },
+    {
+      "epoch": 0.5313408053134081,
+      "grad_norm": 0.7354275031634365,
+      "learning_rate": 2.0230581352469424e-07,
+      "loss": 1.4789,
+      "step": 2560
+    },
+    {
+      "epoch": 0.5315483603154836,
+      "grad_norm": 0.6923837539251685,
+      "learning_rate": 2.0218901615921982e-07,
+      "loss": 1.4758,
+      "step": 2561
+    },
+    {
+      "epoch": 0.5317559153175592,
+      "grad_norm": 1.069232011110929,
+      "learning_rate": 2.0207222636650286e-07,
+      "loss": 1.4795,
+      "step": 2562
+    },
+    {
+      "epoch": 0.5319634703196348,
+      "grad_norm": 0.7616210867512286,
+      "learning_rate": 2.019554441961993e-07,
+      "loss": 1.5087,
+      "step": 2563
+    },
+    {
+      "epoch": 0.5321710253217102,
+      "grad_norm": 0.688797360280766,
+      "learning_rate": 2.018386696979618e-07,
+      "loss": 1.4573,
+      "step": 2564
+    },
+    {
+      "epoch": 0.5323785803237858,
+      "grad_norm": 0.7302463794444253,
+      "learning_rate": 2.017219029214398e-07,
+      "loss": 1.6092,
+      "step": 2565
+    },
+    {
+      "epoch": 0.5325861353258613,
+      "grad_norm": 0.7844837275607282,
+      "learning_rate": 2.0160514391627945e-07,
+      "loss": 1.4471,
+      "step": 2566
+    },
+    {
+      "epoch": 0.5327936903279369,
+      "grad_norm": 1.2505709916627947,
+      "learning_rate": 2.014883927321235e-07,
+      "loss": 1.4994,
+      "step": 2567
+    },
+    {
+      "epoch": 0.5330012453300125,
+      "grad_norm": 0.8899884888286768,
+      "learning_rate": 2.013716494186115e-07,
+      "loss": 1.5106,
+      "step": 2568
+    },
+    {
+      "epoch": 0.533208800332088,
+      "grad_norm": 0.7967945892716943,
+      "learning_rate": 2.0125491402537972e-07,
+      "loss": 1.5715,
+      "step": 2569
+    },
+    {
+      "epoch": 0.5334163553341635,
+      "grad_norm": 0.6324024271307198,
+      "learning_rate": 2.0113818660206072e-07,
+      "loss": 1.437,
+      "step": 2570
+    },
+    {
+      "epoch": 0.5336239103362391,
+      "grad_norm": 4.3694699066472005,
+      "learning_rate": 2.0102146719828404e-07,
+      "loss": 1.4952,
+      "step": 2571
+    },
+    {
+      "epoch": 0.5338314653383146,
+      "grad_norm": 0.8139766291392232,
+      "learning_rate": 2.009047558636757e-07,
+      "loss": 1.5178,
+      "step": 2572
+    },
+    {
+      "epoch": 0.5340390203403902,
+      "grad_norm": 0.7086424050980737,
+      "learning_rate": 2.0078805264785822e-07,
+      "loss": 1.4464,
+      "step": 2573
+    },
+    {
+      "epoch": 0.5342465753424658,
+      "grad_norm": 1.3878930183771134,
+      "learning_rate": 2.0067135760045065e-07,
+      "loss": 1.505,
+      "step": 2574
+    },
+    {
+      "epoch": 0.5344541303445413,
+      "grad_norm": 1.294481870827835,
+      "learning_rate": 2.0055467077106876e-07,
+      "loss": 1.6181,
+      "step": 2575
+    },
+    {
+      "epoch": 0.5346616853466168,
+      "grad_norm": 1.7689129760619762,
+      "learning_rate": 2.0043799220932453e-07,
+      "loss": 1.564,
+      "step": 2576
+    },
+    {
+      "epoch": 0.5348692403486924,
+      "grad_norm": 1.0651299072503784,
+      "learning_rate": 2.0032132196482668e-07,
+      "loss": 1.5243,
+      "step": 2577
+    },
+    {
+      "epoch": 0.5350767953507679,
+      "grad_norm": 1.3711884382001511,
+      "learning_rate": 2.002046600871804e-07,
+      "loss": 1.4992,
+      "step": 2578
+    },
+    {
+      "epoch": 0.5352843503528435,
+      "grad_norm": 0.7601701275761795,
+      "learning_rate": 2.00088006625987e-07,
+      "loss": 1.4679,
+      "step": 2579
+    },
+    {
+      "epoch": 0.5354919053549191,
+      "grad_norm": 0.7558283542940453,
+      "learning_rate": 1.999713616308446e-07,
+      "loss": 1.4558,
+      "step": 2580
+    },
+    {
+      "epoch": 0.5356994603569946,
+      "grad_norm": 0.7442850091165438,
+      "learning_rate": 1.9985472515134752e-07,
+      "loss": 1.5164,
+      "step": 2581
+    },
+    {
+      "epoch": 0.5359070153590701,
+      "grad_norm": 0.7766475963883929,
+      "learning_rate": 1.9973809723708642e-07,
+      "loss": 1.4894,
+      "step": 2582
+    },
+    {
+      "epoch": 0.5361145703611457,
+      "grad_norm": 0.7115226359159237,
+      "learning_rate": 1.9962147793764847e-07,
+      "loss": 1.5156,
+      "step": 2583
+    },
+    {
+      "epoch": 0.5363221253632212,
+      "grad_norm": 0.7948940522129384,
+      "learning_rate": 1.9950486730261714e-07,
+      "loss": 1.5183,
+      "step": 2584
+    },
+    {
+      "epoch": 0.5365296803652968,
+      "grad_norm": 0.8025008089632144,
+      "learning_rate": 1.9938826538157208e-07,
+      "loss": 1.4771,
+      "step": 2585
+    },
+    {
+      "epoch": 0.5367372353673724,
+      "grad_norm": 0.8724798253088313,
+      "learning_rate": 1.992716722240893e-07,
+      "loss": 1.5546,
+      "step": 2586
+    },
+    {
+      "epoch": 0.5369447903694479,
+      "grad_norm": 0.6927123344277129,
+      "learning_rate": 1.9915508787974127e-07,
+      "loss": 1.5162,
+      "step": 2587
+    },
+    {
+      "epoch": 0.5371523453715235,
+      "grad_norm": 0.7849352002811072,
+      "learning_rate": 1.9903851239809645e-07,
+      "loss": 1.577,
+      "step": 2588
+    },
+    {
+      "epoch": 0.537359900373599,
+      "grad_norm": 1.001484972223679,
+      "learning_rate": 1.9892194582871964e-07,
+      "loss": 1.4782,
+      "step": 2589
+    },
+    {
+      "epoch": 0.5375674553756745,
+      "grad_norm": 0.814298982399508,
+      "learning_rate": 1.9880538822117194e-07,
+      "loss": 1.4753,
+      "step": 2590
+    },
+    {
+      "epoch": 0.5377750103777501,
+      "grad_norm": 0.7479124766756335,
+      "learning_rate": 1.9868883962501043e-07,
+      "loss": 1.5312,
+      "step": 2591
+    },
+    {
+      "epoch": 0.5379825653798257,
+      "grad_norm": 1.5220964061406288,
+      "learning_rate": 1.985723000897885e-07,
+      "loss": 1.5761,
+      "step": 2592
+    },
+    {
+      "epoch": 0.5381901203819012,
+      "grad_norm": 0.6334335219464433,
+      "learning_rate": 1.9845576966505578e-07,
+      "loss": 1.5532,
+      "step": 2593
+    },
+    {
+      "epoch": 0.5383976753839768,
+      "grad_norm": 0.9226501087900151,
+      "learning_rate": 1.9833924840035773e-07,
+      "loss": 1.5375,
+      "step": 2594
+    },
+    {
+      "epoch": 0.5386052303860523,
+      "grad_norm": 0.7677883613361626,
+      "learning_rate": 1.9822273634523627e-07,
+      "loss": 1.458,
+      "step": 2595
+    },
+    {
+      "epoch": 0.5388127853881278,
+      "grad_norm": 0.7881348689251744,
+      "learning_rate": 1.9810623354922922e-07,
+      "loss": 1.602,
+      "step": 2596
+    },
+    {
+      "epoch": 0.5390203403902034,
+      "grad_norm": 0.9850253695586121,
+      "learning_rate": 1.9798974006187033e-07,
+      "loss": 1.5089,
+      "step": 2597
+    },
+    {
+      "epoch": 0.539227895392279,
+      "grad_norm": 1.001149933615508,
+      "learning_rate": 1.9787325593268962e-07,
+      "loss": 1.5188,
+      "step": 2598
+    },
+    {
+      "epoch": 0.5394354503943545,
+      "grad_norm": 0.8002193561602818,
+      "learning_rate": 1.9775678121121308e-07,
+      "loss": 1.5664,
+      "step": 2599
+    },
+    {
+      "epoch": 0.5396430053964301,
+      "grad_norm": 0.6582182484610896,
+      "learning_rate": 1.9764031594696266e-07,
+      "loss": 1.4626,
+      "step": 2600
+    },
+    {
+      "epoch": 0.5398505603985057,
+      "grad_norm": 1.1128223269784405,
+      "learning_rate": 1.9752386018945627e-07,
+      "loss": 1.5202,
+      "step": 2601
+    },
+    {
+      "epoch": 0.5400581154005811,
+      "grad_norm": 1.0129186404957065,
+      "learning_rate": 1.9740741398820783e-07,
+      "loss": 1.5198,
+      "step": 2602
+    },
+    {
+      "epoch": 0.5402656704026567,
+      "grad_norm": 1.0330207638800408,
+      "learning_rate": 1.9729097739272716e-07,
+      "loss": 1.4579,
+      "step": 2603
+    },
+    {
+      "epoch": 0.5404732254047323,
+      "grad_norm": 0.6994365836644076,
+      "learning_rate": 1.9717455045251997e-07,
+      "loss": 1.5533,
+      "step": 2604
+    },
+    {
+      "epoch": 0.5406807804068078,
+      "grad_norm": 0.847396599098467,
+      "learning_rate": 1.9705813321708803e-07,
+      "loss": 1.5015,
+      "step": 2605
+    },
+    {
+      "epoch": 0.5408883354088834,
+      "grad_norm": 1.0987093683386875,
+      "learning_rate": 1.9694172573592872e-07,
+      "loss": 1.5718,
+      "step": 2606
+    },
+    {
+      "epoch": 0.541095890410959,
+      "grad_norm": 1.4286627190603036,
+      "learning_rate": 1.9682532805853542e-07,
+      "loss": 1.4825,
+      "step": 2607
+    },
+    {
+      "epoch": 0.5413034454130344,
+      "grad_norm": 0.6840506621435607,
+      "learning_rate": 1.967089402343975e-07,
+      "loss": 1.4244,
+      "step": 2608
+    },
+    {
+      "epoch": 0.54151100041511,
+      "grad_norm": 3.760057848458652,
+      "learning_rate": 1.9659256231299976e-07,
+      "loss": 1.5407,
+      "step": 2609
+    },
+    {
+      "epoch": 0.5417185554171855,
+      "grad_norm": 0.6711912130269639,
+      "learning_rate": 1.9647619434382317e-07,
+      "loss": 1.5012,
+      "step": 2610
+    },
+    {
+      "epoch": 0.5419261104192611,
+      "grad_norm": 0.7361567980378269,
+      "learning_rate": 1.9635983637634413e-07,
+      "loss": 1.5993,
+      "step": 2611
+    },
+    {
+      "epoch": 0.5421336654213367,
+      "grad_norm": 0.7413045068438774,
+      "learning_rate": 1.9624348846003507e-07,
+      "loss": 1.5273,
+      "step": 2612
+    },
+    {
+      "epoch": 0.5423412204234122,
+      "grad_norm": 0.6876218435027974,
+      "learning_rate": 1.9612715064436402e-07,
+      "loss": 1.4582,
+      "step": 2613
+    },
+    {
+      "epoch": 0.5425487754254877,
+      "grad_norm": 0.6831856447717607,
+      "learning_rate": 1.9601082297879473e-07,
+      "loss": 1.5518,
+      "step": 2614
+    },
+    {
+      "epoch": 0.5427563304275633,
+      "grad_norm": 0.7701202309594065,
+      "learning_rate": 1.9589450551278665e-07,
+      "loss": 1.5647,
+      "step": 2615
+    },
+    {
+      "epoch": 0.5429638854296388,
+      "grad_norm": 0.7506364075714071,
+      "learning_rate": 1.957781982957949e-07,
+      "loss": 1.477,
+      "step": 2616
+    },
+    {
+      "epoch": 0.5431714404317144,
+      "grad_norm": 0.8723358264886831,
+      "learning_rate": 1.9566190137727015e-07,
+      "loss": 1.5253,
+      "step": 2617
+    },
+    {
+      "epoch": 0.54337899543379,
+      "grad_norm": 0.8627677965500178,
+      "learning_rate": 1.9554561480665872e-07,
+      "loss": 1.5289,
+      "step": 2618
+    },
+    {
+      "epoch": 0.5435865504358655,
+      "grad_norm": 0.7130662303568074,
+      "learning_rate": 1.9542933863340277e-07,
+      "loss": 1.5445,
+      "step": 2619
+    },
+    {
+      "epoch": 0.543794105437941,
+      "grad_norm": 1.7699211116355538,
+      "learning_rate": 1.9531307290693966e-07,
+      "loss": 1.4854,
+      "step": 2620
+    },
+    {
+      "epoch": 0.5440016604400166,
+      "grad_norm": 3.7505287485680316,
+      "learning_rate": 1.9519681767670248e-07,
+      "loss": 1.4785,
+      "step": 2621
+    },
+    {
+      "epoch": 0.5442092154420921,
+      "grad_norm": 0.7132708475910746,
+      "learning_rate": 1.9508057299212006e-07,
+      "loss": 1.5088,
+      "step": 2622
+    },
+    {
+      "epoch": 0.5444167704441677,
+      "grad_norm": 0.8104327028031877,
+      "learning_rate": 1.9496433890261637e-07,
+      "loss": 1.6185,
+      "step": 2623
+    },
+    {
+      "epoch": 0.5446243254462433,
+      "grad_norm": 0.7107408496719876,
+      "learning_rate": 1.948481154576111e-07,
+      "loss": 1.49,
+      "step": 2624
+    },
+    {
+      "epoch": 0.5448318804483188,
+      "grad_norm": 0.9847344267353024,
+      "learning_rate": 1.9473190270651946e-07,
+      "loss": 1.5385,
+      "step": 2625
+    },
+    {
+      "epoch": 0.5450394354503943,
+      "grad_norm": 0.8365847136395933,
+      "learning_rate": 1.9461570069875189e-07,
+      "loss": 1.5188,
+      "step": 2626
+    },
+    {
+      "epoch": 0.5452469904524699,
+      "grad_norm": 0.8676422222843052,
+      "learning_rate": 1.9449950948371452e-07,
+      "loss": 1.4884,
+      "step": 2627
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 0.9724202963012527,
+      "learning_rate": 1.943833291108087e-07,
+      "loss": 1.4944,
+      "step": 2628
+    },
+    {
+      "epoch": 0.545662100456621,
+      "grad_norm": 0.938996661882557,
+      "learning_rate": 1.9426715962943124e-07,
+      "loss": 1.5254,
+      "step": 2629
+    },
+    {
+      "epoch": 0.5458696554586966,
+      "grad_norm": 0.7555234906413024,
+      "learning_rate": 1.9415100108897433e-07,
+      "loss": 1.535,
+      "step": 2630
+    },
+    {
+      "epoch": 0.5460772104607721,
+      "grad_norm": 0.8342994351519754,
+      "learning_rate": 1.9403485353882556e-07,
+      "loss": 1.5575,
+      "step": 2631
+    },
+    {
+      "epoch": 0.5462847654628477,
+      "grad_norm": 1.4670984804883267,
+      "learning_rate": 1.9391871702836767e-07,
+      "loss": 1.5168,
+      "step": 2632
+    },
+    {
+      "epoch": 0.5464923204649232,
+      "grad_norm": 0.7000872578901604,
+      "learning_rate": 1.938025916069789e-07,
+      "loss": 1.6288,
+      "step": 2633
+    },
+    {
+      "epoch": 0.5466998754669987,
+      "grad_norm": 0.7460420704703576,
+      "learning_rate": 1.936864773240327e-07,
+      "loss": 1.4856,
+      "step": 2634
+    },
+    {
+      "epoch": 0.5469074304690743,
+      "grad_norm": 0.9654472569021206,
+      "learning_rate": 1.9357037422889775e-07,
+      "loss": 1.5593,
+      "step": 2635
+    },
+    {
+      "epoch": 0.5471149854711499,
+      "grad_norm": 0.6153019995606699,
+      "learning_rate": 1.9345428237093796e-07,
+      "loss": 1.489,
+      "step": 2636
+    },
+    {
+      "epoch": 0.5473225404732254,
+      "grad_norm": 0.6563171925169946,
+      "learning_rate": 1.9333820179951265e-07,
+      "loss": 1.5439,
+      "step": 2637
+    },
+    {
+      "epoch": 0.547530095475301,
+      "grad_norm": 0.7718146302500087,
+      "learning_rate": 1.9322213256397607e-07,
+      "loss": 1.5064,
+      "step": 2638
+    },
+    {
+      "epoch": 0.5477376504773765,
+      "grad_norm": 0.6844069954284052,
+      "learning_rate": 1.9310607471367776e-07,
+      "loss": 1.508,
+      "step": 2639
+    },
+    {
+      "epoch": 0.547945205479452,
+      "grad_norm": 0.835347122756183,
+      "learning_rate": 1.9299002829796253e-07,
+      "loss": 1.5552,
+      "step": 2640
+    },
+    {
+      "epoch": 0.5481527604815276,
+      "grad_norm": 1.1051777780740364,
+      "learning_rate": 1.9287399336617013e-07,
+      "loss": 1.5357,
+      "step": 2641
+    },
+    {
+      "epoch": 0.5483603154836032,
+      "grad_norm": 0.9967497623851438,
+      "learning_rate": 1.927579699676357e-07,
+      "loss": 1.5422,
+      "step": 2642
+    },
+    {
+      "epoch": 0.5485678704856787,
+      "grad_norm": 0.6450612345188861,
+      "learning_rate": 1.9264195815168917e-07,
+      "loss": 1.5541,
+      "step": 2643
+    },
+    {
+      "epoch": 0.5487754254877543,
+      "grad_norm": 0.9532642709277006,
+      "learning_rate": 1.925259579676557e-07,
+      "loss": 1.4528,
+      "step": 2644
+    },
+    {
+      "epoch": 0.5489829804898299,
+      "grad_norm": 0.7023915731412769,
+      "learning_rate": 1.924099694648555e-07,
+      "loss": 1.555,
+      "step": 2645
+    },
+    {
+      "epoch": 0.5491905354919053,
+      "grad_norm": 0.7285824094094605,
+      "learning_rate": 1.922939926926039e-07,
+      "loss": 1.5087,
+      "step": 2646
+    },
+    {
+      "epoch": 0.5493980904939809,
+      "grad_norm": 4.718979803607192,
+      "learning_rate": 1.921780277002109e-07,
+      "loss": 1.4276,
+      "step": 2647
+    },
+    {
+      "epoch": 0.5496056454960565,
+      "grad_norm": 1.626087856849418,
+      "learning_rate": 1.9206207453698196e-07,
+      "loss": 1.4988,
+      "step": 2648
+    },
+    {
+      "epoch": 0.549813200498132,
+      "grad_norm": 0.7161478862989256,
+      "learning_rate": 1.919461332522173e-07,
+      "loss": 1.5385,
+      "step": 2649
+    },
+    {
+      "epoch": 0.5500207555002076,
+      "grad_norm": 1.158163434777051,
+      "learning_rate": 1.918302038952119e-07,
+      "loss": 1.5257,
+      "step": 2650
+    },
+    {
+      "epoch": 0.5502283105022832,
+      "grad_norm": 0.8710489994965636,
+      "learning_rate": 1.9171428651525594e-07,
+      "loss": 1.4933,
+      "step": 2651
+    },
+    {
+      "epoch": 0.5504358655043586,
+      "grad_norm": 0.7084888585141259,
+      "learning_rate": 1.9159838116163445e-07,
+      "loss": 1.5889,
+      "step": 2652
+    },
+    {
+      "epoch": 0.5506434205064342,
+      "grad_norm": 0.7457934447926522,
+      "learning_rate": 1.9148248788362725e-07,
+      "loss": 1.6317,
+      "step": 2653
+    },
+    {
+      "epoch": 0.5508509755085098,
+      "grad_norm": 0.7462125314074037,
+      "learning_rate": 1.9136660673050908e-07,
+      "loss": 1.5664,
+      "step": 2654
+    },
+    {
+      "epoch": 0.5510585305105853,
+      "grad_norm": 0.8513899894602226,
+      "learning_rate": 1.912507377515496e-07,
+      "loss": 1.5258,
+      "step": 2655
+    },
+    {
+      "epoch": 0.5512660855126609,
+      "grad_norm": 1.213717355933964,
+      "learning_rate": 1.9113488099601316e-07,
+      "loss": 1.5186,
+      "step": 2656
+    },
+    {
+      "epoch": 0.5514736405147364,
+      "grad_norm": 0.8398296246797831,
+      "learning_rate": 1.9101903651315903e-07,
+      "loss": 1.5114,
+      "step": 2657
+    },
+    {
+      "epoch": 0.5516811955168119,
+      "grad_norm": 1.2823849352393197,
+      "learning_rate": 1.909032043522411e-07,
+      "loss": 1.5577,
+      "step": 2658
+    },
+    {
+      "epoch": 0.5518887505188875,
+      "grad_norm": 0.9858566423500901,
+      "learning_rate": 1.9078738456250822e-07,
+      "loss": 1.516,
+      "step": 2659
+    },
+    {
+      "epoch": 0.552096305520963,
+      "grad_norm": 1.1533939088957754,
+      "learning_rate": 1.9067157719320398e-07,
+      "loss": 1.5529,
+      "step": 2660
+    },
+    {
+      "epoch": 0.5523038605230386,
+      "grad_norm": 0.7309259460140325,
+      "learning_rate": 1.9055578229356635e-07,
+      "loss": 1.4912,
+      "step": 2661
+    },
+    {
+      "epoch": 0.5525114155251142,
+      "grad_norm": 0.8819161146840586,
+      "learning_rate": 1.9043999991282843e-07,
+      "loss": 1.5093,
+      "step": 2662
+    },
+    {
+      "epoch": 0.5527189705271897,
+      "grad_norm": 1.547214659924875,
+      "learning_rate": 1.9032423010021783e-07,
+      "loss": 1.5139,
+      "step": 2663
+    },
+    {
+      "epoch": 0.5529265255292652,
+      "grad_norm": 0.9963851536672461,
+      "learning_rate": 1.902084729049567e-07,
+      "loss": 1.5319,
+      "step": 2664
+    },
+    {
+      "epoch": 0.5531340805313408,
+      "grad_norm": 0.8097640047372465,
+      "learning_rate": 1.9009272837626193e-07,
+      "loss": 1.5667,
+      "step": 2665
+    },
+    {
+      "epoch": 0.5533416355334163,
+      "grad_norm": 0.8382361931016105,
+      "learning_rate": 1.8997699656334514e-07,
+      "loss": 1.5423,
+      "step": 2666
+    },
+    {
+      "epoch": 0.5535491905354919,
+      "grad_norm": 1.0250922342427287,
+      "learning_rate": 1.898612775154123e-07,
+      "loss": 1.5439,
+      "step": 2667
+    },
+    {
+      "epoch": 0.5537567455375675,
+      "grad_norm": 1.0918296168075359,
+      "learning_rate": 1.8974557128166412e-07,
+      "loss": 1.5251,
+      "step": 2668
+    },
+    {
+      "epoch": 0.553964300539643,
+      "grad_norm": 0.9430203632150722,
+      "learning_rate": 1.8962987791129587e-07,
+      "loss": 1.445,
+      "step": 2669
+    },
+    {
+      "epoch": 0.5541718555417185,
+      "grad_norm": 0.6625904624933437,
+      "learning_rate": 1.895141974534972e-07,
+      "loss": 1.4622,
+      "step": 2670
+    },
+    {
+      "epoch": 0.5543794105437941,
+      "grad_norm": 2.797678347456512,
+      "learning_rate": 1.893985299574524e-07,
+      "loss": 1.4945,
+      "step": 2671
+    },
+    {
+      "epoch": 0.5545869655458696,
+      "grad_norm": 1.7556413512574123,
+      "learning_rate": 1.8928287547234034e-07,
+      "loss": 1.5745,
+      "step": 2672
+    },
+    {
+      "epoch": 0.5547945205479452,
+      "grad_norm": 0.9866484212728677,
+      "learning_rate": 1.8916723404733404e-07,
+      "loss": 1.4825,
+      "step": 2673
+    },
+    {
+      "epoch": 0.5550020755500208,
+      "grad_norm": 1.0127980911074592,
+      "learning_rate": 1.8905160573160127e-07,
+      "loss": 1.5796,
+      "step": 2674
+    },
+    {
+      "epoch": 0.5552096305520963,
+      "grad_norm": 0.7065156647173052,
+      "learning_rate": 1.889359905743042e-07,
+      "loss": 1.5202,
+      "step": 2675
+    },
+    {
+      "epoch": 0.5554171855541719,
+      "grad_norm": 1.2487302686251152,
+      "learning_rate": 1.8882038862459915e-07,
+      "loss": 1.547,
+      "step": 2676
+    },
+    {
+      "epoch": 0.5556247405562474,
+      "grad_norm": 1.478208011918142,
+      "learning_rate": 1.8870479993163704e-07,
+      "loss": 1.484,
+      "step": 2677
+    },
+    {
+      "epoch": 0.5558322955583229,
+      "grad_norm": 0.815336877384556,
+      "learning_rate": 1.8858922454456327e-07,
+      "loss": 1.5019,
+      "step": 2678
+    },
+    {
+      "epoch": 0.5560398505603985,
+      "grad_norm": 1.1044039929512466,
+      "learning_rate": 1.884736625125172e-07,
+      "loss": 1.5718,
+      "step": 2679
+    },
+    {
+      "epoch": 0.5562474055624741,
+      "grad_norm": 0.665007064219597,
+      "learning_rate": 1.88358113884633e-07,
+      "loss": 1.488,
+      "step": 2680
+    },
+    {
+      "epoch": 0.5564549605645496,
+      "grad_norm": 0.6790814995066067,
+      "learning_rate": 1.8824257871003866e-07,
+      "loss": 1.5128,
+      "step": 2681
+    },
+    {
+      "epoch": 0.5566625155666252,
+      "grad_norm": 1.0506260031468775,
+      "learning_rate": 1.8812705703785673e-07,
+      "loss": 1.419,
+      "step": 2682
+    },
+    {
+      "epoch": 0.5568700705687007,
+      "grad_norm": 0.9181952804045278,
+      "learning_rate": 1.8801154891720391e-07,
+      "loss": 1.5431,
+      "step": 2683
+    },
+    {
+      "epoch": 0.5570776255707762,
+      "grad_norm": 0.8247199591341412,
+      "learning_rate": 1.8789605439719134e-07,
+      "loss": 1.5133,
+      "step": 2684
+    },
+    {
+      "epoch": 0.5572851805728518,
+      "grad_norm": 1.1506388715183375,
+      "learning_rate": 1.877805735269241e-07,
+      "loss": 1.515,
+      "step": 2685
+    },
+    {
+      "epoch": 0.5574927355749274,
+      "grad_norm": 0.9291975605696297,
+      "learning_rate": 1.8766510635550157e-07,
+      "loss": 1.4602,
+      "step": 2686
+    },
+    {
+      "epoch": 0.5577002905770029,
+      "grad_norm": 0.8013294926749087,
+      "learning_rate": 1.8754965293201747e-07,
+      "loss": 1.5143,
+      "step": 2687
+    },
+    {
+      "epoch": 0.5579078455790785,
+      "grad_norm": 0.9362542771381518,
+      "learning_rate": 1.874342133055594e-07,
+      "loss": 1.4484,
+      "step": 2688
+    },
+    {
+      "epoch": 0.558115400581154,
+      "grad_norm": 1.190948033288314,
+      "learning_rate": 1.8731878752520922e-07,
+      "loss": 1.5479,
+      "step": 2689
+    },
+    {
+      "epoch": 0.5583229555832295,
+      "grad_norm": 0.8587599737060164,
+      "learning_rate": 1.8720337564004303e-07,
+      "loss": 1.4964,
+      "step": 2690
+    },
+    {
+      "epoch": 0.5585305105853051,
+      "grad_norm": 0.8190566264464891,
+      "learning_rate": 1.870879776991307e-07,
+      "loss": 1.5108,
+      "step": 2691
+    },
+    {
+      "epoch": 0.5587380655873807,
+      "grad_norm": 0.7502089057972806,
+      "learning_rate": 1.8697259375153657e-07,
+      "loss": 1.5149,
+      "step": 2692
+    },
+    {
+      "epoch": 0.5589456205894562,
+      "grad_norm": 0.6261930646183931,
+      "learning_rate": 1.8685722384631872e-07,
+      "loss": 1.5177,
+      "step": 2693
+    },
+    {
+      "epoch": 0.5591531755915318,
+      "grad_norm": 0.6954909957302217,
+      "learning_rate": 1.8674186803252942e-07,
+      "loss": 1.5427,
+      "step": 2694
+    },
+    {
+      "epoch": 0.5593607305936074,
+      "grad_norm": 0.8081707958785462,
+      "learning_rate": 1.8662652635921478e-07,
+      "loss": 1.5188,
+      "step": 2695
+    },
+    {
+      "epoch": 0.5595682855956828,
+      "grad_norm": 1.0336840727140324,
+      "learning_rate": 1.865111988754153e-07,
+      "loss": 1.5569,
+      "step": 2696
+    },
+    {
+      "epoch": 0.5597758405977584,
+      "grad_norm": 0.9004309527194245,
+      "learning_rate": 1.8639588563016483e-07,
+      "loss": 1.5987,
+      "step": 2697
+    },
+    {
+      "epoch": 0.559983395599834,
+      "grad_norm": 0.8600382351053841,
+      "learning_rate": 1.862805866724917e-07,
+      "loss": 1.4599,
+      "step": 2698
+    },
+    {
+      "epoch": 0.5601909506019095,
+      "grad_norm": 0.8325831193058758,
+      "learning_rate": 1.8616530205141795e-07,
+      "loss": 1.5351,
+      "step": 2699
+    },
+    {
+      "epoch": 0.5603985056039851,
+      "grad_norm": 0.7393574097641633,
+      "learning_rate": 1.8605003181595947e-07,
+      "loss": 1.5565,
+      "step": 2700
+    },
+    {
+      "epoch": 0.5606060606060606,
+      "grad_norm": 0.6309010403150376,
+      "learning_rate": 1.8593477601512625e-07,
+      "loss": 1.4587,
+      "step": 2701
+    },
+    {
+      "epoch": 0.5608136156081361,
+      "grad_norm": 1.7628409646136418,
+      "learning_rate": 1.858195346979217e-07,
+      "loss": 1.4865,
+      "step": 2702
+    },
+    {
+      "epoch": 0.5610211706102117,
+      "grad_norm": 0.6912674024261483,
+      "learning_rate": 1.8570430791334367e-07,
+      "loss": 1.5231,
+      "step": 2703
+    },
+    {
+      "epoch": 0.5612287256122872,
+      "grad_norm": 0.6966170370309946,
+      "learning_rate": 1.8558909571038338e-07,
+      "loss": 1.5428,
+      "step": 2704
+    },
+    {
+      "epoch": 0.5614362806143628,
+      "grad_norm": 2.328648662710698,
+      "learning_rate": 1.8547389813802607e-07,
+      "loss": 1.5229,
+      "step": 2705
+    },
+    {
+      "epoch": 0.5616438356164384,
+      "grad_norm": 0.6779571511322753,
+      "learning_rate": 1.8535871524525062e-07,
+      "loss": 1.5633,
+      "step": 2706
+    },
+    {
+      "epoch": 0.5618513906185139,
+      "grad_norm": 0.731410520775825,
+      "learning_rate": 1.852435470810298e-07,
+      "loss": 1.4934,
+      "step": 2707
+    },
+    {
+      "epoch": 0.5620589456205894,
+      "grad_norm": 0.7131051144208489,
+      "learning_rate": 1.8512839369432996e-07,
+      "loss": 1.4917,
+      "step": 2708
+    },
+    {
+      "epoch": 0.562266500622665,
+      "grad_norm": 0.8023585879855598,
+      "learning_rate": 1.8501325513411138e-07,
+      "loss": 1.474,
+      "step": 2709
+    },
+    {
+      "epoch": 0.5624740556247405,
+      "grad_norm": 0.8758225868910381,
+      "learning_rate": 1.8489813144932797e-07,
+      "loss": 1.5054,
+      "step": 2710
+    },
+    {
+      "epoch": 0.5626816106268161,
+      "grad_norm": 0.6229606125937508,
+      "learning_rate": 1.8478302268892704e-07,
+      "loss": 1.4871,
+      "step": 2711
+    },
+    {
+      "epoch": 0.5628891656288917,
+      "grad_norm": 0.6872097271487397,
+      "learning_rate": 1.8466792890184993e-07,
+      "loss": 1.5273,
+      "step": 2712
+    },
+    {
+      "epoch": 0.5630967206309672,
+      "grad_norm": 0.7667222581343445,
+      "learning_rate": 1.8455285013703152e-07,
+      "loss": 1.4982,
+      "step": 2713
+    },
+    {
+      "epoch": 0.5633042756330428,
+      "grad_norm": 0.7748817773205877,
+      "learning_rate": 1.844377864434001e-07,
+      "loss": 1.5322,
+      "step": 2714
+    },
+    {
+      "epoch": 0.5635118306351183,
+      "grad_norm": 0.7815913221597046,
+      "learning_rate": 1.8432273786987774e-07,
+      "loss": 1.5468,
+      "step": 2715
+    },
+    {
+      "epoch": 0.5637193856371938,
+      "grad_norm": 0.722430599278042,
+      "learning_rate": 1.842077044653801e-07,
+      "loss": 1.5262,
+      "step": 2716
+    },
+    {
+      "epoch": 0.5639269406392694,
+      "grad_norm": 0.6839973300725115,
+      "learning_rate": 1.8409268627881623e-07,
+      "loss": 1.4895,
+      "step": 2717
+    },
+    {
+      "epoch": 0.564134495641345,
+      "grad_norm": 0.9516528917229495,
+      "learning_rate": 1.8397768335908887e-07,
+      "loss": 1.4988,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5643420506434205,
+      "grad_norm": 0.6788231883898889,
+      "learning_rate": 1.838626957550943e-07,
+      "loss": 1.4586,
+      "step": 2719
+    },
+    {
+      "epoch": 0.564549605645496,
+      "grad_norm": 0.7480678374192195,
+      "learning_rate": 1.83747723515722e-07,
+      "loss": 1.498,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5647571606475716,
+      "grad_norm": 1.1427546986061294,
+      "learning_rate": 1.8363276668985525e-07,
+      "loss": 1.5274,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5649647156496471,
+      "grad_norm": 1.096417469322258,
+      "learning_rate": 1.8351782532637068e-07,
+      "loss": 1.5141,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5651722706517227,
+      "grad_norm": 1.012014928385167,
+      "learning_rate": 1.8340289947413815e-07,
+      "loss": 1.4919,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5653798256537983,
+      "grad_norm": 0.7639563419830245,
+      "learning_rate": 1.832879891820212e-07,
+      "loss": 1.4295,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5655873806558738,
+      "grad_norm": 1.1925686093072705,
+      "learning_rate": 1.8317309449887662e-07,
+      "loss": 1.4739,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5657949356579494,
+      "grad_norm": 1.037202910007226,
+      "learning_rate": 1.8305821547355448e-07,
+      "loss": 1.539,
+      "step": 2726
+    },
+    {
+      "epoch": 0.566002490660025,
+      "grad_norm": 0.8563013831550119,
+      "learning_rate": 1.8294335215489843e-07,
+      "loss": 1.5632,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5662100456621004,
+      "grad_norm": 0.8427013856278586,
+      "learning_rate": 1.828285045917453e-07,
+      "loss": 1.503,
+      "step": 2728
+    },
+    {
+      "epoch": 0.566417600664176,
+      "grad_norm": 0.7298989480893748,
+      "learning_rate": 1.827136728329251e-07,
+      "loss": 1.5318,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5666251556662516,
+      "grad_norm": 0.7845165546987427,
+      "learning_rate": 1.825988569272613e-07,
+      "loss": 1.5243,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5668327106683271,
+      "grad_norm": 0.6408042649221842,
+      "learning_rate": 1.8248405692357066e-07,
+      "loss": 1.5679,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5670402656704027,
+      "grad_norm": 0.6909984077160943,
+      "learning_rate": 1.8236927287066296e-07,
+      "loss": 1.5052,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5672478206724783,
+      "grad_norm": 0.6951777283489345,
+      "learning_rate": 1.8225450481734144e-07,
+      "loss": 1.5199,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5674553756745537,
+      "grad_norm": 0.7694456173903054,
+      "learning_rate": 1.8213975281240236e-07,
+      "loss": 1.4955,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5676629306766293,
+      "grad_norm": 0.9643460694550313,
+      "learning_rate": 1.8202501690463526e-07,
+      "loss": 1.5716,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5678704856787049,
+      "grad_norm": 1.4823493750423102,
+      "learning_rate": 1.8191029714282276e-07,
+      "loss": 1.4821,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5680780406807804,
+      "grad_norm": 0.9450821946974228,
+      "learning_rate": 1.8179559357574074e-07,
+      "loss": 1.5334,
+      "step": 2737
+    },
+    {
+      "epoch": 0.568285595682856,
+      "grad_norm": 0.7361669666385137,
+      "learning_rate": 1.8168090625215803e-07,
+      "loss": 1.4771,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5684931506849316,
+      "grad_norm": 0.7497096119643377,
+      "learning_rate": 1.815662352208367e-07,
+      "loss": 1.4959,
+      "step": 2739
+    },
+    {
+      "epoch": 0.568700705687007,
+      "grad_norm": 0.7278737759107909,
+      "learning_rate": 1.814515805305318e-07,
+      "loss": 1.5506,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5689082606890826,
+      "grad_norm": 0.7946986738801807,
+      "learning_rate": 1.8133694222999142e-07,
+      "loss": 1.6229,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5691158156911582,
+      "grad_norm": 0.8215902947665022,
+      "learning_rate": 1.8122232036795678e-07,
+      "loss": 1.546,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5693233706932337,
+      "grad_norm": 0.7579123615211772,
+      "learning_rate": 1.8110771499316204e-07,
+      "loss": 1.5423,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5695309256953093,
+      "grad_norm": 0.6938686021950268,
+      "learning_rate": 1.8099312615433432e-07,
+      "loss": 1.4626,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5697384806973848,
+      "grad_norm": 0.8070564091763462,
+      "learning_rate": 1.8087855390019385e-07,
+      "loss": 1.5212,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5699460356994603,
+      "grad_norm": 1.1471650079910272,
+      "learning_rate": 1.8076399827945354e-07,
+      "loss": 1.4707,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5701535907015359,
+      "grad_norm": 1.6139513162128873,
+      "learning_rate": 1.8064945934081958e-07,
+      "loss": 1.5841,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5703611457036114,
+      "grad_norm": 0.7614890714888461,
+      "learning_rate": 1.8053493713299082e-07,
+      "loss": 1.5646,
+      "step": 2748
+    },
+    {
+      "epoch": 0.570568700705687,
+      "grad_norm": 1.023191354272193,
+      "learning_rate": 1.8042043170465902e-07,
+      "loss": 1.5557,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5707762557077626,
+      "grad_norm": 1.2261344768306712,
+      "learning_rate": 1.8030594310450886e-07,
+      "loss": 1.5399,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5709838107098381,
+      "grad_norm": 0.723985990597229,
+      "learning_rate": 1.8019147138121794e-07,
+      "loss": 1.5442,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5711913657119136,
+      "grad_norm": 0.949264869721745,
+      "learning_rate": 1.800770165834565e-07,
+      "loss": 1.5218,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5713989207139892,
+      "grad_norm": 0.8808337471237722,
+      "learning_rate": 1.799625787598877e-07,
+      "loss": 1.5304,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5716064757160647,
+      "grad_norm": 0.7164568810738093,
+      "learning_rate": 1.7984815795916753e-07,
+      "loss": 1.5109,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5718140307181403,
+      "grad_norm": 0.8489271133851851,
+      "learning_rate": 1.7973375422994456e-07,
+      "loss": 1.4667,
+      "step": 2755
+    },
+    {
+      "epoch": 0.5720215857202159,
+      "grad_norm": 1.0910126301979972,
+      "learning_rate": 1.796193676208603e-07,
+      "loss": 1.4165,
+      "step": 2756
+    },
+    {
+      "epoch": 0.5722291407222914,
+      "grad_norm": 3.135914016248933,
+      "learning_rate": 1.795049981805489e-07,
+      "loss": 1.489,
+      "step": 2757
+    },
+    {
+      "epoch": 0.572436695724367,
+      "grad_norm": 0.7124518973801518,
+      "learning_rate": 1.7939064595763714e-07,
+      "loss": 1.5295,
+      "step": 2758
+    },
+    {
+      "epoch": 0.5726442507264425,
+      "grad_norm": 0.9999219194852379,
+      "learning_rate": 1.7927631100074466e-07,
+      "loss": 1.4793,
+      "step": 2759
+    },
+    {
+      "epoch": 0.572851805728518,
+      "grad_norm": 0.7301464314158208,
+      "learning_rate": 1.791619933584836e-07,
+      "loss": 1.4797,
+      "step": 2760
+    },
+    {
+      "epoch": 0.5730593607305936,
+      "grad_norm": 0.7173427879646389,
+      "learning_rate": 1.790476930794587e-07,
+      "loss": 1.5339,
+      "step": 2761
+    },
+    {
+      "epoch": 0.5732669157326692,
+      "grad_norm": 0.6558477164196588,
+      "learning_rate": 1.7893341021226753e-07,
+      "loss": 1.5162,
+      "step": 2762
+    },
+    {
+      "epoch": 0.5734744707347447,
+      "grad_norm": 0.7544021447305143,
+      "learning_rate": 1.7881914480550014e-07,
+      "loss": 1.563,
+      "step": 2763
+    },
+    {
+      "epoch": 0.5736820257368203,
+      "grad_norm": 0.7627394445166236,
+      "learning_rate": 1.7870489690773904e-07,
+      "loss": 1.5377,
+      "step": 2764
+    },
+    {
+      "epoch": 0.5738895807388958,
+      "grad_norm": 2.1990430715287927,
+      "learning_rate": 1.785906665675594e-07,
+      "loss": 1.4977,
+      "step": 2765
+    },
+    {
+      "epoch": 0.5740971357409713,
+      "grad_norm": 0.8377752614424979,
+      "learning_rate": 1.7847645383352906e-07,
+      "loss": 1.505,
+      "step": 2766
+    },
+    {
+      "epoch": 0.5743046907430469,
+      "grad_norm": 0.9162713093019611,
+      "learning_rate": 1.783622587542081e-07,
+      "loss": 1.5722,
+      "step": 2767
+    },
+    {
+      "epoch": 0.5745122457451225,
+      "grad_norm": 0.6948015757648495,
+      "learning_rate": 1.7824808137814933e-07,
+      "loss": 1.5344,
+      "step": 2768
+    },
+    {
+      "epoch": 0.574719800747198,
+      "grad_norm": 1.0860076720655756,
+      "learning_rate": 1.7813392175389797e-07,
+      "loss": 1.5259,
+      "step": 2769
+    },
+    {
+      "epoch": 0.5749273557492736,
+      "grad_norm": 0.6762461867833881,
+      "learning_rate": 1.7801977992999148e-07,
+      "loss": 1.5648,
+      "step": 2770
+    },
+    {
+      "epoch": 0.5751349107513491,
+      "grad_norm": 0.6887915169174699,
+      "learning_rate": 1.7790565595496006e-07,
+      "loss": 1.5001,
+      "step": 2771
+    },
+    {
+      "epoch": 0.5753424657534246,
+      "grad_norm": 0.8224992999908507,
+      "learning_rate": 1.7779154987732627e-07,
+      "loss": 1.475,
+      "step": 2772
+    },
+    {
+      "epoch": 0.5755500207555002,
+      "grad_norm": 0.8216569245861504,
+      "learning_rate": 1.7767746174560482e-07,
+      "loss": 1.5263,
+      "step": 2773
+    },
+    {
+      "epoch": 0.5757575757575758,
+      "grad_norm": 0.7748509807248156,
+      "learning_rate": 1.7756339160830307e-07,
+      "loss": 1.5521,
+      "step": 2774
+    },
+    {
+      "epoch": 0.5759651307596513,
+      "grad_norm": 0.8065867638140796,
+      "learning_rate": 1.7744933951392062e-07,
+      "loss": 1.5457,
+      "step": 2775
+    },
+    {
+      "epoch": 0.5761726857617269,
+      "grad_norm": 0.7945712872010535,
+      "learning_rate": 1.7733530551094932e-07,
+      "loss": 1.4163,
+      "step": 2776
+    },
+    {
+      "epoch": 0.5763802407638025,
+      "grad_norm": 0.7227156515783685,
+      "learning_rate": 1.7722128964787338e-07,
+      "loss": 1.5796,
+      "step": 2777
+    },
+    {
+      "epoch": 0.5765877957658779,
+      "grad_norm": 1.1558996117114002,
+      "learning_rate": 1.771072919731695e-07,
+      "loss": 1.5096,
+      "step": 2778
+    },
+    {
+      "epoch": 0.5767953507679535,
+      "grad_norm": 0.69368801528994,
+      "learning_rate": 1.7699331253530624e-07,
+      "loss": 1.4507,
+      "step": 2779
+    },
+    {
+      "epoch": 0.5770029057700291,
+      "grad_norm": 0.7049065143249547,
+      "learning_rate": 1.7687935138274474e-07,
+      "loss": 1.493,
+      "step": 2780
+    },
+    {
+      "epoch": 0.5772104607721046,
+      "grad_norm": 0.880759807237636,
+      "learning_rate": 1.767654085639383e-07,
+      "loss": 1.5459,
+      "step": 2781
+    },
+    {
+      "epoch": 0.5774180157741802,
+      "grad_norm": 0.7924430722882332,
+      "learning_rate": 1.7665148412733229e-07,
+      "loss": 1.5162,
+      "step": 2782
+    },
+    {
+      "epoch": 0.5776255707762558,
+      "grad_norm": 0.735618198447245,
+      "learning_rate": 1.765375781213643e-07,
+      "loss": 1.5227,
+      "step": 2783
+    },
+    {
+      "epoch": 0.5778331257783312,
+      "grad_norm": 0.6610780007571779,
+      "learning_rate": 1.7642369059446435e-07,
+      "loss": 1.506,
+      "step": 2784
+    },
+    {
+      "epoch": 0.5780406807804068,
+      "grad_norm": 1.3411785743217628,
+      "learning_rate": 1.763098215950542e-07,
+      "loss": 1.5726,
+      "step": 2785
+    },
+    {
+      "epoch": 0.5782482357824824,
+      "grad_norm": 0.6685683317640891,
+      "learning_rate": 1.7619597117154807e-07,
+      "loss": 1.4742,
+      "step": 2786
+    },
+    {
+      "epoch": 0.5784557907845579,
+      "grad_norm": 0.6239754715574964,
+      "learning_rate": 1.7608213937235203e-07,
+      "loss": 1.4952,
+      "step": 2787
+    },
+    {
+      "epoch": 0.5786633457866335,
+      "grad_norm": 0.957657533000173,
+      "learning_rate": 1.7596832624586438e-07,
+      "loss": 1.5211,
+      "step": 2788
+    },
+    {
+      "epoch": 0.578870900788709,
+      "grad_norm": 0.7612572146783587,
+      "learning_rate": 1.758545318404755e-07,
+      "loss": 1.4679,
+      "step": 2789
+    },
+    {
+      "epoch": 0.5790784557907845,
+      "grad_norm": 1.6899257873814624,
+      "learning_rate": 1.757407562045676e-07,
+      "loss": 1.5534,
+      "step": 2790
+    },
+    {
+      "epoch": 0.5792860107928601,
+      "grad_norm": 0.7248436953699403,
+      "learning_rate": 1.756269993865151e-07,
+      "loss": 1.4489,
+      "step": 2791
+    },
+    {
+      "epoch": 0.5794935657949356,
+      "grad_norm": 1.2729965354170167,
+      "learning_rate": 1.755132614346846e-07,
+      "loss": 1.5434,
+      "step": 2792
+    },
+    {
+      "epoch": 0.5797011207970112,
+      "grad_norm": 1.0516639089080668,
+      "learning_rate": 1.7539954239743416e-07,
+      "loss": 1.5201,
+      "step": 2793
+    },
+    {
+      "epoch": 0.5799086757990868,
+      "grad_norm": 0.7181568457073596,
+      "learning_rate": 1.752858423231142e-07,
+      "loss": 1.5394,
+      "step": 2794
+    },
+    {
+      "epoch": 0.5801162308011623,
+      "grad_norm": 0.8539085521713808,
+      "learning_rate": 1.7517216126006704e-07,
+      "loss": 1.5428,
+      "step": 2795
+    },
+    {
+      "epoch": 0.5803237858032378,
+      "grad_norm": 0.803091835114635,
+      "learning_rate": 1.7505849925662678e-07,
+      "loss": 1.5689,
+      "step": 2796
+    },
+    {
+      "epoch": 0.5805313408053134,
+      "grad_norm": 0.6717985094200712,
+      "learning_rate": 1.7494485636111952e-07,
+      "loss": 1.5273,
+      "step": 2797
+    },
+    {
+      "epoch": 0.5807388958073889,
+      "grad_norm": 1.0031030593762762,
+      "learning_rate": 1.7483123262186314e-07,
+      "loss": 1.5406,
+      "step": 2798
+    },
+    {
+      "epoch": 0.5809464508094645,
+      "grad_norm": 0.7879728523416093,
+      "learning_rate": 1.7471762808716752e-07,
+      "loss": 1.567,
+      "step": 2799
+    },
+    {
+      "epoch": 0.5811540058115401,
+      "grad_norm": 0.8509554724376943,
+      "learning_rate": 1.7460404280533422e-07,
+      "loss": 1.3852,
+      "step": 2800
+    },
+    {
+      "epoch": 0.5813615608136156,
+      "grad_norm": 0.8771775591357174,
+      "learning_rate": 1.7449047682465685e-07,
+      "loss": 1.6476,
+      "step": 2801
+    },
+    {
+      "epoch": 0.5815691158156912,
+      "grad_norm": 0.694974595372767,
+      "learning_rate": 1.743769301934204e-07,
+      "loss": 1.4773,
+      "step": 2802
+    },
+    {
+      "epoch": 0.5817766708177667,
+      "grad_norm": 0.6686018603524398,
+      "learning_rate": 1.7426340295990208e-07,
+      "loss": 1.516,
+      "step": 2803
+    },
+    {
+      "epoch": 0.5819842258198422,
+      "grad_norm": 0.6899907104155022,
+      "learning_rate": 1.7414989517237054e-07,
+      "loss": 1.4854,
+      "step": 2804
+    },
+    {
+      "epoch": 0.5821917808219178,
+      "grad_norm": 0.6745164516360551,
+      "learning_rate": 1.7403640687908637e-07,
+      "loss": 1.5393,
+      "step": 2805
+    },
+    {
+      "epoch": 0.5823993358239934,
+      "grad_norm": 0.7095786756685708,
+      "learning_rate": 1.7392293812830164e-07,
+      "loss": 1.5309,
+      "step": 2806
+    },
+    {
+      "epoch": 0.5826068908260689,
+      "grad_norm": 0.8871783972376093,
+      "learning_rate": 1.7380948896826048e-07,
+      "loss": 1.5734,
+      "step": 2807
+    },
+    {
+      "epoch": 0.5828144458281445,
+      "grad_norm": 0.6666461280326869,
+      "learning_rate": 1.7369605944719822e-07,
+      "loss": 1.6016,
+      "step": 2808
+    },
+    {
+      "epoch": 0.58302200083022,
+      "grad_norm": 0.785056674409609,
+      "learning_rate": 1.7358264961334217e-07,
+      "loss": 1.5668,
+      "step": 2809
+    },
+    {
+      "epoch": 0.5832295558322955,
+      "grad_norm": 0.8790225430828947,
+      "learning_rate": 1.7346925951491124e-07,
+      "loss": 1.5746,
+      "step": 2810
+    },
+    {
+      "epoch": 0.5834371108343711,
+      "grad_norm": 0.8908804070811525,
+      "learning_rate": 1.7335588920011582e-07,
+      "loss": 1.5646,
+      "step": 2811
+    },
+    {
+      "epoch": 0.5836446658364467,
+      "grad_norm": 0.6455106310486284,
+      "learning_rate": 1.7324253871715802e-07,
+      "loss": 1.4157,
+      "step": 2812
+    },
+    {
+      "epoch": 0.5838522208385222,
+      "grad_norm": 1.2852366321554003,
+      "learning_rate": 1.731292081142314e-07,
+      "loss": 1.5639,
+      "step": 2813
+    },
+    {
+      "epoch": 0.5840597758405978,
+      "grad_norm": 0.736461843347046,
+      "learning_rate": 1.7301589743952115e-07,
+      "loss": 1.5274,
+      "step": 2814
+    },
+    {
+      "epoch": 0.5842673308426733,
+      "grad_norm": 1.6491995037109397,
+      "learning_rate": 1.7290260674120388e-07,
+      "loss": 1.5498,
+      "step": 2815
+    },
+    {
+      "epoch": 0.5844748858447488,
+      "grad_norm": 0.6870443890116192,
+      "learning_rate": 1.7278933606744794e-07,
+      "loss": 1.4876,
+      "step": 2816
+    },
+    {
+      "epoch": 0.5846824408468244,
+      "grad_norm": 0.7150788592562356,
+      "learning_rate": 1.7267608546641292e-07,
+      "loss": 1.5086,
+      "step": 2817
+    },
+    {
+      "epoch": 0.5848899958489,
+      "grad_norm": 0.8112294611213976,
+      "learning_rate": 1.7256285498624994e-07,
+      "loss": 1.5116,
+      "step": 2818
+    },
+    {
+      "epoch": 0.5850975508509755,
+      "grad_norm": 1.056554728121192,
+      "learning_rate": 1.724496446751017e-07,
+      "loss": 1.5391,
+      "step": 2819
+    },
+    {
+      "epoch": 0.5853051058530511,
+      "grad_norm": 0.778416612575924,
+      "learning_rate": 1.7233645458110208e-07,
+      "loss": 1.5524,
+      "step": 2820
+    },
+    {
+      "epoch": 0.5855126608551267,
+      "grad_norm": 1.7858654314912314,
+      "learning_rate": 1.722232847523766e-07,
+      "loss": 1.4902,
+      "step": 2821
+    },
+    {
+      "epoch": 0.5857202158572021,
+      "grad_norm": 0.7264239887411612,
+      "learning_rate": 1.7211013523704213e-07,
+      "loss": 1.5418,
+      "step": 2822
+    },
+    {
+      "epoch": 0.5859277708592777,
+      "grad_norm": 0.9124515378803588,
+      "learning_rate": 1.7199700608320664e-07,
+      "loss": 1.4999,
+      "step": 2823
+    },
+    {
+      "epoch": 0.5861353258613533,
+      "grad_norm": 0.7789997235064122,
+      "learning_rate": 1.7188389733896975e-07,
+      "loss": 1.4683,
+      "step": 2824
+    },
+    {
+      "epoch": 0.5863428808634288,
+      "grad_norm": 0.6846733302363334,
+      "learning_rate": 1.717708090524224e-07,
+      "loss": 1.4486,
+      "step": 2825
+    },
+    {
+      "epoch": 0.5865504358655044,
+      "grad_norm": 0.828452113655679,
+      "learning_rate": 1.7165774127164654e-07,
+      "loss": 1.6132,
+      "step": 2826
+    },
+    {
+      "epoch": 0.58675799086758,
+      "grad_norm": 1.0659231556445423,
+      "learning_rate": 1.715446940447157e-07,
+      "loss": 1.5378,
+      "step": 2827
+    },
+    {
+      "epoch": 0.5869655458696554,
+      "grad_norm": 1.2705205199548841,
+      "learning_rate": 1.714316674196946e-07,
+      "loss": 1.4992,
+      "step": 2828
+    },
+    {
+      "epoch": 0.587173100871731,
+      "grad_norm": 0.7733321292250058,
+      "learning_rate": 1.7131866144463905e-07,
+      "loss": 1.5633,
+      "step": 2829
+    },
+    {
+      "epoch": 0.5873806558738066,
+      "grad_norm": 0.7222036600015269,
+      "learning_rate": 1.7120567616759618e-07,
+      "loss": 1.5132,
+      "step": 2830
+    },
+    {
+      "epoch": 0.5875882108758821,
+      "grad_norm": 0.9664975459056787,
+      "learning_rate": 1.710927116366045e-07,
+      "loss": 1.6256,
+      "step": 2831
+    },
+    {
+      "epoch": 0.5877957658779577,
+      "grad_norm": 1.0454087338726974,
+      "learning_rate": 1.7097976789969332e-07,
+      "loss": 1.5205,
+      "step": 2832
+    },
+    {
+      "epoch": 0.5880033208800332,
+      "grad_norm": 0.6614692515709273,
+      "learning_rate": 1.7086684500488353e-07,
+      "loss": 1.4166,
+      "step": 2833
+    },
+    {
+      "epoch": 0.5882108758821087,
+      "grad_norm": 0.8483585823548292,
+      "learning_rate": 1.7075394300018674e-07,
+      "loss": 1.5012,
+      "step": 2834
+    },
+    {
+      "epoch": 0.5884184308841843,
+      "grad_norm": 0.9826675761446627,
+      "learning_rate": 1.7064106193360597e-07,
+      "loss": 1.5342,
+      "step": 2835
+    },
+    {
+      "epoch": 0.5886259858862598,
+      "grad_norm": 0.7255508303526077,
+      "learning_rate": 1.7052820185313533e-07,
+      "loss": 1.5023,
+      "step": 2836
+    },
+    {
+      "epoch": 0.5888335408883354,
+      "grad_norm": 0.6619937536364876,
+      "learning_rate": 1.7041536280675976e-07,
+      "loss": 1.5608,
+      "step": 2837
+    },
+    {
+      "epoch": 0.589041095890411,
+      "grad_norm": 0.8762113991446207,
+      "learning_rate": 1.7030254484245558e-07,
+      "loss": 1.5317,
+      "step": 2838
+    },
+    {
+      "epoch": 0.5892486508924865,
+      "grad_norm": 0.6825757964039468,
+      "learning_rate": 1.7018974800819002e-07,
+      "loss": 1.4647,
+      "step": 2839
+    },
+    {
+      "epoch": 0.589456205894562,
+      "grad_norm": 1.3496877116127028,
+      "learning_rate": 1.7007697235192115e-07,
+      "loss": 1.5164,
+      "step": 2840
+    },
+    {
+      "epoch": 0.5896637608966376,
+      "grad_norm": 0.9429121077855189,
+      "learning_rate": 1.6996421792159818e-07,
+      "loss": 1.5441,
+      "step": 2841
+    },
+    {
