diff --git "a/trainer_state.json" "b/trainer_state.json"
--- "a/trainer_state.json"
+++ "b/trainer_state.json"
@@ -1,17233 +1,17149 @@
 {
   "best_metric": null,
   "best_model_checkpoint": null,
-  "epoch": 1.8004395604395604,
+  "epoch": 1.8000883262181657,
   "eval_steps": 500,
-  "global_step": 12288,
+  "global_step": 12228,
   "is_hyper_param_search": false,
   "is_local_process_zero": true,
   "is_world_process_zero": true,
   "log_history": [
     {
-      "epoch": 0.00014652014652014652,
-      "grad_norm": 2.4159312562266457,
-      "learning_rate": 9.765625e-09,
-      "loss": 0.6521,
+      "epoch": 0.00014721036360959813,
+      "grad_norm": 2.728368243686464,
+      "learning_rate": 9.813542688910697e-09,
+      "loss": 0.6728,
       "step": 1
     },
     {
-      "epoch": 0.0007326007326007326,
-      "grad_norm": 2.242931374608585,
-      "learning_rate": 4.8828125e-08,
-      "loss": 0.6751,
+      "epoch": 0.0007360518180479906,
+      "grad_norm": 2.36436263451345,
+      "learning_rate": 4.9067713444553494e-08,
+      "loss": 0.6616,
       "step": 5
     },
     {
-      "epoch": 0.0014652014652014652,
-      "grad_norm": 2.0137413845879233,
-      "learning_rate": 9.765625e-08,
-      "loss": 0.6577,
+      "epoch": 0.0014721036360959812,
+      "grad_norm": 2.435737051160779,
+      "learning_rate": 9.813542688910699e-08,
+      "loss": 0.6802,
       "step": 10
     },
     {
-      "epoch": 0.002197802197802198,
-      "grad_norm": 2.4615827770889775,
-      "learning_rate": 1.4648437500000001e-07,
-      "loss": 0.6693,
+      "epoch": 0.002208155454143972,
+      "grad_norm": 2.4940211075102177,
+      "learning_rate": 1.4720314033366047e-07,
+      "loss": 0.6746,
       "step": 15
     },
     {
-      "epoch": 0.0029304029304029304,
-      "grad_norm": 2.0564880448453433,
-      "learning_rate": 1.953125e-07,
-      "loss": 0.6672,
+      "epoch": 0.0029442072721919624,
+      "grad_norm": 2.3800685236179864,
+      "learning_rate": 1.9627085377821398e-07,
+      "loss": 0.6693,
       "step": 20
     },
     {
-      "epoch": 0.003663003663003663,
-      "grad_norm": 2.0796070464691887,
-      "learning_rate": 2.44140625e-07,
-      "loss": 0.6476,
+      "epoch": 0.003680259090239953,
+      "grad_norm": 1.7854052217014194,
+      "learning_rate": 2.4533856722276743e-07,
+      "loss": 0.6401,
       "step": 25
     },
     {
-      "epoch": 0.004395604395604396,
-      "grad_norm": 1.6194878179523382,
-      "learning_rate": 2.9296875000000003e-07,
-      "loss": 0.6567,
+      "epoch": 0.004416310908287944,
+      "grad_norm": 1.8729460237639912,
+      "learning_rate": 2.9440628066732094e-07,
+      "loss": 0.6544,
       "step": 30
     },
     {
-      "epoch": 0.005128205128205128,
-      "grad_norm": 1.5050734017216072,
-      "learning_rate": 3.4179687500000005e-07,
-      "loss": 0.6363,
+      "epoch": 0.005152362726335934,
+      "grad_norm": 1.3749340111576625,
+      "learning_rate": 3.434739941118744e-07,
+      "loss": 0.6246,
       "step": 35
     },
     {
-      "epoch": 0.005860805860805861,
-      "grad_norm": 1.4347029801195341,
-      "learning_rate": 3.90625e-07,
-      "loss": 0.6413,
+      "epoch": 0.005888414544383925,
+      "grad_norm": 1.548613812409993,
+      "learning_rate": 3.9254170755642795e-07,
+      "loss": 0.6268,
       "step": 40
     },
     {
-      "epoch": 0.006593406593406593,
-      "grad_norm": 1.1673856057290772,
-      "learning_rate": 4.3945312500000004e-07,
-      "loss": 0.6301,
+      "epoch": 0.006624466362431915,
+      "grad_norm": 1.2162198549255423,
+      "learning_rate": 4.4160942100098135e-07,
+      "loss": 0.632,
       "step": 45
     },
     {
-      "epoch": 0.007326007326007326,
-      "grad_norm": 1.09382389234995,
-      "learning_rate": 4.8828125e-07,
-      "loss": 0.5953,
+      "epoch": 0.007360518180479906,
+      "grad_norm": 1.1344896752663411,
+      "learning_rate": 4.906771344455349e-07,
+      "loss": 0.5938,
       "step": 50
     },
     {
-      "epoch": 0.00805860805860806,
-      "grad_norm": 0.95912234695919,
-      "learning_rate": 5.37109375e-07,
-      "loss": 0.5915,
+      "epoch": 0.008096569998527897,
+      "grad_norm": 1.0153229071086318,
+      "learning_rate": 5.397448478900884e-07,
+      "loss": 0.6086,
       "step": 55
     },
     {
-      "epoch": 0.008791208791208791,
-      "grad_norm": 0.8028087842621966,
-      "learning_rate": 5.859375000000001e-07,
-      "loss": 0.5964,
+      "epoch": 0.008832621816575887,
+      "grad_norm": 0.8911853226301731,
+      "learning_rate": 5.888125613346419e-07,
+      "loss": 0.5814,
       "step": 60
     },
     {
-      "epoch": 0.009523809523809525,
-      "grad_norm": 0.7618383092986746,
-      "learning_rate": 6.347656250000001e-07,
-      "loss": 0.5709,
+      "epoch": 0.009568673634623878,
+      "grad_norm": 0.7253317192060568,
+      "learning_rate": 6.378802747791954e-07,
+      "loss": 0.5606,
       "step": 65
     },
     {
-      "epoch": 0.010256410256410256,
-      "grad_norm": 0.6938568928414701,
-      "learning_rate": 6.835937500000001e-07,
-      "loss": 0.5726,
+      "epoch": 0.010304725452671868,
+      "grad_norm": 0.7901414889935009,
+      "learning_rate": 6.869479882237488e-07,
+      "loss": 0.5596,
       "step": 70
     },
     {
-      "epoch": 0.01098901098901099,
-      "grad_norm": 0.6871846255942565,
-      "learning_rate": 7.32421875e-07,
-      "loss": 0.5876,
+      "epoch": 0.011040777270719858,
+      "grad_norm": 0.7005773323630946,
+      "learning_rate": 7.360157016683022e-07,
+      "loss": 0.5637,
       "step": 75
     },
     {
-      "epoch": 0.011721611721611722,
-      "grad_norm": 0.6696248589211605,
-      "learning_rate": 7.8125e-07,
-      "loss": 0.5407,
+      "epoch": 0.01177682908876785,
+      "grad_norm": 0.7482787893329246,
+      "learning_rate": 7.850834151128559e-07,
+      "loss": 0.5626,
       "step": 80
     },
     {
-      "epoch": 0.012454212454212455,
-      "grad_norm": 0.6969345857899127,
-      "learning_rate": 8.300781250000001e-07,
-      "loss": 0.5514,
+      "epoch": 0.01251288090681584,
+      "grad_norm": 0.6103751688535668,
+      "learning_rate": 8.341511285574093e-07,
+      "loss": 0.5411,
       "step": 85
     },
     {
-      "epoch": 0.013186813186813187,
-      "grad_norm": 0.7080168030720344,
-      "learning_rate": 8.789062500000001e-07,
-      "loss": 0.5593,
+      "epoch": 0.01324893272486383,
+      "grad_norm": 0.6140626656209314,
+      "learning_rate": 8.832188420019627e-07,
+      "loss": 0.5437,
       "step": 90
     },
     {
-      "epoch": 0.01391941391941392,
-      "grad_norm": 0.6234493870275205,
-      "learning_rate": 9.277343750000001e-07,
-      "loss": 0.5358,
+      "epoch": 0.01398498454291182,
+      "grad_norm": 0.6232501535340139,
+      "learning_rate": 9.322865554465163e-07,
+      "loss": 0.5418,
       "step": 95
     },
     {
-      "epoch": 0.014652014652014652,
-      "grad_norm": 0.6423243268742604,
-      "learning_rate": 9.765625e-07,
-      "loss": 0.5233,
+      "epoch": 0.014721036360959812,
+      "grad_norm": 0.6084600926693361,
+      "learning_rate": 9.813542688910697e-07,
+      "loss": 0.5351,
       "step": 100
     },
     {
-      "epoch": 0.015384615384615385,
-      "grad_norm": 0.6445591567631446,
-      "learning_rate": 1.0253906250000001e-06,
-      "loss": 0.5232,
+      "epoch": 0.015457088179007802,
+      "grad_norm": 0.6302523646733236,
+      "learning_rate": 1.0304219823356233e-06,
+      "loss": 0.5262,
       "step": 105
     },
     {
-      "epoch": 0.01611721611721612,
-      "grad_norm": 0.5895225400800269,
-      "learning_rate": 1.07421875e-06,
-      "loss": 0.5491,
+      "epoch": 0.016193139997055794,
+      "grad_norm": 0.6096300857012811,
+      "learning_rate": 1.0794896957801768e-06,
+      "loss": 0.5136,
       "step": 110
     },
     {
-      "epoch": 0.01684981684981685,
-      "grad_norm": 0.6024446639287304,
-      "learning_rate": 1.1230468750000002e-06,
-      "loss": 0.5249,
+      "epoch": 0.016929191815103783,
+      "grad_norm": 0.6011309359210671,
+      "learning_rate": 1.1285574092247302e-06,
+      "loss": 0.5163,
       "step": 115
     },
     {
-      "epoch": 0.017582417582417582,
-      "grad_norm": 0.6352242041819796,
-      "learning_rate": 1.1718750000000001e-06,
-      "loss": 0.5268,
+      "epoch": 0.017665243633151775,
+      "grad_norm": 0.6104693046051632,
+      "learning_rate": 1.1776251226692837e-06,
+      "loss": 0.5253,
       "step": 120
     },
     {
-      "epoch": 0.018315018315018316,
-      "grad_norm": 0.5743873313770402,
-      "learning_rate": 1.220703125e-06,
-      "loss": 0.5359,
+      "epoch": 0.018401295451199763,
+      "grad_norm": 0.6049006080080384,
+      "learning_rate": 1.2266928361138373e-06,
+      "loss": 0.5071,
       "step": 125
     },
     {
-      "epoch": 0.01904761904761905,
-      "grad_norm": 0.6295853399822839,
-      "learning_rate": 1.2695312500000002e-06,
-      "loss": 0.5086,
+      "epoch": 0.019137347269247755,
+      "grad_norm": 0.6135220924932349,
+      "learning_rate": 1.2757605495583909e-06,
+      "loss": 0.5341,
       "step": 130
     },
     {
-      "epoch": 0.01978021978021978,
-      "grad_norm": 0.5942595247940665,
-      "learning_rate": 1.318359375e-06,
-      "loss": 0.5291,
+      "epoch": 0.019873399087295747,
+      "grad_norm": 0.5741753057894418,
+      "learning_rate": 1.324828263002944e-06,
+      "loss": 0.5184,
       "step": 135
     },
     {
-      "epoch": 0.020512820512820513,
-      "grad_norm": 0.5970756645818605,
-      "learning_rate": 1.3671875000000002e-06,
-      "loss": 0.5179,
+      "epoch": 0.020609450905343735,
+      "grad_norm": 0.5682793875483976,
+      "learning_rate": 1.3738959764474976e-06,
+      "loss": 0.5146,
       "step": 140
     },
     {
-      "epoch": 0.021245421245421246,
-      "grad_norm": 0.6016303784928376,
-      "learning_rate": 1.4160156250000001e-06,
-      "loss": 0.5169,
+      "epoch": 0.021345502723391727,
+      "grad_norm": 0.5801604412057352,
+      "learning_rate": 1.4229636898920513e-06,
+      "loss": 0.5006,
       "step": 145
     },
     {
-      "epoch": 0.02197802197802198,
-      "grad_norm": 0.6041929264356795,
-      "learning_rate": 1.46484375e-06,
-      "loss": 0.5093,
+      "epoch": 0.022081554541439716,
+      "grad_norm": 0.5716040330610084,
+      "learning_rate": 1.4720314033366045e-06,
+      "loss": 0.4972,
       "step": 150
     },
     {
-      "epoch": 0.02271062271062271,
-      "grad_norm": 0.6056640633462855,
-      "learning_rate": 1.5136718750000002e-06,
-      "loss": 0.5233,
+      "epoch": 0.022817606359487708,
+      "grad_norm": 0.6383262495009793,
+      "learning_rate": 1.521099116781158e-06,
+      "loss": 0.5134,
       "step": 155
     },
     {
-      "epoch": 0.023443223443223443,
-      "grad_norm": 0.5955769301786994,
-      "learning_rate": 1.5625e-06,
-      "loss": 0.4934,
+      "epoch": 0.0235536581775357,
+      "grad_norm": 0.574501385772374,
+      "learning_rate": 1.5701668302257118e-06,
+      "loss": 0.5305,
       "step": 160
     },
     {
-      "epoch": 0.024175824175824177,
-      "grad_norm": 0.5834862533300711,
-      "learning_rate": 1.6113281250000002e-06,
-      "loss": 0.515,
+      "epoch": 0.024289709995583688,
+      "grad_norm": 0.624077819074781,
+      "learning_rate": 1.619234543670265e-06,
+      "loss": 0.5052,
       "step": 165
     },
     {
-      "epoch": 0.02490842490842491,
-      "grad_norm": 0.5741046251131711,
-      "learning_rate": 1.6601562500000001e-06,
-      "loss": 0.5026,
+      "epoch": 0.02502576181363168,
+      "grad_norm": 0.5798590790669421,
+      "learning_rate": 1.6683022571148185e-06,
+      "loss": 0.501,
       "step": 170
     },
     {
-      "epoch": 0.02564102564102564,
-      "grad_norm": 0.6100284855554404,
-      "learning_rate": 1.708984375e-06,
-      "loss": 0.5162,
+      "epoch": 0.025761813631679672,
+      "grad_norm": 0.58108200938206,
+      "learning_rate": 1.717369970559372e-06,
+      "loss": 0.5067,
       "step": 175
     },
     {
-      "epoch": 0.026373626373626374,
-      "grad_norm": 0.5674946945185096,
-      "learning_rate": 1.7578125000000002e-06,
-      "loss": 0.4726,
+      "epoch": 0.02649786544972766,
+      "grad_norm": 0.591325748468392,
+      "learning_rate": 1.7664376840039254e-06,
+      "loss": 0.5129,
       "step": 180
     },
     {
-      "epoch": 0.027106227106227107,
-      "grad_norm": 0.6049877055480813,
-      "learning_rate": 1.806640625e-06,
-      "loss": 0.4993,
+      "epoch": 0.027233917267775652,
+      "grad_norm": 0.6038913900244864,
+      "learning_rate": 1.815505397448479e-06,
+      "loss": 0.4816,
       "step": 185
     },
     {
-      "epoch": 0.02783882783882784,
-      "grad_norm": 0.6305774506688272,
-      "learning_rate": 1.8554687500000002e-06,
-      "loss": 0.5038,
+      "epoch": 0.02796996908582364,
+      "grad_norm": 0.5775965197667471,
+      "learning_rate": 1.8645731108930325e-06,
+      "loss": 0.4966,
       "step": 190
     },
     {
-      "epoch": 0.02857142857142857,
-      "grad_norm": 0.5922615313022872,
-      "learning_rate": 1.9042968750000001e-06,
-      "loss": 0.5033,
+      "epoch": 0.028706020903871633,
+      "grad_norm": 0.6133721061273785,
+      "learning_rate": 1.913640824337586e-06,
+      "loss": 0.5071,
       "step": 195
     },
     {
-      "epoch": 0.029304029304029304,
-      "grad_norm": 0.5836180906085862,
-      "learning_rate": 1.953125e-06,
-      "loss": 0.4966,
+      "epoch": 0.029442072721919624,
+      "grad_norm": 0.5922070936085064,
+      "learning_rate": 1.9627085377821394e-06,
+      "loss": 0.4921,
       "step": 200
     },
     {
-      "epoch": 0.030036630036630037,
-      "grad_norm": 0.6174338649362718,
-      "learning_rate": 2.001953125e-06,
-      "loss": 0.504,
+      "epoch": 0.030178124539967613,
+      "grad_norm": 0.5917430267113624,
+      "learning_rate": 2.011776251226693e-06,
+      "loss": 0.4921,
       "step": 205
     },
     {
-      "epoch": 0.03076923076923077,
-      "grad_norm": 0.576153770808032,
-      "learning_rate": 2.0507812500000003e-06,
-      "loss": 0.4887,
+      "epoch": 0.030914176358015605,
+      "grad_norm": 0.6221645979156015,
+      "learning_rate": 2.0608439646712466e-06,
+      "loss": 0.4859,
       "step": 210
     },
     {
-      "epoch": 0.0315018315018315,
-      "grad_norm": 0.6030157714600828,
-      "learning_rate": 2.099609375e-06,
-      "loss": 0.5043,
+      "epoch": 0.0316502281760636,
+      "grad_norm": 0.554467637353109,
+      "learning_rate": 2.1099116781158e-06,
+      "loss": 0.5035,
       "step": 215
     },
     {
-      "epoch": 0.03223443223443224,
-      "grad_norm": 0.5486182063481856,
-      "learning_rate": 2.1484375e-06,
-      "loss": 0.465,
+      "epoch": 0.03238627999411159,
+      "grad_norm": 0.6227780076273214,
+      "learning_rate": 2.1589793915603537e-06,
+      "loss": 0.4956,
       "step": 220
     },
     {
-      "epoch": 0.03296703296703297,
-      "grad_norm": 0.5825942750922115,
-      "learning_rate": 2.1972656250000003e-06,
-      "loss": 0.4991,
+      "epoch": 0.033122331812159574,
+      "grad_norm": 0.5912071689774797,
+      "learning_rate": 2.208047105004907e-06,
+      "loss": 0.5034,
       "step": 225
     },
     {
-      "epoch": 0.0336996336996337,
-      "grad_norm": 0.5771727389731394,
-      "learning_rate": 2.2460937500000004e-06,
-      "loss": 0.5067,
+      "epoch": 0.033858383630207566,
+      "grad_norm": 0.5710270435221912,
+      "learning_rate": 2.2571148184494604e-06,
+      "loss": 0.5062,
       "step": 230
     },
     {
-      "epoch": 0.034432234432234435,
-      "grad_norm": 0.6439214652767457,
-      "learning_rate": 2.294921875e-06,
-      "loss": 0.5181,
+      "epoch": 0.03459443544825556,
+      "grad_norm": 0.5723517153838513,
+      "learning_rate": 2.306182531894014e-06,
+      "loss": 0.4973,
       "step": 235
     },
     {
-      "epoch": 0.035164835164835165,
-      "grad_norm": 0.5874429834389425,
-      "learning_rate": 2.3437500000000002e-06,
-      "loss": 0.4784,
+      "epoch": 0.03533048726630355,
+      "grad_norm": 0.6239526253603609,
+      "learning_rate": 2.3552502453385675e-06,
+      "loss": 0.4966,
       "step": 240
     },
     {
-      "epoch": 0.035897435897435895,
-      "grad_norm": 0.6170127311772752,
-      "learning_rate": 2.3925781250000003e-06,
-      "loss": 0.5006,
+      "epoch": 0.03606653908435154,
+      "grad_norm": 0.5840342287619873,
+      "learning_rate": 2.404317958783121e-06,
+      "loss": 0.4749,
       "step": 245
     },
     {
-      "epoch": 0.03663003663003663,
-      "grad_norm": 0.613101392672586,
-      "learning_rate": 2.44140625e-06,
-      "loss": 0.4976,
+      "epoch": 0.036802590902399526,
+      "grad_norm": 0.5853430651319891,
+      "learning_rate": 2.4533856722276746e-06,
+      "loss": 0.4985,
       "step": 250
     },
     {
-      "epoch": 0.03736263736263736,
-      "grad_norm": 0.5753422378290689,
-      "learning_rate": 2.490234375e-06,
-      "loss": 0.4885,
+      "epoch": 0.03753864272044752,
+      "grad_norm": 0.583836894163469,
+      "learning_rate": 2.5024533856722278e-06,
+      "loss": 0.5003,
       "step": 255
     },
     {
-      "epoch": 0.0380952380952381,
-      "grad_norm": 0.6133895496160483,
-      "learning_rate": 2.5390625000000003e-06,
-      "loss": 0.5008,
+      "epoch": 0.03827469453849551,
+      "grad_norm": 0.6068565691460917,
+      "learning_rate": 2.5515210991167817e-06,
+      "loss": 0.4752,
       "step": 260
     },
     {
-      "epoch": 0.03882783882783883,
-      "grad_norm": 0.5964068990574745,
-      "learning_rate": 2.587890625e-06,
-      "loss": 0.5151,
+      "epoch": 0.0390107463565435,
+      "grad_norm": 0.6260409519393997,
+      "learning_rate": 2.600588812561335e-06,
+      "loss": 0.5041,
       "step": 265
     },
     {
-      "epoch": 0.03956043956043956,
-      "grad_norm": 0.6182569257932172,
-      "learning_rate": 2.63671875e-06,
-      "loss": 0.4983,
+      "epoch": 0.039746798174591494,
+      "grad_norm": 0.6125760382103062,
+      "learning_rate": 2.649656526005888e-06,
+      "loss": 0.5063,
       "step": 270
     },
     {
-      "epoch": 0.040293040293040296,
-      "grad_norm": 0.5736263115770767,
-      "learning_rate": 2.6855468750000003e-06,
-      "loss": 0.4892,
+      "epoch": 0.04048284999263948,
+      "grad_norm": 0.5944574529155787,
+      "learning_rate": 2.698724239450442e-06,
+      "loss": 0.4808,
       "step": 275
     },
     {
-      "epoch": 0.041025641025641026,
-      "grad_norm": 0.619235921760934,
-      "learning_rate": 2.7343750000000004e-06,
-      "loss": 0.4738,
+      "epoch": 0.04121890181068747,
+      "grad_norm": 0.6456076986145468,
+      "learning_rate": 2.747791952894995e-06,
+      "loss": 0.4952,
       "step": 280
     },
     {
-      "epoch": 0.041758241758241756,
-      "grad_norm": 0.6539149573620286,
-      "learning_rate": 2.783203125e-06,
-      "loss": 0.4922,
+      "epoch": 0.04195495362873546,
+      "grad_norm": 0.9915769812640363,
+      "learning_rate": 2.7968596663395487e-06,
+      "loss": 0.501,
       "step": 285
     },
     {
-      "epoch": 0.04249084249084249,
-      "grad_norm": 0.61689092854405,
-      "learning_rate": 2.8320312500000002e-06,
-      "loss": 0.4976,
+      "epoch": 0.042691005446783455,
+      "grad_norm": 0.5920463349789087,
+      "learning_rate": 2.8459273797841027e-06,
+      "loss": 0.4931,
       "step": 290
     },
     {
-      "epoch": 0.04322344322344322,
-      "grad_norm": 0.628660225181727,
-      "learning_rate": 2.8808593750000004e-06,
-      "loss": 0.4826,
+      "epoch": 0.043427057264831447,
+      "grad_norm": 0.6575309659536147,
+      "learning_rate": 2.894995093228656e-06,
+      "loss": 0.4983,
       "step": 295
     },
     {
-      "epoch": 0.04395604395604396,
-      "grad_norm": 0.6617557087521471,
-      "learning_rate": 2.9296875e-06,
-      "loss": 0.4932,
+      "epoch": 0.04416310908287943,
+      "grad_norm": 0.5667173217760384,
+      "learning_rate": 2.944062806673209e-06,
+      "loss": 0.4904,
       "step": 300
     },
     {
-      "epoch": 0.04468864468864469,
-      "grad_norm": 0.5622158080515355,
-      "learning_rate": 2.978515625e-06,
-      "loss": 0.4662,
+      "epoch": 0.04489916090092742,
+      "grad_norm": 0.5821180180751933,
+      "learning_rate": 2.993130520117763e-06,
+      "loss": 0.4919,
       "step": 305
     },
     {
-      "epoch": 0.04542124542124542,
-      "grad_norm": 0.6464987442594975,
-      "learning_rate": 3.0273437500000003e-06,
-      "loss": 0.4891,
+      "epoch": 0.045635212718975415,
+      "grad_norm": 0.6231991536660229,
+      "learning_rate": 3.042198233562316e-06,
+      "loss": 0.4888,
       "step": 310
     },
     {
-      "epoch": 0.046153846153846156,
-      "grad_norm": 0.6216580067513613,
-      "learning_rate": 3.0761718750000004e-06,
-      "loss": 0.4831,
+      "epoch": 0.04637126453702341,
+      "grad_norm": 0.59956643814679,
+      "learning_rate": 3.0912659470068696e-06,
+      "loss": 0.4598,
       "step": 315
     },
     {
-      "epoch": 0.046886446886446886,
-      "grad_norm": 0.5807795396390456,
-      "learning_rate": 3.125e-06,
-      "loss": 0.4658,
+      "epoch": 0.0471073163550714,
+      "grad_norm": 0.5779052883486501,
+      "learning_rate": 3.1403336604514236e-06,
+      "loss": 0.4779,
       "step": 320
     },
     {
-      "epoch": 0.047619047619047616,
-      "grad_norm": 0.6324392259846056,
-      "learning_rate": 3.1738281250000003e-06,
-      "loss": 0.506,
+      "epoch": 0.04784336817311939,
+      "grad_norm": 0.6056761160995231,
+      "learning_rate": 3.1894013738959768e-06,
+      "loss": 0.4994,
       "step": 325
     },
     {
-      "epoch": 0.04835164835164835,
-      "grad_norm": 0.5599658664516196,
-      "learning_rate": 3.2226562500000004e-06,
-      "loss": 0.4736,
+      "epoch": 0.048579419991167376,
+      "grad_norm": 0.599035762897119,
+      "learning_rate": 3.23846908734053e-06,
+      "loss": 0.4785,
       "step": 330
     },
     {
-      "epoch": 0.04908424908424908,
-      "grad_norm": 0.6260406463036369,
-      "learning_rate": 3.271484375e-06,
-      "loss": 0.4856,
+      "epoch": 0.04931547180921537,
+      "grad_norm": 0.6302501472689347,
+      "learning_rate": 3.287536800785084e-06,
+      "loss": 0.4761,
       "step": 335
     },
     {
-      "epoch": 0.04981684981684982,
-      "grad_norm": 0.6106796656996849,
-      "learning_rate": 3.3203125000000002e-06,
-      "loss": 0.4745,
+      "epoch": 0.05005152362726336,
+      "grad_norm": 0.6243951085037258,
+      "learning_rate": 3.336604514229637e-06,
+      "loss": 0.4978,
       "step": 340
     },
     {
-      "epoch": 0.05054945054945055,
-      "grad_norm": 0.5770205864950204,
-      "learning_rate": 3.3691406250000004e-06,
-      "loss": 0.4841,
+      "epoch": 0.05078757544531135,
+      "grad_norm": 0.6106406332182736,
+      "learning_rate": 3.3856722276741906e-06,
+      "loss": 0.4869,
       "step": 345
     },
     {
-      "epoch": 0.05128205128205128,
-      "grad_norm": 0.5935897914996285,
-      "learning_rate": 3.41796875e-06,
-      "loss": 0.4745,
+      "epoch": 0.051523627263359344,
+      "grad_norm": 0.6749576604983416,
+      "learning_rate": 3.434739941118744e-06,
+      "loss": 0.4839,
       "step": 350
     },
     {
-      "epoch": 0.05201465201465202,
-      "grad_norm": 0.5871039026588047,
-      "learning_rate": 3.466796875e-06,
-      "loss": 0.4853,
+      "epoch": 0.05225967908140733,
+      "grad_norm": 0.6327724849785807,
+      "learning_rate": 3.4838076545632977e-06,
+      "loss": 0.4885,
       "step": 355
     },
     {
-      "epoch": 0.05274725274725275,
-      "grad_norm": 0.5882570675337384,
-      "learning_rate": 3.5156250000000003e-06,
-      "loss": 0.4748,
+      "epoch": 0.05299573089945532,
+      "grad_norm": 0.5768211285301043,
+      "learning_rate": 3.532875368007851e-06,
+      "loss": 0.4615,
       "step": 360
     },
     {
-      "epoch": 0.05347985347985348,
-      "grad_norm": 0.6008326193899312,
-      "learning_rate": 3.5644531250000005e-06,
-      "loss": 0.4793,
+      "epoch": 0.05373178271750331,
+      "grad_norm": 0.5919583609982937,
+      "learning_rate": 3.581943081452405e-06,
+      "loss": 0.4765,
       "step": 365
     },
     {
-      "epoch": 0.054212454212454214,
-      "grad_norm": 0.5964595164304082,
-      "learning_rate": 3.61328125e-06,
-      "loss": 0.4792,
+      "epoch": 0.054467834535551304,
+      "grad_norm": 0.5871500912160249,
+      "learning_rate": 3.631010794896958e-06,
+      "loss": 0.4806,
       "step": 370
     },
     {
-      "epoch": 0.054945054945054944,
-      "grad_norm": 0.613105851356729,
-      "learning_rate": 3.6621093750000003e-06,
-      "loss": 0.4778,
+      "epoch": 0.055203886353599296,
+      "grad_norm": 0.6106136580240804,
+      "learning_rate": 3.6800785083415115e-06,
+      "loss": 0.4826,
       "step": 375
     },
     {
-      "epoch": 0.05567765567765568,
-      "grad_norm": 0.6294016643255824,
-      "learning_rate": 3.7109375000000004e-06,
-      "loss": 0.4755,
+      "epoch": 0.05593993817164728,
+      "grad_norm": 0.5941821734755603,
+      "learning_rate": 3.729146221786065e-06,
+      "loss": 0.4762,
       "step": 380
     },
     {
-      "epoch": 0.05641025641025641,
-      "grad_norm": 0.5905210786928944,
-      "learning_rate": 3.759765625e-06,
-      "loss": 0.4751,
+      "epoch": 0.05667598998969527,
+      "grad_norm": 0.6119338751306809,
+      "learning_rate": 3.7782139352306186e-06,
+      "loss": 0.4984,
       "step": 385
     },
     {
-      "epoch": 0.05714285714285714,
-      "grad_norm": 0.572125914373807,
-      "learning_rate": 3.8085937500000002e-06,
-      "loss": 0.4857,
+      "epoch": 0.057412041807743265,
+      "grad_norm": 0.5792835900775687,
+      "learning_rate": 3.827281648675172e-06,
+      "loss": 0.4742,
       "step": 390
     },
     {
-      "epoch": 0.05787545787545788,
-      "grad_norm": 0.5797543518358743,
-      "learning_rate": 3.857421875e-06,
-      "loss": 0.473,
+      "epoch": 0.05814809362579126,
+      "grad_norm": 0.5857316557188246,
+      "learning_rate": 3.876349362119725e-06,
+      "loss": 0.487,
       "step": 395
     },
     {
-      "epoch": 0.05860805860805861,
-      "grad_norm": 0.6375925669051277,
-      "learning_rate": 3.90625e-06,
-      "loss": 0.4856,
+      "epoch": 0.05888414544383925,
+      "grad_norm": 0.5871381247935139,
+      "learning_rate": 3.925417075564279e-06,
+      "loss": 0.4664,
       "step": 400
     },
     {
-      "epoch": 0.05934065934065934,
-      "grad_norm": 0.5984195282406634,
-      "learning_rate": 3.955078125000001e-06,
-      "loss": 0.4912,
+      "epoch": 0.059620197261887234,
+      "grad_norm": 0.5996770410892924,
+      "learning_rate": 3.9744847890088324e-06,
+      "loss": 0.4818,
       "step": 405
     },
     {
-      "epoch": 0.060073260073260075,
-      "grad_norm": 0.6610747118232397,
-      "learning_rate": 4.00390625e-06,
-      "loss": 0.4769,
+      "epoch": 0.060356249079935226,
+      "grad_norm": 0.6078308418061451,
+      "learning_rate": 4.023552502453386e-06,
+      "loss": 0.4723,
       "step": 410
     },
     {
-      "epoch": 0.060805860805860805,
-      "grad_norm": 0.5884293595488747,
-      "learning_rate": 4.052734375e-06,
-      "loss": 0.4761,
+      "epoch": 0.06109230089798322,
+      "grad_norm": 0.6072936425940079,
+      "learning_rate": 4.0726202158979396e-06,
+      "loss": 0.4947,
       "step": 415
     },
     {
-      "epoch": 0.06153846153846154,
-      "grad_norm": 0.6043701478533586,
-      "learning_rate": 4.101562500000001e-06,
-      "loss": 0.4793,
+      "epoch": 0.06182835271603121,
+      "grad_norm": 0.5854312934656148,
+      "learning_rate": 4.121687929342493e-06,
+      "loss": 0.4721,
       "step": 420
     },
     {
-      "epoch": 0.06227106227106227,
-      "grad_norm": 0.5822030559431073,
-      "learning_rate": 4.150390625e-06,
-      "loss": 0.4633,
+      "epoch": 0.0625644045340792,
+      "grad_norm": 0.5912636415377042,
+      "learning_rate": 4.170755642787047e-06,
+      "loss": 0.483,
       "step": 425
     },
     {
-      "epoch": 0.063003663003663,
-      "grad_norm": 0.5959155128039212,
-      "learning_rate": 4.19921875e-06,
-      "loss": 0.4733,
+      "epoch": 0.0633004563521272,
+      "grad_norm": 0.5368794998653236,
+      "learning_rate": 4.2198233562316e-06,
+      "loss": 0.4685,
       "step": 430
     },
     {
-      "epoch": 0.06373626373626373,
-      "grad_norm": 0.6188697338791965,
-      "learning_rate": 4.2480468750000006e-06,
-      "loss": 0.491,
+      "epoch": 0.06403650817017519,
+      "grad_norm": 0.5951573811609723,
+      "learning_rate": 4.268891069676153e-06,
+      "loss": 0.4773,
       "step": 435
     },
     {
-      "epoch": 0.06446886446886448,
-      "grad_norm": 0.6097716090590869,
-      "learning_rate": 4.296875e-06,
-      "loss": 0.474,
+      "epoch": 0.06477255998822318,
+      "grad_norm": 0.5891969626171352,
+      "learning_rate": 4.317958783120707e-06,
+      "loss": 0.4511,
       "step": 440
     },
     {
-      "epoch": 0.0652014652014652,
-      "grad_norm": 0.6123749603827808,
-      "learning_rate": 4.345703125e-06,
-      "loss": 0.4624,
+      "epoch": 0.06550861180627116,
+      "grad_norm": 0.6085822185086244,
+      "learning_rate": 4.36702649656526e-06,