+      "epoch": 0.5898713158987131,
+      "grad_norm": 0.7896162106585859,
+      "learning_rate": 1.6985148476516148e-07,
+      "loss": 1.4951,
+      "step": 2842
+    },
+    {
+      "epoch": 0.5900788709007887,
+      "grad_norm": 0.7501720645012195,
+      "learning_rate": 1.6973877293054209e-07,
+      "loss": 1.5096,
+      "step": 2843
+    },
+    {
+      "epoch": 0.5902864259028643,
+      "grad_norm": 0.7885291660381466,
+      "learning_rate": 1.6962608246566205e-07,
+      "loss": 1.5437,
+      "step": 2844
+    },
+    {
+      "epoch": 0.5904939809049398,
+      "grad_norm": 1.1839814403887827,
+      "learning_rate": 1.6951341341843444e-07,
+      "loss": 1.53,
+      "step": 2845
+    },
+    {
+      "epoch": 0.5907015359070154,
+      "grad_norm": 6.715340988043474,
+      "learning_rate": 1.69400765836763e-07,
+      "loss": 1.4748,
+      "step": 2846
+    },
+    {
+      "epoch": 0.5909090909090909,
+      "grad_norm": 0.9135874048048324,
+      "learning_rate": 1.6928813976854267e-07,
+      "loss": 1.5122,
+      "step": 2847
+    },
+    {
+      "epoch": 0.5911166459111664,
+      "grad_norm": 1.0383435094613938,
+      "learning_rate": 1.6917553526165897e-07,
+      "loss": 1.4801,
+      "step": 2848
+    },
+    {
+      "epoch": 0.591324200913242,
+      "grad_norm": 1.087779549046269,
+      "learning_rate": 1.6906295236398837e-07,
+      "loss": 1.4537,
+      "step": 2849
+    },
+    {
+      "epoch": 0.5915317559153176,
+      "grad_norm": 0.8004657366491701,
+      "learning_rate": 1.6895039112339812e-07,
+      "loss": 1.6085,
+      "step": 2850
+    },
+    {
+      "epoch": 0.5917393109173931,
+      "grad_norm": 0.818620190195881,
+      "learning_rate": 1.688378515877463e-07,
+      "loss": 1.4738,
+      "step": 2851
+    },
+    {
+      "epoch": 0.5919468659194687,
+      "grad_norm": 0.9754305122536338,
+      "learning_rate": 1.6872533380488166e-07,
+      "loss": 1.472,
+      "step": 2852
+    },
+    {
+      "epoch": 0.5921544209215442,
+      "grad_norm": 0.8282078067858415,
+      "learning_rate": 1.6861283782264382e-07,
+      "loss": 1.4543,
+      "step": 2853
+    },
+    {
+      "epoch": 0.5923619759236197,
+      "grad_norm": 0.7777232443399855,
+      "learning_rate": 1.6850036368886315e-07,
+      "loss": 1.5037,
+      "step": 2854
+    },
+    {
+      "epoch": 0.5925695309256953,
+      "grad_norm": 1.0011051100462725,
+      "learning_rate": 1.6838791145136054e-07,
+      "loss": 1.5132,
+      "step": 2855
+    },
+    {
+      "epoch": 0.5927770859277709,
+      "grad_norm": 0.7061602178315153,
+      "learning_rate": 1.6827548115794773e-07,
+      "loss": 1.4822,
+      "step": 2856
+    },
+    {
+      "epoch": 0.5929846409298464,
+      "grad_norm": 0.8615442269748376,
+      "learning_rate": 1.6816307285642725e-07,
+      "loss": 1.4596,
+      "step": 2857
+    },
+    {
+      "epoch": 0.593192195931922,
+      "grad_norm": 0.7410582504633436,
+      "learning_rate": 1.6805068659459188e-07,
+      "loss": 1.4865,
+      "step": 2858
+    },
+    {
+      "epoch": 0.5933997509339975,
+      "grad_norm": 0.6648874348989493,
+      "learning_rate": 1.6793832242022544e-07,
+      "loss": 1.5381,
+      "step": 2859
+    },
+    {
+      "epoch": 0.593607305936073,
+      "grad_norm": 0.6694704310744576,
+      "learning_rate": 1.678259803811022e-07,
+      "loss": 1.4673,
+      "step": 2860
+    },
+    {
+      "epoch": 0.5938148609381486,
+      "grad_norm": 0.6576054240198026,
+      "learning_rate": 1.6771366052498686e-07,
+      "loss": 1.5323,
+      "step": 2861
+    },
+    {
+      "epoch": 0.5940224159402242,
+      "grad_norm": 0.9291845406302139,
+      "learning_rate": 1.6760136289963497e-07,
+      "loss": 1.5228,
+      "step": 2862
+    },
+    {
+      "epoch": 0.5942299709422997,
+      "grad_norm": 1.0831472246572866,
+      "learning_rate": 1.6748908755279252e-07,
+      "loss": 1.5669,
+      "step": 2863
+    },
+    {
+      "epoch": 0.5944375259443753,
+      "grad_norm": 0.7082129139775857,
+      "learning_rate": 1.673768345321959e-07,
+      "loss": 1.6155,
+      "step": 2864
+    },
+    {
+      "epoch": 0.5946450809464509,
+      "grad_norm": 0.7382426941565197,
+      "learning_rate": 1.672646038855722e-07,
+      "loss": 1.5399,
+      "step": 2865
+    },
+    {
+      "epoch": 0.5948526359485263,
+      "grad_norm": 2.157114757113836,
+      "learning_rate": 1.671523956606389e-07,
+      "loss": 1.5971,
+      "step": 2866
+    },
+    {
+      "epoch": 0.5950601909506019,
+      "grad_norm": 0.7760571163502992,
+      "learning_rate": 1.670402099051039e-07,
+      "loss": 1.5534,
+      "step": 2867
+    },
+    {
+      "epoch": 0.5952677459526775,
+      "grad_norm": 0.6557502436454352,
+      "learning_rate": 1.6692804666666565e-07,
+      "loss": 1.48,
+      "step": 2868
+    },
+    {
+      "epoch": 0.595475300954753,
+      "grad_norm": 0.9619804404913119,
+      "learning_rate": 1.6681590599301302e-07,
+      "loss": 1.4867,
+      "step": 2869
+    },
+    {
+      "epoch": 0.5956828559568286,
+      "grad_norm": 0.8467657154933467,
+      "learning_rate": 1.6670378793182516e-07,
+      "loss": 1.5047,
+      "step": 2870
+    },
+    {
+      "epoch": 0.5958904109589042,
+      "grad_norm": 0.9265342656401355,
+      "learning_rate": 1.665916925307717e-07,
+      "loss": 1.5953,
+      "step": 2871
+    },
+    {
+      "epoch": 0.5960979659609796,
+      "grad_norm": 0.6655733530141834,
+      "learning_rate": 1.664796198375128e-07,
+      "loss": 1.4923,
+      "step": 2872
+    },
+    {
+      "epoch": 0.5963055209630552,
+      "grad_norm": 0.7456955983870813,
+      "learning_rate": 1.6636756989969857e-07,
+      "loss": 1.4876,
+      "step": 2873
+    },
+    {
+      "epoch": 0.5965130759651308,
+      "grad_norm": 0.7998449652257649,
+      "learning_rate": 1.6625554276496976e-07,
+      "loss": 1.4516,
+      "step": 2874
+    },
+    {
+      "epoch": 0.5967206309672063,
+      "grad_norm": 0.6742621585339764,
+      "learning_rate": 1.6614353848095738e-07,
+      "loss": 1.5065,
+      "step": 2875
+    },
+    {
+      "epoch": 0.5969281859692819,
+      "grad_norm": 0.6570486109804717,
+      "learning_rate": 1.6603155709528257e-07,
+      "loss": 1.4571,
+      "step": 2876
+    },
+    {
+      "epoch": 0.5971357409713575,
+      "grad_norm": 2.4041789544860936,
+      "learning_rate": 1.6591959865555688e-07,
+      "loss": 1.5265,
+      "step": 2877
+    },
+    {
+      "epoch": 0.5973432959734329,
+      "grad_norm": 0.6247341425283687,
+      "learning_rate": 1.6580766320938214e-07,
+      "loss": 1.4432,
+      "step": 2878
+    },
+    {
+      "epoch": 0.5975508509755085,
+      "grad_norm": 0.668155820819984,
+      "learning_rate": 1.6569575080435027e-07,
+      "loss": 1.4596,
+      "step": 2879
+    },
+    {
+      "epoch": 0.597758405977584,
+      "grad_norm": 0.7758182196691806,
+      "learning_rate": 1.655838614880435e-07,
+      "loss": 1.6065,
+      "step": 2880
+    },
+    {
+      "epoch": 0.5979659609796596,
+      "grad_norm": 1.8880447355405923,
+      "learning_rate": 1.6547199530803414e-07,
+      "loss": 1.4297,
+      "step": 2881
+    },
+    {
+      "epoch": 0.5981735159817352,
+      "grad_norm": 0.7129015100545596,
+      "learning_rate": 1.6536015231188464e-07,
+      "loss": 1.5703,
+      "step": 2882
+    },
+    {
+      "epoch": 0.5983810709838107,
+      "grad_norm": 0.6491171660152925,
+      "learning_rate": 1.652483325471479e-07,
+      "loss": 1.4965,
+      "step": 2883
+    },
+    {
+      "epoch": 0.5985886259858862,
+      "grad_norm": 0.6944868828223861,
+      "learning_rate": 1.6513653606136652e-07,
+      "loss": 1.4748,
+      "step": 2884
+    },
+    {
+      "epoch": 0.5987961809879618,
+      "grad_norm": 0.7715275444741108,
+      "learning_rate": 1.6502476290207349e-07,
+      "loss": 1.522,
+      "step": 2885
+    },
+    {
+      "epoch": 0.5990037359900373,
+      "grad_norm": 0.6678937476334154,
+      "learning_rate": 1.6491301311679177e-07,
+      "loss": 1.5166,
+      "step": 2886
+    },
+    {
+      "epoch": 0.5992112909921129,
+      "grad_norm": 1.3858146481318514,
+      "learning_rate": 1.648012867530344e-07,
+      "loss": 1.5477,
+      "step": 2887
+    },
+    {
+      "epoch": 0.5994188459941885,
+      "grad_norm": 1.0642207640173227,
+      "learning_rate": 1.646895838583044e-07,
+      "loss": 1.5675,
+      "step": 2888
+    },
+    {
+      "epoch": 0.599626400996264,
+      "grad_norm": 1.135044182499458,
+      "learning_rate": 1.6457790448009502e-07,
+      "loss": 1.4908,
+      "step": 2889
+    },
+    {
+      "epoch": 0.5998339559983396,
+      "grad_norm": 1.1370257180738672,
+      "learning_rate": 1.6446624866588922e-07,
+      "loss": 1.5351,
+      "step": 2890
+    },
+    {
+      "epoch": 0.6000415110004151,
+      "grad_norm": 0.6809034803841085,
+      "learning_rate": 1.6435461646316013e-07,
+      "loss": 1.4821,
+      "step": 2891
+    },
+    {
+      "epoch": 0.6002490660024906,
+      "grad_norm": 0.7834256155977669,
+      "learning_rate": 1.6424300791937088e-07,
+      "loss": 1.5516,
+      "step": 2892
+    },
+    {
+      "epoch": 0.6004566210045662,
+      "grad_norm": 1.408629801948699,
+      "learning_rate": 1.641314230819744e-07,
+      "loss": 1.5239,
+      "step": 2893
+    },
+    {
+      "epoch": 0.6006641760066418,
+      "grad_norm": 1.137349098778385,
+      "learning_rate": 1.6401986199841354e-07,
+      "loss": 1.4822,
+      "step": 2894
+    },
+    {
+      "epoch": 0.6008717310087173,
+      "grad_norm": 0.7530387197886126,
+      "learning_rate": 1.6390832471612125e-07,
+      "loss": 1.5264,
+      "step": 2895
+    },
+    {
+      "epoch": 0.6010792860107929,
+      "grad_norm": 0.7325961648805188,
+      "learning_rate": 1.637968112825201e-07,
+      "loss": 1.555,
+      "step": 2896
+    },
+    {
+      "epoch": 0.6012868410128684,
+      "grad_norm": 0.7195609503537667,
+      "learning_rate": 1.636853217450227e-07,
+      "loss": 1.6114,
+      "step": 2897
+    },
+    {
+      "epoch": 0.6014943960149439,
+      "grad_norm": 0.6988482173041103,
+      "learning_rate": 1.6357385615103141e-07,
+      "loss": 1.5373,
+      "step": 2898
+    },
+    {
+      "epoch": 0.6017019510170195,
+      "grad_norm": 0.8099116133934874,
+      "learning_rate": 1.6346241454793844e-07,
+      "loss": 1.5275,
+      "step": 2899
+    },
+    {
+      "epoch": 0.6019095060190951,
+      "grad_norm": 0.7748074177026885,
+      "learning_rate": 1.633509969831258e-07,
+      "loss": 1.5355,
+      "step": 2900
+    },
+    {
+      "epoch": 0.6021170610211706,
+      "grad_norm": 1.6981014660023268,
+      "learning_rate": 1.6323960350396532e-07,
+      "loss": 1.5169,
+      "step": 2901
+    },
+    {
+      "epoch": 0.6023246160232462,
+      "grad_norm": 0.808798216846251,
+      "learning_rate": 1.6312823415781858e-07,
+      "loss": 1.4549,
+      "step": 2902
+    },
+    {
+      "epoch": 0.6025321710253217,
+      "grad_norm": 0.6800704743294834,
+      "learning_rate": 1.6301688899203673e-07,
+      "loss": 1.5851,
+      "step": 2903
+    },
+    {
+      "epoch": 0.6027397260273972,
+      "grad_norm": 0.9308861948639792,
+      "learning_rate": 1.629055680539609e-07,
+      "loss": 1.5198,
+      "step": 2904
+    },
+    {
+      "epoch": 0.6029472810294728,
+      "grad_norm": 0.9177301727407988,
+      "learning_rate": 1.627942713909218e-07,
+      "loss": 1.5081,
+      "step": 2905
+    },
+    {
+      "epoch": 0.6031548360315484,
+      "grad_norm": 0.8365186296662164,
+      "learning_rate": 1.6268299905023967e-07,
+      "loss": 1.4431,
+      "step": 2906
+    },
+    {
+      "epoch": 0.6033623910336239,
+      "grad_norm": 0.8972887384886354,
+      "learning_rate": 1.6257175107922482e-07,
+      "loss": 1.5495,
+      "step": 2907
+    },
+    {
+      "epoch": 0.6035699460356995,
+      "grad_norm": 0.7223700493904724,
+      "learning_rate": 1.624605275251767e-07,
+      "loss": 1.5725,
+      "step": 2908
+    },
+    {
+      "epoch": 0.603777501037775,
+      "grad_norm": 0.7406390612337366,
+      "learning_rate": 1.6234932843538464e-07,
+      "loss": 1.537,
+      "step": 2909
+    },
+    {
+      "epoch": 0.6039850560398505,
+      "grad_norm": 0.6683050447709474,
+      "learning_rate": 1.6223815385712773e-07,
+      "loss": 1.4619,
+      "step": 2910
+    },
+    {
+      "epoch": 0.6041926110419261,
+      "grad_norm": 0.9611396145243545,
+      "learning_rate": 1.6212700383767418e-07,
+      "loss": 1.4987,
+      "step": 2911
+    },
+    {
+      "epoch": 0.6044001660440017,
+      "grad_norm": 0.9775699785624652,
+      "learning_rate": 1.6201587842428216e-07,
+      "loss": 1.5025,
+      "step": 2912
+    },
+    {
+      "epoch": 0.6046077210460772,
+      "grad_norm": 0.8942348272111529,
+      "learning_rate": 1.6190477766419935e-07,
+      "loss": 1.4386,
+      "step": 2913
+    },
+    {
+      "epoch": 0.6048152760481528,
+      "grad_norm": 0.6786349585185361,
+      "learning_rate": 1.6179370160466262e-07,
+      "loss": 1.4386,
+      "step": 2914
+    },
+    {
+      "epoch": 0.6050228310502284,
+      "grad_norm": 0.7774037627875939,
+      "learning_rate": 1.6168265029289868e-07,
+      "loss": 1.5356,
+      "step": 2915
+    },
+    {
+      "epoch": 0.6052303860523038,
+      "grad_norm": 1.2456115676074435,
+      "learning_rate": 1.6157162377612368e-07,
+      "loss": 1.5196,
+      "step": 2916
+    },
+    {
+      "epoch": 0.6054379410543794,
+      "grad_norm": 0.8000454177858142,
+      "learning_rate": 1.6146062210154302e-07,
+      "loss": 1.4826,
+      "step": 2917
+    },
+    {
+      "epoch": 0.605645496056455,
+      "grad_norm": 0.8326334827030125,
+      "learning_rate": 1.6134964531635173e-07,
+      "loss": 1.5512,
+      "step": 2918
+    },
+    {
+      "epoch": 0.6058530510585305,
+      "grad_norm": 1.2482680568309497,
+      "learning_rate": 1.6123869346773416e-07,
+      "loss": 1.4681,
+      "step": 2919
+    },
+    {
+      "epoch": 0.6060606060606061,
+      "grad_norm": 0.9322448888291446,
+      "learning_rate": 1.611277666028641e-07,
+      "loss": 1.4428,
+      "step": 2920
+    },
+    {
+      "epoch": 0.6062681610626817,
+      "grad_norm": 0.9798057768306832,
+      "learning_rate": 1.6101686476890467e-07,
+      "loss": 1.4984,
+      "step": 2921
+    },
+    {
+      "epoch": 0.6064757160647571,
+      "grad_norm": 0.6722285410165774,
+      "learning_rate": 1.6090598801300855e-07,
+      "loss": 1.4506,
+      "step": 2922
+    },
+    {
+      "epoch": 0.6066832710668327,
+      "grad_norm": 3.149224533660752,
+      "learning_rate": 1.6079513638231737e-07,
+      "loss": 1.5731,
+      "step": 2923
+    },
+    {
+      "epoch": 0.6068908260689082,
+      "grad_norm": 0.672411064248572,
+      "learning_rate": 1.6068430992396249e-07,
+      "loss": 1.5029,
+      "step": 2924
+    },
+    {
+      "epoch": 0.6070983810709838,
+      "grad_norm": 0.8382347554066111,
+      "learning_rate": 1.605735086850642e-07,
+      "loss": 1.528,
+      "step": 2925
+    },
+    {
+      "epoch": 0.6073059360730594,
+      "grad_norm": 0.6505680097557937,
+      "learning_rate": 1.604627327127323e-07,
+      "loss": 1.5694,
+      "step": 2926
+    },
+    {
+      "epoch": 0.6075134910751349,
+      "grad_norm": 0.6451655225633504,
+      "learning_rate": 1.6035198205406592e-07,
+      "loss": 1.502,
+      "step": 2927
+    },
+    {
+      "epoch": 0.6077210460772104,
+      "grad_norm": 2.1242372854079092,
+      "learning_rate": 1.6024125675615316e-07,
+      "loss": 1.509,
+      "step": 2928
+    },
+    {
+      "epoch": 0.607928601079286,
+      "grad_norm": 0.7388999339118123,
+      "learning_rate": 1.6013055686607152e-07,
+      "loss": 1.5902,
+      "step": 2929
+    },
+    {
+      "epoch": 0.6081361560813615,
+      "grad_norm": 2.2249876806638897,
+      "learning_rate": 1.600198824308877e-07,
+      "loss": 1.5577,
+      "step": 2930
+    },
+    {
+      "epoch": 0.6083437110834371,
+      "grad_norm": 0.7387184105390527,
+      "learning_rate": 1.599092334976574e-07,
+      "loss": 1.52,
+      "step": 2931
+    },
+    {
+      "epoch": 0.6085512660855127,
+      "grad_norm": 0.8797744193713711,
+      "learning_rate": 1.5979861011342573e-07,
+      "loss": 1.4782,
+      "step": 2932
+    },
+    {
+      "epoch": 0.6087588210875882,
+      "grad_norm": 1.466926016085479,
+      "learning_rate": 1.596880123252269e-07,
+      "loss": 1.4952,
+      "step": 2933
+    },
+    {
+      "epoch": 0.6089663760896638,
+      "grad_norm": 0.7110906563008449,
+      "learning_rate": 1.5957744018008392e-07,
+      "loss": 1.3958,
+      "step": 2934
+    },
+    {
+      "epoch": 0.6091739310917393,
+      "grad_norm": 6.768917304805736,
+      "learning_rate": 1.594668937250092e-07,
+      "loss": 1.5072,
+      "step": 2935
+    },
+    {
+      "epoch": 0.6093814860938148,
+      "grad_norm": 1.4932472007636786,
+      "learning_rate": 1.5935637300700434e-07,
+      "loss": 1.3895,
+      "step": 2936
+    },
+    {
+      "epoch": 0.6095890410958904,
+      "grad_norm": 0.7327965232562699,
+      "learning_rate": 1.592458780730596e-07,
+      "loss": 1.5003,
+      "step": 2937
+    },
+    {
+      "epoch": 0.609796596097966,
+      "grad_norm": 1.2306496974469834,
+      "learning_rate": 1.591354089701546e-07,
+      "loss": 1.5329,
+      "step": 2938
+    },
+    {
+      "epoch": 0.6100041511000415,
+      "grad_norm": 0.7948189873014087,
+      "learning_rate": 1.590249657452579e-07,
+      "loss": 1.5602,
+      "step": 2939
+    },
+    {
+      "epoch": 0.6102117061021171,
+      "grad_norm": 0.9783436064365502,
+      "learning_rate": 1.5891454844532688e-07,
+      "loss": 1.5619,
+      "step": 2940
+    },
+    {
+      "epoch": 0.6104192611041926,
+      "grad_norm": 1.688177212488804,
+      "learning_rate": 1.5880415711730812e-07,
+      "loss": 1.4981,
+      "step": 2941
+    },
+    {
+      "epoch": 0.6106268161062681,
+      "grad_norm": 1.2624233649307406,
+      "learning_rate": 1.5869379180813716e-07,
+      "loss": 1.4939,
+      "step": 2942
+    },
+    {
+      "epoch": 0.6108343711083437,
+      "grad_norm": 1.8929786269348523,
+      "learning_rate": 1.5858345256473832e-07,
+      "loss": 1.5054,
+      "step": 2943
+    },
+    {
+      "epoch": 0.6110419261104193,
+      "grad_norm": 1.215104235857268,
+      "learning_rate": 1.584731394340249e-07,
+      "loss": 1.5444,
+      "step": 2944
+    },
+    {
+      "epoch": 0.6112494811124948,
+      "grad_norm": 1.0269210213539417,
+      "learning_rate": 1.5836285246289918e-07,
+      "loss": 1.4612,
+      "step": 2945
+    },
+    {
+      "epoch": 0.6114570361145704,
+      "grad_norm": 1.4481906241643254,
+      "learning_rate": 1.5825259169825223e-07,
+      "loss": 1.6106,
+      "step": 2946
+    },
+    {
+      "epoch": 0.611664591116646,
+      "grad_norm": 4.674749038767822,
+      "learning_rate": 1.5814235718696393e-07,
+      "loss": 1.5739,
+      "step": 2947
+    },
+    {
+      "epoch": 0.6118721461187214,
+      "grad_norm": 0.8921273018258443,
+      "learning_rate": 1.580321489759032e-07,
+      "loss": 1.5583,
+      "step": 2948
+    },
+    {
+      "epoch": 0.612079701120797,
+      "grad_norm": 1.9281278345913548,
+      "learning_rate": 1.5792196711192753e-07,
+      "loss": 1.5313,
+      "step": 2949
+    },
+    {
+      "epoch": 0.6122872561228726,
+      "grad_norm": 0.6416354612437689,
+      "learning_rate": 1.5781181164188335e-07,
+      "loss": 1.4934,
+      "step": 2950
+    },
+    {
+      "epoch": 0.6124948111249481,
+      "grad_norm": 1.4831932699606183,
+      "learning_rate": 1.5770168261260594e-07,
+      "loss": 1.5226,
+      "step": 2951
+    },
+    {
+      "epoch": 0.6127023661270237,
+      "grad_norm": 0.743602973028861,
+      "learning_rate": 1.5759158007091906e-07,
+      "loss": 1.5191,
+      "step": 2952
+    },
+    {
+      "epoch": 0.6129099211290993,
+      "grad_norm": 1.0871439948594217,
+      "learning_rate": 1.5748150406363553e-07,
+      "loss": 1.36,
+      "step": 2953
+    },
+    {
+      "epoch": 0.6131174761311747,
+      "grad_norm": 1.053034430069096,
+      "learning_rate": 1.573714546375567e-07,
+      "loss": 1.4987,
+      "step": 2954
+    },
+    {
+      "epoch": 0.6133250311332503,
+      "grad_norm": 0.7272434227605858,
+      "learning_rate": 1.5726143183947267e-07,
+      "loss": 1.4755,
+      "step": 2955
+    },
+    {
+      "epoch": 0.6135325861353259,
+      "grad_norm": 0.6521455351751879,
+      "learning_rate": 1.5715143571616217e-07,
+      "loss": 1.4769,
+      "step": 2956
+    },
+    {
+      "epoch": 0.6137401411374014,
+      "grad_norm": 1.838780836723481,
+      "learning_rate": 1.5704146631439272e-07,
+      "loss": 1.4806,
+      "step": 2957
+    },
+    {
+      "epoch": 0.613947696139477,
+      "grad_norm": 0.662858179250531,
+      "learning_rate": 1.569315236809203e-07,
+      "loss": 1.4577,
+      "step": 2958
+    },
+    {
+      "epoch": 0.6141552511415526,
+      "grad_norm": 1.1895464943054594,
+      "learning_rate": 1.5682160786248963e-07,
+      "loss": 1.5234,
+      "step": 2959
+    },
+    {
+      "epoch": 0.614362806143628,
+      "grad_norm": 0.7644677955325846,
+      "learning_rate": 1.5671171890583404e-07,
+      "loss": 1.4675,
+      "step": 2960
+    },
+    {
+      "epoch": 0.6145703611457036,
+      "grad_norm": 1.0595897828837544,
+      "learning_rate": 1.5660185685767538e-07,
+      "loss": 1.5153,
+      "step": 2961
+    },
+    {
+      "epoch": 0.6147779161477792,
+      "grad_norm": 0.7390601781416092,
+      "learning_rate": 1.56492021764724e-07,
+      "loss": 1.5099,
+      "step": 2962
+    },
+    {
+      "epoch": 0.6149854711498547,
+      "grad_norm": 0.6700285618159091,
+      "learning_rate": 1.5638221367367898e-07,
+      "loss": 1.5058,
+      "step": 2963
+    },
+    {
+      "epoch": 0.6151930261519303,
+      "grad_norm": 0.958994073180166,
+      "learning_rate": 1.5627243263122774e-07,
+      "loss": 1.4962,
+      "step": 2964
+    },
+    {
+      "epoch": 0.6154005811540059,
+      "grad_norm": 0.8059355602514806,
+      "learning_rate": 1.5616267868404625e-07,
+      "loss": 1.5718,
+      "step": 2965
+    },
+    {
+      "epoch": 0.6156081361560813,
+      "grad_norm": 2.709452211139489,
+      "learning_rate": 1.5605295187879905e-07,
+      "loss": 1.5221,
+      "step": 2966
+    },
+    {
+      "epoch": 0.6158156911581569,
+      "grad_norm": 0.7243686287520321,
+      "learning_rate": 1.5594325226213893e-07,
+      "loss": 1.4802,
+      "step": 2967
+    },
+    {
+      "epoch": 0.6160232461602324,
+      "grad_norm": 0.9907973640748218,
+      "learning_rate": 1.5583357988070743e-07,
+      "loss": 1.527,
+      "step": 2968
+    },
+    {
+      "epoch": 0.616230801162308,
+      "grad_norm": 1.7532901521611337,
+      "learning_rate": 1.5572393478113415e-07,
+      "loss": 1.5197,
+      "step": 2969
+    },
+    {
+      "epoch": 0.6164383561643836,
+      "grad_norm": 1.0510418178863643,
+      "learning_rate": 1.5561431701003738e-07,
+      "loss": 1.577,
+      "step": 2970
+    },
+    {
+      "epoch": 0.6166459111664591,
+      "grad_norm": 0.8542329308784178,
+      "learning_rate": 1.555047266140238e-07,
+      "loss": 1.4479,
+      "step": 2971
+    },
+    {
+      "epoch": 0.6168534661685346,
+      "grad_norm": 0.6714615169200676,
+      "learning_rate": 1.553951636396881e-07,
+      "loss": 1.4762,
+      "step": 2972
+    },
+    {
+      "epoch": 0.6170610211706102,
+      "grad_norm": 1.0805127152235228,
+      "learning_rate": 1.5528562813361363e-07,
+      "loss": 1.4916,
+      "step": 2973
+    },
+    {
+      "epoch": 0.6172685761726857,
+      "grad_norm": 0.7570564020218621,
+      "learning_rate": 1.551761201423721e-07,
+      "loss": 1.4371,
+      "step": 2974
+    },
+    {
+      "epoch": 0.6174761311747613,
+      "grad_norm": 0.954274125501317,
+      "learning_rate": 1.5506663971252328e-07,
+      "loss": 1.4952,
+      "step": 2975
+    },
+    {
+      "epoch": 0.6176836861768369,
+      "grad_norm": 0.8922934207568118,
+      "learning_rate": 1.549571868906153e-07,
+      "loss": 1.5672,
+      "step": 2976
+    },
+    {
+      "epoch": 0.6178912411789124,
+      "grad_norm": 0.7830545483179003,
+      "learning_rate": 1.5484776172318478e-07,
+      "loss": 1.5184,
+      "step": 2977
+    },
+    {
+      "epoch": 0.618098796180988,
+      "grad_norm": 0.8595535477644503,
+      "learning_rate": 1.5473836425675622e-07,
+      "loss": 1.4959,
+      "step": 2978
+    },
+    {
+      "epoch": 0.6183063511830635,
+      "grad_norm": 0.7796394384734499,
+      "learning_rate": 1.5462899453784255e-07,
+      "loss": 1.5353,
+      "step": 2979
+    },
+    {
+      "epoch": 0.618513906185139,
+      "grad_norm": 0.6474049726223324,
+      "learning_rate": 1.5451965261294495e-07,
+      "loss": 1.5632,
+      "step": 2980
+    },
+    {
+      "epoch": 0.6187214611872146,
+      "grad_norm": 1.1461043069125916,
+      "learning_rate": 1.544103385285527e-07,
+      "loss": 1.5501,
+      "step": 2981
+    },
+    {
+      "epoch": 0.6189290161892902,
+      "grad_norm": 0.7409050555421258,
+      "learning_rate": 1.543010523311431e-07,
+      "loss": 1.6144,
+      "step": 2982
+    },
+    {
+      "epoch": 0.6191365711913657,
+      "grad_norm": 0.7185504687166825,
+      "learning_rate": 1.541917940671819e-07,
+      "loss": 1.5007,
+      "step": 2983
+    },
+    {
+      "epoch": 0.6193441261934413,
+      "grad_norm": 0.7694250443273994,
+      "learning_rate": 1.5408256378312266e-07,
+      "loss": 1.5075,
+      "step": 2984
+    },
+    {
+      "epoch": 0.6195516811955168,
+      "grad_norm": 0.7959331632363079,
+      "learning_rate": 1.5397336152540737e-07,
+      "loss": 1.5204,
+      "step": 2985
+    },
+    {
+      "epoch": 0.6197592361975923,
+      "grad_norm": 0.9809280978612099,
+      "learning_rate": 1.5386418734046592e-07,
+      "loss": 1.4766,
+      "step": 2986
+    },
+    {
+      "epoch": 0.6199667911996679,
+      "grad_norm": 0.907104566372969,
+      "learning_rate": 1.5375504127471614e-07,
+      "loss": 1.5986,
+      "step": 2987
+    },
+    {
+      "epoch": 0.6201743462017435,
+      "grad_norm": 1.1534459656647642,
+      "learning_rate": 1.5364592337456404e-07,
+      "loss": 1.5532,
+      "step": 2988
+    },
+    {
+      "epoch": 0.620381901203819,
+      "grad_norm": 1.0827506933797901,
+      "learning_rate": 1.5353683368640385e-07,
+      "loss": 1.5647,
+      "step": 2989
+    },
+    {
+      "epoch": 0.6205894562058946,
+      "grad_norm": 1.0170255044891223,
+      "learning_rate": 1.5342777225661743e-07,
+      "loss": 1.4406,
+      "step": 2990
+    },
+    {
+      "epoch": 0.6207970112079702,
+      "grad_norm": 3.197352124616111,
+      "learning_rate": 1.533187391315748e-07,
+      "loss": 1.5702,
+      "step": 2991
+    },
+    {
+      "epoch": 0.6210045662100456,
+      "grad_norm": 0.7667294217408318,
+      "learning_rate": 1.532097343576341e-07,
+      "loss": 1.536,
+      "step": 2992
+    },
+    {
+      "epoch": 0.6212121212121212,
+      "grad_norm": 1.037351167495545,
+      "learning_rate": 1.5310075798114106e-07,
+      "loss": 1.5387,
+      "step": 2993
+    },
+    {
+      "epoch": 0.6214196762141968,
+      "grad_norm": 2.85295211655403,
+      "learning_rate": 1.5299181004842966e-07,
+      "loss": 1.4951,
+      "step": 2994
+    },
+    {
+      "epoch": 0.6216272312162723,
+      "grad_norm": 0.9479156853728501,
+      "learning_rate": 1.5288289060582176e-07,
+      "loss": 1.505,
+      "step": 2995
+    },
+    {
+      "epoch": 0.6218347862183479,
+      "grad_norm": 0.8104359866162321,
+      "learning_rate": 1.527739996996268e-07,
+      "loss": 1.5047,
+      "step": 2996
+    },
+    {
+      "epoch": 0.6220423412204235,
+      "grad_norm": 1.058396992585166,
+      "learning_rate": 1.5266513737614255e-07,
+      "loss": 1.5996,
+      "step": 2997
+    },
+    {
+      "epoch": 0.6222498962224989,
+      "grad_norm": 1.1725620408714186,
+      "learning_rate": 1.5255630368165418e-07,
+      "loss": 1.4962,
+      "step": 2998
+    },
+    {
+      "epoch": 0.6224574512245745,
+      "grad_norm": 0.7782880880570476,
+      "learning_rate": 1.5244749866243495e-07,
+      "loss": 1.4797,
+      "step": 2999
+    },
+    {
+      "epoch": 0.6226650062266501,
+      "grad_norm": 0.6842865401473562,
+      "learning_rate": 1.5233872236474583e-07,
+      "loss": 1.4745,
+      "step": 3000
+    },
+    {
+      "epoch": 0.6228725612287256,
+      "grad_norm": 0.8813136241253349,
+      "learning_rate": 1.5222997483483577e-07,
+      "loss": 1.4815,
+      "step": 3001
+    },
+    {
+      "epoch": 0.6230801162308012,
+      "grad_norm": 0.6537692674582971,
+      "learning_rate": 1.521212561189411e-07,
+      "loss": 1.5509,
+      "step": 3002
+    },
+    {
+      "epoch": 0.6232876712328768,
+      "grad_norm": 0.633355514782118,
+      "learning_rate": 1.5201256626328628e-07,
+      "loss": 1.4821,
+      "step": 3003
+    },
+    {
+      "epoch": 0.6234952262349522,
+      "grad_norm": 0.9616770146352793,
+      "learning_rate": 1.5190390531408342e-07,
+      "loss": 1.5782,
+      "step": 3004
+    },
+    {
+      "epoch": 0.6237027812370278,
+      "grad_norm": 0.9021480042707577,
+      "learning_rate": 1.517952733175321e-07,
+      "loss": 1.4727,
+      "step": 3005
+    },
+    {
+      "epoch": 0.6239103362391034,
+      "grad_norm": 1.9363805692078735,
+      "learning_rate": 1.516866703198198e-07,
+      "loss": 1.5533,
+      "step": 3006
+    },
+    {
+      "epoch": 0.6241178912411789,
+      "grad_norm": 0.7041329497262746,
+      "learning_rate": 1.515780963671217e-07,
+      "loss": 1.5163,
+      "step": 3007
+    },
+    {
+      "epoch": 0.6243254462432545,
+      "grad_norm": 0.7197669810289279,
+      "learning_rate": 1.5146955150560054e-07,
+      "loss": 1.5024,
+      "step": 3008
+    },
+    {
+      "epoch": 0.6245330012453301,
+      "grad_norm": 0.7821402677509988,
+      "learning_rate": 1.5136103578140666e-07,
+      "loss": 1.5532,
+      "step": 3009
+    },
+    {
+      "epoch": 0.6247405562474055,
+      "grad_norm": 0.783264355980897,
+      "learning_rate": 1.5125254924067813e-07,
+      "loss": 1.5301,
+      "step": 3010
+    },
+    {
+      "epoch": 0.6249481112494811,
+      "grad_norm": 0.7170160991048493,
+      "learning_rate": 1.511440919295405e-07,
+      "loss": 1.5392,
+      "step": 3011
+    },
+    {
+      "epoch": 0.6251556662515566,
+      "grad_norm": 1.3743728472889096,
+      "learning_rate": 1.5103566389410701e-07,
+      "loss": 1.4351,
+      "step": 3012
+    },
+    {
+      "epoch": 0.6253632212536322,
+      "grad_norm": 0.689914167923522,
+      "learning_rate": 1.509272651804783e-07,
+      "loss": 1.4229,
+      "step": 3013
+    },
+    {
+      "epoch": 0.6255707762557078,
+      "grad_norm": 0.8364211499808111,
+      "learning_rate": 1.5081889583474264e-07,
+      "loss": 1.5598,
+      "step": 3014
+    },
+    {
+      "epoch": 0.6257783312577833,
+      "grad_norm": 1.762970507255435,
+      "learning_rate": 1.5071055590297585e-07,
+      "loss": 1.54,
+      "step": 3015
+    },
+    {
+      "epoch": 0.6259858862598588,
+      "grad_norm": 0.7169625524201716,
+      "learning_rate": 1.5060224543124113e-07,
+      "loss": 1.534,
+      "step": 3016
+    },
+    {
+      "epoch": 0.6261934412619344,
+      "grad_norm": 0.6420430959541432,
+      "learning_rate": 1.504939644655893e-07,
+      "loss": 1.5144,
+      "step": 3017
+    },
+    {
+      "epoch": 0.6264009962640099,
+      "grad_norm": 0.6569486396456943,
+      "learning_rate": 1.5038571305205846e-07,
+      "loss": 1.4862,
+      "step": 3018
+    },
+    {
+      "epoch": 0.6266085512660855,
+      "grad_norm": 0.7611164666440157,
+      "learning_rate": 1.502774912366743e-07,
+      "loss": 1.5575,
+      "step": 3019
+    },
+    {
+      "epoch": 0.6268161062681611,
+      "grad_norm": 0.7973821180491858,
+      "learning_rate": 1.5016929906544978e-07,
+      "loss": 1.5356,
+      "step": 3020
+    },
+    {
+      "epoch": 0.6270236612702366,
+      "grad_norm": 0.7342341745643031,
+      "learning_rate": 1.5006113658438545e-07,
+      "loss": 1.4536,
+      "step": 3021
+    },
+    {
+      "epoch": 0.6272312162723122,
+      "grad_norm": 1.6485937584072101,
+      "learning_rate": 1.49953003839469e-07,
+      "loss": 1.54,
+      "step": 3022
+    },
+    {
+      "epoch": 0.6274387712743877,
+      "grad_norm": 0.7086689226127876,
+      "learning_rate": 1.4984490087667575e-07,
+      "loss": 1.482,
+      "step": 3023
+    },
+    {
+      "epoch": 0.6276463262764632,
+      "grad_norm": 1.1438750820212285,
+      "learning_rate": 1.4973682774196817e-07,
+      "loss": 1.5799,
+      "step": 3024
+    },
+    {
+      "epoch": 0.6278538812785388,
+      "grad_norm": 0.8542738698924801,
+      "learning_rate": 1.4962878448129596e-07,
+      "loss": 1.4845,
+      "step": 3025
+    },
+    {
+      "epoch": 0.6280614362806144,
+      "grad_norm": 0.7921797464277974,
+      "learning_rate": 1.4952077114059635e-07,
+      "loss": 1.5561,
+      "step": 3026
+    },
+    {
+      "epoch": 0.6282689912826899,
+      "grad_norm": 0.9463312455449134,
+      "learning_rate": 1.4941278776579384e-07,
+      "loss": 1.5487,
+      "step": 3027
+    },
+    {
+      "epoch": 0.6284765462847655,
+      "grad_norm": 0.7336887354061277,
+      "learning_rate": 1.4930483440279988e-07,
+      "loss": 1.4212,
+      "step": 3028
+    },
+    {
+      "epoch": 0.628684101286841,
+      "grad_norm": 0.6651588954425106,
+      "learning_rate": 1.4919691109751348e-07,
+      "loss": 1.4534,
+      "step": 3029
+    },
+    {
+      "epoch": 0.6288916562889165,
+      "grad_norm": 0.6058886376895825,
+      "learning_rate": 1.4908901789582086e-07,
+      "loss": 1.5682,
+      "step": 3030
+    },
+    {
+      "epoch": 0.6290992112909921,
+      "grad_norm": 0.7547659377690231,
+      "learning_rate": 1.4898115484359516e-07,
+      "loss": 1.5295,
+      "step": 3031
+    },
+    {
+      "epoch": 0.6293067662930677,
+      "grad_norm": 1.6267872869079583,
+      "learning_rate": 1.4887332198669702e-07,
+      "loss": 1.4697,
+      "step": 3032
+    },
+    {
+      "epoch": 0.6295143212951432,
+      "grad_norm": 0.9037490609311014,
+      "learning_rate": 1.487655193709741e-07,
+      "loss": 1.5508,
+      "step": 3033
+    },
+    {
+      "epoch": 0.6297218762972188,
+      "grad_norm": 1.4427007845775373,
+      "learning_rate": 1.486577470422611e-07,
+      "loss": 1.4478,
+      "step": 3034
+    },
+    {
+      "epoch": 0.6299294312992944,
+      "grad_norm": 1.4216909229705654,
+      "learning_rate": 1.4855000504638e-07,
+      "loss": 1.5714,
+      "step": 3035
+    },
+    {
+      "epoch": 0.6301369863013698,
+      "grad_norm": 0.7270077252043348,
+      "learning_rate": 1.4844229342913996e-07,
+      "loss": 1.5215,
+      "step": 3036
+    },
+    {
+      "epoch": 0.6303445413034454,
+      "grad_norm": 0.7634573109246932,
+      "learning_rate": 1.4833461223633697e-07,
+      "loss": 1.6044,
+      "step": 3037
+    },
+    {
+      "epoch": 0.630552096305521,
+      "grad_norm": 0.7004038343649741,
+      "learning_rate": 1.482269615137542e-07,
+      "loss": 1.4663,
+      "step": 3038
+    },
+    {
+      "epoch": 0.6307596513075965,
+      "grad_norm": 7.898637847027786,
+      "learning_rate": 1.4811934130716202e-07,
+      "loss": 1.5076,
+      "step": 3039
+    },
+    {
+      "epoch": 0.6309672063096721,
+      "grad_norm": 0.7015718235904259,
+      "learning_rate": 1.4801175166231752e-07,
+      "loss": 1.4908,
+      "step": 3040
+    },
+    {
+      "epoch": 0.6311747613117477,
+      "grad_norm": 0.6355451229397927,
+      "learning_rate": 1.4790419262496508e-07,
+      "loss": 1.5284,
+      "step": 3041
+    },
+    {
+      "epoch": 0.6313823163138231,
+      "grad_norm": 0.7914333897141325,
+      "learning_rate": 1.4779666424083593e-07,
+      "loss": 1.428,
+      "step": 3042
+    },
+    {
+      "epoch": 0.6315898713158987,
+      "grad_norm": 1.0819444153853983,
+      "learning_rate": 1.4768916655564815e-07,
+      "loss": 1.5306,
+      "step": 3043
+    },
+    {
+      "epoch": 0.6317974263179743,
+      "grad_norm": 1.0141038537035725,
+      "learning_rate": 1.475816996151071e-07,
+      "loss": 1.5528,
+      "step": 3044
+    },
+    {
+      "epoch": 0.6320049813200498,
+      "grad_norm": 2.5570138409444287,
+      "learning_rate": 1.4747426346490474e-07,
+      "loss": 1.5156,
+      "step": 3045
+    },
+    {
+      "epoch": 0.6322125363221254,
+      "grad_norm": 0.758829756857622,
+      "learning_rate": 1.473668581507201e-07,
+      "loss": 1.5179,
+      "step": 3046
+    },
+    {
+      "epoch": 0.632420091324201,
+      "grad_norm": 0.7781761515774269,
+      "learning_rate": 1.4725948371821905e-07,
+      "loss": 1.4828,
+      "step": 3047
+    },
+    {
+      "epoch": 0.6326276463262764,
+      "grad_norm": 0.8349775520107943,
+      "learning_rate": 1.4715214021305443e-07,
+      "loss": 1.5643,
+      "step": 3048
+    },
+    {
+      "epoch": 0.632835201328352,
+      "grad_norm": 0.807621338388221,
+      "learning_rate": 1.4704482768086574e-07,
+      "loss": 1.5033,
+      "step": 3049
+    },
+    {
+      "epoch": 0.6330427563304276,
+      "grad_norm": 0.7113634675541607,
+      "learning_rate": 1.4693754616727954e-07,
+      "loss": 1.4864,
+      "step": 3050
+    },
+    {
+      "epoch": 0.6332503113325031,
+      "grad_norm": 0.7703297269058391,
+      "learning_rate": 1.4683029571790898e-07,
+      "loss": 1.5122,
+      "step": 3051
+    },
+    {
+      "epoch": 0.6334578663345787,
+      "grad_norm": 1.126413866694997,
+      "learning_rate": 1.467230763783541e-07,
+      "loss": 1.4837,
+      "step": 3052
+    },
+    {
+      "epoch": 0.6336654213366543,
+      "grad_norm": 0.663847763203691,
+      "learning_rate": 1.4661588819420172e-07,
+      "loss": 1.5925,
+      "step": 3053
+    },
+    {
+      "epoch": 0.6338729763387297,
+      "grad_norm": 0.796988765377266,
+      "learning_rate": 1.465087312110256e-07,
+      "loss": 1.5467,
+      "step": 3054
+    },
+    {
+      "epoch": 0.6340805313408053,
+      "grad_norm": 0.7037528947052286,
+      "learning_rate": 1.4640160547438574e-07,
+      "loss": 1.5069,
+      "step": 3055
+    },
+    {
+      "epoch": 0.6342880863428808,
+      "grad_norm": 0.7113551970199539,
+      "learning_rate": 1.462945110298294e-07,
+      "loss": 1.5677,
+      "step": 3056
+    },
+    {
+      "epoch": 0.6344956413449564,
+      "grad_norm": 0.7389917451791563,
+      "learning_rate": 1.4618744792289014e-07,
+      "loss": 1.5606,
+      "step": 3057
+    },
+    {
+      "epoch": 0.634703196347032,
+      "grad_norm": 1.2961475558266493,
+      "learning_rate": 1.4608041619908837e-07,
+      "loss": 1.5365,
+      "step": 3058
+    },
+    {
+      "epoch": 0.6349107513491075,
+      "grad_norm": 0.6944269981154831,
+      "learning_rate": 1.4597341590393125e-07,
+      "loss": 1.529,
+      "step": 3059
+    },
+    {
+      "epoch": 0.635118306351183,
+      "grad_norm": 0.6791333898669759,
+      "learning_rate": 1.4586644708291235e-07,
+      "loss": 1.4323,
+      "step": 3060
+    },
+    {
+      "epoch": 0.6353258613532586,
+      "grad_norm": 0.8779002714428632,
+      "learning_rate": 1.4575950978151209e-07,
+      "loss": 1.5115,
+      "step": 3061
+    },
+    {
+      "epoch": 0.6355334163553341,
+      "grad_norm": 0.7422274739584223,
+      "learning_rate": 1.4565260404519726e-07,
+      "loss": 1.5069,
+      "step": 3062
+    },
+    {
+      "epoch": 0.6357409713574097,
+      "grad_norm": 1.0041976754773174,
+      "learning_rate": 1.4554572991942133e-07,
+      "loss": 1.546,
+      "step": 3063
+    },
+    {
+      "epoch": 0.6359485263594853,
+      "grad_norm": 0.8688180392046078,
+      "learning_rate": 1.454388874496244e-07,
+      "loss": 1.505,
+      "step": 3064
+    },
+    {
+      "epoch": 0.6361560813615608,
+      "grad_norm": 0.7305615683718213,
+      "learning_rate": 1.453320766812331e-07,
+      "loss": 1.5375,
+      "step": 3065
+    },
+    {
+      "epoch": 0.6363636363636364,
+      "grad_norm": 0.746878551233616,
+      "learning_rate": 1.4522529765966044e-07,
+      "loss": 1.5304,
+      "step": 3066
+    },
+    {
+      "epoch": 0.6365711913657119,
+      "grad_norm": 0.7294817896273393,
+      "learning_rate": 1.4511855043030608e-07,
+      "loss": 1.47,
+      "step": 3067
+    },
+    {
+      "epoch": 0.6367787463677874,
+      "grad_norm": 0.7024606832797239,
+      "learning_rate": 1.4501183503855608e-07,
+      "loss": 1.5076,
+      "step": 3068
+    },
+    {
+      "epoch": 0.636986301369863,
+      "grad_norm": 2.7014978284920983,
+      "learning_rate": 1.44905151529783e-07,
+      "loss": 1.4818,
+      "step": 3069
+    },
+    {
+      "epoch": 0.6371938563719386,
+      "grad_norm": 0.7405074455054956,
+      "learning_rate": 1.4479849994934577e-07,
+      "loss": 1.5395,
+      "step": 3070
+    },
+    {
+      "epoch": 0.6374014113740141,
+      "grad_norm": 2.35265288440181,
+      "learning_rate": 1.4469188034259002e-07,
+      "loss": 1.53,
+      "step": 3071
+    },
+    {
+      "epoch": 0.6376089663760897,
+      "grad_norm": 1.3202621993779977,
+      "learning_rate": 1.4458529275484725e-07,
+      "loss": 1.548,
+      "step": 3072
+    },
+    {
+      "epoch": 0.6378165213781652,
+      "grad_norm": 1.5101447094630214,
+      "learning_rate": 1.44478737231436e-07,
+      "loss": 1.5202,
+      "step": 3073
+    },
+    {
+      "epoch": 0.6380240763802407,
+      "grad_norm": 0.9474628256486164,
+      "learning_rate": 1.4437221381766062e-07,
+      "loss": 1.5159,
+      "step": 3074
+    },
+    {
+      "epoch": 0.6382316313823163,
+      "grad_norm": 1.4882513096778087,
+      "learning_rate": 1.442657225588122e-07,
+      "loss": 1.4976,
+      "step": 3075
+    },
+    {
+      "epoch": 0.6384391863843919,
+      "grad_norm": 1.3156103895085105,
+      "learning_rate": 1.4415926350016778e-07,
+      "loss": 1.4427,
+      "step": 3076
+    },
+    {
+      "epoch": 0.6386467413864674,
+      "grad_norm": 0.8089426620932525,
+      "learning_rate": 1.4405283668699127e-07,
+      "loss": 1.5182,
+      "step": 3077
+    },
+    {
+      "epoch": 0.638854296388543,
+      "grad_norm": 0.7426577416740147,
+      "learning_rate": 1.4394644216453213e-07,
+      "loss": 1.5289,
+      "step": 3078
+    },
+    {
+      "epoch": 0.6390618513906186,
+      "grad_norm": 0.8748941427895027,
+      "learning_rate": 1.4384007997802674e-07,
+      "loss": 1.4783,
+      "step": 3079
+    },
+    {
+      "epoch": 0.639269406392694,
+      "grad_norm": 0.6665043031229163,
+      "learning_rate": 1.4373375017269745e-07,
+      "loss": 1.5631,
+      "step": 3080
+    },
+    {
+      "epoch": 0.6394769613947696,
+      "grad_norm": 0.9841742310367197,
+      "learning_rate": 1.4362745279375274e-07,
+      "loss": 1.5146,
+      "step": 3081
+    },
+    {
+      "epoch": 0.6396845163968452,
+      "grad_norm": 0.6311251374540982,
+      "learning_rate": 1.435211878863876e-07,
+      "loss": 1.5031,
+      "step": 3082
+    },
+    {
+      "epoch": 0.6398920713989207,
+      "grad_norm": 0.8117753554171052,
+      "learning_rate": 1.4341495549578296e-07,
+      "loss": 1.5217,
+      "step": 3083
+    },
+    {
+      "epoch": 0.6400996264009963,
+      "grad_norm": 0.9764619582407291,
+      "learning_rate": 1.4330875566710606e-07,
+      "loss": 1.534,
+      "step": 3084
+    },
+    {
+      "epoch": 0.6403071814030719,
+      "grad_norm": 0.9553728325129893,
+      "learning_rate": 1.432025884455101e-07,
+      "loss": 1.5403,
+      "step": 3085
+    },
+    {
+      "epoch": 0.6405147364051473,
+      "grad_norm": 0.7000982075225894,
+      "learning_rate": 1.4309645387613488e-07,
+      "loss": 1.4807,
+      "step": 3086
+    },
+    {
+      "epoch": 0.6407222914072229,
+      "grad_norm": 1.6149960427753751,
+      "learning_rate": 1.429903520041056e-07,
+      "loss": 1.5269,
+      "step": 3087
+    },
+    {
+      "epoch": 0.6409298464092985,
+      "grad_norm": 0.7146758156091039,
+      "learning_rate": 1.4288428287453424e-07,
+      "loss": 1.473,
+      "step": 3088
+    },
+    {
+      "epoch": 0.641137401411374,
+      "grad_norm": 0.6970781604322636,
+      "learning_rate": 1.427782465325185e-07,
+      "loss": 1.5905,
+      "step": 3089
+    },
+    {
+      "epoch": 0.6413449564134496,
+      "grad_norm": 0.7296286315195898,
+      "learning_rate": 1.4267224302314221e-07,
+      "loss": 1.4691,
+      "step": 3090
+    },
+    {
+      "epoch": 0.6415525114155252,
+      "grad_norm": 1.2317732571851574,
+      "learning_rate": 1.4256627239147522e-07,
+      "loss": 1.4712,
+      "step": 3091
+    },
+    {
+      "epoch": 0.6417600664176006,
+      "grad_norm": 0.856851020931563,
+      "learning_rate": 1.4246033468257355e-07,
+      "loss": 1.432,
+      "step": 3092
+    },
+    {
+      "epoch": 0.6419676214196762,
+      "grad_norm": 0.8347545856184144,
+      "learning_rate": 1.4235442994147887e-07,
+      "loss": 1.4874,
+      "step": 3093
+    },
+    {
+      "epoch": 0.6421751764217518,
+      "grad_norm": 2.410892699957657,
+      "learning_rate": 1.422485582132193e-07,
+      "loss": 1.5744,
+      "step": 3094
+    },
+    {
+      "epoch": 0.6423827314238273,
+      "grad_norm": 0.7254458479768149,
+      "learning_rate": 1.4214271954280856e-07,
+      "loss": 1.5756,
+      "step": 3095
+    },
+    {
+      "epoch": 0.6425902864259029,
+      "grad_norm": 0.7842689685340327,
+      "learning_rate": 1.4203691397524646e-07,
+      "loss": 1.5175,
+      "step": 3096
+    },
+    {
+      "epoch": 0.6427978414279785,
+      "grad_norm": 0.8292723896025624,
+      "learning_rate": 1.419311415555188e-07,
+      "loss": 1.5536,
+      "step": 3097
+    },
+    {
+      "epoch": 0.6430053964300539,
+      "grad_norm": 1.1061180140130233,
+      "learning_rate": 1.4182540232859713e-07,
+      "loss": 1.4814,
+      "step": 3098
+    },
+    {
+      "epoch": 0.6432129514321295,
+      "grad_norm": 0.7542346971289389,
+      "learning_rate": 1.4171969633943896e-07,
+      "loss": 1.5254,
+      "step": 3099
+    },
+    {
+      "epoch": 0.6434205064342051,
+      "grad_norm": 0.6603891545657087,
+      "learning_rate": 1.416140236329876e-07,
+      "loss": 1.4529,
+      "step": 3100
+    },
+    {
+      "epoch": 0.6436280614362806,
+      "grad_norm": 0.6625168165180937,
+      "learning_rate": 1.4150838425417253e-07,
+      "loss": 1.439,
+      "step": 3101
+    },
+    {
+      "epoch": 0.6438356164383562,
+      "grad_norm": 0.8494791070221424,
+      "learning_rate": 1.414027782479085e-07,
+      "loss": 1.5402,
+      "step": 3102
+    },
+    {
+      "epoch": 0.6440431714404317,
+      "grad_norm": 0.8003417404459681,
+      "learning_rate": 1.4129720565909664e-07,
+      "loss": 1.6088,
+      "step": 3103
+    },
+    {
+      "epoch": 0.6442507264425072,
+      "grad_norm": 1.8149013968738987,
+      "learning_rate": 1.411916665326234e-07,
+      "loss": 1.5488,
+      "step": 3104
+    },
+    {
+      "epoch": 0.6444582814445828,
+      "grad_norm": 0.7907987128264112,
+      "learning_rate": 1.4108616091336134e-07,
+      "loss": 1.5051,
+      "step": 3105
+    },
+    {
+      "epoch": 0.6446658364466583,
+      "grad_norm": 0.8082073328103767,
+      "learning_rate": 1.409806888461686e-07,
+      "loss": 1.512,
+      "step": 3106
+    },
+    {
+      "epoch": 0.6448733914487339,
+      "grad_norm": 0.9518176879682078,
+      "learning_rate": 1.4087525037588908e-07,
+      "loss": 1.4735,
+      "step": 3107
+    },
+    {
+      "epoch": 0.6450809464508095,
+      "grad_norm": 0.80544846305097,
+      "learning_rate": 1.407698455473525e-07,
+      "loss": 1.5938,
+      "step": 3108
+    },
+    {
+      "epoch": 0.645288501452885,
+      "grad_norm": 0.8651528823161663,
+      "learning_rate": 1.4066447440537416e-07,
+      "loss": 1.5049,
+      "step": 3109
+    },
+    {
+      "epoch": 0.6454960564549606,
+      "grad_norm": 0.6726388140363898,
+      "learning_rate": 1.405591369947551e-07,
+      "loss": 1.4747,
+      "step": 3110
+    },
+    {
+      "epoch": 0.6457036114570361,
+      "grad_norm": 0.9257356011821511,
+      "learning_rate": 1.4045383336028184e-07,
+      "loss": 1.5203,
+      "step": 3111
+    },
+    {
+      "epoch": 0.6459111664591116,
+      "grad_norm": 0.7458646426687328,
+      "learning_rate": 1.4034856354672698e-07,
+      "loss": 1.5565,
+      "step": 3112
+    },
+    {
+      "epoch": 0.6461187214611872,
+      "grad_norm": 1.211695814577115,
+      "learning_rate": 1.4024332759884807e-07,
+      "loss": 1.4715,
+      "step": 3113
+    },
+    {
+      "epoch": 0.6463262764632628,
+      "grad_norm": 0.7086060899643317,
+      "learning_rate": 1.4013812556138896e-07,
+      "loss": 1.4939,
+      "step": 3114
+    },
+    {
+      "epoch": 0.6465338314653383,
+      "grad_norm": 1.5452079901734623,
+      "learning_rate": 1.4003295747907866e-07,
+      "loss": 1.4349,
+      "step": 3115
+    },
+    {
+      "epoch": 0.6467413864674139,
+      "grad_norm": 0.8409333296750025,
+      "learning_rate": 1.3992782339663186e-07,
+      "loss": 1.5196,
+      "step": 3116
+    },
+    {
+      "epoch": 0.6469489414694894,
+      "grad_norm": 0.7396965118502287,
+      "learning_rate": 1.3982272335874865e-07,
+      "loss": 1.4621,
+      "step": 3117
+    },
+    {
+      "epoch": 0.6471564964715649,
+      "grad_norm": 0.7334179538865436,
+      "learning_rate": 1.3971765741011496e-07,
+      "loss": 1.5172,
+      "step": 3118
+    },
+    {
+      "epoch": 0.6473640514736405,
+      "grad_norm": 0.9689462051889443,
+      "learning_rate": 1.3961262559540194e-07,
+      "loss": 1.5153,
+      "step": 3119
+    },
+    {
+      "epoch": 0.6475716064757161,
+      "grad_norm": 1.2226145756928242,
+      "learning_rate": 1.395076279592664e-07,
+      "loss": 1.4683,
+      "step": 3120
+    },
+    {
+      "epoch": 0.6477791614777916,
+      "grad_norm": 0.9055456898843436,
+      "learning_rate": 1.3940266454635046e-07,
+      "loss": 1.4859,
+      "step": 3121
+    },
+    {
+      "epoch": 0.6479867164798672,
+      "grad_norm": 0.7520262766254164,
+      "learning_rate": 1.3929773540128178e-07,
+      "loss": 1.4896,
+      "step": 3122
+    },
+    {
+      "epoch": 0.6481942714819428,
+      "grad_norm": 0.8497141338664903,
+      "learning_rate": 1.3919284056867354e-07,
+      "loss": 1.5648,
+      "step": 3123
+    },
+    {
+      "epoch": 0.6484018264840182,
+      "grad_norm": 0.9689861309046328,
+      "learning_rate": 1.3908798009312417e-07,
+      "loss": 1.5345,
+      "step": 3124
+    },
+    {
+      "epoch": 0.6486093814860938,
+      "grad_norm": 0.7166738617384836,
+      "learning_rate": 1.3898315401921753e-07,
+      "loss": 1.4839,
+      "step": 3125
+    },
+    {
+      "epoch": 0.6488169364881694,
+      "grad_norm": 0.7047849979525533,
+      "learning_rate": 1.3887836239152295e-07,
+      "loss": 1.544,
+      "step": 3126
+    },
+    {
+      "epoch": 0.6490244914902449,
+      "grad_norm": 0.9955286228651691,
+      "learning_rate": 1.3877360525459512e-07,
+      "loss": 1.5037,
+      "step": 3127
+    },
+    {
+      "epoch": 0.6492320464923205,
+      "grad_norm": 0.6827715321456042,
+      "learning_rate": 1.3866888265297373e-07,
+      "loss": 1.4684,
+      "step": 3128
+    },
+    {
+      "epoch": 0.6494396014943961,
+      "grad_norm": 1.1616458049172922,
+      "learning_rate": 1.3856419463118435e-07,
+      "loss": 1.5075,
+      "step": 3129
+    },
+    {
+      "epoch": 0.6496471564964715,
+      "grad_norm": 0.6312473488223415,
+      "learning_rate": 1.3845954123373735e-07,
+      "loss": 1.5471,
+      "step": 3130
+    },
+    {
+      "epoch": 0.6498547114985471,
+      "grad_norm": 1.354544474937245,
+      "learning_rate": 1.383549225051287e-07,
+      "loss": 1.4566,
+      "step": 3131
+    },
+    {
+      "epoch": 0.6500622665006227,
+      "grad_norm": 0.907029290232451,
+      "learning_rate": 1.3825033848983933e-07,
+      "loss": 1.548,
+      "step": 3132
+    },
+    {
+      "epoch": 0.6502698215026982,
+      "grad_norm": 0.8544662443232702,
+      "learning_rate": 1.381457892323358e-07,
+      "loss": 1.5051,
+      "step": 3133
+    },
+    {
+      "epoch": 0.6504773765047738,
+      "grad_norm": 0.8201666520277637,
+      "learning_rate": 1.3804127477706956e-07,
+      "loss": 1.5534,
+      "step": 3134
+    },
+    {
+      "epoch": 0.6506849315068494,
+      "grad_norm": 0.6775086341052935,
+      "learning_rate": 1.3793679516847744e-07,
+      "loss": 1.4579,
+      "step": 3135
+    },
+    {
+      "epoch": 0.6508924865089248,
+      "grad_norm": 2.9902079356614486,
+      "learning_rate": 1.3783235045098134e-07,
+      "loss": 1.4855,
+      "step": 3136
+    },
+    {
+      "epoch": 0.6511000415110004,
+      "grad_norm": 0.8620242069011522,
+      "learning_rate": 1.377279406689883e-07,
+      "loss": 1.4511,
+      "step": 3137
+    },
+    {
+      "epoch": 0.651307596513076,
+      "grad_norm": 0.683327038854693,
+      "learning_rate": 1.376235658668908e-07,
+      "loss": 1.5387,
+      "step": 3138
+    },
+    {
+      "epoch": 0.6515151515151515,
+      "grad_norm": 1.1432071442735263,
+      "learning_rate": 1.3751922608906612e-07,
+      "loss": 1.5425,
+      "step": 3139
+    },
+    {
+      "epoch": 0.6517227065172271,
+      "grad_norm": 0.7113947008645579,
+      "learning_rate": 1.3741492137987675e-07,
+      "loss": 1.491,
+      "step": 3140
+    },
+    {
+      "epoch": 0.6519302615193027,
+      "grad_norm": 0.6147124440706151,
+      "learning_rate": 1.3731065178367026e-07,
+      "loss": 1.5743,
+      "step": 3141
+    },
+    {
+      "epoch": 0.6521378165213781,
+      "grad_norm": 0.6824026715321913,
+      "learning_rate": 1.3720641734477946e-07,
+      "loss": 1.5709,
+      "step": 3142
+    },
+    {
+      "epoch": 0.6523453715234537,
+      "grad_norm": 0.6956716778191295,
+      "learning_rate": 1.3710221810752186e-07,
+      "loss": 1.6263,
+      "step": 3143
+    },
+    {
+      "epoch": 0.6525529265255293,
+      "grad_norm": 0.8923387330246622,
+      "learning_rate": 1.3699805411620035e-07,
+      "loss": 1.5206,
+      "step": 3144
+    },
+    {
+      "epoch": 0.6527604815276048,
+      "grad_norm": 0.9645212346817976,
+      "learning_rate": 1.3689392541510266e-07,
+      "loss": 1.5674,
+      "step": 3145
+    },
+    {
+      "epoch": 0.6529680365296804,
+      "grad_norm": 0.635361822058857,
+      "learning_rate": 1.3678983204850153e-07,
+      "loss": 1.5305,
+      "step": 3146
+    },
+    {
+      "epoch": 0.6531755915317559,
+      "grad_norm": 0.7179233590868765,
+      "learning_rate": 1.3668577406065472e-07,
+      "loss": 1.4897,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6533831465338314,
+      "grad_norm": 1.3084336324783978,
+      "learning_rate": 1.3658175149580485e-07,
+      "loss": 1.5444,
+      "step": 3148
+    },
+    {
+      "epoch": 0.653590701535907,
+      "grad_norm": 1.1654637217615773,
+      "learning_rate": 1.3647776439817968e-07,
+      "loss": 1.5048,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6537982565379825,
+      "grad_norm": 4.559807394204516,
+      "learning_rate": 1.363738128119917e-07,
+      "loss": 1.4855,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6540058115400581,
+      "grad_norm": 0.8977383915080426,
+      "learning_rate": 1.3626989678143835e-07,
+      "loss": 1.5687,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6542133665421337,
+      "grad_norm": 0.7481379331951814,
+      "learning_rate": 1.3616601635070196e-07,
+      "loss": 1.4991,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6544209215442092,
+      "grad_norm": 0.7777807813575783,
+      "learning_rate": 1.3606217156394983e-07,
+      "loss": 1.5649,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6546284765462848,
+      "grad_norm": 0.7495343957749625,
+      "learning_rate": 1.359583624653338e-07,
+      "loss": 1.5372,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6548360315483603,
+      "grad_norm": 0.7378338866584575,
+      "learning_rate": 1.3585458909899095e-07,
+      "loss": 1.4468,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6550435865504358,
+      "grad_norm": 0.7977122563890158,
+      "learning_rate": 1.357508515090429e-07,
+      "loss": 1.4674,
+      "step": 3156
+    },
+    {
+      "epoch": 0.6552511415525114,
+      "grad_norm": 2.380994998102106,
+      "learning_rate": 1.356471497395961e-07,
+      "loss": 1.4269,
+      "step": 3157
+    },
+    {
+      "epoch": 0.655458696554587,
+      "grad_norm": 0.7261063638137861,
+      "learning_rate": 1.3554348383474167e-07,
+      "loss": 1.4712,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6556662515566625,
+      "grad_norm": 0.926607805842262,
+      "learning_rate": 1.3543985383855584e-07,
+      "loss": 1.5323,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6558738065587381,
+      "grad_norm": 0.6499893228636157,
+      "learning_rate": 1.3533625979509916e-07,
+      "loss": 1.4999,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6560813615608136,
+      "grad_norm": 0.980387142846143,
+      "learning_rate": 1.3523270174841711e-07,
+      "loss": 1.5735,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6562889165628891,
+      "grad_norm": 0.7930593259630854,
+      "learning_rate": 1.3512917974253998e-07,
+      "loss": 1.4966,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6564964715649647,
+      "grad_norm": 0.6897423656993112,
+      "learning_rate": 1.3502569382148223e-07,
+      "loss": 1.487,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6567040265670403,
+      "grad_norm": 0.8164414402342477,
+      "learning_rate": 1.3492224402924363e-07,
+      "loss": 1.5089,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6569115815691158,
+      "grad_norm": 0.6630990873187498,
+      "learning_rate": 1.3481883040980817e-07,
+      "loss": 1.5283,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6571191365711914,
+      "grad_norm": 0.7270126661653342,
+      "learning_rate": 1.3471545300714464e-07,
+      "loss": 1.5902,
+      "step": 3166
+    },
+    {
+      "epoch": 0.657326691573267,
+      "grad_norm": 0.6601295245188631,
+      "learning_rate": 1.346121118652062e-07,
+      "loss": 1.5154,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6575342465753424,
+      "grad_norm": 1.046206148608879,
+      "learning_rate": 1.3450880702793098e-07,
+      "loss": 1.4581,
+      "step": 3168
+    },
+    {
+      "epoch": 0.657741801577418,
+      "grad_norm": 0.7879608412016764,
+      "learning_rate": 1.3440553853924128e-07,
+      "loss": 1.4875,
+      "step": 3169
+    },
+    {
+      "epoch": 0.6579493565794936,
+      "grad_norm": 0.7294064551758634,
+      "learning_rate": 1.3430230644304424e-07,
+      "loss": 1.6067,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6581569115815691,
+      "grad_norm": 0.6965681069968713,
+      "learning_rate": 1.3419911078323133e-07,
+      "loss": 1.5046,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6583644665836447,
+      "grad_norm": 0.7455234469985699,
+      "learning_rate": 1.3409595160367865e-07,
+      "loss": 1.5454,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6585720215857203,
+      "grad_norm": 0.810065925372056,
+      "learning_rate": 1.3399282894824667e-07,
+      "loss": 1.5366,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6587795765877957,
+      "grad_norm": 0.7022851758788176,
+      "learning_rate": 1.3388974286078048e-07,
+      "loss": 1.4455,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6589871315898713,
+      "grad_norm": 0.7307043044194036,
+      "learning_rate": 1.337866933851096e-07,
+      "loss": 1.511,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6591946865919469,
+      "grad_norm": 0.9138245044277977,
+      "learning_rate": 1.3368368056504774e-07,
+      "loss": 1.4929,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6594022415940224,
+      "grad_norm": 0.9479502615623114,
+      "learning_rate": 1.3358070444439348e-07,
+      "loss": 1.5206,
+      "step": 3177
+    },
+    {
+      "epoch": 0.659609796596098,
+      "grad_norm": 0.9285880312061874,
+      "learning_rate": 1.3347776506692925e-07,
+      "loss": 1.5595,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6598173515981736,
+      "grad_norm": 0.7780464977009807,
+      "learning_rate": 1.3337486247642235e-07,
+      "loss": 1.5104,
+      "step": 3179
+    },
+    {
+      "epoch": 0.660024906600249,
+      "grad_norm": 0.7112987218549393,
+      "learning_rate": 1.3327199671662417e-07,
+      "loss": 1.5283,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6602324616023246,
+      "grad_norm": 0.6875460376572219,
+      "learning_rate": 1.331691678312705e-07,
+      "loss": 1.5727,
+      "step": 3181
+    },
+    {
+      "epoch": 0.6604400166044002,
+      "grad_norm": 0.7420930274666655,
+      "learning_rate": 1.3306637586408133e-07,
+      "loss": 1.544,
+      "step": 3182
+    },
+    {
+      "epoch": 0.6606475716064757,
+      "grad_norm": 0.6554675119831339,
+      "learning_rate": 1.3296362085876136e-07,
+      "loss": 1.5081,
+      "step": 3183
+    },
+    {
+      "epoch": 0.6608551266085513,
+      "grad_norm": 9.337618882141618,
+      "learning_rate": 1.3286090285899896e-07,
+      "loss": 1.4557,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6610626816106269,
+      "grad_norm": 1.4940346493915295,
+      "learning_rate": 1.3275822190846733e-07,
+      "loss": 1.5046,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6612702366127023,
+      "grad_norm": 0.7160809580436963,
+      "learning_rate": 1.3265557805082362e-07,
+      "loss": 1.5568,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6614777916147779,
+      "grad_norm": 0.6435534078430243,
+      "learning_rate": 1.3255297132970915e-07,
+      "loss": 1.5377,
+      "step": 3187
+    },
+    {
+      "epoch": 0.6616853466168535,
+      "grad_norm": 0.7875195225061454,
+      "learning_rate": 1.3245040178874977e-07,
+      "loss": 1.4831,
+      "step": 3188
+    },
+    {
+      "epoch": 0.661892901618929,
+      "grad_norm": 0.664765418963022,
+      "learning_rate": 1.3234786947155528e-07,
+      "loss": 1.4932,
+      "step": 3189
+    },
+    {
+      "epoch": 0.6621004566210046,
+      "grad_norm": 0.8601205547924736,
+      "learning_rate": 1.322453744217196e-07,
+      "loss": 1.5301,
+      "step": 3190
+    },
+    {
+      "epoch": 0.6623080116230801,
+      "grad_norm": 0.7227998101842118,
+      "learning_rate": 1.32142916682821e-07,
+      "loss": 1.5086,
+      "step": 3191
+    },
+    {
+      "epoch": 0.6625155666251556,
+      "grad_norm": 0.6778281131795043,
+      "learning_rate": 1.3204049629842173e-07,
+      "loss": 1.5216,
+      "step": 3192
+    },
+    {
+      "epoch": 0.6627231216272312,
+      "grad_norm": 0.7300134584675743,
+      "learning_rate": 1.3193811331206818e-07,
+      "loss": 1.5261,
+      "step": 3193
+    },
+    {
+      "epoch": 0.6629306766293067,
+      "grad_norm": 0.7680107298221206,
+      "learning_rate": 1.318357677672911e-07,
+      "loss": 1.5422,
+      "step": 3194
+    },
+    {
+      "epoch": 0.6631382316313823,
+      "grad_norm": 8.887202603643503,
+      "learning_rate": 1.317334597076048e-07,
+      "loss": 1.5642,
+      "step": 3195
+    },
+    {
+      "epoch": 0.6633457866334579,
+      "grad_norm": 0.6408879200284127,
+      "learning_rate": 1.3163118917650813e-07,
+      "loss": 1.5604,
+      "step": 3196
+    },
+    {
+      "epoch": 0.6635533416355334,
+      "grad_norm": 1.4135421621674917,
+      "learning_rate": 1.3152895621748377e-07,
+      "loss": 1.4877,
+      "step": 3197
+    },
+    {
+      "epoch": 0.663760896637609,
+      "grad_norm": 0.6786000783019871,
+      "learning_rate": 1.3142676087399846e-07,
+      "loss": 1.5575,
+      "step": 3198
+    },
+    {
+      "epoch": 0.6639684516396845,
+      "grad_norm": 0.9303123804865102,
+      "learning_rate": 1.3132460318950288e-07,
+      "loss": 1.5206,
+      "step": 3199
+    },
+    {
+      "epoch": 0.66417600664176,
+      "grad_norm": 1.4928343968334328,
+      "learning_rate": 1.3122248320743187e-07,
+      "loss": 1.4816,
+      "step": 3200
+    },
+    {
+      "epoch": 0.6643835616438356,
+      "grad_norm": 0.824543269765013,
+      "learning_rate": 1.3112040097120408e-07,
+      "loss": 1.4694,
+      "step": 3201
+    },
+    {
+      "epoch": 0.6645911166459112,
+      "grad_norm": 0.6939174251708157,
+      "learning_rate": 1.3101835652422216e-07,
+      "loss": 1.5584,
+      "step": 3202
+    },
+    {
+      "epoch": 0.6647986716479867,
+      "grad_norm": 0.9524793869028257,
+      "learning_rate": 1.3091634990987284e-07,
+      "loss": 1.4583,
+      "step": 3203
+    },
+    {
+      "epoch": 0.6650062266500623,
+      "grad_norm": 0.6432380631008792,
+      "learning_rate": 1.3081438117152637e-07,
+      "loss": 1.3888,
+      "step": 3204
+    },
+    {
+      "epoch": 0.6652137816521378,
+      "grad_norm": 0.6818048942278007,
+      "learning_rate": 1.3071245035253734e-07,
+      "loss": 1.4895,
+      "step": 3205
+    },
+    {
+      "epoch": 0.6654213366542133,
+      "grad_norm": 0.7490778227804438,
+      "learning_rate": 1.3061055749624395e-07,
+      "loss": 1.4659,
+      "step": 3206
+    },
+    {
+      "epoch": 0.6656288916562889,
+      "grad_norm": 1.311766790008918,
+      "learning_rate": 1.305087026459684e-07,
+      "loss": 1.5891,
+      "step": 3207
+    },
+    {
+      "epoch": 0.6658364466583645,
+      "grad_norm": 1.3812133017818926,
+      "learning_rate": 1.3040688584501652e-07,
+      "loss": 1.4967,
+      "step": 3208
+    },
+    {
+      "epoch": 0.66604400166044,
+      "grad_norm": 0.9923122562684247,
+      "learning_rate": 1.3030510713667837e-07,
+      "loss": 1.4708,
+      "step": 3209
+    },
+    {
+      "epoch": 0.6662515566625156,
+      "grad_norm": 0.7990055602099573,
+      "learning_rate": 1.3020336656422725e-07,
+      "loss": 1.4704,
+      "step": 3210
+    },
+    {
+      "epoch": 0.6664591116645912,
+      "grad_norm": 0.9175499547845192,
+      "learning_rate": 1.3010166417092077e-07,
+      "loss": 1.4834,
+      "step": 3211
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.7487437399919964,
+      "learning_rate": 1.3000000000000005e-07,
+      "loss": 1.5701,
+      "step": 3212
+    },
+    {
+      "epoch": 0.6668742216687422,
+      "grad_norm": 0.6955458046657547,
+      "learning_rate": 1.2989837409468986e-07,
+      "loss": 1.5305,
+      "step": 3213
+    },
+    {
+      "epoch": 0.6670817766708178,
+      "grad_norm": 0.7623711556317108,
+      "learning_rate": 1.2979678649819902e-07,
+      "loss": 1.4678,
+      "step": 3214
+    },
+    {
+      "epoch": 0.6672893316728933,
+      "grad_norm": 1.5874400225238208,
+      "learning_rate": 1.2969523725371984e-07,
+      "loss": 1.5153,
+      "step": 3215
+    },
+    {
+      "epoch": 0.6674968866749689,
+      "grad_norm": 0.9161302428319901,
+      "learning_rate": 1.2959372640442827e-07,
+      "loss": 1.585,
+      "step": 3216
+    },
+    {
+      "epoch": 0.6677044416770445,
+      "grad_norm": 0.7816286979283945,
+      "learning_rate": 1.2949225399348404e-07,
+      "loss": 1.4338,
+      "step": 3217
+    },
+    {
+      "epoch": 0.6679119966791199,
+      "grad_norm": 4.376219216468093,
+      "learning_rate": 1.293908200640307e-07,
+      "loss": 1.4981,
+      "step": 3218
+    },
+    {
+      "epoch": 0.6681195516811955,
+      "grad_norm": 0.7901166312069904,
+      "learning_rate": 1.29289424659195e-07,
+      "loss": 1.5413,
+      "step": 3219
+    },
+    {
+      "epoch": 0.6683271066832711,
+      "grad_norm": 0.6159446063658047,
+      "learning_rate": 1.2918806782208778e-07,
+      "loss": 1.5397,
+      "step": 3220
+    },
+    {
+      "epoch": 0.6685346616853466,
+      "grad_norm": 0.7342524373657747,
+      "learning_rate": 1.290867495958032e-07,
+      "loss": 1.517,
+      "step": 3221
+    },
+    {
+      "epoch": 0.6687422166874222,
+      "grad_norm": 0.783300181565538,
+      "learning_rate": 1.2898547002341906e-07,
+      "loss": 1.5466,
+      "step": 3222
+    },
+    {
+      "epoch": 0.6689497716894978,
+      "grad_norm": 0.6550201276590841,
+      "learning_rate": 1.2888422914799676e-07,
+      "loss": 1.5131,
+      "step": 3223
+    },
+    {
+      "epoch": 0.6691573266915732,
+      "grad_norm": 0.9689901735166166,
+      "learning_rate": 1.2878302701258123e-07,
+      "loss": 1.5053,
+      "step": 3224
+    },
+    {
+      "epoch": 0.6693648816936488,
+      "grad_norm": 0.8456313707287444,
+      "learning_rate": 1.28681863660201e-07,
+      "loss": 1.4502,
+      "step": 3225
+    },
+    {
+      "epoch": 0.6695724366957244,
+      "grad_norm": 0.6863489930367463,
+      "learning_rate": 1.2858073913386793e-07,
+      "loss": 1.4779,
+      "step": 3226
+    },
+    {
+      "epoch": 0.6697799916977999,
+      "grad_norm": 0.7022832351060266,
+      "learning_rate": 1.2847965347657756e-07,
+      "loss": 1.4645,
+      "step": 3227
+    },
+    {
+      "epoch": 0.6699875466998755,
+      "grad_norm": 0.7552765311622475,
+      "learning_rate": 1.2837860673130878e-07,
+      "loss": 1.5234,
+      "step": 3228
+    },
+    {
+      "epoch": 0.6701951017019511,
+      "grad_norm": 0.8719040512638067,
+      "learning_rate": 1.2827759894102404e-07,
+      "loss": 1.4879,
+      "step": 3229
+    },
+    {
+      "epoch": 0.6704026567040265,
+      "grad_norm": 0.7289068680645504,
+      "learning_rate": 1.2817663014866914e-07,
+      "loss": 1.5704,
+      "step": 3230
+    },
+    {
+      "epoch": 0.6706102117061021,
+      "grad_norm": 1.6309012453266398,
+      "learning_rate": 1.280757003971733e-07,
+      "loss": 1.4617,
+      "step": 3231
+    },
+    {
+      "epoch": 0.6708177667081777,
+      "grad_norm": 0.6820231108481851,
+      "learning_rate": 1.2797480972944916e-07,
+      "loss": 1.571,
+      "step": 3232
+    },
+    {
+      "epoch": 0.6710253217102532,
+      "grad_norm": 0.7473218877225032,
+      "learning_rate": 1.2787395818839292e-07,
+      "loss": 1.4453,
+      "step": 3233
+    },
+    {
+      "epoch": 0.6712328767123288,
+      "grad_norm": 0.9916478133260138,
+      "learning_rate": 1.2777314581688375e-07,
+      "loss": 1.5689,
+      "step": 3234
+    },
+    {
+      "epoch": 0.6714404317144043,
+      "grad_norm": 1.0789071126265648,
+      "learning_rate": 1.2767237265778462e-07,
+      "loss": 1.5831,
+      "step": 3235
+    },
+    {
+      "epoch": 0.6716479867164799,
+      "grad_norm": 1.0820537657310563,
+      "learning_rate": 1.275716387539413e-07,
+      "loss": 1.5257,
+      "step": 3236
+    },
+    {
+      "epoch": 0.6718555417185554,
+      "grad_norm": 0.6828000204932493,
+      "learning_rate": 1.2747094414818344e-07,
+      "loss": 1.4904,
+      "step": 3237
+    },
+    {
+      "epoch": 0.6720630967206309,
+      "grad_norm": 1.0937862147446618,
+      "learning_rate": 1.2737028888332364e-07,
+      "loss": 1.5616,
+      "step": 3238
+    },
+    {
+      "epoch": 0.6722706517227065,
+      "grad_norm": 0.7931219142100963,
+      "learning_rate": 1.2726967300215774e-07,
+      "loss": 1.5742,
+      "step": 3239
+    },
+    {
+      "epoch": 0.6724782067247821,
+      "grad_norm": 0.979294804966168,
+      "learning_rate": 1.271690965474651e-07,
+      "loss": 1.5872,
+      "step": 3240
+    },
+    {
+      "epoch": 0.6726857617268576,
+      "grad_norm": 1.1234927210135381,
+      "learning_rate": 1.2706855956200808e-07,
+      "loss": 1.4989,
+      "step": 3241
+    },
+    {
+      "epoch": 0.6728933167289332,
+      "grad_norm": 0.9893755042487972,
+      "learning_rate": 1.2696806208853238e-07,
+      "loss": 1.5166,
+      "step": 3242
+    },
+    {
+      "epoch": 0.6731008717310087,
+      "grad_norm": 0.9406993242009124,
+      "learning_rate": 1.2686760416976673e-07,
+      "loss": 1.5746,
+      "step": 3243
+    },
+    {
+      "epoch": 0.6733084267330842,
+      "grad_norm": 0.8901696763855884,
+      "learning_rate": 1.2676718584842337e-07,
+      "loss": 1.5241,
+      "step": 3244
+    },
+    {
+      "epoch": 0.6735159817351598,
+      "grad_norm": 0.804322396485139,
+      "learning_rate": 1.2666680716719728e-07,
+      "loss": 1.4835,
+      "step": 3245
+    },
+    {
+      "epoch": 0.6737235367372354,
+      "grad_norm": 0.7513484901499353,
+      "learning_rate": 1.2656646816876703e-07,
+      "loss": 1.4895,
+      "step": 3246
+    },
+    {
+      "epoch": 0.6739310917393109,
+      "grad_norm": 0.6183502696902251,
+      "learning_rate": 1.2646616889579392e-07,
+      "loss": 1.5037,
+      "step": 3247
+    },
+    {
+      "epoch": 0.6741386467413865,
+      "grad_norm": 0.8730405953354398,
+      "learning_rate": 1.263659093909227e-07,
+      "loss": 1.5379,
+      "step": 3248
+    },
+    {
+      "epoch": 0.674346201743462,
+      "grad_norm": 0.7194889363693627,
+      "learning_rate": 1.2626568969678085e-07,
+      "loss": 1.6299,
+      "step": 3249
+    },
+    {
+      "epoch": 0.6745537567455375,
+      "grad_norm": 1.1877046457534175,
+      "learning_rate": 1.2616550985597932e-07,
+      "loss": 1.5008,
+      "step": 3250
+    },
+    {
+      "epoch": 0.6747613117476131,
+      "grad_norm": 1.5876821453458347,
+      "learning_rate": 1.2606536991111178e-07,
+      "loss": 1.5053,
+      "step": 3251
+    },
+    {
+      "epoch": 0.6749688667496887,
+      "grad_norm": 0.7438657559550432,
+      "learning_rate": 1.2596526990475522e-07,
+      "loss": 1.547,
+      "step": 3252
+    },
+    {
+      "epoch": 0.6751764217517642,
+      "grad_norm": 0.671664921082297,
+      "learning_rate": 1.2586520987946935e-07,
+      "loss": 1.5518,
+      "step": 3253
+    },
+    {
+      "epoch": 0.6753839767538398,
+      "grad_norm": 6.126359824906304,
+      "learning_rate": 1.2576518987779706e-07,
+      "loss": 1.4918,
+      "step": 3254
+    },
+    {
+      "epoch": 0.6755915317559154,
+      "grad_norm": 0.7836723018809235,
+      "learning_rate": 1.256652099422643e-07,
+      "loss": 1.5205,
+      "step": 3255
+    },
+    {
+      "epoch": 0.6757990867579908,
+      "grad_norm": 0.69736765216155,
+      "learning_rate": 1.2556527011537986e-07,
+      "loss": 1.4826,
+      "step": 3256
+    },
+    {
+      "epoch": 0.6760066417600664,
+      "grad_norm": 0.7649587894359001,
+      "learning_rate": 1.2546537043963544e-07,
+      "loss": 1.5124,
+      "step": 3257
+    },
+    {
+      "epoch": 0.676214196762142,
+      "grad_norm": 0.945064961123417,
+      "learning_rate": 1.2536551095750567e-07,
+      "loss": 1.5389,
+      "step": 3258
+    },
+    {
+      "epoch": 0.6764217517642175,
+      "grad_norm": 0.8416403467807277,
+      "learning_rate": 1.2526569171144838e-07,
+      "loss": 1.5379,
+      "step": 3259
+    },
+    {
+      "epoch": 0.6766293067662931,
+      "grad_norm": 0.7401161134509093,
+      "learning_rate": 1.251659127439038e-07,
+      "loss": 1.5646,
+      "step": 3260
+    },
+    {
+      "epoch": 0.6768368617683687,
+      "grad_norm": 0.6790074303755887,
+      "learning_rate": 1.2506617409729548e-07,
+      "loss": 1.4979,
+      "step": 3261
+    },
+    {
+      "epoch": 0.6770444167704441,
+      "grad_norm": 1.0133838218497426,
+      "learning_rate": 1.2496647581402964e-07,
+      "loss": 1.5388,
+      "step": 3262
+    },
+    {
+      "epoch": 0.6772519717725197,
+      "grad_norm": 0.7243290771177783,
+      "learning_rate": 1.2486681793649522e-07,
+      "loss": 1.5235,
+      "step": 3263
+    },
+    {
+      "epoch": 0.6774595267745953,
+      "grad_norm": 1.3742502898355515,
+      "learning_rate": 1.2476720050706413e-07,
+      "loss": 1.5499,
+      "step": 3264
+    },
+    {
+      "epoch": 0.6776670817766708,
+      "grad_norm": 0.7587012980669989,
+      "learning_rate": 1.2466762356809115e-07,
+      "loss": 1.5038,
+      "step": 3265
+    },
+    {
+      "epoch": 0.6778746367787464,
+      "grad_norm": 0.6989098005015468,
+      "learning_rate": 1.2456808716191371e-07,
+      "loss": 1.4801,
+      "step": 3266
+    },
+    {
+      "epoch": 0.678082191780822,
+      "grad_norm": 1.2411159625764818,
+      "learning_rate": 1.2446859133085194e-07,
+      "loss": 1.5044,
+      "step": 3267
+    },
+    {
+      "epoch": 0.6782897467828974,
+      "grad_norm": 0.9122984882729779,
+      "learning_rate": 1.243691361172091e-07,
+      "loss": 1.4728,
+      "step": 3268
+    },
+    {
+      "epoch": 0.678497301784973,
+      "grad_norm": 6.165429613297105,
+      "learning_rate": 1.242697215632706e-07,
+      "loss": 1.4999,
+      "step": 3269
+    },
+    {
+      "epoch": 0.6787048567870486,
+      "grad_norm": 0.8134231438380896,
+      "learning_rate": 1.2417034771130502e-07,
+      "loss": 1.4891,
+      "step": 3270
+    },
+    {
+      "epoch": 0.6789124117891241,
+      "grad_norm": 0.829237266367444,
+      "learning_rate": 1.2407101460356346e-07,
+      "loss": 1.5122,
+      "step": 3271
+    },
+    {
+      "epoch": 0.6791199667911997,
+      "grad_norm": 0.744338325002923,
+      "learning_rate": 1.2397172228227973e-07,
+      "loss": 1.4311,
+      "step": 3272
+    },
+    {
+      "epoch": 0.6793275217932753,
+      "grad_norm": 1.2027530577161256,
+      "learning_rate": 1.2387247078967023e-07,
+      "loss": 1.5459,
+      "step": 3273
+    },
+    {
+      "epoch": 0.6795350767953507,
+      "grad_norm": 0.6697714662395956,
+      "learning_rate": 1.237732601679342e-07,
+      "loss": 1.5082,
+      "step": 3274
+    },
+    {
+      "epoch": 0.6797426317974263,
+      "grad_norm": 0.6558015891459825,
+      "learning_rate": 1.236740904592532e-07,
+      "loss": 1.5443,
+      "step": 3275
+    },
+    {
+      "epoch": 0.6799501867995019,
+      "grad_norm": 0.8529060953687726,
+      "learning_rate": 1.2357496170579167e-07,
+      "loss": 1.4878,
+      "step": 3276
+    },
+    {
+      "epoch": 0.6801577418015774,
+      "grad_norm": 0.9428887057006291,
+      "learning_rate": 1.2347587394969647e-07,
+      "loss": 1.5471,
+      "step": 3277
+    },
+    {
+      "epoch": 0.680365296803653,
+      "grad_norm": 0.694225720923425,
+      "learning_rate": 1.2337682723309716e-07,
+      "loss": 1.4984,
+      "step": 3278
+    },
+    {
+      "epoch": 0.6805728518057285,
+      "grad_norm": 0.7773502221000645,
+      "learning_rate": 1.2327782159810562e-07,
+      "loss": 1.5351,
+      "step": 3279
+    },
+    {
+      "epoch": 0.680780406807804,
+      "grad_norm": 0.6685616537262326,
+      "learning_rate": 1.231788570868165e-07,
+      "loss": 1.6079,
+      "step": 3280
+    },
+    {
+      "epoch": 0.6809879618098796,
+      "grad_norm": 0.9936856335858999,
+      "learning_rate": 1.2307993374130694e-07,
+      "loss": 1.5254,
+      "step": 3281
+    },
+    {
+      "epoch": 0.6811955168119551,
+      "grad_norm": 0.8182567346760663,
+      "learning_rate": 1.2298105160363648e-07,
+      "loss": 1.5026,
+      "step": 3282
+    },
+    {
+      "epoch": 0.6814030718140307,
+      "grad_norm": 0.814675612880644,
+      "learning_rate": 1.2288221071584715e-07,
+      "loss": 1.5434,
+      "step": 3283
+    },
+    {
+      "epoch": 0.6816106268161063,
+      "grad_norm": 0.9567834636478196,
+      "learning_rate": 1.2278341111996345e-07,
+      "loss": 1.5131,
+      "step": 3284
+    },
+    {
+      "epoch": 0.6818181818181818,
+      "grad_norm": 0.7737183380493936,
+      "learning_rate": 1.226846528579925e-07,
+      "loss": 1.5125,
+      "step": 3285
+    },
+    {
+      "epoch": 0.6820257368202574,
+      "grad_norm": 1.0249993811725902,
+      "learning_rate": 1.2258593597192347e-07,
+      "loss": 1.4963,
+      "step": 3286
+    },
+    {
+      "epoch": 0.6822332918223329,
+      "grad_norm": 0.7429199143125298,
+      "learning_rate": 1.2248726050372832e-07,
+      "loss": 1.507,
+      "step": 3287
+    },
+    {
+      "epoch": 0.6824408468244084,
+      "grad_norm": 0.8296113467026645,
+      "learning_rate": 1.2238862649536116e-07,
+      "loss": 1.4945,
+      "step": 3288
+    },
+    {
+      "epoch": 0.682648401826484,
+      "grad_norm": 0.7774667067970688,
+      "learning_rate": 1.222900339887586e-07,
+      "loss": 1.4813,
+      "step": 3289
+    },
+    {
+      "epoch": 0.6828559568285596,
+      "grad_norm": 0.81425265512606,
+      "learning_rate": 1.221914830258395e-07,
+      "loss": 1.5438,
+      "step": 3290
+    },
+    {
+      "epoch": 0.6830635118306351,
+      "grad_norm": 1.328499975100338,
+      "learning_rate": 1.220929736485052e-07,
+      "loss": 1.4945,
+      "step": 3291
+    },
+    {
+      "epoch": 0.6832710668327107,
+      "grad_norm": 0.7367649751691349,
+      "learning_rate": 1.2199450589863923e-07,
+      "loss": 1.4778,
+      "step": 3292
+    },
+    {
+      "epoch": 0.6834786218347862,
+      "grad_norm": 0.6885923992190716,
+      "learning_rate": 1.2189607981810734e-07,
+      "loss": 1.5139,
+      "step": 3293
+    },
+    {
+      "epoch": 0.6836861768368617,
+      "grad_norm": 1.0397848070298938,
+      "learning_rate": 1.2179769544875797e-07,
+      "loss": 1.5217,
+      "step": 3294
+    },
+    {
+      "epoch": 0.6838937318389373,
+      "grad_norm": 0.7619277465693344,
+      "learning_rate": 1.2169935283242122e-07,
+      "loss": 1.4922,
+      "step": 3295
+    },
+    {
+      "epoch": 0.6841012868410129,
+      "grad_norm": 0.6614874402393118,
+      "learning_rate": 1.2160105201091e-07,
+      "loss": 1.5332,
+      "step": 3296
+    },
+    {
+      "epoch": 0.6843088418430884,
+      "grad_norm": 0.9768741763711837,
+      "learning_rate": 1.2150279302601915e-07,
+      "loss": 1.5489,
+      "step": 3297
+    },
+    {
+      "epoch": 0.684516396845164,
+      "grad_norm": 1.4033432752226545,
+      "learning_rate": 1.214045759195257e-07,
+      "loss": 1.5793,
+      "step": 3298
+    },
+    {
+      "epoch": 0.6847239518472396,
+      "grad_norm": 0.7967113361811957,
+      "learning_rate": 1.2130640073318895e-07,
+      "loss": 1.4329,
+      "step": 3299
+    },
+    {
+      "epoch": 0.684931506849315,
+      "grad_norm": 0.6020155266308058,
+      "learning_rate": 1.212082675087506e-07,
+      "loss": 1.4975,
+      "step": 3300
+    },
+    {
+      "epoch": 0.6851390618513906,
+      "grad_norm": 1.7521579011489161,
+      "learning_rate": 1.2111017628793398e-07,
+      "loss": 1.495,
+      "step": 3301
+    },
+    {
+      "epoch": 0.6853466168534662,
+      "grad_norm": 0.9102278581022684,
+      "learning_rate": 1.210121271124451e-07,
+      "loss": 1.5043,
+      "step": 3302
+    },
+    {
+      "epoch": 0.6855541718555417,
+      "grad_norm": 0.8018568702051945,
+      "learning_rate": 1.2091412002397178e-07,
+      "loss": 1.6131,
+      "step": 3303
+    },
+    {
+      "epoch": 0.6857617268576173,
+      "grad_norm": 0.68363912999568,
+      "learning_rate": 1.2081615506418407e-07,
+      "loss": 1.432,
+      "step": 3304
+    },
+    {
+      "epoch": 0.6859692818596929,
+      "grad_norm": 0.7161736225818417,
+      "learning_rate": 1.2071823227473398e-07,
+      "loss": 1.4266,
+      "step": 3305
+    },
+    {
+      "epoch": 0.6861768368617683,
+      "grad_norm": 0.8980527616086746,
+      "learning_rate": 1.206203516972558e-07,
+      "loss": 1.5259,
+      "step": 3306
+    },
+    {
+      "epoch": 0.6863843918638439,
+      "grad_norm": 0.8532044066108526,
+      "learning_rate": 1.205225133733657e-07,
+      "loss": 1.5253,
+      "step": 3307
+    },
+    {
+      "epoch": 0.6865919468659195,
+      "grad_norm": 1.1209907349542791,
+      "learning_rate": 1.2042471734466186e-07,
+      "loss": 1.5127,
+      "step": 3308
+    },
+    {
+      "epoch": 0.686799501867995,
+      "grad_norm": 0.8125529387242744,
+      "learning_rate": 1.2032696365272477e-07,
+      "loss": 1.5256,
+      "step": 3309
+    },
+    {
+      "epoch": 0.6870070568700706,
+      "grad_norm": 0.7587835731626091,
+      "learning_rate": 1.2022925233911644e-07,
+      "loss": 1.4907,
+      "step": 3310
+    },
+    {
+      "epoch": 0.6872146118721462,
+      "grad_norm": 0.8429248957642861,
+      "learning_rate": 1.201315834453813e-07,
+      "loss": 1.5301,
+      "step": 3311
+    },
+    {
+      "epoch": 0.6874221668742216,
+      "grad_norm": 0.8750320770033471,
+      "learning_rate": 1.2003395701304553e-07,
+      "loss": 1.5367,
+      "step": 3312
+    },
+    {
+      "epoch": 0.6876297218762972,
+      "grad_norm": 0.7535733375252992,
+      "learning_rate": 1.1993637308361732e-07,
+      "loss": 1.5162,
+      "step": 3313
+    },
+    {
+      "epoch": 0.6878372768783728,
+      "grad_norm": 0.7918701842023438,
+      "learning_rate": 1.1983883169858665e-07,
+      "loss": 1.4967,
+      "step": 3314
+    },
+    {
+      "epoch": 0.6880448318804483,
+      "grad_norm": 1.5252905589841776,
+      "learning_rate": 1.1974133289942575e-07,
+      "loss": 1.5304,
+      "step": 3315
+    },
+    {
+      "epoch": 0.6882523868825239,
+      "grad_norm": 12.902680839597537,
+      "learning_rate": 1.1964387672758823e-07,
+      "loss": 1.5442,
+      "step": 3316
+    },
+    {
+      "epoch": 0.6884599418845995,
+      "grad_norm": 0.6834085475986064,
+      "learning_rate": 1.195464632245101e-07,
+      "loss": 1.4972,
+      "step": 3317
+    },
+    {
+      "epoch": 0.688667496886675,
+      "grad_norm": 2.3087423028078775,
+      "learning_rate": 1.1944909243160894e-07,
+      "loss": 1.4164,
+      "step": 3318
+    },
+    {
+      "epoch": 0.6888750518887505,
+      "grad_norm": 0.925745961665308,
+      "learning_rate": 1.1935176439028413e-07,
+      "loss": 1.5181,
+      "step": 3319
+    },
+    {
+      "epoch": 0.6890826068908261,
+      "grad_norm": 0.8419661533384216,
+      "learning_rate": 1.192544791419171e-07,
+      "loss": 1.5236,
+      "step": 3320
+    },
+    {
+      "epoch": 0.6892901618929016,
+      "grad_norm": 0.768152891059504,
+      "learning_rate": 1.191572367278709e-07,
+      "loss": 1.5419,
+      "step": 3321
+    },
+    {
+      "epoch": 0.6894977168949772,
+      "grad_norm": 0.6787632415645909,
+      "learning_rate": 1.190600371894904e-07,
+      "loss": 1.4516,
+      "step": 3322
+    },
+    {
+      "epoch": 0.6897052718970528,
+      "grad_norm": 0.8858615369776436,
+      "learning_rate": 1.1896288056810224e-07,
+      "loss": 1.5259,
+      "step": 3323
+    },
+    {
+      "epoch": 0.6899128268991283,
+      "grad_norm": 0.8590260454942608,
+      "learning_rate": 1.18865766905015e-07,
+      "loss": 1.5474,
+      "step": 3324
+    },
+    {
+      "epoch": 0.6901203819012038,
+      "grad_norm": 0.759344397446281,
+      "learning_rate": 1.187686962415186e-07,
+      "loss": 1.5552,
+      "step": 3325
+    },
+    {
+      "epoch": 0.6903279369032793,
+      "grad_norm": 0.7050968244115102,
+      "learning_rate": 1.1867166861888512e-07,
+      "loss": 1.5404,
+      "step": 3326
+    },
+    {
+      "epoch": 0.6905354919053549,
+      "grad_norm": 0.783512390219713,
+      "learning_rate": 1.185746840783679e-07,
+      "loss": 1.562,
+      "step": 3327
+    },
+    {
+      "epoch": 0.6907430469074305,
+      "grad_norm": 1.2303305702469816,
+      "learning_rate": 1.1847774266120243e-07,
+      "loss": 1.5431,
+      "step": 3328
+    },
+    {
+      "epoch": 0.690950601909506,
+      "grad_norm": 0.7077933651680663,
+      "learning_rate": 1.1838084440860547e-07,
+      "loss": 1.5424,
+      "step": 3329
+    },
+    {
+      "epoch": 0.6911581569115816,
+      "grad_norm": 0.6707646089645365,
+      "learning_rate": 1.1828398936177557e-07,
+      "loss": 1.553,
+      "step": 3330
+    },
+    {
+      "epoch": 0.6913657119136571,
+      "grad_norm": 0.6945010132617898,
+      "learning_rate": 1.1818717756189309e-07,
+      "loss": 1.5995,
+      "step": 3331
+    },
+    {
+      "epoch": 0.6915732669157326,
+      "grad_norm": 0.7113711241103959,
+      "learning_rate": 1.1809040905011972e-07,
+      "loss": 1.5542,
+      "step": 3332
+    },
+    {
+      "epoch": 0.6917808219178082,
+      "grad_norm": 0.741018194954576,
+      "learning_rate": 1.1799368386759885e-07,
+      "loss": 1.5128,
+      "step": 3333
+    },
+    {
+      "epoch": 0.6919883769198838,
+      "grad_norm": 0.7578699710656742,
+      "learning_rate": 1.1789700205545543e-07,
+      "loss": 1.5334,
+      "step": 3334
+    },
+    {
+      "epoch": 0.6921959319219593,
+      "grad_norm": 0.6931746645977676,
+      "learning_rate": 1.1780036365479622e-07,
+      "loss": 1.4796,
+      "step": 3335
+    },
+    {
+      "epoch": 0.6924034869240349,
+      "grad_norm": 0.8950187295375504,
+      "learning_rate": 1.1770376870670897e-07,
+      "loss": 1.4896,
+      "step": 3336
+    },
+    {
+      "epoch": 0.6926110419261104,
+      "grad_norm": 0.9078363976774274,
+      "learning_rate": 1.1760721725226352e-07,
+      "loss": 1.5204,
+      "step": 3337
+    },
+    {
+      "epoch": 0.6928185969281859,
+      "grad_norm": 0.9472658491168165,
+      "learning_rate": 1.1751070933251095e-07,
+      "loss": 1.4614,
+      "step": 3338
+    },
+    {
+      "epoch": 0.6930261519302615,
+      "grad_norm": 0.9016770323418097,
+      "learning_rate": 1.174142449884838e-07,
+      "loss": 1.5571,
+      "step": 3339
+    },
+    {
+      "epoch": 0.6932337069323371,
+      "grad_norm": 0.7569346722844374,
+      "learning_rate": 1.1731782426119614e-07,
+      "loss": 1.5435,
+      "step": 3340
+    },
+    {
+      "epoch": 0.6934412619344126,
+      "grad_norm": 0.8050539807043565,
+      "learning_rate": 1.1722144719164369e-07,
+      "loss": 1.5405,
+      "step": 3341
+    },
+    {
+      "epoch": 0.6936488169364882,
+      "grad_norm": 0.7095210306297901,
+      "learning_rate": 1.1712511382080313e-07,
+      "loss": 1.4438,
+      "step": 3342
+    },
+    {
+      "epoch": 0.6938563719385638,
+      "grad_norm": 0.7265093608447354,
+      "learning_rate": 1.1702882418963304e-07,
+      "loss": 1.5281,
+      "step": 3343
+    },
+    {
+      "epoch": 0.6940639269406392,
+      "grad_norm": 0.9488333394268295,
+      "learning_rate": 1.169325783390732e-07,
+      "loss": 1.5596,
+      "step": 3344
+    },
+    {
+      "epoch": 0.6942714819427148,
+      "grad_norm": 0.7796546526634055,
+      "learning_rate": 1.168363763100447e-07,
+      "loss": 1.4799,
+      "step": 3345
+    },
+    {
+      "epoch": 0.6944790369447904,
+      "grad_norm": 0.6962087211105117,
+      "learning_rate": 1.1674021814345025e-07,
+      "loss": 1.4983,
+      "step": 3346
+    },
+    {
+      "epoch": 0.6946865919468659,
+      "grad_norm": 0.7024505788531472,
+      "learning_rate": 1.1664410388017363e-07,
+      "loss": 1.4934,
+      "step": 3347
+    },
+    {
+      "epoch": 0.6948941469489415,
+      "grad_norm": 0.8371346310222159,
+      "learning_rate": 1.165480335610801e-07,
+      "loss": 1.537,
+      "step": 3348
+    },
+    {
+      "epoch": 0.6951017019510171,
+      "grad_norm": 5.766918684986479,
+      "learning_rate": 1.1645200722701611e-07,
+      "loss": 1.4984,
+      "step": 3349
+    },
+    {
+      "epoch": 0.6953092569530925,
+      "grad_norm": 0.6324679222504724,
+      "learning_rate": 1.1635602491880979e-07,
+      "loss": 1.5612,
+      "step": 3350
+    },
+    {
+      "epoch": 0.6955168119551681,
+      "grad_norm": 0.6374458169376153,
+      "learning_rate": 1.1626008667726992e-07,
+      "loss": 1.4523,
+      "step": 3351
+    },
+    {
+      "epoch": 0.6957243669572437,
+      "grad_norm": 0.6733894863582579,
+      "learning_rate": 1.1616419254318712e-07,
+      "loss": 1.51,
+      "step": 3352
+    },
+    {
+      "epoch": 0.6959319219593192,
+      "grad_norm": 0.9030733028122518,
+      "learning_rate": 1.1606834255733298e-07,
+      "loss": 1.4622,
+      "step": 3353
+    },
+    {
+      "epoch": 0.6961394769613948,
+      "grad_norm": 0.7261027838379019,
+      "learning_rate": 1.1597253676046034e-07,
+      "loss": 1.5268,
+      "step": 3354
+    },
+    {
+      "epoch": 0.6963470319634704,
+      "grad_norm": 0.820286833260227,
+      "learning_rate": 1.1587677519330323e-07,
+      "loss": 1.5128,
+      "step": 3355
+    },
+    {
+      "epoch": 0.6965545869655458,
+      "grad_norm": 0.6825507510364497,
+      "learning_rate": 1.1578105789657709e-07,
+      "loss": 1.5899,
+      "step": 3356
+    },
+    {
+      "epoch": 0.6967621419676214,
+      "grad_norm": 0.6792810922204955,
+      "learning_rate": 1.1568538491097824e-07,
+      "loss": 1.4961,
+      "step": 3357
+    },
+    {
+      "epoch": 0.696969696969697,
+      "grad_norm": 1.0137170584353286,
+      "learning_rate": 1.1558975627718435e-07,
+      "loss": 1.5192,
+      "step": 3358
+    },
+    {
+      "epoch": 0.6971772519717725,
+      "grad_norm": 0.7233152589245139,
+      "learning_rate": 1.1549417203585417e-07,
+      "loss": 1.5873,
+      "step": 3359
+    },
+    {
+      "epoch": 0.6973848069738481,
+      "grad_norm": 0.7756375127051249,
+      "learning_rate": 1.1539863222762747e-07,
+      "loss": 1.5078,
+      "step": 3360
+    },
+    {
+      "epoch": 0.6975923619759237,
+      "grad_norm": 1.080794483931704,
+      "learning_rate": 1.1530313689312542e-07,
+      "loss": 1.4975,
+      "step": 3361
+    },
+    {
+      "epoch": 0.6977999169779991,
+      "grad_norm": 0.8838499449263395,
+      "learning_rate": 1.1520768607295002e-07,
+      "loss": 1.6161,
+      "step": 3362
+    },
+    {
+      "epoch": 0.6980074719800747,
+      "grad_norm": 0.792520258563017,
+      "learning_rate": 1.1511227980768441e-07,
+      "loss": 1.5134,
+      "step": 3363
+    },
+    {
+      "epoch": 0.6982150269821503,
+      "grad_norm": 0.7875389931040292,
+      "learning_rate": 1.1501691813789278e-07,
+      "loss": 1.4809,
+      "step": 3364
+    },
+    {
+      "epoch": 0.6984225819842258,
+      "grad_norm": 0.7777011386628931,
+      "learning_rate": 1.1492160110412053e-07,
+      "loss": 1.5742,
+      "step": 3365
+    },
+    {
+      "epoch": 0.6986301369863014,
+      "grad_norm": 1.0701278615616723,
+      "learning_rate": 1.1482632874689367e-07,
+      "loss": 1.5065,
+      "step": 3366
+    },
+    {
+      "epoch": 0.698837691988377,
+      "grad_norm": 1.1254047769170288,
+      "learning_rate": 1.1473110110671971e-07,
+      "loss": 1.5128,
+      "step": 3367
+    },
+    {
+      "epoch": 0.6990452469904525,
+      "grad_norm": 0.6609901295752498,
+      "learning_rate": 1.1463591822408683e-07,
+      "loss": 1.5167,
+      "step": 3368
+    },
+    {
+      "epoch": 0.699252801992528,
+      "grad_norm": 0.835998566213407,
+      "learning_rate": 1.1454078013946424e-07,
+      "loss": 1.5469,
+      "step": 3369
+    },
+    {
+      "epoch": 0.6994603569946035,
+      "grad_norm": 1.156215741766386,
+      "learning_rate": 1.1444568689330218e-07,
+      "loss": 1.5647,
+      "step": 3370
+    },
+    {
+      "epoch": 0.6996679119966791,
+      "grad_norm": 0.7445806399656314,
+      "learning_rate": 1.1435063852603168e-07,
+      "loss": 1.513,
+      "step": 3371
+    },
+    {
+      "epoch": 0.6998754669987547,
+      "grad_norm": 1.1131369662749901,
+      "learning_rate": 1.1425563507806489e-07,
+      "loss": 1.5949,
+      "step": 3372
+    },
+    {
+      "epoch": 0.7000830220008302,
+      "grad_norm": 0.8551760962085827,
+      "learning_rate": 1.1416067658979473e-07,
+      "loss": 1.4022,
+      "step": 3373
+    },
+    {
+      "epoch": 0.7002905770029058,
+      "grad_norm": 0.781042201990767,
+      "learning_rate": 1.14065763101595e-07,
+      "loss": 1.518,
+      "step": 3374
+    },
+    {
+      "epoch": 0.7004981320049813,
+      "grad_norm": 0.7313206751075126,
+      "learning_rate": 1.1397089465382035e-07,
+      "loss": 1.4871,
+      "step": 3375
+    },
+    {
+      "epoch": 0.7007056870070568,
+      "grad_norm": 0.8077026460850166,
+      "learning_rate": 1.138760712868065e-07,
+      "loss": 1.4924,
+      "step": 3376
+    },
+    {
+      "epoch": 0.7009132420091324,
+      "grad_norm": 1.0750350606176107,
+      "learning_rate": 1.1378129304086959e-07,
+      "loss": 1.5342,
+      "step": 3377
+    },
+    {
+      "epoch": 0.701120797011208,
+      "grad_norm": 5.540574661631686,
+      "learning_rate": 1.1368655995630698e-07,
+      "loss": 1.5273,
+      "step": 3378
+    },
+    {
+      "epoch": 0.7013283520132835,
+      "grad_norm": 0.894835760034304,
+      "learning_rate": 1.1359187207339665e-07,
+      "loss": 1.4689,
+      "step": 3379
+    },
+    {
+      "epoch": 0.7015359070153591,
+      "grad_norm": 0.7055104736417149,
+      "learning_rate": 1.1349722943239731e-07,
+      "loss": 1.566,
+      "step": 3380
+    },
+    {
+      "epoch": 0.7017434620174346,
+      "grad_norm": 0.910771424105918,
+      "learning_rate": 1.1340263207354845e-07,
+      "loss": 1.5529,
+      "step": 3381
+    },
+    {
+      "epoch": 0.7019510170195101,
+      "grad_norm": 0.8460787939571708,
+      "learning_rate": 1.1330808003707053e-07,
+      "loss": 1.4856,
+      "step": 3382
+    },
+    {
+      "epoch": 0.7021585720215857,
+      "grad_norm": 0.6502972224687663,
+      "learning_rate": 1.1321357336316443e-07,
+      "loss": 1.5609,
+      "step": 3383
+    },
+    {
+      "epoch": 0.7023661270236613,
+      "grad_norm": 0.797498816732242,
+      "learning_rate": 1.1311911209201195e-07,
+      "loss": 1.5528,
+      "step": 3384
+    },
+    {
+      "epoch": 0.7025736820257368,
+      "grad_norm": 0.6798084192841621,
+      "learning_rate": 1.1302469626377549e-07,
+      "loss": 1.5647,
+      "step": 3385
+    },
+    {
+      "epoch": 0.7027812370278124,
+      "grad_norm": 0.9007867850587731,
+      "learning_rate": 1.1293032591859808e-07,
+      "loss": 1.5564,
+      "step": 3386
+    },
+    {
+      "epoch": 0.702988792029888,
+      "grad_norm": 0.8667662349900563,
+      "learning_rate": 1.1283600109660359e-07,
+      "loss": 1.495,
+      "step": 3387
+    },
+    {
+      "epoch": 0.7031963470319634,
+      "grad_norm": 1.8222121137183087,
+      "learning_rate": 1.1274172183789641e-07,
+      "loss": 1.5639,
+      "step": 3388
+    },
+    {
+      "epoch": 0.703403902034039,
+      "grad_norm": 0.8155154265528134,
+      "learning_rate": 1.1264748818256155e-07,
+      "loss": 1.5187,
+      "step": 3389
+    },
+    {
+      "epoch": 0.7036114570361146,
+      "grad_norm": 1.0159828103052049,
+      "learning_rate": 1.1255330017066458e-07,
+      "loss": 1.4738,
+      "step": 3390
+    },
+    {
+      "epoch": 0.7038190120381901,
+      "grad_norm": 0.9553159692328533,
+      "learning_rate": 1.1245915784225202e-07,
+      "loss": 1.5887,
+      "step": 3391
+    },
+    {
+      "epoch": 0.7040265670402657,
+      "grad_norm": 0.8003320912955327,
+      "learning_rate": 1.123650612373503e-07,
+      "loss": 1.5768,
+      "step": 3392
+    },
+    {
+      "epoch": 0.7042341220423413,
+      "grad_norm": 1.4264326332110921,
+      "learning_rate": 1.1227101039596712e-07,
+      "loss": 1.4663,
+      "step": 3393
+    },
+    {
+      "epoch": 0.7044416770444167,
+      "grad_norm": 0.7296481548596724,
+      "learning_rate": 1.1217700535809025e-07,
+      "loss": 1.5041,
+      "step": 3394
+    },
+    {
+      "epoch": 0.7046492320464923,
+      "grad_norm": 0.7214947984550868,
+      "learning_rate": 1.1208304616368816e-07,
+      "loss": 1.5116,
+      "step": 3395
+    },
+    {
+      "epoch": 0.7048567870485679,
+      "grad_norm": 3.6798468607206267,
+      "learning_rate": 1.1198913285270975e-07,
+      "loss": 1.5258,
+      "step": 3396
+    },
+    {
+      "epoch": 0.7050643420506434,
+      "grad_norm": 0.6495861512302876,
+      "learning_rate": 1.1189526546508458e-07,
+      "loss": 1.4752,
+      "step": 3397
+    },
+    {
+      "epoch": 0.705271897052719,
+      "grad_norm": 0.8641430139462923,
+      "learning_rate": 1.1180144404072253e-07,
+      "loss": 1.5091,
+      "step": 3398
+    },
+    {
+      "epoch": 0.7054794520547946,
+      "grad_norm": 1.2306453664608021,
+      "learning_rate": 1.1170766861951389e-07,
+      "loss": 1.4493,
+      "step": 3399
+    },
+    {
+      "epoch": 0.70568700705687,
+      "grad_norm": 0.7239116576551664,
+      "learning_rate": 1.1161393924132972e-07,
+      "loss": 1.4877,
+      "step": 3400
+    },
+    {
+      "epoch": 0.7058945620589456,
+      "grad_norm": 1.2269671774796294,
+      "learning_rate": 1.1152025594602093e-07,
+      "loss": 1.5184,
+      "step": 3401
+    },
+    {
+      "epoch": 0.7061021170610212,
+      "grad_norm": 0.6758102647485595,
+      "learning_rate": 1.1142661877341942e-07,
+      "loss": 1.5247,
+      "step": 3402
+    },
+    {
+      "epoch": 0.7063096720630967,
+      "grad_norm": 0.8147318006256409,
+      "learning_rate": 1.1133302776333713e-07,
+      "loss": 1.5336,
+      "step": 3403
+    },
+    {
+      "epoch": 0.7065172270651723,
+      "grad_norm": 1.178490597600748,
+      "learning_rate": 1.112394829555665e-07,
+      "loss": 1.4254,
+      "step": 3404
+    },
+    {
+      "epoch": 0.7067247820672479,
+      "grad_norm": 0.8676264160748454,
+      "learning_rate": 1.1114598438988023e-07,
+      "loss": 1.4749,
+      "step": 3405
+    },