+      "loss": 0.4742,
       "step": 445
     },
     {
-      "epoch": 0.06593406593406594,
-      "grad_norm": 0.5873918820746148,
-      "learning_rate": 4.3945312500000005e-06,
-      "loss": 0.4831,
+      "epoch": 0.06624466362431915,
+      "grad_norm": 0.5767003298984905,
+      "learning_rate": 4.416094210009814e-06,
+      "loss": 0.4721,
       "step": 450
     },
     {
-      "epoch": 0.06666666666666667,
-      "grad_norm": 0.5894168087718852,
-      "learning_rate": 4.443359375e-06,
-      "loss": 0.4898,
+      "epoch": 0.06698071544236714,
+      "grad_norm": 0.5902641756754682,
+      "learning_rate": 4.465161923454367e-06,
+      "loss": 0.465,
       "step": 455
     },
     {
-      "epoch": 0.0673992673992674,
-      "grad_norm": 0.5678993089140597,
-      "learning_rate": 4.492187500000001e-06,
-      "loss": 0.4717,
+      "epoch": 0.06771676726041513,
+      "grad_norm": 0.5689500427434829,
+      "learning_rate": 4.514229636898921e-06,
+      "loss": 0.4501,
       "step": 460
     },
     {
-      "epoch": 0.06813186813186813,
-      "grad_norm": 0.6204084821274443,
-      "learning_rate": 4.5410156250000005e-06,
-      "loss": 0.4733,
+      "epoch": 0.06845281907846312,
+      "grad_norm": 0.6103508671952677,
+      "learning_rate": 4.563297350343474e-06,
+      "loss": 0.4821,
       "step": 465
     },
     {
-      "epoch": 0.06886446886446887,
-      "grad_norm": 0.5906487740302222,
-      "learning_rate": 4.58984375e-06,
-      "loss": 0.493,
+      "epoch": 0.06918887089651111,
+      "grad_norm": 0.6380519885930872,
+      "learning_rate": 4.612365063788028e-06,
+      "loss": 0.4709,
       "step": 470
     },
     {
-      "epoch": 0.0695970695970696,
-      "grad_norm": 0.5953739735003926,
-      "learning_rate": 4.638671875000001e-06,
-      "loss": 0.4949,
+      "epoch": 0.0699249227145591,
+      "grad_norm": 0.6310313258517151,
+      "learning_rate": 4.6614327772325814e-06,
+      "loss": 0.4927,
       "step": 475
     },
     {
-      "epoch": 0.07032967032967033,
-      "grad_norm": 0.5728951469349395,
-      "learning_rate": 4.6875000000000004e-06,
-      "loss": 0.4738,
+      "epoch": 0.0706609745326071,
+      "grad_norm": 0.6181666793600091,
+      "learning_rate": 4.710500490677135e-06,
+      "loss": 0.5014,
       "step": 480
     },
     {
-      "epoch": 0.07106227106227106,
-      "grad_norm": 0.6061175641189019,
-      "learning_rate": 4.736328125e-06,
-      "loss": 0.4758,
+      "epoch": 0.07139702635065509,
+      "grad_norm": 0.585653829471279,
+      "learning_rate": 4.7595682041216886e-06,
+      "loss": 0.4846,
       "step": 485
     },
     {
-      "epoch": 0.07179487179487179,
-      "grad_norm": 0.6124214135481375,
-      "learning_rate": 4.785156250000001e-06,
-      "loss": 0.5028,
+      "epoch": 0.07213307816870308,
+      "grad_norm": 0.5922456519304234,
+      "learning_rate": 4.808635917566242e-06,
+      "loss": 0.4759,
       "step": 490
     },
     {
-      "epoch": 0.07252747252747253,
-      "grad_norm": 0.6074050788211793,
-      "learning_rate": 4.833984375e-06,
-      "loss": 0.4825,
+      "epoch": 0.07286912998675106,
+      "grad_norm": 0.5891457619852838,
+      "learning_rate": 4.857703631010796e-06,
+      "loss": 0.4624,
       "step": 495
     },
     {
-      "epoch": 0.07326007326007326,
-      "grad_norm": 0.5641153492172148,
-      "learning_rate": 4.8828125e-06,
-      "loss": 0.4791,
+      "epoch": 0.07360518180479905,
+      "grad_norm": 0.5845369377813938,
+      "learning_rate": 4.906771344455349e-06,
+      "loss": 0.4669,
       "step": 500
     },
     {
-      "epoch": 0.073992673992674,
-      "grad_norm": 0.589145847070636,
-      "learning_rate": 4.931640625000001e-06,
-      "loss": 0.4775,
+      "epoch": 0.07434123362284704,
+      "grad_norm": 0.56432460580195,
+      "learning_rate": 4.955839057899902e-06,
+      "loss": 0.4937,
       "step": 505
     },
     {
-      "epoch": 0.07472527472527472,
-      "grad_norm": 0.6081676703097589,
-      "learning_rate": 4.98046875e-06,
-      "loss": 0.48,
+      "epoch": 0.07507728544089504,
+      "grad_norm": 0.5659440877842081,
+      "learning_rate": 5.0049067713444555e-06,
+      "loss": 0.4657,
       "step": 510
     },
     {
-      "epoch": 0.07545787545787545,
-      "grad_norm": 0.5812914515901026,
-      "learning_rate": 5.029296875e-06,
-      "loss": 0.48,
+      "epoch": 0.07581333725894303,
+      "grad_norm": 0.5505888303562148,
+      "learning_rate": 5.053974484789009e-06,
+      "loss": 0.467,
       "step": 515
     },
     {
-      "epoch": 0.0761904761904762,
-      "grad_norm": 0.6167465872303627,
-      "learning_rate": 5.078125000000001e-06,
-      "loss": 0.479,
+      "epoch": 0.07654938907699102,
+      "grad_norm": 0.5868518238421928,
+      "learning_rate": 5.1030421982335635e-06,
+      "loss": 0.4826,
       "step": 520
     },
     {
-      "epoch": 0.07692307692307693,
-      "grad_norm": 0.6100423988314366,
-      "learning_rate": 5.126953125e-06,
-      "loss": 0.4855,
+      "epoch": 0.07728544089503901,
+      "grad_norm": 0.6102754783922901,
+      "learning_rate": 5.152109911678116e-06,
+      "loss": 0.4727,
       "step": 525
     },
     {
-      "epoch": 0.07765567765567766,
-      "grad_norm": 0.5873756076460772,
-      "learning_rate": 5.17578125e-06,
-      "loss": 0.4705,
+      "epoch": 0.078021492713087,
+      "grad_norm": 0.5513354797822135,
+      "learning_rate": 5.20117762512267e-06,
+      "loss": 0.4825,
       "step": 530
     },
     {
-      "epoch": 0.07838827838827839,
-      "grad_norm": 0.6379594116334866,
-      "learning_rate": 5.2246093750000006e-06,
-      "loss": 0.4789,
+      "epoch": 0.078757544531135,
+      "grad_norm": 0.5952133347377497,
+      "learning_rate": 5.250245338567223e-06,
+      "loss": 0.4666,
       "step": 535
     },
     {
-      "epoch": 0.07912087912087912,
-      "grad_norm": 0.6043795568282792,
-      "learning_rate": 5.2734375e-06,
-      "loss": 0.4904,
+      "epoch": 0.07949359634918299,
+      "grad_norm": 0.6057839823052446,
+      "learning_rate": 5.299313052011776e-06,
+      "loss": 0.4859,
       "step": 540
     },
     {
-      "epoch": 0.07985347985347985,
-      "grad_norm": 0.5627871797826735,
-      "learning_rate": 5.322265625000001e-06,
-      "loss": 0.4732,
+      "epoch": 0.08022964816723098,
+      "grad_norm": 0.6149832759183071,
+      "learning_rate": 5.3483807654563304e-06,
+      "loss": 0.4627,
       "step": 545
     },
     {
-      "epoch": 0.08058608058608059,
-      "grad_norm": 0.5772953048482563,
-      "learning_rate": 5.3710937500000005e-06,
-      "loss": 0.4714,
+      "epoch": 0.08096569998527896,
+      "grad_norm": 0.5625685874675828,
+      "learning_rate": 5.397448478900884e-06,
+      "loss": 0.4795,
       "step": 550
     },
     {
-      "epoch": 0.08131868131868132,
-      "grad_norm": 0.5989279963185531,
-      "learning_rate": 5.419921875e-06,
-      "loss": 0.48,
+      "epoch": 0.08170175180332695,
+      "grad_norm": 0.5695634420904787,
+      "learning_rate": 5.446516192345437e-06,
+      "loss": 0.4917,
       "step": 555
     },
     {
-      "epoch": 0.08205128205128205,
-      "grad_norm": 0.5603152309391928,
-      "learning_rate": 5.468750000000001e-06,
-      "loss": 0.48,
+      "epoch": 0.08243780362137494,
+      "grad_norm": 0.5882207985851441,
+      "learning_rate": 5.49558390578999e-06,
+      "loss": 0.4904,
       "step": 560
     },
     {
-      "epoch": 0.08278388278388278,
-      "grad_norm": 0.548779034619482,
-      "learning_rate": 5.5175781250000005e-06,
-      "loss": 0.4693,
+      "epoch": 0.08317385543942293,
+      "grad_norm": 0.565711510623109,
+      "learning_rate": 5.544651619234545e-06,
+      "loss": 0.4682,
       "step": 565
     },
     {
-      "epoch": 0.08351648351648351,
-      "grad_norm": 0.5994436565592611,
-      "learning_rate": 5.56640625e-06,
-      "loss": 0.4813,
+      "epoch": 0.08390990725747093,
+      "grad_norm": 0.5825425744408332,
+      "learning_rate": 5.593719332679097e-06,
+      "loss": 0.47,
       "step": 570
     },
     {
-      "epoch": 0.08424908424908426,
-      "grad_norm": 0.6314474035157444,
-      "learning_rate": 5.615234375000001e-06,
-      "loss": 0.4784,
+      "epoch": 0.08464595907551892,
+      "grad_norm": 0.5816544517858435,
+      "learning_rate": 5.642787046123651e-06,
+      "loss": 0.4719,
       "step": 575
     },
     {
-      "epoch": 0.08498168498168499,
-      "grad_norm": 0.6246412942815367,
-      "learning_rate": 5.6640625000000005e-06,
-      "loss": 0.4741,
+      "epoch": 0.08538201089356691,
+      "grad_norm": 0.5896285603216627,
+      "learning_rate": 5.691854759568205e-06,
+      "loss": 0.4912,
       "step": 580
     },
     {
-      "epoch": 0.08571428571428572,
-      "grad_norm": 0.5866396752374676,
-      "learning_rate": 5.712890625e-06,
-      "loss": 0.4802,
+      "epoch": 0.0861180627116149,
+      "grad_norm": 0.6117545163317246,
+      "learning_rate": 5.740922473012758e-06,
+      "loss": 0.4792,
       "step": 585
     },
     {
-      "epoch": 0.08644688644688645,
-      "grad_norm": 0.5904192652410284,
-      "learning_rate": 5.761718750000001e-06,
-      "loss": 0.4565,
+      "epoch": 0.08685411452966289,
+      "grad_norm": 0.5773620560230253,
+      "learning_rate": 5.789990186457312e-06,
+      "loss": 0.4877,
       "step": 590
     },
     {
-      "epoch": 0.08717948717948718,
-      "grad_norm": 0.5519252352856109,
-      "learning_rate": 5.810546875e-06,
-      "loss": 0.4475,
+      "epoch": 0.08759016634771088,
+      "grad_norm": 0.6601154780260813,
+      "learning_rate": 5.839057899901865e-06,
+      "loss": 0.4691,
       "step": 595
     },
     {
-      "epoch": 0.08791208791208792,
-      "grad_norm": 0.5948751037148314,
-      "learning_rate": 5.859375e-06,
-      "loss": 0.4717,
+      "epoch": 0.08832621816575886,
+      "grad_norm": 0.5633119014086018,
+      "learning_rate": 5.888125613346418e-06,
+      "loss": 0.4685,
       "step": 600
     },
     {
-      "epoch": 0.08864468864468865,
-      "grad_norm": 0.5827642155370751,
-      "learning_rate": 5.908203125000001e-06,
-      "loss": 0.47,
+      "epoch": 0.08906226998380685,
+      "grad_norm": 0.5798833556082699,
+      "learning_rate": 5.937193326790972e-06,
+      "loss": 0.4845,
       "step": 605
     },
     {
-      "epoch": 0.08937728937728938,
-      "grad_norm": 0.5904415040624938,
-      "learning_rate": 5.95703125e-06,
-      "loss": 0.4668,
+      "epoch": 0.08979832180185485,
+      "grad_norm": 0.6091643398421105,
+      "learning_rate": 5.986261040235526e-06,
+      "loss": 0.4524,
       "step": 610
     },
     {
-      "epoch": 0.09010989010989011,
-      "grad_norm": 0.5761997438099439,
-      "learning_rate": 6.005859375e-06,
-      "loss": 0.4752,
+      "epoch": 0.09053437361990284,
+      "grad_norm": 0.5565575718179494,
+      "learning_rate": 6.035328753680079e-06,
+      "loss": 0.4557,
       "step": 615
     },
     {
-      "epoch": 0.09084249084249084,
-      "grad_norm": 0.5489736943110936,
-      "learning_rate": 6.054687500000001e-06,
-      "loss": 0.4566,
+      "epoch": 0.09127042543795083,
+      "grad_norm": 0.5644936245127977,
+      "learning_rate": 6.084396467124632e-06,
+      "loss": 0.4732,
       "step": 620
     },
     {
-      "epoch": 0.09157509157509157,
-      "grad_norm": 0.5597636269927064,
-      "learning_rate": 6.103515625e-06,
-      "loss": 0.4813,
+      "epoch": 0.09200647725599882,
+      "grad_norm": 0.5790776725696446,
+      "learning_rate": 6.1334641805691865e-06,
+      "loss": 0.478,
       "step": 625
     },
     {
-      "epoch": 0.09230769230769231,
-      "grad_norm": 0.5714338556216557,
-      "learning_rate": 6.152343750000001e-06,
-      "loss": 0.4735,
+      "epoch": 0.09274252907404681,
+      "grad_norm": 0.5643831879058268,
+      "learning_rate": 6.182531894013739e-06,
+      "loss": 0.4586,
       "step": 630
     },
     {
-      "epoch": 0.09304029304029304,
-      "grad_norm": 0.5952178258097818,
-      "learning_rate": 6.201171875000001e-06,
-      "loss": 0.4826,
+      "epoch": 0.0934785808920948,
+      "grad_norm": 0.5837458919285372,
+      "learning_rate": 6.231599607458293e-06,
+      "loss": 0.4866,
       "step": 635
     },
     {
-      "epoch": 0.09377289377289377,
-      "grad_norm": 0.5884694242344789,
-      "learning_rate": 6.25e-06,
-      "loss": 0.4646,
+      "epoch": 0.0942146327101428,
+      "grad_norm": 0.5566898853661357,
+      "learning_rate": 6.280667320902847e-06,
+      "loss": 0.4685,
       "step": 640
     },
     {
-      "epoch": 0.0945054945054945,
-      "grad_norm": 0.5933386033268863,
-      "learning_rate": 6.298828125000001e-06,
-      "loss": 0.4608,
+      "epoch": 0.09495068452819079,
+      "grad_norm": 0.5812564266093952,
+      "learning_rate": 6.3297350343474e-06,
+      "loss": 0.4749,
       "step": 645
     },
     {
-      "epoch": 0.09523809523809523,
-      "grad_norm": 0.6037443715273233,
-      "learning_rate": 6.3476562500000006e-06,
-      "loss": 0.4637,
+      "epoch": 0.09568673634623878,
+      "grad_norm": 0.5903324398235352,
+      "learning_rate": 6.3788027477919535e-06,
+      "loss": 0.4765,
       "step": 650
     },
     {
-      "epoch": 0.09597069597069598,
-      "grad_norm": 0.6163005967307136,
-      "learning_rate": 6.396484375e-06,
-      "loss": 0.4731,
+      "epoch": 0.09642278816428676,
+      "grad_norm": 0.6304091374610179,
+      "learning_rate": 6.427870461236507e-06,
+      "loss": 0.4756,
       "step": 655
     },
     {
-      "epoch": 0.0967032967032967,
-      "grad_norm": 0.5453447185366223,
-      "learning_rate": 6.445312500000001e-06,
-      "loss": 0.469,
+      "epoch": 0.09715883998233475,
+      "grad_norm": 0.586486674113841,
+      "learning_rate": 6.47693817468106e-06,
+      "loss": 0.4681,
       "step": 660
     },
     {
-      "epoch": 0.09743589743589744,
-      "grad_norm": 0.6212943621659923,
-      "learning_rate": 6.4941406250000005e-06,
-      "loss": 0.466,
+      "epoch": 0.09789489180038274,
+      "grad_norm": 0.5893707646862878,
+      "learning_rate": 6.526005888125614e-06,
+      "loss": 0.4815,
       "step": 665
     },
     {
-      "epoch": 0.09816849816849817,
-      "grad_norm": 0.5309135046878034,
-      "learning_rate": 6.54296875e-06,
-      "loss": 0.4626,
+      "epoch": 0.09863094361843074,
+      "grad_norm": 0.5642675537976032,
+      "learning_rate": 6.575073601570168e-06,
+      "loss": 0.4681,
       "step": 670
     },
     {
-      "epoch": 0.0989010989010989,
-      "grad_norm": 0.5382683389928992,
-      "learning_rate": 6.591796875000001e-06,
-      "loss": 0.4838,
+      "epoch": 0.09936699543647873,
+      "grad_norm": 0.5612248341901227,
+      "learning_rate": 6.6241413150147205e-06,
+      "loss": 0.4846,
       "step": 675
     },
     {
-      "epoch": 0.09963369963369964,
-      "grad_norm": 0.5827291346723401,
-      "learning_rate": 6.6406250000000005e-06,
-      "loss": 0.4905,
+      "epoch": 0.10010304725452672,
+      "grad_norm": 0.5834664312871857,
+      "learning_rate": 6.673209028459274e-06,
+      "loss": 0.4834,
       "step": 680
     },
     {
-      "epoch": 0.10036630036630037,
-      "grad_norm": 0.532784806632675,
-      "learning_rate": 6.689453125e-06,
-      "loss": 0.4744,
+      "epoch": 0.10083909907257471,
+      "grad_norm": 0.629128155687521,
+      "learning_rate": 6.722276741903828e-06,
+      "loss": 0.4722,
       "step": 685
     },
     {
-      "epoch": 0.1010989010989011,
-      "grad_norm": 0.5838167742764417,
-      "learning_rate": 6.738281250000001e-06,
-      "loss": 0.4703,
+      "epoch": 0.1015751508906227,
+      "grad_norm": 0.5828728916316157,
+      "learning_rate": 6.771344455348381e-06,
+      "loss": 0.4645,
       "step": 690
     },
     {
-      "epoch": 0.10183150183150183,
-      "grad_norm": 0.5376363462627745,
-      "learning_rate": 6.7871093750000004e-06,
-      "loss": 0.4661,
+      "epoch": 0.1023112027086707,
+      "grad_norm": 0.6490735802843812,
+      "learning_rate": 6.820412168792935e-06,
+      "loss": 0.4759,
       "step": 695
     },
     {
-      "epoch": 0.10256410256410256,
-      "grad_norm": 0.5239263973505636,
-      "learning_rate": 6.8359375e-06,
-      "loss": 0.4579,
+      "epoch": 0.10304725452671869,
+      "grad_norm": 0.5807842256629684,
+      "learning_rate": 6.869479882237488e-06,
+      "loss": 0.472,
       "step": 700
     },
     {
-      "epoch": 0.10329670329670329,
-      "grad_norm": 0.5905645370609754,
-      "learning_rate": 6.884765625000001e-06,
-      "loss": 0.4619,
+      "epoch": 0.10378330634476667,
+      "grad_norm": 0.6076870406724216,
+      "learning_rate": 6.918547595682041e-06,
+      "loss": 0.4553,
       "step": 705
     },
     {
-      "epoch": 0.10402930402930403,
-      "grad_norm": 0.5700211959784208,
-      "learning_rate": 6.93359375e-06,
-      "loss": 0.4557,
+      "epoch": 0.10451935816281466,
+      "grad_norm": 0.5648532124295207,
+      "learning_rate": 6.967615309126595e-06,
+      "loss": 0.4807,
       "step": 710
     },
     {
-      "epoch": 0.10476190476190476,
-      "grad_norm": 0.5903415816955252,
-      "learning_rate": 6.982421875000001e-06,
-      "loss": 0.4571,
+      "epoch": 0.10525540998086265,
+      "grad_norm": 0.5969143940699317,
+      "learning_rate": 7.016683022571149e-06,
+      "loss": 0.473,
       "step": 715
     },
     {
-      "epoch": 0.1054945054945055,
-      "grad_norm": 0.5574565092969472,
-      "learning_rate": 7.031250000000001e-06,
-      "loss": 0.4606,
+      "epoch": 0.10599146179891064,
+      "grad_norm": 0.6022013779305591,
+      "learning_rate": 7.065750736015702e-06,
+      "loss": 0.4619,
       "step": 720
     },
     {
-      "epoch": 0.10622710622710622,
-      "grad_norm": 0.5861130971423574,
-      "learning_rate": 7.080078125e-06,
-      "loss": 0.4594,
+      "epoch": 0.10672751361695863,
+      "grad_norm": 0.5458678918831419,
+      "learning_rate": 7.114818449460256e-06,
+      "loss": 0.4833,
       "step": 725
     },
     {
-      "epoch": 0.10695970695970695,
-      "grad_norm": 0.5486960784949657,
-      "learning_rate": 7.128906250000001e-06,
-      "loss": 0.4984,
+      "epoch": 0.10746356543500662,
+      "grad_norm": 0.5355193744561354,
+      "learning_rate": 7.16388616290481e-06,
+      "loss": 0.4761,
       "step": 730
     },
     {
-      "epoch": 0.1076923076923077,
-      "grad_norm": 0.557733033021095,
-      "learning_rate": 7.177734375000001e-06,
-      "loss": 0.4655,
+      "epoch": 0.10819961725305462,
+      "grad_norm": 0.573431459253995,
+      "learning_rate": 7.212953876349362e-06,
+      "loss": 0.4681,
       "step": 735
     },
     {
-      "epoch": 0.10842490842490843,
-      "grad_norm": 0.5436179967978834,
-      "learning_rate": 7.2265625e-06,
-      "loss": 0.4686,
+      "epoch": 0.10893566907110261,
+      "grad_norm": 0.5288123343977066,
+      "learning_rate": 7.262021589793916e-06,
+      "loss": 0.4631,
       "step": 740
     },
     {
-      "epoch": 0.10915750915750916,
-      "grad_norm": 0.5724129400787233,
-      "learning_rate": 7.275390625000001e-06,
-      "loss": 0.4661,
+      "epoch": 0.1096717208891506,
+      "grad_norm": 0.5697904981825487,
+      "learning_rate": 7.31108930323847e-06,
+      "loss": 0.4715,
       "step": 745
     },
     {
-      "epoch": 0.10989010989010989,
-      "grad_norm": 0.5466105215938837,
-      "learning_rate": 7.3242187500000006e-06,
-      "loss": 0.4731,
+      "epoch": 0.11040777270719859,
+      "grad_norm": 0.5208534768593922,
+      "learning_rate": 7.360157016683023e-06,
+      "loss": 0.4665,
       "step": 750
     },
     {
-      "epoch": 0.11062271062271062,
-      "grad_norm": 0.5768341262782312,
-      "learning_rate": 7.373046875e-06,
-      "loss": 0.4822,
+      "epoch": 0.11114382452524657,
+      "grad_norm": 0.5498377158829724,
+      "learning_rate": 7.4092247301275766e-06,
+      "loss": 0.4714,
       "step": 755
     },
     {
-      "epoch": 0.11135531135531136,
-      "grad_norm": 0.6159551594829057,
-      "learning_rate": 7.421875000000001e-06,
-      "loss": 0.476,
+      "epoch": 0.11187987634329456,
+      "grad_norm": 0.5675244832493291,
+      "learning_rate": 7.45829244357213e-06,
+      "loss": 0.457,
       "step": 760
     },
     {
-      "epoch": 0.11208791208791209,
-      "grad_norm": 0.6072974251996277,
-      "learning_rate": 7.4707031250000005e-06,
-      "loss": 0.4782,
+      "epoch": 0.11261592816134255,
+      "grad_norm": 0.5347391538454078,
+      "learning_rate": 7.507360157016683e-06,
+      "loss": 0.4657,
       "step": 765
     },
     {
-      "epoch": 0.11282051282051282,
-      "grad_norm": 0.5871442389262,
-      "learning_rate": 7.51953125e-06,
-      "loss": 0.4494,
+      "epoch": 0.11335197997939055,
+      "grad_norm": 0.5583228209281907,
+      "learning_rate": 7.556427870461237e-06,
+      "loss": 0.4764,
       "step": 770
     },
     {
-      "epoch": 0.11355311355311355,
-      "grad_norm": 0.602204799204848,
-      "learning_rate": 7.568359375000001e-06,
-      "loss": 0.4719,
+      "epoch": 0.11408803179743854,
+      "grad_norm": 0.5461223520236801,
+      "learning_rate": 7.605495583905791e-06,
+      "loss": 0.4667,
       "step": 775
     },
     {
-      "epoch": 0.11428571428571428,
-      "grad_norm": 0.5813186170396549,
-      "learning_rate": 7.6171875000000005e-06,
-      "loss": 0.4788,
+      "epoch": 0.11482408361548653,
+      "grad_norm": 0.5663027522324963,
+      "learning_rate": 7.654563297350344e-06,
+      "loss": 0.4701,
       "step": 780
     },
     {
-      "epoch": 0.11501831501831501,
-      "grad_norm": 0.6304842458401531,
-      "learning_rate": 7.666015625000001e-06,
-      "loss": 0.492,
+      "epoch": 0.11556013543353452,
+      "grad_norm": 0.5556645456848583,
+      "learning_rate": 7.703631010794898e-06,
+      "loss": 0.475,
       "step": 785
     },
     {
-      "epoch": 0.11575091575091576,
-      "grad_norm": 0.5975861262153394,
-      "learning_rate": 7.71484375e-06,
-      "loss": 0.4716,
+      "epoch": 0.11629618725158251,
+      "grad_norm": 0.5202037674063839,
+      "learning_rate": 7.75269872423945e-06,
+      "loss": 0.4582,
       "step": 790
     },
     {
-      "epoch": 0.11648351648351649,
-      "grad_norm": 0.5356317621558694,
-      "learning_rate": 7.763671875e-06,
-      "loss": 0.4504,
+      "epoch": 0.1170322390696305,
+      "grad_norm": 0.5621800088114947,
+      "learning_rate": 7.801766437684003e-06,
+      "loss": 0.4745,
       "step": 795
     },
     {
-      "epoch": 0.11721611721611722,
-      "grad_norm": 0.5559709134630877,
-      "learning_rate": 7.8125e-06,
-      "loss": 0.4712,
+      "epoch": 0.1177682908876785,
+      "grad_norm": 0.5357544493332862,
+      "learning_rate": 7.850834151128558e-06,
+      "loss": 0.4654,
       "step": 800
     },
     {
-      "epoch": 0.11794871794871795,
-      "grad_norm": 0.5629653822120899,
-      "learning_rate": 7.861328125e-06,
-      "loss": 0.4678,
+      "epoch": 0.11850434270572649,
+      "grad_norm": 0.5419172883253839,
+      "learning_rate": 7.899901864573112e-06,
+      "loss": 0.4735,
       "step": 805
     },
     {
-      "epoch": 0.11868131868131868,
-      "grad_norm": 0.5403593741271583,
-      "learning_rate": 7.910156250000001e-06,
-      "loss": 0.4771,
+      "epoch": 0.11924039452377447,
+      "grad_norm": 0.5525060417009166,
+      "learning_rate": 7.948969578017665e-06,
+      "loss": 0.4709,
       "step": 810
     },
     {
-      "epoch": 0.11941391941391942,
-      "grad_norm": 0.5386233851491723,
-      "learning_rate": 7.958984375000001e-06,
-      "loss": 0.4762,
+      "epoch": 0.11997644634182246,
+      "grad_norm": 0.5448040937044886,
+      "learning_rate": 7.998037291462218e-06,
+      "loss": 0.4664,
       "step": 815
     },
     {
-      "epoch": 0.12014652014652015,
-      "grad_norm": 0.5134313568682624,
-      "learning_rate": 8.0078125e-06,
-      "loss": 0.4646,
+      "epoch": 0.12071249815987045,
+      "grad_norm": 0.5486116584488951,
+      "learning_rate": 8.047105004906772e-06,
+      "loss": 0.4677,
       "step": 820
     },
     {
-      "epoch": 0.12087912087912088,
-      "grad_norm": 0.5707071480675963,
-      "learning_rate": 8.056640625e-06,
-      "loss": 0.4813,
+      "epoch": 0.12144854997791844,
+      "grad_norm": 0.5965699643929866,
+      "learning_rate": 8.096172718351325e-06,
+      "loss": 0.4677,
       "step": 825
     },
     {
-      "epoch": 0.12161172161172161,
-      "grad_norm": 1.057465377362691,
-      "learning_rate": 8.10546875e-06,
-      "loss": 0.4772,
+      "epoch": 0.12218460179596644,
+      "grad_norm": 0.7738268323135488,
+      "learning_rate": 8.145240431795879e-06,
+      "loss": 0.4896,
       "step": 830
     },
     {
-      "epoch": 0.12234432234432234,
-      "grad_norm": 0.7969984191209428,
-      "learning_rate": 8.154296875000001e-06,
-      "loss": 0.4616,
+      "epoch": 0.12292065361401443,
+      "grad_norm": 0.5502489638196433,
+      "learning_rate": 8.194308145240434e-06,
+      "loss": 0.4704,
       "step": 835
     },
     {
-      "epoch": 0.12307692307692308,
-      "grad_norm": 0.5508120508001506,
-      "learning_rate": 8.203125000000001e-06,
-      "loss": 0.4805,
+      "epoch": 0.12365670543206242,
+      "grad_norm": 0.5641949189118746,
+      "learning_rate": 8.243375858684986e-06,
+      "loss": 0.4764,
       "step": 840
     },
     {
-      "epoch": 0.12380952380952381,
-      "grad_norm": 0.5450863996688757,
-      "learning_rate": 8.251953125000001e-06,
-      "loss": 0.4853,
+      "epoch": 0.12439275725011041,
+      "grad_norm": 0.5289076252333902,
+      "learning_rate": 8.292443572129539e-06,
+      "loss": 0.4751,
       "step": 845
     },
     {
-      "epoch": 0.12454212454212454,
-      "grad_norm": 0.548584060543633,
-      "learning_rate": 8.30078125e-06,
-      "loss": 0.4635,
+      "epoch": 0.1251288090681584,
+      "grad_norm": 0.5476369574245666,
+      "learning_rate": 8.341511285574093e-06,
+      "loss": 0.4683,
       "step": 850
     },
     {
-      "epoch": 0.12527472527472527,
-      "grad_norm": 0.54496105370757,
-      "learning_rate": 8.349609375e-06,
-      "loss": 0.4769,
+      "epoch": 0.12586486088620638,
+      "grad_norm": 0.5498288749893551,
+      "learning_rate": 8.390578999018646e-06,
+      "loss": 0.494,
       "step": 855
     },
     {
-      "epoch": 0.126007326007326,
-      "grad_norm": 0.6056793412035149,
-      "learning_rate": 8.3984375e-06,
-      "loss": 0.4714,
+      "epoch": 0.1266009127042544,
+      "grad_norm": 0.5249158510086049,
+      "learning_rate": 8.4396467124632e-06,
+      "loss": 0.4874,
       "step": 860
     },
     {
-      "epoch": 0.12673992673992673,
-      "grad_norm": 0.5649345221606744,
-      "learning_rate": 8.447265625000001e-06,
-      "loss": 0.483,
+      "epoch": 0.12733696452230236,
+      "grad_norm": 0.5826503894307695,
+      "learning_rate": 8.488714425907753e-06,
+      "loss": 0.4794,
       "step": 865
     },
     {
-      "epoch": 0.12747252747252746,
-      "grad_norm": 0.5616348293821443,
-      "learning_rate": 8.496093750000001e-06,
-      "loss": 0.4757,
+      "epoch": 0.12807301634035037,
+      "grad_norm": 0.6067612119654865,
+      "learning_rate": 8.537782139352306e-06,
+      "loss": 0.465,
       "step": 870
     },
     {
-      "epoch": 0.1282051282051282,
-      "grad_norm": 0.5611441794672326,
-      "learning_rate": 8.544921875e-06,
-      "loss": 0.469,
+      "epoch": 0.12880906815839835,
+      "grad_norm": 0.5417276894889486,
+      "learning_rate": 8.58684985279686e-06,
+      "loss": 0.4674,
       "step": 875
     },
     {
-      "epoch": 0.12893772893772895,
-      "grad_norm": 0.5405096788454254,
-      "learning_rate": 8.59375e-06,
-      "loss": 0.4847,
+      "epoch": 0.12954511997644635,
+      "grad_norm": 0.523838175362927,
+      "learning_rate": 8.635917566241415e-06,
+      "loss": 0.469,
       "step": 880
     },
     {
-      "epoch": 0.12967032967032968,
-      "grad_norm": 0.5593871863023577,
-      "learning_rate": 8.642578125e-06,
-      "loss": 0.4709,
+      "epoch": 0.13028117179449433,
+      "grad_norm": 0.5409098167007628,
+      "learning_rate": 8.684985279685967e-06,
+      "loss": 0.4749,
       "step": 885
     },
     {
-      "epoch": 0.1304029304029304,
-      "grad_norm": 0.5219134179461249,
-      "learning_rate": 8.69140625e-06,
-      "loss": 0.4763,
+      "epoch": 0.1310172236125423,
+      "grad_norm": 0.5404589881617674,
+      "learning_rate": 8.73405299313052e-06,
+      "loss": 0.4659,
       "step": 890
     },
     {
-      "epoch": 0.13113553113553114,
-      "grad_norm": 0.5402339547204121,
-      "learning_rate": 8.740234375000001e-06,
-      "loss": 0.4665,
+      "epoch": 0.13175327543059032,
+      "grad_norm": 0.5783971691158564,
+      "learning_rate": 8.783120706575075e-06,
+      "loss": 0.4452,
       "step": 895
     },
     {
-      "epoch": 0.13186813186813187,
-      "grad_norm": 0.5017739581429062,
-      "learning_rate": 8.789062500000001e-06,
-      "loss": 0.4655,
+      "epoch": 0.1324893272486383,
+      "grad_norm": 0.5182432824868518,
+      "learning_rate": 8.832188420019627e-06,
+      "loss": 0.4683,
       "step": 900
     },
     {
-      "epoch": 0.1326007326007326,
-      "grad_norm": 0.5398065464611551,
-      "learning_rate": 8.837890625e-06,