+    {
+      "epoch": 0.7069323370693233,
+      "grad_norm": 0.8391086821680002,
+      "learning_rate": 1.1105253210603163e-07,
+      "loss": 1.5734,
+      "step": 3406
+    },
+    {
+      "epoch": 0.7071398920713989,
+      "grad_norm": 0.7113711131535023,
+      "learning_rate": 1.1095912614375384e-07,
+      "loss": 1.5361,
+      "step": 3407
+    },
+    {
+      "epoch": 0.7073474470734745,
+      "grad_norm": 0.6918596428886382,
+      "learning_rate": 1.1086576654276081e-07,
+      "loss": 1.5083,
+      "step": 3408
+    },
+    {
+      "epoch": 0.70755500207555,
+      "grad_norm": 0.7689142284314142,
+      "learning_rate": 1.1077245334274649e-07,
+      "loss": 1.557,
+      "step": 3409
+    },
+    {
+      "epoch": 0.7077625570776256,
+      "grad_norm": 0.7011428368836177,
+      "learning_rate": 1.106791865833852e-07,
+      "loss": 1.5308,
+      "step": 3410
+    },
+    {
+      "epoch": 0.7079701120797012,
+      "grad_norm": 1.5249158426303928,
+      "learning_rate": 1.1058596630433141e-07,
+      "loss": 1.4557,
+      "step": 3411
+    },
+    {
+      "epoch": 0.7081776670817767,
+      "grad_norm": 0.6853799161808818,
+      "learning_rate": 1.1049279254522002e-07,
+      "loss": 1.4946,
+      "step": 3412
+    },
+    {
+      "epoch": 0.7083852220838522,
+      "grad_norm": 0.6944368727896674,
+      "learning_rate": 1.1039966534566598e-07,
+      "loss": 1.4359,
+      "step": 3413
+    },
+    {
+      "epoch": 0.7085927770859277,
+      "grad_norm": 0.8965955726509883,
+      "learning_rate": 1.1030658474526453e-07,
+      "loss": 1.5478,
+      "step": 3414
+    },
+    {
+      "epoch": 0.7088003320880033,
+      "grad_norm": 1.8059073891547446,
+      "learning_rate": 1.1021355078359105e-07,
+      "loss": 1.5069,
+      "step": 3415
+    },
+    {
+      "epoch": 0.7090078870900789,
+      "grad_norm": 0.9499686968573035,
+      "learning_rate": 1.1012056350020103e-07,
+      "loss": 1.5489,
+      "step": 3416
+    },
+    {
+      "epoch": 0.7092154420921544,
+      "grad_norm": 1.6790197019262487,
+      "learning_rate": 1.1002762293463041e-07,
+      "loss": 1.4396,
+      "step": 3417
+    },
+    {
+      "epoch": 0.70942299709423,
+      "grad_norm": 1.6726372354046732,
+      "learning_rate": 1.0993472912639483e-07,
+      "loss": 1.5242,
+      "step": 3418
+    },
+    {
+      "epoch": 0.7096305520963055,
+      "grad_norm": 0.6870671077589678,
+      "learning_rate": 1.0984188211499037e-07,
+      "loss": 1.4819,
+      "step": 3419
+    },
+    {
+      "epoch": 0.709838107098381,
+      "grad_norm": 0.7986800560958398,
+      "learning_rate": 1.0974908193989314e-07,
+      "loss": 1.4839,
+      "step": 3420
+    },
+    {
+      "epoch": 0.7100456621004566,
+      "grad_norm": 0.701225474841331,
+      "learning_rate": 1.0965632864055923e-07,
+      "loss": 1.5095,
+      "step": 3421
+    },
+    {
+      "epoch": 0.7102532171025322,
+      "grad_norm": 0.9000052779641537,
+      "learning_rate": 1.095636222564249e-07,
+      "loss": 1.4381,
+      "step": 3422
+    },
+    {
+      "epoch": 0.7104607721046077,
+      "grad_norm": 0.9760847601231115,
+      "learning_rate": 1.0947096282690651e-07,
+      "loss": 1.5768,
+      "step": 3423
+    },
+    {
+      "epoch": 0.7106683271066833,
+      "grad_norm": 0.7177336090841828,
+      "learning_rate": 1.0937835039140036e-07,
+      "loss": 1.4964,
+      "step": 3424
+    },
+    {
+      "epoch": 0.7108758821087588,
+      "grad_norm": 1.1542762072593606,
+      "learning_rate": 1.0928578498928271e-07,
+      "loss": 1.6029,
+      "step": 3425
+    },
+    {
+      "epoch": 0.7110834371108343,
+      "grad_norm": 2.087455477893475,
+      "learning_rate": 1.0919326665991014e-07,
+      "loss": 1.5359,
+      "step": 3426
+    },
+    {
+      "epoch": 0.7112909921129099,
+      "grad_norm": 0.7889164796682878,
+      "learning_rate": 1.0910079544261867e-07,
+      "loss": 1.588,
+      "step": 3427
+    },
+    {
+      "epoch": 0.7114985471149855,
+      "grad_norm": 0.7167140538992997,
+      "learning_rate": 1.0900837137672487e-07,
+      "loss": 1.5114,
+      "step": 3428
+    },
+    {
+      "epoch": 0.711706102117061,
+      "grad_norm": 0.7436177106588999,
+      "learning_rate": 1.0891599450152488e-07,
+      "loss": 1.5227,
+      "step": 3429
+    },
+    {
+      "epoch": 0.7119136571191366,
+      "grad_norm": 0.7285124348259286,
+      "learning_rate": 1.0882366485629493e-07,
+      "loss": 1.5462,
+      "step": 3430
+    },
+    {
+      "epoch": 0.7121212121212122,
+      "grad_norm": 1.6351554078354409,
+      "learning_rate": 1.0873138248029103e-07,
+      "loss": 1.5069,
+      "step": 3431
+    },
+    {
+      "epoch": 0.7123287671232876,
+      "grad_norm": 1.0571591945968657,
+      "learning_rate": 1.0863914741274944e-07,
+      "loss": 1.6116,
+      "step": 3432
+    },
+    {
+      "epoch": 0.7125363221253632,
+      "grad_norm": 0.7444166242792438,
+      "learning_rate": 1.0854695969288576e-07,
+      "loss": 1.4893,
+      "step": 3433
+    },
+    {
+      "epoch": 0.7127438771274388,
+      "grad_norm": 0.7116820816687842,
+      "learning_rate": 1.0845481935989598e-07,
+      "loss": 1.5422,
+      "step": 3434
+    },
+    {
+      "epoch": 0.7129514321295143,
+      "grad_norm": 1.1038395448310774,
+      "learning_rate": 1.0836272645295567e-07,
+      "loss": 1.5807,
+      "step": 3435
+    },
+    {
+      "epoch": 0.7131589871315899,
+      "grad_norm": 1.0974574100948222,
+      "learning_rate": 1.082706810112202e-07,
+      "loss": 1.4822,
+      "step": 3436
+    },
+    {
+      "epoch": 0.7133665421336655,
+      "grad_norm": 0.7683395681093691,
+      "learning_rate": 1.0817868307382499e-07,
+      "loss": 1.5411,
+      "step": 3437
+    },
+    {
+      "epoch": 0.7135740971357409,
+      "grad_norm": 0.7559514818352393,
+      "learning_rate": 1.0808673267988501e-07,
+      "loss": 1.5509,
+      "step": 3438
+    },
+    {
+      "epoch": 0.7137816521378165,
+      "grad_norm": 1.1196120224305883,
+      "learning_rate": 1.0799482986849517e-07,
+      "loss": 1.5054,
+      "step": 3439
+    },
+    {
+      "epoch": 0.7139892071398921,
+      "grad_norm": 0.6249036262364656,
+      "learning_rate": 1.0790297467873006e-07,
+      "loss": 1.4911,
+      "step": 3440
+    },
+    {
+      "epoch": 0.7141967621419676,
+      "grad_norm": 0.67064140653349,
+      "learning_rate": 1.0781116714964425e-07,
+      "loss": 1.5888,
+      "step": 3441
+    },
+    {
+      "epoch": 0.7144043171440432,
+      "grad_norm": 0.7229316229756434,
+      "learning_rate": 1.0771940732027158e-07,
+      "loss": 1.5085,
+      "step": 3442
+    },
+    {
+      "epoch": 0.7146118721461188,
+      "grad_norm": 1.0515234357392989,
+      "learning_rate": 1.0762769522962609e-07,
+      "loss": 1.5273,
+      "step": 3443
+    },
+    {
+      "epoch": 0.7148194271481942,
+      "grad_norm": 0.8279528336345802,
+      "learning_rate": 1.075360309167013e-07,
+      "loss": 1.5189,
+      "step": 3444
+    },
+    {
+      "epoch": 0.7150269821502698,
+      "grad_norm": 0.6841549818736408,
+      "learning_rate": 1.0744441442047038e-07,
+      "loss": 1.5353,
+      "step": 3445
+    },
+    {
+      "epoch": 0.7152345371523454,
+      "grad_norm": 1.0487289234961572,
+      "learning_rate": 1.0735284577988624e-07,
+      "loss": 1.6083,
+      "step": 3446
+    },
+    {
+      "epoch": 0.7154420921544209,
+      "grad_norm": 0.9674360105650714,
+      "learning_rate": 1.0726132503388157e-07,
+      "loss": 1.5686,
+      "step": 3447
+    },
+    {
+      "epoch": 0.7156496471564965,
+      "grad_norm": 0.765445488702532,
+      "learning_rate": 1.0716985222136833e-07,
+      "loss": 1.5377,
+      "step": 3448
+    },
+    {
+      "epoch": 0.7158572021585721,
+      "grad_norm": 0.7100025050864576,
+      "learning_rate": 1.0707842738123853e-07,
+      "loss": 1.5377,
+      "step": 3449
+    },
+    {
+      "epoch": 0.7160647571606475,
+      "grad_norm": 0.7078256190335965,
+      "learning_rate": 1.0698705055236347e-07,
+      "loss": 1.491,
+      "step": 3450
+    },
+    {
+      "epoch": 0.7162723121627231,
+      "grad_norm": 0.8198441173871723,
+      "learning_rate": 1.0689572177359419e-07,
+      "loss": 1.5363,
+      "step": 3451
+    },
+    {
+      "epoch": 0.7164798671647987,
+      "grad_norm": 2.647231521430732,
+      "learning_rate": 1.0680444108376128e-07,
+      "loss": 1.5115,
+      "step": 3452
+    },
+    {
+      "epoch": 0.7166874221668742,
+      "grad_norm": 0.6459301821916063,
+      "learning_rate": 1.0671320852167487e-07,
+      "loss": 1.5117,
+      "step": 3453
+    },
+    {
+      "epoch": 0.7168949771689498,
+      "grad_norm": 1.2455895148294254,
+      "learning_rate": 1.066220241261246e-07,
+      "loss": 1.6328,
+      "step": 3454
+    },
+    {
+      "epoch": 0.7171025321710254,
+      "grad_norm": 1.0920924776240006,
+      "learning_rate": 1.0653088793587959e-07,
+      "loss": 1.6182,
+      "step": 3455
+    },
+    {
+      "epoch": 0.7173100871731009,
+      "grad_norm": 0.6943053006762021,
+      "learning_rate": 1.0643979998968874e-07,
+      "loss": 1.5335,
+      "step": 3456
+    },
+    {
+      "epoch": 0.7175176421751764,
+      "grad_norm": 0.8267328563694016,
+      "learning_rate": 1.0634876032627995e-07,
+      "loss": 1.4867,
+      "step": 3457
+    },
+    {
+      "epoch": 0.7177251971772519,
+      "grad_norm": 0.6788692500886616,
+      "learning_rate": 1.0625776898436109e-07,
+      "loss": 1.4977,
+      "step": 3458
+    },
+    {
+      "epoch": 0.7179327521793275,
+      "grad_norm": 1.2025370458462237,
+      "learning_rate": 1.0616682600261908e-07,
+      "loss": 1.5043,
+      "step": 3459
+    },
+    {
+      "epoch": 0.7181403071814031,
+      "grad_norm": 0.9235647677862718,
+      "learning_rate": 1.0607593141972065e-07,
+      "loss": 1.4992,
+      "step": 3460
+    },
+    {
+      "epoch": 0.7183478621834786,
+      "grad_norm": 0.7197334293487501,
+      "learning_rate": 1.059850852743116e-07,
+      "loss": 1.4912,
+      "step": 3461
+    },
+    {
+      "epoch": 0.7185554171855542,
+      "grad_norm": 0.7836750158583452,
+      "learning_rate": 1.0589428760501735e-07,
+      "loss": 1.5512,
+      "step": 3462
+    },
+    {
+      "epoch": 0.7187629721876297,
+      "grad_norm": 0.9897917405017157,
+      "learning_rate": 1.0580353845044272e-07,
+      "loss": 1.4891,
+      "step": 3463
+    },
+    {
+      "epoch": 0.7189705271897052,
+      "grad_norm": 0.7329757449722253,
+      "learning_rate": 1.057128378491718e-07,
+      "loss": 1.5052,
+      "step": 3464
+    },
+    {
+      "epoch": 0.7191780821917808,
+      "grad_norm": 1.3481118957459333,
+      "learning_rate": 1.0562218583976808e-07,
+      "loss": 1.467,
+      "step": 3465
+    },
+    {
+      "epoch": 0.7193856371938564,
+      "grad_norm": 0.8198451687075072,
+      "learning_rate": 1.0553158246077432e-07,
+      "loss": 1.5052,
+      "step": 3466
+    },
+    {
+      "epoch": 0.7195931921959319,
+      "grad_norm": 0.7314624809062473,
+      "learning_rate": 1.0544102775071286e-07,
+      "loss": 1.5383,
+      "step": 3467
+    },
+    {
+      "epoch": 0.7198007471980075,
+      "grad_norm": 1.2523254175250966,
+      "learning_rate": 1.053505217480849e-07,
+      "loss": 1.5512,
+      "step": 3468
+    },
+    {
+      "epoch": 0.720008302200083,
+      "grad_norm": 0.7628103383523822,
+      "learning_rate": 1.052600644913714e-07,
+      "loss": 1.5053,
+      "step": 3469
+    },
+    {
+      "epoch": 0.7202158572021585,
+      "grad_norm": 0.7908765517854673,
+      "learning_rate": 1.0516965601903229e-07,
+      "loss": 1.5232,
+      "step": 3470
+    },
+    {
+      "epoch": 0.7204234122042341,
+      "grad_norm": 0.7443546871985859,
+      "learning_rate": 1.0507929636950686e-07,
+      "loss": 1.5312,
+      "step": 3471
+    },
+    {
+      "epoch": 0.7206309672063097,
+      "grad_norm": 5.823675098534744,
+      "learning_rate": 1.0498898558121361e-07,
+      "loss": 1.5589,
+      "step": 3472
+    },
+    {
+      "epoch": 0.7208385222083852,
+      "grad_norm": 0.9633263239700761,
+      "learning_rate": 1.0489872369255044e-07,
+      "loss": 1.4893,
+      "step": 3473
+    },
+    {
+      "epoch": 0.7210460772104608,
+      "grad_norm": 0.9445556149735829,
+      "learning_rate": 1.0480851074189405e-07,
+      "loss": 1.5584,
+      "step": 3474
+    },
+    {
+      "epoch": 0.7212536322125364,
+      "grad_norm": 0.7408658572192602,
+      "learning_rate": 1.047183467676008e-07,
+      "loss": 1.5427,
+      "step": 3475
+    },
+    {
+      "epoch": 0.7214611872146118,
+      "grad_norm": 0.7182503479605851,
+      "learning_rate": 1.0462823180800592e-07,
+      "loss": 1.5319,
+      "step": 3476
+    },
+    {
+      "epoch": 0.7216687422166874,
+      "grad_norm": 0.6768545648627804,
+      "learning_rate": 1.045381659014239e-07,
+      "loss": 1.4635,
+      "step": 3477
+    },
+    {
+      "epoch": 0.721876297218763,
+      "grad_norm": 0.9073556237571551,
+      "learning_rate": 1.0444814908614847e-07,
+      "loss": 1.5581,
+      "step": 3478
+    },
+    {
+      "epoch": 0.7220838522208385,
+      "grad_norm": 0.7015626924716445,
+      "learning_rate": 1.0435818140045232e-07,
+      "loss": 1.4892,
+      "step": 3479
+    },
+    {
+      "epoch": 0.7222914072229141,
+      "grad_norm": 0.927316628980158,
+      "learning_rate": 1.0426826288258732e-07,
+      "loss": 1.5168,
+      "step": 3480
+    },
+    {
+      "epoch": 0.7224989622249897,
+      "grad_norm": 0.6851525828446593,
+      "learning_rate": 1.0417839357078438e-07,
+      "loss": 1.498,
+      "step": 3481
+    },
+    {
+      "epoch": 0.7227065172270651,
+      "grad_norm": 1.3886246675256522,
+      "learning_rate": 1.0408857350325373e-07,
+      "loss": 1.4486,
+      "step": 3482
+    },
+    {
+      "epoch": 0.7229140722291407,
+      "grad_norm": 0.9593852785656315,
+      "learning_rate": 1.039988027181842e-07,
+      "loss": 1.5566,
+      "step": 3483
+    },
+    {
+      "epoch": 0.7231216272312163,
+      "grad_norm": 0.7535439580518706,
+      "learning_rate": 1.039090812537442e-07,
+      "loss": 1.5244,
+      "step": 3484
+    },
+    {
+      "epoch": 0.7233291822332918,
+      "grad_norm": 1.0377846126660604,
+      "learning_rate": 1.0381940914808079e-07,
+      "loss": 1.5072,
+      "step": 3485
+    },
+    {
+      "epoch": 0.7235367372353674,
+      "grad_norm": 0.6944576851705965,
+      "learning_rate": 1.0372978643932017e-07,
+      "loss": 1.4807,
+      "step": 3486
+    },
+    {
+      "epoch": 0.723744292237443,
+      "grad_norm": 1.365358618798681,
+      "learning_rate": 1.0364021316556753e-07,
+      "loss": 1.5216,
+      "step": 3487
+    },
+    {
+      "epoch": 0.7239518472395184,
+      "grad_norm": 1.837688085476273,
+      "learning_rate": 1.0355068936490714e-07,
+      "loss": 1.6805,
+      "step": 3488
+    },
+    {
+      "epoch": 0.724159402241594,
+      "grad_norm": 1.0460044808421052,
+      "learning_rate": 1.0346121507540207e-07,
+      "loss": 1.5613,
+      "step": 3489
+    },
+    {
+      "epoch": 0.7243669572436696,
+      "grad_norm": 1.3950140663041959,
+      "learning_rate": 1.033717903350944e-07,
+      "loss": 1.5042,
+      "step": 3490
+    },
+    {
+      "epoch": 0.7245745122457451,
+      "grad_norm": 0.7513321151996887,
+      "learning_rate": 1.032824151820052e-07,
+      "loss": 1.4694,
+      "step": 3491
+    },
+    {
+      "epoch": 0.7247820672478207,
+      "grad_norm": 0.7629524847712228,
+      "learning_rate": 1.0319308965413432e-07,
+      "loss": 1.4591,
+      "step": 3492
+    },
+    {
+      "epoch": 0.7249896222498963,
+      "grad_norm": 0.9518255905102921,
+      "learning_rate": 1.0310381378946073e-07,
+      "loss": 1.5944,
+      "step": 3493
+    },
+    {
+      "epoch": 0.7251971772519717,
+      "grad_norm": 0.6162696343292873,
+      "learning_rate": 1.0301458762594211e-07,
+      "loss": 1.5008,
+      "step": 3494
+    },
+    {
+      "epoch": 0.7254047322540473,
+      "grad_norm": 0.8022839055270045,
+      "learning_rate": 1.0292541120151504e-07,
+      "loss": 1.511,
+      "step": 3495
+    },
+    {
+      "epoch": 0.7256122872561229,
+      "grad_norm": 1.0940071065748838,
+      "learning_rate": 1.0283628455409492e-07,
+      "loss": 1.5909,
+      "step": 3496
+    },
+    {
+      "epoch": 0.7258198422581984,
+      "grad_norm": 1.1709218608870133,
+      "learning_rate": 1.0274720772157621e-07,
+      "loss": 1.5549,
+      "step": 3497
+    },
+    {
+      "epoch": 0.726027397260274,
+      "grad_norm": 0.9878535934356821,
+      "learning_rate": 1.0265818074183173e-07,
+      "loss": 1.461,
+      "step": 3498
+    },
+    {
+      "epoch": 0.7262349522623496,
+      "grad_norm": 0.9975102925754061,
+      "learning_rate": 1.0256920365271364e-07,
+      "loss": 1.5375,
+      "step": 3499
+    },
+    {
+      "epoch": 0.726442507264425,
+      "grad_norm": 0.8032812596782868,
+      "learning_rate": 1.0248027649205258e-07,
+      "loss": 1.5211,
+      "step": 3500
+    },
+    {
+      "epoch": 0.7266500622665006,
+      "grad_norm": 1.060000298412728,
+      "learning_rate": 1.0239139929765794e-07,
+      "loss": 1.5059,
+      "step": 3501
+    },
+    {
+      "epoch": 0.7268576172685761,
+      "grad_norm": 1.0244438171236696,
+      "learning_rate": 1.0230257210731798e-07,
+      "loss": 1.5762,
+      "step": 3502
+    },
+    {
+      "epoch": 0.7270651722706517,
+      "grad_norm": 0.9245184548745388,
+      "learning_rate": 1.0221379495879964e-07,
+      "loss": 1.5557,
+      "step": 3503
+    },
+    {
+      "epoch": 0.7272727272727273,
+      "grad_norm": 0.7810132847151354,
+      "learning_rate": 1.0212506788984869e-07,
+      "loss": 1.5716,
+      "step": 3504
+    },
+    {
+      "epoch": 0.7274802822748028,
+      "grad_norm": 0.7138421150836209,
+      "learning_rate": 1.0203639093818948e-07,
+      "loss": 1.5515,
+      "step": 3505
+    },
+    {
+      "epoch": 0.7276878372768784,
+      "grad_norm": 1.1715676253441438,
+      "learning_rate": 1.0194776414152508e-07,
+      "loss": 1.5042,
+      "step": 3506
+    },
+    {
+      "epoch": 0.727895392278954,
+      "grad_norm": 0.9403212754332279,
+      "learning_rate": 1.018591875375372e-07,
+      "loss": 1.4847,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7281029472810294,
+      "grad_norm": 0.8960275640053457,
+      "learning_rate": 1.0177066116388644e-07,
+      "loss": 1.5453,
+      "step": 3508
+    },
+    {
+      "epoch": 0.728310502283105,
+      "grad_norm": 0.6534933979268344,
+      "learning_rate": 1.016821850582116e-07,
+      "loss": 1.4915,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7285180572851806,
+      "grad_norm": 0.7161478778129308,
+      "learning_rate": 1.0159375925813058e-07,
+      "loss": 1.4892,
+      "step": 3510
+    },
+    {
+      "epoch": 0.7287256122872561,
+      "grad_norm": 0.7409351012650822,
+      "learning_rate": 1.0150538380123963e-07,
+      "loss": 1.5268,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7289331672893317,
+      "grad_norm": 0.6564319732829017,
+      "learning_rate": 1.0141705872511366e-07,
+      "loss": 1.5091,
+      "step": 3512
+    },
+    {
+      "epoch": 0.7291407222914073,
+      "grad_norm": 0.8775484123968684,
+      "learning_rate": 1.0132878406730601e-07,
+      "loss": 1.5611,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7293482772934827,
+      "grad_norm": 0.7175917073983599,
+      "learning_rate": 1.0124055986534894e-07,
+      "loss": 1.494,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7295558322955583,
+      "grad_norm": 0.8688244746594186,
+      "learning_rate": 1.0115238615675287e-07,
+      "loss": 1.411,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7297633872976339,
+      "grad_norm": 0.9891404415327963,
+      "learning_rate": 1.0106426297900705e-07,
+      "loss": 1.4801,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7299709422997094,
+      "grad_norm": 0.8288369506330701,
+      "learning_rate": 1.0097619036957903e-07,
+      "loss": 1.5297,
+      "step": 3517
+    },
+    {
+      "epoch": 0.730178497301785,
+      "grad_norm": 0.6572737414427684,
+      "learning_rate": 1.008881683659149e-07,
+      "loss": 1.564,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7303860523038606,
+      "grad_norm": 1.139785157304229,
+      "learning_rate": 1.0080019700543938e-07,
+      "loss": 1.6262,
+      "step": 3519
+    },
+    {
+      "epoch": 0.730593607305936,
+      "grad_norm": 0.7304149356373166,
+      "learning_rate": 1.007122763255555e-07,
+      "loss": 1.5035,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7308011623080116,
+      "grad_norm": 0.7315328606683559,
+      "learning_rate": 1.0062440636364487e-07,
+      "loss": 1.4751,
+      "step": 3521
+    },
+    {
+      "epoch": 0.7310087173100872,
+      "grad_norm": 0.6663624177177068,
+      "learning_rate": 1.0053658715706732e-07,
+      "loss": 1.5196,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7312162723121627,
+      "grad_norm": 0.8689010473345146,
+      "learning_rate": 1.0044881874316147e-07,
+      "loss": 1.4941,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7314238273142383,
+      "grad_norm": 0.8804783563845247,
+      "learning_rate": 1.0036110115924388e-07,
+      "loss": 1.5124,
+      "step": 3524
+    },
+    {
+      "epoch": 0.7316313823163139,
+      "grad_norm": 0.7191593442233523,
+      "learning_rate": 1.0027343444260989e-07,
+      "loss": 1.5318,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7318389373183893,
+      "grad_norm": 0.8240079618409041,
+      "learning_rate": 1.0018581863053302e-07,
+      "loss": 1.5371,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7320464923204649,
+      "grad_norm": 0.8446927467521336,
+      "learning_rate": 1.0009825376026517e-07,
+      "loss": 1.5352,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7322540473225405,
+      "grad_norm": 0.7817214014588983,
+      "learning_rate": 1.0001073986903655e-07,
+      "loss": 1.4841,
+      "step": 3528
+    },
+    {
+      "epoch": 0.732461602324616,
+      "grad_norm": 0.7302446948987851,
+      "learning_rate": 9.992327699405587e-08,
+      "loss": 1.5353,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7326691573266916,
+      "grad_norm": 0.8350525115135822,
+      "learning_rate": 9.983586517250995e-08,
+      "loss": 1.552,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7328767123287672,
+      "grad_norm": 0.6675648454335639,
+      "learning_rate": 9.974850444156393e-08,
+      "loss": 1.5087,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7330842673308426,
+      "grad_norm": 0.7316053247659076,
+      "learning_rate": 9.966119483836144e-08,
+      "loss": 1.5407,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7332918223329182,
+      "grad_norm": 0.781601385051515,
+      "learning_rate": 9.957393640002398e-08,
+      "loss": 1.4458,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7334993773349938,
+      "grad_norm": 0.7570949871519662,
+      "learning_rate": 9.948672916365172e-08,
+      "loss": 1.514,
+      "step": 3534
+    },
+    {
+      "epoch": 0.7337069323370693,
+      "grad_norm": 0.7145193620209511,
+      "learning_rate": 9.939957316632273e-08,
+      "loss": 1.5104,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7339144873391449,
+      "grad_norm": 0.7024855382939132,
+      "learning_rate": 9.931246844509349e-08,
+      "loss": 1.6003,
+      "step": 3536
+    },
+    {
+      "epoch": 0.7341220423412205,
+      "grad_norm": 0.7871894616169899,
+      "learning_rate": 9.922541503699854e-08,
+      "loss": 1.571,
+      "step": 3537
+    },
+    {
+      "epoch": 0.734329597343296,
+      "grad_norm": 1.0107601811326696,
+      "learning_rate": 9.913841297905084e-08,
+      "loss": 1.5042,
+      "step": 3538
+    },
+    {
+      "epoch": 0.7345371523453715,
+      "grad_norm": 0.6877909911620147,
+      "learning_rate": 9.905146230824111e-08,
+      "loss": 1.4833,
+      "step": 3539
+    },
+    {
+      "epoch": 0.7347447073474471,
+      "grad_norm": 0.6940409940112109,
+      "learning_rate": 9.89645630615387e-08,
+      "loss": 1.4972,
+      "step": 3540
+    },
+    {
+      "epoch": 0.7349522623495226,
+      "grad_norm": 0.9914106869704771,
+      "learning_rate": 9.887771527589076e-08,
+      "loss": 1.5252,
+      "step": 3541
+    },
+    {
+      "epoch": 0.7351598173515982,
+      "grad_norm": 0.7530933879533854,
+      "learning_rate": 9.879091898822259e-08,
+      "loss": 1.5458,
+      "step": 3542
+    },
+    {
+      "epoch": 0.7353673723536738,
+      "grad_norm": 0.7820889112215141,
+      "learning_rate": 9.870417423543783e-08,
+      "loss": 1.5081,
+      "step": 3543
+    },
+    {
+      "epoch": 0.7355749273557493,
+      "grad_norm": 7.343112071419662,
+      "learning_rate": 9.861748105441796e-08,
+      "loss": 1.5927,
+      "step": 3544
+    },
+    {
+      "epoch": 0.7357824823578248,
+      "grad_norm": 1.1255535248113777,
+      "learning_rate": 9.853083948202262e-08,
+      "loss": 1.557,
+      "step": 3545
+    },
+    {
+      "epoch": 0.7359900373599004,
+      "grad_norm": 0.7464328021477331,
+      "learning_rate": 9.844424955508946e-08,
+      "loss": 1.5543,
+      "step": 3546
+    },
+    {
+      "epoch": 0.7361975923619759,
+      "grad_norm": 1.1609289607772046,
+      "learning_rate": 9.835771131043437e-08,
+      "loss": 1.5586,
+      "step": 3547
+    },
+    {
+      "epoch": 0.7364051473640515,
+      "grad_norm": 0.745131720083381,
+      "learning_rate": 9.827122478485091e-08,
+      "loss": 1.4728,
+      "step": 3548
+    },
+    {
+      "epoch": 0.736612702366127,
+      "grad_norm": 1.1781646569331006,
+      "learning_rate": 9.818479001511108e-08,
+      "loss": 1.5383,
+      "step": 3549
+    },
+    {
+      "epoch": 0.7368202573682026,
+      "grad_norm": 0.916623850817447,
+      "learning_rate": 9.809840703796436e-08,
+      "loss": 1.5477,
+      "step": 3550
+    },
+    {
+      "epoch": 0.7370278123702781,
+      "grad_norm": 1.1912368837048912,
+      "learning_rate": 9.801207589013877e-08,
+      "loss": 1.4877,
+      "step": 3551
+    },
+    {
+      "epoch": 0.7372353673723536,
+      "grad_norm": 0.6688069036968155,
+      "learning_rate": 9.79257966083399e-08,
+      "loss": 1.5085,
+      "step": 3552
+    },
+    {
+      "epoch": 0.7374429223744292,
+      "grad_norm": 0.8349703665188563,
+      "learning_rate": 9.783956922925143e-08,
+      "loss": 1.4892,
+      "step": 3553
+    },
+    {
+      "epoch": 0.7376504773765048,
+      "grad_norm": 0.7333367595107295,
+      "learning_rate": 9.775339378953489e-08,
+      "loss": 1.583,
+      "step": 3554
+    },
+    {
+      "epoch": 0.7378580323785803,
+      "grad_norm": 1.0229688919066944,
+      "learning_rate": 9.766727032582991e-08,
+      "loss": 1.5252,
+      "step": 3555
+    },
+    {
+      "epoch": 0.7380655873806559,
+      "grad_norm": 0.9033812777674429,
+      "learning_rate": 9.758119887475383e-08,
+      "loss": 1.4958,
+      "step": 3556
+    },
+    {
+      "epoch": 0.7382731423827315,
+      "grad_norm": 1.0052481750564066,
+      "learning_rate": 9.74951794729019e-08,
+      "loss": 1.4912,
+      "step": 3557
+    },
+    {
+      "epoch": 0.7384806973848069,
+      "grad_norm": 0.6936452152040502,
+      "learning_rate": 9.740921215684746e-08,
+      "loss": 1.4876,
+      "step": 3558
+    },
+    {
+      "epoch": 0.7386882523868825,
+      "grad_norm": 0.6390009641008648,
+      "learning_rate": 9.732329696314128e-08,
+      "loss": 1.5676,
+      "step": 3559
+    },
+    {
+      "epoch": 0.7388958073889581,
+      "grad_norm": 0.7461845916104995,
+      "learning_rate": 9.723743392831242e-08,
+      "loss": 1.5522,
+      "step": 3560
+    },
+    {
+      "epoch": 0.7391033623910336,
+      "grad_norm": 1.1391861126335914,
+      "learning_rate": 9.715162308886748e-08,
+      "loss": 1.4372,
+      "step": 3561
+    },
+    {
+      "epoch": 0.7393109173931092,
+      "grad_norm": 1.0590953767184115,
+      "learning_rate": 9.706586448129098e-08,
+      "loss": 1.5627,
+      "step": 3562
+    },
+    {
+      "epoch": 0.7395184723951848,
+      "grad_norm": 1.0695954128895118,
+      "learning_rate": 9.698015814204513e-08,
+      "loss": 1.4766,
+      "step": 3563
+    },
+    {
+      "epoch": 0.7397260273972602,
+      "grad_norm": 0.730742232955605,
+      "learning_rate": 9.689450410757015e-08,
+      "loss": 1.4653,
+      "step": 3564
+    },
+    {
+      "epoch": 0.7399335823993358,
+      "grad_norm": 1.2244434674195188,
+      "learning_rate": 9.680890241428366e-08,
+      "loss": 1.5253,
+      "step": 3565
+    },
+    {
+      "epoch": 0.7401411374014114,
+      "grad_norm": 0.6902041798656504,
+      "learning_rate": 9.672335309858136e-08,
+      "loss": 1.5975,
+      "step": 3566
+    },
+    {
+      "epoch": 0.7403486924034869,
+      "grad_norm": 0.9870079348127876,
+      "learning_rate": 9.663785619683654e-08,
+      "loss": 1.5879,
+      "step": 3567
+    },
+    {
+      "epoch": 0.7405562474055625,
+      "grad_norm": 0.8590093097119726,
+      "learning_rate": 9.65524117454001e-08,
+      "loss": 1.5186,
+      "step": 3568
+    },
+    {
+      "epoch": 0.7407638024076381,
+      "grad_norm": 0.9991589484120666,
+      "learning_rate": 9.646701978060092e-08,
+      "loss": 1.5371,
+      "step": 3569
+    },
+    {
+      "epoch": 0.7409713574097135,
+      "grad_norm": 0.8045653839935082,
+      "learning_rate": 9.63816803387453e-08,
+      "loss": 1.5223,
+      "step": 3570
+    },
+    {
+      "epoch": 0.7411789124117891,
+      "grad_norm": 1.055883259237904,
+      "learning_rate": 9.629639345611733e-08,
+      "loss": 1.5017,
+      "step": 3571
+    },
+    {
+      "epoch": 0.7413864674138647,
+      "grad_norm": 0.6599155835299202,
+      "learning_rate": 9.621115916897863e-08,
+      "loss": 1.5045,
+      "step": 3572
+    },
+    {
+      "epoch": 0.7415940224159402,
+      "grad_norm": 0.7489401737430197,
+      "learning_rate": 9.612597751356881e-08,
+      "loss": 1.5073,
+      "step": 3573
+    },
+    {
+      "epoch": 0.7418015774180158,
+      "grad_norm": 1.0100408628632684,
+      "learning_rate": 9.604084852610455e-08,
+      "loss": 1.4853,
+      "step": 3574
+    },
+    {
+      "epoch": 0.7420091324200914,
+      "grad_norm": 1.3088106546240301,
+      "learning_rate": 9.59557722427806e-08,
+      "loss": 1.5599,
+      "step": 3575
+    },
+    {
+      "epoch": 0.7422166874221668,
+      "grad_norm": 1.0287162147288085,
+      "learning_rate": 9.587074869976913e-08,
+      "loss": 1.4606,
+      "step": 3576
+    },
+    {
+      "epoch": 0.7424242424242424,
+      "grad_norm": 1.8692419602648973,
+      "learning_rate": 9.578577793321987e-08,
+      "loss": 1.5265,
+      "step": 3577
+    },
+    {
+      "epoch": 0.742631797426318,
+      "grad_norm": 0.7643858945820965,
+      "learning_rate": 9.570085997926007e-08,
+      "loss": 1.6107,
+      "step": 3578
+    },
+    {
+      "epoch": 0.7428393524283935,
+      "grad_norm": 1.2367791797192762,
+      "learning_rate": 9.561599487399477e-08,
+      "loss": 1.5758,
+      "step": 3579
+    },
+    {
+      "epoch": 0.7430469074304691,
+      "grad_norm": 1.1170930395952943,
+      "learning_rate": 9.553118265350612e-08,
+      "loss": 1.5485,
+      "step": 3580
+    },
+    {
+      "epoch": 0.7432544624325447,
+      "grad_norm": 0.8440723816098336,
+      "learning_rate": 9.54464233538542e-08,
+      "loss": 1.562,
+      "step": 3581
+    },
+    {
+      "epoch": 0.7434620174346201,
+      "grad_norm": 0.7638712575843175,
+      "learning_rate": 9.536171701107638e-08,
+      "loss": 1.47,
+      "step": 3582
+    },
+    {
+      "epoch": 0.7436695724366957,
+      "grad_norm": 0.805192030151996,
+      "learning_rate": 9.527706366118746e-08,
+      "loss": 1.5099,
+      "step": 3583
+    },
+    {
+      "epoch": 0.7438771274387713,
+      "grad_norm": 1.01454666943058,
+      "learning_rate": 9.519246334017993e-08,
+      "loss": 1.6119,
+      "step": 3584
+    },
+    {
+      "epoch": 0.7440846824408468,
+      "grad_norm": 0.7878363968586458,
+      "learning_rate": 9.510791608402352e-08,
+      "loss": 1.5805,
+      "step": 3585
+    },
+    {
+      "epoch": 0.7442922374429224,
+      "grad_norm": 0.793042311915429,
+      "learning_rate": 9.50234219286655e-08,
+      "loss": 1.5931,
+      "step": 3586
+    },
+    {
+      "epoch": 0.744499792444998,
+      "grad_norm": 0.6847976949602224,
+      "learning_rate": 9.493898091003047e-08,
+      "loss": 1.4501,
+      "step": 3587
+    },
+    {
+      "epoch": 0.7447073474470735,
+      "grad_norm": 0.6412388438618505,
+      "learning_rate": 9.485459306402071e-08,
+      "loss": 1.5356,
+      "step": 3588
+    },
+    {
+      "epoch": 0.744914902449149,
+      "grad_norm": 0.7173748124955761,
+      "learning_rate": 9.477025842651545e-08,
+      "loss": 1.4932,
+      "step": 3589
+    },
+    {
+      "epoch": 0.7451224574512246,
+      "grad_norm": 1.001471632627153,
+      "learning_rate": 9.468597703337168e-08,
+      "loss": 1.554,
+      "step": 3590
+    },
+    {
+      "epoch": 0.7453300124533001,
+      "grad_norm": 0.6349163338583317,
+      "learning_rate": 9.460174892042359e-08,
+      "loss": 1.5331,
+      "step": 3591
+    },
+    {
+      "epoch": 0.7455375674553757,
+      "grad_norm": 0.7513834034170537,
+      "learning_rate": 9.451757412348272e-08,
+      "loss": 1.5469,
+      "step": 3592
+    },
+    {
+      "epoch": 0.7457451224574512,
+      "grad_norm": 0.8243141395715089,
+      "learning_rate": 9.443345267833797e-08,
+      "loss": 1.5075,
+      "step": 3593
+    },
+    {
+      "epoch": 0.7459526774595268,
+      "grad_norm": 0.7663924863171154,
+      "learning_rate": 9.434938462075545e-08,
+      "loss": 1.4336,
+      "step": 3594
+    },
+    {
+      "epoch": 0.7461602324616023,
+      "grad_norm": 0.7371219286925582,
+      "learning_rate": 9.426536998647886e-08,
+      "loss": 1.5257,
+      "step": 3595
+    },
+    {
+      "epoch": 0.7463677874636778,
+      "grad_norm": 0.6943179125038027,
+      "learning_rate": 9.418140881122886e-08,
+      "loss": 1.5305,
+      "step": 3596
+    },
+    {
+      "epoch": 0.7465753424657534,
+      "grad_norm": 1.5726619247420668,
+      "learning_rate": 9.409750113070357e-08,
+      "loss": 1.4289,
+      "step": 3597
+    },
+    {
+      "epoch": 0.746782897467829,
+      "grad_norm": 0.7556701318492306,
+      "learning_rate": 9.401364698057818e-08,
+      "loss": 1.4868,
+      "step": 3598
+    },
+    {
+      "epoch": 0.7469904524699045,
+      "grad_norm": 1.1105610790603844,
+      "learning_rate": 9.392984639650546e-08,
+      "loss": 1.5944,
+      "step": 3599
+    },
+    {
+      "epoch": 0.7471980074719801,
+      "grad_norm": 0.7832087145495985,
+      "learning_rate": 9.3846099414115e-08,
+      "loss": 1.5109,
+      "step": 3600
+    },
+    {
+      "epoch": 0.7474055624740557,
+      "grad_norm": 0.6840864812485508,
+      "learning_rate": 9.37624060690139e-08,
+      "loss": 1.4812,
+      "step": 3601
+    },
+    {
+      "epoch": 0.7476131174761311,
+      "grad_norm": 0.8426543652504688,
+      "learning_rate": 9.367876639678635e-08,
+      "loss": 1.4512,
+      "step": 3602
+    },
+    {
+      "epoch": 0.7478206724782067,
+      "grad_norm": 1.7193620263692206,
+      "learning_rate": 9.359518043299368e-08,
+      "loss": 1.5339,
+      "step": 3603
+    },
+    {
+      "epoch": 0.7480282274802823,
+      "grad_norm": 1.0330931124169942,
+      "learning_rate": 9.351164821317436e-08,
+      "loss": 1.5384,
+      "step": 3604
+    },
+    {
+      "epoch": 0.7482357824823578,
+      "grad_norm": 0.8152406280891313,
+      "learning_rate": 9.342816977284423e-08,
+      "loss": 1.526,
+      "step": 3605
+    },
+    {
+      "epoch": 0.7484433374844334,
+      "grad_norm": 0.6346991091453212,
+      "learning_rate": 9.3344745147496e-08,
+      "loss": 1.5611,
+      "step": 3606
+    },
+    {
+      "epoch": 0.748650892486509,
+      "grad_norm": 0.7762644819740941,
+      "learning_rate": 9.326137437259963e-08,
+      "loss": 1.5273,
+      "step": 3607
+    },
+    {
+      "epoch": 0.7488584474885844,
+      "grad_norm": 0.7579529789462294,
+      "learning_rate": 9.317805748360216e-08,
+      "loss": 1.4517,
+      "step": 3608
+    },
+    {
+      "epoch": 0.74906600249066,
+      "grad_norm": 0.8055395691430277,
+      "learning_rate": 9.309479451592766e-08,
+      "loss": 1.6014,
+      "step": 3609
+    },
+    {
+      "epoch": 0.7492735574927356,
+      "grad_norm": 0.6772453086448496,
+      "learning_rate": 9.301158550497743e-08,
+      "loss": 1.4379,
+      "step": 3610
+    },
+    {
+      "epoch": 0.7494811124948111,
+      "grad_norm": 1.1633975361715683,
+      "learning_rate": 9.292843048612973e-08,
+      "loss": 1.4662,
+      "step": 3611
+    },
+    {
+      "epoch": 0.7496886674968867,
+      "grad_norm": 0.7628559056287824,
+      "learning_rate": 9.284532949473979e-08,
+      "loss": 1.5149,
+      "step": 3612
+    },
+    {
+      "epoch": 0.7498962224989623,
+      "grad_norm": 0.7102182469626219,
+      "learning_rate": 9.276228256613996e-08,
+      "loss": 1.4979,
+      "step": 3613
+    },
+    {
+      "epoch": 0.7501037775010377,
+      "grad_norm": 0.7879041883352768,
+      "learning_rate": 9.26792897356397e-08,
+      "loss": 1.5298,
+      "step": 3614
+    },
+    {
+      "epoch": 0.7503113325031133,
+      "grad_norm": 0.7259513279103693,
+      "learning_rate": 9.259635103852517e-08,
+      "loss": 1.5131,
+      "step": 3615
+    },
+    {
+      "epoch": 0.7505188875051889,
+      "grad_norm": 0.7684661547124279,
+      "learning_rate": 9.251346651005985e-08,
+      "loss": 1.4974,
+      "step": 3616
+    },
+    {
+      "epoch": 0.7507264425072644,
+      "grad_norm": 0.9703851415539535,
+      "learning_rate": 9.243063618548402e-08,
+      "loss": 1.5228,
+      "step": 3617
+    },
+    {
+      "epoch": 0.75093399750934,
+      "grad_norm": 0.9346028824908276,
+      "learning_rate": 9.23478601000149e-08,
+      "loss": 1.54,
+      "step": 3618
+    },
+    {
+      "epoch": 0.7511415525114156,
+      "grad_norm": 0.7843313760590046,
+      "learning_rate": 9.226513828884662e-08,
+      "loss": 1.5666,
+      "step": 3619
+    },
+    {
+      "epoch": 0.751349107513491,
+      "grad_norm": 0.6549349648424905,
+      "learning_rate": 9.218247078715045e-08,
+      "loss": 1.6034,
+      "step": 3620
+    },
+    {
+      "epoch": 0.7515566625155666,
+      "grad_norm": 1.294863237446233,
+      "learning_rate": 9.209985763007435e-08,
+      "loss": 1.4559,
+      "step": 3621
+    },
+    {
+      "epoch": 0.7517642175176422,
+      "grad_norm": 0.6597904381426672,
+      "learning_rate": 9.20172988527432e-08,
+      "loss": 1.4973,
+      "step": 3622
+    },
+    {
+      "epoch": 0.7519717725197177,
+      "grad_norm": 0.7819001676580447,
+      "learning_rate": 9.193479449025885e-08,
+      "loss": 1.5146,
+      "step": 3623
+    },
+    {
+      "epoch": 0.7521793275217933,
+      "grad_norm": 0.9259703191820416,
+      "learning_rate": 9.185234457769988e-08,
+      "loss": 1.4776,
+      "step": 3624
+    },
+    {
+      "epoch": 0.7523868825238689,
+      "grad_norm": 2.3432664924602506,
+      "learning_rate": 9.176994915012192e-08,
+      "loss": 1.5577,
+      "step": 3625
+    },
+    {
+      "epoch": 0.7525944375259443,
+      "grad_norm": 0.9119218311963444,
+      "learning_rate": 9.168760824255727e-08,
+      "loss": 1.4676,
+      "step": 3626
+    },
+    {
+      "epoch": 0.7528019925280199,
+      "grad_norm": 0.7482764664301839,
+      "learning_rate": 9.160532189001508e-08,
+      "loss": 1.5307,
+      "step": 3627
+    },
+    {
+      "epoch": 0.7530095475300955,
+      "grad_norm": 1.1246070961082144,
+      "learning_rate": 9.15230901274813e-08,
+      "loss": 1.4735,
+      "step": 3628
+    },
+    {
+      "epoch": 0.753217102532171,
+      "grad_norm": 1.0573383598738648,
+      "learning_rate": 9.144091298991885e-08,
+      "loss": 1.5765,
+      "step": 3629
+    },
+    {
+      "epoch": 0.7534246575342466,
+      "grad_norm": 0.7302090915243562,
+      "learning_rate": 9.135879051226703e-08,
+      "loss": 1.555,
+      "step": 3630
+    },
+    {
+      "epoch": 0.7536322125363222,
+      "grad_norm": 0.698951332245233,
+      "learning_rate": 9.12767227294423e-08,
+      "loss": 1.4702,
+      "step": 3631
+    },
+    {
+      "epoch": 0.7538397675383977,
+      "grad_norm": 0.9210042144448171,
+      "learning_rate": 9.119470967633767e-08,
+      "loss": 1.5207,
+      "step": 3632
+    },
+    {
+      "epoch": 0.7540473225404732,
+      "grad_norm": 0.7837425659381553,
+      "learning_rate": 9.111275138782288e-08,
+      "loss": 1.4977,
+      "step": 3633
+    },
+    {
+      "epoch": 0.7542548775425488,
+      "grad_norm": 0.8845803645583383,
+      "learning_rate": 9.103084789874439e-08,
+      "loss": 1.5369,
+      "step": 3634
+    },
+    {
+      "epoch": 0.7544624325446243,
+      "grad_norm": 0.7447126493612226,
+      "learning_rate": 9.09489992439255e-08,
+      "loss": 1.5491,
+      "step": 3635
+    },
+    {
+      "epoch": 0.7546699875466999,
+      "grad_norm": 0.7301421253401057,
+      "learning_rate": 9.086720545816603e-08,
+      "loss": 1.3985,
+      "step": 3636
+    },
+    {
+      "epoch": 0.7548775425487754,
+      "grad_norm": 0.8743236769031292,
+      "learning_rate": 9.078546657624251e-08,
+      "loss": 1.5653,
+      "step": 3637
+    },
+    {
+      "epoch": 0.755085097550851,
+      "grad_norm": 0.9437770186127424,
+      "learning_rate": 9.070378263290813e-08,
+      "loss": 1.5162,
+      "step": 3638
+    },
+    {
+      "epoch": 0.7552926525529265,
+      "grad_norm": 1.0029874017489346,
+      "learning_rate": 9.062215366289272e-08,
+      "loss": 1.4785,
+      "step": 3639
+    },
+    {
+      "epoch": 0.755500207555002,
+      "grad_norm": 1.0329217893835239,
+      "learning_rate": 9.054057970090291e-08,
+      "loss": 1.4984,
+      "step": 3640
+    },
+    {
+      "epoch": 0.7557077625570776,
+      "grad_norm": 0.9337780088755829,
+      "learning_rate": 9.045906078162154e-08,
+      "loss": 1.4553,
+      "step": 3641
+    },
+    {
+      "epoch": 0.7559153175591532,
+      "grad_norm": 3.12311257996803,
+      "learning_rate": 9.037759693970843e-08,
+      "loss": 1.5756,
+      "step": 3642
+    },
+    {
+      "epoch": 0.7561228725612287,
+      "grad_norm": 0.6561975771699147,
+      "learning_rate": 9.029618820979987e-08,
+      "loss": 1.5128,
+      "step": 3643
+    },
+    {
+      "epoch": 0.7563304275633043,
+      "grad_norm": 1.265134904146197,
+      "learning_rate": 9.02148346265086e-08,
+      "loss": 1.547,
+      "step": 3644
+    },
+    {
+      "epoch": 0.7565379825653799,
+      "grad_norm": 1.3010035821836963,
+      "learning_rate": 9.013353622442403e-08,
+      "loss": 1.4933,
+      "step": 3645
+    },
+    {
+      "epoch": 0.7567455375674553,
+      "grad_norm": 2.9548535906281033,
+      "learning_rate": 9.005229303811214e-08,
+      "loss": 1.4702,
+      "step": 3646
+    },
+    {
+      "epoch": 0.7569530925695309,
+      "grad_norm": 0.7062970208458317,
+      "learning_rate": 8.997110510211532e-08,
+      "loss": 1.4853,
+      "step": 3647
+    },
+    {
+      "epoch": 0.7571606475716065,
+      "grad_norm": 0.7453569534830744,
+      "learning_rate": 8.988997245095249e-08,
+      "loss": 1.4645,
+      "step": 3648
+    },
+    {
+      "epoch": 0.757368202573682,
+      "grad_norm": 0.9781340107487748,
+      "learning_rate": 8.980889511911929e-08,
+      "loss": 1.4859,
+      "step": 3649
+    },
+    {
+      "epoch": 0.7575757575757576,
+      "grad_norm": 0.9825116589942944,
+      "learning_rate": 8.972787314108736e-08,
+      "loss": 1.4699,
+      "step": 3650
+    },
+    {
+      "epoch": 0.7577833125778332,
+      "grad_norm": 0.6445189638333721,
+      "learning_rate": 8.964690655130533e-08,
+      "loss": 1.5362,
+      "step": 3651
+    },
+    {
+      "epoch": 0.7579908675799086,
+      "grad_norm": 0.7373274396811409,
+      "learning_rate": 8.956599538419791e-08,
+      "loss": 1.4911,
+      "step": 3652
+    },
+    {
+      "epoch": 0.7581984225819842,
+      "grad_norm": 1.1659684112711577,
+      "learning_rate": 8.948513967416648e-08,
+      "loss": 1.4624,
+      "step": 3653
+    },
+    {
+      "epoch": 0.7584059775840598,
+      "grad_norm": 0.9417157911081553,
+      "learning_rate": 8.940433945558858e-08,
+      "loss": 1.5196,
+      "step": 3654
+    },
+    {
+      "epoch": 0.7586135325861353,
+      "grad_norm": 0.7096003176473102,
+      "learning_rate": 8.932359476281857e-08,
+      "loss": 1.5523,
+      "step": 3655
+    },
+    {
+      "epoch": 0.7588210875882109,
+      "grad_norm": 0.9762969946562743,
+      "learning_rate": 8.924290563018669e-08,
+      "loss": 1.5188,
+      "step": 3656
+    },
+    {
+      "epoch": 0.7590286425902865,
+      "grad_norm": 0.9150895164904417,
+      "learning_rate": 8.9162272092e-08,
+      "loss": 1.5406,
+      "step": 3657
+    },
+    {
+      "epoch": 0.7592361975923619,
+      "grad_norm": 2.9040983958213284,
+      "learning_rate": 8.908169418254164e-08,
+      "loss": 1.5876,
+      "step": 3658
+    },
+    {
+      "epoch": 0.7594437525944375,
+      "grad_norm": 0.7344825206485066,
+      "learning_rate": 8.900117193607128e-08,
+      "loss": 1.504,
+      "step": 3659
+    },
+    {
+      "epoch": 0.7596513075965131,
+      "grad_norm": 0.7751599626545403,
+      "learning_rate": 8.89207053868247e-08,
+      "loss": 1.5777,
+      "step": 3660
+    },
+    {
+      "epoch": 0.7598588625985886,
+      "grad_norm": 0.6869096818130381,
+      "learning_rate": 8.884029456901428e-08,
+      "loss": 1.5168,
+      "step": 3661
+    },
+    {
+      "epoch": 0.7600664176006642,
+      "grad_norm": 0.9347348822598524,
+      "learning_rate": 8.875993951682855e-08,
+      "loss": 1.5377,
+      "step": 3662
+    },
+    {
+      "epoch": 0.7602739726027398,
+      "grad_norm": 0.91803327954358,
+      "learning_rate": 8.867964026443223e-08,
+      "loss": 1.5877,
+      "step": 3663
+    },
+    {
+      "epoch": 0.7604815276048152,
+      "grad_norm": 0.6701587500166564,
+      "learning_rate": 8.859939684596662e-08,
+      "loss": 1.4773,
+      "step": 3664
+    },
+    {
+      "epoch": 0.7606890826068908,
+      "grad_norm": 0.92745092959334,
+      "learning_rate": 8.851920929554888e-08,
+      "loss": 1.5237,
+      "step": 3665
+    },
+    {
+      "epoch": 0.7608966376089664,
+      "grad_norm": 0.762215020846363,
+      "learning_rate": 8.843907764727279e-08,
+      "loss": 1.4637,
+      "step": 3666
+    },
+    {
+      "epoch": 0.7611041926110419,
+      "grad_norm": 0.8446275863309926,
+      "learning_rate": 8.835900193520813e-08,
+      "loss": 1.5173,
+      "step": 3667
+    },
+    {
+      "epoch": 0.7613117476131175,
+      "grad_norm": 0.6371087787526062,
+      "learning_rate": 8.8278982193401e-08,
+      "loss": 1.4613,
+      "step": 3668
+    },
+    {
+      "epoch": 0.7615193026151931,
+      "grad_norm": 0.6543524756484076,
+      "learning_rate": 8.819901845587356e-08,
+      "loss": 1.4976,
+      "step": 3669
+    },
+    {
+      "epoch": 0.7617268576172685,
+      "grad_norm": 1.1075388492482394,
+      "learning_rate": 8.81191107566245e-08,
+      "loss": 1.5791,
+      "step": 3670
+    },
+    {
+      "epoch": 0.7619344126193441,
+      "grad_norm": 0.8167834355485803,
+      "learning_rate": 8.803925912962817e-08,
+      "loss": 1.5175,
+      "step": 3671
+    },
+    {
+      "epoch": 0.7621419676214197,
+      "grad_norm": 0.9714348238066599,
+      "learning_rate": 8.795946360883558e-08,
+      "loss": 1.5938,
+      "step": 3672
+    },
+    {
+      "epoch": 0.7623495226234952,
+      "grad_norm": 0.7166402771149134,
+      "learning_rate": 8.787972422817357e-08,
+      "loss": 1.4976,
+      "step": 3673
+    },
+    {
+      "epoch": 0.7625570776255708,
+      "grad_norm": 0.8369066453950968,
+      "learning_rate": 8.780004102154518e-08,
+      "loss": 1.4956,
+      "step": 3674
+    },
+    {
+      "epoch": 0.7627646326276464,
+      "grad_norm": 1.316099403142914,
+      "learning_rate": 8.772041402282968e-08,
+      "loss": 1.5624,
+      "step": 3675
+    },
+    {
+      "epoch": 0.7629721876297219,
+      "grad_norm": 1.719596367736537,
+      "learning_rate": 8.76408432658823e-08,
+      "loss": 1.4942,
+      "step": 3676
+    },
+    {
+      "epoch": 0.7631797426317974,
+      "grad_norm": 0.6515883407703427,
+      "learning_rate": 8.756132878453446e-08,
+      "loss": 1.4572,
+      "step": 3677
+    },
+    {
+      "epoch": 0.763387297633873,
+      "grad_norm": 0.7904696550358193,
+      "learning_rate": 8.748187061259352e-08,
+      "loss": 1.5274,
+      "step": 3678
+    },
+    {
+      "epoch": 0.7635948526359485,
+      "grad_norm": 0.711986518451615,
+      "learning_rate": 8.740246878384313e-08,
+      "loss": 1.5569,
+      "step": 3679
+    },
+    {
+      "epoch": 0.7638024076380241,
+      "grad_norm": 0.7103109837342821,
+      "learning_rate": 8.732312333204264e-08,
+      "loss": 1.5171,
+      "step": 3680
+    },
+    {
+      "epoch": 0.7640099626400996,
+      "grad_norm": 0.6564687552264464,
+      "learning_rate": 8.724383429092786e-08,
+      "loss": 1.5703,
+      "step": 3681
+    },
+    {
+      "epoch": 0.7642175176421752,
+      "grad_norm": 0.7106015365302101,
+      "learning_rate": 8.716460169421013e-08,
+      "loss": 1.455,
+      "step": 3682
+    },
+    {
+      "epoch": 0.7644250726442507,
+      "grad_norm": 0.6628386709799882,
+      "learning_rate": 8.708542557557725e-08,
+      "loss": 1.4875,
+      "step": 3683
+    },
+    {
+      "epoch": 0.7646326276463262,
+      "grad_norm": 0.9369132019515705,
+      "learning_rate": 8.700630596869274e-08,
+      "loss": 1.5483,
+      "step": 3684
+    },
+    {
+      "epoch": 0.7648401826484018,
+      "grad_norm": 0.9708001105081899,
+      "learning_rate": 8.692724290719615e-08,
+      "loss": 1.5552,
+      "step": 3685
+    },
+    {
+      "epoch": 0.7650477376504774,
+      "grad_norm": 0.6799600387927957,
+      "learning_rate": 8.684823642470294e-08,
+      "loss": 1.464,
+      "step": 3686
+    },
+    {
+      "epoch": 0.7652552926525529,
+      "grad_norm": 0.9400216422936265,
+      "learning_rate": 8.676928655480467e-08,
+      "loss": 1.4919,
+      "step": 3687
+    },
+    {
+      "epoch": 0.7654628476546285,
+      "grad_norm": 0.6616351345272905,
+      "learning_rate": 8.669039333106869e-08,
+      "loss": 1.5028,
+      "step": 3688
+    },
+    {
+      "epoch": 0.765670402656704,
+      "grad_norm": 1.0346354200029755,
+      "learning_rate": 8.661155678703824e-08,
+      "loss": 1.4399,
+      "step": 3689
+    },
+    {
+      "epoch": 0.7658779576587795,
+      "grad_norm": 1.1564937823973798,
+      "learning_rate": 8.653277695623269e-08,
+      "loss": 1.5234,
+      "step": 3690
+    },
+    {
+      "epoch": 0.7660855126608551,
+      "grad_norm": 0.9097324875657153,
+      "learning_rate": 8.645405387214691e-08,
+      "loss": 1.5455,
+      "step": 3691
+    },
+    {
+      "epoch": 0.7662930676629307,
+      "grad_norm": 0.8511477824718381,
+      "learning_rate": 8.637538756825204e-08,
+      "loss": 1.5288,
+      "step": 3692
+    },
+    {
+      "epoch": 0.7665006226650062,
+      "grad_norm": 0.7056324365479921,
+      "learning_rate": 8.629677807799484e-08,
+      "loss": 1.4495,
+      "step": 3693
+    },
+    {
+      "epoch": 0.7667081776670818,
+      "grad_norm": 0.8962247940387371,
+      "learning_rate": 8.6218225434798e-08,
+      "loss": 1.4997,
+      "step": 3694
+    },
+    {
+      "epoch": 0.7669157326691574,
+      "grad_norm": 0.9727264386207618,
+      "learning_rate": 8.613972967205993e-08,
+      "loss": 1.4703,
+      "step": 3695
+    },
+    {
+      "epoch": 0.7671232876712328,
+      "grad_norm": 1.2402968049705936,
+      "learning_rate": 8.606129082315514e-08,
+      "loss": 1.4662,
+      "step": 3696
+    },
+    {
+      "epoch": 0.7673308426733084,
+      "grad_norm": 0.9480454389682683,
+      "learning_rate": 8.59829089214335e-08,
+      "loss": 1.5176,
+      "step": 3697
+    },
+    {
+      "epoch": 0.767538397675384,
+      "grad_norm": 0.7791763063083721,
+      "learning_rate": 8.590458400022109e-08,
+      "loss": 1.562,
+      "step": 3698
+    },
+    {
+      "epoch": 0.7677459526774595,
+      "grad_norm": 0.8795455326700431,
+      "learning_rate": 8.582631609281954e-08,
+      "loss": 1.6003,
+      "step": 3699
+    },
+    {
+      "epoch": 0.7679535076795351,
+      "grad_norm": 0.9303963120593105,
+      "learning_rate": 8.574810523250622e-08,
+      "loss": 1.4352,
+      "step": 3700
+    },
+    {
+      "epoch": 0.7681610626816107,
+      "grad_norm": 0.745958652852297,
+      "learning_rate": 8.56699514525344e-08,
+      "loss": 1.5844,
+      "step": 3701
+    },
+    {
+      "epoch": 0.7683686176836861,
+      "grad_norm": 1.5248799469735759,
+      "learning_rate": 8.559185478613299e-08,
+      "loss": 1.5652,
+      "step": 3702
+    },
+    {
+      "epoch": 0.7685761726857617,
+      "grad_norm": 0.9504135251306799,
+      "learning_rate": 8.551381526650658e-08,
+      "loss": 1.5243,
+      "step": 3703
+    },
+    {
+      "epoch": 0.7687837276878373,
+      "grad_norm": 0.6910282403052183,
+      "learning_rate": 8.543583292683547e-08,
+      "loss": 1.5383,
+      "step": 3704
+    },
+    {
+      "epoch": 0.7689912826899128,
+      "grad_norm": 0.6499442373461638,
+      "learning_rate": 8.535790780027582e-08,
+      "loss": 1.5775,
+      "step": 3705
+    },
+    {
+      "epoch": 0.7691988376919884,
+      "grad_norm": 0.9535545726573528,
+      "learning_rate": 8.528003991995914e-08,
+      "loss": 1.5118,
+      "step": 3706
+    },
+    {
+      "epoch": 0.769406392694064,
+      "grad_norm": 0.9369225524492816,
+      "learning_rate": 8.520222931899292e-08,
+      "loss": 1.4968,
+      "step": 3707
+    },
+    {
+      "epoch": 0.7696139476961394,
+      "grad_norm": 1.4228802776735283,
+      "learning_rate": 8.512447603046011e-08,
+      "loss": 1.5087,
+      "step": 3708
+    },
+    {
+      "epoch": 0.769821502698215,
+      "grad_norm": 0.7752328550439567,
+      "learning_rate": 8.504678008741936e-08,
+      "loss": 1.4768,
+      "step": 3709
+    },
+    {
+      "epoch": 0.7700290577002906,
+      "grad_norm": 1.0805202104476346,
+      "learning_rate": 8.496914152290485e-08,
+      "loss": 1.5587,
+      "step": 3710
+    },
+    {
+      "epoch": 0.7702366127023661,
+      "grad_norm": 0.6949638840831297,
+      "learning_rate": 8.489156036992655e-08,
+      "loss": 1.4693,
+      "step": 3711
+    },
+    {
+      "epoch": 0.7704441677044417,
+      "grad_norm": 1.4180052788818878,
+      "learning_rate": 8.481403666146987e-08,
+      "loss": 1.5181,
+      "step": 3712
+    },
+    {
+      "epoch": 0.7706517227065173,
+      "grad_norm": 0.7642757731984444,
+      "learning_rate": 8.47365704304958e-08,
+      "loss": 1.4588,
+      "step": 3713
+    },
+    {
+      "epoch": 0.7708592777085927,
+      "grad_norm": 2.261112460955011,
+      "learning_rate": 8.465916170994094e-08,
+      "loss": 1.5696,
+      "step": 3714
+    },
+    {
+      "epoch": 0.7710668327106683,
+      "grad_norm": 0.9840937688868856,
+      "learning_rate": 8.45818105327174e-08,
+      "loss": 1.5552,
+      "step": 3715
+    },
+    {
+      "epoch": 0.7712743877127439,
+      "grad_norm": 0.9044154791431672,
+      "learning_rate": 8.45045169317129e-08,
+      "loss": 1.4646,
+      "step": 3716
+    },
+    {
+      "epoch": 0.7714819427148194,
+      "grad_norm": 0.7233508299839303,
+      "learning_rate": 8.442728093979061e-08,
+      "loss": 1.5773,
+      "step": 3717
+    },
+    {
+      "epoch": 0.771689497716895,
+      "grad_norm": 0.6547539358787329,
+      "learning_rate": 8.435010258978922e-08,
+      "loss": 1.5136,
+      "step": 3718
+    },
+    {
+      "epoch": 0.7718970527189706,
+      "grad_norm": 0.8478591806594814,
+      "learning_rate": 8.427298191452285e-08,
+      "loss": 1.5112,
+      "step": 3719
+    },
+    {
+      "epoch": 0.772104607721046,
+      "grad_norm": 0.8263975586927169,
+      "learning_rate": 8.419591894678134e-08,
+      "loss": 1.5382,
+      "step": 3720
+    },
+    {
+      "epoch": 0.7723121627231216,
+      "grad_norm": 0.8752860535749272,
+      "learning_rate": 8.411891371932958e-08,
+      "loss": 1.4869,
+      "step": 3721
+    },
+    {
+      "epoch": 0.7725197177251972,
+      "grad_norm": 0.7368123606567933,
+      "learning_rate": 8.404196626490831e-08,
+      "loss": 1.5335,
+      "step": 3722
+    },
+    {
+      "epoch": 0.7727272727272727,
+      "grad_norm": 2.2823409655822764,
+      "learning_rate": 8.396507661623354e-08,
+      "loss": 1.5371,
+      "step": 3723
+    },
+    {
+      "epoch": 0.7729348277293483,
+      "grad_norm": 0.8785549336127969,
+      "learning_rate": 8.388824480599664e-08,
+      "loss": 1.5247,
+      "step": 3724
+    },
+    {
+      "epoch": 0.7731423827314239,
+      "grad_norm": 0.6689860470059837,
+      "learning_rate": 8.381147086686444e-08,
+      "loss": 1.4838,
+      "step": 3725
+    },
+    {
+      "epoch": 0.7733499377334994,
+      "grad_norm": 0.6419935459006911,
+      "learning_rate": 8.373475483147929e-08,
+      "loss": 1.5401,
+      "step": 3726
+    },
+    {
+      "epoch": 0.773557492735575,
+      "grad_norm": 0.6435875932662984,
+      "learning_rate": 8.365809673245872e-08,
+      "loss": 1.458,
+      "step": 3727
+    },
+    {
+      "epoch": 0.7737650477376504,
+      "grad_norm": 1.094980297827521,
+      "learning_rate": 8.358149660239578e-08,
+      "loss": 1.512,
+      "step": 3728
+    },
+    {
+      "epoch": 0.773972602739726,
+      "grad_norm": 1.279880254885234,
+      "learning_rate": 8.350495447385878e-08,
+      "loss": 1.5499,
+      "step": 3729
+    },
+    {
+      "epoch": 0.7741801577418016,
+      "grad_norm": 0.7391367560584082,
+      "learning_rate": 8.342847037939133e-08,
+      "loss": 1.5163,
+      "step": 3730
+    },
+    {
+      "epoch": 0.7743877127438771,
+      "grad_norm": 0.8069549910756837,
+      "learning_rate": 8.335204435151262e-08,
+      "loss": 1.4291,
+      "step": 3731
+    },
+    {
+      "epoch": 0.7745952677459527,
+      "grad_norm": 1.3074851644422694,
+      "learning_rate": 8.327567642271676e-08,
+      "loss": 1.5477,
+      "step": 3732
+    },
+    {
+      "epoch": 0.7748028227480283,
+      "grad_norm": 1.4786036555935946,
+      "learning_rate": 8.319936662547349e-08,
+      "loss": 1.5801,
+      "step": 3733
+    },
+    {
+      "epoch": 0.7750103777501037,
+      "grad_norm": 0.8115404428015045,
+      "learning_rate": 8.31231149922277e-08,
+      "loss": 1.498,
+      "step": 3734
+    },
+    {
+      "epoch": 0.7752179327521793,
+      "grad_norm": 0.7048538488265006,
+      "learning_rate": 8.304692155539952e-08,
+      "loss": 1.486,
+      "step": 3735
+    },
+    {
+      "epoch": 0.7754254877542549,
+      "grad_norm": 1.3901521643293466,
+      "learning_rate": 8.297078634738436e-08,
+      "loss": 1.5362,
+      "step": 3736
+    },
+    {
+      "epoch": 0.7756330427563304,
+      "grad_norm": 0.7722605542096388,
+      "learning_rate": 8.289470940055297e-08,
+      "loss": 1.5278,
+      "step": 3737
+    },
+    {
+      "epoch": 0.775840597758406,
+      "grad_norm": 0.9071026501168962,
+      "learning_rate": 8.28186907472512e-08,
+      "loss": 1.5438,
+      "step": 3738
+    },
+    {
+      "epoch": 0.7760481527604816,
+      "grad_norm": 0.7428620731943374,
+      "learning_rate": 8.27427304198002e-08,
+      "loss": 1.489,
+      "step": 3739
+    },
+    {
+      "epoch": 0.776255707762557,
+      "grad_norm": 0.6993922725125539,
+      "learning_rate": 8.266682845049621e-08,
+      "loss": 1.5453,
+      "step": 3740
+    },
+    {
+      "epoch": 0.7764632627646326,
+      "grad_norm": 0.8349717976196509,
+      "learning_rate": 8.259098487161076e-08,
+      "loss": 1.5133,
+      "step": 3741
+    },
+    {
+      "epoch": 0.7766708177667082,
+      "grad_norm": 0.6512627588323215,
+      "learning_rate": 8.251519971539057e-08,
+      "loss": 1.5113,
+      "step": 3742
+    },
+    {
+      "epoch": 0.7768783727687837,
+      "grad_norm": 0.7237646398392438,
+      "learning_rate": 8.243947301405745e-08,
+      "loss": 1.5484,
+      "step": 3743
+    },
+    {
+      "epoch": 0.7770859277708593,
+      "grad_norm": 0.6834301678937142,
+      "learning_rate": 8.236380479980838e-08,
+      "loss": 1.5154,
+      "step": 3744
+    },
+    {
+      "epoch": 0.7772934827729349,
+      "grad_norm": 2.676292870574534,
+      "learning_rate": 8.228819510481544e-08,
+      "loss": 1.4527,
+      "step": 3745
+    },
+    {
+      "epoch": 0.7775010377750103,
+      "grad_norm": 0.7187387132947246,
+      "learning_rate": 8.221264396122598e-08,
+      "loss": 1.5916,
+      "step": 3746
+    },
+    {
+      "epoch": 0.7777085927770859,
+      "grad_norm": 0.9796527956740613,
+      "learning_rate": 8.213715140116217e-08,
+      "loss": 1.5333,
+      "step": 3747
+    },
+    {
+      "epoch": 0.7779161477791615,
+      "grad_norm": 0.7220937755834875,
+      "learning_rate": 8.20617174567216e-08,
+      "loss": 1.5335,
+      "step": 3748
+    },
+    {
+      "epoch": 0.778123702781237,
+      "grad_norm": 0.7611049438253926,
+      "learning_rate": 8.198634215997669e-08,
+      "loss": 1.5446,
+      "step": 3749
+    },
+    {
+      "epoch": 0.7783312577833126,
+      "grad_norm": 0.7040971316583947,
+      "learning_rate": 8.191102554297505e-08,
+      "loss": 1.5059,
+      "step": 3750
+    },
+    {
+      "epoch": 0.7785388127853882,
+      "grad_norm": 0.8601770693392301,
+      "learning_rate": 8.183576763773925e-08,
+      "loss": 1.4694,
+      "step": 3751
+    },
+    {
+      "epoch": 0.7787463677874636,
+      "grad_norm": 0.8470768345754324,
+      "learning_rate": 8.176056847626705e-08,
+      "loss": 1.5573,
+      "step": 3752
+    },
+    {
+      "epoch": 0.7789539227895392,
+      "grad_norm": 0.7332075382870787,
+      "learning_rate": 8.168542809053108e-08,
+      "loss": 1.5627,
+      "step": 3753
+    },
+    {
+      "epoch": 0.7791614777916148,
+      "grad_norm": 0.639110979242346,
+      "learning_rate": 8.161034651247895e-08,
+      "loss": 1.4819,
+      "step": 3754
+    },
+    {
+      "epoch": 0.7793690327936903,
+      "grad_norm": 0.7283092367073781,
+      "learning_rate": 8.15353237740336e-08,
+      "loss": 1.5144,
+      "step": 3755
+    },
+    {
+      "epoch": 0.7795765877957659,
+      "grad_norm": 0.895981108138558,
+      "learning_rate": 8.146035990709246e-08,
+      "loss": 1.5767,
+      "step": 3756
+    },
+    {
+      "epoch": 0.7797841427978415,
+      "grad_norm": 0.6467164460809052,
+      "learning_rate": 8.138545494352828e-08,
+      "loss": 1.5692,
+      "step": 3757
+    },
+    {
+      "epoch": 0.779991697799917,
+      "grad_norm": 0.7404725669024991,
+      "learning_rate": 8.131060891518869e-08,
+      "loss": 1.4578,
+      "step": 3758
+    },
+    {
+      "epoch": 0.7801992528019925,
+      "grad_norm": 0.6595972029259889,
+      "learning_rate": 8.123582185389616e-08,
+      "loss": 1.5813,
+      "step": 3759
+    },
+    {
+      "epoch": 0.7804068078040681,
+      "grad_norm": 0.7221318755264007,
+      "learning_rate": 8.116109379144817e-08,
+      "loss": 1.5387,
+      "step": 3760
+    },
+    {
+      "epoch": 0.7806143628061436,
+      "grad_norm": 0.6215279182105392,
+      "learning_rate": 8.108642475961725e-08,
+      "loss": 1.5125,
+      "step": 3761
+    },
+    {
+      "epoch": 0.7808219178082192,
+      "grad_norm": 0.67488476820223,
+      "learning_rate": 8.101181479015043e-08,
+      "loss": 1.4738,
+      "step": 3762
+    },
+    {
+      "epoch": 0.7810294728102948,
+      "grad_norm": 1.2130955024336798,
+      "learning_rate": 8.093726391477011e-08,
+      "loss": 1.4904,
+      "step": 3763
+    },
+    {
+      "epoch": 0.7812370278123703,
+      "grad_norm": 0.7563625260734779,
+      "learning_rate": 8.086277216517327e-08,
+      "loss": 1.5131,
+      "step": 3764
+    },
+    {
+      "epoch": 0.7814445828144458,
+      "grad_norm": 1.5034617378440747,
+      "learning_rate": 8.078833957303184e-08,
+      "loss": 1.4702,
+      "step": 3765
+    },
+    {
+      "epoch": 0.7816521378165214,
+      "grad_norm": 0.6873898163346628,
+      "learning_rate": 8.071396616999248e-08,
+      "loss": 1.4982,
+      "step": 3766
+    },
+    {
+      "epoch": 0.7818596928185969,
+      "grad_norm": 1.4986158607595441,
+      "learning_rate": 8.063965198767692e-08,
+      "loss": 1.4948,
+      "step": 3767
+    },
+    {
+      "epoch": 0.7820672478206725,
+      "grad_norm": 0.726055707165679,
+      "learning_rate": 8.056539705768155e-08,
+      "loss": 1.5342,
+      "step": 3768
+    },
+    {
+      "epoch": 0.7822748028227481,
+      "grad_norm": 0.8321359651169101,
+      "learning_rate": 8.049120141157752e-08,
+      "loss": 1.5281,
+      "step": 3769
+    },
+    {
+      "epoch": 0.7824823578248236,
+      "grad_norm": 0.669062632940801,
+      "learning_rate": 8.041706508091102e-08,
+      "loss": 1.5371,
+      "step": 3770
+    },
+    {
+      "epoch": 0.7826899128268991,
+      "grad_norm": 0.7072107789837956,
+      "learning_rate": 8.034298809720259e-08,
+      "loss": 1.4802,
+      "step": 3771
+    },
+    {
+      "epoch": 0.7828974678289746,
+      "grad_norm": 0.9694839930411056,
+      "learning_rate": 8.026897049194805e-08,
+      "loss": 1.5827,
+      "step": 3772
+    },
+    {
+      "epoch": 0.7831050228310502,
+      "grad_norm": 2.2323015066991014,
+      "learning_rate": 8.019501229661753e-08,
+      "loss": 1.5571,
+      "step": 3773
+    },
+    {
+      "epoch": 0.7833125778331258,
+      "grad_norm": 0.8829954343640335,
+      "learning_rate": 8.01211135426562e-08,
+      "loss": 1.513,
+      "step": 3774
+    },
+    {
+      "epoch": 0.7835201328352013,
+      "grad_norm": 0.762528386982982,
+      "learning_rate": 8.004727426148384e-08,
+      "loss": 1.427,
+      "step": 3775
+    },
+    {
+      "epoch": 0.7837276878372769,
+      "grad_norm": 1.5006911856192857,
+      "learning_rate": 7.997349448449491e-08,
+      "loss": 1.5287,
+      "step": 3776
+    },
+    {
+      "epoch": 0.7839352428393525,
+      "grad_norm": 0.7332664802109321,
+      "learning_rate": 7.989977424305859e-08,
+      "loss": 1.4988,
+      "step": 3777
+    },
+    {
+      "epoch": 0.7841427978414279,
+      "grad_norm": 0.6956308212824712,
+      "learning_rate": 7.982611356851887e-08,
+      "loss": 1.596,
+      "step": 3778
+    },
+    {
+      "epoch": 0.7843503528435035,
+      "grad_norm": 1.3954923716870664,
+      "learning_rate": 7.975251249219424e-08,
+      "loss": 1.4597,
+      "step": 3779
+    },
+    {
+      "epoch": 0.7845579078455791,
+      "grad_norm": 0.828086102970035,
+      "learning_rate": 7.967897104537788e-08,
+      "loss": 1.5027,
+      "step": 3780
+    },
+    {
+      "epoch": 0.7847654628476546,
+      "grad_norm": 0.8836328427525264,
+      "learning_rate": 7.960548925933786e-08,
+      "loss": 1.5138,
+      "step": 3781
+    },
+    {
+      "epoch": 0.7849730178497302,
+      "grad_norm": 1.0363897182174382,
+      "learning_rate": 7.953206716531639e-08,
+      "loss": 1.5149,
+      "step": 3782
+    },
+    {
+      "epoch": 0.7851805728518058,
+      "grad_norm": 0.7921597815198995,
+      "learning_rate": 7.945870479453084e-08,
+      "loss": 1.5505,
+      "step": 3783
+    },
+    {
+      "epoch": 0.7853881278538812,
+      "grad_norm": 2.0687899270821166,
+      "learning_rate": 7.938540217817285e-08,
+      "loss": 1.5512,
+      "step": 3784
+    },
+    {
+      "epoch": 0.7855956828559568,
+      "grad_norm": 0.7272525902992321,
+      "learning_rate": 7.931215934740873e-08,
+      "loss": 1.5294,
+      "step": 3785
+    },
+    {
+      "epoch": 0.7858032378580324,
+      "grad_norm": 0.8660182828580697,
+      "learning_rate": 7.923897633337939e-08,
+      "loss": 1.4758,
+      "step": 3786
+    },
+    {
+      "epoch": 0.7860107928601079,
+      "grad_norm": 1.2755352483770819,
+      "learning_rate": 7.916585316720039e-08,
+      "loss": 1.4728,
+      "step": 3787
+    },
+    {
+      "epoch": 0.7862183478621835,
+      "grad_norm": 0.8623628404181108,
+      "learning_rate": 7.909278987996157e-08,
+      "loss": 1.5123,
+      "step": 3788
+    },
+    {
+      "epoch": 0.7864259028642591,
+      "grad_norm": 0.7458981809530579,
+      "learning_rate": 7.901978650272772e-08,
+      "loss": 1.5909,
+      "step": 3789
+    },
+    {
+      "epoch": 0.7866334578663345,
+      "grad_norm": 0.6640707473838863,
+      "learning_rate": 7.894684306653781e-08,
+      "loss": 1.5525,
+      "step": 3790
+    },
+    {
+      "epoch": 0.7868410128684101,
+      "grad_norm": 0.8683760806748698,
+      "learning_rate": 7.887395960240548e-08,
+      "loss": 1.5188,
+      "step": 3791
+    },
+    {
+      "epoch": 0.7870485678704857,
+      "grad_norm": 0.9084728902160711,
+      "learning_rate": 7.88011361413188e-08,
+      "loss": 1.4716,
+      "step": 3792
+    },
+    {
+      "epoch": 0.7872561228725612,
+      "grad_norm": 0.8931162535612022,
+      "learning_rate": 7.872837271424044e-08,
+      "loss": 1.5544,
+      "step": 3793
+    },
+    {
+      "epoch": 0.7874636778746368,
+      "grad_norm": 0.6460004877130135,
+      "learning_rate": 7.86556693521075e-08,
+      "loss": 1.4741,
+      "step": 3794
+    },
+    {
+      "epoch": 0.7876712328767124,
+      "grad_norm": 0.7034857915995364,
+      "learning_rate": 7.858302608583138e-08,
+      "loss": 1.4803,
+      "step": 3795
+    },
+    {
+      "epoch": 0.7878787878787878,
+      "grad_norm": 0.6623982003852003,
+      "learning_rate": 7.85104429462983e-08,
+      "loss": 1.5303,
+      "step": 3796
+    },
+    {
+      "epoch": 0.7880863428808634,
+      "grad_norm": 1.1133133388108711,
+      "learning_rate": 7.843791996436841e-08,
+      "loss": 1.5157,
+      "step": 3797
+    },
+    {
+      "epoch": 0.788293897882939,
+      "grad_norm": 1.1949157436516982,
+      "learning_rate": 7.836545717087675e-08,
+      "loss": 1.5005,
+      "step": 3798
+    },
+    {
+      "epoch": 0.7885014528850145,
+      "grad_norm": 1.3589645995798711,
+      "learning_rate": 7.829305459663253e-08,
+      "loss": 1.4688,
+      "step": 3799
+    },
+    {
+      "epoch": 0.7887090078870901,
+      "grad_norm": 1.1256971724652678,
+      "learning_rate": 7.82207122724194e-08,
+      "loss": 1.5563,
+      "step": 3800
+    },
+    {
+      "epoch": 0.7889165628891657,
+      "grad_norm": 0.6952310962030432,
+      "learning_rate": 7.814843022899531e-08,
+      "loss": 1.5187,
+      "step": 3801
+    },
+    {
+      "epoch": 0.7891241178912412,
+      "grad_norm": 0.9310453197307528,
+      "learning_rate": 7.807620849709286e-08,
+      "loss": 1.5251,
+      "step": 3802
+    },
+    {
+      "epoch": 0.7893316728933167,
+      "grad_norm": 1.000908443559828,
+      "learning_rate": 7.800404710741857e-08,
+      "loss": 1.5833,
+      "step": 3803
+    },
+    {
+      "epoch": 0.7895392278953923,
+      "grad_norm": 0.8475468738785007,
+      "learning_rate": 7.793194609065373e-08,
+      "loss": 1.4907,
+      "step": 3804
+    },
+    {
+      "epoch": 0.7897467828974678,
+      "grad_norm": 0.7078086153324371,
+      "learning_rate": 7.785990547745374e-08,
+      "loss": 1.4704,
+      "step": 3805
+    },
+    {
+      "epoch": 0.7899543378995434,
+      "grad_norm": 0.7119022791600783,
+      "learning_rate": 7.778792529844826e-08,
+      "loss": 1.5093,
+      "step": 3806
+    },
+    {
+      "epoch": 0.790161892901619,
+      "grad_norm": 0.6738375791146888,
+      "learning_rate": 7.771600558424152e-08,
+      "loss": 1.5654,
+      "step": 3807
+    },
+    {
+      "epoch": 0.7903694479036945,
+      "grad_norm": 1.1868599223865224,
+      "learning_rate": 7.764414636541175e-08,
+      "loss": 1.4883,
+      "step": 3808
+    },
+    {
+      "epoch": 0.79057700290577,
+      "grad_norm": 0.6224598939070389,
+      "learning_rate": 7.757234767251159e-08,
+      "loss": 1.5634,
+      "step": 3809
+    },
+    {
+      "epoch": 0.7907845579078456,
+      "grad_norm": 1.4109504168100024,
+      "learning_rate": 7.750060953606795e-08,
+      "loss": 1.5388,
+      "step": 3810
+    },
+    {
+      "epoch": 0.7909921129099211,
+      "grad_norm": 0.750672852855631,
+      "learning_rate": 7.742893198658207e-08,
+      "loss": 1.5594,
+      "step": 3811
+    },
+    {
+      "epoch": 0.7911996679119967,
+      "grad_norm": 0.9601149668161083,
+      "learning_rate": 7.735731505452916e-08,
+      "loss": 1.5171,
+      "step": 3812
+    },
+    {
+      "epoch": 0.7914072229140723,
+      "grad_norm": 0.6934940716436475,
+      "learning_rate": 7.7285758770359e-08,
+      "loss": 1.5707,
+      "step": 3813
+    },
+    {
+      "epoch": 0.7916147779161478,
+      "grad_norm": 0.9586831429557638,
+      "learning_rate": 7.721426316449538e-08,
+      "loss": 1.503,
+      "step": 3814
+    },
+    {
+      "epoch": 0.7918223329182233,
+      "grad_norm": 0.727117044470924,
+      "learning_rate": 7.714282826733627e-08,
+      "loss": 1.573,
+      "step": 3815
+    },
+    {
+      "epoch": 0.7920298879202988,
+      "grad_norm": 0.7258883514207821,
+      "learning_rate": 7.707145410925397e-08,
+      "loss": 1.5284,
+      "step": 3816
+    },
+    {
+      "epoch": 0.7922374429223744,
+      "grad_norm": 12.323516611009495,
+      "learning_rate": 7.70001407205948e-08,
+      "loss": 1.4886,
+      "step": 3817
+    },
+    {
+      "epoch": 0.79244499792445,
+      "grad_norm": 0.6711876057145292,
+      "learning_rate": 7.692888813167942e-08,
+      "loss": 1.5204,
+      "step": 3818
+    },
+    {
+      "epoch": 0.7926525529265255,
+      "grad_norm": 1.0533549166405607,
+      "learning_rate": 7.685769637280246e-08,
+      "loss": 1.4556,
+      "step": 3819
+    },
+    {
+      "epoch": 0.7928601079286011,
+      "grad_norm": 0.972836022817293,
+      "learning_rate": 7.678656547423282e-08,
+      "loss": 1.59,
+      "step": 3820
+    },
+    {
+      "epoch": 0.7930676629306767,
+      "grad_norm": 0.767392440444129,
+      "learning_rate": 7.671549546621337e-08,
+      "loss": 1.4897,
+      "step": 3821
+    },
+    {
+      "epoch": 0.7932752179327521,
+      "grad_norm": 0.7178476520542515,
+      "learning_rate": 7.664448637896135e-08,
+      "loss": 1.4294,
+      "step": 3822
+    },
+    {
+      "epoch": 0.7934827729348277,
+      "grad_norm": 0.796196181175966,
+      "learning_rate": 7.657353824266777e-08,
+      "loss": 1.5647,
+      "step": 3823
+    },
+    {
+      "epoch": 0.7936903279369033,
+      "grad_norm": 0.8611464441889768,
+      "learning_rate": 7.6502651087498e-08,
+      "loss": 1.519,
+      "step": 3824
+    },
+    {
+      "epoch": 0.7938978829389788,
+      "grad_norm": 0.788630629261394,
+      "learning_rate": 7.643182494359137e-08,
+      "loss": 1.6052,
+      "step": 3825
+    },
+    {
+      "epoch": 0.7941054379410544,
+      "grad_norm": 0.6879220984840951,
+      "learning_rate": 7.636105984106125e-08,
+      "loss": 1.4583,
+      "step": 3826
+    },
+    {
+      "epoch": 0.79431299294313,
+      "grad_norm": 0.7485399682769517,
+      "learning_rate": 7.629035580999504e-08,
+      "loss": 1.5529,
+      "step": 3827
+    },
+    {
+      "epoch": 0.7945205479452054,
+      "grad_norm": 0.9942223646125695,
+      "learning_rate": 7.621971288045436e-08,
+      "loss": 1.5415,
+      "step": 3828
+    },
+    {
+      "epoch": 0.794728102947281,
+      "grad_norm": 0.6681319118955458,
+      "learning_rate": 7.614913108247451e-08,
+      "loss": 1.4598,
+      "step": 3829
+    },
+    {
+      "epoch": 0.7949356579493566,
+      "grad_norm": 1.1967036992304174,
+      "learning_rate": 7.607861044606516e-08,
+      "loss": 1.467,
+      "step": 3830
+    },
+    {
+      "epoch": 0.7951432129514321,
+      "grad_norm": 1.6029952828306122,
+      "learning_rate": 7.600815100120977e-08,
+      "loss": 1.489,
+      "step": 3831
+    },
+    {
+      "epoch": 0.7953507679535077,
+      "grad_norm": 0.7758285658518153,
+      "learning_rate": 7.593775277786572e-08,
+      "loss": 1.5299,
+      "step": 3832
+    },
+    {
+      "epoch": 0.7955583229555833,
+      "grad_norm": 1.236228425852546,
+      "learning_rate": 7.586741580596464e-08,
+      "loss": 1.4974,
+      "step": 3833
+    },
+    {
+      "epoch": 0.7957658779576587,
+      "grad_norm": 0.8086451800153249,
+      "learning_rate": 7.579714011541183e-08,
+      "loss": 1.5466,
+      "step": 3834
+    },
+    {
+      "epoch": 0.7959734329597343,
+      "grad_norm": 1.9761647944911873,
+      "learning_rate": 7.572692573608667e-08,
+      "loss": 1.5511,
+      "step": 3835
+    },
+    {
+      "epoch": 0.7961809879618099,
+      "grad_norm": 0.7472347685052534,
+      "learning_rate": 7.565677269784243e-08,
+      "loss": 1.499,
+      "step": 3836
+    },
+    {
+      "epoch": 0.7963885429638854,
+      "grad_norm": 0.7078437594225262,
+      "learning_rate": 7.558668103050643e-08,
+      "loss": 1.4638,
+      "step": 3837
+    },
+    {
+      "epoch": 0.796596097965961,
+      "grad_norm": 0.671621701620675,
+      "learning_rate": 7.55166507638796e-08,
+      "loss": 1.4281,
+      "step": 3838
+    },
+    {
+      "epoch": 0.7968036529680366,
+      "grad_norm": 0.6390390284844484,
+      "learning_rate": 7.544668192773712e-08,
+      "loss": 1.5204,
+      "step": 3839
+    },
+    {
+      "epoch": 0.797011207970112,
+      "grad_norm": 1.104077219766796,
+      "learning_rate": 7.53767745518278e-08,
+      "loss": 1.5223,
+      "step": 3840
+    },
+    {
+      "epoch": 0.7972187629721876,
+      "grad_norm": 0.9119077366866154,
+      "learning_rate": 7.53069286658744e-08,
+      "loss": 1.4567,
+      "step": 3841
+    },
+    {
+      "epoch": 0.7974263179742632,
+      "grad_norm": 0.9714383685205503,
+      "learning_rate": 7.523714429957351e-08,
+      "loss": 1.5393,
+      "step": 3842
+    },
+    {
+      "epoch": 0.7976338729763387,
+      "grad_norm": 0.8706994516262466,
+      "learning_rate": 7.516742148259568e-08,
+      "loss": 1.52,
+      "step": 3843
+    },
+    {
+      "epoch": 0.7978414279784143,
+      "grad_norm": 0.8319415582822318,
+      "learning_rate": 7.509776024458514e-08,
+      "loss": 1.5033,
+      "step": 3844
+    },
+    {
+      "epoch": 0.7980489829804899,
+      "grad_norm": 0.6780742323181039,
+      "learning_rate": 7.502816061516002e-08,
+      "loss": 1.4795,
+      "step": 3845
+    },
+    {
+      "epoch": 0.7982565379825654,
+      "grad_norm": 0.8182092940025123,
+      "learning_rate": 7.495862262391222e-08,
+      "loss": 1.4866,
+      "step": 3846
+    },
+    {
+      "epoch": 0.7984640929846409,
+      "grad_norm": 0.7862229344316162,
+      "learning_rate": 7.488914630040737e-08,
+      "loss": 1.4307,
+      "step": 3847
+    },
+    {
+      "epoch": 0.7986716479867165,
+      "grad_norm": 1.1802221959396697,
+      "learning_rate": 7.481973167418512e-08,
+      "loss": 1.6294,
+      "step": 3848
+    },
+    {
+      "epoch": 0.798879202988792,
+      "grad_norm": 0.803247594134049,
+      "learning_rate": 7.475037877475863e-08,
+      "loss": 1.5033,
+      "step": 3849
+    },
+    {
+      "epoch": 0.7990867579908676,
+      "grad_norm": 1.510249894192323,
+      "learning_rate": 7.468108763161495e-08,
+      "loss": 1.5315,
+      "step": 3850
+    },
+    {
+      "epoch": 0.7992943129929432,
+      "grad_norm": 0.6407938632163084,
+      "learning_rate": 7.461185827421475e-08,
+      "loss": 1.4926,
+      "step": 3851
+    },
+    {
+      "epoch": 0.7995018679950187,
+      "grad_norm": 0.6622262503774939,
+      "learning_rate": 7.454269073199267e-08,
+      "loss": 1.4781,
+      "step": 3852
+    },
+    {
+      "epoch": 0.7997094229970942,
+      "grad_norm": 1.0209070709245331,
+      "learning_rate": 7.447358503435673e-08,
+      "loss": 1.5373,
+      "step": 3853
+    },
+    {
+      "epoch": 0.7999169779991698,
+      "grad_norm": 0.6329284519458666,
+      "learning_rate": 7.440454121068895e-08,
+      "loss": 1.5231,
+      "step": 3854
+    },
+    {
+      "epoch": 0.8001245330012453,
+      "grad_norm": 0.9136660302900956,
+      "learning_rate": 7.433555929034493e-08,
+      "loss": 1.5606,
+      "step": 3855
+    },
+    {
+      "epoch": 0.8003320880033209,
+      "grad_norm": 0.6295169325286589,
+      "learning_rate": 7.426663930265394e-08,
+      "loss": 1.5154,
+      "step": 3856
+    },
+    {
+      "epoch": 0.8005396430053965,
+      "grad_norm": 1.1273239063433547,
+      "learning_rate": 7.419778127691885e-08,
+      "loss": 1.5534,
+      "step": 3857
+    },
+    {
+      "epoch": 0.800747198007472,
+      "grad_norm": 0.8445575302503551,
+      "learning_rate": 7.41289852424164e-08,
+      "loss": 1.4837,
+      "step": 3858
+    },
+    {
+      "epoch": 0.8009547530095475,
+      "grad_norm": 1.2778901830517406,
+      "learning_rate": 7.406025122839674e-08,
+      "loss": 1.5638,
+      "step": 3859
+    },
+    {
+      "epoch": 0.801162308011623,
+      "grad_norm": 0.7747781489593895,
+      "learning_rate": 7.399157926408379e-08,
+      "loss": 1.4585,
+      "step": 3860
+    },
+    {
+      "epoch": 0.8013698630136986,
+      "grad_norm": 0.9269615278458849,
+      "learning_rate": 7.392296937867504e-08,
+      "loss": 1.5848,
+      "step": 3861
+    },
+    {
+      "epoch": 0.8015774180157742,
+      "grad_norm": 0.6722431753258891,
+      "learning_rate": 7.385442160134154e-08,
+      "loss": 1.538,
+      "step": 3862
+    },
+    {
+      "epoch": 0.8017849730178497,
+      "grad_norm": 0.8534815699612766,
+      "learning_rate": 7.37859359612281e-08,
+      "loss": 1.5587,
+      "step": 3863
+    },
+    {
+      "epoch": 0.8019925280199253,
+      "grad_norm": 0.7334316412176408,
+      "learning_rate": 7.371751248745287e-08,
+      "loss": 1.5953,
+      "step": 3864
+    },
+    {
+      "epoch": 0.8022000830220009,
+      "grad_norm": 0.7292637896641967,
+      "learning_rate": 7.364915120910777e-08,
+      "loss": 1.6417,
+      "step": 3865
+    },
+    {
+      "epoch": 0.8024076380240763,
+      "grad_norm": 0.7643331973665636,
+      "learning_rate": 7.35808521552582e-08,
+      "loss": 1.5008,
+      "step": 3866
+    },
+    {
+      "epoch": 0.8026151930261519,
+      "grad_norm": 0.9534567074707053,
+      "learning_rate": 7.351261535494309e-08,
+      "loss": 1.5734,
+      "step": 3867
+    },
+    {
+      "epoch": 0.8028227480282275,
+      "grad_norm": 0.8442508637811588,
+      "learning_rate": 7.344444083717483e-08,
+      "loss": 1.4857,
+      "step": 3868
+    },
+    {
+      "epoch": 0.803030303030303,
+      "grad_norm": 1.035905981812537,
+      "learning_rate": 7.337632863093956e-08,
+      "loss": 1.4805,
+      "step": 3869
+    },
+    {
+      "epoch": 0.8032378580323786,
+      "grad_norm": 0.7324116497096257,
+      "learning_rate": 7.330827876519674e-08,
+      "loss": 1.4544,
+      "step": 3870
+    },
+    {
+      "epoch": 0.8034454130344542,
+      "grad_norm": 0.9948183654442785,
+      "learning_rate": 7.324029126887934e-08,
+      "loss": 1.5091,
+      "step": 3871
+    },
+    {
+      "epoch": 0.8036529680365296,
+      "grad_norm": 0.6639923825482803,
+      "learning_rate": 7.317236617089384e-08,
+      "loss": 1.445,
+      "step": 3872
+    },
+    {
+      "epoch": 0.8038605230386052,
+      "grad_norm": 0.7544995742017733,
+      "learning_rate": 7.310450350012014e-08,
+      "loss": 1.4993,
+      "step": 3873
+    },
+    {
+      "epoch": 0.8040680780406808,
+      "grad_norm": 0.8374010144518952,
+      "learning_rate": 7.303670328541174e-08,
+      "loss": 1.4786,
+      "step": 3874
+    },
+    {
+      "epoch": 0.8042756330427563,
+      "grad_norm": 0.7765825548957686,
+      "learning_rate": 7.296896555559545e-08,
+      "loss": 1.5372,
+      "step": 3875
+    },
+    {
+      "epoch": 0.8044831880448319,
+      "grad_norm": 0.8292215954590633,
+      "learning_rate": 7.290129033947157e-08,
+      "loss": 1.5174,
+      "step": 3876
+    },
+    {
+      "epoch": 0.8046907430469075,
+      "grad_norm": 0.6371192683135943,
+      "learning_rate": 7.283367766581374e-08,
+      "loss": 1.5826,
+      "step": 3877
+    },
+    {
+      "epoch": 0.8048982980489829,
+      "grad_norm": 6.161995203666757,
+      "learning_rate": 7.27661275633692e-08,
+      "loss": 1.477,
+      "step": 3878
+    },
+    {
+      "epoch": 0.8051058530510585,
+      "grad_norm": 0.6754767166600503,
+      "learning_rate": 7.269864006085828e-08,
+      "loss": 1.4497,
+      "step": 3879
+    },
+    {
+      "epoch": 0.8053134080531341,
+      "grad_norm": 0.7662155229934128,
+      "learning_rate": 7.263121518697504e-08,
+      "loss": 1.5367,
+      "step": 3880
+    },
+    {
+      "epoch": 0.8055209630552096,
+      "grad_norm": 0.6607989901290842,
+      "learning_rate": 7.256385297038669e-08,
+      "loss": 1.4935,
+      "step": 3881
+    },
+    {
+      "epoch": 0.8057285180572852,
+      "grad_norm": 0.9536220755827555,
+      "learning_rate": 7.249655343973384e-08,
+      "loss": 1.5654,
+      "step": 3882
+    },
+    {
+      "epoch": 0.8059360730593608,
+      "grad_norm": 0.750685502810434,
+      "learning_rate": 7.242931662363043e-08,
+      "loss": 1.5163,
+      "step": 3883
+    },
+    {
+      "epoch": 0.8061436280614362,
+      "grad_norm": 2.5670962030730626,
+      "learning_rate": 7.236214255066387e-08,
+      "loss": 1.4778,
+      "step": 3884
+    },
+    {
+      "epoch": 0.8063511830635118,
+      "grad_norm": 0.9602932267573792,
+      "learning_rate": 7.229503124939474e-08,
+      "loss": 1.5177,
+      "step": 3885
+    },
+    {
+      "epoch": 0.8065587380655874,
+      "grad_norm": 1.4307181956918524,
+      "learning_rate": 7.222798274835691e-08,
+      "loss": 1.4839,
+      "step": 3886
+    },
+    {
+      "epoch": 0.8067662930676629,
+      "grad_norm": 1.5139680655817953,
+      "learning_rate": 7.21609970760578e-08,
+      "loss": 1.5649,
+      "step": 3887
+    },
+    {
+      "epoch": 0.8069738480697385,
+      "grad_norm": 0.7507957692202775,
+      "learning_rate": 7.209407426097771e-08,
+      "loss": 1.5233,
+      "step": 3888
+    },
+    {
+      "epoch": 0.8071814030718141,
+      "grad_norm": 0.7970275094463274,
+      "learning_rate": 7.202721433157065e-08,
+      "loss": 1.4786,
+      "step": 3889
+    },
+    {
+      "epoch": 0.8073889580738896,
+      "grad_norm": 1.5403777399492706,
+      "learning_rate": 7.196041731626357e-08,
+      "loss": 1.4664,
+      "step": 3890
+    },
+    {
+      "epoch": 0.8075965130759651,
+      "grad_norm": 0.7170626884402947,
+      "learning_rate": 7.189368324345684e-08,
+      "loss": 1.4986,
+      "step": 3891
+    },
+    {
+      "epoch": 0.8078040680780407,
+      "grad_norm": 0.8497841324387797,
+      "learning_rate": 7.182701214152393e-08,
+      "loss": 1.4893,
+      "step": 3892
+    },
+    {
+      "epoch": 0.8080116230801162,
+      "grad_norm": 1.7606896202113627,
+      "learning_rate": 7.17604040388118e-08,
+      "loss": 1.5014,
+      "step": 3893
+    },
+    {
+      "epoch": 0.8082191780821918,
+      "grad_norm": 2.380417792949077,
+      "learning_rate": 7.169385896364024e-08,
+      "loss": 1.6079,
+      "step": 3894
+    },
+    {
+      "epoch": 0.8084267330842674,
+      "grad_norm": 1.2069511101542256,
+      "learning_rate": 7.162737694430258e-08,
+      "loss": 1.4392,
+      "step": 3895
+    },
+    {
+      "epoch": 0.8086342880863429,
+      "grad_norm": 1.2657653724173512,
+      "learning_rate": 7.156095800906519e-08,
+      "loss": 1.5376,
+      "step": 3896
+    },
+    {
+      "epoch": 0.8088418430884184,
+      "grad_norm": 0.881473986770486,
+      "learning_rate": 7.149460218616762e-08,
+      "loss": 1.5203,
+      "step": 3897
+    },
+    {
+      "epoch": 0.809049398090494,
+      "grad_norm": 0.939380879352767,
+      "learning_rate": 7.142830950382255e-08,
+      "loss": 1.4858,
+      "step": 3898
+    },
+    {
+      "epoch": 0.8092569530925695,
+      "grad_norm": 0.6894016139318221,
+      "learning_rate": 7.136207999021598e-08,
+      "loss": 1.5571,
+      "step": 3899
+    },
+    {
+      "epoch": 0.8094645080946451,
+      "grad_norm": 0.9955442421253108,
+      "learning_rate": 7.129591367350687e-08,
+      "loss": 1.605,
+      "step": 3900
+    },
+    {
+      "epoch": 0.8096720630967207,
+      "grad_norm": 0.9466330902587341,
+      "learning_rate": 7.122981058182738e-08,
+      "loss": 1.5101,
+      "step": 3901
+    },
+    {
+      "epoch": 0.8098796180987962,
+      "grad_norm": 0.9201279792541084,
+      "learning_rate": 7.116377074328286e-08,
+      "loss": 1.4535,
+      "step": 3902
+    },
+    {
+      "epoch": 0.8100871731008717,
+      "grad_norm": 0.8987459809397822,
+      "learning_rate": 7.109779418595156e-08,
+      "loss": 1.5099,
+      "step": 3903
+    },
+    {
+      "epoch": 0.8102947281029472,
+      "grad_norm": 0.8394512381360564,
+      "learning_rate": 7.103188093788514e-08,
+      "loss": 1.5202,
+      "step": 3904
+    },
+    {
+      "epoch": 0.8105022831050228,
+      "grad_norm": 1.0019271931051434,
+      "learning_rate": 7.096603102710801e-08,
+      "loss": 1.4988,
+      "step": 3905
+    },
+    {
+      "epoch": 0.8107098381070984,
+      "grad_norm": 0.8486595373804554,
+      "learning_rate": 7.090024448161787e-08,
+      "loss": 1.4988,
+      "step": 3906
+    },
+    {
+      "epoch": 0.8109173931091739,
+      "grad_norm": 0.627107944819042,
+      "learning_rate": 7.08345213293854e-08,
+      "loss": 1.4776,
+      "step": 3907
+    },
+    {
+      "epoch": 0.8111249481112495,
+      "grad_norm": 0.7180127528183975,
+      "learning_rate": 7.076886159835437e-08,
+      "loss": 1.4889,
+      "step": 3908
+    },
+    {
+      "epoch": 0.811332503113325,
+      "grad_norm": 0.8250597194821029,
+      "learning_rate": 7.070326531644147e-08,
+      "loss": 1.5134,
+      "step": 3909
+    },
+    {
+      "epoch": 0.8115400581154005,
+      "grad_norm": 0.7623502901965159,
+      "learning_rate": 7.063773251153657e-08,
+      "loss": 1.5209,
+      "step": 3910
+    },
+    {
+      "epoch": 0.8117476131174761,
+      "grad_norm": 0.8853965008680758,
+      "learning_rate": 7.057226321150249e-08,
+      "loss": 1.4935,
+      "step": 3911
+    },
+    {
+      "epoch": 0.8119551681195517,
+      "grad_norm": 0.658447897728513,
+      "learning_rate": 7.050685744417497e-08,
+      "loss": 1.5537,
+      "step": 3912
+    },
+    {
+      "epoch": 0.8121627231216272,
+      "grad_norm": 0.772900224816853,
+      "learning_rate": 7.044151523736295e-08,
+      "loss": 1.5726,
+      "step": 3913
+    },
+    {
+      "epoch": 0.8123702781237028,
+      "grad_norm": 0.7762644944873829,
+      "learning_rate": 7.037623661884798e-08,
+      "loss": 1.5239,
+      "step": 3914
+    },
+    {
+      "epoch": 0.8125778331257784,
+      "grad_norm": 1.7418646295764761,
+      "learning_rate": 7.031102161638496e-08,
+      "loss": 1.5462,
+      "step": 3915
+    },
+    {
+      "epoch": 0.8127853881278538,
+      "grad_norm": 1.037776632386954,
+      "learning_rate": 7.024587025770154e-08,
+      "loss": 1.4912,
+      "step": 3916
+    },
+    {
+      "epoch": 0.8129929431299294,
+      "grad_norm": 0.8405495028598344,
+      "learning_rate": 7.018078257049836e-08,
+      "loss": 1.503,
+      "step": 3917
+    },
+    {
+      "epoch": 0.813200498132005,
+      "grad_norm": 1.2601807883092309,
+      "learning_rate": 7.011575858244889e-08,
+      "loss": 1.5968,
+      "step": 3918
+    },
+    {
+      "epoch": 0.8134080531340805,
+      "grad_norm": 0.8631842028497757,
+      "learning_rate": 7.005079832119977e-08,
+      "loss": 1.5123,
+      "step": 3919
+    },
+    {
+      "epoch": 0.8136156081361561,
+      "grad_norm": 0.8109946111067116,
+      "learning_rate": 6.998590181437018e-08,
+      "loss": 1.5462,
+      "step": 3920
+    },
+    {
+      "epoch": 0.8138231631382317,
+      "grad_norm": 0.7499304668430169,
+      "learning_rate": 6.992106908955253e-08,
+      "loss": 1.5361,
+      "step": 3921
+    },
+    {
+      "epoch": 0.8140307181403071,
+      "grad_norm": 0.6435543150522368,
+      "learning_rate": 6.985630017431195e-08,
+      "loss": 1.5213,
+      "step": 3922
+    },
+    {
+      "epoch": 0.8142382731423827,
+      "grad_norm": 3.350248213726436,
+      "learning_rate": 6.97915950961864e-08,
+      "loss": 1.5266,
+      "step": 3923
+    },
+    {
+      "epoch": 0.8144458281444583,
+      "grad_norm": 0.7466573640325941,
+      "learning_rate": 6.972695388268683e-08,
+      "loss": 1.431,
+      "step": 3924
+    },
+    {
+      "epoch": 0.8146533831465338,
+      "grad_norm": 0.6616118246346304,
+      "learning_rate": 6.966237656129699e-08,
+      "loss": 1.5716,
+      "step": 3925
+    },
+    {
+      "epoch": 0.8148609381486094,
+      "grad_norm": 0.7911929576274094,
+      "learning_rate": 6.959786315947337e-08,
+      "loss": 1.4813,
+      "step": 3926
+    },
+    {
+      "epoch": 0.815068493150685,
+      "grad_norm": 0.6657740916108909,
+      "learning_rate": 6.953341370464534e-08,
+      "loss": 1.5693,
+      "step": 3927
+    },
+    {
+      "epoch": 0.8152760481527604,
+      "grad_norm": 0.8922021290033862,
+      "learning_rate": 6.946902822421523e-08,
+      "loss": 1.4883,
+      "step": 3928
+    },
+    {
+      "epoch": 0.815483603154836,
+      "grad_norm": 0.6748956994472686,
+      "learning_rate": 6.940470674555787e-08,
+      "loss": 1.5426,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8156911581569116,
+      "grad_norm": 0.8180035705269678,
+      "learning_rate": 6.934044929602118e-08,
+      "loss": 1.4922,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8158987131589871,
+      "grad_norm": 0.6885504284129773,
+      "learning_rate": 6.927625590292562e-08,
+      "loss": 1.5001,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8161062681610627,
+      "grad_norm": 0.7493248564437155,
+      "learning_rate": 6.921212659356462e-08,
+      "loss": 1.4672,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8163138231631383,
+      "grad_norm": 0.7009465607093497,
+      "learning_rate": 6.914806139520412e-08,
+      "loss": 1.5368,
+      "step": 3933
+    },
+    {
+      "epoch": 0.8165213781652138,
+      "grad_norm": 0.6555749643294975,
+      "learning_rate": 6.908406033508311e-08,
+      "loss": 1.5656,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8167289331672893,
+      "grad_norm": 2.862404783126632,
+      "learning_rate": 6.9020123440413e-08,
+      "loss": 1.5035,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8169364881693649,
+      "grad_norm": 1.8440126188902186,
+      "learning_rate": 6.895625073837813e-08,
+      "loss": 1.4635,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8171440431714404,
+      "grad_norm": 0.7583126981058725,
+      "learning_rate": 6.889244225613549e-08,
+      "loss": 1.582,
+      "step": 3937
+    },
+    {
+      "epoch": 0.817351598173516,
+      "grad_norm": 0.9761642608640844,
+      "learning_rate": 6.882869802081463e-08,
+      "loss": 1.493,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8175591531755916,
+      "grad_norm": 0.6784036569302359,
+      "learning_rate": 6.87650180595181e-08,
+      "loss": 1.5166,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8177667081776671,
+      "grad_norm": 1.537508250417813,
+      "learning_rate": 6.870140239932081e-08,
+      "loss": 1.5095,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8179742631797426,
+      "grad_norm": 0.8462411484393287,
+      "learning_rate": 6.863785106727044e-08,
+      "loss": 1.5169,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8181818181818182,
+      "grad_norm": 1.1495082592010728,
+      "learning_rate": 6.857436409038737e-08,
+      "loss": 1.544,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8183893731838937,
+      "grad_norm": 1.2199801385914453,
+      "learning_rate": 6.851094149566463e-08,
+      "loss": 1.5085,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8185969281859693,
+      "grad_norm": 1.1344819929576584,
+      "learning_rate": 6.844758331006767e-08,
+      "loss": 1.6313,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8188044831880449,
+      "grad_norm": 0.7719865711365146,
+      "learning_rate": 6.838428956053484e-08,
+      "loss": 1.4869,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8190120381901204,
+      "grad_norm": 0.7716814754133661,
+      "learning_rate": 6.832106027397692e-08,
+      "loss": 1.4697,
+      "step": 3946
+    },
+    {
+      "epoch": 0.819219593192196,
+      "grad_norm": 0.7200453306772558,
+      "learning_rate": 6.825789547727734e-08,
+      "loss": 1.5536,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8194271481942715,
+      "grad_norm": 0.9333078881532242,
+      "learning_rate": 6.819479519729203e-08,
+      "loss": 1.5322,
+      "step": 3948
+    },
+    {
+      "epoch": 0.819634703196347,
+      "grad_norm": 1.0854579305499146,
+      "learning_rate": 6.813175946084964e-08,
+      "loss": 1.4809,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8198422581984226,
+      "grad_norm": 0.8929740088085423,
+      "learning_rate": 6.806878829475126e-08,
+      "loss": 1.577,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8200498132004981,
+      "grad_norm": 0.8738077849303773,
+      "learning_rate": 6.800588172577056e-08,
+      "loss": 1.4622,
+      "step": 3951
+    },
+    {
+      "epoch": 0.8202573682025737,
+      "grad_norm": 0.7658881722551253,
+      "learning_rate": 6.79430397806537e-08,
+      "loss": 1.5241,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8204649232046493,
+      "grad_norm": 0.7887480792608634,
+      "learning_rate": 6.788026248611943e-08,
+      "loss": 1.5417,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8206724782067247,
+      "grad_norm": 1.0391172925583911,
+      "learning_rate": 6.781754986885908e-08,
+      "loss": 1.5583,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8208800332088003,
+      "grad_norm": 0.6797121632495665,
+      "learning_rate": 6.775490195553623e-08,
+      "loss": 1.5143,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8210875882108759,
+      "grad_norm": 0.930437352774962,
+      "learning_rate": 6.769231877278722e-08,
+      "loss": 1.483,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8212951432129514,
+      "grad_norm": 0.977655021724759,
+      "learning_rate": 6.762980034722074e-08,
+      "loss": 1.556,
+      "step": 3957
+    },
+    {
+      "epoch": 0.821502698215027,
+      "grad_norm": 1.1061971670670543,
+      "learning_rate": 6.756734670541796e-08,
+      "loss": 1.5894,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8217102532171026,
+      "grad_norm": 0.7392699414131901,
+      "learning_rate": 6.750495787393246e-08,
+      "loss": 1.4559,
+      "step": 3959
+    },
+    {
+      "epoch": 0.821917808219178,
+      "grad_norm": 0.7135311695772597,
+      "learning_rate": 6.744263387929043e-08,
+      "loss": 1.588,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8221253632212536,
+      "grad_norm": 0.83828066088895,
+      "learning_rate": 6.738037474799024e-08,
+      "loss": 1.5301,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8223329182233292,
+      "grad_norm": 0.678534812742404,
+      "learning_rate": 6.731818050650291e-08,
+      "loss": 1.5302,
+      "step": 3962
+    },
+    {
+      "epoch": 0.8225404732254047,
+      "grad_norm": 0.7335522427644094,
+      "learning_rate": 6.725605118127178e-08,
+      "loss": 1.5232,
+      "step": 3963
+    },
+    {
+      "epoch": 0.8227480282274803,
+      "grad_norm": 0.6417187060611894,
+      "learning_rate": 6.719398679871251e-08,
+      "loss": 1.4722,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8229555832295559,
+      "grad_norm": 0.7662115995940865,
+      "learning_rate": 6.713198738521333e-08,
+      "loss": 1.5787,
+      "step": 3965
+    },
+    {
+      "epoch": 0.8231631382316313,
+      "grad_norm": 1.8027099372868731,
+      "learning_rate": 6.707005296713468e-08,
+      "loss": 1.4893,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8233706932337069,
+      "grad_norm": 0.8526968726829183,
+      "learning_rate": 6.700818357080946e-08,
+      "loss": 1.5642,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8235782482357825,
+      "grad_norm": 0.7536337905438388,
+      "learning_rate": 6.694637922254285e-08,
+      "loss": 1.5283,
+      "step": 3968
+    },
+    {
+      "epoch": 0.823785803237858,
+      "grad_norm": 0.6901380135153965,
+      "learning_rate": 6.688463994861256e-08,
+      "loss": 1.5496,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8239933582399336,
+      "grad_norm": 0.9394012051448764,
+      "learning_rate": 6.682296577526825e-08,
+      "loss": 1.5479,
+      "step": 3970
+    },
+    {
+      "epoch": 0.8242009132420092,
+      "grad_norm": 3.338279899333179,
+      "learning_rate": 6.676135672873235e-08,
+      "loss": 1.5816,
+      "step": 3971