-      "loss": 0.4685,
+      "epoch": 0.1332253790666863,
+      "grad_norm": 0.5385422506386383,
+      "learning_rate": 8.881256133464182e-06,
+      "loss": 0.4607,
       "step": 905
     },
     {
-      "epoch": 0.13333333333333333,
-      "grad_norm": 0.539672747550422,
-      "learning_rate": 8.88671875e-06,
-      "loss": 0.4646,
+      "epoch": 0.13396143088473428,
+      "grad_norm": 0.570114872389286,
+      "learning_rate": 8.930323846908734e-06,
+      "loss": 0.4725,
       "step": 910
     },
     {
-      "epoch": 0.13406593406593406,
-      "grad_norm": 0.5649222602250018,
-      "learning_rate": 8.935546875e-06,
-      "loss": 0.4712,
+      "epoch": 0.13469748270278228,
+      "grad_norm": 0.5638061126176769,
+      "learning_rate": 8.979391560353287e-06,
+      "loss": 0.4894,
       "step": 915
     },
     {
-      "epoch": 0.1347985347985348,
-      "grad_norm": 0.555978425517813,
-      "learning_rate": 8.984375000000002e-06,
-      "loss": 0.4797,
+      "epoch": 0.13543353452083026,
+      "grad_norm": 0.5284798481462458,
+      "learning_rate": 9.028459273797842e-06,
+      "loss": 0.4599,
       "step": 920
     },
     {
-      "epoch": 0.13553113553113552,
-      "grad_norm": 0.5451150959046317,
-      "learning_rate": 9.033203125000001e-06,
-      "loss": 0.4608,
+      "epoch": 0.13616958633887827,
+      "grad_norm": 0.545300441979569,
+      "learning_rate": 9.077526987242396e-06,
+      "loss": 0.4736,
       "step": 925
     },
     {
-      "epoch": 0.13626373626373625,
-      "grad_norm": 0.6136799089282742,
-      "learning_rate": 9.082031250000001e-06,
-      "loss": 0.465,
+      "epoch": 0.13690563815692625,
+      "grad_norm": 0.5443738398973984,
+      "learning_rate": 9.126594700686949e-06,
+      "loss": 0.475,
       "step": 930
     },
     {
-      "epoch": 0.136996336996337,
-      "grad_norm": 0.5392919758975416,
-      "learning_rate": 9.130859375e-06,
-      "loss": 0.4737,
+      "epoch": 0.13764168997497425,
+      "grad_norm": 0.5282020418222877,
+      "learning_rate": 9.175662414131501e-06,
+      "loss": 0.4671,
       "step": 935
     },
     {
-      "epoch": 0.13772893772893774,
-      "grad_norm": 0.5433299746227306,
-      "learning_rate": 9.1796875e-06,
-      "loss": 0.4618,
+      "epoch": 0.13837774179302223,
+      "grad_norm": 0.5228118303753154,
+      "learning_rate": 9.224730127576056e-06,
+      "loss": 0.4723,
       "step": 940
     },
     {
-      "epoch": 0.13846153846153847,
-      "grad_norm": 0.5269797392518154,
-      "learning_rate": 9.228515625e-06,
-      "loss": 0.4663,
+      "epoch": 0.1391137936110702,
+      "grad_norm": 0.5191901819627673,
+      "learning_rate": 9.273797841020608e-06,
+      "loss": 0.4808,
       "step": 945
     },
     {
-      "epoch": 0.1391941391941392,
-      "grad_norm": 0.5303507680071278,
-      "learning_rate": 9.277343750000001e-06,
-      "loss": 0.4793,
+      "epoch": 0.1398498454291182,
+      "grad_norm": 0.5293478366717721,
+      "learning_rate": 9.322865554465163e-06,
+      "loss": 0.4766,
       "step": 950
     },
     {
-      "epoch": 0.13992673992673993,
-      "grad_norm": 0.5739728476514406,
-      "learning_rate": 9.326171875000001e-06,
-      "loss": 0.4891,
+      "epoch": 0.1405858972471662,
+      "grad_norm": 0.5354175876597523,
+      "learning_rate": 9.371933267909716e-06,
+      "loss": 0.4813,
       "step": 955
     },
     {
-      "epoch": 0.14065934065934066,
-      "grad_norm": 0.5720121486255587,
-      "learning_rate": 9.375000000000001e-06,
-      "loss": 0.4781,
+      "epoch": 0.1413219490652142,
+      "grad_norm": 0.5605277944405026,
+      "learning_rate": 9.42100098135427e-06,
+      "loss": 0.4662,
       "step": 960
     },
     {
-      "epoch": 0.1413919413919414,
-      "grad_norm": 0.5226798686905856,
-      "learning_rate": 9.423828125e-06,
-      "loss": 0.4633,
+      "epoch": 0.14205800088326218,
+      "grad_norm": 0.5248861064318012,
+      "learning_rate": 9.470068694798823e-06,
+      "loss": 0.4652,
       "step": 965
     },
     {
-      "epoch": 0.14212454212454212,
-      "grad_norm": 0.5183949318801034,
-      "learning_rate": 9.47265625e-06,
-      "loss": 0.4488,
+      "epoch": 0.14279405270131018,
+      "grad_norm": 0.515307674108716,
+      "learning_rate": 9.519136408243377e-06,
+      "loss": 0.4571,
       "step": 970
     },
     {
-      "epoch": 0.14285714285714285,
-      "grad_norm": 0.530336983746849,
-      "learning_rate": 9.521484375e-06,
-      "loss": 0.4526,
+      "epoch": 0.14353010451935816,
+      "grad_norm": 0.5538399997823142,
+      "learning_rate": 9.56820412168793e-06,
+      "loss": 0.4708,
       "step": 975
     },
     {
-      "epoch": 0.14358974358974358,
-      "grad_norm": 0.5436835706956455,
-      "learning_rate": 9.570312500000001e-06,
-      "loss": 0.4716,
+      "epoch": 0.14426615633740617,
+      "grad_norm": 0.5093828223702961,
+      "learning_rate": 9.617271835132484e-06,
+      "loss": 0.4546,
       "step": 980
     },
     {
-      "epoch": 0.1443223443223443,
-      "grad_norm": 0.6066464639380588,
-      "learning_rate": 9.619140625000001e-06,
-      "loss": 0.4836,
+      "epoch": 0.14500220815545414,
+      "grad_norm": 0.5508697416192083,
+      "learning_rate": 9.666339548577037e-06,
+      "loss": 0.4627,
       "step": 985
     },
     {
-      "epoch": 0.14505494505494507,
-      "grad_norm": 0.5554488216735801,
-      "learning_rate": 9.66796875e-06,
-      "loss": 0.4524,
+      "epoch": 0.14573825997350212,
+      "grad_norm": 0.5480456861919574,
+      "learning_rate": 9.715407262021591e-06,
+      "loss": 0.4589,
       "step": 990
     },
     {
-      "epoch": 0.1457875457875458,
-      "grad_norm": 0.5478213717376119,
-      "learning_rate": 9.716796875e-06,
-      "loss": 0.4821,
+      "epoch": 0.14647431179155013,
+      "grad_norm": 0.5357831384000477,
+      "learning_rate": 9.764474975466144e-06,
+      "loss": 0.4737,
       "step": 995
     },
     {
-      "epoch": 0.14652014652014653,
-      "grad_norm": 0.5471748160993797,
-      "learning_rate": 9.765625e-06,
-      "loss": 0.488,
+      "epoch": 0.1472103636095981,
+      "grad_norm": 0.5743875070998096,
+      "learning_rate": 9.813542688910698e-06,
+      "loss": 0.4664,
       "step": 1000
     },
     {
-      "epoch": 0.14725274725274726,
-      "grad_norm": 0.5200793639005971,
-      "learning_rate": 9.814453125000002e-06,
-      "loss": 0.4639,
+      "epoch": 0.1479464154276461,
+      "grad_norm": 0.5285569772929545,
+      "learning_rate": 9.862610402355251e-06,
+      "loss": 0.4684,
       "step": 1005
     },
     {
-      "epoch": 0.147985347985348,
-      "grad_norm": 0.5416684777484602,
-      "learning_rate": 9.863281250000001e-06,
-      "loss": 0.4661,
+      "epoch": 0.1486824672456941,
+      "grad_norm": 0.5658895722799637,
+      "learning_rate": 9.911678115799804e-06,
+      "loss": 0.4727,
       "step": 1010
     },
     {
-      "epoch": 0.14871794871794872,
-      "grad_norm": 0.5416594075142821,
-      "learning_rate": 9.912109375000001e-06,
-      "loss": 0.4564,
+      "epoch": 0.1494185190637421,
+      "grad_norm": 0.5046840364272455,
+      "learning_rate": 9.960745829244358e-06,
+      "loss": 0.4691,
       "step": 1015
     },
     {
-      "epoch": 0.14945054945054945,
-      "grad_norm": 0.5303288476884193,
-      "learning_rate": 9.9609375e-06,
-      "loss": 0.461,
+      "epoch": 0.15015457088179007,
+      "grad_norm": 0.5063812234251259,
+      "learning_rate": 1.0009813542688911e-05,
+      "loss": 0.4661,
       "step": 1020
     },
     {
-      "epoch": 0.15018315018315018,
-      "grad_norm": 0.5571714281369051,
-      "learning_rate": 1.0009765625e-05,
-      "loss": 0.4854,
+      "epoch": 0.15089062269983808,
+      "grad_norm": 0.5282738769028222,
+      "learning_rate": 1.0058881256133465e-05,
+      "loss": 0.4872,
       "step": 1025
     },
     {
-      "epoch": 0.1509157509157509,
-      "grad_norm": 0.6193668370642765,
-      "learning_rate": 1.005859375e-05,
-      "loss": 0.4803,
+      "epoch": 0.15162667451788606,
+      "grad_norm": 0.6513678207446305,
+      "learning_rate": 1.0107948969578018e-05,
+      "loss": 0.4664,
       "step": 1030
     },
     {
-      "epoch": 0.15164835164835164,
-      "grad_norm": 0.5383083934721935,
-      "learning_rate": 1.0107421875000002e-05,
-      "loss": 0.4604,
+      "epoch": 0.15236272633593406,
+      "grad_norm": 0.5162653018671886,
+      "learning_rate": 1.0157016683022571e-05,
+      "loss": 0.458,
       "step": 1035
     },
     {
-      "epoch": 0.1523809523809524,
-      "grad_norm": 0.5663562446139142,
-      "learning_rate": 1.0156250000000001e-05,
-      "loss": 0.4758,
+      "epoch": 0.15309877815398204,
+      "grad_norm": 0.5153888899075103,
+      "learning_rate": 1.0206084396467127e-05,
+      "loss": 0.4868,
       "step": 1040
     },
     {
-      "epoch": 0.15311355311355312,
-      "grad_norm": 0.5668737710162076,
-      "learning_rate": 1.0205078125000001e-05,
-      "loss": 0.4543,
+      "epoch": 0.15383482997203002,
+      "grad_norm": 0.5622003345509735,
+      "learning_rate": 1.025515210991168e-05,
+      "loss": 0.4898,
       "step": 1045
     },
     {
-      "epoch": 0.15384615384615385,
-      "grad_norm": 0.5453268015424697,
-      "learning_rate": 1.025390625e-05,
-      "loss": 0.4595,
+      "epoch": 0.15457088179007802,
+      "grad_norm": 0.5053927687579358,
+      "learning_rate": 1.0304219823356232e-05,
+      "loss": 0.4604,
       "step": 1050
     },
     {
-      "epoch": 0.15457875457875458,
-      "grad_norm": 0.5671479089435382,
-      "learning_rate": 1.0302734375e-05,
-      "loss": 0.4682,
+      "epoch": 0.155306933608126,
+      "grad_norm": 0.510564493666499,
+      "learning_rate": 1.0353287536800787e-05,
+      "loss": 0.4585,
       "step": 1055
     },
     {
-      "epoch": 0.15531135531135531,
-      "grad_norm": 0.559268834190801,
-      "learning_rate": 1.03515625e-05,
-      "loss": 0.4856,
+      "epoch": 0.156042985426174,
+      "grad_norm": 0.5342815365135547,
+      "learning_rate": 1.040235525024534e-05,
+      "loss": 0.4658,
       "step": 1060
     },
     {
-      "epoch": 0.15604395604395604,
-      "grad_norm": 0.5276944813849548,
-      "learning_rate": 1.0400390625000001e-05,
-      "loss": 0.4461,
+      "epoch": 0.15677903724422199,
+      "grad_norm": 0.5303796569780999,
+      "learning_rate": 1.0451422963689892e-05,
+      "loss": 0.468,
       "step": 1065
     },
     {
-      "epoch": 0.15677655677655677,
-      "grad_norm": 0.5155281991905493,
-      "learning_rate": 1.0449218750000001e-05,
-      "loss": 0.4772,
+      "epoch": 0.15751508906227,
+      "grad_norm": 0.5435896518221282,
+      "learning_rate": 1.0500490677134447e-05,
+      "loss": 0.4664,
       "step": 1070
     },
     {
-      "epoch": 0.1575091575091575,
-      "grad_norm": 0.5184923436594232,
-      "learning_rate": 1.0498046875000001e-05,
-      "loss": 0.4556,
+      "epoch": 0.15825114088031797,
+      "grad_norm": 0.5127641089890831,
+      "learning_rate": 1.0549558390579e-05,
+      "loss": 0.4613,
       "step": 1075
     },
     {
-      "epoch": 0.15824175824175823,
-      "grad_norm": 0.5029598122850704,
-      "learning_rate": 1.0546875e-05,
-      "loss": 0.4613,
+      "epoch": 0.15898719269836598,
+      "grad_norm": 0.580969232971526,
+      "learning_rate": 1.0598626104023552e-05,
+      "loss": 0.4552,
       "step": 1080
     },
     {
-      "epoch": 0.15897435897435896,
-      "grad_norm": 0.5286665458480232,
-      "learning_rate": 1.0595703125e-05,
-      "loss": 0.4559,
+      "epoch": 0.15972324451641395,
+      "grad_norm": 0.49770589445853974,
+      "learning_rate": 1.0647693817468108e-05,
+      "loss": 0.4667,
       "step": 1085
     },
     {
-      "epoch": 0.1597069597069597,
-      "grad_norm": 0.5047397009957366,
-      "learning_rate": 1.0644531250000002e-05,
-      "loss": 0.4586,
+      "epoch": 0.16045929633446196,
+      "grad_norm": 0.5434030253983654,
+      "learning_rate": 1.0696761530912661e-05,
+      "loss": 0.4681,
       "step": 1090
     },
     {
-      "epoch": 0.16043956043956045,
-      "grad_norm": 0.7228349992269166,
-      "learning_rate": 1.0693359375000001e-05,
-      "loss": 0.4678,
+      "epoch": 0.16119534815250994,
+      "grad_norm": 0.5222809668848786,
+      "learning_rate": 1.0745829244357214e-05,
+      "loss": 0.4753,
       "step": 1095
     },
     {
-      "epoch": 0.16117216117216118,
-      "grad_norm": 0.5229916349907864,
-      "learning_rate": 1.0742187500000001e-05,
-      "loss": 0.4904,
+      "epoch": 0.16193139997055792,
+      "grad_norm": 0.5224958676817909,
+      "learning_rate": 1.0794896957801768e-05,
+      "loss": 0.4828,
       "step": 1100
     },
     {
-      "epoch": 0.1619047619047619,
-      "grad_norm": 0.5036353858856911,
-      "learning_rate": 1.0791015625e-05,
-      "loss": 0.4688,
+      "epoch": 0.16266745178860592,
+      "grad_norm": 0.5087197860162646,
+      "learning_rate": 1.084396467124632e-05,
+      "loss": 0.465,
       "step": 1105
     },
     {
-      "epoch": 0.16263736263736264,
-      "grad_norm": 0.5122903895008505,
-      "learning_rate": 1.083984375e-05,
-      "loss": 0.4757,
+      "epoch": 0.1634035036066539,
+      "grad_norm": 0.5258709460092567,
+      "learning_rate": 1.0893032384690873e-05,
+      "loss": 0.478,
       "step": 1110
     },
     {
-      "epoch": 0.16336996336996337,
-      "grad_norm": 0.497493881419283,
-      "learning_rate": 1.0888671875e-05,
-      "loss": 0.4535,
+      "epoch": 0.1641395554247019,
+      "grad_norm": 0.5013511959010691,
+      "learning_rate": 1.0942100098135428e-05,
+      "loss": 0.462,
       "step": 1115
     },
     {
-      "epoch": 0.1641025641025641,
-      "grad_norm": 0.5287816022219953,
-      "learning_rate": 1.0937500000000002e-05,
-      "loss": 0.4546,
+      "epoch": 0.16487560724274988,
+      "grad_norm": 0.523215723727446,
+      "learning_rate": 1.099116781157998e-05,
+      "loss": 0.4622,
       "step": 1120
     },
     {
-      "epoch": 0.16483516483516483,
-      "grad_norm": 0.5393853431608534,
-      "learning_rate": 1.0986328125000001e-05,
-      "loss": 0.4613,
+      "epoch": 0.1656116590607979,
+      "grad_norm": 0.4993724186877467,
+      "learning_rate": 1.1040235525024535e-05,
+      "loss": 0.4878,
       "step": 1125
     },
     {
-      "epoch": 0.16556776556776556,
-      "grad_norm": 0.5175948331294432,
-      "learning_rate": 1.1035156250000001e-05,
-      "loss": 0.4819,
+      "epoch": 0.16634771087884587,
+      "grad_norm": 0.5149066945719283,
+      "learning_rate": 1.108930323846909e-05,
+      "loss": 0.4719,
       "step": 1130
     },
     {
-      "epoch": 0.1663003663003663,
-      "grad_norm": 0.5202860471821462,
-      "learning_rate": 1.1083984375e-05,
-      "loss": 0.4674,
+      "epoch": 0.16708376269689387,
+      "grad_norm": 0.5324143493051305,
+      "learning_rate": 1.1138370951913642e-05,
+      "loss": 0.4589,
       "step": 1135
     },
     {
-      "epoch": 0.16703296703296702,
-      "grad_norm": 0.7149468935941481,
-      "learning_rate": 1.11328125e-05,
-      "loss": 0.4721,
+      "epoch": 0.16781981451494185,
+      "grad_norm": 0.5104651877601035,
+      "learning_rate": 1.1187438665358195e-05,
+      "loss": 0.4578,
       "step": 1140
     },
     {
-      "epoch": 0.16776556776556775,
-      "grad_norm": 0.5208583785619997,
-      "learning_rate": 1.1181640625e-05,
-      "loss": 0.4848,
+      "epoch": 0.16855586633298983,
+      "grad_norm": 0.5136825933725584,
+      "learning_rate": 1.123650637880275e-05,
+      "loss": 0.4602,
       "step": 1145
     },
     {
-      "epoch": 0.1684981684981685,
-      "grad_norm": 0.5240976003205658,
-      "learning_rate": 1.1230468750000002e-05,
-      "loss": 0.4805,
+      "epoch": 0.16929191815103783,
+      "grad_norm": 0.5932405914639006,
+      "learning_rate": 1.1285574092247302e-05,
+      "loss": 0.4806,
       "step": 1150
     },
     {
-      "epoch": 0.16923076923076924,
-      "grad_norm": 0.6131879110902204,
-      "learning_rate": 1.1279296875000001e-05,
-      "loss": 0.4747,
+      "epoch": 0.1700279699690858,
+      "grad_norm": 0.4987604697030047,
+      "learning_rate": 1.1334641805691855e-05,
+      "loss": 0.4708,
       "step": 1155
     },
     {
-      "epoch": 0.16996336996336997,
-      "grad_norm": 0.5194993444884493,
-      "learning_rate": 1.1328125000000001e-05,
-      "loss": 0.4723,
+      "epoch": 0.17076402178713382,
+      "grad_norm": 0.5120281390736848,
+      "learning_rate": 1.138370951913641e-05,
+      "loss": 0.4747,
       "step": 1160
     },
     {
-      "epoch": 0.1706959706959707,
-      "grad_norm": 0.5366643091240741,
-      "learning_rate": 1.1376953125e-05,
-      "loss": 0.4599,
+      "epoch": 0.1715000736051818,
+      "grad_norm": 0.5160816666730633,
+      "learning_rate": 1.1432777232580963e-05,
+      "loss": 0.4636,
       "step": 1165
     },
     {
-      "epoch": 0.17142857142857143,
-      "grad_norm": 0.5307151335800977,
-      "learning_rate": 1.142578125e-05,
-      "loss": 0.4764,
+      "epoch": 0.1722361254232298,
+      "grad_norm": 0.5050377908699278,
+      "learning_rate": 1.1481844946025516e-05,
+      "loss": 0.4571,
       "step": 1170
     },
     {
-      "epoch": 0.17216117216117216,
-      "grad_norm": 0.5404046903160628,
-      "learning_rate": 1.1474609375000002e-05,
-      "loss": 0.4746,
+      "epoch": 0.17297217724127778,
+      "grad_norm": 0.5056895497995904,
+      "learning_rate": 1.153091265947007e-05,
+      "loss": 0.4632,
       "step": 1175
     },
     {
-      "epoch": 0.1728937728937729,
-      "grad_norm": 0.5051709515526519,
-      "learning_rate": 1.1523437500000001e-05,
-      "loss": 0.4521,
+      "epoch": 0.17370822905932579,
+      "grad_norm": 0.5335108204710897,
+      "learning_rate": 1.1579980372914623e-05,
+      "loss": 0.4746,
       "step": 1180
     },
     {
-      "epoch": 0.17362637362637362,
-      "grad_norm": 0.5868257177183442,
-      "learning_rate": 1.1572265625000001e-05,
-      "loss": 0.4842,
+      "epoch": 0.17444428087737376,
+      "grad_norm": 0.516518365889526,
+      "learning_rate": 1.1629048086359176e-05,
+      "loss": 0.457,
       "step": 1185
     },
     {
-      "epoch": 0.17435897435897435,
-      "grad_norm": 0.6728971850067752,
-      "learning_rate": 1.162109375e-05,
-      "loss": 0.4532,
+      "epoch": 0.17518033269542177,
+      "grad_norm": 0.5052625349340393,
+      "learning_rate": 1.167811579980373e-05,
+      "loss": 0.4698,
       "step": 1190
     },
     {
-      "epoch": 0.17509157509157508,
-      "grad_norm": 0.5583627103602447,
-      "learning_rate": 1.1669921875e-05,
-      "loss": 0.4647,
+      "epoch": 0.17591638451346975,
+      "grad_norm": 0.5451467203953263,
+      "learning_rate": 1.1727183513248283e-05,
+      "loss": 0.4629,
       "step": 1195
     },
     {
-      "epoch": 0.17582417582417584,
-      "grad_norm": 0.5167154137386244,
-      "learning_rate": 1.171875e-05,
-      "loss": 0.4612,
+      "epoch": 0.17665243633151773,
+      "grad_norm": 0.5335149299750542,
+      "learning_rate": 1.1776251226692836e-05,
+      "loss": 0.4553,
       "step": 1200
     },
     {
-      "epoch": 0.17655677655677657,
-      "grad_norm": 0.4849028929167034,
-      "learning_rate": 1.1767578125000002e-05,
-      "loss": 0.467,
+      "epoch": 0.17738848814956573,
+      "grad_norm": 0.5652693128753509,
+      "learning_rate": 1.1825318940137392e-05,
+      "loss": 0.4659,
       "step": 1205
     },
     {
-      "epoch": 0.1772893772893773,
-      "grad_norm": 0.49902210105050515,
-      "learning_rate": 1.1816406250000001e-05,
-      "loss": 0.4617,
+      "epoch": 0.1781245399676137,
+      "grad_norm": 0.5068344293397069,
+      "learning_rate": 1.1874386653581945e-05,
+      "loss": 0.4499,
       "step": 1210
     },
     {
-      "epoch": 0.17802197802197803,
-      "grad_norm": 0.4823129924518987,
-      "learning_rate": 1.1865234375000001e-05,
-      "loss": 0.4626,
+      "epoch": 0.17886059178566172,
+      "grad_norm": 0.48440022275092465,
+      "learning_rate": 1.1923454367026497e-05,
+      "loss": 0.4591,
       "step": 1215
     },
     {
-      "epoch": 0.17875457875457876,
-      "grad_norm": 0.5300239765467367,
-      "learning_rate": 1.19140625e-05,
-      "loss": 0.4659,
+      "epoch": 0.1795966436037097,
+      "grad_norm": 0.8941425323645812,
+      "learning_rate": 1.1972522080471052e-05,
+      "loss": 0.4764,
       "step": 1220
     },
     {
-      "epoch": 0.1794871794871795,
-      "grad_norm": 0.6088645484486718,
-      "learning_rate": 1.1962890625e-05,
-      "loss": 0.4694,
+      "epoch": 0.1803326954217577,
+      "grad_norm": 0.5367647624912888,
+      "learning_rate": 1.2021589793915604e-05,
+      "loss": 0.4443,
       "step": 1225
     },
     {
-      "epoch": 0.18021978021978022,
-      "grad_norm": 0.5455240800251795,
-      "learning_rate": 1.201171875e-05,
-      "loss": 0.4845,
+      "epoch": 0.18106874723980568,
+      "grad_norm": 0.5514766921493407,
+      "learning_rate": 1.2070657507360157e-05,
+      "loss": 0.4774,
       "step": 1230
     },
     {
-      "epoch": 0.18095238095238095,
-      "grad_norm": 0.5424304480123273,
-      "learning_rate": 1.2060546875000002e-05,
-      "loss": 0.4668,
+      "epoch": 0.18180479905785368,
+      "grad_norm": 0.5388082425030042,
+      "learning_rate": 1.2119725220804712e-05,
+      "loss": 0.4594,
       "step": 1235
     },
     {
-      "epoch": 0.18168498168498168,
-      "grad_norm": 0.5203368741702424,
-      "learning_rate": 1.2109375000000001e-05,
-      "loss": 0.4862,
+      "epoch": 0.18254085087590166,
+      "grad_norm": 0.5068247119187569,
+      "learning_rate": 1.2168792934249264e-05,
+      "loss": 0.4802,
       "step": 1240
     },
     {
-      "epoch": 0.1824175824175824,
-      "grad_norm": 0.6584898777407595,
-      "learning_rate": 1.2158203125000001e-05,
-      "loss": 0.4531,
+      "epoch": 0.18327690269394967,
+      "grad_norm": 0.5227838084909681,
+      "learning_rate": 1.2217860647693817e-05,
+      "loss": 0.4691,
       "step": 1245
     },
     {
-      "epoch": 0.18315018315018314,
-      "grad_norm": 0.5327149440973011,
-      "learning_rate": 1.220703125e-05,
-      "loss": 0.4563,
+      "epoch": 0.18401295451199765,
+      "grad_norm": 0.4938061071127395,
+      "learning_rate": 1.2266928361138373e-05,
+      "loss": 0.4746,
       "step": 1250
     },
     {
-      "epoch": 0.1838827838827839,
-      "grad_norm": 0.5277598765377318,
-      "learning_rate": 1.2255859375e-05,
-      "loss": 0.4699,
+      "epoch": 0.18474900633004562,
+      "grad_norm": 0.5238294647370162,
+      "learning_rate": 1.2315996074582926e-05,
+      "loss": 0.4668,
       "step": 1255
     },
     {
-      "epoch": 0.18461538461538463,
-      "grad_norm": 0.5167807191476441,
-      "learning_rate": 1.2304687500000002e-05,
-      "loss": 0.4652,
+      "epoch": 0.18548505814809363,
+      "grad_norm": 0.46614497230235846,
+      "learning_rate": 1.2365063788027479e-05,
+      "loss": 0.4612,
       "step": 1260
     },
     {
-      "epoch": 0.18534798534798536,
-      "grad_norm": 0.4808989318720859,
-      "learning_rate": 1.2353515625000001e-05,
-      "loss": 0.4825,
+      "epoch": 0.1862211099661416,
+      "grad_norm": 0.4985552878100185,
+      "learning_rate": 1.2414131501472033e-05,
+      "loss": 0.4817,
       "step": 1265
     },
     {
-      "epoch": 0.18608058608058609,
-      "grad_norm": 0.5097882629770788,
-      "learning_rate": 1.2402343750000001e-05,
-      "loss": 0.4759,
+      "epoch": 0.1869571617841896,
+      "grad_norm": 0.5156607573559511,
+      "learning_rate": 1.2463199214916586e-05,
+      "loss": 0.4766,
       "step": 1270
     },
     {
-      "epoch": 0.18681318681318682,
-      "grad_norm": 0.5456506028065036,
-      "learning_rate": 1.2451171875000001e-05,
-      "loss": 0.4788,
+      "epoch": 0.1876932136022376,
+      "grad_norm": 0.4894863185953375,
+      "learning_rate": 1.2512266928361138e-05,
+      "loss": 0.4651,
       "step": 1275
     },
     {
-      "epoch": 0.18754578754578755,
-      "grad_norm": 0.5107241981918527,
-      "learning_rate": 1.25e-05,
-      "loss": 0.483,
+      "epoch": 0.1884292654202856,
+      "grad_norm": 0.5231954839271257,
+      "learning_rate": 1.2561334641805694e-05,
+      "loss": 0.4826,
       "step": 1280
     },
     {
-      "epoch": 0.18827838827838828,
-      "grad_norm": 0.49376810757312745,
-      "learning_rate": 1.2548828125e-05,
-      "loss": 0.4593,
+      "epoch": 0.18916531723833357,
+      "grad_norm": 0.49809529774347094,
+      "learning_rate": 1.2610402355250247e-05,
+      "loss": 0.464,
       "step": 1285
     },
     {
-      "epoch": 0.189010989010989,
-      "grad_norm": 0.5133797804113845,
-      "learning_rate": 1.2597656250000002e-05,
-      "loss": 0.4631,
+      "epoch": 0.18990136905638158,
+      "grad_norm": 0.5234554810941109,
+      "learning_rate": 1.26594700686948e-05,
+      "loss": 0.4733,
       "step": 1290
     },
     {
-      "epoch": 0.18974358974358974,
-      "grad_norm": 0.5008331295590094,
-      "learning_rate": 1.2646484375000001e-05,
-      "loss": 0.4802,
+      "epoch": 0.19063742087442956,
+      "grad_norm": 0.5051586556741555,
+      "learning_rate": 1.2708537782139354e-05,
+      "loss": 0.4738,
       "step": 1295
     },
     {
-      "epoch": 0.19047619047619047,
-      "grad_norm": 0.5675292443089186,
-      "learning_rate": 1.2695312500000001e-05,
-      "loss": 0.4888,
+      "epoch": 0.19137347269247756,
+      "grad_norm": 0.49517034893999295,
+      "learning_rate": 1.2757605495583907e-05,
+      "loss": 0.47,
       "step": 1300
     },
     {
-      "epoch": 0.1912087912087912,
-      "grad_norm": 0.5005818050999626,
-      "learning_rate": 1.2744140625e-05,
-      "loss": 0.4689,
+      "epoch": 0.19210952451052554,
+      "grad_norm": 0.5915470507381195,
+      "learning_rate": 1.280667320902846e-05,
+      "loss": 0.4781,
       "step": 1305
     },
     {
-      "epoch": 0.19194139194139195,
-      "grad_norm": 0.5283882256001375,
-      "learning_rate": 1.279296875e-05,
-      "loss": 0.4465,
+      "epoch": 0.19284557632857352,
+      "grad_norm": 0.508338735620736,
+      "learning_rate": 1.2855740922473014e-05,
+      "loss": 0.4707,
       "step": 1310
     },
     {
-      "epoch": 0.19267399267399268,
-      "grad_norm": 0.5193448027983114,
-      "learning_rate": 1.2841796875e-05,
-      "loss": 0.4678,
+      "epoch": 0.19358162814662153,
+      "grad_norm": 0.4990887547861896,
+      "learning_rate": 1.2904808635917567e-05,
+      "loss": 0.4835,
       "step": 1315
     },
     {
-      "epoch": 0.1934065934065934,
-      "grad_norm": 0.6159454505341463,
-      "learning_rate": 1.2890625000000002e-05,
-      "loss": 0.4606,
+      "epoch": 0.1943176799646695,
+      "grad_norm": 0.48691875103080917,
+      "learning_rate": 1.295387634936212e-05,
+      "loss": 0.4674,
       "step": 1320
     },
     {
-      "epoch": 0.19413919413919414,
-      "grad_norm": 0.5086536030516351,
-      "learning_rate": 1.2939453125000001e-05,
-      "loss": 0.4589,
+      "epoch": 0.1950537317827175,
+      "grad_norm": 0.4841195176880159,
+      "learning_rate": 1.3002944062806676e-05,
+      "loss": 0.4729,
       "step": 1325
     },
     {
-      "epoch": 0.19487179487179487,
-      "grad_norm": 0.48860047105264637,
-      "learning_rate": 1.2988281250000001e-05,
-      "loss": 0.4541,
+      "epoch": 0.1957897836007655,
+      "grad_norm": 0.4992178518108465,
+      "learning_rate": 1.3052011776251228e-05,
+      "loss": 0.4688,
       "step": 1330
     },
     {
-      "epoch": 0.1956043956043956,
-      "grad_norm": 0.46316718079861624,
-      "learning_rate": 1.3037109375e-05,
-      "loss": 0.4544,
+      "epoch": 0.1965258354188135,
+      "grad_norm": 0.47943066480788776,
+      "learning_rate": 1.3101079489695781e-05,
+      "loss": 0.4699,
       "step": 1335
     },
     {
-      "epoch": 0.19633699633699633,
-      "grad_norm": 0.49587456775643657,
-      "learning_rate": 1.30859375e-05,
-      "loss": 0.4623,
+      "epoch": 0.19726188723686147,
+      "grad_norm": 0.5012063847835244,
+      "learning_rate": 1.3150147203140335e-05,
+      "loss": 0.4661,
       "step": 1340
     },
     {
-      "epoch": 0.19706959706959706,
-      "grad_norm": 0.501454631805098,
-      "learning_rate": 1.3134765625000002e-05,
-      "loss": 0.4507,
+      "epoch": 0.19799793905490948,
+      "grad_norm": 0.535826158477013,
+      "learning_rate": 1.3199214916584888e-05,
+      "loss": 0.4706,
       "step": 1345
     },
     {
-      "epoch": 0.1978021978021978,
-      "grad_norm": 0.5000449976017295,
-      "learning_rate": 1.3183593750000002e-05,
-      "loss": 0.4512,
+      "epoch": 0.19873399087295746,
+      "grad_norm": 0.5071529638069743,
+      "learning_rate": 1.3248282630029441e-05,
+      "loss": 0.4625,
       "step": 1350
     },
     {
-      "epoch": 0.19853479853479852,
-      "grad_norm": 0.4839379008548153,
-      "learning_rate": 1.3232421875000001e-05,
-      "loss": 0.4862,
+      "epoch": 0.19947004269100543,
+      "grad_norm": 0.5179383619569404,
+      "learning_rate": 1.3297350343473995e-05,
+      "loss": 0.4759,
       "step": 1355
     },
     {
-      "epoch": 0.19926739926739928,
-      "grad_norm": 0.5208252130809964,
-      "learning_rate": 1.3281250000000001e-05,
-      "loss": 0.4676,