+    },
+    {
+      "epoch": 0.8244084682440846,
+      "grad_norm": 0.837861545862268,
+      "learning_rate": 6.66998128351993e-08,
+      "loss": 1.5243,
+      "step": 3972
+    },
+    {
+      "epoch": 0.8246160232461602,
+      "grad_norm": 0.8099423206670616,
+      "learning_rate": 6.663833412083594e-08,
+      "loss": 1.5165,
+      "step": 3973
+    },
+    {
+      "epoch": 0.8248235782482358,
+      "grad_norm": 0.7002104648336371,
+      "learning_rate": 6.657692061178135e-08,
+      "loss": 1.5257,
+      "step": 3974
+    },
+    {
+      "epoch": 0.8250311332503113,
+      "grad_norm": 0.8528257669780135,
+      "learning_rate": 6.651557233414701e-08,
+      "loss": 1.6136,
+      "step": 3975
+    },
+    {
+      "epoch": 0.8252386882523869,
+      "grad_norm": 0.9462603281941957,
+      "learning_rate": 6.645428931401654e-08,
+      "loss": 1.5316,
+      "step": 3976
+    },
+    {
+      "epoch": 0.8254462432544625,
+      "grad_norm": 0.6783504465284893,
+      "learning_rate": 6.639307157744584e-08,
+      "loss": 1.5971,
+      "step": 3977
+    },
+    {
+      "epoch": 0.825653798256538,
+      "grad_norm": 1.014143670423745,
+      "learning_rate": 6.633191915046308e-08,
+      "loss": 1.4433,
+      "step": 3978
+    },
+    {
+      "epoch": 0.8258613532586135,
+      "grad_norm": 0.6574601956332877,
+      "learning_rate": 6.627083205906858e-08,
+      "loss": 1.4916,
+      "step": 3979
+    },
+    {
+      "epoch": 0.8260689082606891,
+      "grad_norm": 0.7508584673423091,
+      "learning_rate": 6.620981032923507e-08,
+      "loss": 1.5214,
+      "step": 3980
+    },
+    {
+      "epoch": 0.8262764632627646,
+      "grad_norm": 1.9789408390185967,
+      "learning_rate": 6.614885398690731e-08,
+      "loss": 1.5309,
+      "step": 3981
+    },
+    {
+      "epoch": 0.8264840182648402,
+      "grad_norm": 0.756981072224144,
+      "learning_rate": 6.608796305800233e-08,
+      "loss": 1.5772,
+      "step": 3982
+    },
+    {
+      "epoch": 0.8266915732669158,
+      "grad_norm": 0.7642241825056885,
+      "learning_rate": 6.602713756840925e-08,
+      "loss": 1.5674,
+      "step": 3983
+    },
+    {
+      "epoch": 0.8268991282689913,
+      "grad_norm": 0.6732728997199778,
+      "learning_rate": 6.596637754398964e-08,
+      "loss": 1.5036,
+      "step": 3984
+    },
+    {
+      "epoch": 0.8271066832710668,
+      "grad_norm": 1.2061167295007007,
+      "learning_rate": 6.590568301057684e-08,
+      "loss": 1.5607,
+      "step": 3985
+    },
+    {
+      "epoch": 0.8273142382731424,
+      "grad_norm": 0.8267377943104486,
+      "learning_rate": 6.584505399397671e-08,
+      "loss": 1.4657,
+      "step": 3986
+    },
+    {
+      "epoch": 0.8275217932752179,
+      "grad_norm": 0.7452701686992949,
+      "learning_rate": 6.578449051996704e-08,
+      "loss": 1.5487,
+      "step": 3987
+    },
+    {
+      "epoch": 0.8277293482772935,
+      "grad_norm": 0.7849842204686843,
+      "learning_rate": 6.572399261429779e-08,
+      "loss": 1.5224,
+      "step": 3988
+    },
+    {
+      "epoch": 0.8279369032793691,
+      "grad_norm": 0.6890583808368584,
+      "learning_rate": 6.566356030269107e-08,
+      "loss": 1.5383,
+      "step": 3989
+    },
+    {
+      "epoch": 0.8281444582814446,
+      "grad_norm": 0.6631750490967718,
+      "learning_rate": 6.560319361084113e-08,
+      "loss": 1.5068,
+      "step": 3990
+    },
+    {
+      "epoch": 0.8283520132835201,
+      "grad_norm": 0.6934382234608076,
+      "learning_rate": 6.554289256441428e-08,
+      "loss": 1.5096,
+      "step": 3991
+    },
+    {
+      "epoch": 0.8285595682855957,
+      "grad_norm": 0.7029998840001955,
+      "learning_rate": 6.548265718904885e-08,
+      "loss": 1.4434,
+      "step": 3992
+    },
+    {
+      "epoch": 0.8287671232876712,
+      "grad_norm": 0.7602917098136772,
+      "learning_rate": 6.542248751035549e-08,
+      "loss": 1.534,
+      "step": 3993
+    },
+    {
+      "epoch": 0.8289746782897468,
+      "grad_norm": 0.6970463824668981,
+      "learning_rate": 6.536238355391653e-08,
+      "loss": 1.5488,
+      "step": 3994
+    },
+    {
+      "epoch": 0.8291822332918223,
+      "grad_norm": 0.8554720805167183,
+      "learning_rate": 6.530234534528678e-08,
+      "loss": 1.4676,
+      "step": 3995
+    },
+    {
+      "epoch": 0.8293897882938979,
+      "grad_norm": 0.7399200491620423,
+      "learning_rate": 6.524237290999273e-08,
+      "loss": 1.5223,
+      "step": 3996
+    },
+    {
+      "epoch": 0.8295973432959735,
+      "grad_norm": 1.109589195801519,
+      "learning_rate": 6.518246627353316e-08,
+      "loss": 1.5512,
+      "step": 3997
+    },
+    {
+      "epoch": 0.8298048982980489,
+      "grad_norm": 0.8261470761235,
+      "learning_rate": 6.512262546137879e-08,
+      "loss": 1.4925,
+      "step": 3998
+    },
+    {
+      "epoch": 0.8300124533001245,
+      "grad_norm": 0.7896728165008715,
+      "learning_rate": 6.506285049897226e-08,
+      "loss": 1.5333,
+      "step": 3999
+    },
+    {
+      "epoch": 0.8302200083022001,
+      "grad_norm": 0.8472332292152489,
+      "learning_rate": 6.500314141172835e-08,
+      "loss": 1.5613,
+      "step": 4000
+    },
+    {
+      "epoch": 0.8304275633042756,
+      "grad_norm": 0.7266928084604373,
+      "learning_rate": 6.49434982250338e-08,
+      "loss": 1.6294,
+      "step": 4001
+    },
+    {
+      "epoch": 0.8306351183063512,
+      "grad_norm": 0.9058703164168328,
+      "learning_rate": 6.488392096424731e-08,
+      "loss": 1.4667,
+      "step": 4002
+    },
+    {
+      "epoch": 0.8308426733084268,
+      "grad_norm": 0.7157024306252904,
+      "learning_rate": 6.482440965469952e-08,
+      "loss": 1.4278,
+      "step": 4003
+    },
+    {
+      "epoch": 0.8310502283105022,
+      "grad_norm": 0.8619017286337939,
+      "learning_rate": 6.476496432169305e-08,
+      "loss": 1.4644,
+      "step": 4004
+    },
+    {
+      "epoch": 0.8312577833125778,
+      "grad_norm": 0.8196160308817438,
+      "learning_rate": 6.470558499050247e-08,
+      "loss": 1.5187,
+      "step": 4005
+    },
+    {
+      "epoch": 0.8314653383146534,
+      "grad_norm": 1.0416874361096575,
+      "learning_rate": 6.464627168637437e-08,
+      "loss": 1.4788,
+      "step": 4006
+    },
+    {
+      "epoch": 0.8316728933167289,
+      "grad_norm": 0.6446496176647841,
+      "learning_rate": 6.458702443452712e-08,
+      "loss": 1.4749,
+      "step": 4007
+    },
+    {
+      "epoch": 0.8318804483188045,
+      "grad_norm": 0.9902385284942313,
+      "learning_rate": 6.452784326015112e-08,
+      "loss": 1.5644,
+      "step": 4008
+    },
+    {
+      "epoch": 0.8320880033208801,
+      "grad_norm": 0.6637870809850734,
+      "learning_rate": 6.446872818840857e-08,
+      "loss": 1.5092,
+      "step": 4009
+    },
+    {
+      "epoch": 0.8322955583229555,
+      "grad_norm": 1.3557226768033919,
+      "learning_rate": 6.440967924443376e-08,
+      "loss": 1.464,
+      "step": 4010
+    },
+    {
+      "epoch": 0.8325031133250311,
+      "grad_norm": 0.8224373065000923,
+      "learning_rate": 6.435069645333255e-08,
+      "loss": 1.5493,
+      "step": 4011
+    },
+    {
+      "epoch": 0.8327106683271067,
+      "grad_norm": 0.6673155421788557,
+      "learning_rate": 6.429177984018299e-08,
+      "loss": 1.492,
+      "step": 4012
+    },
+    {
+      "epoch": 0.8329182233291822,
+      "grad_norm": 1.1966523136035498,
+      "learning_rate": 6.423292943003483e-08,
+      "loss": 1.4943,
+      "step": 4013
+    },
+    {
+      "epoch": 0.8331257783312578,
+      "grad_norm": 1.1958785207033091,
+      "learning_rate": 6.417414524790972e-08,
+      "loss": 1.5035,
+      "step": 4014
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 0.683083340107233,
+      "learning_rate": 6.411542731880104e-08,
+      "loss": 1.4767,
+      "step": 4015
+    },
+    {
+      "epoch": 0.8335408883354088,
+      "grad_norm": 0.8023184503380092,
+      "learning_rate": 6.405677566767422e-08,
+      "loss": 1.5772,
+      "step": 4016
+    },
+    {
+      "epoch": 0.8337484433374844,
+      "grad_norm": 1.9828963594388602,
+      "learning_rate": 6.39981903194663e-08,
+      "loss": 1.49,
+      "step": 4017
+    },
+    {
+      "epoch": 0.83395599833956,
+      "grad_norm": 0.6552785754007138,
+      "learning_rate": 6.393967129908623e-08,
+      "loss": 1.5242,
+      "step": 4018
+    },
+    {
+      "epoch": 0.8341635533416355,
+      "grad_norm": 1.115460944958447,
+      "learning_rate": 6.388121863141485e-08,
+      "loss": 1.481,
+      "step": 4019
+    },
+    {
+      "epoch": 0.8343711083437111,
+      "grad_norm": 2.3952332672823013,
+      "learning_rate": 6.382283234130449e-08,
+      "loss": 1.5793,
+      "step": 4020
+    },
+    {
+      "epoch": 0.8345786633457867,
+      "grad_norm": 0.7691691239866543,
+      "learning_rate": 6.37645124535796e-08,
+      "loss": 1.5019,
+      "step": 4021
+    },
+    {
+      "epoch": 0.8347862183478622,
+      "grad_norm": 1.0999299882569888,
+      "learning_rate": 6.370625899303619e-08,
+      "loss": 1.4645,
+      "step": 4022
+    },
+    {
+      "epoch": 0.8349937733499377,
+      "grad_norm": 1.1305404759027466,
+      "learning_rate": 6.36480719844421e-08,
+      "loss": 1.4836,
+      "step": 4023
+    },
+    {
+      "epoch": 0.8352013283520133,
+      "grad_norm": 0.7426489174245947,
+      "learning_rate": 6.358995145253684e-08,
+      "loss": 1.4689,
+      "step": 4024
+    },
+    {
+      "epoch": 0.8354088833540888,
+      "grad_norm": 3.657741647753389,
+      "learning_rate": 6.353189742203186e-08,
+      "loss": 1.5608,
+      "step": 4025
+    },
+    {
+      "epoch": 0.8356164383561644,
+      "grad_norm": 1.619263971695883,
+      "learning_rate": 6.347390991761001e-08,
+      "loss": 1.5022,
+      "step": 4026
+    },
+    {
+      "epoch": 0.83582399335824,
+      "grad_norm": 0.7463013493328976,
+      "learning_rate": 6.341598896392622e-08,
+      "loss": 1.5812,
+      "step": 4027
+    },
+    {
+      "epoch": 0.8360315483603155,
+      "grad_norm": 0.8919105757367832,
+      "learning_rate": 6.33581345856068e-08,
+      "loss": 1.5046,
+      "step": 4028
+    },
+    {
+      "epoch": 0.836239103362391,
+      "grad_norm": 1.045769014945916,
+      "learning_rate": 6.330034680724994e-08,
+      "loss": 1.5252,
+      "step": 4029
+    },
+    {
+      "epoch": 0.8364466583644666,
+      "grad_norm": 1.1356060064209288,
+      "learning_rate": 6.324262565342551e-08,
+      "loss": 1.5143,
+      "step": 4030
+    },
+    {
+      "epoch": 0.8366542133665421,
+      "grad_norm": 0.863722223247178,
+      "learning_rate": 6.318497114867496e-08,
+      "loss": 1.5349,
+      "step": 4031
+    },
+    {
+      "epoch": 0.8368617683686177,
+      "grad_norm": 0.7454296691034005,
+      "learning_rate": 6.312738331751151e-08,
+      "loss": 1.4336,
+      "step": 4032
+    },
+    {
+      "epoch": 0.8370693233706933,
+      "grad_norm": 0.7981852137376518,
+      "learning_rate": 6.306986218441989e-08,
+      "loss": 1.5751,
+      "step": 4033
+    },
+    {
+      "epoch": 0.8372768783727688,
+      "grad_norm": 0.9196672384985635,
+      "learning_rate": 6.301240777385668e-08,
+      "loss": 1.5902,
+      "step": 4034
+    },
+    {
+      "epoch": 0.8374844333748444,
+      "grad_norm": 0.7031730723823548,
+      "learning_rate": 6.295502011024982e-08,
+      "loss": 1.5013,
+      "step": 4035
+    },
+    {
+      "epoch": 0.8376919883769199,
+      "grad_norm": 0.7216430684092466,
+      "learning_rate": 6.289769921799917e-08,
+      "loss": 1.512,
+      "step": 4036
+    },
+    {
+      "epoch": 0.8378995433789954,
+      "grad_norm": 2.729593799347814,
+      "learning_rate": 6.284044512147594e-08,
+      "loss": 1.5635,
+      "step": 4037
+    },
+    {
+      "epoch": 0.838107098381071,
+      "grad_norm": 0.6598622031005454,
+      "learning_rate": 6.278325784502313e-08,
+      "loss": 1.5532,
+      "step": 4038
+    },
+    {
+      "epoch": 0.8383146533831465,
+      "grad_norm": 0.7528190074016754,
+      "learning_rate": 6.272613741295521e-08,
+      "loss": 1.4635,
+      "step": 4039
+    },
+    {
+      "epoch": 0.8385222083852221,
+      "grad_norm": 0.6400501362935004,
+      "learning_rate": 6.266908384955827e-08,
+      "loss": 1.5204,
+      "step": 4040
+    },
+    {
+      "epoch": 0.8387297633872977,
+      "grad_norm": 1.813795407042309,
+      "learning_rate": 6.261209717908995e-08,
+      "loss": 1.5519,
+      "step": 4041
+    },
+    {
+      "epoch": 0.8389373183893731,
+      "grad_norm": 0.7276495992698796,
+      "learning_rate": 6.255517742577952e-08,
+      "loss": 1.5482,
+      "step": 4042
+    },
+    {
+      "epoch": 0.8391448733914487,
+      "grad_norm": 0.8518484520662879,
+      "learning_rate": 6.249832461382775e-08,
+      "loss": 1.4735,
+      "step": 4043
+    },
+    {
+      "epoch": 0.8393524283935243,
+      "grad_norm": 2.598369620965311,
+      "learning_rate": 6.244153876740686e-08,
+      "loss": 1.545,
+      "step": 4044
+    },
+    {
+      "epoch": 0.8395599833955998,
+      "grad_norm": 0.9537866917180096,
+      "learning_rate": 6.238481991066085e-08,
+      "loss": 1.4609,
+      "step": 4045
+    },
+    {
+      "epoch": 0.8397675383976754,
+      "grad_norm": 0.6656352183916423,
+      "learning_rate": 6.232816806770487e-08,
+      "loss": 1.4427,
+      "step": 4046
+    },
+    {
+      "epoch": 0.839975093399751,
+      "grad_norm": 0.7057138708790681,
+      "learning_rate": 6.227158326262591e-08,
+      "loss": 1.4061,
+      "step": 4047
+    },
+    {
+      "epoch": 0.8401826484018264,
+      "grad_norm": 0.7004764760883524,
+      "learning_rate": 6.221506551948233e-08,
+      "loss": 1.5352,
+      "step": 4048
+    },
+    {
+      "epoch": 0.840390203403902,
+      "grad_norm": 0.8454728395337476,
+      "learning_rate": 6.215861486230392e-08,
+      "loss": 1.5024,
+      "step": 4049
+    },
+    {
+      "epoch": 0.8405977584059776,
+      "grad_norm": 1.8347602893046522,
+      "learning_rate": 6.210223131509197e-08,
+      "loss": 1.4925,
+      "step": 4050
+    },
+    {
+      "epoch": 0.8408053134080531,
+      "grad_norm": 1.0991041899686835,
+      "learning_rate": 6.204591490181941e-08,
+      "loss": 1.5169,
+      "step": 4051
+    },
+    {
+      "epoch": 0.8410128684101287,
+      "grad_norm": 0.6797311961478647,
+      "learning_rate": 6.198966564643031e-08,
+      "loss": 1.5296,
+      "step": 4052
+    },
+    {
+      "epoch": 0.8412204234122043,
+      "grad_norm": 0.6797181790647425,
+      "learning_rate": 6.193348357284048e-08,
+      "loss": 1.5148,
+      "step": 4053
+    },
+    {
+      "epoch": 0.8414279784142797,
+      "grad_norm": 2.276414250337877,
+      "learning_rate": 6.187736870493699e-08,
+      "loss": 1.5085,
+      "step": 4054
+    },
+    {
+      "epoch": 0.8416355334163553,
+      "grad_norm": 0.8307526684781883,
+      "learning_rate": 6.182132106657839e-08,
+      "loss": 1.5162,
+      "step": 4055
+    },
+    {
+      "epoch": 0.8418430884184309,
+      "grad_norm": 3.8172882819525,
+      "learning_rate": 6.176534068159471e-08,
+      "loss": 1.4723,
+      "step": 4056
+    },
+    {
+      "epoch": 0.8420506434205064,
+      "grad_norm": 0.6544425682927617,
+      "learning_rate": 6.170942757378728e-08,
+      "loss": 1.5078,
+      "step": 4057
+    },
+    {
+      "epoch": 0.842258198422582,
+      "grad_norm": 0.9381897686102861,
+      "learning_rate": 6.165358176692885e-08,
+      "loss": 1.5589,
+      "step": 4058
+    },
+    {
+      "epoch": 0.8424657534246576,
+      "grad_norm": 0.8562378369972459,
+      "learning_rate": 6.159780328476358e-08,
+      "loss": 1.5122,
+      "step": 4059
+    },
+    {
+      "epoch": 0.842673308426733,
+      "grad_norm": 0.7729498223687795,
+      "learning_rate": 6.154209215100709e-08,
+      "loss": 1.5176,
+      "step": 4060
+    },
+    {
+      "epoch": 0.8428808634288086,
+      "grad_norm": 0.7682316854340769,
+      "learning_rate": 6.14864483893461e-08,
+      "loss": 1.4993,
+      "step": 4061
+    },
+    {
+      "epoch": 0.8430884184308842,
+      "grad_norm": 0.7712317870670445,
+      "learning_rate": 6.1430872023439e-08,
+      "loss": 1.4995,
+      "step": 4062
+    },
+    {
+      "epoch": 0.8432959734329597,
+      "grad_norm": 0.8989037497727227,
+      "learning_rate": 6.137536307691535e-08,
+      "loss": 1.5372,
+      "step": 4063
+    },
+    {
+      "epoch": 0.8435035284350353,
+      "grad_norm": 0.6893389098040597,
+      "learning_rate": 6.131992157337608e-08,
+      "loss": 1.5526,
+      "step": 4064
+    },
+    {
+      "epoch": 0.8437110834371109,
+      "grad_norm": 0.7643914499723264,
+      "learning_rate": 6.126454753639342e-08,
+      "loss": 1.4557,
+      "step": 4065
+    },
+    {
+      "epoch": 0.8439186384391864,
+      "grad_norm": 0.6529769647658299,
+      "learning_rate": 6.120924098951102e-08,
+      "loss": 1.4867,
+      "step": 4066
+    },
+    {
+      "epoch": 0.8441261934412619,
+      "grad_norm": 0.8971531483074574,
+      "learning_rate": 6.115400195624363e-08,
+      "loss": 1.5656,
+      "step": 4067
+    },
+    {
+      "epoch": 0.8443337484433375,
+      "grad_norm": 0.7366126592976487,
+      "learning_rate": 6.109883046007749e-08,
+      "loss": 1.4596,
+      "step": 4068
+    },
+    {
+      "epoch": 0.844541303445413,
+      "grad_norm": 0.7260828512171599,
+      "learning_rate": 6.10437265244701e-08,
+      "loss": 1.4864,
+      "step": 4069
+    },
+    {
+      "epoch": 0.8447488584474886,
+      "grad_norm": 1.2388362343258517,
+      "learning_rate": 6.098869017285007e-08,
+      "loss": 1.5321,
+      "step": 4070
+    },
+    {
+      "epoch": 0.8449564134495642,
+      "grad_norm": 0.6813257330196466,
+      "learning_rate": 6.09337214286175e-08,
+      "loss": 1.4887,
+      "step": 4071
+    },
+    {
+      "epoch": 0.8451639684516397,
+      "grad_norm": 0.7391326802325049,
+      "learning_rate": 6.087882031514364e-08,
+      "loss": 1.5583,
+      "step": 4072
+    },
+    {
+      "epoch": 0.8453715234537152,
+      "grad_norm": 0.8887157733844627,
+      "learning_rate": 6.082398685577094e-08,
+      "loss": 1.4544,
+      "step": 4073
+    },
+    {
+      "epoch": 0.8455790784557908,
+      "grad_norm": 0.7114636308472719,
+      "learning_rate": 6.07692210738131e-08,
+      "loss": 1.5509,
+      "step": 4074
+    },
+    {
+      "epoch": 0.8457866334578663,
+      "grad_norm": 0.9378076225994207,
+      "learning_rate": 6.071452299255522e-08,
+      "loss": 1.4951,
+      "step": 4075
+    },
+    {
+      "epoch": 0.8459941884599419,
+      "grad_norm": 0.7049201342496771,
+      "learning_rate": 6.065989263525329e-08,
+      "loss": 1.4885,
+      "step": 4076
+    },
+    {
+      "epoch": 0.8462017434620175,
+      "grad_norm": 0.8435285100069949,
+      "learning_rate": 6.060533002513481e-08,
+      "loss": 1.4984,
+      "step": 4077
+    },
+    {
+      "epoch": 0.846409298464093,
+      "grad_norm": 4.0280121079449716,
+      "learning_rate": 6.055083518539831e-08,
+      "loss": 1.4455,
+      "step": 4078
+    },
+    {
+      "epoch": 0.8466168534661686,
+      "grad_norm": 1.5884707723620994,
+      "learning_rate": 6.049640813921356e-08,
+      "loss": 1.5793,
+      "step": 4079
+    },
+    {
+      "epoch": 0.8468244084682441,
+      "grad_norm": 0.7859921295499089,
+      "learning_rate": 6.044204890972144e-08,
+      "loss": 1.4924,
+      "step": 4080
+    },
+    {
+      "epoch": 0.8470319634703196,
+      "grad_norm": 1.0117214614418315,
+      "learning_rate": 6.038775752003415e-08,
+      "loss": 1.4763,
+      "step": 4081
+    },
+    {
+      "epoch": 0.8472395184723952,
+      "grad_norm": 0.8687583311362441,
+      "learning_rate": 6.033353399323491e-08,
+      "loss": 1.5534,
+      "step": 4082
+    },
+    {
+      "epoch": 0.8474470734744707,
+      "grad_norm": 0.8520361933348216,
+      "learning_rate": 6.027937835237808e-08,
+      "loss": 1.4673,
+      "step": 4083
+    },
+    {
+      "epoch": 0.8476546284765463,
+      "grad_norm": 0.8579679305047718,
+      "learning_rate": 6.022529062048925e-08,
+      "loss": 1.4336,
+      "step": 4084
+    },
+    {
+      "epoch": 0.8478621834786219,
+      "grad_norm": 2.903900444805852,
+      "learning_rate": 6.017127082056505e-08,
+      "loss": 1.5954,
+      "step": 4085
+    },
+    {
+      "epoch": 0.8480697384806973,
+      "grad_norm": 2.115870460475413,
+      "learning_rate": 6.011731897557333e-08,
+      "loss": 1.5902,
+      "step": 4086
+    },
+    {
+      "epoch": 0.8482772934827729,
+      "grad_norm": 1.3854393479840557,
+      "learning_rate": 6.006343510845288e-08,
+      "loss": 1.5892,
+      "step": 4087
+    },
+    {
+      "epoch": 0.8484848484848485,
+      "grad_norm": 0.9180946898742948,
+      "learning_rate": 6.00096192421138e-08,
+      "loss": 1.5293,
+      "step": 4088
+    },
+    {
+      "epoch": 0.848692403486924,
+      "grad_norm": 0.7679457186793486,
+      "learning_rate": 5.995587139943708e-08,
+      "loss": 1.5339,
+      "step": 4089
+    },
+    {
+      "epoch": 0.8488999584889996,
+      "grad_norm": 0.7799100023818198,
+      "learning_rate": 5.990219160327494e-08,
+      "loss": 1.5169,
+      "step": 4090
+    },
+    {
+      "epoch": 0.8491075134910752,
+      "grad_norm": 0.8760153198603559,
+      "learning_rate": 5.984857987645054e-08,
+      "loss": 1.533,
+      "step": 4091
+    },
+    {
+      "epoch": 0.8493150684931506,
+      "grad_norm": 0.6843072573169289,
+      "learning_rate": 5.97950362417582e-08,
+      "loss": 1.5652,
+      "step": 4092
+    },
+    {
+      "epoch": 0.8495226234952262,
+      "grad_norm": 0.9997743818174234,
+      "learning_rate": 5.974156072196326e-08,
+      "loss": 1.5145,
+      "step": 4093
+    },
+    {
+      "epoch": 0.8497301784973018,
+      "grad_norm": 0.9971245097302768,
+      "learning_rate": 5.96881533398021e-08,
+      "loss": 1.4771,
+      "step": 4094
+    },
+    {
+      "epoch": 0.8499377334993773,
+      "grad_norm": 0.9221102322880435,
+      "learning_rate": 5.96348141179821e-08,
+      "loss": 1.527,
+      "step": 4095
+    },
+    {
+      "epoch": 0.8501452885014529,
+      "grad_norm": 1.0828798853149835,
+      "learning_rate": 5.958154307918165e-08,
+      "loss": 1.5759,
+      "step": 4096
+    },
+    {
+      "epoch": 0.8503528435035285,
+      "grad_norm": 1.8271044083505592,
+      "learning_rate": 5.9528340246050245e-08,
+      "loss": 1.4829,
+      "step": 4097
+    },
+    {
+      "epoch": 0.8505603985056039,
+      "grad_norm": 0.6920566831847973,
+      "learning_rate": 5.947520564120829e-08,
+      "loss": 1.444,
+      "step": 4098
+    },
+    {
+      "epoch": 0.8507679535076795,
+      "grad_norm": 0.6859673802190966,
+      "learning_rate": 5.94221392872472e-08,
+      "loss": 1.5767,
+      "step": 4099
+    },
+    {
+      "epoch": 0.8509755085097551,
+      "grad_norm": 0.9842430714096108,
+      "learning_rate": 5.9369141206729345e-08,
+      "loss": 1.478,
+      "step": 4100
+    },
+    {
+      "epoch": 0.8511830635118306,
+      "grad_norm": 0.8889043794707988,
+      "learning_rate": 5.9316211422188224e-08,
+      "loss": 1.6093,
+      "step": 4101
+    },
+    {
+      "epoch": 0.8513906185139062,
+      "grad_norm": 1.3491699629533125,
+      "learning_rate": 5.926334995612802e-08,
+      "loss": 1.5002,
+      "step": 4102
+    },
+    {
+      "epoch": 0.8515981735159818,
+      "grad_norm": 0.659643775828377,
+      "learning_rate": 5.9210556831024115e-08,
+      "loss": 1.4739,
+      "step": 4103
+    },
+    {
+      "epoch": 0.8518057285180572,
+      "grad_norm": 0.7217318821852536,
+      "learning_rate": 5.9157832069322763e-08,
+      "loss": 1.5892,
+      "step": 4104
+    },
+    {
+      "epoch": 0.8520132835201328,
+      "grad_norm": 0.8551784757463411,
+      "learning_rate": 5.910517569344108e-08,
+      "loss": 1.4656,
+      "step": 4105
+    },
+    {
+      "epoch": 0.8522208385222084,
+      "grad_norm": 0.8819767414008842,
+      "learning_rate": 5.905258772576714e-08,
+      "loss": 1.5238,
+      "step": 4106
+    },
+    {
+      "epoch": 0.8524283935242839,
+      "grad_norm": 2.072011996681699,
+      "learning_rate": 5.900006818866003e-08,
+      "loss": 1.5415,
+      "step": 4107
+    },
+    {
+      "epoch": 0.8526359485263595,
+      "grad_norm": 1.3649564615315766,
+      "learning_rate": 5.894761710444961e-08,
+      "loss": 1.5169,
+      "step": 4108
+    },
+    {
+      "epoch": 0.8528435035284351,
+      "grad_norm": 0.6393359145198483,
+      "learning_rate": 5.889523449543672e-08,
+      "loss": 1.5239,
+      "step": 4109
+    },
+    {
+      "epoch": 0.8530510585305106,
+      "grad_norm": 0.7792328380483428,
+      "learning_rate": 5.8842920383893014e-08,
+      "loss": 1.4813,
+      "step": 4110
+    },
+    {
+      "epoch": 0.8532586135325861,
+      "grad_norm": 1.0054966380207717,
+      "learning_rate": 5.879067479206107e-08,
+      "loss": 1.5708,
+      "step": 4111
+    },
+    {
+      "epoch": 0.8534661685346617,
+      "grad_norm": 1.6157621894184846,
+      "learning_rate": 5.873849774215435e-08,
+      "loss": 1.5545,
+      "step": 4112
+    },
+    {
+      "epoch": 0.8536737235367372,
+      "grad_norm": 0.6940989261812918,
+      "learning_rate": 5.8686389256357164e-08,
+      "loss": 1.5008,
+      "step": 4113
+    },
+    {
+      "epoch": 0.8538812785388128,
+      "grad_norm": 0.6655757647220574,
+      "learning_rate": 5.863434935682461e-08,
+      "loss": 1.5302,
+      "step": 4114
+    },
+    {
+      "epoch": 0.8540888335408884,
+      "grad_norm": 1.6014266122475473,
+      "learning_rate": 5.858237806568267e-08,
+      "loss": 1.5292,
+      "step": 4115
+    },
+    {
+      "epoch": 0.8542963885429639,
+      "grad_norm": 1.5053212959295068,
+      "learning_rate": 5.853047540502826e-08,
+      "loss": 1.4585,
+      "step": 4116
+    },
+    {
+      "epoch": 0.8545039435450394,
+      "grad_norm": 1.3008247275325335,
+      "learning_rate": 5.847864139692886e-08,
+      "loss": 1.4977,
+      "step": 4117
+    },
+    {
+      "epoch": 0.854711498547115,
+      "grad_norm": 1.116950383213203,
+      "learning_rate": 5.842687606342301e-08,
+      "loss": 1.4599,
+      "step": 4118
+    },
+    {
+      "epoch": 0.8549190535491905,
+      "grad_norm": 0.8278884111697892,
+      "learning_rate": 5.837517942651996e-08,
+      "loss": 1.4832,
+      "step": 4119
+    },
+    {
+      "epoch": 0.8551266085512661,
+      "grad_norm": 0.8400076822517751,
+      "learning_rate": 5.8323551508199717e-08,
+      "loss": 1.5074,
+      "step": 4120
+    },
+    {
+      "epoch": 0.8553341635533417,
+      "grad_norm": 0.665323811090533,
+      "learning_rate": 5.827199233041306e-08,
+      "loss": 1.5407,
+      "step": 4121
+    },
+    {
+      "epoch": 0.8555417185554172,
+      "grad_norm": 1.086483612788927,
+      "learning_rate": 5.8220501915081674e-08,
+      "loss": 1.5669,
+      "step": 4122
+    },
+    {
+      "epoch": 0.8557492735574928,
+      "grad_norm": 0.9649567223562108,
+      "learning_rate": 5.816908028409787e-08,
+      "loss": 1.514,
+      "step": 4123
+    },
+    {
+      "epoch": 0.8559568285595683,
+      "grad_norm": 0.9446684263406112,
+      "learning_rate": 5.8117727459324726e-08,
+      "loss": 1.4992,
+      "step": 4124
+    },
+    {
+      "epoch": 0.8561643835616438,
+      "grad_norm": 0.997418081391977,
+      "learning_rate": 5.8066443462596216e-08,
+      "loss": 1.4549,
+      "step": 4125
+    },
+    {
+      "epoch": 0.8563719385637194,
+      "grad_norm": 0.6849424593675485,
+      "learning_rate": 5.801522831571677e-08,
+      "loss": 1.5443,
+      "step": 4126
+    },
+    {
+      "epoch": 0.8565794935657949,
+      "grad_norm": 0.651125153577864,
+      "learning_rate": 5.796408204046186e-08,
+      "loss": 1.4817,
+      "step": 4127
+    },
+    {
+      "epoch": 0.8567870485678705,
+      "grad_norm": 2.123893077361472,
+      "learning_rate": 5.791300465857741e-08,
+      "loss": 1.4618,
+      "step": 4128
+    },
+    {
+      "epoch": 0.8569946035699461,
+      "grad_norm": 0.8021709304260402,
+      "learning_rate": 5.786199619178023e-08,
+      "loss": 1.5456,
+      "step": 4129
+    },
+    {
+      "epoch": 0.8572021585720215,
+      "grad_norm": 0.8451273955406761,
+      "learning_rate": 5.781105666175776e-08,
+      "loss": 1.5068,
+      "step": 4130
+    },
+    {
+      "epoch": 0.8574097135740971,
+      "grad_norm": 0.637188116898242,
+      "learning_rate": 5.7760186090168115e-08,
+      "loss": 1.5736,
+      "step": 4131
+    },
+    {
+      "epoch": 0.8576172685761727,
+      "grad_norm": 1.0107153983033976,
+      "learning_rate": 5.770938449864009e-08,
+      "loss": 1.4868,
+      "step": 4132
+    },
+    {
+      "epoch": 0.8578248235782482,
+      "grad_norm": 2.365642913075964,
+      "learning_rate": 5.7658651908773256e-08,
+      "loss": 1.5215,
+      "step": 4133
+    },
+    {
+      "epoch": 0.8580323785803238,
+      "grad_norm": 0.8824901576220718,
+      "learning_rate": 5.7607988342137715e-08,
+      "loss": 1.3926,
+      "step": 4134
+    },
+    {
+      "epoch": 0.8582399335823994,
+      "grad_norm": 0.7803924668582455,
+      "learning_rate": 5.7557393820274204e-08,
+      "loss": 1.4852,
+      "step": 4135
+    },
+    {
+      "epoch": 0.8584474885844748,
+      "grad_norm": 0.6677374403236563,
+      "learning_rate": 5.750686836469433e-08,
+      "loss": 1.5503,
+      "step": 4136
+    },
+    {
+      "epoch": 0.8586550435865504,
+      "grad_norm": 0.6167136380753402,
+      "learning_rate": 5.745641199688001e-08,
+      "loss": 1.5213,
+      "step": 4137
+    },
+    {
+      "epoch": 0.858862598588626,
+      "grad_norm": 1.0645362874290583,
+      "learning_rate": 5.740602473828402e-08,
+      "loss": 1.5123,
+      "step": 4138
+    },
+    {
+      "epoch": 0.8590701535907015,
+      "grad_norm": 0.6499849215289663,
+      "learning_rate": 5.7355706610329716e-08,
+      "loss": 1.5587,
+      "step": 4139
+    },
+    {
+      "epoch": 0.8592777085927771,
+      "grad_norm": 1.1319038296417008,
+      "learning_rate": 5.7305457634411e-08,
+      "loss": 1.5284,
+      "step": 4140
+    },
+    {
+      "epoch": 0.8594852635948527,
+      "grad_norm": 0.804954701781392,
+      "learning_rate": 5.725527783189239e-08,
+      "loss": 1.4607,
+      "step": 4141
+    },
+    {
+      "epoch": 0.8596928185969281,
+      "grad_norm": 0.7438323490016324,
+      "learning_rate": 5.720516722410904e-08,
+      "loss": 1.448,
+      "step": 4142
+    },
+    {
+      "epoch": 0.8599003735990037,
+      "grad_norm": 0.7976820323740148,
+      "learning_rate": 5.71551258323666e-08,
+      "loss": 1.4899,
+      "step": 4143
+    },
+    {
+      "epoch": 0.8601079286010793,
+      "grad_norm": 2.458745574904749,
+      "learning_rate": 5.7105153677941375e-08,
+      "loss": 1.4524,
+      "step": 4144
+    },
+    {
+      "epoch": 0.8603154836031548,
+      "grad_norm": 0.9145255952623502,
+      "learning_rate": 5.705525078208022e-08,
+      "loss": 1.4915,
+      "step": 4145
+    },
+    {
+      "epoch": 0.8605230386052304,
+      "grad_norm": 0.6765272493454979,
+      "learning_rate": 5.700541716600048e-08,
+      "loss": 1.4882,
+      "step": 4146
+    },
+    {
+      "epoch": 0.860730593607306,
+      "grad_norm": 0.7432066502522412,
+      "learning_rate": 5.6955652850890076e-08,
+      "loss": 1.5213,
+      "step": 4147
+    },
+    {
+      "epoch": 0.8609381486093814,
+      "grad_norm": 0.8804565519988562,
+      "learning_rate": 5.690595785790753e-08,
+      "loss": 1.5755,
+      "step": 4148
+    },
+    {
+      "epoch": 0.861145703611457,
+      "grad_norm": 0.8229643361269555,
+      "learning_rate": 5.68563322081818e-08,
+      "loss": 1.509,
+      "step": 4149
+    },
+    {
+      "epoch": 0.8613532586135326,
+      "grad_norm": 0.8743217596728453,
+      "learning_rate": 5.6806775922812364e-08,
+      "loss": 1.4992,
+      "step": 4150
+    },
+    {
+      "epoch": 0.8615608136156081,
+      "grad_norm": 1.0385558238892758,
+      "learning_rate": 5.6757289022869346e-08,
+      "loss": 1.4313,
+      "step": 4151
+    },
+    {
+      "epoch": 0.8617683686176837,
+      "grad_norm": 0.8545273338243716,
+      "learning_rate": 5.670787152939311e-08,
+      "loss": 1.4924,
+      "step": 4152
+    },
+    {
+      "epoch": 0.8619759236197593,
+      "grad_norm": 0.7217691153188052,
+      "learning_rate": 5.6658523463394766e-08,
+      "loss": 1.5483,
+      "step": 4153
+    },
+    {
+      "epoch": 0.8621834786218348,
+      "grad_norm": 0.732358806800885,
+      "learning_rate": 5.660924484585579e-08,
+      "loss": 1.4881,
+      "step": 4154
+    },
+    {
+      "epoch": 0.8623910336239103,
+      "grad_norm": 1.2570711275175508,
+      "learning_rate": 5.6560035697728123e-08,
+      "loss": 1.5465,
+      "step": 4155
+    },
+    {
+      "epoch": 0.8625985886259859,
+      "grad_norm": 0.7396997836660634,
+      "learning_rate": 5.651089603993415e-08,
+      "loss": 1.5714,
+      "step": 4156
+    },
+    {
+      "epoch": 0.8628061436280614,
+      "grad_norm": 1.0294913668569072,
+      "learning_rate": 5.6461825893366874e-08,
+      "loss": 1.4659,
+      "step": 4157
+    },
+    {
+      "epoch": 0.863013698630137,
+      "grad_norm": 2.3175215153323983,
+      "learning_rate": 5.641282527888947e-08,
+      "loss": 1.4965,
+      "step": 4158
+    },
+    {
+      "epoch": 0.8632212536322126,
+      "grad_norm": 0.9819983911736111,
+      "learning_rate": 5.6363894217335803e-08,
+      "loss": 1.5411,
+      "step": 4159
+    },
+    {
+      "epoch": 0.8634288086342881,
+      "grad_norm": 0.7072225792616519,
+      "learning_rate": 5.631503272951001e-08,
+      "loss": 1.4452,
+      "step": 4160
+    },
+    {
+      "epoch": 0.8636363636363636,
+      "grad_norm": 0.9733887170555072,
+      "learning_rate": 5.626624083618668e-08,
+      "loss": 1.6014,
+      "step": 4161
+    },
+    {
+      "epoch": 0.8638439186384392,
+      "grad_norm": 0.813595502246141,
+      "learning_rate": 5.62175185581109e-08,
+      "loss": 1.4301,
+      "step": 4162
+    },
+    {
+      "epoch": 0.8640514736405147,
+      "grad_norm": 0.8372090441523727,
+      "learning_rate": 5.616886591599806e-08,
+      "loss": 1.5035,
+      "step": 4163
+    },
+    {
+      "epoch": 0.8642590286425903,
+      "grad_norm": 0.8064299937919819,
+      "learning_rate": 5.6120282930533944e-08,
+      "loss": 1.5675,
+      "step": 4164
+    },
+    {
+      "epoch": 0.8644665836446659,
+      "grad_norm": 0.9033349856205647,
+      "learning_rate": 5.6071769622374746e-08,
+      "loss": 1.5799,
+      "step": 4165
+    },
+    {
+      "epoch": 0.8646741386467414,
+      "grad_norm": 1.4560557057536148,
+      "learning_rate": 5.602332601214711e-08,
+      "loss": 1.5069,
+      "step": 4166
+    },
+    {
+      "epoch": 0.864881693648817,
+      "grad_norm": 0.7661794136719537,
+      "learning_rate": 5.5974952120447875e-08,
+      "loss": 1.5017,
+      "step": 4167
+    },
+    {
+      "epoch": 0.8650892486508925,
+      "grad_norm": 0.7517363879885167,
+      "learning_rate": 5.59266479678444e-08,
+      "loss": 1.5414,
+      "step": 4168
+    },
+    {
+      "epoch": 0.865296803652968,
+      "grad_norm": 0.6986495433826778,
+      "learning_rate": 5.587841357487431e-08,
+      "loss": 1.5208,
+      "step": 4169
+    },
+    {
+      "epoch": 0.8655043586550436,
+      "grad_norm": 0.7529383038820391,
+      "learning_rate": 5.583024896204559e-08,
+      "loss": 1.6272,
+      "step": 4170
+    },
+    {
+      "epoch": 0.8657119136571192,
+      "grad_norm": 0.6211978460501443,
+      "learning_rate": 5.5782154149836514e-08,
+      "loss": 1.4486,
+      "step": 4171
+    },
+    {
+      "epoch": 0.8659194686591947,
+      "grad_norm": 0.7761169013788679,
+      "learning_rate": 5.5734129158695844e-08,
+      "loss": 1.5324,
+      "step": 4172
+    },
+    {
+      "epoch": 0.8661270236612703,
+      "grad_norm": 0.9257611442368944,
+      "learning_rate": 5.5686174009042364e-08,
+      "loss": 1.5136,
+      "step": 4173
+    },
+    {
+      "epoch": 0.8663345786633457,
+      "grad_norm": 1.0235908485719383,
+      "learning_rate": 5.5638288721265427e-08,
+      "loss": 1.5922,
+      "step": 4174
+    },
+    {
+      "epoch": 0.8665421336654213,
+      "grad_norm": 0.7283903810303131,
+      "learning_rate": 5.5590473315724595e-08,
+      "loss": 1.4859,
+      "step": 4175
+    },
+    {
+      "epoch": 0.8667496886674969,
+      "grad_norm": 0.9405004054817377,
+      "learning_rate": 5.5542727812749645e-08,
+      "loss": 1.4551,
+      "step": 4176
+    },
+    {
+      "epoch": 0.8669572436695724,
+      "grad_norm": 1.1197242467483994,
+      "learning_rate": 5.549505223264081e-08,
+      "loss": 1.4883,
+      "step": 4177
+    },
+    {
+      "epoch": 0.867164798671648,
+      "grad_norm": 0.7700679216421907,
+      "learning_rate": 5.5447446595668327e-08,
+      "loss": 1.5388,
+      "step": 4178
+    },
+    {
+      "epoch": 0.8673723536737236,
+      "grad_norm": 0.6823895391875299,
+      "learning_rate": 5.539991092207296e-08,
+      "loss": 1.4758,
+      "step": 4179
+    },
+    {
+      "epoch": 0.867579908675799,
+      "grad_norm": 1.0629000749707977,
+      "learning_rate": 5.535244523206559e-08,
+      "loss": 1.5405,
+      "step": 4180
+    },
+    {
+      "epoch": 0.8677874636778746,
+      "grad_norm": 0.6778443767062662,
+      "learning_rate": 5.5305049545827366e-08,
+      "loss": 1.4496,
+      "step": 4181
+    },
+    {
+      "epoch": 0.8679950186799502,
+      "grad_norm": 0.9285642258982703,
+      "learning_rate": 5.5257723883509645e-08,
+      "loss": 1.5067,
+      "step": 4182
+    },
+    {
+      "epoch": 0.8682025736820257,
+      "grad_norm": 0.8395923263599718,
+      "learning_rate": 5.521046826523413e-08,
+      "loss": 1.5268,
+      "step": 4183
+    },
+    {
+      "epoch": 0.8684101286841013,
+      "grad_norm": 0.8162135082748619,
+      "learning_rate": 5.516328271109253e-08,
+      "loss": 1.5262,
+      "step": 4184
+    },
+    {
+      "epoch": 0.8686176836861769,
+      "grad_norm": 1.0771736494419826,
+      "learning_rate": 5.5116167241147003e-08,
+      "loss": 1.4944,
+      "step": 4185
+    },
+    {
+      "epoch": 0.8688252386882523,
+      "grad_norm": 0.8292095731176862,
+      "learning_rate": 5.5069121875429724e-08,
+      "loss": 1.5152,
+      "step": 4186
+    },
+    {
+      "epoch": 0.8690327936903279,
+      "grad_norm": 2.0213811479064447,
+      "learning_rate": 5.5022146633943136e-08,
+      "loss": 1.5375,
+      "step": 4187
+    },
+    {
+      "epoch": 0.8692403486924035,
+      "grad_norm": 0.6859199044064056,
+      "learning_rate": 5.4975241536659914e-08,
+      "loss": 1.5366,
+      "step": 4188
+    },
+    {
+      "epoch": 0.869447903694479,
+      "grad_norm": 2.126172541469436,
+      "learning_rate": 5.492840660352285e-08,
+      "loss": 1.5683,
+      "step": 4189
+    },
+    {
+      "epoch": 0.8696554586965546,
+      "grad_norm": 0.7433443582528955,
+      "learning_rate": 5.4881641854444885e-08,
+      "loss": 1.5988,
+      "step": 4190
+    },
+    {
+      "epoch": 0.8698630136986302,
+      "grad_norm": 0.8145355296085374,
+      "learning_rate": 5.483494730930911e-08,
+      "loss": 1.4664,
+      "step": 4191
+    },
+    {
+      "epoch": 0.8700705687007056,
+      "grad_norm": 7.258279082738799,
+      "learning_rate": 5.478832298796895e-08,
+      "loss": 1.4882,
+      "step": 4192
+    },
+    {
+      "epoch": 0.8702781237027812,
+      "grad_norm": 1.081515734963167,
+      "learning_rate": 5.4741768910247665e-08,
+      "loss": 1.4902,
+      "step": 4193
+    },
+    {
+      "epoch": 0.8704856787048568,
+      "grad_norm": 0.9896179450721011,
+      "learning_rate": 5.4695285095938905e-08,
+      "loss": 1.5437,
+      "step": 4194
+    },
+    {
+      "epoch": 0.8706932337069323,
+      "grad_norm": 0.6525901152976705,
+      "learning_rate": 5.464887156480633e-08,
+      "loss": 1.4839,
+      "step": 4195
+    },
+    {
+      "epoch": 0.8709007887090079,
+      "grad_norm": 0.8832141859934202,
+      "learning_rate": 5.460252833658374e-08,
+      "loss": 1.5311,
+      "step": 4196
+    },
+    {
+      "epoch": 0.8711083437110835,
+      "grad_norm": 1.0599630386727497,
+      "learning_rate": 5.455625543097503e-08,
+      "loss": 1.5021,
+      "step": 4197
+    },
+    {
+      "epoch": 0.871315898713159,
+      "grad_norm": 1.42664385688762,
+      "learning_rate": 5.451005286765424e-08,
+      "loss": 1.5206,
+      "step": 4198
+    },
+    {
+      "epoch": 0.8715234537152345,
+      "grad_norm": 0.7265401135034683,
+      "learning_rate": 5.4463920666265493e-08,
+      "loss": 1.5428,
+      "step": 4199
+    },
+    {
+      "epoch": 0.8717310087173101,
+      "grad_norm": 0.7619546583017525,
+      "learning_rate": 5.4417858846422924e-08,
+      "loss": 1.6151,
+      "step": 4200
+    },
+    {
+      "epoch": 0.8719385637193856,
+      "grad_norm": 1.0970789377494663,
+      "learning_rate": 5.437186742771083e-08,
+      "loss": 1.4679,
+      "step": 4201
+    },
+    {
+      "epoch": 0.8721461187214612,
+      "grad_norm": 1.142686229205972,
+      "learning_rate": 5.432594642968349e-08,
+      "loss": 1.5342,
+      "step": 4202
+    },
+    {
+      "epoch": 0.8723536737235368,
+      "grad_norm": 0.663847238892591,
+      "learning_rate": 5.428009587186535e-08,
+      "loss": 1.5082,
+      "step": 4203
+    },
+    {
+      "epoch": 0.8725612287256123,
+      "grad_norm": 4.310129815916642,
+      "learning_rate": 5.423431577375085e-08,
+      "loss": 1.4768,
+      "step": 4204
+    },
+    {
+      "epoch": 0.8727687837276878,
+      "grad_norm": 0.9347674646692149,
+      "learning_rate": 5.418860615480445e-08,
+      "loss": 1.5613,
+      "step": 4205
+    },
+    {
+      "epoch": 0.8729763387297634,
+      "grad_norm": 0.7720927812734257,
+      "learning_rate": 5.414296703446063e-08,
+      "loss": 1.4526,
+      "step": 4206
+    },
+    {
+      "epoch": 0.8731838937318389,
+      "grad_norm": 0.6868919417199573,
+      "learning_rate": 5.409739843212406e-08,
+      "loss": 1.4771,
+      "step": 4207
+    },
+    {
+      "epoch": 0.8733914487339145,
+      "grad_norm": 0.8804363251874378,
+      "learning_rate": 5.405190036716912e-08,
+      "loss": 1.4903,
+      "step": 4208
+    },
+    {
+      "epoch": 0.8735990037359901,
+      "grad_norm": 0.9938283284415154,
+      "learning_rate": 5.4006472858940535e-08,
+      "loss": 1.5267,
+      "step": 4209
+    },
+    {
+      "epoch": 0.8738065587380656,
+      "grad_norm": 0.6702845584703695,
+      "learning_rate": 5.3961115926752786e-08,
+      "loss": 1.497,
+      "step": 4210
+    },
+    {
+      "epoch": 0.8740141137401412,
+      "grad_norm": 0.6828171091381926,
+      "learning_rate": 5.391582958989047e-08,
+      "loss": 1.6409,
+      "step": 4211
+    },
+    {
+      "epoch": 0.8742216687422167,
+      "grad_norm": 0.8883174027937882,
+      "learning_rate": 5.3870613867608104e-08,
+      "loss": 1.5631,
+      "step": 4212
+    },
+    {
+      "epoch": 0.8744292237442922,
+      "grad_norm": 0.6266442412496519,
+      "learning_rate": 5.382546877913026e-08,
+      "loss": 1.536,
+      "step": 4213
+    },
+    {
+      "epoch": 0.8746367787463678,
+      "grad_norm": 2.356246620684954,
+      "learning_rate": 5.378039434365143e-08,
+      "loss": 1.5815,
+      "step": 4214
+    },
+    {
+      "epoch": 0.8748443337484434,
+      "grad_norm": 0.9445070211315894,
+      "learning_rate": 5.3735390580336024e-08,
+      "loss": 1.5322,
+      "step": 4215
+    },
+    {
+      "epoch": 0.8750518887505189,
+      "grad_norm": 0.7320135726769582,
+      "learning_rate": 5.3690457508318464e-08,
+      "loss": 1.4941,
+      "step": 4216
+    },
+    {
+      "epoch": 0.8752594437525945,
+      "grad_norm": 0.7033963475871209,
+      "learning_rate": 5.3645595146703085e-08,
+      "loss": 1.4586,
+      "step": 4217
+    },
+    {
+      "epoch": 0.8754669987546699,
+      "grad_norm": 0.9536858445805583,
+      "learning_rate": 5.3600803514564255e-08,
+      "loss": 1.5208,
+      "step": 4218
+    },
+    {
+      "epoch": 0.8756745537567455,
+      "grad_norm": 0.9280960572717364,
+      "learning_rate": 5.355608263094607e-08,
+      "loss": 1.5253,
+      "step": 4219
+    },
+    {
+      "epoch": 0.8758821087588211,
+      "grad_norm": 0.7964899446883809,
+      "learning_rate": 5.351143251486271e-08,
+      "loss": 1.5894,
+      "step": 4220
+    },
+    {
+      "epoch": 0.8760896637608966,
+      "grad_norm": 0.6131247619568557,
+      "learning_rate": 5.3466853185298265e-08,
+      "loss": 1.5307,
+      "step": 4221
+    },
+    {
+      "epoch": 0.8762972187629722,
+      "grad_norm": 0.8473869228543437,
+      "learning_rate": 5.342234466120662e-08,
+      "loss": 1.5252,
+      "step": 4222
+    },
+    {
+      "epoch": 0.8765047737650478,
+      "grad_norm": 0.7112192747153537,
+      "learning_rate": 5.337790696151162e-08,
+      "loss": 1.5337,
+      "step": 4223
+    },
+    {
+      "epoch": 0.8767123287671232,
+      "grad_norm": 0.979106709730265,
+      "learning_rate": 5.333354010510703e-08,
+      "loss": 1.5374,
+      "step": 4224
+    },
+    {
+      "epoch": 0.8769198837691988,
+      "grad_norm": 0.672210708302282,
+      "learning_rate": 5.3289244110856456e-08,
+      "loss": 1.4724,
+      "step": 4225
+    },
+    {
+      "epoch": 0.8771274387712744,
+      "grad_norm": 0.7382489583601339,
+      "learning_rate": 5.324501899759336e-08,
+      "loss": 1.5052,
+      "step": 4226
+    },
+    {
+      "epoch": 0.8773349937733499,
+      "grad_norm": 0.648877679246535,
+      "learning_rate": 5.32008647841211e-08,
+      "loss": 1.4578,
+      "step": 4227
+    },
+    {
+      "epoch": 0.8775425487754255,
+      "grad_norm": 0.6891513737866265,
+      "learning_rate": 5.315678148921284e-08,
+      "loss": 1.5207,
+      "step": 4228
+    },
+    {
+      "epoch": 0.8777501037775011,
+      "grad_norm": 0.6335163523907139,
+      "learning_rate": 5.311276913161169e-08,
+      "loss": 1.5482,
+      "step": 4229
+    },
+    {
+      "epoch": 0.8779576587795765,
+      "grad_norm": 0.6845397792008311,
+      "learning_rate": 5.306882773003047e-08,
+      "loss": 1.462,
+      "step": 4230
+    },
+    {
+      "epoch": 0.8781652137816521,
+      "grad_norm": 0.9797956229546194,
+      "learning_rate": 5.302495730315196e-08,
+      "loss": 1.4755,
+      "step": 4231
+    },
+    {
+      "epoch": 0.8783727687837277,
+      "grad_norm": 2.4637326093339227,
+      "learning_rate": 5.2981157869628646e-08,
+      "loss": 1.5182,
+      "step": 4232
+    },
+    {
+      "epoch": 0.8785803237858032,
+      "grad_norm": 0.6523987461597316,
+      "learning_rate": 5.293742944808296e-08,
+      "loss": 1.5461,
+      "step": 4233
+    },
+    {
+      "epoch": 0.8787878787878788,
+      "grad_norm": 0.9102005775740785,
+      "learning_rate": 5.2893772057106945e-08,
+      "loss": 1.5329,
+      "step": 4234
+    },
+    {
+      "epoch": 0.8789954337899544,
+      "grad_norm": 0.6878377862214508,
+      "learning_rate": 5.285018571526266e-08,
+      "loss": 1.4771,
+      "step": 4235
+    },
+    {
+      "epoch": 0.8792029887920298,
+      "grad_norm": 1.0666386433823898,
+      "learning_rate": 5.280667044108184e-08,
+      "loss": 1.4533,
+      "step": 4236
+    },
+    {
+      "epoch": 0.8794105437941054,
+      "grad_norm": 0.9152677055223459,
+      "learning_rate": 5.276322625306601e-08,
+      "loss": 1.5399,
+      "step": 4237
+    },
+    {
+      "epoch": 0.879618098796181,
+      "grad_norm": 0.7648227800062127,
+      "learning_rate": 5.271985316968647e-08,
+      "loss": 1.5596,
+      "step": 4238
+    },
+    {
+      "epoch": 0.8798256537982565,
+      "grad_norm": 0.7166525771998916,
+      "learning_rate": 5.267655120938434e-08,
+      "loss": 1.5551,
+      "step": 4239
+    },
+    {
+      "epoch": 0.8800332088003321,
+      "grad_norm": 0.8567833446848079,
+      "learning_rate": 5.263332039057048e-08,
+      "loss": 1.5274,
+      "step": 4240
+    },
+    {
+      "epoch": 0.8802407638024077,
+      "grad_norm": 0.759213438469565,
+      "learning_rate": 5.259016073162541e-08,
+      "loss": 1.5716,
+      "step": 4241
+    },
+    {
+      "epoch": 0.8804483188044832,
+      "grad_norm": 0.9688616544910612,
+      "learning_rate": 5.254707225089958e-08,
+      "loss": 1.5298,
+      "step": 4242
+    },
+    {
+      "epoch": 0.8806558738065587,
+      "grad_norm": 0.8737294001400681,
+      "learning_rate": 5.250405496671296e-08,
+      "loss": 1.463,
+      "step": 4243
+    },
+    {
+      "epoch": 0.8808634288086343,
+      "grad_norm": 0.8053355337454994,
+      "learning_rate": 5.246110889735541e-08,
+      "loss": 1.4607,
+      "step": 4244
+    },
+    {
+      "epoch": 0.8810709838107098,
+      "grad_norm": 1.295230301483268,
+      "learning_rate": 5.2418234061086476e-08,
+      "loss": 1.4871,
+      "step": 4245
+    },
+    {
+      "epoch": 0.8812785388127854,
+      "grad_norm": 0.678528741831449,
+      "learning_rate": 5.2375430476135374e-08,
+      "loss": 1.4612,
+      "step": 4246
+    },
+    {
+      "epoch": 0.881486093814861,
+      "grad_norm": 0.9622688739992937,
+      "learning_rate": 5.2332698160701e-08,
+      "loss": 1.4754,
+      "step": 4247
+    },
+    {
+      "epoch": 0.8816936488169365,
+      "grad_norm": 0.7119298611054514,
+      "learning_rate": 5.229003713295213e-08,
+      "loss": 1.5235,
+      "step": 4248
+    },
+    {
+      "epoch": 0.881901203819012,
+      "grad_norm": 1.05046913972188,
+      "learning_rate": 5.224744741102697e-08,
+      "loss": 1.5191,
+      "step": 4249
+    },
+    {
+      "epoch": 0.8821087588210876,
+      "grad_norm": 0.7666164712001088,
+      "learning_rate": 5.220492901303362e-08,
+      "loss": 1.492,
+      "step": 4250
+    },
+    {
+      "epoch": 0.8823163138231631,
+      "grad_norm": 0.7700859781958104,
+      "learning_rate": 5.2162481957049754e-08,
+      "loss": 1.4919,
+      "step": 4251
+    },
+    {
+      "epoch": 0.8825238688252387,
+      "grad_norm": 0.7334359996695763,
+      "learning_rate": 5.212010626112271e-08,
+      "loss": 1.535,
+      "step": 4252
+    },
+    {
+      "epoch": 0.8827314238273143,
+      "grad_norm": 0.7693179701976535,
+      "learning_rate": 5.207780194326952e-08,
+      "loss": 1.4861,
+      "step": 4253
+    },
+    {
+      "epoch": 0.8829389788293898,
+      "grad_norm": 0.7190063652592567,
+      "learning_rate": 5.2035569021476866e-08,