+      "epoch": 0.20020609450905344,
+      "grad_norm": 0.48405439133748607,
+      "learning_rate": 1.3346418056918548e-05,
+      "loss": 0.4761,
       "step": 1360
     },
     {
-      "epoch": 0.2,
-      "grad_norm": 0.5745959281687488,
-      "learning_rate": 1.3330078125e-05,
-      "loss": 0.4762,
+      "epoch": 0.20094214632710142,
+      "grad_norm": 0.5154094066647193,
+      "learning_rate": 1.33954857703631e-05,
+      "loss": 0.478,
       "step": 1365
     },
     {
-      "epoch": 0.20073260073260074,
-      "grad_norm": 0.6043469530792722,
-      "learning_rate": 1.337890625e-05,
-      "loss": 0.4704,
+      "epoch": 0.20167819814514942,
+      "grad_norm": 0.5039502624189186,
+      "learning_rate": 1.3444553483807657e-05,
+      "loss": 0.4465,
       "step": 1370
     },
     {
-      "epoch": 0.20146520146520147,
-      "grad_norm": 0.5334372627977081,
-      "learning_rate": 1.3427734375000002e-05,
-      "loss": 0.4716,
+      "epoch": 0.2024142499631974,
+      "grad_norm": 0.5107787451227336,
+      "learning_rate": 1.349362119725221e-05,
+      "loss": 0.4544,
       "step": 1375
     },
     {
-      "epoch": 0.2021978021978022,
-      "grad_norm": 0.5390143553108346,
-      "learning_rate": 1.3476562500000001e-05,
-      "loss": 0.4735,
+      "epoch": 0.2031503017812454,
+      "grad_norm": 0.5182898230453957,
+      "learning_rate": 1.3542688910696762e-05,
+      "loss": 0.4776,
       "step": 1380
     },
     {
-      "epoch": 0.20293040293040293,
-      "grad_norm": 0.4736959203098974,
-      "learning_rate": 1.3525390625000001e-05,
-      "loss": 0.4774,
+      "epoch": 0.20388635359929339,
+      "grad_norm": 0.5025863501160254,
+      "learning_rate": 1.3591756624141317e-05,
+      "loss": 0.4681,
       "step": 1385
     },
     {
-      "epoch": 0.20366300366300366,
-      "grad_norm": 0.49280583683920803,
-      "learning_rate": 1.3574218750000001e-05,
-      "loss": 0.4652,
+      "epoch": 0.2046224054173414,
+      "grad_norm": 0.4894363132630262,
+      "learning_rate": 1.364082433758587e-05,
+      "loss": 0.4675,
       "step": 1390
     },
     {
-      "epoch": 0.2043956043956044,
-      "grad_norm": 0.525459455539279,
-      "learning_rate": 1.3623046875e-05,
-      "loss": 0.4822,
+      "epoch": 0.20535845723538937,
+      "grad_norm": 0.5433477313290025,
+      "learning_rate": 1.3689892051030422e-05,
+      "loss": 0.4811,
       "step": 1395
     },
     {
-      "epoch": 0.20512820512820512,
-      "grad_norm": 0.48689775993494483,
-      "learning_rate": 1.3671875e-05,
-      "loss": 0.4719,
+      "epoch": 0.20609450905343737,
+      "grad_norm": 0.4967075561653421,
+      "learning_rate": 1.3738959764474977e-05,
+      "loss": 0.4458,
       "step": 1400
     },
     {
-      "epoch": 0.20586080586080585,
-      "grad_norm": 0.5212181067359885,
-      "learning_rate": 1.3720703125000002e-05,
-      "loss": 0.4709,
+      "epoch": 0.20683056087148535,
+      "grad_norm": 0.5244064638537597,
+      "learning_rate": 1.378802747791953e-05,
+      "loss": 0.469,
       "step": 1405
     },
     {
-      "epoch": 0.20659340659340658,
-      "grad_norm": 0.508003690634916,
-      "learning_rate": 1.3769531250000001e-05,
-      "loss": 0.4737,
+      "epoch": 0.20756661268953333,
+      "grad_norm": 0.5204337598099966,
+      "learning_rate": 1.3837095191364082e-05,
+      "loss": 0.4758,
       "step": 1410
     },
     {
-      "epoch": 0.20732600732600734,
-      "grad_norm": 0.5073601130842522,
-      "learning_rate": 1.3818359375000001e-05,
-      "loss": 0.4799,
+      "epoch": 0.20830266450758134,
+      "grad_norm": 0.5080676476650419,
+      "learning_rate": 1.3886162904808638e-05,
+      "loss": 0.473,
       "step": 1415
     },
     {
-      "epoch": 0.20805860805860807,
-      "grad_norm": 0.5355556977525563,
-      "learning_rate": 1.38671875e-05,
-      "loss": 0.474,
+      "epoch": 0.20903871632562931,
+      "grad_norm": 0.47828452516531716,
+      "learning_rate": 1.393523061825319e-05,
+      "loss": 0.4816,
       "step": 1420
     },
     {
-      "epoch": 0.2087912087912088,
-      "grad_norm": 0.4990582643295359,
-      "learning_rate": 1.3916015625e-05,
-      "loss": 0.4854,
+      "epoch": 0.20977476814367732,
+      "grad_norm": 0.48532948614032556,
+      "learning_rate": 1.3984298331697743e-05,
+      "loss": 0.4714,
       "step": 1425
     },
     {
-      "epoch": 0.20952380952380953,
-      "grad_norm": 0.5093226857332248,
-      "learning_rate": 1.3964843750000002e-05,
-      "loss": 0.4555,
+      "epoch": 0.2105108199617253,
+      "grad_norm": 0.5086062877140418,
+      "learning_rate": 1.4033366045142298e-05,
+      "loss": 0.4667,
       "step": 1430
     },
     {
-      "epoch": 0.21025641025641026,
-      "grad_norm": 0.502584313631928,
-      "learning_rate": 1.4013671875000002e-05,
-      "loss": 0.4732,
+      "epoch": 0.2112468717797733,
+      "grad_norm": 0.5050717781580496,
+      "learning_rate": 1.408243375858685e-05,
+      "loss": 0.478,
       "step": 1435
     },
     {
-      "epoch": 0.210989010989011,
-      "grad_norm": 0.526309212370889,
-      "learning_rate": 1.4062500000000001e-05,
-      "loss": 0.4728,
+      "epoch": 0.21198292359782128,
+      "grad_norm": 0.48821477274253927,
+      "learning_rate": 1.4131501472031403e-05,
+      "loss": 0.4588,
       "step": 1440
     },
     {
-      "epoch": 0.21172161172161172,
-      "grad_norm": 0.5176074132776409,
-      "learning_rate": 1.4111328125000001e-05,
-      "loss": 0.4664,
+      "epoch": 0.2127189754158693,
+      "grad_norm": 0.48266677376219413,
+      "learning_rate": 1.418056918547596e-05,
+      "loss": 0.4557,
       "step": 1445
     },
     {
-      "epoch": 0.21245421245421245,
-      "grad_norm": 0.4879365025716744,
-      "learning_rate": 1.416015625e-05,
-      "loss": 0.467,
+      "epoch": 0.21345502723391727,
+      "grad_norm": 0.5512455514452225,
+      "learning_rate": 1.4229636898920512e-05,
+      "loss": 0.4672,
       "step": 1450
     },
     {
-      "epoch": 0.21318681318681318,
-      "grad_norm": 1.9306549484896947,
-      "learning_rate": 1.4208984375e-05,
-      "loss": 0.4649,
+      "epoch": 0.21419107905196527,
+      "grad_norm": 0.5083481320196631,
+      "learning_rate": 1.4278704612365065e-05,
+      "loss": 0.4661,
       "step": 1455
     },
     {
-      "epoch": 0.2139194139194139,
-      "grad_norm": 0.4955066630748684,
-      "learning_rate": 1.4257812500000002e-05,
-      "loss": 0.4601,
+      "epoch": 0.21492713087001325,
+      "grad_norm": 0.5595410982557921,
+      "learning_rate": 1.432777232580962e-05,
+      "loss": 0.4708,
       "step": 1460
     },
     {
-      "epoch": 0.21465201465201464,
-      "grad_norm": 0.5025180598567222,
-      "learning_rate": 1.4306640625000002e-05,
-      "loss": 0.4608,
+      "epoch": 0.21566318268806123,
+      "grad_norm": 0.5224743589963625,
+      "learning_rate": 1.4376840039254172e-05,
+      "loss": 0.4759,
       "step": 1465
     },
     {
-      "epoch": 0.2153846153846154,
-      "grad_norm": 0.49793528536390486,
-      "learning_rate": 1.4355468750000001e-05,
-      "loss": 0.4799,
+      "epoch": 0.21639923450610923,
+      "grad_norm": 0.5195549182987274,
+      "learning_rate": 1.4425907752698725e-05,
+      "loss": 0.4649,
       "step": 1470
     },
     {
-      "epoch": 0.21611721611721613,
-      "grad_norm": 1.8297240222248374,
-      "learning_rate": 1.4404296875000001e-05,
-      "loss": 0.4689,
+      "epoch": 0.2171352863241572,
+      "grad_norm": 0.5320486190932664,
+      "learning_rate": 1.4474975466143279e-05,
+      "loss": 0.4527,
       "step": 1475
     },
     {
-      "epoch": 0.21684981684981686,
-      "grad_norm": 0.4705591944470831,
-      "learning_rate": 1.4453125e-05,
-      "loss": 0.4828,
+      "epoch": 0.21787133814220522,
+      "grad_norm": 0.5152466166857504,
+      "learning_rate": 1.4524043179587832e-05,
+      "loss": 0.4706,
       "step": 1480
     },
     {
-      "epoch": 0.2175824175824176,
-      "grad_norm": 0.4835356918254959,
-      "learning_rate": 1.4501953125e-05,
-      "loss": 0.4661,
+      "epoch": 0.2186073899602532,
+      "grad_norm": 0.4971011294748948,
+      "learning_rate": 1.4573110893032384e-05,
+      "loss": 0.4701,
       "step": 1485
     },
     {
-      "epoch": 0.21831501831501832,
-      "grad_norm": 0.510433847664988,
-      "learning_rate": 1.4550781250000002e-05,
-      "loss": 0.4748,
+      "epoch": 0.2193434417783012,
+      "grad_norm": 0.468694375011289,
+      "learning_rate": 1.462217860647694e-05,
+      "loss": 0.4653,
       "step": 1490
     },
     {
-      "epoch": 0.21904761904761905,
-      "grad_norm": 0.5216647874966567,
-      "learning_rate": 1.4599609375000001e-05,
-      "loss": 0.475,
+      "epoch": 0.22007949359634918,
+      "grad_norm": 0.47946916582322097,
+      "learning_rate": 1.4671246319921493e-05,
+      "loss": 0.4674,
       "step": 1495
     },
     {
-      "epoch": 0.21978021978021978,
-      "grad_norm": 0.4944348257107384,
-      "learning_rate": 1.4648437500000001e-05,
-      "loss": 0.4768,
+      "epoch": 0.22081554541439719,
+      "grad_norm": 0.5032322510285128,
+      "learning_rate": 1.4720314033366046e-05,
+      "loss": 0.4583,
       "step": 1500
     },
     {
-      "epoch": 0.2205128205128205,
-      "grad_norm": 0.4945475138346755,
-      "learning_rate": 1.4697265625000001e-05,
-      "loss": 0.4655,
+      "epoch": 0.22155159723244516,
+      "grad_norm": 0.5256011707393451,
+      "learning_rate": 1.47693817468106e-05,
+      "loss": 0.4835,
       "step": 1505
     },
     {
-      "epoch": 0.22124542124542124,
-      "grad_norm": 0.5006384611474771,
-      "learning_rate": 1.474609375e-05,
-      "loss": 0.4794,
+      "epoch": 0.22228764905049314,
+      "grad_norm": 0.5339877468693773,
+      "learning_rate": 1.4818449460255153e-05,
+      "loss": 0.4807,
       "step": 1510
     },
     {
-      "epoch": 0.22197802197802197,
-      "grad_norm": 0.5466308398968402,
-      "learning_rate": 1.4794921875000002e-05,
-      "loss": 0.4753,
+      "epoch": 0.22302370086854115,
+      "grad_norm": 0.5160851206234418,
+      "learning_rate": 1.4867517173699706e-05,
+      "loss": 0.4785,
       "step": 1515
     },
     {
-      "epoch": 0.22271062271062272,
-      "grad_norm": 0.5348130721491957,
-      "learning_rate": 1.4843750000000002e-05,
-      "loss": 0.4655,
+      "epoch": 0.22375975268658913,
+      "grad_norm": 0.47860088752170427,
+      "learning_rate": 1.491658488714426e-05,
+      "loss": 0.4794,
       "step": 1520
     },
     {
-      "epoch": 0.22344322344322345,
-      "grad_norm": 0.48173601863057236,
-      "learning_rate": 1.4892578125000001e-05,
-      "loss": 0.471,
+      "epoch": 0.22449580450463713,
+      "grad_norm": 0.5243776797101112,
+      "learning_rate": 1.4965652600588813e-05,
+      "loss": 0.4767,
       "step": 1525
     },
     {
-      "epoch": 0.22417582417582418,
-      "grad_norm": 0.5217530092063394,
-      "learning_rate": 1.4941406250000001e-05,
-      "loss": 0.4598,
+      "epoch": 0.2252318563226851,
+      "grad_norm": 0.5179508095461343,
+      "learning_rate": 1.5014720314033366e-05,
+      "loss": 0.4847,
       "step": 1530
     },
     {
-      "epoch": 0.22490842490842491,
-      "grad_norm": 0.4802076358174074,
-      "learning_rate": 1.4990234375e-05,
-      "loss": 0.4695,
+      "epoch": 0.22596790814073311,
+      "grad_norm": 0.5001243630855052,
+      "learning_rate": 1.5063788027477922e-05,
+      "loss": 0.4675,
       "step": 1535
     },
     {
-      "epoch": 0.22564102564102564,
-      "grad_norm": 0.49840080694366656,
-      "learning_rate": 1.50390625e-05,
-      "loss": 0.4668,
+      "epoch": 0.2267039599587811,
+      "grad_norm": 0.48838181870103403,
+      "learning_rate": 1.5112855740922475e-05,
+      "loss": 0.4765,
       "step": 1540
     },
     {
-      "epoch": 0.22637362637362637,
-      "grad_norm": 0.48517238694140485,
-      "learning_rate": 1.5087890625000002e-05,
-      "loss": 0.4545,
+      "epoch": 0.2274400117768291,
+      "grad_norm": 0.5093776483261092,
+      "learning_rate": 1.5161923454367027e-05,
+      "loss": 0.4804,
       "step": 1545
     },
     {
-      "epoch": 0.2271062271062271,
-      "grad_norm": 0.5112587584402484,
-      "learning_rate": 1.5136718750000002e-05,
-      "loss": 0.4818,
+      "epoch": 0.22817606359487708,
+      "grad_norm": 0.49757765788276226,
+      "learning_rate": 1.5210991167811582e-05,
+      "loss": 0.4602,
       "step": 1550
     },
     {
-      "epoch": 0.22783882783882783,
-      "grad_norm": 0.4622437817284304,
-      "learning_rate": 1.5185546875000001e-05,
-      "loss": 0.4778,
+      "epoch": 0.22891211541292508,
+      "grad_norm": 0.6042817011562104,
+      "learning_rate": 1.5260058881256136e-05,
+      "loss": 0.4679,
       "step": 1555
     },
     {
-      "epoch": 0.22857142857142856,
-      "grad_norm": 0.4998516140224748,
-      "learning_rate": 1.5234375000000001e-05,
-      "loss": 0.4523,
+      "epoch": 0.22964816723097306,
+      "grad_norm": 0.5333395009103001,
+      "learning_rate": 1.5309126594700687e-05,
+      "loss": 0.4774,
       "step": 1560
     },
     {
-      "epoch": 0.2293040293040293,
-      "grad_norm": 0.5313317202642523,
-      "learning_rate": 1.5283203125e-05,
-      "loss": 0.4691,
+      "epoch": 0.23038421904902104,
+      "grad_norm": 0.4880655438801243,
+      "learning_rate": 1.535819430814524e-05,
+      "loss": 0.4656,
       "step": 1565
     },
     {
-      "epoch": 0.23003663003663002,
-      "grad_norm": 0.5042407860552259,
-      "learning_rate": 1.5332031250000002e-05,
-      "loss": 0.4845,
+      "epoch": 0.23112027086706904,
+      "grad_norm": 0.49098355944447114,
+      "learning_rate": 1.5407262021589796e-05,
+      "loss": 0.4675,
       "step": 1570
     },
     {
-      "epoch": 0.23076923076923078,
-      "grad_norm": 0.49760097793140207,
-      "learning_rate": 1.5380859375e-05,
-      "loss": 0.4682,
+      "epoch": 0.23185632268511702,
+      "grad_norm": 0.5067158803640502,
+      "learning_rate": 1.5456329735034347e-05,
+      "loss": 0.4609,
       "step": 1575
     },
     {
-      "epoch": 0.2315018315018315,
-      "grad_norm": 0.4826040802476265,
-      "learning_rate": 1.54296875e-05,
-      "loss": 0.4577,
+      "epoch": 0.23259237450316503,
+      "grad_norm": 0.512616066005067,
+      "learning_rate": 1.55053974484789e-05,
+      "loss": 0.4768,
       "step": 1580
     },
     {
-      "epoch": 0.23223443223443224,
-      "grad_norm": 0.4785399152650899,
-      "learning_rate": 1.5478515625000003e-05,
-      "loss": 0.4667,
+      "epoch": 0.233328426321213,
+      "grad_norm": 0.7154629590386294,
+      "learning_rate": 1.5554465161923456e-05,
+      "loss": 0.4737,
       "step": 1585
     },
     {
-      "epoch": 0.23296703296703297,
-      "grad_norm": 0.5071525975487288,
-      "learning_rate": 1.552734375e-05,
-      "loss": 0.448,
+      "epoch": 0.234064478139261,
+      "grad_norm": 0.47648906066367264,
+      "learning_rate": 1.5603532875368007e-05,
+      "loss": 0.4498,
       "step": 1590
     },
     {
-      "epoch": 0.2336996336996337,
-      "grad_norm": 0.4765305380475655,
-      "learning_rate": 1.5576171875000002e-05,
-      "loss": 0.4486,
+      "epoch": 0.234800529957309,
+      "grad_norm": 0.4982638347719357,
+      "learning_rate": 1.5652600588812565e-05,
+      "loss": 0.4487,
       "step": 1595
     },
     {
-      "epoch": 0.23443223443223443,
-      "grad_norm": 0.506737028891706,
-      "learning_rate": 1.5625e-05,
-      "loss": 0.4594,
+      "epoch": 0.235536581775357,
+      "grad_norm": 0.5214363004304297,
+      "learning_rate": 1.5701668302257116e-05,
+      "loss": 0.4567,
       "step": 1600
     },
     {
-      "epoch": 0.23516483516483516,
-      "grad_norm": 0.5033365268409765,
-      "learning_rate": 1.5673828125000002e-05,
-      "loss": 0.4857,
+      "epoch": 0.23627263359340497,
+      "grad_norm": 0.5281970704860975,
+      "learning_rate": 1.575073601570167e-05,
+      "loss": 0.4871,
       "step": 1605
     },
     {
-      "epoch": 0.2358974358974359,
-      "grad_norm": 0.5533174142914994,
-      "learning_rate": 1.572265625e-05,
-      "loss": 0.4713,
+      "epoch": 0.23700868541145298,
+      "grad_norm": 0.47438263533599084,
+      "learning_rate": 1.5799803729146224e-05,
+      "loss": 0.4744,
       "step": 1610
     },
     {
-      "epoch": 0.23663003663003662,
-      "grad_norm": 0.49950173686881316,
-      "learning_rate": 1.5771484375e-05,
-      "loss": 0.4835,
+      "epoch": 0.23774473722950096,
+      "grad_norm": 0.49674603424275676,
+      "learning_rate": 1.5848871442590775e-05,
+      "loss": 0.4779,
       "step": 1615
     },
     {
-      "epoch": 0.23736263736263735,
-      "grad_norm": 0.477522695620854,
-      "learning_rate": 1.5820312500000003e-05,
-      "loss": 0.5004,
+      "epoch": 0.23848078904754894,
+      "grad_norm": 0.5051623027599355,
+      "learning_rate": 1.589793915603533e-05,
+      "loss": 0.4652,
       "step": 1620
     },
     {
-      "epoch": 0.23809523809523808,
-      "grad_norm": 0.4596092750042302,
-      "learning_rate": 1.5869140625e-05,
-      "loss": 0.4735,
+      "epoch": 0.23921684086559694,
+      "grad_norm": 0.5229696639451223,
+      "learning_rate": 1.5947006869479884e-05,
+      "loss": 0.4788,
       "step": 1625
     },
     {
-      "epoch": 0.23882783882783884,
-      "grad_norm": 0.4864862846797984,
-      "learning_rate": 1.5917968750000002e-05,
-      "loss": 0.48,
+      "epoch": 0.23995289268364492,
+      "grad_norm": 0.5036740790381153,
+      "learning_rate": 1.5996074582924435e-05,
+      "loss": 0.4763,
       "step": 1630
     },
     {
-      "epoch": 0.23956043956043957,
-      "grad_norm": 0.47289872856899984,
-      "learning_rate": 1.5966796875e-05,
-      "loss": 0.4747,
+      "epoch": 0.24068894450169293,
+      "grad_norm": 0.5035354462935039,
+      "learning_rate": 1.604514229636899e-05,
+      "loss": 0.4678,
       "step": 1635
     },
     {
-      "epoch": 0.2402930402930403,
-      "grad_norm": 0.46942665161369906,
-      "learning_rate": 1.6015625e-05,
-      "loss": 0.4725,
+      "epoch": 0.2414249963197409,
+      "grad_norm": 0.4900212380645429,
+      "learning_rate": 1.6094210009813544e-05,
+      "loss": 0.4696,
       "step": 1640
     },
     {
-      "epoch": 0.24102564102564103,
-      "grad_norm": 0.4781748865843972,
-      "learning_rate": 1.6064453125000003e-05,
-      "loss": 0.4726,
+      "epoch": 0.2421610481377889,
+      "grad_norm": 0.5150211975147684,
+      "learning_rate": 1.61432777232581e-05,
+      "loss": 0.4548,
       "step": 1645
     },
     {
-      "epoch": 0.24175824175824176,
-      "grad_norm": 0.4957955173984609,
-      "learning_rate": 1.611328125e-05,
-      "loss": 0.4618,
+      "epoch": 0.2428970999558369,
+      "grad_norm": 0.4981942201415043,
+      "learning_rate": 1.619234543670265e-05,
+      "loss": 0.4836,
       "step": 1650
     },
     {
-      "epoch": 0.2424908424908425,
-      "grad_norm": 0.4751687434528722,
-      "learning_rate": 1.6162109375000002e-05,
-      "loss": 0.4771,
+      "epoch": 0.2436331517738849,
+      "grad_norm": 0.5226568976328225,
+      "learning_rate": 1.6241413150147204e-05,
+      "loss": 0.4708,
       "step": 1655
     },
     {
-      "epoch": 0.24322344322344322,
-      "grad_norm": 0.5229761322805487,
-      "learning_rate": 1.62109375e-05,
-      "loss": 0.4827,
+      "epoch": 0.24436920359193287,
+      "grad_norm": 0.5157607231064318,
+      "learning_rate": 1.6290480863591758e-05,
+      "loss": 0.4601,
       "step": 1660
     },
     {
-      "epoch": 0.24395604395604395,
-      "grad_norm": 0.5022996752707977,
-      "learning_rate": 1.6259765625e-05,
-      "loss": 0.4686,
+      "epoch": 0.24510525540998085,
+      "grad_norm": 0.5395851933711503,
+      "learning_rate": 1.633954857703631e-05,
+      "loss": 0.4752,
       "step": 1665
     },
     {
-      "epoch": 0.24468864468864468,
-      "grad_norm": 0.4641070216865701,
-      "learning_rate": 1.6308593750000003e-05,
-      "loss": 0.4697,
+      "epoch": 0.24584130722802885,
+      "grad_norm": 0.4751495401729902,
+      "learning_rate": 1.6388616290480867e-05,
+      "loss": 0.4705,
       "step": 1670
     },
     {
-      "epoch": 0.2454212454212454,
-      "grad_norm": 0.49275154239134633,
-      "learning_rate": 1.6357421875e-05,
-      "loss": 0.474,
+      "epoch": 0.24657735904607683,
+      "grad_norm": 0.5237146203386382,
+      "learning_rate": 1.6437684003925418e-05,
+      "loss": 0.4816,
       "step": 1675
     },
     {
-      "epoch": 0.24615384615384617,
-      "grad_norm": 0.4661966887748862,
-      "learning_rate": 1.6406250000000002e-05,
-      "loss": 0.4929,
+      "epoch": 0.24731341086412484,
+      "grad_norm": 0.5033688649754753,
+      "learning_rate": 1.6486751717369972e-05,
+      "loss": 0.4712,
       "step": 1680
     },
     {
-      "epoch": 0.2468864468864469,
-      "grad_norm": 0.5012839182422457,
-      "learning_rate": 1.6455078125e-05,
-      "loss": 0.4713,
+      "epoch": 0.24804946268217282,
+      "grad_norm": 0.514702849413646,
+      "learning_rate": 1.6535819430814527e-05,
+      "loss": 0.4837,
       "step": 1685
     },
     {
-      "epoch": 0.24761904761904763,
-      "grad_norm": 0.49505852648352355,
-      "learning_rate": 1.6503906250000002e-05,
-      "loss": 0.459,
+      "epoch": 0.24878551450022082,
+      "grad_norm": 0.5023097725691634,
+      "learning_rate": 1.6584887144259078e-05,
+      "loss": 0.4738,
       "step": 1690
     },
     {
-      "epoch": 0.24835164835164836,
-      "grad_norm": 0.574214037509089,
-      "learning_rate": 1.6552734375e-05,
-      "loss": 0.4909,
+      "epoch": 0.2495215663182688,
+      "grad_norm": 0.5183562617108541,
+      "learning_rate": 1.6633954857703632e-05,
+      "loss": 0.4683,
       "step": 1695
     },
     {
-      "epoch": 0.2490842490842491,
-      "grad_norm": 0.5046382196659095,
-      "learning_rate": 1.66015625e-05,
-      "loss": 0.4499,
+      "epoch": 0.2502576181363168,
+      "grad_norm": 0.5360790864899261,
+      "learning_rate": 1.6683022571148187e-05,
+      "loss": 0.4743,
       "step": 1700
     },
     {
-      "epoch": 0.24981684981684982,
-      "grad_norm": 0.49535699131303024,
-      "learning_rate": 1.6650390625000003e-05,
-      "loss": 0.4721,
+      "epoch": 0.2509936699543648,
+      "grad_norm": 0.4907804756360193,
+      "learning_rate": 1.6732090284592738e-05,
+      "loss": 0.4676,
       "step": 1705
     },
     {
-      "epoch": 0.25054945054945055,
-      "grad_norm": 0.528858810537546,
-      "learning_rate": 1.669921875e-05,
-      "loss": 0.4783,
+      "epoch": 0.25172972177241276,
+      "grad_norm": 0.46744276329021295,
+      "learning_rate": 1.6781157998037292e-05,
+      "loss": 0.4716,
       "step": 1710
     },
     {
-      "epoch": 0.2512820512820513,
-      "grad_norm": 0.4971493188600044,
-      "learning_rate": 1.6748046875000002e-05,
-      "loss": 0.4619,
+      "epoch": 0.2524657735904608,
+      "grad_norm": 0.47718721326723956,
+      "learning_rate": 1.6830225711481847e-05,
+      "loss": 0.4871,
       "step": 1715
     },
     {
-      "epoch": 0.252014652014652,
-      "grad_norm": 0.4871153410339573,
-      "learning_rate": 1.6796875e-05,
-      "loss": 0.4639,
+      "epoch": 0.2532018254085088,
+      "grad_norm": 0.5248079760485729,
+      "learning_rate": 1.68792934249264e-05,
+      "loss": 0.4566,
       "step": 1720
     },
     {
-      "epoch": 0.25274725274725274,
-      "grad_norm": 0.4753062852302058,
-      "learning_rate": 1.6845703125e-05,
-      "loss": 0.4678,
+      "epoch": 0.25393787722655675,
+      "grad_norm": 0.49252041030933036,
+      "learning_rate": 1.6928361138370952e-05,
+      "loss": 0.4711,
       "step": 1725
     },
     {
-      "epoch": 0.25347985347985347,
-      "grad_norm": 0.47868759947721007,
-      "learning_rate": 1.6894531250000003e-05,
-      "loss": 0.4608,
+      "epoch": 0.25467392904460473,
+      "grad_norm": 0.4869784177371722,
+      "learning_rate": 1.6977428851815506e-05,
+      "loss": 0.4677,
       "step": 1730
     },
     {
-      "epoch": 0.2542124542124542,
-      "grad_norm": 0.4698689149576196,
-      "learning_rate": 1.6943359375e-05,
-      "loss": 0.4544,
+      "epoch": 0.2554099808626527,
+      "grad_norm": 0.5497380707317164,
+      "learning_rate": 1.702649656526006e-05,
+      "loss": 0.4673,
       "step": 1735
     },
     {
-      "epoch": 0.2549450549450549,
-      "grad_norm": 0.48882593711349437,
-      "learning_rate": 1.6992187500000002e-05,
-      "loss": 0.4728,
+      "epoch": 0.25614603268070074,
+      "grad_norm": 0.4677135024107484,
+      "learning_rate": 1.7075564278704612e-05,
+      "loss": 0.4561,
       "step": 1740
     },
     {
-      "epoch": 0.25567765567765566,
-      "grad_norm": 0.500011982347738,
-      "learning_rate": 1.7041015625e-05,
-      "loss": 0.4677,
+      "epoch": 0.2568820844987487,
+      "grad_norm": 0.5031203169744454,
+      "learning_rate": 1.7124631992149166e-05,
+      "loss": 0.4571,
       "step": 1745
     },
     {
-      "epoch": 0.2564102564102564,
-      "grad_norm": 0.49598465314699264,
-      "learning_rate": 1.708984375e-05,
-      "loss": 0.4772,
+      "epoch": 0.2576181363167967,
+      "grad_norm": 0.6191590969953277,
+      "learning_rate": 1.717369970559372e-05,
+      "loss": 0.4568,
       "step": 1750
     },
     {
-      "epoch": 0.2571428571428571,
-      "grad_norm": 0.4329945364963151,
-      "learning_rate": 1.7138671875000003e-05,
-      "loss": 0.4648,
+      "epoch": 0.2583541881348447,
+      "grad_norm": 0.4908953840225422,
+      "learning_rate": 1.722276741903827e-05,
+      "loss": 0.4819,
       "step": 1755
     },
     {
-      "epoch": 0.2578754578754579,
-      "grad_norm": 0.48232118635122256,
-      "learning_rate": 1.71875e-05,
-      "loss": 0.4954,
+      "epoch": 0.2590902399528927,
+      "grad_norm": 0.47019441958924985,
+      "learning_rate": 1.727183513248283e-05,
+      "loss": 0.4773,
       "step": 1760
     },
     {
-      "epoch": 0.25860805860805863,
-      "grad_norm": 0.4922967812176913,
-      "learning_rate": 1.7236328125000002e-05,
-      "loss": 0.493,
+      "epoch": 0.2598262917709407,
+      "grad_norm": 0.4745018245300634,
+      "learning_rate": 1.732090284592738e-05,
+      "loss": 0.4685,
       "step": 1765
     },
     {
-      "epoch": 0.25934065934065936,
-      "grad_norm": 0.49419218230058926,
-      "learning_rate": 1.728515625e-05,
-      "loss": 0.4778,
+      "epoch": 0.26056234358898867,
+      "grad_norm": 0.5598599406025719,
+      "learning_rate": 1.7369970559371935e-05,
+      "loss": 0.4841,
       "step": 1770
     },
     {
-      "epoch": 0.2600732600732601,
-      "grad_norm": 0.4610463309322283,
-      "learning_rate": 1.7333984375000002e-05,
-      "loss": 0.458,
+      "epoch": 0.26129839540703664,
+      "grad_norm": 0.4657021893443309,
+      "learning_rate": 1.741903827281649e-05,
+      "loss": 0.471,
       "step": 1775
     },
     {
-      "epoch": 0.2608058608058608,
-      "grad_norm": 0.6298389104020712,
-      "learning_rate": 1.73828125e-05,
-      "loss": 0.4724,
+      "epoch": 0.2620344472250846,
+      "grad_norm": 0.4731011863702337,
+      "learning_rate": 1.746810598626104e-05,
+      "loss": 0.4721,
       "step": 1780
     },
     {
-      "epoch": 0.26153846153846155,
-      "grad_norm": 0.44564141424220477,
-      "learning_rate": 1.7431640625e-05,
-      "loss": 0.4547,
+      "epoch": 0.26277049904313265,
+      "grad_norm": 0.4811101419462721,
+      "learning_rate": 1.7517173699705595e-05,
+      "loss": 0.4698,
       "step": 1785
     },
     {
-      "epoch": 0.2622710622710623,
-      "grad_norm": 0.4943282192514032,
-      "learning_rate": 1.7480468750000003e-05,
-      "loss": 0.4682,
+      "epoch": 0.26350655086118063,
+      "grad_norm": 0.493675477012419,
+      "learning_rate": 1.756624141315015e-05,
+      "loss": 0.4847,
       "step": 1790
     },
     {
-      "epoch": 0.263003663003663,
-      "grad_norm": 0.5296359933117363,
-      "learning_rate": 1.7529296875e-05,
-      "loss": 0.4668,
+      "epoch": 0.2642426026792286,
+      "grad_norm": 0.5003651025617639,
+      "learning_rate": 1.76153091265947e-05,
+      "loss": 0.4877,
       "step": 1795
     },
     {
-      "epoch": 0.26373626373626374,
-      "grad_norm": 0.47784329131477293,
-      "learning_rate": 1.7578125000000002e-05,
-      "loss": 0.4623,
+      "epoch": 0.2649786544972766,
+      "grad_norm": 0.6143919387458073,
+      "learning_rate": 1.7664376840039255e-05,
+      "loss": 0.4648,
       "step": 1800
     },
     {
-      "epoch": 0.2644688644688645,
-      "grad_norm": 0.474034871482463,
-      "learning_rate": 1.7626953125e-05,
-      "loss": 0.4637,
+      "epoch": 0.2657147063153246,
+      "grad_norm": 0.4927430508144176,
+      "learning_rate": 1.771344455348381e-05,
+      "loss": 0.4813,
       "step": 1805
     },
     {
-      "epoch": 0.2652014652014652,
-      "grad_norm": 0.46948293169273314,
-      "learning_rate": 1.767578125e-05,
-      "loss": 0.4659,
+      "epoch": 0.2664507581333726,
+      "grad_norm": 0.48103388838746014,
+      "learning_rate": 1.7762512266928363e-05,
+      "loss": 0.4674,
       "step": 1810
     },
     {
-      "epoch": 0.26593406593406593,
-      "grad_norm": 0.45953721175077844,
-      "learning_rate": 1.7724609375000003e-05,
-      "loss": 0.4866,
+      "epoch": 0.2671868099514206,
+      "grad_norm": 0.4499372204828534,
+      "learning_rate": 1.7811579980372914e-05,
+      "loss": 0.4642,
       "step": 1815
     },
     {
-      "epoch": 0.26666666666666666,
-      "grad_norm": 0.5001978170812945,
-      "learning_rate": 1.77734375e-05,
-      "loss": 0.4675,
+      "epoch": 0.26792286176946856,
+      "grad_norm": 0.4971451354459214,
+      "learning_rate": 1.786064769381747e-05,
+      "loss": 0.4616,
       "step": 1820
     },
     {
-      "epoch": 0.2673992673992674,
-      "grad_norm": 0.4749703185073931,
-      "learning_rate": 1.7822265625000002e-05,
-      "loss": 0.4985,
+      "epoch": 0.26865891358751653,
+      "grad_norm": 0.458431302762374,
+      "learning_rate": 1.7909715407262023e-05,
+      "loss": 0.4688,
       "step": 1825
     },
     {
-      "epoch": 0.2681318681318681,
-      "grad_norm": 0.49066995039881806,
-      "learning_rate": 1.787109375e-05,
-      "loss": 0.4921,
+      "epoch": 0.26939496540556457,
+      "grad_norm": 0.4969338641891289,
+      "learning_rate": 1.7958783120706574e-05,
+      "loss": 0.4754,
       "step": 1830
     },
     {
-      "epoch": 0.26886446886446885,
-      "grad_norm": 0.49405211470211396,
-      "learning_rate": 1.7919921875e-05,
-      "loss": 0.4892,
+      "epoch": 0.27013101722361255,
+      "grad_norm": 0.4600582361624074,
+      "learning_rate": 1.8007850834151132e-05,
+      "loss": 0.4709,
       "step": 1835
     },
     {
-      "epoch": 0.2695970695970696,
-      "grad_norm": 0.49291065382018007,
-      "learning_rate": 1.7968750000000003e-05,
-      "loss": 0.4742,
+      "epoch": 0.2708670690416605,
+      "grad_norm": 0.5108060982398376,
+      "learning_rate": 1.8056918547595683e-05,
+      "loss": 0.4747,
       "step": 1840
     },
     {
-      "epoch": 0.2703296703296703,
-      "grad_norm": 0.4494775993205646,
-      "learning_rate": 1.8017578125e-05,
-      "loss": 0.4536,
+      "epoch": 0.2716031208597085,
+      "grad_norm": 0.4460882058437207,
+      "learning_rate": 1.8105986261040237e-05,
+      "loss": 0.479,
       "step": 1845
     },
     {
-      "epoch": 0.27106227106227104,
-      "grad_norm": 0.4488385326855789,
-      "learning_rate": 1.8066406250000002e-05,
-      "loss": 0.4906,
+      "epoch": 0.27233917267775654,
+      "grad_norm": 0.4723634907192592,
+      "learning_rate": 1.8155053974484792e-05,
+      "loss": 0.4683,
       "step": 1850
     },
     {
-      "epoch": 0.2717948717948718,
-      "grad_norm": 0.46467724220767814,
-      "learning_rate": 1.8115234375e-05,
-      "loss": 0.4752,
+      "epoch": 0.2730752244958045,
+      "grad_norm": 0.46938877208903224,
+      "learning_rate": 1.8204121687929343e-05,
+      "loss": 0.4674,
       "step": 1855
     },
     {
-      "epoch": 0.2725274725274725,
-      "grad_norm": 0.4676235296258921,
-      "learning_rate": 1.8164062500000002e-05,
-      "loss": 0.4689,
+      "epoch": 0.2738112763138525,
+      "grad_norm": 0.551824744493149,
+      "learning_rate": 1.8253189401373897e-05,
+      "loss": 0.4763,
       "step": 1860
     },
     {
-      "epoch": 0.27326007326007323,
-      "grad_norm": 0.5393139967406888,
-      "learning_rate": 1.8212890625e-05,
-      "loss": 0.4697,
+      "epoch": 0.27454732813190047,
+      "grad_norm": 0.48403544448997265,
+      "learning_rate": 1.830225711481845e-05,
+      "loss": 0.4756,
       "step": 1865
     },
     {
-      "epoch": 0.273992673992674,
-      "grad_norm": 0.5861718805708979,
-      "learning_rate": 1.826171875e-05,
-      "loss": 0.4635,
+      "epoch": 0.2752833799499485,
+      "grad_norm": 0.46983314682824356,
+      "learning_rate": 1.8351324828263003e-05,
+      "loss": 0.4819,
       "step": 1870
     },
     {
-      "epoch": 0.27472527472527475,
-      "grad_norm": 0.8103200245272049,
-      "learning_rate": 1.8310546875000003e-05,
-      "loss": 0.4793,
+      "epoch": 0.2760194317679965,
+      "grad_norm": 0.4576607658725547,
+      "learning_rate": 1.8400392541707557e-05,
+      "loss": 0.4455,
       "step": 1875
     },
     {
-      "epoch": 0.2754578754578755,
-      "grad_norm": 0.4880340601172291,
-      "learning_rate": 1.8359375e-05,
-      "loss": 0.466,
+      "epoch": 0.27675548358604446,
+      "grad_norm": 0.47118086211235977,
+      "learning_rate": 1.844946025515211e-05,
+      "loss": 0.4694,
       "step": 1880
     },
     {
-      "epoch": 0.2761904761904762,
-      "grad_norm": 0.4790298158701428,
-      "learning_rate": 1.8408203125000002e-05,
-      "loss": 0.4923,
+      "epoch": 0.27749153540409244,
+      "grad_norm": 0.49239146345942914,
+      "learning_rate": 1.8498527968596666e-05,
+      "loss": 0.4805,
       "step": 1885
     },
     {
-      "epoch": 0.27692307692307694,
-      "grad_norm": 0.6603329163970332,
-      "learning_rate": 1.845703125e-05,
-      "loss": 0.4881,
+      "epoch": 0.2782275872221404,
+      "grad_norm": 0.5594939354842817,
+      "learning_rate": 1.8547595682041217e-05,
+      "loss": 0.4754,
       "step": 1890
     },
     {
-      "epoch": 0.27765567765567767,
-      "grad_norm": 0.493241053754715,
-      "learning_rate": 1.8505859375e-05,
-      "loss": 0.4857,
+      "epoch": 0.27896363904018845,
+      "grad_norm": 0.47899677266183416,
+      "learning_rate": 1.859666339548577e-05,
+      "loss": 0.471,
       "step": 1895
     },
     {
-      "epoch": 0.2783882783882784,
-      "grad_norm": 0.4681650600538861,
-      "learning_rate": 1.8554687500000003e-05,
-      "loss": 0.4762,
+      "epoch": 0.2796996908582364,
+      "grad_norm": 0.6334955053175579,
+      "learning_rate": 1.8645731108930326e-05,
+      "loss": 0.4782,
       "step": 1900
     },
     {
-      "epoch": 0.27912087912087913,
-      "grad_norm": 0.4633678790580011,
-      "learning_rate": 1.8603515625e-05,
-      "loss": 0.4609,
+      "epoch": 0.2804357426762844,
+      "grad_norm": 0.46276680124375486,
+      "learning_rate": 1.8694798822374877e-05,
+      "loss": 0.47,
       "step": 1905
     },
     {
-      "epoch": 0.27985347985347986,
-      "grad_norm": 0.4735824848064331,
-      "learning_rate": 1.8652343750000002e-05,
-      "loss": 0.4459,
+      "epoch": 0.2811717944943324,
+      "grad_norm": 0.46323911418641445,
+      "learning_rate": 1.874386653581943e-05,
+      "loss": 0.4599,
       "step": 1910
     },
     {
-      "epoch": 0.2805860805860806,
-      "grad_norm": 0.46703566408803404,
-      "learning_rate": 1.8701171875e-05,
-      "loss": 0.4624,
+      "epoch": 0.2819078463123804,
+      "grad_norm": 0.476336202980536,
+      "learning_rate": 1.8792934249263986e-05,
+      "loss": 0.4801,
       "step": 1915
     },
     {
-      "epoch": 0.2813186813186813,
-      "grad_norm": 0.4720185289135887,
-      "learning_rate": 1.8750000000000002e-05,
-      "loss": 0.4775,
+      "epoch": 0.2826438981304284,
+      "grad_norm": 0.4498718281230744,
+      "learning_rate": 1.884200196270854e-05,
+      "loss": 0.4563,
       "step": 1920
     },
     {
-      "epoch": 0.28205128205128205,
-      "grad_norm": 0.47311678768321663,
-      "learning_rate": 1.8798828125000003e-05,
-      "loss": 0.4738,
+      "epoch": 0.2833799499484764,
+      "grad_norm": 0.4671774377376596,
+      "learning_rate": 1.8891069676153094e-05,
+      "loss": 0.4792,
       "step": 1925
     },
     {
-      "epoch": 0.2827838827838828,
-      "grad_norm": 0.4572599200236833,
-      "learning_rate": 1.884765625e-05,
-      "loss": 0.48,
+      "epoch": 0.28411600176652435,
+      "grad_norm": 0.4500480347092648,
+      "learning_rate": 1.8940137389597645e-05,
+      "loss": 0.4743,
       "step": 1930
     },
     {
-      "epoch": 0.2835164835164835,
-      "grad_norm": 0.4818085410272587,
-      "learning_rate": 1.8896484375000003e-05,
-      "loss": 0.4726,
+      "epoch": 0.28485205358457233,
+      "grad_norm": 0.5107626573555243,
+      "learning_rate": 1.89892051030422e-05,
+      "loss": 0.48,
       "step": 1935
     },
     {
-      "epoch": 0.28424908424908424,
-      "grad_norm": 0.4629029786842517,
-      "learning_rate": 1.89453125e-05,
-      "loss": 0.4715,
+      "epoch": 0.28558810540262036,
+      "grad_norm": 0.517903581914278,
+      "learning_rate": 1.9038272816486754e-05,
+      "loss": 0.4892,
       "step": 1940
     },
     {
-      "epoch": 0.28498168498168497,
-      "grad_norm": 0.4873451515455,
-      "learning_rate": 1.8994140625000002e-05,
-      "loss": 0.4564,
+      "epoch": 0.28632415722066834,
+      "grad_norm": 0.4766111115870996,
+      "learning_rate": 1.9087340529931305e-05,
+      "loss": 0.4692,
       "step": 1945
     },
     {
-      "epoch": 0.2857142857142857,
-      "grad_norm": 0.4664172453174398,
-      "learning_rate": 1.904296875e-05,
-      "loss": 0.4585,
+      "epoch": 0.2870602090387163,
+      "grad_norm": 0.4825649575499548,
+      "learning_rate": 1.913640824337586e-05,
+      "loss": 0.49,
       "step": 1950
     },
     {
-      "epoch": 0.28644688644688643,
-      "grad_norm": 0.4696566885216601,
-      "learning_rate": 1.9091796875e-05,
-      "loss": 0.4603,
+      "epoch": 0.2877962608567643,
+      "grad_norm": 0.4724648176127981,
+      "learning_rate": 1.9185475956820414e-05,
+      "loss": 0.4932,
       "step": 1955
     },
     {
-      "epoch": 0.28717948717948716,
-      "grad_norm": 0.46259581185803655,
-      "learning_rate": 1.9140625000000003e-05,
-      "loss": 0.4616,
+      "epoch": 0.28853231267481233,
+      "grad_norm": 0.47881279516123754,
+      "learning_rate": 1.923454367026497e-05,
+      "loss": 0.4801,
       "step": 1960
     },
     {
-      "epoch": 0.2879120879120879,
-      "grad_norm": 0.46970335979866645,
-      "learning_rate": 1.9189453125e-05,
-      "loss": 0.4736,
+      "epoch": 0.2892683644928603,
+      "grad_norm": 0.45468198120850223,
+      "learning_rate": 1.9283611383709523e-05,
+      "loss": 0.4806,
       "step": 1965
     },
     {
-      "epoch": 0.2886446886446886,
-      "grad_norm": 0.47296502986396777,
-      "learning_rate": 1.9238281250000002e-05,
-      "loss": 0.4831,
+      "epoch": 0.2900044163109083,
+      "grad_norm": 0.44628501808614907,
+      "learning_rate": 1.9332679097154074e-05,
+      "loss": 0.4642,
       "step": 1970
     },
     {
-      "epoch": 0.2893772893772894,
-      "grad_norm": 0.5115242848392085,
-      "learning_rate": 1.9287109375e-05,
-      "loss": 0.4844,
+      "epoch": 0.29074046812895626,
+      "grad_norm": 0.44328776399808534,
+      "learning_rate": 1.9381746810598628e-05,
+      "loss": 0.4578,
       "step": 1975
     },
     {
-      "epoch": 0.29010989010989013,
-      "grad_norm": 0.4705181315677594,
-      "learning_rate": 1.93359375e-05,
-      "loss": 0.4854,
+      "epoch": 0.29147651994700424,
+      "grad_norm": 0.4684544494472312,
+      "learning_rate": 1.9430814524043183e-05,
+      "loss": 0.4675,
       "step": 1980
     },
     {
-      "epoch": 0.29084249084249086,
-      "grad_norm": 0.47843755815200834,
-      "learning_rate": 1.9384765625000003e-05,
-      "loss": 0.4663,
+      "epoch": 0.2922125717650523,
+      "grad_norm": 0.4352182628674537,
+      "learning_rate": 1.9479882237487734e-05,
+      "loss": 0.4735,
       "step": 1985
     },
     {
-      "epoch": 0.2915750915750916,
-      "grad_norm": 0.45771995006184946,
-      "learning_rate": 1.943359375e-05,
-      "loss": 0.4613,
+      "epoch": 0.29294862358310025,
+      "grad_norm": 0.45561466921902033,
+      "learning_rate": 1.9528949950932288e-05,
+      "loss": 0.4741,
       "step": 1990
     },
     {
-      "epoch": 0.2923076923076923,
-      "grad_norm": 0.48868203559391954,
-      "learning_rate": 1.9482421875000002e-05,
-      "loss": 0.4744,
+      "epoch": 0.29368467540114823,
+      "grad_norm": 0.43295581864349225,
+      "learning_rate": 1.9578017664376843e-05,
+      "loss": 0.4602,
       "step": 1995
     },
     {
-      "epoch": 0.29304029304029305,
-      "grad_norm": 0.46424047982029926,
-      "learning_rate": 1.953125e-05,
-      "loss": 0.4742,
+      "epoch": 0.2944207272191962,
+      "grad_norm": 0.4316839171230612,
+      "learning_rate": 1.9627085377821397e-05,
+      "loss": 0.467,
       "step": 2000
     },
     {
-      "epoch": 0.2937728937728938,
-      "grad_norm": 0.48114634444665394,
-      "learning_rate": 1.9580078125000002e-05,
-      "loss": 0.453,
+      "epoch": 0.29515677903724424,
+      "grad_norm": 0.4569846908548342,
+      "learning_rate": 1.9676153091265948e-05,
+      "loss": 0.4721,
       "step": 2005
     },
     {
-      "epoch": 0.2945054945054945,
-      "grad_norm": 0.4356027962155649,
-      "learning_rate": 1.9628906250000003e-05,
-      "loss": 0.4662,
+      "epoch": 0.2958928308552922,
+      "grad_norm": 0.4616736815476582,
+      "learning_rate": 1.9725220804710502e-05,
+      "loss": 0.4711,
       "step": 2010
     },
     {
-      "epoch": 0.29523809523809524,
-      "grad_norm": 0.45641072197702454,
-      "learning_rate": 1.9677734375e-05,
-      "loss": 0.4703,
+      "epoch": 0.2966288826733402,
+      "grad_norm": 0.4335580397496052,
+      "learning_rate": 1.9774288518155057e-05,
+      "loss": 0.4773,
       "step": 2015
     },
     {
-      "epoch": 0.295970695970696,
-      "grad_norm": 0.47139505467873855,
-      "learning_rate": 1.9726562500000003e-05,
-      "loss": 0.4728,
+      "epoch": 0.2973649344913882,
+      "grad_norm": 0.5072829009356221,
+      "learning_rate": 1.9823356231599608e-05,
+      "loss": 0.4618,
       "step": 2020
     },
     {
-      "epoch": 0.2967032967032967,
-      "grad_norm": 0.5130247107540543,
-      "learning_rate": 1.9775390625e-05,
-      "loss": 0.4827,
+      "epoch": 0.2981009863094362,
+      "grad_norm": 0.4553026268639488,
+      "learning_rate": 1.9872423945044162e-05,
+      "loss": 0.487,
       "step": 2025
     },
     {
-      "epoch": 0.29743589743589743,
-      "grad_norm": 0.47811321943368856,
-      "learning_rate": 1.9824218750000002e-05,
-      "loss": 0.4734,
+      "epoch": 0.2988370381274842,
+      "grad_norm": 0.46481213419200135,
+      "learning_rate": 1.9921491658488717e-05,
+      "loss": 0.4682,
       "step": 2030
     },
     {
-      "epoch": 0.29816849816849816,
-      "grad_norm": 0.47001518772631856,
-      "learning_rate": 1.9873046875e-05,
-      "loss": 0.4782,
+      "epoch": 0.29957308994553217,
+      "grad_norm": 0.4350216051138614,
+      "learning_rate": 1.9970559371933268e-05,
+      "loss": 0.4869,
       "step": 2035
     },
     {
-      "epoch": 0.2989010989010989,
-      "grad_norm": 0.4691655671055533,
-      "learning_rate": 1.9921875e-05,
-      "loss": 0.4534,
+      "epoch": 0.30030914176358015,
+      "grad_norm": 0.44647682822227974,
+      "learning_rate": 1.9999999413208634e-05,
+      "loss": 0.4687,
       "step": 2040
     },
     {
-      "epoch": 0.2996336996336996,
-      "grad_norm": 0.46638939567865445,
-      "learning_rate": 1.9970703125000003e-05,
-      "loss": 0.4727,
+      "epoch": 0.3010451935816281,
+      "grad_norm": 0.4715234199947958,
+      "learning_rate": 1.9999992811806558e-05,
+      "loss": 0.4681,
       "step": 2045
     },
     {
-      "epoch": 0.30036630036630035,
-      "grad_norm": 0.4849584243099219,
-      "learning_rate": 1.9999999418673043e-05,
-      "loss": 0.486,
+      "epoch": 0.30178124539967616,
+      "grad_norm": 0.4332618051435828,
+      "learning_rate": 1.9999978875518057e-05,
+      "loss": 0.4492,
       "step": 2050
     },
     {
-      "epoch": 0.3010989010989011,
-      "grad_norm": 0.46608548906747865,
-      "learning_rate": 1.9999992878745508e-05,
-      "loss": 0.4817,
+      "epoch": 0.30251729721772413,
+      "grad_norm": 0.48395635327674913,
+      "learning_rate": 1.9999957604353346e-05,
+      "loss": 0.485,
       "step": 2055
     },
     {
-      "epoch": 0.3018315018315018,
-      "grad_norm": 0.4582597387899196,
-      "learning_rate": 1.9999979072236512e-05,
-      "loss": 0.4655,
+      "epoch": 0.3032533490357721,
+      "grad_norm": 0.4799025445525673,
+      "learning_rate": 1.999992899832804e-05,
+      "loss": 0.4728,
       "step": 2060
     },
     {
-      "epoch": 0.30256410256410254,
-      "grad_norm": 0.4498963606835548,
-      "learning_rate": 1.9999957999156077e-05,
-      "loss": 0.4648,
+      "epoch": 0.3039894008538201,
+      "grad_norm": 0.4596658733934185,
+      "learning_rate": 1.999989305746311e-05,
+      "loss": 0.452,
       "step": 2065
     },
     {
-      "epoch": 0.3032967032967033,
-      "grad_norm": 0.4881316781230909,
-      "learning_rate": 1.9999929659519525e-05,
-      "loss": 0.4869,
+      "epoch": 0.3047254526718681,
+      "grad_norm": 0.46794477322830674,
+      "learning_rate": 1.999984978178492e-05,
+      "loss": 0.4726,
       "step": 2070
     },
     {
-      "epoch": 0.304029304029304,
-      "grad_norm": 0.47231720891197954,
-      "learning_rate": 1.9999894053347445e-05,
-      "loss": 0.4633,
+      "epoch": 0.3054615044899161,
+      "grad_norm": 0.47232714732023295,
+      "learning_rate": 1.999979917132522e-05,
+      "loss": 0.4924,
       "step": 2075
     },
     {
-      "epoch": 0.3047619047619048,
-      "grad_norm": 0.440552640547169,
-      "learning_rate": 1.999985118066571e-05,
-      "loss": 0.4732,
+      "epoch": 0.3061975563079641,
+      "grad_norm": 0.4756519580313491,
+      "learning_rate": 1.9999741226121124e-05,
+      "loss": 0.4748,
       "step": 2080
     },
     {
-      "epoch": 0.3054945054945055,
-      "grad_norm": 0.4430343759090646,
-      "learning_rate": 1.999980104150548e-05,
-      "loss": 0.4857,
+      "epoch": 0.30693360812601206,
+      "grad_norm": 0.5423454091606669,
+      "learning_rate": 1.9999675946215134e-05,
+      "loss": 0.4536,
       "step": 2085
     },
     {
-      "epoch": 0.30622710622710625,
-      "grad_norm": 0.4694057908906834,
-      "learning_rate": 1.999974363590318e-05,
-      "loss": 0.5014,
+      "epoch": 0.30766965994406004,
+      "grad_norm": 0.47056906278754096,
+      "learning_rate": 1.9999603331655143e-05,
+      "loss": 0.4932,
       "step": 2090
     },
     {
-      "epoch": 0.306959706959707,
-      "grad_norm": 0.429012323732836,
-      "learning_rate": 1.9999678963900533e-05,
-      "loss": 0.471,
+      "epoch": 0.30840571176210807,
+      "grad_norm": 0.43973383442548275,
+      "learning_rate": 1.9999523382494397e-05,
+      "loss": 0.4736,
       "step": 2095
     },
     {
-      "epoch": 0.3076923076923077,
-      "grad_norm": 0.4650830657293966,
-      "learning_rate": 1.9999607025544525e-05,
-      "loss": 0.4774,
+      "epoch": 0.30914176358015605,
+      "grad_norm": 0.46132412885474183,
+      "learning_rate": 1.999943609879155e-05,
+      "loss": 0.478,
       "step": 2100
     },
     {
-      "epoch": 0.30842490842490844,
-      "grad_norm": 0.49607542309669084,
-      "learning_rate": 1.9999527820887437e-05,
-      "loss": 0.4641,
+      "epoch": 0.309877815398204,
+      "grad_norm": 0.43501549406649326,
+      "learning_rate": 1.999934148061062e-05,
+      "loss": 0.4678,
       "step": 2105
     },
     {
-      "epoch": 0.30915750915750917,
-      "grad_norm": 0.44564274747775723,
-      "learning_rate": 1.999944134998682e-05,
-      "loss": 0.489,
+      "epoch": 0.310613867216252,
+      "grad_norm": 0.4401058980449031,
+      "learning_rate": 1.999923952802101e-05,
+      "loss": 0.4765,
       "step": 2110
     },
     {
-      "epoch": 0.3098901098901099,
-      "grad_norm": 0.4544573494793725,
-      "learning_rate": 1.999934761290551e-05,
-      "loss": 0.4845,
+      "epoch": 0.31134991903430004,
+      "grad_norm": 0.4394249742033585,
+      "learning_rate": 1.999913024109749e-05,
+      "loss": 0.4856,
       "step": 2115
     },
     {
-      "epoch": 0.31062271062271063,
-      "grad_norm": 0.47085329248475644,
-      "learning_rate": 1.9999246609711624e-05,
-      "loss": 0.4879,
+      "epoch": 0.312085970852348,
+      "grad_norm": 0.4435086725351014,
+      "learning_rate": 1.999901361992024e-05,
+      "loss": 0.4754,
       "step": 2120
     },
     {
-      "epoch": 0.31135531135531136,
-      "grad_norm": 0.4776391928865398,
-      "learning_rate": 1.9999138340478554e-05,
-      "loss": 0.4801,
+      "epoch": 0.312822022670396,
+      "grad_norm": 0.4541298259035805,
+      "learning_rate": 1.9998889664574785e-05,
+      "loss": 0.4743,
       "step": 2125
     },
     {
-      "epoch": 0.3120879120879121,
-      "grad_norm": 0.4703485930609568,
-      "learning_rate": 1.9999022805284977e-05,
-      "loss": 0.4695,
+      "epoch": 0.31355807448844397,
+      "grad_norm": 0.4567073617700798,
+      "learning_rate": 1.9998758375152052e-05,
+      "loss": 0.4769,
       "step": 2130
     },
     {
-      "epoch": 0.3128205128205128,
-      "grad_norm": 0.47262214629039767,
-      "learning_rate": 1.9998900004214845e-05,
-      "loss": 0.4788,
+      "epoch": 0.31429412630649195,
+      "grad_norm": 0.48915117920176165,
+      "learning_rate": 1.9998619751748338e-05,
+      "loss": 0.4576,
       "step": 2135
     },
     {
-      "epoch": 0.31355311355311355,
-      "grad_norm": 0.46721083494958504,
-      "learning_rate": 1.9998769937357394e-05,
-      "loss": 0.4781,
+      "epoch": 0.31503017812454,
+      "grad_norm": 0.44619255283557624,
+      "learning_rate": 1.999847379446532e-05,
+      "loss": 0.4562,
       "step": 2140
     },
     {
-      "epoch": 0.3142857142857143,
-      "grad_norm": 0.46564165100287497,
-      "learning_rate": 1.9998632604807138e-05,
-      "loss": 0.4591,
+      "epoch": 0.31576622994258796,
+      "grad_norm": 0.4450124222386723,
+      "learning_rate": 1.9998320503410064e-05,
+      "loss": 0.4905,
       "step": 2145
     },
     {
-      "epoch": 0.315018315018315,
-      "grad_norm": 0.46052659281533487,
-      "learning_rate": 1.999848800666387e-05,
-      "loss": 0.4615,
+      "epoch": 0.31650228176063594,
+      "grad_norm": 0.4404465778071448,
+      "learning_rate": 1.9998159878694997e-05,
+      "loss": 0.4815,
       "step": 2150
     },
     {
-      "epoch": 0.31575091575091574,
-      "grad_norm": 0.4523921077473231,
-      "learning_rate": 1.999833614303267e-05,
-      "loss": 0.4945,
+      "epoch": 0.3172383335786839,
+      "grad_norm": 0.5675612692588065,
+      "learning_rate": 1.9997991920437943e-05,
+      "loss": 0.4999,
       "step": 2155
     },
     {
-      "epoch": 0.31648351648351647,
-      "grad_norm": 0.6224759141299284,
-      "learning_rate": 1.999817701402388e-05,
-      "loss": 0.4865,
+      "epoch": 0.31797438539673195,
+      "grad_norm": 0.43280904474802123,
+      "learning_rate": 1.999781662876209e-05,
+      "loss": 0.476,
       "step": 2160
     },
     {
-      "epoch": 0.3172161172161172,
-      "grad_norm": 0.4549954778580932,
-      "learning_rate": 1.9998010619753138e-05,
-      "loss": 0.4703,
+      "epoch": 0.31871043721477993,
+      "grad_norm": 0.4235352582007157,
+      "learning_rate": 1.9997634003796025e-05,
+      "loss": 0.4675,
       "step": 2165
     },
     {
-      "epoch": 0.31794871794871793,
-      "grad_norm": 0.44869160471700487,
-      "learning_rate": 1.9997836960341356e-05,
-      "loss": 0.4916,
+      "epoch": 0.3194464890328279,
+      "grad_norm": 0.4688991793652292,
+      "learning_rate": 1.999744404567369e-05,
+      "loss": 0.4742,
       "step": 2170
     },
     {
-      "epoch": 0.31868131868131866,
-      "grad_norm": 0.48138120519581834,
-      "learning_rate": 1.999765603591473e-05,
-      "loss": 0.4735,
+      "epoch": 0.3201825408508759,
+      "grad_norm": 0.48925261025186434,
+      "learning_rate": 1.999724675453442e-05,
+      "loss": 0.4888,
       "step": 2175
     },
     {
-      "epoch": 0.3194139194139194,
-      "grad_norm": 0.4606953753029427,
-      "learning_rate": 1.999746784660472e-05,
-      "loss": 0.4655,
+      "epoch": 0.3209185926689239,
+      "grad_norm": 0.4495508697602979,
+      "learning_rate": 1.999704213052293e-05,
+      "loss": 0.4768,
       "step": 2180
     },
     {
-      "epoch": 0.3201465201465201,
-      "grad_norm": 0.4960164215656271,
-      "learning_rate": 1.999727239254808e-05,
-      "loss": 0.4874,
+      "epoch": 0.3216546444869719,
+      "grad_norm": 0.4609517564296356,
+      "learning_rate": 1.99968301737893e-05,
+      "loss": 0.4664,
       "step": 2185
     },
     {
-      "epoch": 0.3208791208791209,
-      "grad_norm": 0.4970242069390641,
-      "learning_rate": 1.9997069673886843e-05,
-      "loss": 0.4989,
+      "epoch": 0.3223906963050199,
+      "grad_norm": 0.4467866715727226,
+      "learning_rate": 1.999661088448901e-05,
+      "loss": 0.4849,
       "step": 2190
     },
     {
-      "epoch": 0.32161172161172163,
-      "grad_norm": 0.4453023673170512,
-      "learning_rate": 1.9996859690768307e-05,
-      "loss": 0.4887,
+      "epoch": 0.32312674812306785,
+      "grad_norm": 0.4652266644750633,
+      "learning_rate": 1.99963842627829e-05,
+      "loss": 0.472,
       "step": 2195
     },
     {
-      "epoch": 0.32234432234432236,
-      "grad_norm": 0.4762842181398665,
-      "learning_rate": 1.9996642443345067e-05,
-      "loss": 0.467,
+      "epoch": 0.32386279994111583,
+      "grad_norm": 0.4687520169018355,
+      "learning_rate": 1.9996150308837194e-05,
+      "loss": 0.4689,
       "step": 2200
     },
     {
-      "epoch": 0.3230769230769231,
-      "grad_norm": 0.45617671867468856,
-      "learning_rate": 1.9996417931774986e-05,
-      "loss": 0.4616,
+      "epoch": 0.32459885175916386,
+      "grad_norm": 0.4562403617598054,
+      "learning_rate": 1.9995909022823497e-05,
+      "loss": 0.4636,
       "step": 2205
     },
     {
-      "epoch": 0.3238095238095238,
-      "grad_norm": 0.4577829813582058,
-      "learning_rate": 1.9996186156221202e-05,
-      "loss": 0.4681,
+      "epoch": 0.32533490357721184,
+      "grad_norm": 0.46461524609801136,
+      "learning_rate": 1.9995660404918787e-05,
+      "loss": 0.4887,
       "step": 2210
     },
     {
-      "epoch": 0.32454212454212455,
-      "grad_norm": 0.4667373496580333,
-      "learning_rate": 1.999594711685214e-05,
-      "loss": 0.4811,
+      "epoch": 0.3260709553952598,
+      "grad_norm": 0.44292774850519984,
+      "learning_rate": 1.9995404455305426e-05,
+      "loss": 0.4836,
       "step": 2215
     },
     {
-      "epoch": 0.3252747252747253,
-      "grad_norm": 0.44437406789561146,
-      "learning_rate": 1.99957008138415e-05,
-      "loss": 0.4898,
+      "epoch": 0.3268070072133078,
+      "grad_norm": 0.40967463773562945,
+      "learning_rate": 1.999514117417115e-05,
+      "loss": 0.4792,
       "step": 2220
     },
     {
-      "epoch": 0.326007326007326,
-      "grad_norm": 0.4748532696317031,
-      "learning_rate": 1.9995447247368265e-05,
-      "loss": 0.4617,
+      "epoch": 0.32754305903135583,
+      "grad_norm": 0.4363305863192649,
+      "learning_rate": 1.9994870561709064e-05,
+      "loss": 0.4901,
       "step": 2225
     },
     {
-      "epoch": 0.32673992673992674,
-      "grad_norm": 0.4589811457098767,
-      "learning_rate": 1.9995186417616682e-05,
-      "loss": 0.4863,
+      "epoch": 0.3282791108494038,
+      "grad_norm": 0.4440205284030912,
+      "learning_rate": 1.9994592618117673e-05,
+      "loss": 0.4678,
       "step": 2230
     },
     {
-      "epoch": 0.3274725274725275,
-      "grad_norm": 0.4493781747279523,
-      "learning_rate": 1.999491832477629e-05,
-      "loss": 0.4594,
+      "epoch": 0.3290151626674518,
+      "grad_norm": 0.4637447567088612,
+      "learning_rate": 1.9994307343600838e-05,
+      "loss": 0.4536,
       "step": 2235
     },
     {
-      "epoch": 0.3282051282051282,
-      "grad_norm": 0.4471943486082735,
-      "learning_rate": 1.9994642969041902e-05,
-      "loss": 0.4692,
+      "epoch": 0.32975121448549977,
+      "grad_norm": 0.457851261301874,
+      "learning_rate": 1.9994014738367806e-05,
+      "loss": 0.4691,
       "step": 2240
     },
     {
-      "epoch": 0.32893772893772893,
-      "grad_norm": 0.4471860716440031,
-      "learning_rate": 1.9994360350613605e-05,
-      "loss": 0.4541,
+      "epoch": 0.33048726630354774,
+      "grad_norm": 0.4639933895918874,
+      "learning_rate": 1.99937148026332e-05,
+      "loss": 0.4733,
       "step": 2245
     },
     {
-      "epoch": 0.32967032967032966,
-      "grad_norm": 0.4415395734304109,
-      "learning_rate": 1.9994070469696768e-05,
-      "loss": 0.4606,
+      "epoch": 0.3312233181215958,
+      "grad_norm": 0.5048664352458206,
+      "learning_rate": 1.999340753661702e-05,
+      "loss": 0.4726,
       "step": 2250
     },
     {
-      "epoch": 0.3304029304029304,
-      "grad_norm": 0.44838585468171893,
-      "learning_rate": 1.9993773326502037e-05,
-      "loss": 0.4647,
+      "epoch": 0.33195936993964376,
+      "grad_norm": 0.4796488629165394,
+      "learning_rate": 1.9993092940544637e-05,
+      "loss": 0.4654,
       "step": 2255
     },
     {
-      "epoch": 0.3311355311355311,
-      "grad_norm": 0.48306770352552575,
-      "learning_rate": 1.9993468921245332e-05,
-      "loss": 0.4702,
+      "epoch": 0.33269542175769173,
+      "grad_norm": 0.46805300188648025,
+      "learning_rate": 1.9992771014646816e-05,
+      "loss": 0.458,
       "step": 2260
     },
     {
-      "epoch": 0.33186813186813185,
-      "grad_norm": 0.46716439355838063,
-      "learning_rate": 1.9993157254147847e-05,
-      "loss": 0.4748,
+      "epoch": 0.3334314735757397,
+      "grad_norm": 0.43093945222637614,
+      "learning_rate": 1.9992441759159673e-05,
+      "loss": 0.4428,
       "step": 2265