+      "loss": 1.5472,
+      "step": 4254
+    },
+    {
+      "epoch": 0.8831465338314654,
+      "grad_norm": 0.6821158850330902,
+      "learning_rate": 5.199340751370107e-08,
+      "loss": 1.4748,
+      "step": 4255
+    },
+    {
+      "epoch": 0.8833540888335409,
+      "grad_norm": 1.5571151444271683,
+      "learning_rate": 5.195131743786807e-08,
+      "loss": 1.5677,
+      "step": 4256
+    },
+    {
+      "epoch": 0.8835616438356164,
+      "grad_norm": 0.8142576940205603,
+      "learning_rate": 5.1909298811873484e-08,
+      "loss": 1.4703,
+      "step": 4257
+    },
+    {
+      "epoch": 0.883769198837692,
+      "grad_norm": 0.7192879281080775,
+      "learning_rate": 5.1867351653582434e-08,
+      "loss": 1.4792,
+      "step": 4258
+    },
+    {
+      "epoch": 0.8839767538397676,
+      "grad_norm": 0.8739303422913324,
+      "learning_rate": 5.182547598082983e-08,
+      "loss": 1.5629,
+      "step": 4259
+    },
+    {
+      "epoch": 0.8841843088418431,
+      "grad_norm": 0.6737205626905679,
+      "learning_rate": 5.178367181142007e-08,
+      "loss": 1.5414,
+      "step": 4260
+    },
+    {
+      "epoch": 0.8843918638439187,
+      "grad_norm": 0.7835482444865381,
+      "learning_rate": 5.174193916312717e-08,
+      "loss": 1.5058,
+      "step": 4261
+    },
+    {
+      "epoch": 0.8845994188459941,
+      "grad_norm": 1.3818172229022785,
+      "learning_rate": 5.1700278053694724e-08,
+      "loss": 1.5235,
+      "step": 4262
+    },
+    {
+      "epoch": 0.8848069738480697,
+      "grad_norm": 0.8367357269564156,
+      "learning_rate": 5.1658688500835937e-08,
+      "loss": 1.467,
+      "step": 4263
+    },
+    {
+      "epoch": 0.8850145288501453,
+      "grad_norm": 0.7080073868049961,
+      "learning_rate": 5.161717052223358e-08,
+      "loss": 1.5468,
+      "step": 4264
+    },
+    {
+      "epoch": 0.8852220838522208,
+      "grad_norm": 0.6422585614531965,
+      "learning_rate": 5.1575724135540044e-08,
+      "loss": 1.6077,
+      "step": 4265
+    },
+    {
+      "epoch": 0.8854296388542964,
+      "grad_norm": 0.7018188776745039,
+      "learning_rate": 5.153434935837717e-08,
+      "loss": 1.5084,
+      "step": 4266
+    },
+    {
+      "epoch": 0.885637193856372,
+      "grad_norm": 0.7996096745563676,
+      "learning_rate": 5.1493046208336434e-08,
+      "loss": 1.4675,
+      "step": 4267
+    },
+    {
+      "epoch": 0.8858447488584474,
+      "grad_norm": 0.8817282853164217,
+      "learning_rate": 5.145181470297888e-08,
+      "loss": 1.4802,
+      "step": 4268
+    },
+    {
+      "epoch": 0.886052303860523,
+      "grad_norm": 3.3772671493344615,
+      "learning_rate": 5.141065485983497e-08,
+      "loss": 1.4752,
+      "step": 4269
+    },
+    {
+      "epoch": 0.8862598588625986,
+      "grad_norm": 0.8372472219496861,
+      "learning_rate": 5.136956669640486e-08,
+      "loss": 1.4755,
+      "step": 4270
+    },
+    {
+      "epoch": 0.8864674138646741,
+      "grad_norm": 0.7506953327093302,
+      "learning_rate": 5.132855023015809e-08,
+      "loss": 1.5137,
+      "step": 4271
+    },
+    {
+      "epoch": 0.8866749688667497,
+      "grad_norm": 0.8936346567465829,
+      "learning_rate": 5.128760547853379e-08,
+      "loss": 1.486,
+      "step": 4272
+    },
+    {
+      "epoch": 0.8868825238688253,
+      "grad_norm": 1.296900704664816,
+      "learning_rate": 5.124673245894058e-08,
+      "loss": 1.4206,
+      "step": 4273
+    },
+    {
+      "epoch": 0.8870900788709007,
+      "grad_norm": 0.8835573287414936,
+      "learning_rate": 5.120593118875661e-08,
+      "loss": 1.4375,
+      "step": 4274
+    },
+    {
+      "epoch": 0.8872976338729763,
+      "grad_norm": 0.8481941939830817,
+      "learning_rate": 5.116520168532942e-08,
+      "loss": 1.525,
+      "step": 4275
+    },
+    {
+      "epoch": 0.8875051888750519,
+      "grad_norm": 2.2552084923334887,
+      "learning_rate": 5.11245439659762e-08,
+      "loss": 1.4724,
+      "step": 4276
+    },
+    {
+      "epoch": 0.8877127438771274,
+      "grad_norm": 0.8629327342298814,
+      "learning_rate": 5.1083958047983486e-08,
+      "loss": 1.5254,
+      "step": 4277
+    },
+    {
+      "epoch": 0.887920298879203,
+      "grad_norm": 0.9498131123080511,
+      "learning_rate": 5.104344394860736e-08,
+      "loss": 1.5334,
+      "step": 4278
+    },
+    {
+      "epoch": 0.8881278538812786,
+      "grad_norm": 1.7009220910789542,
+      "learning_rate": 5.100300168507328e-08,
+      "loss": 1.505,
+      "step": 4279
+    },
+    {
+      "epoch": 0.888335408883354,
+      "grad_norm": 0.7809552615299162,
+      "learning_rate": 5.096263127457631e-08,
+      "loss": 1.5391,
+      "step": 4280
+    },
+    {
+      "epoch": 0.8885429638854296,
+      "grad_norm": 0.7497246029049855,
+      "learning_rate": 5.0922332734280836e-08,
+      "loss": 1.4543,
+      "step": 4281
+    },
+    {
+      "epoch": 0.8887505188875052,
+      "grad_norm": 0.782328456157103,
+      "learning_rate": 5.0882106081320694e-08,
+      "loss": 1.5103,
+      "step": 4282
+    },
+    {
+      "epoch": 0.8889580738895807,
+      "grad_norm": 0.7978758306197025,
+      "learning_rate": 5.084195133279927e-08,
+      "loss": 1.5552,
+      "step": 4283
+    },
+    {
+      "epoch": 0.8891656288916563,
+      "grad_norm": 1.1181287179920762,
+      "learning_rate": 5.0801868505789205e-08,
+      "loss": 1.5197,
+      "step": 4284
+    },
+    {
+      "epoch": 0.8893731838937319,
+      "grad_norm": 0.6434283809211656,
+      "learning_rate": 5.0761857617332705e-08,
+      "loss": 1.5485,
+      "step": 4285
+    },
+    {
+      "epoch": 0.8895807388958074,
+      "grad_norm": 0.8840134229109191,
+      "learning_rate": 5.0721918684441356e-08,
+      "loss": 1.5788,
+      "step": 4286
+    },
+    {
+      "epoch": 0.8897882938978829,
+      "grad_norm": 0.6680170235180583,
+      "learning_rate": 5.0682051724096084e-08,
+      "loss": 1.5123,
+      "step": 4287
+    },
+    {
+      "epoch": 0.8899958488999585,
+      "grad_norm": 0.6609811219082681,
+      "learning_rate": 5.0642256753247256e-08,
+      "loss": 1.5164,
+      "step": 4288
+    },
+    {
+      "epoch": 0.890203403902034,
+      "grad_norm": 1.2699579751064243,
+      "learning_rate": 5.060253378881474e-08,
+      "loss": 1.5474,
+      "step": 4289
+    },
+    {
+      "epoch": 0.8904109589041096,
+      "grad_norm": 0.946140842710006,
+      "learning_rate": 5.056288284768753e-08,
+      "loss": 1.5962,
+      "step": 4290
+    },
+    {
+      "epoch": 0.8906185139061852,
+      "grad_norm": 0.8625508751684643,
+      "learning_rate": 5.052330394672428e-08,
+      "loss": 1.4837,
+      "step": 4291
+    },
+    {
+      "epoch": 0.8908260689082607,
+      "grad_norm": 3.111788008305834,
+      "learning_rate": 5.0483797102752834e-08,
+      "loss": 1.5295,
+      "step": 4292
+    },
+    {
+      "epoch": 0.8910336239103362,
+      "grad_norm": 0.8459879573160543,
+      "learning_rate": 5.044436233257042e-08,
+      "loss": 1.5793,
+      "step": 4293
+    },
+    {
+      "epoch": 0.8912411789124118,
+      "grad_norm": 0.7554870079899078,
+      "learning_rate": 5.0404999652943735e-08,
+      "loss": 1.5324,
+      "step": 4294
+    },
+    {
+      "epoch": 0.8914487339144873,
+      "grad_norm": 0.7717021565993143,
+      "learning_rate": 5.036570908060871e-08,
+      "loss": 1.515,
+      "step": 4295
+    },
+    {
+      "epoch": 0.8916562889165629,
+      "grad_norm": 0.7529179131308154,
+      "learning_rate": 5.0326490632270656e-08,
+      "loss": 1.5593,
+      "step": 4296
+    },
+    {
+      "epoch": 0.8918638439186385,
+      "grad_norm": 0.6936682708787681,
+      "learning_rate": 5.028734432460418e-08,
+      "loss": 1.5714,
+      "step": 4297
+    },
+    {
+      "epoch": 0.892071398920714,
+      "grad_norm": 1.3458158128488977,
+      "learning_rate": 5.024827017425331e-08,
+      "loss": 1.5148,
+      "step": 4298
+    },
+    {
+      "epoch": 0.8922789539227896,
+      "grad_norm": 0.7696684986420448,
+      "learning_rate": 5.020926819783128e-08,
+      "loss": 1.5205,
+      "step": 4299
+    },
+    {
+      "epoch": 0.8924865089248651,
+      "grad_norm": 0.8115595098586696,
+      "learning_rate": 5.0170338411920745e-08,
+      "loss": 1.4317,
+      "step": 4300
+    },
+    {
+      "epoch": 0.8926940639269406,
+      "grad_norm": 0.7534127642869912,
+      "learning_rate": 5.01314808330736e-08,
+      "loss": 1.5335,
+      "step": 4301
+    },
+    {
+      "epoch": 0.8929016189290162,
+      "grad_norm": 0.8573846778963077,
+      "learning_rate": 5.0092695477811057e-08,
+      "loss": 1.6471,
+      "step": 4302
+    },
+    {
+      "epoch": 0.8931091739310918,
+      "grad_norm": 1.4395934480749717,
+      "learning_rate": 5.005398236262358e-08,
+      "loss": 1.5053,
+      "step": 4303
+    },
+    {
+      "epoch": 0.8933167289331673,
+      "grad_norm": 0.9472802638245992,
+      "learning_rate": 5.001534150397101e-08,
+      "loss": 1.5163,
+      "step": 4304
+    },
+    {
+      "epoch": 0.8935242839352429,
+      "grad_norm": 1.284152263952106,
+      "learning_rate": 4.9976772918282406e-08,
+      "loss": 1.5548,
+      "step": 4305
+    },
+    {
+      "epoch": 0.8937318389373183,
+      "grad_norm": 0.8089470550323055,
+      "learning_rate": 4.99382766219561e-08,
+      "loss": 1.4239,
+      "step": 4306
+    },
+    {
+      "epoch": 0.8939393939393939,
+      "grad_norm": 0.7458479682006363,
+      "learning_rate": 4.989985263135968e-08,
+      "loss": 1.4975,
+      "step": 4307
+    },
+    {
+      "epoch": 0.8941469489414695,
+      "grad_norm": 0.8121655497119886,
+      "learning_rate": 4.9861500962830014e-08,
+      "loss": 1.5591,
+      "step": 4308
+    },
+    {
+      "epoch": 0.894354503943545,
+      "grad_norm": 0.8018688010174477,
+      "learning_rate": 4.982322163267326e-08,
+      "loss": 1.5427,
+      "step": 4309
+    },
+    {
+      "epoch": 0.8945620589456206,
+      "grad_norm": 0.6794273813769126,
+      "learning_rate": 4.978501465716468e-08,
+      "loss": 1.5165,
+      "step": 4310
+    },
+    {
+      "epoch": 0.8947696139476962,
+      "grad_norm": 0.725473907679013,
+      "learning_rate": 4.9746880052548935e-08,
+      "loss": 1.4709,
+      "step": 4311
+    },
+    {
+      "epoch": 0.8949771689497716,
+      "grad_norm": 0.9676880458658392,
+      "learning_rate": 4.970881783503982e-08,
+      "loss": 1.589,
+      "step": 4312
+    },
+    {
+      "epoch": 0.8951847239518472,
+      "grad_norm": 0.7389128274610071,
+      "learning_rate": 4.967082802082041e-08,
+      "loss": 1.581,
+      "step": 4313
+    },
+    {
+      "epoch": 0.8953922789539228,
+      "grad_norm": 0.8186399866900692,
+      "learning_rate": 4.96329106260429e-08,
+      "loss": 1.5291,
+      "step": 4314
+    },
+    {
+      "epoch": 0.8955998339559983,
+      "grad_norm": 1.6937377868442203,
+      "learning_rate": 4.9595065666828864e-08,
+      "loss": 1.5178,
+      "step": 4315
+    },
+    {
+      "epoch": 0.8958073889580739,
+      "grad_norm": 1.1622595347845388,
+      "learning_rate": 4.955729315926886e-08,
+      "loss": 1.5696,
+      "step": 4316
+    },
+    {
+      "epoch": 0.8960149439601495,
+      "grad_norm": 0.7733428171985816,
+      "learning_rate": 4.9519593119422846e-08,
+      "loss": 1.5337,
+      "step": 4317
+    },
+    {
+      "epoch": 0.896222498962225,
+      "grad_norm": 0.8729457673286071,
+      "learning_rate": 4.948196556331982e-08,
+      "loss": 1.5659,
+      "step": 4318
+    },
+    {
+      "epoch": 0.8964300539643005,
+      "grad_norm": 0.636443077605352,
+      "learning_rate": 4.944441050695802e-08,
+      "loss": 1.516,
+      "step": 4319
+    },
+    {
+      "epoch": 0.8966376089663761,
+      "grad_norm": 0.8005755156950772,
+      "learning_rate": 4.940692796630491e-08,
+      "loss": 1.4816,
+      "step": 4320
+    },
+    {
+      "epoch": 0.8968451639684516,
+      "grad_norm": 0.9880259238605373,
+      "learning_rate": 4.936951795729704e-08,
+      "loss": 1.5539,
+      "step": 4321
+    },
+    {
+      "epoch": 0.8970527189705272,
+      "grad_norm": 0.692931295818655,
+      "learning_rate": 4.9332180495840136e-08,
+      "loss": 1.5093,
+      "step": 4322
+    },
+    {
+      "epoch": 0.8972602739726028,
+      "grad_norm": 0.738273417602925,
+      "learning_rate": 4.929491559780911e-08,
+      "loss": 1.5313,
+      "step": 4323
+    },
+    {
+      "epoch": 0.8974678289746783,
+      "grad_norm": 0.8776384646402227,
+      "learning_rate": 4.925772327904805e-08,
+      "loss": 1.5398,
+      "step": 4324
+    },
+    {
+      "epoch": 0.8976753839767538,
+      "grad_norm": 1.7551359055823745,
+      "learning_rate": 4.922060355537005e-08,
+      "loss": 1.4829,
+      "step": 4325
+    },
+    {
+      "epoch": 0.8978829389788294,
+      "grad_norm": 3.2203706962528575,
+      "learning_rate": 4.918355644255752e-08,
+      "loss": 1.529,
+      "step": 4326
+    },
+    {
+      "epoch": 0.8980904939809049,
+      "grad_norm": 1.6691959832247925,
+      "learning_rate": 4.9146581956361864e-08,
+      "loss": 1.5196,
+      "step": 4327
+    },
+    {
+      "epoch": 0.8982980489829805,
+      "grad_norm": 0.7695873416536831,
+      "learning_rate": 4.910968011250366e-08,
+      "loss": 1.5075,
+      "step": 4328
+    },
+    {
+      "epoch": 0.8985056039850561,
+      "grad_norm": 0.7215491884709342,
+      "learning_rate": 4.9072850926672564e-08,
+      "loss": 1.4768,
+      "step": 4329
+    },
+    {
+      "epoch": 0.8987131589871316,
+      "grad_norm": 13.009121681595024,
+      "learning_rate": 4.9036094414527416e-08,
+      "loss": 1.5481,
+      "step": 4330
+    },
+    {
+      "epoch": 0.8989207139892071,
+      "grad_norm": 0.7662619366580208,
+      "learning_rate": 4.899941059169611e-08,
+      "loss": 1.5058,
+      "step": 4331
+    },
+    {
+      "epoch": 0.8991282689912827,
+      "grad_norm": 0.9438494218201771,
+      "learning_rate": 4.8962799473775576e-08,
+      "loss": 1.5007,
+      "step": 4332
+    },
+    {
+      "epoch": 0.8993358239933582,
+      "grad_norm": 1.8020124466171614,
+      "learning_rate": 4.8926261076331954e-08,
+      "loss": 1.443,
+      "step": 4333
+    },
+    {
+      "epoch": 0.8995433789954338,
+      "grad_norm": 1.343314488193815,
+      "learning_rate": 4.888979541490035e-08,
+      "loss": 1.4764,
+      "step": 4334
+    },
+    {
+      "epoch": 0.8997509339975094,
+      "grad_norm": 0.9158077720735733,
+      "learning_rate": 4.8853402504985026e-08,
+      "loss": 1.4863,
+      "step": 4335
+    },
+    {
+      "epoch": 0.8999584889995849,
+      "grad_norm": 0.7047215744402243,
+      "learning_rate": 4.88170823620593e-08,
+      "loss": 1.5387,
+      "step": 4336
+    },
+    {
+      "epoch": 0.9001660440016604,
+      "grad_norm": 0.6709317440385352,
+      "learning_rate": 4.878083500156548e-08,
+      "loss": 1.4327,
+      "step": 4337
+    },
+    {
+      "epoch": 0.900373599003736,
+      "grad_norm": 0.8814860386970792,
+      "learning_rate": 4.8744660438914985e-08,
+      "loss": 1.4798,
+      "step": 4338
+    },
+    {
+      "epoch": 0.9005811540058115,
+      "grad_norm": 0.7987821173935699,
+      "learning_rate": 4.870855868948837e-08,
+      "loss": 1.5077,
+      "step": 4339
+    },
+    {
+      "epoch": 0.9007887090078871,
+      "grad_norm": 0.756473745359541,
+      "learning_rate": 4.867252976863499e-08,
+      "loss": 1.4111,
+      "step": 4340
+    },
+    {
+      "epoch": 0.9009962640099627,
+      "grad_norm": 0.6532856202952518,
+      "learning_rate": 4.863657369167351e-08,
+      "loss": 1.4461,
+      "step": 4341
+    },
+    {
+      "epoch": 0.9012038190120382,
+      "grad_norm": 0.6751058802893536,
+      "learning_rate": 4.860069047389147e-08,
+      "loss": 1.5752,
+      "step": 4342
+    },
+    {
+      "epoch": 0.9014113740141138,
+      "grad_norm": 1.3548043307678506,
+      "learning_rate": 4.856488013054543e-08,
+      "loss": 1.482,
+      "step": 4343
+    },
+    {
+      "epoch": 0.9016189290161893,
+      "grad_norm": 0.7146013599961456,
+      "learning_rate": 4.852914267686099e-08,
+      "loss": 1.4116,
+      "step": 4344
+    },
+    {
+      "epoch": 0.9018264840182648,
+      "grad_norm": 0.6792004305762552,
+      "learning_rate": 4.849347812803281e-08,
+      "loss": 1.5328,
+      "step": 4345
+    },
+    {
+      "epoch": 0.9020340390203404,
+      "grad_norm": 1.1622597544407174,
+      "learning_rate": 4.8457886499224496e-08,
+      "loss": 1.5107,
+      "step": 4346
+    },
+    {
+      "epoch": 0.902241594022416,
+      "grad_norm": 0.7438104866212684,
+      "learning_rate": 4.842236780556864e-08,
+      "loss": 1.4275,
+      "step": 4347
+    },
+    {
+      "epoch": 0.9024491490244915,
+      "grad_norm": 0.7361792801280723,
+      "learning_rate": 4.838692206216692e-08,
+      "loss": 1.4535,
+      "step": 4348
+    },
+    {
+      "epoch": 0.9026567040265671,
+      "grad_norm": 0.7286256834292965,
+      "learning_rate": 4.83515492840898e-08,
+      "loss": 1.4551,
+      "step": 4349
+    },
+    {
+      "epoch": 0.9028642590286425,
+      "grad_norm": 0.7656211505890863,
+      "learning_rate": 4.8316249486377e-08,
+      "loss": 1.5635,
+      "step": 4350
+    },
+    {
+      "epoch": 0.9030718140307181,
+      "grad_norm": 0.6991044769892722,
+      "learning_rate": 4.82810226840369e-08,
+      "loss": 1.5301,
+      "step": 4351
+    },
+    {
+      "epoch": 0.9032793690327937,
+      "grad_norm": 0.718697650781546,
+      "learning_rate": 4.824586889204711e-08,
+      "loss": 1.5897,
+      "step": 4352
+    },
+    {
+      "epoch": 0.9034869240348692,
+      "grad_norm": 3.9182666062595737,
+      "learning_rate": 4.8210788125354046e-08,
+      "loss": 1.4579,
+      "step": 4353
+    },
+    {
+      "epoch": 0.9036944790369448,
+      "grad_norm": 0.7970377583883338,
+      "learning_rate": 4.8175780398873125e-08,
+      "loss": 1.5459,
+      "step": 4354
+    },
+    {
+      "epoch": 0.9039020340390204,
+      "grad_norm": 0.6824980370755335,
+      "learning_rate": 4.814084572748869e-08,
+      "loss": 1.3993,
+      "step": 4355
+    },
+    {
+      "epoch": 0.9041095890410958,
+      "grad_norm": 0.7123105604691621,
+      "learning_rate": 4.810598412605407e-08,
+      "loss": 1.493,
+      "step": 4356
+    },
+    {
+      "epoch": 0.9043171440431714,
+      "grad_norm": 1.7779741114233782,
+      "learning_rate": 4.807119560939146e-08,
+      "loss": 1.5502,
+      "step": 4357
+    },
+    {
+      "epoch": 0.904524699045247,
+      "grad_norm": 0.7390264874137673,
+      "learning_rate": 4.803648019229204e-08,
+      "loss": 1.5403,
+      "step": 4358
+    },
+    {
+      "epoch": 0.9047322540473225,
+      "grad_norm": 0.7988710148506181,
+      "learning_rate": 4.8001837889515864e-08,
+      "loss": 1.465,
+      "step": 4359
+    },
+    {
+      "epoch": 0.9049398090493981,
+      "grad_norm": 0.8442705619661417,
+      "learning_rate": 4.796726871579192e-08,
+      "loss": 1.4121,
+      "step": 4360
+    },
+    {
+      "epoch": 0.9051473640514737,
+      "grad_norm": 0.9624261441467818,
+      "learning_rate": 4.793277268581811e-08,
+      "loss": 1.4451,
+      "step": 4361
+    },
+    {
+      "epoch": 0.9053549190535491,
+      "grad_norm": 0.9471352482548796,
+      "learning_rate": 4.789834981426124e-08,
+      "loss": 1.4793,
+      "step": 4362
+    },
+    {
+      "epoch": 0.9055624740556247,
+      "grad_norm": 0.7201880730417008,
+      "learning_rate": 4.7864000115757e-08,
+      "loss": 1.4578,
+      "step": 4363
+    },
+    {
+      "epoch": 0.9057700290577003,
+      "grad_norm": 0.786463621275674,
+      "learning_rate": 4.782972360490992e-08,
+      "loss": 1.5344,
+      "step": 4364
+    },
+    {
+      "epoch": 0.9059775840597758,
+      "grad_norm": 0.7245361586970211,
+      "learning_rate": 4.7795520296293553e-08,
+      "loss": 1.5746,
+      "step": 4365
+    },
+    {
+      "epoch": 0.9061851390618514,
+      "grad_norm": 1.2410605215183514,
+      "learning_rate": 4.776139020445016e-08,
+      "loss": 1.5514,
+      "step": 4366
+    },
+    {
+      "epoch": 0.906392694063927,
+      "grad_norm": 0.7714388110442567,
+      "learning_rate": 4.7727333343890974e-08,
+      "loss": 1.5348,
+      "step": 4367
+    },
+    {
+      "epoch": 0.9066002490660025,
+      "grad_norm": 0.8684083889947328,
+      "learning_rate": 4.769334972909608e-08,
+      "loss": 1.506,
+      "step": 4368
+    },
+    {
+      "epoch": 0.906807804068078,
+      "grad_norm": 0.8549172423272855,
+      "learning_rate": 4.765943937451439e-08,
+      "loss": 1.483,
+      "step": 4369
+    },
+    {
+      "epoch": 0.9070153590701536,
+      "grad_norm": 0.6870905683449732,
+      "learning_rate": 4.762560229456369e-08,
+      "loss": 1.5133,
+      "step": 4370
+    },
+    {
+      "epoch": 0.9072229140722291,
+      "grad_norm": 0.8551743345917164,
+      "learning_rate": 4.759183850363062e-08,
+      "loss": 1.606,
+      "step": 4371
+    },
+    {
+      "epoch": 0.9074304690743047,
+      "grad_norm": 1.6168826855365306,
+      "learning_rate": 4.755814801607065e-08,
+      "loss": 1.5089,
+      "step": 4372
+    },
+    {
+      "epoch": 0.9076380240763803,
+      "grad_norm": 0.6422541228698613,
+      "learning_rate": 4.752453084620806e-08,
+      "loss": 1.5377,
+      "step": 4373
+    },
+    {
+      "epoch": 0.9078455790784558,
+      "grad_norm": 0.7689861188126695,
+      "learning_rate": 4.749098700833603e-08,
+      "loss": 1.483,
+      "step": 4374
+    },
+    {
+      "epoch": 0.9080531340805313,
+      "grad_norm": 0.6978198724548461,
+      "learning_rate": 4.7457516516716414e-08,
+      "loss": 1.5303,
+      "step": 4375
+    },
+    {
+      "epoch": 0.9082606890826069,
+      "grad_norm": 1.0136882312727724,
+      "learning_rate": 4.7424119385580055e-08,
+      "loss": 1.5679,
+      "step": 4376
+    },
+    {
+      "epoch": 0.9084682440846824,
+      "grad_norm": 0.6774693709160086,
+      "learning_rate": 4.739079562912651e-08,
+      "loss": 1.5668,
+      "step": 4377
+    },
+    {
+      "epoch": 0.908675799086758,
+      "grad_norm": 1.1064660155775856,
+      "learning_rate": 4.7357545261524166e-08,
+      "loss": 1.5283,
+      "step": 4378
+    },
+    {
+      "epoch": 0.9088833540888336,
+      "grad_norm": 0.727948921900572,
+      "learning_rate": 4.732436829691015e-08,
+      "loss": 1.5968,
+      "step": 4379
+    },
+    {
+      "epoch": 0.9090909090909091,
+      "grad_norm": 0.7389830990205652,
+      "learning_rate": 4.729126474939049e-08,
+      "loss": 1.5618,
+      "step": 4380
+    },
+    {
+      "epoch": 0.9092984640929846,
+      "grad_norm": 0.7569380462722427,
+      "learning_rate": 4.725823463303986e-08,
+      "loss": 1.5163,
+      "step": 4381
+    },
+    {
+      "epoch": 0.9095060190950602,
+      "grad_norm": 0.6847799619270005,
+      "learning_rate": 4.7225277961901865e-08,
+      "loss": 1.5,
+      "step": 4382
+    },
+    {
+      "epoch": 0.9097135740971357,
+      "grad_norm": 0.9145135943101996,
+      "learning_rate": 4.719239474998875e-08,
+      "loss": 1.5592,
+      "step": 4383
+    },
+    {
+      "epoch": 0.9099211290992113,
+      "grad_norm": 0.7878203473287412,
+      "learning_rate": 4.7159585011281635e-08,
+      "loss": 1.5378,
+      "step": 4384
+    },
+    {
+      "epoch": 0.9101286841012869,
+      "grad_norm": 0.6200429902153209,
+      "learning_rate": 4.712684875973028e-08,
+      "loss": 1.4947,
+      "step": 4385
+    },
+    {
+      "epoch": 0.9103362391033624,
+      "grad_norm": 1.3305272680240208,
+      "learning_rate": 4.7094186009253336e-08,
+      "loss": 1.5159,
+      "step": 4386
+    },
+    {
+      "epoch": 0.910543794105438,
+      "grad_norm": 0.7071682533579644,
+      "learning_rate": 4.706159677373812e-08,
+      "loss": 1.4443,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9107513491075135,
+      "grad_norm": 1.8221901111293077,
+      "learning_rate": 4.7029081067040666e-08,
+      "loss": 1.5619,
+      "step": 4388
+    },
+    {
+      "epoch": 0.910958904109589,
+      "grad_norm": 1.0410088942310942,
+      "learning_rate": 4.6996638902985883e-08,
+      "loss": 1.5018,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9111664591116646,
+      "grad_norm": 1.2034139004525068,
+      "learning_rate": 4.696427029536721e-08,
+      "loss": 1.4891,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9113740141137402,
+      "grad_norm": 0.7881662048489497,
+      "learning_rate": 4.6931975257946985e-08,
+      "loss": 1.5807,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9115815691158157,
+      "grad_norm": 1.8912080648551375,
+      "learning_rate": 4.689975380445619e-08,
+      "loss": 1.4143,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9117891241178913,
+      "grad_norm": 0.7055284395849184,
+      "learning_rate": 4.686760594859454e-08,
+      "loss": 1.5589,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9119966791199668,
+      "grad_norm": 1.0525802899971646,
+      "learning_rate": 4.6835531704030396e-08,
+      "loss": 1.5496,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9122042341220423,
+      "grad_norm": 0.6931283373747346,
+      "learning_rate": 4.680353108440098e-08,
+      "loss": 1.5093,
+      "step": 4395
+    },
+    {
+      "epoch": 0.9124117891241179,
+      "grad_norm": 0.828222493701835,
+      "learning_rate": 4.677160410331199e-08,
+      "loss": 1.4444,
+      "step": 4396
+    },
+    {
+      "epoch": 0.9126193441261934,
+      "grad_norm": 0.7121260738688048,
+      "learning_rate": 4.6739750774338006e-08,
+      "loss": 1.6235,
+      "step": 4397
+    },
+    {
+      "epoch": 0.912826899128269,
+      "grad_norm": 0.8476768761382423,
+      "learning_rate": 4.6707971111022194e-08,
+      "loss": 1.5912,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9130344541303446,
+      "grad_norm": 2.139204798723291,
+      "learning_rate": 4.6676265126876386e-08,
+      "loss": 1.455,
+      "step": 4399
+    },
+    {
+      "epoch": 0.91324200913242,
+      "grad_norm": 0.720492289256062,
+      "learning_rate": 4.664463283538122e-08,
+      "loss": 1.5401,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9134495641344956,
+      "grad_norm": 0.7742920956672075,
+      "learning_rate": 4.661307424998579e-08,
+      "loss": 1.5944,
+      "step": 4401
+    },
+    {
+      "epoch": 0.9136571191365712,
+      "grad_norm": 1.2314237156426655,
+      "learning_rate": 4.658158938410809e-08,
+      "loss": 1.4757,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9138646741386467,
+      "grad_norm": 0.7277071162954373,
+      "learning_rate": 4.6550178251134556e-08,
+      "loss": 1.5239,
+      "step": 4403
+    },
+    {
+      "epoch": 0.9140722291407223,
+      "grad_norm": 0.8245201853069499,
+      "learning_rate": 4.6518840864420406e-08,
+      "loss": 1.4398,
+      "step": 4404
+    },
+    {
+      "epoch": 0.9142797841427979,
+      "grad_norm": 1.3045080232070827,
+      "learning_rate": 4.6487577237289424e-08,
+      "loss": 1.5627,
+      "step": 4405
+    },
+    {
+      "epoch": 0.9144873391448733,
+      "grad_norm": 0.7077548303895802,
+      "learning_rate": 4.645638738303416e-08,
+      "loss": 1.5299,
+      "step": 4406
+    },
+    {
+      "epoch": 0.9146948941469489,
+      "grad_norm": 0.8372044048656329,
+      "learning_rate": 4.642527131491562e-08,
+      "loss": 1.5105,
+      "step": 4407
+    },
+    {
+      "epoch": 0.9149024491490245,
+      "grad_norm": 0.9066838762925182,
+      "learning_rate": 4.6394229046163595e-08,
+      "loss": 1.5328,
+      "step": 4408
+    },
+    {
+      "epoch": 0.9151100041511,
+      "grad_norm": 0.7633254579902081,
+      "learning_rate": 4.6363260589976394e-08,
+      "loss": 1.5325,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9153175591531756,
+      "grad_norm": 0.9357551080452527,
+      "learning_rate": 4.633236595952098e-08,
+      "loss": 1.5307,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9155251141552512,
+      "grad_norm": 0.9034473550145689,
+      "learning_rate": 4.630154516793299e-08,
+      "loss": 1.5008,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9157326691573267,
+      "grad_norm": 0.7076322259730242,
+      "learning_rate": 4.627079822831653e-08,
+      "loss": 1.4441,
+      "step": 4412
+    },
+    {
+      "epoch": 0.9159402241594022,
+      "grad_norm": 0.9026288025768455,
+      "learning_rate": 4.624012515374444e-08,
+      "loss": 1.4793,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9161477791614778,
+      "grad_norm": 0.6874945841040789,
+      "learning_rate": 4.620952595725803e-08,
+      "loss": 1.4911,
+      "step": 4414
+    },
+    {
+      "epoch": 0.9163553341635533,
+      "grad_norm": 1.0048841346332247,
+      "learning_rate": 4.617900065186737e-08,
+      "loss": 1.509,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9165628891656289,
+      "grad_norm": 0.6583404609257811,
+      "learning_rate": 4.614854925055089e-08,
+      "loss": 1.5402,
+      "step": 4416
+    },
+    {
+      "epoch": 0.9167704441677045,
+      "grad_norm": 1.6156209724642725,
+      "learning_rate": 4.611817176625581e-08,
+      "loss": 1.5041,
+      "step": 4417
+    },
+    {
+      "epoch": 0.91697799916978,
+      "grad_norm": 6.318478167668886,
+      "learning_rate": 4.6087868211897785e-08,
+      "loss": 1.4936,
+      "step": 4418
+    },
+    {
+      "epoch": 0.9171855541718555,
+      "grad_norm": 0.987265882766978,
+      "learning_rate": 4.605763860036108e-08,
+      "loss": 1.5105,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9173931091739311,
+      "grad_norm": 0.7562715301866854,
+      "learning_rate": 4.602748294449855e-08,
+      "loss": 1.5229,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9176006641760066,
+      "grad_norm": 0.7504194150683303,
+      "learning_rate": 4.599740125713158e-08,
+      "loss": 1.483,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9178082191780822,
+      "grad_norm": 0.650245880355937,
+      "learning_rate": 4.596739355105005e-08,
+      "loss": 1.5077,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9180157741801578,
+      "grad_norm": 1.2719317873896605,
+      "learning_rate": 4.593745983901249e-08,
+      "loss": 1.6539,
+      "step": 4423
+    },
+    {
+      "epoch": 0.9182233291822333,
+      "grad_norm": 1.1612098878021526,
+      "learning_rate": 4.590760013374593e-08,
+      "loss": 1.5066,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9184308841843088,
+      "grad_norm": 0.7345640594038324,
+      "learning_rate": 4.587781444794588e-08,
+      "loss": 1.5099,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9186384391863844,
+      "grad_norm": 0.7000581237384536,
+      "learning_rate": 4.5848102794276454e-08,
+      "loss": 1.4999,
+      "step": 4426
+    },
+    {
+      "epoch": 0.9188459941884599,
+      "grad_norm": 1.3712296747194503,
+      "learning_rate": 4.581846518537025e-08,
+      "loss": 1.4945,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9190535491905355,
+      "grad_norm": 1.0385679378460553,
+      "learning_rate": 4.578890163382841e-08,
+      "loss": 1.4996,
+      "step": 4428
+    },
+    {
+      "epoch": 0.9192611041926111,
+      "grad_norm": 0.8330339988495031,
+      "learning_rate": 4.575941215222051e-08,
+      "loss": 1.5503,
+      "step": 4429
+    },
+    {
+      "epoch": 0.9194686591946866,
+      "grad_norm": 0.8311479711107309,
+      "learning_rate": 4.5729996753084796e-08,
+      "loss": 1.5274,
+      "step": 4430
+    },
+    {
+      "epoch": 0.9196762141967622,
+      "grad_norm": 1.044686514166469,
+      "learning_rate": 4.5700655448927805e-08,
+      "loss": 1.5151,
+      "step": 4431
+    },
+    {
+      "epoch": 0.9198837691988377,
+      "grad_norm": 1.2782924304520138,
+      "learning_rate": 4.567138825222475e-08,
+      "loss": 1.5257,
+      "step": 4432
+    },
+    {
+      "epoch": 0.9200913242009132,
+      "grad_norm": 2.04488419397859,
+      "learning_rate": 4.564219517541926e-08,
+      "loss": 1.532,
+      "step": 4433
+    },
+    {
+      "epoch": 0.9202988792029888,
+      "grad_norm": 0.9057868866762059,
+      "learning_rate": 4.561307623092343e-08,
+      "loss": 1.5895,
+      "step": 4434
+    },
+    {
+      "epoch": 0.9205064342050644,
+      "grad_norm": 0.8695854615889368,
+      "learning_rate": 4.558403143111788e-08,
+      "loss": 1.5115,
+      "step": 4435
+    },
+    {
+      "epoch": 0.9207139892071399,
+      "grad_norm": 0.7957806638665895,
+      "learning_rate": 4.5555060788351695e-08,
+      "loss": 1.5068,
+      "step": 4436
+    },
+    {
+      "epoch": 0.9209215442092155,
+      "grad_norm": 0.9864650291143578,
+      "learning_rate": 4.552616431494241e-08,
+      "loss": 1.5335,
+      "step": 4437
+    },
+    {
+      "epoch": 0.921129099211291,
+      "grad_norm": 0.8536182415804855,
+      "learning_rate": 4.549734202317604e-08,
+      "loss": 1.499,
+      "step": 4438
+    },
+    {
+      "epoch": 0.9213366542133665,
+      "grad_norm": 1.3897147528792686,
+      "learning_rate": 4.5468593925307064e-08,
+      "loss": 1.4869,
+      "step": 4439
+    },
+    {
+      "epoch": 0.9215442092154421,
+      "grad_norm": 0.8138634818407828,
+      "learning_rate": 4.5439920033558376e-08,
+      "loss": 1.5216,
+      "step": 4440
+    },
+    {
+      "epoch": 0.9217517642175176,
+      "grad_norm": 0.8327162883012469,
+      "learning_rate": 4.541132036012145e-08,
+      "loss": 1.5013,
+      "step": 4441
+    },
+    {
+      "epoch": 0.9219593192195932,
+      "grad_norm": 0.7633244001951912,
+      "learning_rate": 4.5382794917155963e-08,
+      "loss": 1.3942,
+      "step": 4442
+    },
+    {
+      "epoch": 0.9221668742216688,
+      "grad_norm": 1.2163544308978829,
+      "learning_rate": 4.535434371679029e-08,
+      "loss": 1.5024,
+      "step": 4443
+    },
+    {
+      "epoch": 0.9223744292237442,
+      "grad_norm": 1.1302019272247883,
+      "learning_rate": 4.532596677112111e-08,
+      "loss": 1.4594,
+      "step": 4444
+    },
+    {
+      "epoch": 0.9225819842258198,
+      "grad_norm": 0.7303077583103256,
+      "learning_rate": 4.529766409221351e-08,
+      "loss": 1.5442,
+      "step": 4445
+    },
+    {
+      "epoch": 0.9227895392278954,
+      "grad_norm": 1.3541486344650249,
+      "learning_rate": 4.526943569210102e-08,
+      "loss": 1.4617,
+      "step": 4446
+    },
+    {
+      "epoch": 0.9229970942299709,
+      "grad_norm": 0.9248102150018691,
+      "learning_rate": 4.524128158278569e-08,
+      "loss": 1.4724,
+      "step": 4447
+    },
+    {
+      "epoch": 0.9232046492320465,
+      "grad_norm": 0.7525579147298637,
+      "learning_rate": 4.521320177623784e-08,
+      "loss": 1.5254,
+      "step": 4448
+    },
+    {
+      "epoch": 0.9234122042341221,
+      "grad_norm": 2.0492512128914906,
+      "learning_rate": 4.5185196284396255e-08,
+      "loss": 1.4891,
+      "step": 4449
+    },
+    {
+      "epoch": 0.9236197592361975,
+      "grad_norm": 1.1699401786400996,
+      "learning_rate": 4.515726511916815e-08,
+      "loss": 1.4862,
+      "step": 4450
+    },
+    {
+      "epoch": 0.9238273142382731,
+      "grad_norm": 1.190228317672257,
+      "learning_rate": 4.5129408292429097e-08,
+      "loss": 1.4071,
+      "step": 4451
+    },
+    {
+      "epoch": 0.9240348692403487,
+      "grad_norm": 1.228477886821952,
+      "learning_rate": 4.510162581602309e-08,
+      "loss": 1.5247,
+      "step": 4452
+    },
+    {
+      "epoch": 0.9242424242424242,
+      "grad_norm": 1.2742643777187674,
+      "learning_rate": 4.50739177017625e-08,
+      "loss": 1.5843,
+      "step": 4453
+    },
+    {
+      "epoch": 0.9244499792444998,
+      "grad_norm": 3.340894858996516,
+      "learning_rate": 4.5046283961428095e-08,
+      "loss": 1.506,
+      "step": 4454
+    },
+    {
+      "epoch": 0.9246575342465754,
+      "grad_norm": 0.6847573499275801,
+      "learning_rate": 4.501872460676895e-08,
+      "loss": 1.5187,
+      "step": 4455
+    },
+    {
+      "epoch": 0.9248650892486509,
+      "grad_norm": 1.9571099592762493,
+      "learning_rate": 4.499123964950266e-08,
+      "loss": 1.5183,
+      "step": 4456
+    },
+    {
+      "epoch": 0.9250726442507264,
+      "grad_norm": 1.1258851354934087,
+      "learning_rate": 4.496382910131502e-08,
+      "loss": 1.5013,
+      "step": 4457
+    },
+    {
+      "epoch": 0.925280199252802,
+      "grad_norm": 0.8065146095577503,
+      "learning_rate": 4.493649297386033e-08,
+      "loss": 1.5093,
+      "step": 4458
+    },
+    {
+      "epoch": 0.9254877542548775,
+      "grad_norm": 0.6998701312752142,
+      "learning_rate": 4.490923127876115e-08,
+      "loss": 1.4497,
+      "step": 4459
+    },
+    {
+      "epoch": 0.9256953092569531,
+      "grad_norm": 2.77959072017406,
+      "learning_rate": 4.488204402760843e-08,
+      "loss": 1.4958,
+      "step": 4460
+    },
+    {
+      "epoch": 0.9259028642590287,
+      "grad_norm": 0.9388118863179752,
+      "learning_rate": 4.485493123196144e-08,
+      "loss": 1.5113,
+      "step": 4461
+    },
+    {
+      "epoch": 0.9261104192611042,
+      "grad_norm": 2.2936421316010955,
+      "learning_rate": 4.482789290334789e-08,
+      "loss": 1.5426,
+      "step": 4462
+    },
+    {
+      "epoch": 0.9263179742631797,
+      "grad_norm": 0.702620256173264,
+      "learning_rate": 4.4800929053263714e-08,
+      "loss": 1.4416,
+      "step": 4463
+    },
+    {
+      "epoch": 0.9265255292652553,
+      "grad_norm": 0.6802808803940142,
+      "learning_rate": 4.477403969317323e-08,
+      "loss": 1.6134,
+      "step": 4464
+    },
+    {
+      "epoch": 0.9267330842673308,
+      "grad_norm": 0.9739628418057668,
+      "learning_rate": 4.47472248345091e-08,
+      "loss": 1.4952,
+      "step": 4465
+    },
+    {
+      "epoch": 0.9269406392694064,
+      "grad_norm": 0.9030463701204992,
+      "learning_rate": 4.472048448867225e-08,
+      "loss": 1.4775,
+      "step": 4466
+    },
+    {
+      "epoch": 0.927148194271482,
+      "grad_norm": 0.6996844976075115,
+      "learning_rate": 4.4693818667032e-08,
+      "loss": 1.5166,
+      "step": 4467
+    },
+    {
+      "epoch": 0.9273557492735575,
+      "grad_norm": 0.7571359355755798,
+      "learning_rate": 4.4667227380925945e-08,
+      "loss": 1.5114,
+      "step": 4468
+    },
+    {
+      "epoch": 0.927563304275633,
+      "grad_norm": 0.6672144682852137,
+      "learning_rate": 4.464071064165998e-08,
+      "loss": 1.5223,
+      "step": 4469
+    },
+    {
+      "epoch": 0.9277708592777086,
+      "grad_norm": 0.7942084696423082,
+      "learning_rate": 4.461426846050831e-08,
+      "loss": 1.4548,
+      "step": 4470
+    },
+    {
+      "epoch": 0.9279784142797841,
+      "grad_norm": 0.658150520069637,
+      "learning_rate": 4.45879008487135e-08,
+      "loss": 1.507,
+      "step": 4471
+    },
+    {
+      "epoch": 0.9281859692818597,
+      "grad_norm": 0.7082920604084778,
+      "learning_rate": 4.4561607817486283e-08,
+      "loss": 1.5786,
+      "step": 4472
+    },
+    {
+      "epoch": 0.9283935242839353,
+      "grad_norm": 0.7296630743884838,
+      "learning_rate": 4.453538937800581e-08,
+      "loss": 1.4484,
+      "step": 4473
+    },
+    {
+      "epoch": 0.9286010792860108,
+      "grad_norm": 0.7233965176401961,
+      "learning_rate": 4.450924554141948e-08,
+      "loss": 1.5375,
+      "step": 4474
+    },
+    {
+      "epoch": 0.9288086342880864,
+      "grad_norm": 0.8294890166298233,
+      "learning_rate": 4.448317631884292e-08,
+      "loss": 1.5272,
+      "step": 4475
+    },
+    {
+      "epoch": 0.9290161892901619,
+      "grad_norm": 0.7815812696781098,
+      "learning_rate": 4.4457181721360046e-08,
+      "loss": 1.5248,
+      "step": 4476
+    },
+    {
+      "epoch": 0.9292237442922374,
+      "grad_norm": 0.699343076994957,
+      "learning_rate": 4.4431261760023145e-08,
+      "loss": 1.5114,
+      "step": 4477
+    },
+    {
+      "epoch": 0.929431299294313,
+      "grad_norm": 0.7149120920009966,
+      "learning_rate": 4.4405416445852646e-08,
+      "loss": 1.5059,
+      "step": 4478
+    },
+    {
+      "epoch": 0.9296388542963886,
+      "grad_norm": 1.4018707708059848,
+      "learning_rate": 4.437964578983729e-08,
+      "loss": 1.5493,
+      "step": 4479
+    },
+    {
+      "epoch": 0.9298464092984641,
+      "grad_norm": 0.6541224093567589,
+      "learning_rate": 4.4353949802934124e-08,
+      "loss": 1.527,
+      "step": 4480
+    },
+    {
+      "epoch": 0.9300539643005397,
+      "grad_norm": 0.7940169131847479,
+      "learning_rate": 4.4328328496068323e-08,
+      "loss": 1.5077,
+      "step": 4481
+    },
+    {
+      "epoch": 0.9302615193026152,
+      "grad_norm": 1.3773760777592015,
+      "learning_rate": 4.430278188013347e-08,
+      "loss": 1.5272,
+      "step": 4482
+    },
+    {
+      "epoch": 0.9304690743046907,
+      "grad_norm": 0.8855725936660176,
+      "learning_rate": 4.4277309965991236e-08,
+      "loss": 1.4773,
+      "step": 4483
+    },
+    {
+      "epoch": 0.9306766293067663,
+      "grad_norm": 0.8090724338889203,
+      "learning_rate": 4.4251912764471656e-08,
+      "loss": 1.5352,
+      "step": 4484
+    },
+    {
+      "epoch": 0.9308841843088418,
+      "grad_norm": 0.8313199205017122,
+      "learning_rate": 4.422659028637291e-08,
+      "loss": 1.4942,
+      "step": 4485
+    },
+    {
+      "epoch": 0.9310917393109174,
+      "grad_norm": 1.8701926240939024,
+      "learning_rate": 4.420134254246144e-08,
+      "loss": 1.5,
+      "step": 4486
+    },
+    {
+      "epoch": 0.931299294312993,
+      "grad_norm": 1.313129285023572,
+      "learning_rate": 4.417616954347194e-08,
+      "loss": 1.5197,
+      "step": 4487
+    },
+    {
+      "epoch": 0.9315068493150684,
+      "grad_norm": 0.6029976800612497,
+      "learning_rate": 4.4151071300107296e-08,
+      "loss": 1.5494,
+      "step": 4488
+    },
+    {
+      "epoch": 0.931714404317144,
+      "grad_norm": 0.7550787286907598,
+      "learning_rate": 4.412604782303862e-08,
+      "loss": 1.4717,
+      "step": 4489
+    },
+    {
+      "epoch": 0.9319219593192196,
+      "grad_norm": 0.7139714846761307,
+      "learning_rate": 4.410109912290521e-08,
+      "loss": 1.4775,
+      "step": 4490
+    },
+    {
+      "epoch": 0.9321295143212951,
+      "grad_norm": 0.901604077229633,
+      "learning_rate": 4.407622521031462e-08,
+      "loss": 1.434,
+      "step": 4491
+    },
+    {
+      "epoch": 0.9323370693233707,
+      "grad_norm": 0.9259749545321699,
+      "learning_rate": 4.405142609584253e-08,
+      "loss": 1.5553,
+      "step": 4492
+    },
+    {
+      "epoch": 0.9325446243254463,
+      "grad_norm": 1.0513330819349027,
+      "learning_rate": 4.402670179003292e-08,
+      "loss": 1.4863,
+      "step": 4493
+    },
+    {
+      "epoch": 0.9327521793275217,
+      "grad_norm": 1.4548972607670645,
+      "learning_rate": 4.4002052303397874e-08,
+      "loss": 1.4719,
+      "step": 4494
+    },
+    {
+      "epoch": 0.9329597343295973,
+      "grad_norm": 2.2831818349145996,
+      "learning_rate": 4.3977477646417714e-08,
+      "loss": 1.5404,
+      "step": 4495
+    },
+    {
+      "epoch": 0.9331672893316729,
+      "grad_norm": 0.7028828157279985,
+      "learning_rate": 4.395297782954091e-08,
+      "loss": 1.4762,
+      "step": 4496
+    },
+    {
+      "epoch": 0.9333748443337484,
+      "grad_norm": 0.7034628389845509,
+      "learning_rate": 4.392855286318419e-08,
+      "loss": 1.5368,
+      "step": 4497
+    },
+    {
+      "epoch": 0.933582399335824,
+      "grad_norm": 0.8850500424659637,
+      "learning_rate": 4.390420275773232e-08,
+      "loss": 1.5443,
+      "step": 4498
+    },
+    {
+      "epoch": 0.9337899543378996,
+      "grad_norm": 0.72343686302418,
+      "learning_rate": 4.3879927523538366e-08,
+      "loss": 1.5123,
+      "step": 4499
+    },
+    {
+      "epoch": 0.933997509339975,
+      "grad_norm": 0.6702840746003527,
+      "learning_rate": 4.385572717092353e-08,
+      "loss": 1.5527,
+      "step": 4500
+    },
+    {
+      "epoch": 0.9342050643420506,
+      "grad_norm": 0.8031675225917405,
+      "learning_rate": 4.3831601710177126e-08,
+      "loss": 1.4747,
+      "step": 4501
+    },
+    {
+      "epoch": 0.9344126193441262,
+      "grad_norm": 0.721780908832118,
+      "learning_rate": 4.380755115155666e-08,
+      "loss": 1.5649,
+      "step": 4502
+    },
+    {
+      "epoch": 0.9346201743462017,
+      "grad_norm": 0.7118293261803608,
+      "learning_rate": 4.378357550528781e-08,
+      "loss": 1.6267,
+      "step": 4503
+    },
+    {
+      "epoch": 0.9348277293482773,
+      "grad_norm": 0.8200651442469405,
+      "learning_rate": 4.375967478156437e-08,
+      "loss": 1.4658,
+      "step": 4504
+    },
+    {
+      "epoch": 0.9350352843503529,
+      "grad_norm": 0.7175265262449997,
+      "learning_rate": 4.3735848990548274e-08,
+      "loss": 1.5391,
+      "step": 4505
+    },
+    {
+      "epoch": 0.9352428393524284,
+      "grad_norm": 0.7726362716001752,
+      "learning_rate": 4.3712098142369694e-08,
+      "loss": 1.5698,
+      "step": 4506
+    },
+    {
+      "epoch": 0.935450394354504,
+      "grad_norm": 0.9930249508022962,
+      "learning_rate": 4.368842224712677e-08,
+      "loss": 1.564,
+      "step": 4507
+    },
+    {
+      "epoch": 0.9356579493565795,
+      "grad_norm": 0.8468251701210507,
+      "learning_rate": 4.36648213148859e-08,
+      "loss": 1.5035,
+      "step": 4508
+    },
+    {
+      "epoch": 0.935865504358655,
+      "grad_norm": 1.0619639715631535,
+      "learning_rate": 4.364129535568159e-08,
+      "loss": 1.4753,
+      "step": 4509
+    },
+    {
+      "epoch": 0.9360730593607306,
+      "grad_norm": 0.7544660832297827,
+      "learning_rate": 4.3617844379516424e-08,
+      "loss": 1.5263,
+      "step": 4510
+    },
+    {
+      "epoch": 0.9362806143628062,
+      "grad_norm": 0.7925913449770202,
+      "learning_rate": 4.359446839636114e-08,
+      "loss": 1.5363,
+      "step": 4511
+    },
+    {
+      "epoch": 0.9364881693648817,
+      "grad_norm": 1.2563411978038899,
+      "learning_rate": 4.357116741615463e-08,
+      "loss": 1.5103,
+      "step": 4512
+    },
+    {
+      "epoch": 0.9366957243669572,
+      "grad_norm": 1.205123315595704,
+      "learning_rate": 4.3547941448803794e-08,
+      "loss": 1.5272,
+      "step": 4513
+    },
+    {
+      "epoch": 0.9369032793690328,
+      "grad_norm": 0.6445394889101358,
+      "learning_rate": 4.3524790504183716e-08,
+      "loss": 1.6343,
+      "step": 4514
+    },
+    {
+      "epoch": 0.9371108343711083,
+      "grad_norm": 2.3546511600674465,
+      "learning_rate": 4.3501714592137555e-08,
+      "loss": 1.4781,
+      "step": 4515
+    },
+    {
+      "epoch": 0.9373183893731839,
+      "grad_norm": 0.9086640031498722,
+      "learning_rate": 4.3478713722476587e-08,
+      "loss": 1.4756,
+      "step": 4516
+    },
+    {
+      "epoch": 0.9375259443752595,
+      "grad_norm": 0.6822374602427158,
+      "learning_rate": 4.345578790498019e-08,
+      "loss": 1.5227,
+      "step": 4517
+    },
+    {
+      "epoch": 0.937733499377335,
+      "grad_norm": 1.1028352204265481,
+      "learning_rate": 4.3432937149395786e-08,
+      "loss": 1.5268,
+      "step": 4518
+    },
+    {
+      "epoch": 0.9379410543794106,
+      "grad_norm": 0.9282871089810995,
+      "learning_rate": 4.341016146543892e-08,
+      "loss": 1.4393,
+      "step": 4519
+    },
+    {
+      "epoch": 0.9381486093814861,
+      "grad_norm": 0.7863922224590355,
+      "learning_rate": 4.338746086279317e-08,
+      "loss": 1.4801,
+      "step": 4520
+    },
+    {
+      "epoch": 0.9383561643835616,
+      "grad_norm": 0.8555634570162212,
+      "learning_rate": 4.336483535111032e-08,
+      "loss": 1.5035,
+      "step": 4521
+    },
+    {
+      "epoch": 0.9385637193856372,
+      "grad_norm": 1.6854804852988134,
+      "learning_rate": 4.334228494001006e-08,
+      "loss": 1.6103,
+      "step": 4522
+    },
+    {
+      "epoch": 0.9387712743877128,
+      "grad_norm": 0.6385385269374086,
+      "learning_rate": 4.331980963908024e-08,
+      "loss": 1.5397,
+      "step": 4523
+    },
+    {
+      "epoch": 0.9389788293897883,
+      "grad_norm": 0.9300061362751645,
+      "learning_rate": 4.3297409457876784e-08,
+      "loss": 1.5995,
+      "step": 4524
+    },
+    {
+      "epoch": 0.9391863843918639,
+      "grad_norm": 0.9375377838751968,
+      "learning_rate": 4.327508440592362e-08,
+      "loss": 1.5737,
+      "step": 4525
+    },
+    {
+      "epoch": 0.9393939393939394,
+      "grad_norm": 0.6737914930537501,
+      "learning_rate": 4.325283449271279e-08,
+      "loss": 1.5171,
+      "step": 4526
+    },
+    {
+      "epoch": 0.9396014943960149,
+      "grad_norm": 1.136481574205765,
+      "learning_rate": 4.323065972770438e-08,
+      "loss": 1.5142,
+      "step": 4527
+    },
+    {
+      "epoch": 0.9398090493980905,
+      "grad_norm": 1.0050799981727405,
+      "learning_rate": 4.3208560120326475e-08,
+      "loss": 1.4586,
+      "step": 4528
+    },
+    {
+      "epoch": 0.940016604400166,
+      "grad_norm": 0.9607987975142752,
+      "learning_rate": 4.318653567997527e-08,
+      "loss": 1.5697,
+      "step": 4529
+    },
+    {
+      "epoch": 0.9402241594022416,
+      "grad_norm": 1.0471143595993813,
+      "learning_rate": 4.316458641601497e-08,
+      "loss": 1.5214,
+      "step": 4530
+    },
+    {
+      "epoch": 0.9404317144043172,
+      "grad_norm": 0.706593188312328,
+      "learning_rate": 4.3142712337777806e-08,
+      "loss": 1.552,
+      "step": 4531
+    },
+    {
+      "epoch": 0.9406392694063926,
+      "grad_norm": 0.7029313367378158,
+      "learning_rate": 4.3120913454564064e-08,
+      "loss": 1.5417,
+      "step": 4532
+    },
+    {
+      "epoch": 0.9408468244084682,
+      "grad_norm": 0.8704239289503982,
+      "learning_rate": 4.3099189775642044e-08,
+      "loss": 1.5143,
+      "step": 4533
+    },
+    {
+      "epoch": 0.9410543794105438,
+      "grad_norm": 1.2689383914240466,
+      "learning_rate": 4.307754131024808e-08,
+      "loss": 1.5231,
+      "step": 4534
+    },
+    {
+      "epoch": 0.9412619344126193,
+      "grad_norm": 0.9087786551018489,
+      "learning_rate": 4.305596806758655e-08,
+      "loss": 1.4596,
+      "step": 4535
+    },
+    {
+      "epoch": 0.9414694894146949,
+      "grad_norm": 0.7299322704651798,
+      "learning_rate": 4.3034470056829774e-08,
+      "loss": 1.546,
+      "step": 4536
+    },
+    {
+      "epoch": 0.9416770444167705,
+      "grad_norm": 0.837026096713149,