     },
     {
-      "epoch": 0.3326007326007326,
-      "grad_norm": 0.4395836964883837,
-      "learning_rate": 1.9992838325436062e-05,
-      "loss": 0.4801,
+      "epoch": 0.33416752539378775,
+      "grad_norm": 0.426941370229423,
+      "learning_rate": 1.9992105174324722e-05,
+      "loss": 0.4851,
       "step": 2270
     },
     {
-      "epoch": 0.3333333333333333,
-      "grad_norm": 0.4516346952215036,
-      "learning_rate": 1.9992512135341735e-05,
-      "loss": 0.4907,
+      "epoch": 0.3349035772118357,
+      "grad_norm": 0.4591334473999698,
+      "learning_rate": 1.999176126038884e-05,
+      "loss": 0.4898,
       "step": 2275
     },
     {
-      "epoch": 0.33406593406593404,
-      "grad_norm": 0.47750780474177107,
-      "learning_rate": 1.9992178684101885e-05,
-      "loss": 0.4744,
+      "epoch": 0.3356396290298837,
+      "grad_norm": 0.4251506129306176,
+      "learning_rate": 1.9991410017604283e-05,
+      "loss": 0.4658,
       "step": 2280
     },
     {
-      "epoch": 0.3347985347985348,
-      "grad_norm": 0.46827910368482406,
-      "learning_rate": 1.9991837971958823e-05,
-      "loss": 0.4682,
+      "epoch": 0.3363756808479317,
+      "grad_norm": 0.4782181713266746,
+      "learning_rate": 1.999105144622869e-05,
+      "loss": 0.4666,
       "step": 2285
     },
     {
-      "epoch": 0.3355311355311355,
-      "grad_norm": 0.4284253365569513,
-      "learning_rate": 1.999148999916013e-05,
-      "loss": 0.4682,
+      "epoch": 0.33711173266597966,
+      "grad_norm": 0.4479398981455512,
+      "learning_rate": 1.9990685546525065e-05,
+      "loss": 0.4619,
       "step": 2290
     },
     {
-      "epoch": 0.3362637362637363,
-      "grad_norm": 0.44894863625304804,
-      "learning_rate": 1.999113476595866e-05,
-      "loss": 0.4641,
+      "epoch": 0.3378477844840277,
+      "grad_norm": 0.48245664385321907,
+      "learning_rate": 1.9990312318761788e-05,
+      "loss": 0.4861,
       "step": 2295
     },
     {
-      "epoch": 0.336996336996337,
-      "grad_norm": 0.4319941290239276,
-      "learning_rate": 1.9990772272612545e-05,
-      "loss": 0.4766,
+      "epoch": 0.33858383630207567,
+      "grad_norm": 0.43048953402489026,
+      "learning_rate": 1.9989931763212628e-05,
+      "loss": 0.4792,
       "step": 2300
     },
     {
-      "epoch": 0.33772893772893775,
-      "grad_norm": 0.43325498368336673,
-      "learning_rate": 1.9990402519385204e-05,
-      "loss": 0.4652,
+      "epoch": 0.33931988812012365,
+      "grad_norm": 0.5985939942681147,
+      "learning_rate": 1.998954388015671e-05,
+      "loss": 0.4781,
       "step": 2305
     },
     {
-      "epoch": 0.3384615384615385,
-      "grad_norm": 0.44734423832754516,
-      "learning_rate": 1.9990025506545315e-05,
-      "loss": 0.4717,
+      "epoch": 0.3400559399381716,
+      "grad_norm": 0.43754557176144243,
+      "learning_rate": 1.9989148669878537e-05,
+      "loss": 0.4747,
       "step": 2310
     },
     {
-      "epoch": 0.3391941391941392,
-      "grad_norm": 0.4331449566526826,
-      "learning_rate": 1.998964123436683e-05,
-      "loss": 0.4785,
+      "epoch": 0.34079199175621966,
+      "grad_norm": 0.4373253784689262,
+      "learning_rate": 1.9988746132668006e-05,
+      "loss": 0.4438,
       "step": 2315
     },
     {
-      "epoch": 0.33992673992673994,
-      "grad_norm": 0.4439465535289741,
-      "learning_rate": 1.9989249703128998e-05,
-      "loss": 0.4612,
+      "epoch": 0.34152804357426764,
+      "grad_norm": 0.4768187921022628,
+      "learning_rate": 1.9988336268820365e-05,
+      "loss": 0.4595,
       "step": 2320
     },
     {
-      "epoch": 0.34065934065934067,
-      "grad_norm": 0.4318055127992744,
-      "learning_rate": 1.998885091311632e-05,
-      "loss": 0.4707,
+      "epoch": 0.3422640953923156,
+      "grad_norm": 0.4470593284289592,
+      "learning_rate": 1.9987919078636242e-05,
+      "loss": 0.4702,
       "step": 2325
     },
     {
-      "epoch": 0.3413919413919414,
-      "grad_norm": 0.4412776315633937,
-      "learning_rate": 1.998844486461858e-05,
-      "loss": 0.4844,
+      "epoch": 0.3430001472103636,
+      "grad_norm": 0.44728625386910265,
+      "learning_rate": 1.998749456242165e-05,
+      "loss": 0.4918,
       "step": 2330
     },
     {
-      "epoch": 0.34212454212454213,
-      "grad_norm": 0.4496185459470404,
-      "learning_rate": 1.998803155793084e-05,
-      "loss": 0.4755,
+      "epoch": 0.3437361990284116,
+      "grad_norm": 0.4514444055802256,
+      "learning_rate": 1.9987062720487956e-05,
+      "loss": 0.4826,
       "step": 2335
     },
     {
-      "epoch": 0.34285714285714286,
-      "grad_norm": 0.46607308652238927,
-      "learning_rate": 1.9987610993353436e-05,
-      "loss": 0.4609,
+      "epoch": 0.3444722508464596,
+      "grad_norm": 0.4535206317389082,
+      "learning_rate": 1.998662355315192e-05,
+      "loss": 0.4675,
       "step": 2340
     },
     {
-      "epoch": 0.3435897435897436,
-      "grad_norm": 0.44165834082679384,
-      "learning_rate": 1.9987183171191966e-05,
-      "loss": 0.4552,
+      "epoch": 0.3452083026645076,
+      "grad_norm": 0.4189857578216396,
+      "learning_rate": 1.9986177060735666e-05,
+      "loss": 0.4737,
       "step": 2345
     },
     {
-      "epoch": 0.3443223443223443,
-      "grad_norm": 0.45333703568193734,
-      "learning_rate": 1.9986748091757315e-05,
-      "loss": 0.4684,
+      "epoch": 0.34594435448255556,
+      "grad_norm": 0.5019851500218824,
+      "learning_rate": 1.998572324356669e-05,
+      "loss": 0.4694,
       "step": 2350
     },
     {
-      "epoch": 0.34505494505494505,
-      "grad_norm": 0.4567861600300713,
-      "learning_rate": 1.9986305755365638e-05,
-      "loss": 0.4701,
+      "epoch": 0.34668040630060354,
+      "grad_norm": 0.6105382979046713,
+      "learning_rate": 1.998526210197786e-05,
+      "loss": 0.471,
       "step": 2355
     },
     {
-      "epoch": 0.3457875457875458,
-      "grad_norm": 0.42216281070142847,
-      "learning_rate": 1.9985856162338365e-05,
-      "loss": 0.4814,
+      "epoch": 0.34741645811865157,
+      "grad_norm": 0.4403970102714735,
+      "learning_rate": 1.998479363630742e-05,
+      "loss": 0.451,
       "step": 2360
     },
     {
-      "epoch": 0.3465201465201465,
-      "grad_norm": 0.4614463501828505,
-      "learning_rate": 1.998539931300219e-05,
-      "loss": 0.4724,
+      "epoch": 0.34815250993669955,
+      "grad_norm": 0.4360708417057406,
+      "learning_rate": 1.998431784689898e-05,
+      "loss": 0.4756,
       "step": 2365
     },
     {
-      "epoch": 0.34725274725274724,
-      "grad_norm": 0.48197598208732645,
-      "learning_rate": 1.998493520768909e-05,
-      "loss": 0.4759,
+      "epoch": 0.34888856175474753,
+      "grad_norm": 0.4342094719915312,
+      "learning_rate": 1.998383473410154e-05,
+      "loss": 0.4816,
       "step": 2370
     },
     {
-      "epoch": 0.34798534798534797,
-      "grad_norm": 0.43198815185153344,
-      "learning_rate": 1.9984463846736313e-05,
-      "loss": 0.4602,
+      "epoch": 0.3496246135727955,
+      "grad_norm": 0.4558318975401987,
+      "learning_rate": 1.998334429826944e-05,
+      "loss": 0.4605,
       "step": 2375
     },
     {
-      "epoch": 0.3487179487179487,
-      "grad_norm": 0.44829414019934727,
-      "learning_rate": 1.9983985230486372e-05,
-      "loss": 0.4643,
+      "epoch": 0.35036066539084354,
+      "grad_norm": 0.4459264432189479,
+      "learning_rate": 1.9982846539762422e-05,
+      "loss": 0.4838,
       "step": 2380
     },
     {
-      "epoch": 0.34945054945054943,
-      "grad_norm": 0.4417248573626436,
-      "learning_rate": 1.9983499359287063e-05,
-      "loss": 0.4658,
+      "epoch": 0.3510967172088915,
+      "grad_norm": 0.657558154988274,
+      "learning_rate": 1.9982341458945584e-05,
+      "loss": 0.4727,
       "step": 2385
     },
     {
-      "epoch": 0.35018315018315016,
-      "grad_norm": 0.42631469243015624,
-      "learning_rate": 1.998300623349145e-05,
-      "loss": 0.466,
+      "epoch": 0.3518327690269395,
+      "grad_norm": 0.41325825632624885,
+      "learning_rate": 1.998182905618939e-05,
+      "loss": 0.4772,
       "step": 2390
     },
     {
-      "epoch": 0.3509157509157509,
-      "grad_norm": 0.40101767103579056,
-      "learning_rate": 1.9982505853457862e-05,
-      "loss": 0.4593,
+      "epoch": 0.3525688208449875,
+      "grad_norm": 0.434298057981475,
+      "learning_rate": 1.9981309331869693e-05,
+      "loss": 0.4701,
       "step": 2395
     },
     {
-      "epoch": 0.3516483516483517,
-      "grad_norm": 0.4405886955018248,
-      "learning_rate": 1.998199821954991e-05,
-      "loss": 0.4866,
+      "epoch": 0.35330487266303545,
+      "grad_norm": 0.44706761089730784,
+      "learning_rate": 1.99807822863677e-05,
+      "loss": 0.4798,
       "step": 2400
     },
     {
-      "epoch": 0.3523809523809524,
-      "grad_norm": 0.5123330356572352,
-      "learning_rate": 1.998148333213646e-05,
-      "loss": 0.4684,
+      "epoch": 0.3540409244810835,
+      "grad_norm": 0.45226798144264396,
+      "learning_rate": 1.9980247920069987e-05,
+      "loss": 0.4805,
       "step": 2405
     },
     {
-      "epoch": 0.35311355311355314,
-      "grad_norm": 0.45916305279498376,
-      "learning_rate": 1.9980961191591674e-05,
-      "loss": 0.4845,
+      "epoch": 0.35477697629913146,
+      "grad_norm": 0.4130884443923589,
+      "learning_rate": 1.9979706233368518e-05,
+      "loss": 0.4731,
       "step": 2410
     },
     {
-      "epoch": 0.35384615384615387,
-      "grad_norm": 0.4934563308167938,
-      "learning_rate": 1.998043179829496e-05,
-      "loss": 0.4559,
+      "epoch": 0.35551302811717944,
+      "grad_norm": 0.42720376029824986,
+      "learning_rate": 1.9979157226660605e-05,
+      "loss": 0.4663,
       "step": 2415
     },
     {
-      "epoch": 0.3545787545787546,
-      "grad_norm": 0.44456323846327644,
-      "learning_rate": 1.9979895152631005e-05,
-      "loss": 0.4708,
+      "epoch": 0.3562490799352274,
+      "grad_norm": 0.4506910508944987,
+      "learning_rate": 1.9978600900348947e-05,
+      "loss": 0.4736,
       "step": 2420
     },
     {
-      "epoch": 0.3553113553113553,
-      "grad_norm": 0.4464955895356425,
-      "learning_rate": 1.9979351254989775e-05,
-      "loss": 0.478,
+      "epoch": 0.35698513175327545,
+      "grad_norm": 0.4936781621144996,
+      "learning_rate": 1.997803725484159e-05,
+      "loss": 0.4715,
       "step": 2425
     },
     {
-      "epoch": 0.35604395604395606,
-      "grad_norm": 0.4634970837041198,
-      "learning_rate": 1.9978800105766493e-05,
-      "loss": 0.4802,
+      "epoch": 0.35772118357132343,
+      "grad_norm": 0.44991510071626706,
+      "learning_rate": 1.9977466290551976e-05,
+      "loss": 0.464,
       "step": 2430
     },
     {
-      "epoch": 0.3567765567765568,
-      "grad_norm": 0.4243954559204476,
-      "learning_rate": 1.997824170536166e-05,
-      "loss": 0.4808,
+      "epoch": 0.3584572353893714,
+      "grad_norm": 0.4373424349432953,
+      "learning_rate": 1.9976888007898893e-05,
+      "loss": 0.473,
       "step": 2435
     },
     {
-      "epoch": 0.3575091575091575,
-      "grad_norm": 0.4451848668333986,
-      "learning_rate": 1.9977676054181033e-05,
-      "loss": 0.4589,
+      "epoch": 0.3591932872074194,
+      "grad_norm": 0.43508431133521286,
+      "learning_rate": 1.997630240730651e-05,
+      "loss": 0.4728,
       "step": 2440
     },
     {
-      "epoch": 0.35824175824175825,
-      "grad_norm": 0.44164023449977263,
-      "learning_rate": 1.9977103152635656e-05,
-      "loss": 0.4761,
+      "epoch": 0.35992933902546737,
+      "grad_norm": 0.44562067928514576,
+      "learning_rate": 1.997570948920435e-05,
+      "loss": 0.4781,
       "step": 2445
     },
     {
-      "epoch": 0.358974358974359,
-      "grad_norm": 0.4592316037675001,
-      "learning_rate": 1.9976523001141835e-05,
-      "loss": 0.4728,
+      "epoch": 0.3606653908435154,
+      "grad_norm": 0.4361128978836239,
+      "learning_rate": 1.9975109254027318e-05,
+      "loss": 0.4596,
       "step": 2450
     },
     {
-      "epoch": 0.3597069597069597,
-      "grad_norm": 0.4592825783574398,
-      "learning_rate": 1.9975935600121134e-05,
-      "loss": 0.4673,
+      "epoch": 0.3614014426615634,
+      "grad_norm": 0.46848167821114756,
+      "learning_rate": 1.9974501702215683e-05,
+      "loss": 0.4802,
       "step": 2455
     },
     {
-      "epoch": 0.36043956043956044,
-      "grad_norm": 0.4588396167146063,
-      "learning_rate": 1.99753409500004e-05,
-      "loss": 0.4605,
+      "epoch": 0.36213749447961135,
+      "grad_norm": 0.4507629685143818,
+      "learning_rate": 1.9973886834215076e-05,
+      "loss": 0.4765,
       "step": 2460
     },
     {
-      "epoch": 0.36117216117216117,
-      "grad_norm": 0.4523330376762571,
-      "learning_rate": 1.9974739051211738e-05,
-      "loss": 0.4793,
+      "epoch": 0.36287354629765933,
+      "grad_norm": 0.44318965289648626,
+      "learning_rate": 1.997326465047649e-05,
+      "loss": 0.4729,
       "step": 2465
     },
     {
-      "epoch": 0.3619047619047619,
-      "grad_norm": 0.4505742161300693,
-      "learning_rate": 1.9974129904192525e-05,
-      "loss": 0.4757,
+      "epoch": 0.36360959811570737,
+      "grad_norm": 0.41925778256104673,
+      "learning_rate": 1.9972635151456296e-05,
+      "loss": 0.4832,
       "step": 2470
     },
     {
-      "epoch": 0.3626373626373626,
-      "grad_norm": 0.4171150087533455,
-      "learning_rate": 1.9973513509385394e-05,
-      "loss": 0.455,
+      "epoch": 0.36434564993375534,
+      "grad_norm": 0.43678667449158226,
+      "learning_rate": 1.9971998337616222e-05,
+      "loss": 0.4744,
       "step": 2475
     },
     {
-      "epoch": 0.36336996336996336,
-      "grad_norm": 0.42463363919184804,
-      "learning_rate": 1.9972889867238264e-05,
-      "loss": 0.461,
+      "epoch": 0.3650817017518033,
+      "grad_norm": 0.4715447103249681,
+      "learning_rate": 1.9971354209423368e-05,
+      "loss": 0.4699,
       "step": 2480
     },
     {
-      "epoch": 0.3641025641025641,
-      "grad_norm": 0.47370760239418574,
-      "learning_rate": 1.9972258978204304e-05,
-      "loss": 0.474,
+      "epoch": 0.3658177535698513,
+      "grad_norm": 0.43313415803213506,
+      "learning_rate": 1.9970702767350194e-05,
+      "loss": 0.4654,
       "step": 2485
     },
     {
-      "epoch": 0.3648351648351648,
-      "grad_norm": 0.4499155574302461,
-      "learning_rate": 1.997162084274196e-05,
-      "loss": 0.4791,
+      "epoch": 0.36655380538789933,
+      "grad_norm": 0.506790783037607,
+      "learning_rate": 1.997004401187452e-05,
+      "loss": 0.4691,
       "step": 2490
     },
     {
-      "epoch": 0.36556776556776555,
-      "grad_norm": 0.4696631194673513,
-      "learning_rate": 1.9970975461314933e-05,
-      "loss": 0.491,
+      "epoch": 0.3672898572059473,
+      "grad_norm": 0.48156503079351776,
+      "learning_rate": 1.9969377943479543e-05,
+      "loss": 0.4737,
       "step": 2495
     },
     {
-      "epoch": 0.3663003663003663,
-      "grad_norm": 0.44712091155797246,
-      "learning_rate": 1.99703228343922e-05,
-      "loss": 0.4902,
+      "epoch": 0.3680259090239953,
+      "grad_norm": 0.44798228820681174,
+      "learning_rate": 1.996870456265381e-05,
+      "loss": 0.4592,
       "step": 2500
     },
     {
-      "epoch": 0.367032967032967,
-      "grad_norm": 0.4348061740364967,
-      "learning_rate": 1.9969662962448e-05,
-      "loss": 0.4746,
+      "epoch": 0.36876196084204327,
+      "grad_norm": 0.4637716878626953,
+      "learning_rate": 1.9968023869891245e-05,
+      "loss": 0.4681,
       "step": 2505
     },
     {
-      "epoch": 0.3677655677655678,
-      "grad_norm": 0.5086079541138988,
-      "learning_rate": 1.9968995845961824e-05,
-      "loss": 0.4529,
+      "epoch": 0.36949801266009125,
+      "grad_norm": 0.4470703626223196,
+      "learning_rate": 1.9967335865691128e-05,
+      "loss": 0.4732,
       "step": 2510
     },
     {
-      "epoch": 0.3684981684981685,
-      "grad_norm": 0.44339962250971227,
-      "learning_rate": 1.996832148541845e-05,
-      "loss": 0.4601,
+      "epoch": 0.3702340644781393,
+      "grad_norm": 0.43703534594365284,
+      "learning_rate": 1.99666405505581e-05,
+      "loss": 0.4648,
       "step": 2515
     },
     {
-      "epoch": 0.36923076923076925,
-      "grad_norm": 0.4327290016860837,
-      "learning_rate": 1.9967639881307897e-05,
-      "loss": 0.4762,
+      "epoch": 0.37097011629618726,
+      "grad_norm": 0.4242609070008135,
+      "learning_rate": 1.9965937925002167e-05,
+      "loss": 0.4647,
       "step": 2520
     },
     {
-      "epoch": 0.36996336996337,
-      "grad_norm": 0.4582585890635615,
-      "learning_rate": 1.9966951034125466e-05,
-      "loss": 0.4788,
+      "epoch": 0.37170616811423524,
+      "grad_norm": 0.44885516226659466,
+      "learning_rate": 1.99652279895387e-05,
+      "loss": 0.4828,
       "step": 2525
     },
     {
-      "epoch": 0.3706959706959707,
-      "grad_norm": 0.4425251767984628,
-      "learning_rate": 1.9966254944371713e-05,
-      "loss": 0.4724,
+      "epoch": 0.3724422199322832,
+      "grad_norm": 0.4359073224351897,
+      "learning_rate": 1.9964510744688422e-05,
+      "loss": 0.4775,
       "step": 2530
     },
     {
-      "epoch": 0.37142857142857144,
-      "grad_norm": 0.41646091581942785,
-      "learning_rate": 1.9965551612552455e-05,
-      "loss": 0.4696,
+      "epoch": 0.37317827175033125,
+      "grad_norm": 0.46498654397417966,
+      "learning_rate": 1.9963786190977434e-05,
+      "loss": 0.4716,
       "step": 2535
     },
     {
-      "epoch": 0.37216117216117217,
-      "grad_norm": 0.40403561466112603,
-      "learning_rate": 1.996484103917878e-05,
-      "loss": 0.4581,
+      "epoch": 0.3739143235683792,
+      "grad_norm": 0.41129957424529356,
+      "learning_rate": 1.9963054328937184e-05,
+      "loss": 0.4751,
       "step": 2540
     },
     {
-      "epoch": 0.3728937728937729,
-      "grad_norm": 0.4138783207320578,
-      "learning_rate": 1.9964123224767023e-05,
-      "loss": 0.4691,
+      "epoch": 0.3746503753864272,
+      "grad_norm": 0.4350576054674013,
+      "learning_rate": 1.9962315159104485e-05,
+      "loss": 0.4567,
       "step": 2545
     },
     {
-      "epoch": 0.37362637362637363,
-      "grad_norm": 0.4582236609088255,
-      "learning_rate": 1.9963398169838794e-05,
-      "loss": 0.4598,
+      "epoch": 0.3753864272044752,
+      "grad_norm": 0.444822730056749,
+      "learning_rate": 1.9961568682021508e-05,
+      "loss": 0.4703,
       "step": 2550
     },
     {
-      "epoch": 0.37435897435897436,
-      "grad_norm": 0.43861877305531866,
-      "learning_rate": 1.9962665874920964e-05,
-      "loss": 0.4715,
+      "epoch": 0.37612247902252316,
+      "grad_norm": 0.43892132977734233,
+      "learning_rate": 1.996081489823579e-05,
+      "loss": 0.4749,
       "step": 2555
     },
     {
-      "epoch": 0.3750915750915751,
-      "grad_norm": 0.426211153181784,
-      "learning_rate": 1.996192634054566e-05,
-      "loss": 0.4752,
+      "epoch": 0.3768585308405712,
+      "grad_norm": 0.42200264359981277,
+      "learning_rate": 1.996005380830022e-05,
+      "loss": 0.4658,
       "step": 2560
     },
     {
-      "epoch": 0.3758241758241758,
-      "grad_norm": 0.4353446290075632,
-      "learning_rate": 1.9961179567250267e-05,
-      "loss": 0.4768,
+      "epoch": 0.37759458265861917,
+      "grad_norm": 0.4557070379670827,
+      "learning_rate": 1.995928541277305e-05,
+      "loss": 0.4542,
       "step": 2565
     },
     {
-      "epoch": 0.37655677655677655,
-      "grad_norm": 0.4218728291102029,
-      "learning_rate": 1.996042555557744e-05,
-      "loss": 0.464,
+      "epoch": 0.37833063447666715,
+      "grad_norm": 0.4463535286579343,
+      "learning_rate": 1.995850971221789e-05,
+      "loss": 0.4859,
       "step": 2570
     },
     {
-      "epoch": 0.3772893772893773,
-      "grad_norm": 0.43312052410700363,
-      "learning_rate": 1.9959664306075084e-05,
-      "loss": 0.4666,
+      "epoch": 0.3790666862947151,
+      "grad_norm": 0.4545292326370332,
+      "learning_rate": 1.9957726707203706e-05,
+      "loss": 0.4818,
       "step": 2575
     },
     {
-      "epoch": 0.378021978021978,
-      "grad_norm": 0.4158441193547152,
-      "learning_rate": 1.9958895819296368e-05,
-      "loss": 0.4506,
+      "epoch": 0.37980273811276316,
+      "grad_norm": 0.48012598712851023,
+      "learning_rate": 1.995693639830483e-05,
+      "loss": 0.4767,
       "step": 2580
     },
     {
-      "epoch": 0.37875457875457874,
-      "grad_norm": 0.45065981507325853,
-      "learning_rate": 1.9958120095799724e-05,
-      "loss": 0.4836,
+      "epoch": 0.38053878993081114,
+      "grad_norm": 0.4360662866815463,
+      "learning_rate": 1.995613878610094e-05,
+      "loss": 0.4742,
       "step": 2585
     },
     {
-      "epoch": 0.37948717948717947,
-      "grad_norm": 0.42564408981825813,
-      "learning_rate": 1.995733713614883e-05,
-      "loss": 0.4658,
+      "epoch": 0.3812748417488591,
+      "grad_norm": 0.4766069613257336,
+      "learning_rate": 1.995533387117707e-05,
+      "loss": 0.4757,
       "step": 2590
     },
     {
-      "epoch": 0.3802197802197802,
-      "grad_norm": 0.4214338330830984,
-      "learning_rate": 1.9956546940912638e-05,
-      "loss": 0.4671,
+      "epoch": 0.3820108935669071,
+      "grad_norm": 0.428997216193116,
+      "learning_rate": 1.9954521654123627e-05,
+      "loss": 0.471,
       "step": 2595
     },
     {
-      "epoch": 0.38095238095238093,
-      "grad_norm": 0.4304888906650203,
-      "learning_rate": 1.995574951066535e-05,
-      "loss": 0.4786,
+      "epoch": 0.38274694538495513,
+      "grad_norm": 0.4214116969762896,
+      "learning_rate": 1.995370213553636e-05,
+      "loss": 0.4641,
       "step": 2600
     },
     {
-      "epoch": 0.38168498168498166,
-      "grad_norm": 0.412768839745328,
-      "learning_rate": 1.9954944845986416e-05,
-      "loss": 0.4554,
+      "epoch": 0.3834829972030031,
+      "grad_norm": 0.41477266318843403,
+      "learning_rate": 1.9952875316016376e-05,
+      "loss": 0.4804,
       "step": 2605
     },
     {
-      "epoch": 0.3824175824175824,
-      "grad_norm": 0.41692261832569244,
-      "learning_rate": 1.9954132947460562e-05,
-      "loss": 0.4724,
+      "epoch": 0.3842190490210511,
+      "grad_norm": 0.4232486648727509,
+      "learning_rate": 1.995204119617014e-05,
+      "loss": 0.4739,
       "step": 2610
     },
     {
-      "epoch": 0.3831501831501832,
-      "grad_norm": 0.42952204546591416,
-      "learning_rate": 1.9953313815677763e-05,
-      "loss": 0.462,
+      "epoch": 0.38495510083909906,
+      "grad_norm": 0.42908297616282554,
+      "learning_rate": 1.995119977660947e-05,
+      "loss": 0.4659,
       "step": 2615
     },
     {
-      "epoch": 0.3838827838827839,
-      "grad_norm": 0.45479760018622895,
-      "learning_rate": 1.995248745123324e-05,
-      "loss": 0.4895,
+      "epoch": 0.38569115265714704,
+      "grad_norm": 0.4367130508945953,
+      "learning_rate": 1.995035105795153e-05,
+      "loss": 0.4646,
       "step": 2620
     },
     {
-      "epoch": 0.38461538461538464,
-      "grad_norm": 0.4106595241460115,
-      "learning_rate": 1.995165385472748e-05,
-      "loss": 0.4516,
+      "epoch": 0.3864272044751951,
+      "grad_norm": 0.43916168441046405,
+      "learning_rate": 1.994949504081886e-05,
+      "loss": 0.4645,
       "step": 2625
     },
     {
-      "epoch": 0.38534798534798537,
-      "grad_norm": 0.46724827471487923,
-      "learning_rate": 1.995081302676623e-05,
-      "loss": 0.4676,
+      "epoch": 0.38716325629324305,
+      "grad_norm": 0.4179134009051416,
+      "learning_rate": 1.9948631725839326e-05,
+      "loss": 0.4818,
       "step": 2630
     },
     {
-      "epoch": 0.3860805860805861,
-      "grad_norm": 0.43642050328749066,
-      "learning_rate": 1.9949964967960472e-05,
-      "loss": 0.4691,
+      "epoch": 0.38789930811129103,
+      "grad_norm": 0.42355447856295925,
+      "learning_rate": 1.9947761113646166e-05,
+      "loss": 0.4632,
       "step": 2635
     },
     {
-      "epoch": 0.3868131868131868,
-      "grad_norm": 0.4342906280447318,
-      "learning_rate": 1.9949109678926472e-05,
-      "loss": 0.5037,
+      "epoch": 0.388635359929339,
+      "grad_norm": 0.4251933072853647,
+      "learning_rate": 1.994688320487797e-05,
+      "loss": 0.4729,
       "step": 2640
     },
     {
-      "epoch": 0.38754578754578756,
-      "grad_norm": 0.4545020175156162,
-      "learning_rate": 1.9948247160285717e-05,
-      "loss": 0.4757,
+      "epoch": 0.38937141174738704,
+      "grad_norm": 0.4467923689627895,
+      "learning_rate": 1.9945998000178663e-05,
+      "loss": 0.4616,
       "step": 2645
     },
     {
-      "epoch": 0.3882783882783883,
-      "grad_norm": 0.4375133844709041,
-      "learning_rate": 1.9947377412664974e-05,
-      "loss": 0.4713,
+      "epoch": 0.390107463565435,
+      "grad_norm": 0.4491808533304436,
+      "learning_rate": 1.994510550019754e-05,
+      "loss": 0.4976,
       "step": 2650
     },
     {
-      "epoch": 0.389010989010989,
-      "grad_norm": 0.4448354117733457,
-      "learning_rate": 1.9946500436696248e-05,
-      "loss": 0.4697,
+      "epoch": 0.390843515383483,
+      "grad_norm": 0.42429793760927365,
+      "learning_rate": 1.994420570558924e-05,
+      "loss": 0.4673,
       "step": 2655
     },
     {
-      "epoch": 0.38974358974358975,
-      "grad_norm": 0.42405644917844576,
-      "learning_rate": 1.9945616233016803e-05,
-      "loss": 0.4746,
+      "epoch": 0.391579567201531,
+      "grad_norm": 0.4322490576948786,
+      "learning_rate": 1.9943298617013755e-05,
+      "loss": 0.4916,
       "step": 2660
     },
     {
-      "epoch": 0.3904761904761905,
-      "grad_norm": 0.43918153514576697,
-      "learning_rate": 1.9944724802269155e-05,
-      "loss": 0.4507,
+      "epoch": 0.39231561901957895,
+      "grad_norm": 0.42197433316013055,
+      "learning_rate": 1.9942384235136418e-05,
+      "loss": 0.4707,
       "step": 2665
     },
     {
-      "epoch": 0.3912087912087912,
-      "grad_norm": 0.437319551455226,
-      "learning_rate": 1.9943826145101065e-05,
-      "loss": 0.4876,
+      "epoch": 0.393051670837627,
+      "grad_norm": 0.4273479256117159,
+      "learning_rate": 1.9941462560627924e-05,
+      "loss": 0.4591,
       "step": 2670
     },
     {
-      "epoch": 0.39194139194139194,
-      "grad_norm": 0.4428182093921229,
-      "learning_rate": 1.9942920262165554e-05,
-      "loss": 0.4669,
+      "epoch": 0.39378772265567497,
+      "grad_norm": 0.4352470069161979,
+      "learning_rate": 1.994053359416431e-05,
+      "loss": 0.4788,
       "step": 2675
     },
     {
-      "epoch": 0.39267399267399267,
-      "grad_norm": 0.4498586125749959,
-      "learning_rate": 1.994200715412089e-05,
-      "loss": 0.457,
+      "epoch": 0.39452377447372294,
+      "grad_norm": 0.42524739307120096,
+      "learning_rate": 1.9939597336426958e-05,
+      "loss": 0.4708,
       "step": 2680
     },
     {
-      "epoch": 0.3934065934065934,
-      "grad_norm": 0.44351385651904046,
-      "learning_rate": 1.9941086821630585e-05,
-      "loss": 0.4838,
+      "epoch": 0.3952598262917709,
+      "grad_norm": 0.4778174570858514,
+      "learning_rate": 1.993865378810261e-05,
+      "loss": 0.4737,
       "step": 2685
     },
     {
-      "epoch": 0.3941391941391941,
-      "grad_norm": 0.4078717462375167,
-      "learning_rate": 1.9940159265363415e-05,
-      "loss": 0.4566,
+      "epoch": 0.39599587810981896,
+      "grad_norm": 0.43701707594895806,
+      "learning_rate": 1.9937702949883343e-05,
+      "loss": 0.454,
       "step": 2690
     },
     {
-      "epoch": 0.39487179487179486,
-      "grad_norm": 0.4011687882917101,
-      "learning_rate": 1.9939224485993393e-05,
-      "loss": 0.4543,
+      "epoch": 0.39673192992786693,
+      "grad_norm": 0.4408193981864741,
+      "learning_rate": 1.993674482246659e-05,
+      "loss": 0.4941,
       "step": 2695
     },
     {
-      "epoch": 0.3956043956043956,
-      "grad_norm": 0.42486716225410437,
-      "learning_rate": 1.993828248419978e-05,
-      "loss": 0.4583,
+      "epoch": 0.3974679817459149,
+      "grad_norm": 0.44887014268199094,
+      "learning_rate": 1.9935779406555126e-05,
+      "loss": 0.4811,
       "step": 2700
     },
     {
-      "epoch": 0.3963369963369963,
-      "grad_norm": 0.42652221818799463,
-      "learning_rate": 1.9937333260667095e-05,
-      "loss": 0.4601,
+      "epoch": 0.3982040335639629,
+      "grad_norm": 0.4313218301535768,
+      "learning_rate": 1.993480670285707e-05,
+      "loss": 0.4619,
       "step": 2705
     },
     {
-      "epoch": 0.39706959706959705,
-      "grad_norm": 0.4768230911766403,
-      "learning_rate": 1.99363768160851e-05,
-      "loss": 0.4732,
+      "epoch": 0.39894008538201087,
+      "grad_norm": 0.4373516450492796,
+      "learning_rate": 1.9933826712085898e-05,
+      "loss": 0.4689,
       "step": 2710
     },
     {
-      "epoch": 0.3978021978021978,
-      "grad_norm": 0.4351470474079274,
-      "learning_rate": 1.9935413151148802e-05,
-      "loss": 0.4642,
+      "epoch": 0.3996761372000589,
+      "grad_norm": 0.39934662785688674,
+      "learning_rate": 1.9932839434960414e-05,
+      "loss": 0.4729,
       "step": 2715
     },
     {
-      "epoch": 0.39853479853479856,
-      "grad_norm": 0.430677839074461,
-      "learning_rate": 1.993444226655846e-05,
-      "loss": 0.4647,
+      "epoch": 0.4004121890181069,
+      "grad_norm": 0.41606780448949654,
+      "learning_rate": 1.993184487220478e-05,
+      "loss": 0.4635,
       "step": 2720
     },
     {
-      "epoch": 0.3992673992673993,
-      "grad_norm": 0.4200612552021059,
-      "learning_rate": 1.9933464163019568e-05,
-      "loss": 0.472,
+      "epoch": 0.40114824083615486,
+      "grad_norm": 0.46256374395781535,
+      "learning_rate": 1.993084302454849e-05,
+      "loss": 0.4734,
       "step": 2725
     },
     {
-      "epoch": 0.4,
-      "grad_norm": 0.4526322616576567,
-      "learning_rate": 1.993247884124288e-05,
-      "loss": 0.4772,
+      "epoch": 0.40188429265420283,
+      "grad_norm": 0.4197100615354211,
+      "learning_rate": 1.9929833892726398e-05,
+      "loss": 0.4627,
       "step": 2730
     },
     {
-      "epoch": 0.40073260073260075,
-      "grad_norm": 0.42669669732174814,
-      "learning_rate": 1.9931486301944388e-05,
-      "loss": 0.4702,
+      "epoch": 0.40262034447225087,
+      "grad_norm": 0.4379080643054496,
+      "learning_rate": 1.9928817477478687e-05,
+      "loss": 0.4611,
       "step": 2735
     },
     {
-      "epoch": 0.4014652014652015,
-      "grad_norm": 0.4315295987179508,
-      "learning_rate": 1.9930486545845324e-05,
-      "loss": 0.4705,
+      "epoch": 0.40335639629029885,
+      "grad_norm": 0.4084960104593173,
+      "learning_rate": 1.9927793779550886e-05,
+      "loss": 0.4558,
       "step": 2740
     },
     {
-      "epoch": 0.4021978021978022,
-      "grad_norm": 0.5545099109486473,
-      "learning_rate": 1.992947957367217e-05,
-      "loss": 0.4567,
+      "epoch": 0.4040924481083468,
+      "grad_norm": 0.42345583706123907,
+      "learning_rate": 1.9926762799693865e-05,
+      "loss": 0.4749,
       "step": 2745
     },
     {
-      "epoch": 0.40293040293040294,
-      "grad_norm": 0.47415426519542037,
-      "learning_rate": 1.992846538615666e-05,
-      "loss": 0.4576,
+      "epoch": 0.4048284999263948,
+      "grad_norm": 0.40237746882884434,
+      "learning_rate": 1.992572453866384e-05,
+      "loss": 0.4521,
       "step": 2750
     },
     {
-      "epoch": 0.40366300366300367,
-      "grad_norm": 0.4510239651090376,
-      "learning_rate": 1.992744398403576e-05,
-      "loss": 0.4796,
+      "epoch": 0.40556455174444284,
+      "grad_norm": 0.43271978633491137,
+      "learning_rate": 1.9924678997222364e-05,
+      "loss": 0.4746,
       "step": 2755
     },
     {
-      "epoch": 0.4043956043956044,
-      "grad_norm": 0.419403060450085,
-      "learning_rate": 1.992641536805167e-05,
-      "loss": 0.4595,
+      "epoch": 0.4063006035624908,
+      "grad_norm": 0.4181052235978989,
+      "learning_rate": 1.992362617613633e-05,
+      "loss": 0.4798,
       "step": 2760
     },
     {
-      "epoch": 0.40512820512820513,
-      "grad_norm": 0.46325220121502125,
-      "learning_rate": 1.992537953895185e-05,
-      "loss": 0.4616,
+      "epoch": 0.4070366553805388,
+      "grad_norm": 0.4524920937481242,
+      "learning_rate": 1.9922566076177964e-05,
+      "loss": 0.4703,
       "step": 2765
     },
     {
-      "epoch": 0.40586080586080586,
-      "grad_norm": 0.41964023342762674,
-      "learning_rate": 1.9924336497489e-05,
-      "loss": 0.4749,
+      "epoch": 0.40777270719858677,
+      "grad_norm": 0.42518121450361884,
+      "learning_rate": 1.9921498698124847e-05,
+      "loss": 0.4523,
       "step": 2770
     },
     {
-      "epoch": 0.4065934065934066,
-      "grad_norm": 0.4140452604597255,
-      "learning_rate": 1.9923286244421045e-05,
-      "loss": 0.4766,
+      "epoch": 0.40850875901663475,
+      "grad_norm": 0.42126470170814195,
+      "learning_rate": 1.992042404275989e-05,
+      "loss": 0.4755,
       "step": 2775
     },
     {
-      "epoch": 0.4073260073260073,
-      "grad_norm": 0.4089490254304085,
-      "learning_rate": 1.9922228780511164e-05,
-      "loss": 0.4761,
+      "epoch": 0.4092448108346828,
+      "grad_norm": 0.44297126766174794,
+      "learning_rate": 1.9919342110871333e-05,
+      "loss": 0.4797,
       "step": 2780
     },
     {
-      "epoch": 0.40805860805860805,
-      "grad_norm": 0.42570100124391697,
-      "learning_rate": 1.9921164106527773e-05,
-      "loss": 0.4649,
+      "epoch": 0.40998086265273076,
+      "grad_norm": 0.4713457406940533,
+      "learning_rate": 1.9918252903252764e-05,
+      "loss": 0.481,
       "step": 2785
     },
     {
-      "epoch": 0.4087912087912088,
-      "grad_norm": 0.42213967282368353,
-      "learning_rate": 1.9920092223244526e-05,
-      "loss": 0.4545,
+      "epoch": 0.41071691447077874,
+      "grad_norm": 0.40506369456948327,
+      "learning_rate": 1.991715642070311e-05,
+      "loss": 0.4617,
       "step": 2790
     },
     {
-      "epoch": 0.4095238095238095,
-      "grad_norm": 0.4229435158499598,
-      "learning_rate": 1.991901313144032e-05,
-      "loss": 0.4672,
+      "epoch": 0.4114529662888267,
+      "grad_norm": 0.41396699426312916,
+      "learning_rate": 1.9916052664026623e-05,
+      "loss": 0.4559,
       "step": 2795
     },
     {
-      "epoch": 0.41025641025641024,
-      "grad_norm": 0.41948017040271546,
-      "learning_rate": 1.9917926831899283e-05,
-      "loss": 0.444,
+      "epoch": 0.41218901810687475,
+      "grad_norm": 0.41040000834914764,
+      "learning_rate": 1.99149416340329e-05,
+      "loss": 0.4859,
       "step": 2800
     },
     {
-      "epoch": 0.41098901098901097,
-      "grad_norm": 0.42482483007703387,
-      "learning_rate": 1.991683332541078e-05,
-      "loss": 0.4749,
+      "epoch": 0.4129250699249227,
+      "grad_norm": 0.4445291093214514,
+      "learning_rate": 1.9913823331536866e-05,
+      "loss": 0.4841,
       "step": 2805
     },
     {
-      "epoch": 0.4117216117216117,
-      "grad_norm": 0.40192909545353095,
-      "learning_rate": 1.9915732612769422e-05,
-      "loss": 0.4655,
+      "epoch": 0.4136611217429707,
+      "grad_norm": 0.48970415093420555,
+      "learning_rate": 1.9912697757358787e-05,
+      "loss": 0.4753,
       "step": 2810
     },
     {
-      "epoch": 0.41245421245421243,
-      "grad_norm": 0.5431509754402679,
-      "learning_rate": 1.9914624694775055e-05,
-      "loss": 0.4629,
+      "epoch": 0.4143971735610187,
+      "grad_norm": 0.4328230899901993,
+      "learning_rate": 1.9911564912324262e-05,
+      "loss": 0.4745,
       "step": 2815
     },
     {
-      "epoch": 0.41318681318681316,
-      "grad_norm": 0.4382504187524325,
-      "learning_rate": 1.9913509572232745e-05,
-      "loss": 0.4775,
+      "epoch": 0.41513322537906666,
+      "grad_norm": 0.4290769164771048,
+      "learning_rate": 1.9910424797264212e-05,
+      "loss": 0.4549,
       "step": 2820
     },
     {
-      "epoch": 0.4139194139194139,
-      "grad_norm": 0.446368115249646,
-      "learning_rate": 1.991238724595282e-05,
-      "loss": 0.4708,
+      "epoch": 0.4158692771971147,
+      "grad_norm": 0.4464613687929015,
+      "learning_rate": 1.9909277413014908e-05,
+      "loss": 0.4618,
       "step": 2825
     },
     {
-      "epoch": 0.4146520146520147,
-      "grad_norm": 0.4597753061899568,
-      "learning_rate": 1.9911257716750816e-05,
-      "loss": 0.4776,
+      "epoch": 0.4166053290151627,
+      "grad_norm": 0.7532038501882797,
+      "learning_rate": 1.990812276041794e-05,
+      "loss": 0.4695,
       "step": 2830
     },
     {
-      "epoch": 0.4153846153846154,
-      "grad_norm": 0.4151626162955841,
-      "learning_rate": 1.9910120985447525e-05,
-      "loss": 0.4601,
+      "epoch": 0.41734138083321065,
+      "grad_norm": 0.4010934042836865,
+      "learning_rate": 1.990696084032023e-05,
+      "loss": 0.4733,
       "step": 2835
     },
     {
-      "epoch": 0.41611721611721614,
-      "grad_norm": 0.46047399733159433,
-      "learning_rate": 1.9908977052868956e-05,
-      "loss": 0.485,
+      "epoch": 0.41807743265125863,
+      "grad_norm": 0.48258845578057696,
+      "learning_rate": 1.990579165357404e-05,
+      "loss": 0.4714,
       "step": 2840
     },
     {
-      "epoch": 0.41684981684981687,
-      "grad_norm": 0.4241080283573777,
-      "learning_rate": 1.990782591984636e-05,
-      "loss": 0.466,
+      "epoch": 0.41881348446930666,
+      "grad_norm": 0.44701405970311925,
+      "learning_rate": 1.9904615201036958e-05,
+      "loss": 0.4538,
       "step": 2845
     },
     {
-      "epoch": 0.4175824175824176,
-      "grad_norm": 0.4809486308461704,
-      "learning_rate": 1.9906667587216216e-05,
-      "loss": 0.4649,
+      "epoch": 0.41954953628735464,
+      "grad_norm": 0.4163801328729629,
+      "learning_rate": 1.9903431483571887e-05,
+      "loss": 0.4654,
       "step": 2850
     },
     {
-      "epoch": 0.4183150183150183,
-      "grad_norm": 0.4281122076011044,
-      "learning_rate": 1.9905502055820237e-05,
-      "loss": 0.4724,
+      "epoch": 0.4202855881054026,
+      "grad_norm": 0.43828348076014584,
+      "learning_rate": 1.990224050204708e-05,
+      "loss": 0.4783,
       "step": 2855
     },
     {
-      "epoch": 0.41904761904761906,
-      "grad_norm": 0.4264996088139275,
-      "learning_rate": 1.990432932650537e-05,
-      "loss": 0.48,
+      "epoch": 0.4210216399234506,
+      "grad_norm": 0.4304881069425013,
+      "learning_rate": 1.9901042257336107e-05,
+      "loss": 0.4826,
       "step": 2860
     },
     {
-      "epoch": 0.4197802197802198,
-      "grad_norm": 0.4300567168814666,
-      "learning_rate": 1.990314940012379e-05,
-      "loss": 0.4689,
+      "epoch": 0.4217576917414986,
+      "grad_norm": 0.45782879456871134,
+      "learning_rate": 1.9899836750317865e-05,
+      "loss": 0.4798,
       "step": 2865
     },
     {
-      "epoch": 0.4205128205128205,
-      "grad_norm": 0.4480456883196755,
-      "learning_rate": 1.990196227753289e-05,
-      "loss": 0.4722,
+      "epoch": 0.4224937435595466,
+      "grad_norm": 0.43195005386091423,
+      "learning_rate": 1.9898623981876587e-05,
+      "loss": 0.4643,
       "step": 2870
     },
     {
-      "epoch": 0.42124542124542125,
-      "grad_norm": 0.4322204562192915,
-      "learning_rate": 1.9900767959595312e-05,
-      "loss": 0.5072,
+      "epoch": 0.4232297953775946,
+      "grad_norm": 0.45983030843494205,
+      "learning_rate": 1.9897403952901822e-05,
+      "loss": 0.4711,
       "step": 2875
     },
     {
-      "epoch": 0.421978021978022,
-      "grad_norm": 0.4050762891489174,
-      "learning_rate": 1.9899566447178914e-05,
-      "loss": 0.4644,
+      "epoch": 0.42396584719564256,
+      "grad_norm": 0.43173486287916685,
+      "learning_rate": 1.9896176664288447e-05,
+      "loss": 0.4624,
       "step": 2880
     },
     {
-      "epoch": 0.4227106227106227,
-      "grad_norm": 0.4381041319495102,
-      "learning_rate": 1.9898357741156788e-05,
-      "loss": 0.4705,
+      "epoch": 0.42470189901369054,
+      "grad_norm": 0.44218985490594753,
+      "learning_rate": 1.9894942116936666e-05,
+      "loss": 0.4754,
       "step": 2885
     },
     {
-      "epoch": 0.42344322344322344,
-      "grad_norm": 0.41659237129980764,
-      "learning_rate": 1.989714184240725e-05,
-      "loss": 0.4513,
+      "epoch": 0.4254379508317386,
+      "grad_norm": 0.3865084959037182,
+      "learning_rate": 1.9893700311752003e-05,
+      "loss": 0.4492,
       "step": 2890
     },
     {
-      "epoch": 0.42417582417582417,
-      "grad_norm": 0.43285927087871323,
-      "learning_rate": 1.9895918751813843e-05,
-      "loss": 0.4599,
+      "epoch": 0.42617400264978655,
+      "grad_norm": 0.41897147151455905,
+      "learning_rate": 1.989245124964531e-05,
+      "loss": 0.4541,
       "step": 2895
     },
     {
-      "epoch": 0.4249084249084249,
-      "grad_norm": 1.1240756278162536,
-      "learning_rate": 1.9894688470265335e-05,
-      "loss": 0.475,
+      "epoch": 0.42691005446783453,
+      "grad_norm": 0.43255821511679104,
+      "learning_rate": 1.9891194931532768e-05,
+      "loss": 0.4645,
       "step": 2900
     },
     {
-      "epoch": 0.4256410256410256,
-      "grad_norm": 0.43100131199062985,
-      "learning_rate": 1.9893450998655723e-05,
-      "loss": 0.4781,
+      "epoch": 0.4276461062858825,
+      "grad_norm": 0.44304923672780167,
+      "learning_rate": 1.988993135833586e-05,
+      "loss": 0.4695,
       "step": 2905
     },
     {
-      "epoch": 0.42637362637362636,
-      "grad_norm": 0.42432854968802913,
-      "learning_rate": 1.9892206337884226e-05,
-      "loss": 0.4711,
+      "epoch": 0.42838215810393054,
+      "grad_norm": 0.41268218699985837,
+      "learning_rate": 1.9888660530981413e-05,
+      "loss": 0.4635,
       "step": 2910
     },
     {
-      "epoch": 0.4271062271062271,
-      "grad_norm": 0.4479110111360452,
-      "learning_rate": 1.9890954488855288e-05,
-      "loss": 0.4697,
+      "epoch": 0.4291182099219785,
+      "grad_norm": 0.4377162556413914,
+      "learning_rate": 1.9887382450401563e-05,
+      "loss": 0.4784,
       "step": 2915
     },
     {
-      "epoch": 0.4278388278388278,
-      "grad_norm": 0.44549474371119435,
-      "learning_rate": 1.9889695452478573e-05,
-      "loss": 0.4618,
+      "epoch": 0.4298542617400265,
+      "grad_norm": 0.4318335871287204,
+      "learning_rate": 1.9886097117533764e-05,
+      "loss": 0.4758,
       "step": 2920
     },
     {
-      "epoch": 0.42857142857142855,
-      "grad_norm": 0.42161761546436355,
-      "learning_rate": 1.9888429229668972e-05,
-      "loss": 0.4613,
+      "epoch": 0.4305903135580745,
+      "grad_norm": 0.4305024561550163,
+      "learning_rate": 1.98848045333208e-05,
+      "loss": 0.4675,
       "step": 2925
     },
     {
-      "epoch": 0.4293040293040293,
-      "grad_norm": 0.4215544190235201,
-      "learning_rate": 1.9887155821346597e-05,
-      "loss": 0.4508,
+      "epoch": 0.43132636537612246,
+      "grad_norm": 0.44025037364780273,
+      "learning_rate": 1.9883504698710764e-05,
+      "loss": 0.4566,
       "step": 2930
     },
     {
-      "epoch": 0.43003663003663006,
-      "grad_norm": 0.4423258878717002,
-      "learning_rate": 1.988587522843678e-05,
-      "loss": 0.4674,
+      "epoch": 0.4320624171941705,
+      "grad_norm": 0.44310453288236495,
+      "learning_rate": 1.9882197614657067e-05,
+      "loss": 0.4607,
       "step": 2935
     },
     {
-      "epoch": 0.4307692307692308,
-      "grad_norm": 0.42574662815844067,
-      "learning_rate": 1.9884587451870082e-05,
-      "loss": 0.4596,
+      "epoch": 0.43279846901221847,
+      "grad_norm": 0.4320374381203412,
+      "learning_rate": 1.988088328211845e-05,
+      "loss": 0.4703,
       "step": 2940
     },
     {
-      "epoch": 0.4315018315018315,
-      "grad_norm": 0.4170768435563345,
-      "learning_rate": 1.9883292492582268e-05,
-      "loss": 0.4608,
+      "epoch": 0.43353452083026645,
+      "grad_norm": 0.4360481931394084,
+      "learning_rate": 1.9879561702058954e-05,
+      "loss": 0.4709,
       "step": 2945
     },
     {
-      "epoch": 0.43223443223443225,
-      "grad_norm": 0.44697757313895964,
-      "learning_rate": 1.9881990351514333e-05,
-      "loss": 0.4673,
+      "epoch": 0.4342705726483144,
+      "grad_norm": 0.3961083811911451,
+      "learning_rate": 1.9878232875447948e-05,
+      "loss": 0.4668,
       "step": 2950
     },
     {
-      "epoch": 0.432967032967033,
-      "grad_norm": 0.41995177242580917,
-      "learning_rate": 1.9880681029612494e-05,
-      "loss": 0.468,
+      "epoch": 0.43500662446636246,
+      "grad_norm": 0.4331840104260549,
+      "learning_rate": 1.987689680326011e-05,
+      "loss": 0.4605,
       "step": 2955
     },
     {
-      "epoch": 0.4336996336996337,
-      "grad_norm": 0.43402289944455125,
-      "learning_rate": 1.9879364527828178e-05,
-      "loss": 0.4517,
+      "epoch": 0.43574267628441044,
+      "grad_norm": 0.43775242213997434,
+      "learning_rate": 1.9875553486475434e-05,
+      "loss": 0.4652,
       "step": 2960
     },
     {
-      "epoch": 0.43443223443223444,
-      "grad_norm": 0.46212386496174346,
-      "learning_rate": 1.987804084711803e-05,
-      "loss": 0.4849,
+      "epoch": 0.4364787281024584,
+      "grad_norm": 0.39381836032976314,
+      "learning_rate": 1.9874202926079226e-05,
+      "loss": 0.4509,
       "step": 2965
     },
     {
-      "epoch": 0.4351648351648352,
-      "grad_norm": 0.5112733415377284,
-      "learning_rate": 1.987670998844392e-05,
-      "loss": 0.4721,
+      "epoch": 0.4372147799205064,
+      "grad_norm": 0.46469421855299614,
+      "learning_rate": 1.9872845123062117e-05,
+      "loss": 0.4649,
       "step": 2970
     },
     {
-      "epoch": 0.4358974358974359,
-      "grad_norm": 0.4285494192095629,
-      "learning_rate": 1.9875371952772922e-05,
-      "loss": 0.4541,
+      "epoch": 0.43795083173855437,
+      "grad_norm": 0.43420109025858905,
+      "learning_rate": 1.987148007842003e-05,
+      "loss": 0.4707,
       "step": 2975
     },
     {
-      "epoch": 0.43663003663003663,
-      "grad_norm": 0.43764286187833223,
-      "learning_rate": 1.9874026741077336e-05,
-      "loss": 0.4777,
+      "epoch": 0.4386868835566024,
+      "grad_norm": 0.41017038513244725,
+      "learning_rate": 1.9870107793154215e-05,
+      "loss": 0.4659,
       "step": 2980
     },
     {
-      "epoch": 0.43736263736263736,
-      "grad_norm": 0.41258038663230506,
-      "learning_rate": 1.9872674354334667e-05,
-      "loss": 0.4538,
+      "epoch": 0.4394229353746504,
+      "grad_norm": 0.42378395928333984,
+      "learning_rate": 1.986872826827123e-05,
+      "loss": 0.4779,
       "step": 2985
     },
     {
-      "epoch": 0.4380952380952381,
-      "grad_norm": 0.4271331651523818,
-      "learning_rate": 1.987131479352764e-05,
-      "loss": 0.475,
+      "epoch": 0.44015898719269836,
+      "grad_norm": 0.41958986445827784,
+      "learning_rate": 1.9867341504782936e-05,
+      "loss": 0.4555,
       "step": 2990
     },
     {
-      "epoch": 0.4388278388278388,
-      "grad_norm": 0.4006631405996531,
-      "learning_rate": 1.9869948059644197e-05,
-      "loss": 0.4584,
+      "epoch": 0.44089503901074634,
+      "grad_norm": 0.44111437989525354,
+      "learning_rate": 1.9865947503706517e-05,
+      "loss": 0.4724,
       "step": 2995
     },
     {
-      "epoch": 0.43956043956043955,
-      "grad_norm": 0.44954746312551097,
-      "learning_rate": 1.986857415367748e-05,
-      "loss": 0.4723,
+      "epoch": 0.44163109082879437,
+      "grad_norm": 0.43488335562692976,
+      "learning_rate": 1.986454626606445e-05,
+      "loss": 0.4525,
       "step": 3000
     },
     {
-      "epoch": 0.4402930402930403,
-      "grad_norm": 0.4097449870862522,
-      "learning_rate": 1.9867193076625852e-05,
-      "loss": 0.4599,
+      "epoch": 0.44236714264684235,
+      "grad_norm": 0.4334718080918571,
+      "learning_rate": 1.9863137792884534e-05,
+      "loss": 0.464,
       "step": 3005
     },
     {
-      "epoch": 0.441025641025641,
-      "grad_norm": 0.4421570174944609,
-      "learning_rate": 1.9865804829492883e-05,
-      "loss": 0.4506,
+      "epoch": 0.4431031944648903,
+      "grad_norm": 0.40573764527765027,
+      "learning_rate": 1.9861722085199866e-05,
+      "loss": 0.474,
       "step": 3010
     },
     {
-      "epoch": 0.44175824175824174,
-      "grad_norm": 0.4265771508229422,
-      "learning_rate": 1.986440941328736e-05,
-      "loss": 0.4611,
+      "epoch": 0.4438392462829383,
+      "grad_norm": 0.416306792206869,
+      "learning_rate": 1.986029914404885e-05,
+      "loss": 0.4616,
       "step": 3015
     },
     {
-      "epoch": 0.4424908424908425,
-      "grad_norm": 0.4185455430733579,
-      "learning_rate": 1.986300682902327e-05,
-      "loss": 0.4593,
+      "epoch": 0.4445752981009863,
+      "grad_norm": 0.4109439567274265,
+      "learning_rate": 1.9858868970475202e-05,
+      "loss": 0.4683,
       "step": 3020
     },
     {
-      "epoch": 0.4432234432234432,
-      "grad_norm": 0.43628127572576264,
-      "learning_rate": 1.9861597077719818e-05,
-      "loss": 0.5056,
+      "epoch": 0.4453113499190343,
+      "grad_norm": 0.4281915217845688,
+      "learning_rate": 1.9857431565527936e-05,
+      "loss": 0.4473,
       "step": 3025
     },
     {
-      "epoch": 0.44395604395604393,
-      "grad_norm": 0.41242286663396627,
-      "learning_rate": 1.98601801604014e-05,
-      "loss": 0.4594,
+      "epoch": 0.4460474017370823,
+      "grad_norm": 0.3945272194503782,
+      "learning_rate": 1.9855986930261374e-05,
+      "loss": 0.4602,
       "step": 3030
     },
     {
-      "epoch": 0.44468864468864466,
-      "grad_norm": 0.4468118929304783,
-      "learning_rate": 1.9858756078097644e-05,
-      "loss": 0.4787,
+      "epoch": 0.44678345355513027,
+      "grad_norm": 0.44107881912248864,
+      "learning_rate": 1.9854535065735137e-05,
+      "loss": 0.4632,
       "step": 3035
     },
     {
-      "epoch": 0.44542124542124545,
-      "grad_norm": 0.4236707817709085,
-      "learning_rate": 1.985732483184337e-05,
-      "loss": 0.4664,
+      "epoch": 0.44751950537317825,
+      "grad_norm": 0.40426276608883266,
+      "learning_rate": 1.985307597301416e-05,
+      "loss": 0.4739,
       "step": 3040
     },
     {
-      "epoch": 0.4461538461538462,
-      "grad_norm": 0.41568305822108564,
-      "learning_rate": 1.9855886422678598e-05,
-      "loss": 0.4556,
+      "epoch": 0.4482555571912263,
+      "grad_norm": 0.4039687613674645,
+      "learning_rate": 1.9851609653168664e-05,
+      "loss": 0.4626,
       "step": 3045
     },
     {
-      "epoch": 0.4468864468864469,
-      "grad_norm": 0.4031732738635877,
-      "learning_rate": 1.9854440851648566e-05,
-      "loss": 0.4639,
+      "epoch": 0.44899160900927426,
+      "grad_norm": 0.42987297721657564,
+      "learning_rate": 1.9850136107274182e-05,
+      "loss": 0.4722,
       "step": 3050
     },
     {
-      "epoch": 0.44761904761904764,
-      "grad_norm": 0.42845350632694124,
-      "learning_rate": 1.9852988119803708e-05,
-      "loss": 0.4816,
+      "epoch": 0.44972766082732224,
+      "grad_norm": 0.4799009985252744,
+      "learning_rate": 1.984865533641154e-05,
+      "loss": 0.4498,
       "step": 3055
     },
     {
-      "epoch": 0.44835164835164837,
-      "grad_norm": 0.4182498366755355,
-      "learning_rate": 1.985152822819967e-05,
-      "loss": 0.4721,
+      "epoch": 0.4504637126453702,
+      "grad_norm": 0.4261866300237137,
+      "learning_rate": 1.984716734166688e-05,
+      "loss": 0.4703,
       "step": 3060
     },
     {
-      "epoch": 0.4490842490842491,
-      "grad_norm": 2.0996781684996297,
-      "learning_rate": 1.985006117789729e-05,
-      "loss": 0.4682,
+      "epoch": 0.45119976446341825,
+      "grad_norm": 0.41273386694642783,
+      "learning_rate": 1.9845672124131613e-05,
+      "loss": 0.4647,
       "step": 3065
     },
     {
-      "epoch": 0.44981684981684983,
-      "grad_norm": 0.41820013155916524,
-      "learning_rate": 1.9848586969962612e-05,
-      "loss": 0.4709,
+      "epoch": 0.45193581628146623,
+      "grad_norm": 0.44425449264384115,
+      "learning_rate": 1.9844169684902473e-05,
+      "loss": 0.4624,
       "step": 3070
     },
     {
-      "epoch": 0.45054945054945056,
-      "grad_norm": 0.4170880105786003,
-      "learning_rate": 1.984710560546688e-05,
-      "loss": 0.4801,
+      "epoch": 0.4526718680995142,
+      "grad_norm": 0.40456413118707896,
+      "learning_rate": 1.9842660025081485e-05,
+      "loss": 0.456,
       "step": 3075
     },
     {
-      "epoch": 0.4512820512820513,
-      "grad_norm": 0.41308092270816765,
-      "learning_rate": 1.984561708548655e-05,
-      "loss": 0.4724,
+      "epoch": 0.4534079199175622,
+      "grad_norm": 0.40770054840443004,
+      "learning_rate": 1.9841143145775966e-05,
+      "loss": 0.4561,
       "step": 3080
     },
     {
-      "epoch": 0.452014652014652,
-      "grad_norm": 0.42432685224665,
-      "learning_rate": 1.984412141110326e-05,
-      "loss": 0.524,
+      "epoch": 0.45414397173561016,
+      "grad_norm": 0.4342418217242728,
+      "learning_rate": 1.983961904809853e-05,
+      "loss": 0.4774,
       "step": 3085
     },
     {
-      "epoch": 0.45274725274725275,
-      "grad_norm": 0.4335010203313694,
-      "learning_rate": 1.9842618583403857e-05,
-      "loss": 0.4665,
+      "epoch": 0.4548800235536582,
+      "grad_norm": 0.41929000606256295,
+      "learning_rate": 1.9838087733167088e-05,
+      "loss": 0.4647,
       "step": 3090
     },
     {
-      "epoch": 0.4534798534798535,
-      "grad_norm": 0.4126926537895775,
-      "learning_rate": 1.984110860348038e-05,
-      "loss": 0.488,
+      "epoch": 0.4556160753717062,
+      "grad_norm": 0.4004519524235956,
+      "learning_rate": 1.983654920210484e-05,
+      "loss": 0.4636,
       "step": 3095
     },
     {
-      "epoch": 0.4542124542124542,
-      "grad_norm": 0.477986799745578,
-      "learning_rate": 1.9839591472430076e-05,
-      "loss": 0.48,
+      "epoch": 0.45635212718975415,
+      "grad_norm": 0.41124674907739295,
+      "learning_rate": 1.983500345604028e-05,
+      "loss": 0.4787,
       "step": 3100
     },
     {
-      "epoch": 0.45494505494505494,
-      "grad_norm": 0.3938638046525444,
-      "learning_rate": 1.9838067191355377e-05,
-      "loss": 0.4686,
+      "epoch": 0.45708817900780213,
+      "grad_norm": 0.4708532248959944,
+      "learning_rate": 1.9833450496107203e-05,
+      "loss": 0.4767,
       "step": 3105
     },
     {
-      "epoch": 0.45567765567765567,
-      "grad_norm": 0.422316248221422,
-      "learning_rate": 1.9836535761363914e-05,
-      "loss": 0.4792,
+      "epoch": 0.45782423082585016,
+      "grad_norm": 0.458182926506385,
+      "learning_rate": 1.9831890323444686e-05,
+      "loss": 0.461,
       "step": 3110
     },
     {
-      "epoch": 0.4564102564102564,
-      "grad_norm": 0.42758442770500454,
-      "learning_rate": 1.9834997183568517e-05,
-      "loss": 0.4652,
+      "epoch": 0.45856028264389814,
+      "grad_norm": 0.4283861415071619,
+      "learning_rate": 1.9830322939197094e-05,
+      "loss": 0.4612,
       "step": 3115
     },
     {
-      "epoch": 0.45714285714285713,
-      "grad_norm": 0.41505650486941165,
-      "learning_rate": 1.9833451459087207e-05,
-      "loss": 0.4694,
+      "epoch": 0.4592963344619461,
+      "grad_norm": 0.4223663432196678,
+      "learning_rate": 1.9828748344514087e-05,
+      "loss": 0.4561,
       "step": 3120
     },
     {
-      "epoch": 0.45787545787545786,
-      "grad_norm": 0.4313708557250346,
-      "learning_rate": 1.9831898589043195e-05,
-      "loss": 0.4721,
+      "epoch": 0.4600323862799941,
+      "grad_norm": 0.4386331225059637,
+      "learning_rate": 1.9827166540550622e-05,
+      "loss": 0.4732,
       "step": 3125
     },
     {
-      "epoch": 0.4586080586080586,
-      "grad_norm": 0.4039460331210853,
-      "learning_rate": 1.983033857456489e-05,
-      "loss": 0.4745,
+      "epoch": 0.4607684380980421,
+      "grad_norm": 0.4128996398421751,
+      "learning_rate": 1.9825577528466925e-05,
+      "loss": 0.4693,
       "step": 3130
     },
     {
-      "epoch": 0.4593406593406593,
-      "grad_norm": 0.41533105180381913,
-      "learning_rate": 1.982877141678589e-05,
-      "loss": 0.4589,
+      "epoch": 0.4615044899160901,
+      "grad_norm": 0.41584310994050183,
+      "learning_rate": 1.9823981309428525e-05,
+      "loss": 0.4742,
       "step": 3135
     },
     {
-      "epoch": 0.46007326007326005,
-      "grad_norm": 0.40794998171916896,
-      "learning_rate": 1.982719711684498e-05,
-      "loss": 0.4809,
+      "epoch": 0.4622405417341381,
+      "grad_norm": 0.40372530977646753,
+      "learning_rate": 1.9822377884606227e-05,
+      "loss": 0.4657,
       "step": 3140
     },
     {
-      "epoch": 0.4608058608058608,
-      "grad_norm": 0.4437074981748962,
-      "learning_rate": 1.9825615675886142e-05,
-      "loss": 0.4666,
+      "epoch": 0.46297659355218607,
+      "grad_norm": 0.4116788800619764,
+      "learning_rate": 1.982076725517613e-05,
+      "loss": 0.4768,
       "step": 3145
     },
     {
-      "epoch": 0.46153846153846156,
-      "grad_norm": 0.41963545077807324,
-      "learning_rate": 1.9824027095058546e-05,
-      "loss": 0.4577,
+      "epoch": 0.46371264537023404,
+      "grad_norm": 0.4164112439477256,
+      "learning_rate": 1.981914942231961e-05,
+      "loss": 0.4615,
       "step": 3150
     },
     {
-      "epoch": 0.4622710622710623,
-      "grad_norm": 0.41695879640414035,
-      "learning_rate": 1.9822431375516543e-05,
-      "loss": 0.4646,
+      "epoch": 0.4644486971882821,
+      "grad_norm": 0.41310344930399845,
+      "learning_rate": 1.9817524387223335e-05,
+      "loss": 0.4496,
       "step": 3155
     },
     {
-      "epoch": 0.463003663003663,
-      "grad_norm": 0.45304494814251517,
-      "learning_rate": 1.982082851841968e-05,
-      "loss": 0.4647,
+      "epoch": 0.46518474900633006,
+      "grad_norm": 0.39542517683227685,
+      "learning_rate": 1.9815892151079242e-05,
+      "loss": 0.4493,
       "step": 3160
     },
     {
-      "epoch": 0.46373626373626375,
-      "grad_norm": 0.42268156465950907,
-      "learning_rate": 1.9819218524932685e-05,
-      "loss": 0.5243,
+      "epoch": 0.46592080082437803,
+      "grad_norm": 0.39877143833113265,
+      "learning_rate": 1.9814252715084567e-05,
+      "loss": 0.4722,
       "step": 3165
     },
     {
-      "epoch": 0.4644688644688645,