+      "learning_rate": 4.301304728711815e-08,
+      "loss": 1.5874,
+      "step": 4537
+    },
+    {
+      "epoch": 0.941884599418846,
+      "grad_norm": 0.7498371578254125,
+      "learning_rate": 4.299169976756013e-08,
+      "loss": 1.4819,
+      "step": 4538
+    },
+    {
+      "epoch": 0.9420921544209215,
+      "grad_norm": 0.9368841752996462,
+      "learning_rate": 4.297042750723203e-08,
+      "loss": 1.4899,
+      "step": 4539
+    },
+    {
+      "epoch": 0.9422997094229971,
+      "grad_norm": 1.4710849738274496,
+      "learning_rate": 4.294923051517828e-08,
+      "loss": 1.578,
+      "step": 4540
+    },
+    {
+      "epoch": 0.9425072644250726,
+      "grad_norm": 1.7118077346986422,
+      "learning_rate": 4.2928108800411314e-08,
+      "loss": 1.4544,
+      "step": 4541
+    },
+    {
+      "epoch": 0.9427148194271482,
+      "grad_norm": 0.8717334193784907,
+      "learning_rate": 4.2907062371911456e-08,
+      "loss": 1.5114,
+      "step": 4542
+    },
+    {
+      "epoch": 0.9429223744292238,
+      "grad_norm": 0.797151938995318,
+      "learning_rate": 4.288609123862715e-08,
+      "loss": 1.543,
+      "step": 4543
+    },
+    {
+      "epoch": 0.9431299294312993,
+      "grad_norm": 0.6492360388504499,
+      "learning_rate": 4.2865195409474765e-08,
+      "loss": 1.5174,
+      "step": 4544
+    },
+    {
+      "epoch": 0.9433374844333748,
+      "grad_norm": 0.8143199366541592,
+      "learning_rate": 4.28443748933386e-08,
+      "loss": 1.522,
+      "step": 4545
+    },
+    {
+      "epoch": 0.9435450394354504,
+      "grad_norm": 0.8687909828584272,
+      "learning_rate": 4.282362969907102e-08,
+      "loss": 1.5406,
+      "step": 4546
+    },
+    {
+      "epoch": 0.9437525944375259,
+      "grad_norm": 0.6432271038911301,
+      "learning_rate": 4.2802959835492335e-08,
+      "loss": 1.5629,
+      "step": 4547
+    },
+    {
+      "epoch": 0.9439601494396015,
+      "grad_norm": 0.7500351491669647,
+      "learning_rate": 4.278236531139082e-08,
+      "loss": 1.5795,
+      "step": 4548
+    },
+    {
+      "epoch": 0.9441677044416771,
+      "grad_norm": 0.9294157260063273,
+      "learning_rate": 4.276184613552269e-08,
+      "loss": 1.5078,
+      "step": 4549
+    },
+    {
+      "epoch": 0.9443752594437526,
+      "grad_norm": 0.8764919926678213,
+      "learning_rate": 4.2741402316612195e-08,
+      "loss": 1.4858,
+      "step": 4550
+    },
+    {
+      "epoch": 0.9445828144458281,
+      "grad_norm": 0.8546642247694494,
+      "learning_rate": 4.2721033863351483e-08,
+      "loss": 1.5149,
+      "step": 4551
+    },
+    {
+      "epoch": 0.9447903694479037,
+      "grad_norm": 0.774177244202855,
+      "learning_rate": 4.270074078440069e-08,
+      "loss": 1.5055,
+      "step": 4552
+    },
+    {
+      "epoch": 0.9449979244499792,
+      "grad_norm": 0.7604176349904409,
+      "learning_rate": 4.2680523088387875e-08,
+      "loss": 1.4702,
+      "step": 4553
+    },
+    {
+      "epoch": 0.9452054794520548,
+      "grad_norm": 2.0082303625924713,
+      "learning_rate": 4.266038078390908e-08,
+      "loss": 1.461,
+      "step": 4554
+    },
+    {
+      "epoch": 0.9454130344541304,
+      "grad_norm": 0.6675684174323314,
+      "learning_rate": 4.26403138795283e-08,
+      "loss": 1.4913,
+      "step": 4555
+    },
+    {
+      "epoch": 0.9456205894562059,
+      "grad_norm": 0.7995072028384712,
+      "learning_rate": 4.262032238377741e-08,
+      "loss": 1.4671,
+      "step": 4556
+    },
+    {
+      "epoch": 0.9458281444582815,
+      "grad_norm": 1.0244577318179149,
+      "learning_rate": 4.2600406305156267e-08,
+      "loss": 1.5108,
+      "step": 4557
+    },
+    {
+      "epoch": 0.946035699460357,
+      "grad_norm": 0.7408491221435469,
+      "learning_rate": 4.258056565213267e-08,
+      "loss": 1.4884,
+      "step": 4558
+    },
+    {
+      "epoch": 0.9462432544624325,
+      "grad_norm": 0.8713103891772991,
+      "learning_rate": 4.256080043314235e-08,
+      "loss": 1.4809,
+      "step": 4559
+    },
+    {
+      "epoch": 0.9464508094645081,
+      "grad_norm": 0.6426473969151258,
+      "learning_rate": 4.254111065658895e-08,
+      "loss": 1.5272,
+      "step": 4560
+    },
+    {
+      "epoch": 0.9466583644665837,
+      "grad_norm": 1.16716621572466,
+      "learning_rate": 4.252149633084403e-08,
+      "loss": 1.5701,
+      "step": 4561
+    },
+    {
+      "epoch": 0.9468659194686592,
+      "grad_norm": 0.7637856511611366,
+      "learning_rate": 4.250195746424709e-08,
+      "loss": 1.4722,
+      "step": 4562
+    },
+    {
+      "epoch": 0.9470734744707348,
+      "grad_norm": 0.9508379784710671,
+      "learning_rate": 4.248249406510552e-08,
+      "loss": 1.4911,
+      "step": 4563
+    },
+    {
+      "epoch": 0.9472810294728103,
+      "grad_norm": 2.8244709877335006,
+      "learning_rate": 4.2463106141694644e-08,
+      "loss": 1.4978,
+      "step": 4564
+    },
+    {
+      "epoch": 0.9474885844748858,
+      "grad_norm": 0.6451189998139304,
+      "learning_rate": 4.2443793702257715e-08,
+      "loss": 1.5645,
+      "step": 4565
+    },
+    {
+      "epoch": 0.9476961394769614,
+      "grad_norm": 1.1247387387131438,
+      "learning_rate": 4.242455675500585e-08,
+      "loss": 1.6368,
+      "step": 4566
+    },
+    {
+      "epoch": 0.947903694479037,
+      "grad_norm": 1.496196664764651,
+      "learning_rate": 4.2405395308118086e-08,
+      "loss": 1.5292,
+      "step": 4567
+    },
+    {
+      "epoch": 0.9481112494811125,
+      "grad_norm": 0.9142829891268238,
+      "learning_rate": 4.238630936974139e-08,
+      "loss": 1.5708,
+      "step": 4568
+    },
+    {
+      "epoch": 0.9483188044831881,
+      "grad_norm": 1.1393848253853893,
+      "learning_rate": 4.236729894799056e-08,
+      "loss": 1.4659,
+      "step": 4569
+    },
+    {
+      "epoch": 0.9485263594852636,
+      "grad_norm": 0.7690035063731331,
+      "learning_rate": 4.2348364050948344e-08,
+      "loss": 1.516,
+      "step": 4570
+    },
+    {
+      "epoch": 0.9487339144873391,
+      "grad_norm": 0.6911298818902055,
+      "learning_rate": 4.232950468666533e-08,
+      "loss": 1.5558,
+      "step": 4571
+    },
+    {
+      "epoch": 0.9489414694894147,
+      "grad_norm": 1.5514773118523946,
+      "learning_rate": 4.231072086316005e-08,
+      "loss": 1.5528,
+      "step": 4572
+    },
+    {
+      "epoch": 0.9491490244914902,
+      "grad_norm": 0.7423324589971126,
+      "learning_rate": 4.2292012588418896e-08,
+      "loss": 1.419,
+      "step": 4573
+    },
+    {
+      "epoch": 0.9493565794935658,
+      "grad_norm": 0.6969655876415753,
+      "learning_rate": 4.2273379870396084e-08,
+      "loss": 1.5177,
+      "step": 4574
+    },
+    {
+      "epoch": 0.9495641344956414,
+      "grad_norm": 0.6856667438061688,
+      "learning_rate": 4.225482271701379e-08,
+      "loss": 1.59,
+      "step": 4575
+    },
+    {
+      "epoch": 0.9497716894977168,
+      "grad_norm": 0.8036229616953146,
+      "learning_rate": 4.2236341136162017e-08,
+      "loss": 1.4783,
+      "step": 4576
+    },
+    {
+      "epoch": 0.9499792444997924,
+      "grad_norm": 1.3426688142416194,
+      "learning_rate": 4.221793513569863e-08,
+      "loss": 1.5896,
+      "step": 4577
+    },
+    {
+      "epoch": 0.950186799501868,
+      "grad_norm": 0.6733771643306685,
+      "learning_rate": 4.219960472344936e-08,
+      "loss": 1.555,
+      "step": 4578
+    },
+    {
+      "epoch": 0.9503943545039435,
+      "grad_norm": 0.6456767579538037,
+      "learning_rate": 4.218134990720785e-08,
+      "loss": 1.5328,
+      "step": 4579
+    },
+    {
+      "epoch": 0.9506019095060191,
+      "grad_norm": 0.9374275557549357,
+      "learning_rate": 4.216317069473555e-08,
+      "loss": 1.4818,
+      "step": 4580
+    },
+    {
+      "epoch": 0.9508094645080947,
+      "grad_norm": 0.6484423525678912,
+      "learning_rate": 4.214506709376175e-08,
+      "loss": 1.5129,
+      "step": 4581
+    },
+    {
+      "epoch": 0.9510170195101701,
+      "grad_norm": 0.7737760479119955,
+      "learning_rate": 4.212703911198366e-08,
+      "loss": 1.5594,
+      "step": 4582
+    },
+    {
+      "epoch": 0.9512245745122457,
+      "grad_norm": 0.7844084207844886,
+      "learning_rate": 4.210908675706626e-08,
+      "loss": 1.4957,
+      "step": 4583
+    },
+    {
+      "epoch": 0.9514321295143213,
+      "grad_norm": 0.9262593197523248,
+      "learning_rate": 4.209121003664245e-08,
+      "loss": 1.5767,
+      "step": 4584
+    },
+    {
+      "epoch": 0.9516396845163968,
+      "grad_norm": 0.7698081820293008,
+      "learning_rate": 4.2073408958312926e-08,
+      "loss": 1.5767,
+      "step": 4585
+    },
+    {
+      "epoch": 0.9518472395184724,
+      "grad_norm": 0.6735198851412284,
+      "learning_rate": 4.205568352964622e-08,
+      "loss": 1.5023,
+      "step": 4586
+    },
+    {
+      "epoch": 0.952054794520548,
+      "grad_norm": 0.8459595848873649,
+      "learning_rate": 4.203803375817872e-08,
+      "loss": 1.4842,
+      "step": 4587
+    },
+    {
+      "epoch": 0.9522623495226235,
+      "grad_norm": 1.0155324317812409,
+      "learning_rate": 4.202045965141468e-08,
+      "loss": 1.6362,
+      "step": 4588
+    },
+    {
+      "epoch": 0.952469904524699,
+      "grad_norm": 0.7090566499052219,
+      "learning_rate": 4.200296121682606e-08,
+      "loss": 1.5566,
+      "step": 4589
+    },
+    {
+      "epoch": 0.9526774595267746,
+      "grad_norm": 0.9737052611134742,
+      "learning_rate": 4.1985538461852796e-08,
+      "loss": 1.4877,
+      "step": 4590
+    },
+    {
+      "epoch": 0.9528850145288501,
+      "grad_norm": 1.0673626424193492,
+      "learning_rate": 4.196819139390257e-08,
+      "loss": 1.5156,
+      "step": 4591
+    },
+    {
+      "epoch": 0.9530925695309257,
+      "grad_norm": 0.7054889880778412,
+      "learning_rate": 4.195092002035089e-08,
+      "loss": 1.4907,
+      "step": 4592
+    },
+    {
+      "epoch": 0.9533001245330013,
+      "grad_norm": 0.6999754730041319,
+      "learning_rate": 4.193372434854108e-08,
+      "loss": 1.5677,
+      "step": 4593
+    },
+    {
+      "epoch": 0.9535076795350768,
+      "grad_norm": 0.8279841744733754,
+      "learning_rate": 4.191660438578428e-08,
+      "loss": 1.6033,
+      "step": 4594
+    },
+    {
+      "epoch": 0.9537152345371523,
+      "grad_norm": 1.3923924407940542,
+      "learning_rate": 4.189956013935945e-08,
+      "loss": 1.5126,
+      "step": 4595
+    },
+    {
+      "epoch": 0.9539227895392279,
+      "grad_norm": 1.681141184342241,
+      "learning_rate": 4.188259161651336e-08,
+      "loss": 1.5012,
+      "step": 4596
+    },
+    {
+      "epoch": 0.9541303445413034,
+      "grad_norm": 0.7265135910015387,
+      "learning_rate": 4.186569882446057e-08,
+      "loss": 1.4503,
+      "step": 4597
+    },
+    {
+      "epoch": 0.954337899543379,
+      "grad_norm": 0.854786428605816,
+      "learning_rate": 4.1848881770383405e-08,
+      "loss": 1.471,
+      "step": 4598
+    },
+    {
+      "epoch": 0.9545454545454546,
+      "grad_norm": 0.8148948557589987,
+      "learning_rate": 4.183214046143212e-08,
+      "loss": 1.5146,
+      "step": 4599
+    },
+    {
+      "epoch": 0.9547530095475301,
+      "grad_norm": 0.7662688981036365,
+      "learning_rate": 4.1815474904724586e-08,
+      "loss": 1.4914,
+      "step": 4600
+    },
+    {
+      "epoch": 0.9549605645496057,
+      "grad_norm": 0.837785386236163,
+      "learning_rate": 4.1798885107346605e-08,
+      "loss": 1.522,
+      "step": 4601
+    },
+    {
+      "epoch": 0.9551681195516812,
+      "grad_norm": 0.8890482116200626,
+      "learning_rate": 4.178237107635171e-08,
+      "loss": 1.498,
+      "step": 4602
+    },
+    {
+      "epoch": 0.9553756745537567,
+      "grad_norm": 0.8753034034433054,
+      "learning_rate": 4.176593281876123e-08,
+      "loss": 1.502,
+      "step": 4603
+    },
+    {
+      "epoch": 0.9555832295558323,
+      "grad_norm": 0.6595086623139345,
+      "learning_rate": 4.1749570341564245e-08,
+      "loss": 1.4684,
+      "step": 4604
+    },
+    {
+      "epoch": 0.9557907845579079,
+      "grad_norm": 1.0800578946187207,
+      "learning_rate": 4.17332836517177e-08,
+      "loss": 1.438,
+      "step": 4605
+    },
+    {
+      "epoch": 0.9559983395599834,
+      "grad_norm": 0.6281388416977238,
+      "learning_rate": 4.1717072756146225e-08,
+      "loss": 1.5573,
+      "step": 4606
+    },
+    {
+      "epoch": 0.956205894562059,
+      "grad_norm": 1.8534936102912238,
+      "learning_rate": 4.170093766174226e-08,
+      "loss": 1.4748,
+      "step": 4607
+    },
+    {
+      "epoch": 0.9564134495641345,
+      "grad_norm": 0.7001395982790451,
+      "learning_rate": 4.1684878375366025e-08,
+      "loss": 1.5377,
+      "step": 4608
+    },
+    {
+      "epoch": 0.95662100456621,
+      "grad_norm": 1.6047256182064333,
+      "learning_rate": 4.1668894903845525e-08,
+      "loss": 1.5357,
+      "step": 4609
+    },
+    {
+      "epoch": 0.9568285595682856,
+      "grad_norm": 0.6796980319489333,
+      "learning_rate": 4.165298725397648e-08,
+      "loss": 1.5431,
+      "step": 4610
+    },
+    {
+      "epoch": 0.9570361145703612,
+      "grad_norm": 1.2916547937961653,
+      "learning_rate": 4.163715543252242e-08,
+      "loss": 1.5561,
+      "step": 4611
+    },
+    {
+      "epoch": 0.9572436695724367,
+      "grad_norm": 0.7948554214906115,
+      "learning_rate": 4.162139944621461e-08,
+      "loss": 1.477,
+      "step": 4612
+    },
+    {
+      "epoch": 0.9574512245745123,
+      "grad_norm": 0.6937265303064192,
+      "learning_rate": 4.160571930175206e-08,
+      "loss": 1.5232,
+      "step": 4613
+    },
+    {
+      "epoch": 0.9576587795765878,
+      "grad_norm": 0.8077156923953326,
+      "learning_rate": 4.159011500580157e-08,
+      "loss": 1.5153,
+      "step": 4614
+    },
+    {
+      "epoch": 0.9578663345786633,
+      "grad_norm": 1.758448154174464,
+      "learning_rate": 4.157458656499767e-08,
+      "loss": 1.4906,
+      "step": 4615
+    },
+    {
+      "epoch": 0.9580738895807389,
+      "grad_norm": 1.1439333501068285,
+      "learning_rate": 4.1559133985942626e-08,
+      "loss": 1.4249,
+      "step": 4616
+    },
+    {
+      "epoch": 0.9582814445828145,
+      "grad_norm": 1.407703602097641,
+      "learning_rate": 4.154375727520648e-08,
+      "loss": 1.5138,
+      "step": 4617
+    },
+    {
+      "epoch": 0.95848899958489,
+      "grad_norm": 1.0606155976012503,
+      "learning_rate": 4.152845643932701e-08,
+      "loss": 1.5539,
+      "step": 4618
+    },
+    {
+      "epoch": 0.9586965545869656,
+      "grad_norm": 0.7122111604792619,
+      "learning_rate": 4.151323148480968e-08,
+      "loss": 1.5537,
+      "step": 4619
+    },
+    {
+      "epoch": 0.958904109589041,
+      "grad_norm": 0.7859059654256734,
+      "learning_rate": 4.149808241812781e-08,
+      "loss": 1.5215,
+      "step": 4620
+    },
+    {
+      "epoch": 0.9591116645911166,
+      "grad_norm": 0.8564996464193283,
+      "learning_rate": 4.1483009245722314e-08,
+      "loss": 1.462,
+      "step": 4621
+    },
+    {
+      "epoch": 0.9593192195931922,
+      "grad_norm": 0.6311425774837578,
+      "learning_rate": 4.1468011974001914e-08,
+      "loss": 1.5691,
+      "step": 4622
+    },
+    {
+      "epoch": 0.9595267745952677,
+      "grad_norm": 0.6202786189785873,
+      "learning_rate": 4.145309060934312e-08,
+      "loss": 1.446,
+      "step": 4623
+    },
+    {
+      "epoch": 0.9597343295973433,
+      "grad_norm": 1.2288408414971486,
+      "learning_rate": 4.1438245158089997e-08,
+      "loss": 1.5343,
+      "step": 4624
+    },
+    {
+      "epoch": 0.9599418845994189,
+      "grad_norm": 0.7311824696214257,
+      "learning_rate": 4.142347562655451e-08,
+      "loss": 1.4818,
+      "step": 4625
+    },
+    {
+      "epoch": 0.9601494396014943,
+      "grad_norm": 1.0596396783461544,
+      "learning_rate": 4.140878202101625e-08,
+      "loss": 1.5183,
+      "step": 4626
+    },
+    {
+      "epoch": 0.9603569946035699,
+      "grad_norm": 1.325654952200408,
+      "learning_rate": 4.139416434772255e-08,
+      "loss": 1.5279,
+      "step": 4627
+    },
+    {
+      "epoch": 0.9605645496056455,
+      "grad_norm": 0.8300291401820142,
+      "learning_rate": 4.1379622612888426e-08,
+      "loss": 1.5105,
+      "step": 4628
+    },
+    {
+      "epoch": 0.960772104607721,
+      "grad_norm": 0.9084762549121753,
+      "learning_rate": 4.13651568226967e-08,
+      "loss": 1.563,
+      "step": 4629
+    },
+    {
+      "epoch": 0.9609796596097966,
+      "grad_norm": 1.089336144976623,
+      "learning_rate": 4.135076698329779e-08,
+      "loss": 1.48,
+      "step": 4630
+    },
+    {
+      "epoch": 0.9611872146118722,
+      "grad_norm": 0.7102430393330588,
+      "learning_rate": 4.1336453100809893e-08,
+      "loss": 1.5693,
+      "step": 4631
+    },
+    {
+      "epoch": 0.9613947696139477,
+      "grad_norm": 0.8714599631287424,
+      "learning_rate": 4.132221518131891e-08,
+      "loss": 1.4737,
+      "step": 4632
+    },
+    {
+      "epoch": 0.9616023246160232,
+      "grad_norm": 0.7782564964660168,
+      "learning_rate": 4.130805323087838e-08,
+      "loss": 1.4724,
+      "step": 4633
+    },
+    {
+      "epoch": 0.9618098796180988,
+      "grad_norm": 0.7500661734896757,
+      "learning_rate": 4.1293967255509624e-08,
+      "loss": 1.5445,
+      "step": 4634
+    },
+    {
+      "epoch": 0.9620174346201743,
+      "grad_norm": 0.8544078344655154,
+      "learning_rate": 4.1279957261201614e-08,
+      "loss": 1.5287,
+      "step": 4635
+    },
+    {
+      "epoch": 0.9622249896222499,
+      "grad_norm": 0.7329942189821124,
+      "learning_rate": 4.1266023253911034e-08,
+      "loss": 1.5819,
+      "step": 4636
+    },
+    {
+      "epoch": 0.9624325446243255,
+      "grad_norm": 0.9364016938946021,
+      "learning_rate": 4.125216523956224e-08,
+      "loss": 1.5064,
+      "step": 4637
+    },
+    {
+      "epoch": 0.962640099626401,
+      "grad_norm": 0.6864980539580976,
+      "learning_rate": 4.123838322404731e-08,
+      "loss": 1.5072,
+      "step": 4638
+    },
+    {
+      "epoch": 0.9628476546284765,
+      "grad_norm": 1.299292949039472,
+      "learning_rate": 4.1224677213225986e-08,
+      "loss": 1.562,
+      "step": 4639
+    },
+    {
+      "epoch": 0.9630552096305521,
+      "grad_norm": 1.077368512042783,
+      "learning_rate": 4.121104721292569e-08,
+      "loss": 1.5629,
+      "step": 4640
+    },
+    {
+      "epoch": 0.9632627646326276,
+      "grad_norm": 0.8967764752786451,
+      "learning_rate": 4.119749322894154e-08,
+      "loss": 1.4243,
+      "step": 4641
+    },
+    {
+      "epoch": 0.9634703196347032,
+      "grad_norm": 0.6675107737060751,
+      "learning_rate": 4.1184015267036336e-08,
+      "loss": 1.5298,
+      "step": 4642
+    },
+    {
+      "epoch": 0.9636778746367788,
+      "grad_norm": 0.6756480050196306,
+      "learning_rate": 4.117061333294053e-08,
+      "loss": 1.5958,
+      "step": 4643
+    },
+    {
+      "epoch": 0.9638854296388543,
+      "grad_norm": 0.7054910490034481,
+      "learning_rate": 4.11572874323523e-08,
+      "loss": 1.5472,
+      "step": 4644
+    },
+    {
+      "epoch": 0.9640929846409299,
+      "grad_norm": 0.7840653708980738,
+      "learning_rate": 4.114403757093744e-08,
+      "loss": 1.5807,
+      "step": 4645
+    },
+    {
+      "epoch": 0.9643005396430054,
+      "grad_norm": 0.6934424233209485,
+      "learning_rate": 4.113086375432947e-08,
+      "loss": 1.5376,
+      "step": 4646
+    },
+    {
+      "epoch": 0.9645080946450809,
+      "grad_norm": 0.7877721217863387,
+      "learning_rate": 4.111776598812951e-08,
+      "loss": 1.5276,
+      "step": 4647
+    },
+    {
+      "epoch": 0.9647156496471565,
+      "grad_norm": 0.7454271418523588,
+      "learning_rate": 4.110474427790641e-08,
+      "loss": 1.484,
+      "step": 4648
+    },
+    {
+      "epoch": 0.9649232046492321,
+      "grad_norm": 0.744986205741643,
+      "learning_rate": 4.109179862919663e-08,
+      "loss": 1.5905,
+      "step": 4649
+    },
+    {
+      "epoch": 0.9651307596513076,
+      "grad_norm": 0.7004663036656038,
+      "learning_rate": 4.1078929047504335e-08,
+      "loss": 1.4938,
+      "step": 4650
+    },
+    {
+      "epoch": 0.9653383146533832,
+      "grad_norm": 1.0713546895606605,
+      "learning_rate": 4.1066135538301335e-08,
+      "loss": 1.5061,
+      "step": 4651
+    },
+    {
+      "epoch": 0.9655458696554587,
+      "grad_norm": 1.566737592602583,
+      "learning_rate": 4.1053418107027064e-08,
+      "loss": 1.5567,
+      "step": 4652
+    },
+    {
+      "epoch": 0.9657534246575342,
+      "grad_norm": 0.8376640221052621,
+      "learning_rate": 4.104077675908867e-08,
+      "loss": 1.5295,
+      "step": 4653
+    },
+    {
+      "epoch": 0.9659609796596098,
+      "grad_norm": 1.0867380990412316,
+      "learning_rate": 4.102821149986086e-08,
+      "loss": 1.5059,
+      "step": 4654
+    },
+    {
+      "epoch": 0.9661685346616854,
+      "grad_norm": 2.167915928889979,
+      "learning_rate": 4.101572233468614e-08,
+      "loss": 1.5103,
+      "step": 4655
+    },
+    {
+      "epoch": 0.9663760896637609,
+      "grad_norm": 0.7660759611693861,
+      "learning_rate": 4.100330926887451e-08,
+      "loss": 1.5275,
+      "step": 4656
+    },
+    {
+      "epoch": 0.9665836446658365,
+      "grad_norm": 1.2109957192797038,
+      "learning_rate": 4.099097230770366e-08,
+      "loss": 1.4999,
+      "step": 4657
+    },
+    {
+      "epoch": 0.966791199667912,
+      "grad_norm": 0.8004370859145361,
+      "learning_rate": 4.097871145641899e-08,
+      "loss": 1.5075,
+      "step": 4658
+    },
+    {
+      "epoch": 0.9669987546699875,
+      "grad_norm": 0.8864657515165125,
+      "learning_rate": 4.096652672023344e-08,
+      "loss": 1.4705,
+      "step": 4659
+    },
+    {
+      "epoch": 0.9672063096720631,
+      "grad_norm": 0.856026531066869,
+      "learning_rate": 4.095441810432769e-08,
+      "loss": 1.5136,
+      "step": 4660
+    },
+    {
+      "epoch": 0.9674138646741387,
+      "grad_norm": 0.6244707290968177,
+      "learning_rate": 4.094238561384995e-08,
+      "loss": 1.5085,
+      "step": 4661
+    },
+    {
+      "epoch": 0.9676214196762142,
+      "grad_norm": 0.7320728434558134,
+      "learning_rate": 4.093042925391615e-08,
+      "loss": 1.5445,
+      "step": 4662
+    },
+    {
+      "epoch": 0.9678289746782898,
+      "grad_norm": 1.1710369472762623,
+      "learning_rate": 4.091854902960979e-08,
+      "loss": 1.5024,
+      "step": 4663
+    },
+    {
+      "epoch": 0.9680365296803652,
+      "grad_norm": 0.6828462775661966,
+      "learning_rate": 4.090674494598206e-08,
+      "loss": 1.5456,
+      "step": 4664
+    },
+    {
+      "epoch": 0.9682440846824408,
+      "grad_norm": 0.9390340505547526,
+      "learning_rate": 4.0895017008051715e-08,
+      "loss": 1.4355,
+      "step": 4665
+    },
+    {
+      "epoch": 0.9684516396845164,
+      "grad_norm": 0.7522428400905802,
+      "learning_rate": 4.088336522080517e-08,
+      "loss": 1.6443,
+      "step": 4666
+    },
+    {
+      "epoch": 0.9686591946865919,
+      "grad_norm": 0.8685731677402596,
+      "learning_rate": 4.087178958919646e-08,
+      "loss": 1.5257,
+      "step": 4667
+    },
+    {
+      "epoch": 0.9688667496886675,
+      "grad_norm": 0.8520921792331272,
+      "learning_rate": 4.086029011814722e-08,
+      "loss": 1.564,
+      "step": 4668
+    },
+    {
+      "epoch": 0.9690743046907431,
+      "grad_norm": 0.8267511840783619,
+      "learning_rate": 4.084886681254676e-08,
+      "loss": 1.5538,
+      "step": 4669
+    },
+    {
+      "epoch": 0.9692818596928185,
+      "grad_norm": 0.7229289608801082,
+      "learning_rate": 4.0837519677251917e-08,
+      "loss": 1.4945,
+      "step": 4670
+    },
+    {
+      "epoch": 0.9694894146948941,
+      "grad_norm": 0.7277180439821617,
+      "learning_rate": 4.082624871708722e-08,
+      "loss": 1.4949,
+      "step": 4671
+    },
+    {
+      "epoch": 0.9696969696969697,
+      "grad_norm": 0.9480371814541704,
+      "learning_rate": 4.0815053936844776e-08,
+      "loss": 1.4791,
+      "step": 4672
+    },
+    {
+      "epoch": 0.9699045246990452,
+      "grad_norm": 4.645480242387703,
+      "learning_rate": 4.080393534128431e-08,
+      "loss": 1.5323,
+      "step": 4673
+    },
+    {
+      "epoch": 0.9701120797011208,
+      "grad_norm": 0.8358385530189513,
+      "learning_rate": 4.079289293513315e-08,
+      "loss": 1.574,
+      "step": 4674
+    },
+    {
+      "epoch": 0.9703196347031964,
+      "grad_norm": 1.9166545887255289,
+      "learning_rate": 4.078192672308623e-08,
+      "loss": 1.5657,
+      "step": 4675
+    },
+    {
+      "epoch": 0.9705271897052719,
+      "grad_norm": 1.5631358596632146,
+      "learning_rate": 4.077103670980611e-08,
+      "loss": 1.4578,
+      "step": 4676
+    },
+    {
+      "epoch": 0.9707347447073474,
+      "grad_norm": 1.0245883091076342,
+      "learning_rate": 4.07602228999229e-08,
+      "loss": 1.5085,
+      "step": 4677
+    },
+    {
+      "epoch": 0.970942299709423,
+      "grad_norm": 0.8479210791056112,
+      "learning_rate": 4.0749485298034395e-08,
+      "loss": 1.4895,
+      "step": 4678
+    },
+    {
+      "epoch": 0.9711498547114985,
+      "grad_norm": 0.7161365790318167,
+      "learning_rate": 4.07388239087059e-08,
+      "loss": 1.4698,
+      "step": 4679
+    },
+    {
+      "epoch": 0.9713574097135741,
+      "grad_norm": 0.982805133534909,
+      "learning_rate": 4.072823873647037e-08,
+      "loss": 1.5524,
+      "step": 4680
+    },
+    {
+      "epoch": 0.9715649647156497,
+      "grad_norm": 1.6696359927983846,
+      "learning_rate": 4.071772978582831e-08,
+      "loss": 1.5761,
+      "step": 4681
+    },
+    {
+      "epoch": 0.9717725197177252,
+      "grad_norm": 1.1304925351904684,
+      "learning_rate": 4.0707297061247894e-08,
+      "loss": 1.4797,
+      "step": 4682
+    },
+    {
+      "epoch": 0.9719800747198007,
+      "grad_norm": 0.974033859429217,
+      "learning_rate": 4.0696940567164815e-08,
+      "loss": 1.4515,
+      "step": 4683
+    },
+    {
+      "epoch": 0.9721876297218763,
+      "grad_norm": 0.9018628731043778,
+      "learning_rate": 4.068666030798237e-08,
+      "loss": 1.586,
+      "step": 4684
+    },
+    {
+      "epoch": 0.9723951847239518,
+      "grad_norm": 0.989033216063998,
+      "learning_rate": 4.0676456288071484e-08,
+      "loss": 1.4983,
+      "step": 4685
+    },
+    {
+      "epoch": 0.9726027397260274,
+      "grad_norm": 0.6397991257323977,
+      "learning_rate": 4.066632851177059e-08,
+      "loss": 1.485,
+      "step": 4686
+    },
+    {
+      "epoch": 0.972810294728103,
+      "grad_norm": 0.7267750542797208,
+      "learning_rate": 4.065627698338581e-08,
+      "loss": 1.5025,
+      "step": 4687
+    },
+    {
+      "epoch": 0.9730178497301785,
+      "grad_norm": 0.9942896945184911,
+      "learning_rate": 4.0646301707190725e-08,
+      "loss": 1.5947,
+      "step": 4688
+    },
+    {
+      "epoch": 0.973225404732254,
+      "grad_norm": 1.204562693698375,
+      "learning_rate": 4.063640268742657e-08,
+      "loss": 1.4865,
+      "step": 4689
+    },
+    {
+      "epoch": 0.9734329597343296,
+      "grad_norm": 0.7031025269493411,
+      "learning_rate": 4.0626579928302184e-08,
+      "loss": 1.4158,
+      "step": 4690
+    },
+    {
+      "epoch": 0.9736405147364051,
+      "grad_norm": 0.8515389410367321,
+      "learning_rate": 4.0616833433993916e-08,
+      "loss": 1.5084,
+      "step": 4691
+    },
+    {
+      "epoch": 0.9738480697384807,
+      "grad_norm": 0.921565007136761,
+      "learning_rate": 4.060716320864572e-08,
+      "loss": 1.5712,
+      "step": 4692
+    },
+    {
+      "epoch": 0.9740556247405563,
+      "grad_norm": 0.677451922128275,
+      "learning_rate": 4.059756925636912e-08,
+      "loss": 1.5606,
+      "step": 4693
+    },
+    {
+      "epoch": 0.9742631797426318,
+      "grad_norm": 0.860466881581266,
+      "learning_rate": 4.058805158124321e-08,
+      "loss": 1.459,
+      "step": 4694
+    },
+    {
+      "epoch": 0.9744707347447074,
+      "grad_norm": 0.6859805359874132,
+      "learning_rate": 4.057861018731464e-08,
+      "loss": 1.5478,
+      "step": 4695
+    },
+    {
+      "epoch": 0.9746782897467829,
+      "grad_norm": 1.4477060233263033,
+      "learning_rate": 4.056924507859765e-08,
+      "loss": 1.4988,
+      "step": 4696
+    },
+    {
+      "epoch": 0.9748858447488584,
+      "grad_norm": 1.101620088514572,
+      "learning_rate": 4.055995625907405e-08,
+      "loss": 1.4904,
+      "step": 4697
+    },
+    {
+      "epoch": 0.975093399750934,
+      "grad_norm": 1.1776161363871063,
+      "learning_rate": 4.055074373269318e-08,
+      "loss": 1.5173,
+      "step": 4698
+    },
+    {
+      "epoch": 0.9753009547530096,
+      "grad_norm": 1.0307666290948483,
+      "learning_rate": 4.054160750337196e-08,
+      "loss": 1.4563,
+      "step": 4699
+    },
+    {
+      "epoch": 0.9755085097550851,
+      "grad_norm": 0.7362611680560055,
+      "learning_rate": 4.05325475749949e-08,
+      "loss": 1.4868,
+      "step": 4700
+    },
+    {
+      "epoch": 0.9757160647571607,
+      "grad_norm": 0.811619979400467,
+      "learning_rate": 4.0523563951414e-08,
+      "loss": 1.5298,
+      "step": 4701
+    },
+    {
+      "epoch": 0.9759236197592362,
+      "grad_norm": 0.6944963343845532,
+      "learning_rate": 4.051465663644888e-08,
+      "loss": 1.4459,
+      "step": 4702
+    },
+    {
+      "epoch": 0.9761311747613117,
+      "grad_norm": 0.6619802049420035,
+      "learning_rate": 4.050582563388671e-08,
+      "loss": 1.5209,
+      "step": 4703
+    },
+    {
+      "epoch": 0.9763387297633873,
+      "grad_norm": 0.7613884963535807,
+      "learning_rate": 4.049707094748217e-08,
+      "loss": 1.4893,
+      "step": 4704
+    },
+    {
+      "epoch": 0.9765462847654629,
+      "grad_norm": 1.6403748813229346,
+      "learning_rate": 4.048839258095754e-08,
+      "loss": 1.4914,
+      "step": 4705
+    },
+    {
+      "epoch": 0.9767538397675384,
+      "grad_norm": 3.1261298282470413,
+      "learning_rate": 4.047979053800262e-08,
+      "loss": 1.5275,
+      "step": 4706
+    },
+    {
+      "epoch": 0.976961394769614,
+      "grad_norm": 1.21190380898979,
+      "learning_rate": 4.0471264822274773e-08,
+      "loss": 1.5578,
+      "step": 4707
+    },
+    {
+      "epoch": 0.9771689497716894,
+      "grad_norm": 0.7222124017937313,
+      "learning_rate": 4.0462815437398894e-08,
+      "loss": 1.492,
+      "step": 4708
+    },
+    {
+      "epoch": 0.977376504773765,
+      "grad_norm": 0.6759149568725001,
+      "learning_rate": 4.045444238696746e-08,
+      "loss": 1.5949,
+      "step": 4709
+    },
+    {
+      "epoch": 0.9775840597758406,
+      "grad_norm": 0.8984520709168742,
+      "learning_rate": 4.044614567454046e-08,
+      "loss": 1.4678,
+      "step": 4710
+    },
+    {
+      "epoch": 0.9777916147779161,
+      "grad_norm": 0.6788739707070109,
+      "learning_rate": 4.043792530364543e-08,
+      "loss": 1.5045,
+      "step": 4711
+    },
+    {
+      "epoch": 0.9779991697799917,
+      "grad_norm": 1.1852897891125689,
+      "learning_rate": 4.0429781277777465e-08,
+      "loss": 1.3628,
+      "step": 4712
+    },
+    {
+      "epoch": 0.9782067247820673,
+      "grad_norm": 0.6682802264781241,
+      "learning_rate": 4.0421713600399195e-08,
+      "loss": 1.5119,
+      "step": 4713
+    },
+    {
+      "epoch": 0.9784142797841427,
+      "grad_norm": 0.8759680400164908,
+      "learning_rate": 4.0413722274940745e-08,
+      "loss": 1.5006,
+      "step": 4714
+    },
+    {
+      "epoch": 0.9786218347862183,
+      "grad_norm": 0.7496164249162688,
+      "learning_rate": 4.040580730479984e-08,
+      "loss": 1.5238,
+      "step": 4715
+    },
+    {
+      "epoch": 0.9788293897882939,
+      "grad_norm": 1.1257328583630644,
+      "learning_rate": 4.039796869334172e-08,
+      "loss": 1.5144,
+      "step": 4716
+    },
+    {
+      "epoch": 0.9790369447903694,
+      "grad_norm": 0.74059874591586,
+      "learning_rate": 4.0390206443899156e-08,
+      "loss": 1.5735,
+      "step": 4717
+    },
+    {
+      "epoch": 0.979244499792445,
+      "grad_norm": 0.643683515685995,
+      "learning_rate": 4.0382520559772454e-08,
+      "loss": 1.5653,
+      "step": 4718
+    },
+    {
+      "epoch": 0.9794520547945206,
+      "grad_norm": 0.7923047515581395,
+      "learning_rate": 4.037491104422941e-08,
+      "loss": 1.5341,
+      "step": 4719
+    },
+    {
+      "epoch": 0.979659609796596,
+      "grad_norm": 0.7166449522443827,
+      "learning_rate": 4.0367377900505434e-08,
+      "loss": 1.532,
+      "step": 4720
+    },
+    {
+      "epoch": 0.9798671647986716,
+      "grad_norm": 1.201917632385006,
+      "learning_rate": 4.0359921131803386e-08,
+      "loss": 1.5161,
+      "step": 4721
+    },
+    {
+      "epoch": 0.9800747198007472,
+      "grad_norm": 0.7587177346956757,
+      "learning_rate": 4.035254074129371e-08,
+      "loss": 1.5741,
+      "step": 4722
+    },
+    {
+      "epoch": 0.9802822748028227,
+      "grad_norm": 2.127737003348551,
+      "learning_rate": 4.034523673211434e-08,
+      "loss": 1.5328,
+      "step": 4723
+    },
+    {
+      "epoch": 0.9804898298048983,
+      "grad_norm": 0.8185401955680965,
+      "learning_rate": 4.033800910737075e-08,
+      "loss": 1.4286,
+      "step": 4724
+    },
+    {
+      "epoch": 0.9806973848069739,
+      "grad_norm": 0.828920286576002,
+      "learning_rate": 4.033085787013591e-08,
+      "loss": 1.5209,
+      "step": 4725
+    },
+    {
+      "epoch": 0.9809049398090494,
+      "grad_norm": 0.8453287934918091,
+      "learning_rate": 4.0323783023450396e-08,
+      "loss": 1.4562,
+      "step": 4726
+    },
+    {
+      "epoch": 0.981112494811125,
+      "grad_norm": 0.6952283768584352,
+      "learning_rate": 4.031678457032218e-08,
+      "loss": 1.5043,
+      "step": 4727
+    },
+    {
+      "epoch": 0.9813200498132005,
+      "grad_norm": 1.2897209690624207,
+      "learning_rate": 4.030986251372687e-08,
+      "loss": 1.4461,
+      "step": 4728
+    },
+    {
+      "epoch": 0.981527604815276,
+      "grad_norm": 1.0469843352990331,
+      "learning_rate": 4.0303016856607495e-08,
+      "loss": 1.4823,
+      "step": 4729
+    },
+    {
+      "epoch": 0.9817351598173516,
+      "grad_norm": 0.9265847389795577,
+      "learning_rate": 4.029624760187468e-08,
+      "loss": 1.5008,
+      "step": 4730
+    },
+    {
+      "epoch": 0.9819427148194272,
+      "grad_norm": 0.7520590136058093,
+      "learning_rate": 4.028955475240653e-08,
+      "loss": 1.4532,
+      "step": 4731
+    },
+    {
+      "epoch": 0.9821502698215027,
+      "grad_norm": 0.8392658881370643,
+      "learning_rate": 4.028293831104865e-08,
+      "loss": 1.548,
+      "step": 4732
+    },
+    {
+      "epoch": 0.9823578248235783,
+      "grad_norm": 0.7241276603863792,
+      "learning_rate": 4.027639828061418e-08,
+      "loss": 1.4409,
+      "step": 4733
+    },
+    {
+      "epoch": 0.9825653798256538,
+      "grad_norm": 0.946193601502734,
+      "learning_rate": 4.026993466388377e-08,
+      "loss": 1.4558,
+      "step": 4734
+    },
+    {
+      "epoch": 0.9827729348277293,
+      "grad_norm": 0.9661476942971086,
+      "learning_rate": 4.026354746360558e-08,
+      "loss": 1.5323,
+      "step": 4735
+    },
+    {
+      "epoch": 0.9829804898298049,
+      "grad_norm": 1.468427866731725,
+      "learning_rate": 4.0257236682495285e-08,
+      "loss": 1.5041,
+      "step": 4736
+    },
+    {
+      "epoch": 0.9831880448318805,
+      "grad_norm": 0.6742847505585995,
+      "learning_rate": 4.025100232323605e-08,
+      "loss": 1.5136,
+      "step": 4737
+    },
+    {
+      "epoch": 0.983395599833956,
+      "grad_norm": 1.0342345747341286,
+      "learning_rate": 4.024484438847856e-08,
+      "loss": 1.5696,
+      "step": 4738
+    },
+    {
+      "epoch": 0.9836031548360316,
+      "grad_norm": 2.1248841534357905,
+      "learning_rate": 4.0238762880841e-08,
+      "loss": 1.4973,
+      "step": 4739
+    },
+    {
+      "epoch": 0.9838107098381071,
+      "grad_norm": 0.717429367986153,
+      "learning_rate": 4.023275780290908e-08,
+      "loss": 1.4209,
+      "step": 4740
+    },
+    {
+      "epoch": 0.9840182648401826,
+      "grad_norm": 0.7871185432417753,
+      "learning_rate": 4.022682915723599e-08,
+      "loss": 1.5679,
+      "step": 4741
+    },
+    {
+      "epoch": 0.9842258198422582,
+      "grad_norm": 0.8522536893074967,
+      "learning_rate": 4.0220976946342444e-08,
+      "loss": 1.4394,
+      "step": 4742
+    },
+    {
+      "epoch": 0.9844333748443338,
+      "grad_norm": 0.8234984704094376,
+      "learning_rate": 4.0215201172716636e-08,
+      "loss": 1.5065,
+      "step": 4743
+    },
+    {
+      "epoch": 0.9846409298464093,
+      "grad_norm": 0.6869708368717611,
+      "learning_rate": 4.0209501838814276e-08,
+      "loss": 1.4532,
+      "step": 4744
+    },
+    {
+      "epoch": 0.9848484848484849,
+      "grad_norm": 0.7509564992042418,
+      "learning_rate": 4.0203878947058566e-08,
+      "loss": 1.6183,
+      "step": 4745
+    },
+    {
+      "epoch": 0.9850560398505604,
+      "grad_norm": 1.0947761123950381,
+      "learning_rate": 4.0198332499840224e-08,
+      "loss": 1.5381,
+      "step": 4746
+    },
+    {
+      "epoch": 0.9852635948526359,
+      "grad_norm": 0.8684484861142557,
+      "learning_rate": 4.01928624995174e-08,
+      "loss": 1.5029,
+      "step": 4747
+    },
+    {
+      "epoch": 0.9854711498547115,
+      "grad_norm": 0.6835330015666372,
+      "learning_rate": 4.018746894841589e-08,
+      "loss": 1.4818,
+      "step": 4748
+    },
+    {
+      "epoch": 0.9856787048567871,
+      "grad_norm": 0.6886924674526166,
+      "learning_rate": 4.0182151848828783e-08,
+      "loss": 1.5899,
+      "step": 4749
+    },
+    {
+      "epoch": 0.9858862598588626,
+      "grad_norm": 1.1383021939772826,
+      "learning_rate": 4.0176911203016855e-08,
+      "loss": 1.5909,
+      "step": 4750
+    },
+    {
+      "epoch": 0.9860938148609382,
+      "grad_norm": 0.645991041376154,
+      "learning_rate": 4.0171747013208234e-08,
+      "loss": 1.4994,
+      "step": 4751
+    },
+    {
+      "epoch": 0.9863013698630136,
+      "grad_norm": 0.8743987096193,
+      "learning_rate": 4.0166659281598603e-08,
+      "loss": 1.5412,
+      "step": 4752
+    },
+    {
+      "epoch": 0.9865089248650892,
+      "grad_norm": 1.0354994519211387,
+      "learning_rate": 4.016164801035116e-08,
+      "loss": 1.5045,
+      "step": 4753
+    },
+    {
+      "epoch": 0.9867164798671648,
+      "grad_norm": 1.5615867852372607,
+      "learning_rate": 4.0156713201596526e-08,
+      "loss": 1.5393,
+      "step": 4754
+    },
+    {
+      "epoch": 0.9869240348692403,
+      "grad_norm": 1.4601826423134665,
+      "learning_rate": 4.015185485743289e-08,
+      "loss": 1.595,
+      "step": 4755
+    },
+    {
+      "epoch": 0.9871315898713159,
+      "grad_norm": 0.659440547783532,
+      "learning_rate": 4.0147072979925864e-08,
+      "loss": 1.5497,
+      "step": 4756
+    },
+    {
+      "epoch": 0.9873391448733915,
+      "grad_norm": 0.7084580914786018,
+      "learning_rate": 4.014236757110858e-08,
+      "loss": 1.5256,
+      "step": 4757
+    },
+    {
+      "epoch": 0.987546699875467,
+      "grad_norm": 0.7138896654617318,
+      "learning_rate": 4.0137738632981674e-08,
+      "loss": 1.4459,
+      "step": 4758
+    },
+    {
+      "epoch": 0.9877542548775425,
+      "grad_norm": 0.6409615288214876,
+      "learning_rate": 4.013318616751322e-08,
+      "loss": 1.5234,
+      "step": 4759
+    },
+    {
+      "epoch": 0.9879618098796181,
+      "grad_norm": 1.050531096764086,
+      "learning_rate": 4.0128710176638817e-08,
+      "loss": 1.5938,
+      "step": 4760
+    },
+    {
+      "epoch": 0.9881693648816936,
+      "grad_norm": 0.795066709337219,
+      "learning_rate": 4.0124310662261526e-08,
+      "loss": 1.539,
+      "step": 4761
+    },
+    {
+      "epoch": 0.9883769198837692,
+      "grad_norm": 1.4681377456511795,
+      "learning_rate": 4.011998762625192e-08,
+      "loss": 1.4879,
+      "step": 4762
+    },
+    {
+      "epoch": 0.9885844748858448,
+      "grad_norm": 0.7427594931128179,
+      "learning_rate": 4.011574107044802e-08,
+      "loss": 1.5562,
+      "step": 4763
+    },
+    {
+      "epoch": 0.9887920298879203,
+      "grad_norm": 0.7979355255280299,
+      "learning_rate": 4.0111570996655386e-08,
+      "loss": 1.4239,
+      "step": 4764
+    },
+    {
+      "epoch": 0.9889995848899958,
+      "grad_norm": 0.6487432575061177,
+      "learning_rate": 4.010747740664698e-08,
+      "loss": 1.4893,
+      "step": 4765
+    },
+    {
+      "epoch": 0.9892071398920714,
+      "grad_norm": 0.6932652546876449,
+      "learning_rate": 4.01034603021633e-08,
+      "loss": 1.5076,
+      "step": 4766
+    },
+    {
+      "epoch": 0.9894146948941469,
+      "grad_norm": 0.8184296285089726,
+      "learning_rate": 4.0099519684912334e-08,
+      "loss": 1.5866,
+      "step": 4767
+    },
+    {
+      "epoch": 0.9896222498962225,
+      "grad_norm": 0.8105581521638762,
+      "learning_rate": 4.009565555656951e-08,
+      "loss": 1.5144,
+      "step": 4768
+    },
+    {
+      "epoch": 0.9898298048982981,
+      "grad_norm": 0.7538387275210242,
+      "learning_rate": 4.009186791877774e-08,
+      "loss": 1.5249,
+      "step": 4769
+    },
+    {
+      "epoch": 0.9900373599003736,
+      "grad_norm": 1.9042359716630708,
+      "learning_rate": 4.0088156773147466e-08,
+      "loss": 1.501,
+      "step": 4770
+    },
+    {
+      "epoch": 0.9902449149024491,
+      "grad_norm": 0.820178119045428,
+      "learning_rate": 4.008452212125652e-08,
+      "loss": 1.4738,
+      "step": 4771
+    },
+    {
+      "epoch": 0.9904524699045247,
+      "grad_norm": 0.6675939545950915,
+      "learning_rate": 4.0080963964650306e-08,
+      "loss": 1.4948,
+      "step": 4772
+    },
+    {
+      "epoch": 0.9906600249066002,
+      "grad_norm": 0.7515560145313551,
+      "learning_rate": 4.007748230484161e-08,
+      "loss": 1.5279,
+      "step": 4773
+    },
+    {
+      "epoch": 0.9908675799086758,
+      "grad_norm": 0.7500708538598676,
+      "learning_rate": 4.007407714331079e-08,
+      "loss": 1.5235,
+      "step": 4774
+    },
+    {
+      "epoch": 0.9910751349107514,
+      "grad_norm": 0.7484725917075831,
+      "learning_rate": 4.0070748481505594e-08,
+      "loss": 1.4688,
+      "step": 4775
+    },
+    {
+      "epoch": 0.9912826899128269,
+      "grad_norm": 0.8655764784855894,
+      "learning_rate": 4.006749632084131e-08,
+      "loss": 1.5187,
+      "step": 4776
+    },
+    {
+      "epoch": 0.9914902449149025,
+      "grad_norm": 0.6952214555759988,
+      "learning_rate": 4.0064320662700635e-08,
+      "loss": 1.4983,
+      "step": 4777
+    },
+    {
+      "epoch": 0.991697799916978,
+      "grad_norm": 0.9166452120885582,
+      "learning_rate": 4.0061221508433795e-08,
+      "loss": 1.5525,
+      "step": 4778
+    },
+    {
+      "epoch": 0.9919053549190535,
+      "grad_norm": 0.7817218910211783,
+      "learning_rate": 4.005819885935846e-08,
+      "loss": 1.5093,
+      "step": 4779
+    },
+    {
+      "epoch": 0.9921129099211291,
+      "grad_norm": 0.7503094567945067,
+      "learning_rate": 4.00552527167598e-08,
+      "loss": 1.532,
+      "step": 4780
+    },
+    {
+      "epoch": 0.9923204649232047,
+      "grad_norm": 0.6992339495425524,
+      "learning_rate": 4.005238308189043e-08,
+      "loss": 1.4616,
+      "step": 4781
+    },
+    {
+      "epoch": 0.9925280199252802,
+      "grad_norm": 0.6570827794463783,
+      "learning_rate": 4.004958995597042e-08,
+      "loss": 1.5701,
+      "step": 4782
+    },
+    {
+      "epoch": 0.9927355749273558,
+      "grad_norm": 0.7246868341128492,
+      "learning_rate": 4.004687334018735e-08,
+      "loss": 1.51,
+      "step": 4783
+    },
+    {
+      "epoch": 0.9929431299294313,
+      "grad_norm": 0.7304689068384665,
+      "learning_rate": 4.004423323569627e-08,
+      "loss": 1.5637,
+      "step": 4784
+    },
+    {
+      "epoch": 0.9931506849315068,
+      "grad_norm": 0.981954952445363,
+      "learning_rate": 4.0041669643619645e-08,
+      "loss": 1.4135,
+      "step": 4785
+    },
+    {
+      "epoch": 0.9933582399335824,
+      "grad_norm": 0.902759387612884,
+      "learning_rate": 4.003918256504748e-08,
+      "loss": 1.5007,
+      "step": 4786
+    },
+    {
+      "epoch": 0.993565794935658,
+      "grad_norm": 0.6947775892465466,
+      "learning_rate": 4.0036772001037195e-08,
+      "loss": 1.6436,
+      "step": 4787
+    },
+    {
+      "epoch": 0.9937733499377335,
+      "grad_norm": 0.6413717353504101,
+      "learning_rate": 4.0034437952613695e-08,
+      "loss": 1.4695,
+      "step": 4788
+    },
+    {
+      "epoch": 0.9939809049398091,
+      "grad_norm": 0.8198940266814553,
+      "learning_rate": 4.0032180420769376e-08,
+      "loss": 1.545,
+      "step": 4789
+    },
+    {
+      "epoch": 0.9941884599418847,
+      "grad_norm": 0.9209396363654392,
+      "learning_rate": 4.002999940646406e-08,
+      "loss": 1.4476,
+      "step": 4790
+    },
+    {
+      "epoch": 0.9943960149439601,
+      "grad_norm": 0.6808047918826174,
+      "learning_rate": 4.002789491062506e-08,
+      "loss": 1.5462,
+      "step": 4791
+    },
+    {
+      "epoch": 0.9946035699460357,
+      "grad_norm": 0.7161441085781081,
+      "learning_rate": 4.0025866934147177e-08,
+      "loss": 1.543,
+      "step": 4792
+    },
+    {
+      "epoch": 0.9948111249481113,
+      "grad_norm": 0.692252847155623,
+      "learning_rate": 4.0023915477892605e-08,
+      "loss": 1.5144,
+      "step": 4793
+    },
+    {
+      "epoch": 0.9950186799501868,
+      "grad_norm": 0.8444795158888794,
+      "learning_rate": 4.002204054269109e-08,
+      "loss": 1.489,
+      "step": 4794
+    },
+    {
+      "epoch": 0.9952262349522624,
+      "grad_norm": 0.7233455211136888,
+      "learning_rate": 4.002024212933979e-08,
+      "loss": 1.578,
+      "step": 4795
+    },
+    {
+      "epoch": 0.9954337899543378,
+      "grad_norm": 0.7538220410667158,
+      "learning_rate": 4.001852023860335e-08,
+      "loss": 1.5155,
+      "step": 4796
+    },
+    {
+      "epoch": 0.9956413449564134,
+      "grad_norm": 1.125191476257224,
+      "learning_rate": 4.001687487121388e-08,
+      "loss": 1.509,
+      "step": 4797
+    },
+    {
+      "epoch": 0.995848899958489,
+      "grad_norm": 0.8026149744749589,
+      "learning_rate": 4.001530602787092e-08,
+      "loss": 1.5738,
+      "step": 4798
+    },
+    {
+      "epoch": 0.9960564549605645,
+      "grad_norm": 0.8292481411587816,
+      "learning_rate": 4.001381370924151e-08,
+      "loss": 1.5108,
+      "step": 4799
+    },
+    {
+      "epoch": 0.9962640099626401,
+      "grad_norm": 0.7459653503900843,
+      "learning_rate": 4.001239791596016e-08,
+      "loss": 1.5017,
+      "step": 4800
+    },
+    {
+      "epoch": 0.9964715649647157,
+      "grad_norm": 1.0819786540329295,
+      "learning_rate": 4.0011058648628806e-08,
+      "loss": 1.61,
+      "step": 4801
+    },
+    {
+      "epoch": 0.9966791199667911,
+      "grad_norm": 0.7525892175417594,
+      "learning_rate": 4.000979590781689e-08,
+      "loss": 1.5622,
+      "step": 4802
+    },
+    {
+      "epoch": 0.9968866749688667,
+      "grad_norm": 0.8743837860522929,
+      "learning_rate": 4.000860969406129e-08,
+      "loss": 1.5628,
+      "step": 4803
+    },
+    {
+      "epoch": 0.9970942299709423,
+      "grad_norm": 0.7750546424710333,
+      "learning_rate": 4.000750000786634e-08,
+      "loss": 1.4099,
+      "step": 4804
+    },
+    {
+      "epoch": 0.9973017849730178,
+      "grad_norm": 0.6353047899187152,
+      "learning_rate": 4.000646684970386e-08,
+      "loss": 1.5222,
+      "step": 4805
+    },
+    {
+      "epoch": 0.9975093399750934,
+      "grad_norm": 0.6470880844559511,
+      "learning_rate": 4.000551022001311e-08,
+      "loss": 1.5562,
+      "step": 4806
+    },
+    {
+      "epoch": 0.997716894977169,
+      "grad_norm": 0.9084872529285581,
+      "learning_rate": 4.0004630119200843e-08,
+      "loss": 1.5464,
+      "step": 4807
+    },
+    {
+      "epoch": 0.9979244499792445,
+      "grad_norm": 1.140572754103038,
+      "learning_rate": 4.0003826547641254e-08,
+      "loss": 1.5235,
+      "step": 4808
+    },
+    {
+      "epoch": 0.99813200498132,
+      "grad_norm": 0.7299495991107142,
+      "learning_rate": 4.000309950567598e-08,
+      "loss": 1.4966,
+      "step": 4809
+    },
+    {
+      "epoch": 0.9983395599833956,
+      "grad_norm": 0.8724224627053094,
+      "learning_rate": 4.000244899361414e-08,
+      "loss": 1.4967,
+      "step": 4810
+    },
+    {
+      "epoch": 0.9985471149854711,
+      "grad_norm": 1.1786718780774377,
+      "learning_rate": 4.000187501173234e-08,
+      "loss": 1.4663,
+      "step": 4811
+    },
+    {
+      "epoch": 0.9987546699875467,
+      "grad_norm": 0.8396401578326474,
+      "learning_rate": 4.000137756027459e-08,
+      "loss": 1.46,
+      "step": 4812
+    },
+    {
+      "epoch": 0.9989622249896223,
+      "grad_norm": 0.6312751894777425,
+      "learning_rate": 4.000095663945242e-08,
+      "loss": 1.6157,
+      "step": 4813
+    },
+    {
+      "epoch": 0.9991697799916978,
+      "grad_norm": 0.7178029286991596,
+      "learning_rate": 4.000061224944478e-08,
+      "loss": 1.4568,
+      "step": 4814
+    },
+    {
+      "epoch": 0.9993773349937733,
+      "grad_norm": 0.8409185300333479,
+      "learning_rate": 4.00003443903981e-08,
+      "loss": 1.5537,
+      "step": 4815
+    },
+    {
+      "epoch": 0.9995848899958489,
+      "grad_norm": 0.6319662812221615,
+      "learning_rate": 4.0000153062426275e-08,
+      "loss": 1.5672,
+      "step": 4816
+    },
+    {
+      "epoch": 0.9997924449979244,
+      "grad_norm": 0.7315460409082969,
+      "learning_rate": 4.000003826561064e-08,
+      "loss": 1.5316,
+      "step": 4817
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.8062079039678556,
+      "learning_rate": 4e-08,
+      "loss": 1.578,
+      "step": 4818
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 4818,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 964,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.3769285113479168e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}