-      "grad_norm": 0.4330357382344699,
-      "learning_rate": 1.981760139622548e-05,
-      "loss": 0.4628,
+      "epoch": 0.466656852642426,
+      "grad_norm": 0.4286432019131809,
+      "learning_rate": 1.981260608044181e-05,
+      "loss": 0.4814,
       "step": 3170
     },
     {
-      "epoch": 0.4652014652014652,
-      "grad_norm": 0.3912091812312883,
-      "learning_rate": 1.9815977133473153e-05,
-      "loss": 0.4655,
+      "epoch": 0.467392904460474,
+      "grad_norm": 0.4228249800132515,
+      "learning_rate": 1.981095224835877e-05,
+      "loss": 0.4496,
       "step": 3175
     },
     {
-      "epoch": 0.46593406593406594,
-      "grad_norm": 0.4117875491925943,
-      "learning_rate": 1.9814345737856e-05,
-      "loss": 0.4706,
+      "epoch": 0.468128956278522,
+      "grad_norm": 0.4002611631369483,
+      "learning_rate": 1.980929122004851e-05,
+      "loss": 0.4708,
       "step": 3180
     },
     {
-      "epoch": 0.4666666666666667,
-      "grad_norm": 0.4194813796894191,
-      "learning_rate": 1.9812707210559482e-05,
-      "loss": 0.4664,
+      "epoch": 0.46886500809657,
+      "grad_norm": 0.4193718325228932,
+      "learning_rate": 1.980762299672937e-05,
+      "loss": 0.4716,
       "step": 3185
     },
     {
-      "epoch": 0.4673992673992674,
-      "grad_norm": 0.44104772544766263,
-      "learning_rate": 1.9811061552774254e-05,
-      "loss": 0.4499,
+      "epoch": 0.469601059914618,
+      "grad_norm": 0.41774347369097076,
+      "learning_rate": 1.9805947579624986e-05,
+      "loss": 0.473,
       "step": 3190
     },
     {
-      "epoch": 0.46813186813186813,
-      "grad_norm": 0.39534418481458466,
-      "learning_rate": 1.980940876569614e-05,
-      "loss": 0.4544,
+      "epoch": 0.47033711173266596,
+      "grad_norm": 0.4180941873133464,
+      "learning_rate": 1.980426496996425e-05,
+      "loss": 0.4617,
       "step": 3195
     },
     {
-      "epoch": 0.46886446886446886,
-      "grad_norm": 0.406074463901355,
-      "learning_rate": 1.980774885052616e-05,
-      "loss": 0.4769,
+      "epoch": 0.471073163550714,
+      "grad_norm": 0.401915602088444,
+      "learning_rate": 1.980257516898134e-05,
+      "loss": 0.4661,
       "step": 3200
     },
     {
-      "epoch": 0.4695970695970696,
-      "grad_norm": 0.44974837591152544,
-      "learning_rate": 1.9806081808470507e-05,
-      "loss": 0.4722,
+      "epoch": 0.47180921536876197,
+      "grad_norm": 0.38887379149008394,
+      "learning_rate": 1.9800878177915705e-05,
+      "loss": 0.456,
       "step": 3205
     },
     {
-      "epoch": 0.4703296703296703,
-      "grad_norm": 0.44829907029092797,
-      "learning_rate": 1.980440764074054e-05,
-      "loss": 0.4734,
+      "epoch": 0.47254526718680995,
+      "grad_norm": 0.4058785283380266,
+      "learning_rate": 1.9799173998012072e-05,
+      "loss": 0.4676,
       "step": 3210
     },
     {
-      "epoch": 0.47106227106227105,
-      "grad_norm": 0.40035861121791594,
-      "learning_rate": 1.980272634855282e-05,
-      "loss": 0.4585,
+      "epoch": 0.4732813190048579,
+      "grad_norm": 0.40844557947363974,
+      "learning_rate": 1.9797462630520437e-05,
+      "loss": 0.464,
       "step": 3215
     },
     {
-      "epoch": 0.4717948717948718,
-      "grad_norm": 0.44003860092573194,
-      "learning_rate": 1.980103793312907e-05,
-      "loss": 0.4666,
+      "epoch": 0.47401737082290596,
+      "grad_norm": 0.4042514657643234,
+      "learning_rate": 1.979574407669607e-05,
+      "loss": 0.4791,
       "step": 3220
     },
     {
-      "epoch": 0.4725274725274725,
-      "grad_norm": 0.7529828149259594,
-      "learning_rate": 1.9799342395696185e-05,
-      "loss": 0.4473,
+      "epoch": 0.47475342264095394,
+      "grad_norm": 0.42841802368536447,
+      "learning_rate": 1.9794018337799517e-05,
+      "loss": 0.476,
       "step": 3225
     },
     {
-      "epoch": 0.47326007326007324,
-      "grad_norm": 0.4151198080685876,
-      "learning_rate": 1.9797639737486246e-05,
-      "loss": 0.4752,
+      "epoch": 0.4754894744590019,
+      "grad_norm": 0.40088970264481133,
+      "learning_rate": 1.9792285415096576e-05,
+      "loss": 0.4481,
       "step": 3230
     },
     {
-      "epoch": 0.473992673992674,
-      "grad_norm": 0.39959614249652226,
-      "learning_rate": 1.9795929959736502e-05,
-      "loss": 0.4599,
+      "epoch": 0.4762255262770499,
+      "grad_norm": 0.40476069582920854,
+      "learning_rate": 1.9790545309858336e-05,
+      "loss": 0.4633,
       "step": 3235
     },
     {
-      "epoch": 0.4747252747252747,
-      "grad_norm": 0.4551376232704039,
-      "learning_rate": 1.9794213063689386e-05,
-      "loss": 0.4625,
+      "epoch": 0.47696157809509787,
+      "grad_norm": 0.4199593822897928,
+      "learning_rate": 1.978879802336115e-05,
+      "loss": 0.468,
       "step": 3240
     },
     {
-      "epoch": 0.47545787545787543,
-      "grad_norm": 0.43462040045084416,
-      "learning_rate": 1.9792489050592486e-05,
-      "loss": 0.4518,
+      "epoch": 0.4776976299131459,
+      "grad_norm": 0.4085284137295676,
+      "learning_rate": 1.9787043556886622e-05,
+      "loss": 0.4604,
       "step": 3245
     },
     {
-      "epoch": 0.47619047619047616,
-      "grad_norm": 0.47841139977456976,
-      "learning_rate": 1.979075792169858e-05,
-      "loss": 0.4641,
+      "epoch": 0.4784336817311939,
+      "grad_norm": 0.44961470043396845,
+      "learning_rate": 1.9785281911721643e-05,
+      "loss": 0.4658,
       "step": 3250
     },
     {
-      "epoch": 0.47692307692307695,
-      "grad_norm": 0.39672143880466443,
-      "learning_rate": 1.9789019678265595e-05,
-      "loss": 0.4545,
+      "epoch": 0.47916973354924186,
+      "grad_norm": 0.4061982836140069,
+      "learning_rate": 1.9783513089158356e-05,
+      "loss": 0.4562,
       "step": 3255
     },
     {
-      "epoch": 0.4776556776556777,
-      "grad_norm": 0.4321837925413481,
-      "learning_rate": 1.978727432155665e-05,
-      "loss": 0.4552,
+      "epoch": 0.47990578536728984,
+      "grad_norm": 0.4195112377167201,
+      "learning_rate": 1.9781737090494176e-05,
+      "loss": 0.4583,
       "step": 3260
     },
     {
-      "epoch": 0.4783882783882784,
-      "grad_norm": 0.4403586310764908,
-      "learning_rate": 1.9785521852840025e-05,
-      "loss": 0.4728,
+      "epoch": 0.4806418371853379,
+      "grad_norm": 0.43364546064085246,
+      "learning_rate": 1.9779953917031776e-05,
+      "loss": 0.4586,
       "step": 3265
     },
     {
-      "epoch": 0.47912087912087914,
-      "grad_norm": 0.40406904762288254,
-      "learning_rate": 1.9783762273389162e-05,
-      "loss": 0.4605,
+      "epoch": 0.48137788900338585,
+      "grad_norm": 0.4115588808981016,
+      "learning_rate": 1.9778163570079095e-05,
+      "loss": 0.4724,
       "step": 3270
     },
     {
-      "epoch": 0.47985347985347987,
-      "grad_norm": 0.4251693515779485,
-      "learning_rate": 1.9781995584482676e-05,
-      "loss": 0.4555,
+      "epoch": 0.48211394082143383,
+      "grad_norm": 0.40816394060622097,
+      "learning_rate": 1.9776366050949333e-05,
+      "loss": 0.4464,
       "step": 3275
     },
     {
-      "epoch": 0.4805860805860806,
-      "grad_norm": 0.41008050156432474,
-      "learning_rate": 1.9780221787404343e-05,
-      "loss": 0.4699,
+      "epoch": 0.4828499926394818,
+      "grad_norm": 0.467843462744796,
+      "learning_rate": 1.9774561360960955e-05,
+      "loss": 0.4725,
       "step": 3280
     },
     {
-      "epoch": 0.48131868131868133,
-      "grad_norm": 0.39802159163681655,
-      "learning_rate": 1.977844088344312e-05,
-      "loss": 0.4763,
+      "epoch": 0.4835860444575298,
+      "grad_norm": 0.41759683676633286,
+      "learning_rate": 1.977274950143768e-05,
+      "loss": 0.4611,
       "step": 3285
     },
     {
-      "epoch": 0.48205128205128206,
-      "grad_norm": 0.4173141548284825,
-      "learning_rate": 1.97766528738931e-05,
-      "loss": 0.4664,
+      "epoch": 0.4843220962755778,
+      "grad_norm": 0.45394616999310433,
+      "learning_rate": 1.977093047370848e-05,
+      "loss": 0.4586,
       "step": 3290
     },
     {
-      "epoch": 0.4827838827838828,
-      "grad_norm": 0.39535352109619026,
-      "learning_rate": 1.977485776005357e-05,
-      "loss": 0.4553,
+      "epoch": 0.4850581480936258,
+      "grad_norm": 0.3971437956948848,
+      "learning_rate": 1.97691042791076e-05,
+      "loss": 0.445,
       "step": 3295
     },
     {
-      "epoch": 0.4835164835164835,
-      "grad_norm": 0.403127280029853,
-      "learning_rate": 1.977305554322895e-05,
-      "loss": 0.4638,
+      "epoch": 0.4857941999116738,
+      "grad_norm": 0.4417562604444158,
+      "learning_rate": 1.9767270918974533e-05,
+      "loss": 0.4631,
       "step": 3300
     },
     {
-      "epoch": 0.48424908424908425,
-      "grad_norm": 0.4315228635100888,
-      "learning_rate": 1.9771246224728853e-05,
-      "loss": 0.4738,
+      "epoch": 0.48653025172972175,
+      "grad_norm": 0.402271545011745,
+      "learning_rate": 1.9765430394654027e-05,
+      "loss": 0.4655,
       "step": 3305
     },
     {
-      "epoch": 0.484981684981685,
-      "grad_norm": 0.44872850655100005,
-      "learning_rate": 1.9769429805868027e-05,
-      "loss": 0.4533,
+      "epoch": 0.4872663035477698,
+      "grad_norm": 0.42477604801229096,
+      "learning_rate": 1.976358270749609e-05,
+      "loss": 0.4688,
       "step": 3310
     },
     {
-      "epoch": 0.4857142857142857,
-      "grad_norm": 0.4079127867380826,
-      "learning_rate": 1.9767606287966384e-05,
-      "loss": 0.445,
+      "epoch": 0.48800235536581776,
+      "grad_norm": 0.4329126934730056,
+      "learning_rate": 1.9761727858855973e-05,
+      "loss": 0.4474,
       "step": 3315
     },
     {
-      "epoch": 0.48644688644688644,
-      "grad_norm": 0.42517621903353997,
-      "learning_rate": 1.9765775672349012e-05,
-      "loss": 0.4701,
+      "epoch": 0.48873840718386574,
+      "grad_norm": 0.43263812510031907,
+      "learning_rate": 1.9759865850094197e-05,
+      "loss": 0.4742,
       "step": 3320
     },
     {
-      "epoch": 0.48717948717948717,
-      "grad_norm": 0.39543291608601666,
-      "learning_rate": 1.9763937960346132e-05,
-      "loss": 0.4561,
+      "epoch": 0.4894744590019137,
+      "grad_norm": 0.41666249042657116,
+      "learning_rate": 1.9757996682576515e-05,
+      "loss": 0.463,
       "step": 3325
     },
     {
-      "epoch": 0.4879120879120879,
-      "grad_norm": 0.43589487175440456,
-      "learning_rate": 1.9762093153293142e-05,
-      "loss": 0.4557,
+      "epoch": 0.4902105108199617,
+      "grad_norm": 0.4265013419951412,
+      "learning_rate": 1.975612035767395e-05,
+      "loss": 0.4717,
       "step": 3330
     },
     {
-      "epoch": 0.48864468864468863,
-      "grad_norm": 0.4129391636519809,
-      "learning_rate": 1.976024125253058e-05,
-      "loss": 0.4765,
+      "epoch": 0.49094656263800973,
+      "grad_norm": 0.4130363534076511,
+      "learning_rate": 1.9754236876762763e-05,
+      "loss": 0.4549,
       "step": 3335
     },
     {
-      "epoch": 0.48937728937728936,
-      "grad_norm": 0.4356046878850061,
-      "learning_rate": 1.975838225940415e-05,
-      "loss": 0.4509,
+      "epoch": 0.4916826144560577,
+      "grad_norm": 0.41706839132995893,
+      "learning_rate": 1.9752346241224466e-05,
+      "loss": 0.4857,
       "step": 3340
     },
     {
-      "epoch": 0.4901098901098901,
-      "grad_norm": 0.39839856694377546,
-      "learning_rate": 1.9756516175264703e-05,
-      "loss": 0.454,
+      "epoch": 0.4924186662741057,
+      "grad_norm": 0.396662264423611,
+      "learning_rate": 1.975044845244582e-05,
+      "loss": 0.4386,
       "step": 3345
     },
     {
-      "epoch": 0.4908424908424908,
-      "grad_norm": 0.4019660081498156,
-      "learning_rate": 1.9754643001468247e-05,
-      "loss": 0.4688,
+      "epoch": 0.49315471809215367,
+      "grad_norm": 0.40161162089841823,
+      "learning_rate": 1.9748543511818835e-05,
+      "loss": 0.4459,
       "step": 3350
     },
     {
-      "epoch": 0.49157509157509155,
-      "grad_norm": 0.42725169158973203,
-      "learning_rate": 1.9752762739375938e-05,
-      "loss": 0.4764,
+      "epoch": 0.4938907699102017,
+      "grad_norm": 0.3915110302234934,
+      "learning_rate": 1.974663142074076e-05,
+      "loss": 0.4625,
       "step": 3355
     },
     {
-      "epoch": 0.49230769230769234,
-      "grad_norm": 0.41230781357917834,
-      "learning_rate": 1.9750875390354083e-05,
-      "loss": 0.4759,
+      "epoch": 0.4946268217282497,
+      "grad_norm": 0.43822507627955265,
+      "learning_rate": 1.9744712180614093e-05,
+      "loss": 0.4733,
       "step": 3360
     },
     {
-      "epoch": 0.49304029304029307,
-      "grad_norm": 0.4358085904919089,
-      "learning_rate": 1.9748980955774148e-05,
-      "loss": 0.4675,
+      "epoch": 0.49536287354629766,
+      "grad_norm": 0.389001674944297,
+      "learning_rate": 1.974278579284658e-05,
+      "loss": 0.4633,
       "step": 3365
     },
     {
-      "epoch": 0.4937728937728938,
-      "grad_norm": 0.41400889915347355,
-      "learning_rate": 1.9747079437012736e-05,
-      "loss": 0.4754,
+      "epoch": 0.49609892536434563,
+      "grad_norm": 0.39998826272002275,
+      "learning_rate": 1.9740852258851203e-05,
+      "loss": 0.4656,
       "step": 3370
     },
     {
-      "epoch": 0.4945054945054945,
-      "grad_norm": 0.39535489205011043,
-      "learning_rate": 1.9745170835451598e-05,
-      "loss": 0.4843,
+      "epoch": 0.49683497718239367,
+      "grad_norm": 0.41163651340635726,
+      "learning_rate": 1.9738911580046185e-05,
+      "loss": 0.4749,
       "step": 3375
     },
     {
-      "epoch": 0.49523809523809526,
-      "grad_norm": 0.3875877035122525,
-      "learning_rate": 1.974325515247764e-05,
-      "loss": 0.4519,
+      "epoch": 0.49757102900044164,
+      "grad_norm": 0.4068628138795416,
+      "learning_rate": 1.9736963757855e-05,
+      "loss": 0.4476,
       "step": 3380
     },
     {
-      "epoch": 0.495970695970696,
-      "grad_norm": 0.40011808170012764,
-      "learning_rate": 1.9741332389482915e-05,
-      "loss": 0.4705,
+      "epoch": 0.4983070808184896,
+      "grad_norm": 0.39802458927173445,
+      "learning_rate": 1.973500879370635e-05,
+      "loss": 0.4591,
       "step": 3385
     },
     {
-      "epoch": 0.4967032967032967,
-      "grad_norm": 0.3875608965577219,
-      "learning_rate": 1.9739402547864605e-05,
-      "loss": 0.4612,
+      "epoch": 0.4990431326365376,
+      "grad_norm": 0.3959109892997859,
+      "learning_rate": 1.9733046689034182e-05,
+      "loss": 0.4696,
       "step": 3390
     },
     {
-      "epoch": 0.49743589743589745,
-      "grad_norm": 0.40605035088734287,
-      "learning_rate": 1.9737465629025054e-05,
-      "loss": 0.4729,
+      "epoch": 0.4997791844545856,
+      "grad_norm": 0.3837127945016217,
+      "learning_rate": 1.9731077445277684e-05,
+      "loss": 0.4583,
       "step": 3395
     },
     {
-      "epoch": 0.4981684981684982,
-      "grad_norm": 0.4076032885761923,
-      "learning_rate": 1.9735521634371734e-05,
-      "loss": 0.4601,
+      "epoch": 0.5005152362726336,
+      "grad_norm": 0.4205751599188156,
+      "learning_rate": 1.972910106388126e-05,
+      "loss": 0.4766,
       "step": 3400
     },
     {
-      "epoch": 0.4989010989010989,
-      "grad_norm": 0.4058404149960834,
-      "learning_rate": 1.973357056531727e-05,
-      "loss": 0.4699,
+      "epoch": 0.5012512880906815,
+      "grad_norm": 0.41793031062362607,
+      "learning_rate": 1.9727117546294573e-05,
+      "loss": 0.4594,
       "step": 3405
     },
     {
-      "epoch": 0.49963369963369964,
-      "grad_norm": 0.38441775513851356,
-      "learning_rate": 1.9731612423279425e-05,
-      "loss": 0.4395,
+      "epoch": 0.5019873399087296,
+      "grad_norm": 0.43789613525420124,
+      "learning_rate": 1.9725126893972514e-05,
+      "loss": 0.456,
       "step": 3410
     },
     {
-      "epoch": 0.5003663003663004,
-      "grad_norm": 0.3856852354737355,
-      "learning_rate": 1.9729647209681095e-05,
-      "loss": 0.4415,
+      "epoch": 0.5027233917267776,
+      "grad_norm": 0.4327081367462604,
+      "learning_rate": 1.9723129108375203e-05,
+      "loss": 0.4528,
       "step": 3415
     },
     {
-      "epoch": 0.5010989010989011,
-      "grad_norm": 0.40733571237223243,
-      "learning_rate": 1.9727674925950322e-05,
-      "loss": 0.4776,
+      "epoch": 0.5034594435448255,
+      "grad_norm": 0.40465484943979296,
+      "learning_rate": 1.9721124190967995e-05,
+      "loss": 0.4635,
       "step": 3420
     },
     {
-      "epoch": 0.5018315018315018,
-      "grad_norm": 0.4160928717740055,
-      "learning_rate": 1.9725695573520286e-05,
-      "loss": 0.4759,
+      "epoch": 0.5041954953628736,
+      "grad_norm": 0.39001525778029733,
+      "learning_rate": 1.9719112143221472e-05,
+      "loss": 0.4515,
       "step": 3425
     },
     {
-      "epoch": 0.5025641025641026,
-      "grad_norm": 0.39267972694890735,
-      "learning_rate": 1.9723709153829295e-05,
-      "loss": 0.4444,
+      "epoch": 0.5049315471809216,
+      "grad_norm": 0.4142644311024722,
+      "learning_rate": 1.971709296661145e-05,
+      "loss": 0.4528,
       "step": 3430
     },
     {
-      "epoch": 0.5032967032967033,
-      "grad_norm": 0.39638394361802154,
-      "learning_rate": 1.97217156683208e-05,
-      "loss": 0.4372,
+      "epoch": 0.5056675989989695,
+      "grad_norm": 0.4233936755271006,
+      "learning_rate": 1.9715066662618974e-05,
+      "loss": 0.4522,
       "step": 3435
     },
     {
-      "epoch": 0.504029304029304,
-      "grad_norm": 0.42218569169595366,
-      "learning_rate": 1.9719715118443386e-05,
-      "loss": 0.4581,
+      "epoch": 0.5064036508170175,
+      "grad_norm": 0.41635368549993773,
+      "learning_rate": 1.9713033232730318e-05,
+      "loss": 0.4635,
       "step": 3440
     },
     {
-      "epoch": 0.5047619047619047,
-      "grad_norm": 0.42583544248704563,
-      "learning_rate": 1.9717707505650767e-05,
-      "loss": 0.4585,
+      "epoch": 0.5071397026350655,
+      "grad_norm": 0.42385806677774274,
+      "learning_rate": 1.971099267843698e-05,
+      "loss": 0.4588,
       "step": 3445
     },
     {
-      "epoch": 0.5054945054945055,
-      "grad_norm": 0.40510679362327034,
-      "learning_rate": 1.9715692831401798e-05,
-      "loss": 0.4595,
+      "epoch": 0.5078757544531135,
+      "grad_norm": 0.39366113592704577,
+      "learning_rate": 1.9708945001235686e-05,
+      "loss": 0.4635,
       "step": 3450
     },
     {
-      "epoch": 0.5062271062271062,
-      "grad_norm": 0.4244850714320826,
-      "learning_rate": 1.9713671097160454e-05,
-      "loss": 0.4637,
+      "epoch": 0.5086118062711615,
+      "grad_norm": 0.4285306151705861,
+      "learning_rate": 1.9706890202628376e-05,
+      "loss": 0.4563,
       "step": 3455
     },
     {
-      "epoch": 0.5069597069597069,
-      "grad_norm": 0.3959484392053746,
-      "learning_rate": 1.971164230439585e-05,
-      "loss": 0.4503,
+      "epoch": 0.5093478580892095,
+      "grad_norm": 0.41035525935027833,
+      "learning_rate": 1.970482828412223e-05,
+      "loss": 0.4585,
       "step": 3460
     },
     {
-      "epoch": 0.5076923076923077,
-      "grad_norm": 0.4020029057429197,
-      "learning_rate": 1.970960645458222e-05,
-      "loss": 0.456,
+      "epoch": 0.5100839099072575,
+      "grad_norm": 0.4001206582628323,
+      "learning_rate": 1.9702759247229647e-05,
+      "loss": 0.4612,
       "step": 3465
     },
     {
-      "epoch": 0.5084249084249084,
-      "grad_norm": 0.3865009439791744,
-      "learning_rate": 1.9707563549198934e-05,
-      "loss": 0.444,
+      "epoch": 0.5108199617253054,
+      "grad_norm": 0.4076277925400673,
+      "learning_rate": 1.9700683093468235e-05,
+      "loss": 0.4656,
       "step": 3470
     },
     {
-      "epoch": 0.5091575091575091,
-      "grad_norm": 0.3824361919325873,
-      "learning_rate": 1.970551358973049e-05,
-      "loss": 0.4502,
+      "epoch": 0.5115560135433534,
+      "grad_norm": 0.40262365010775447,
+      "learning_rate": 1.9698599824360832e-05,
+      "loss": 0.4513,
       "step": 3475
     },
     {
-      "epoch": 0.5098901098901099,
-      "grad_norm": 0.39293252412701707,
-      "learning_rate": 1.9703456577666507e-05,
-      "loss": 0.4575,
+      "epoch": 0.5122920653614015,
+      "grad_norm": 0.547089339379227,
+      "learning_rate": 1.9696509441435498e-05,
+      "loss": 0.4561,
       "step": 3480
     },
     {
-      "epoch": 0.5106227106227106,
-      "grad_norm": 0.37927190798663696,
-      "learning_rate": 1.970139251450173e-05,
-      "loss": 0.4571,
+      "epoch": 0.5130281171794494,
+      "grad_norm": 0.44020060733759014,
+      "learning_rate": 1.9694411946225502e-05,
+      "loss": 0.4565,
       "step": 3485
     },
     {
-      "epoch": 0.5113553113553113,
-      "grad_norm": 0.40747726547713325,
-      "learning_rate": 1.9699321401736028e-05,
-      "loss": 0.4598,
+      "epoch": 0.5137641689974974,
+      "grad_norm": 0.41610521756030733,
+      "learning_rate": 1.9692307340269334e-05,
+      "loss": 0.4832,
       "step": 3490
     },
     {
-      "epoch": 0.512087912087912,
-      "grad_norm": 0.42005617221329955,
-      "learning_rate": 1.9697243240874394e-05,
-      "loss": 0.4582,
+      "epoch": 0.5145002208155454,
+      "grad_norm": 0.41417230996854715,
+      "learning_rate": 1.96901956251107e-05,
+      "loss": 0.446,
       "step": 3495
     },
     {
-      "epoch": 0.5128205128205128,
-      "grad_norm": 0.4117478480032374,
-      "learning_rate": 1.9695158033426945e-05,
-      "loss": 0.4634,
+      "epoch": 0.5152362726335934,
+      "grad_norm": 0.40750264954931703,
+      "learning_rate": 1.9688076802298525e-05,
+      "loss": 0.4465,
       "step": 3500
     },
     {
-      "epoch": 0.5135531135531135,
-      "grad_norm": 0.40341267325790553,
-      "learning_rate": 1.9693065780908907e-05,
-      "loss": 0.4708,
+      "epoch": 0.5159723244516414,
+      "grad_norm": 0.4170889737094148,
+      "learning_rate": 1.9685950873386935e-05,
+      "loss": 0.4539,
       "step": 3505
     },
     {
-      "epoch": 0.5142857142857142,
-      "grad_norm": 0.4306074517094488,
-      "learning_rate": 1.9690966484840645e-05,
-      "loss": 0.4583,
+      "epoch": 0.5167083762696894,
+      "grad_norm": 0.4148491025631686,
+      "learning_rate": 1.9683817839935278e-05,
+      "loss": 0.4607,
       "step": 3510
     },
     {
-      "epoch": 0.515018315018315,
-      "grad_norm": 0.40995384275820096,
-      "learning_rate": 1.9688860146747616e-05,
-      "loss": 0.4667,
+      "epoch": 0.5174444280877374,
+      "grad_norm": 0.4163873194725317,
+      "learning_rate": 1.968167770350812e-05,
+      "loss": 0.4762,
       "step": 3515
     },
     {
-      "epoch": 0.5157509157509158,
-      "grad_norm": 0.4180393839855888,
-      "learning_rate": 1.9686746768160422e-05,
-      "loss": 0.4695,
+      "epoch": 0.5181804799057854,
+      "grad_norm": 0.42547828840980223,
+      "learning_rate": 1.9679530465675213e-05,
+      "loss": 0.458,
       "step": 3520
     },
     {
-      "epoch": 0.5164835164835165,
-      "grad_norm": 0.3939921437930165,
-      "learning_rate": 1.9684626350614756e-05,
-      "loss": 0.4703,
+      "epoch": 0.5189165317238333,
+      "grad_norm": 0.43469177209885107,
+      "learning_rate": 1.967737612801154e-05,
+      "loss": 0.462,
       "step": 3525
     },
     {
-      "epoch": 0.5172161172161173,
-      "grad_norm": 0.39002670431542874,
-      "learning_rate": 1.968249889565145e-05,
-      "loss": 0.4504,
+      "epoch": 0.5196525835418814,
+      "grad_norm": 0.4100192594814962,
+      "learning_rate": 1.9675214692097285e-05,
+      "loss": 0.4417,
       "step": 3530
     },
     {
-      "epoch": 0.517948717948718,
-      "grad_norm": 0.40762066174533135,
-      "learning_rate": 1.968036440481643e-05,
-      "loss": 0.4597,
+      "epoch": 0.5203886353599293,
+      "grad_norm": 0.3912658697938237,
+      "learning_rate": 1.9673046159517838e-05,
+      "loss": 0.4667,
       "step": 3535
     },
     {
-      "epoch": 0.5186813186813187,
-      "grad_norm": 0.3970346723297381,
-      "learning_rate": 1.9678222879660745e-05,
-      "loss": 0.4592,
+      "epoch": 0.5211246871779773,
+      "grad_norm": 0.3903472216559458,
+      "learning_rate": 1.9670870531863795e-05,
+      "loss": 0.4689,
       "step": 3540
     },
     {
-      "epoch": 0.5194139194139195,
-      "grad_norm": 0.3962972174341633,
-      "learning_rate": 1.967607432174055e-05,
-      "loss": 0.4693,
+      "epoch": 0.5218607389960254,
+      "grad_norm": 0.4161083312677932,
+      "learning_rate": 1.966868781073095e-05,
+      "loss": 0.4584,
       "step": 3545
     },
     {
-      "epoch": 0.5201465201465202,
-      "grad_norm": 0.41614465682076274,
-      "learning_rate": 1.9673918732617116e-05,
-      "loss": 0.4396,
+      "epoch": 0.5225967908140733,
+      "grad_norm": 0.3938751548055341,
+      "learning_rate": 1.9666497997720312e-05,
+      "loss": 0.4574,
       "step": 3550
     },
     {
-      "epoch": 0.5208791208791209,
-      "grad_norm": 0.3931314447365937,
-      "learning_rate": 1.967175611385682e-05,
-      "loss": 0.4729,
+      "epoch": 0.5233328426321213,
+      "grad_norm": 0.39683016345756705,
+      "learning_rate": 1.9664301094438082e-05,
+      "loss": 0.454,
       "step": 3555
     },
     {
-      "epoch": 0.5216117216117216,
-      "grad_norm": 0.4307490611740214,
-      "learning_rate": 1.966958646703115e-05,
-      "loss": 0.457,
+      "epoch": 0.5240688944501692,
+      "grad_norm": 0.3798854852370915,
+      "learning_rate": 1.966209710249566e-05,
+      "loss": 0.4635,
       "step": 3560
     },
     {
-      "epoch": 0.5223443223443224,
-      "grad_norm": 0.5263107663262924,
-      "learning_rate": 1.966740979371669e-05,
-      "loss": 0.4663,
+      "epoch": 0.5248049462682173,
+      "grad_norm": 0.37628706188680294,
+      "learning_rate": 1.965988602350966e-05,
+      "loss": 0.4638,
       "step": 3565
     },
     {
-      "epoch": 0.5230769230769231,
-      "grad_norm": 0.40406648280499774,
-      "learning_rate": 1.9665226095495145e-05,
-      "loss": 0.4532,
+      "epoch": 0.5255409980862653,
+      "grad_norm": 0.37866479579838463,
+      "learning_rate": 1.965766785910188e-05,
+      "loss": 0.4651,
       "step": 3570
     },
     {
-      "epoch": 0.5238095238095238,
-      "grad_norm": 0.4000541126808561,
-      "learning_rate": 1.966303537395332e-05,
-      "loss": 0.4739,
+      "epoch": 0.5262770499043132,
+      "grad_norm": 0.40679176520597915,
+      "learning_rate": 1.9655442610899314e-05,
+      "loss": 0.4561,
       "step": 3575
     },
     {
-      "epoch": 0.5245421245421246,
-      "grad_norm": 0.4218513937972455,
-      "learning_rate": 1.9660837630683122e-05,
-      "loss": 0.4599,
+      "epoch": 0.5270131017223613,
+      "grad_norm": 0.40261444513151734,
+      "learning_rate": 1.9653210280534165e-05,
+      "loss": 0.457,
       "step": 3580
     },
     {
-      "epoch": 0.5252747252747253,
-      "grad_norm": 0.39734014409545276,
-      "learning_rate": 1.9658632867281554e-05,
-      "loss": 0.4604,
+      "epoch": 0.5277491535404093,
+      "grad_norm": 0.5761751860941852,
+      "learning_rate": 1.965097086964382e-05,
+      "loss": 0.4637,
       "step": 3585
     },
     {
-      "epoch": 0.526007326007326,
-      "grad_norm": 0.43566685531130656,
-      "learning_rate": 1.965642108535073e-05,
-      "loss": 0.4599,
+      "epoch": 0.5284852053584572,
+      "grad_norm": 0.40247146010633544,
+      "learning_rate": 1.9648724379870864e-05,
+      "loss": 0.4559,
       "step": 3590
     },
     {
-      "epoch": 0.5267399267399268,
-      "grad_norm": 0.3877333961498012,
-      "learning_rate": 1.965420228649786e-05,
-      "loss": 0.4514,
+      "epoch": 0.5292212571765053,
+      "grad_norm": 0.4075287152163133,
+      "learning_rate": 1.9646470812863076e-05,
+      "loss": 0.4677,
       "step": 3595
     },
     {
-      "epoch": 0.5274725274725275,
-      "grad_norm": 0.4037996593369899,
-      "learning_rate": 1.9651976472335253e-05,
-      "loss": 0.4733,
+      "epoch": 0.5299573089945532,
+      "grad_norm": 0.36811491522127043,
+      "learning_rate": 1.9644210170273414e-05,
+      "loss": 0.4343,
       "step": 3600
     },
     {
-      "epoch": 0.5282051282051282,
-      "grad_norm": 0.4097352848327507,
-      "learning_rate": 1.9649743644480322e-05,
-      "loss": 0.4504,
+      "epoch": 0.5306933608126012,
+      "grad_norm": 0.4078901945434575,
+      "learning_rate": 1.964194245376004e-05,
+      "loss": 0.4563,
       "step": 3605
     },
     {
-      "epoch": 0.528937728937729,
-      "grad_norm": 0.40968685362670326,
-      "learning_rate": 1.9647503804555562e-05,
-      "loss": 0.4689,
+      "epoch": 0.5314294126306492,
+      "grad_norm": 0.4002015071997187,
+      "learning_rate": 1.9639667664986303e-05,
+      "loss": 0.4488,
       "step": 3610
     },
     {
-      "epoch": 0.5296703296703297,
-      "grad_norm": 0.40407274799067217,
-      "learning_rate": 1.9645256954188578e-05,
-      "loss": 0.4809,
+      "epoch": 0.5321654644486972,
+      "grad_norm": 0.40452853995295185,
+      "learning_rate": 1.963738580562073e-05,
+      "loss": 0.4767,
       "step": 3615
     },
     {
-      "epoch": 0.5304029304029304,
-      "grad_norm": 0.43432136759647494,
-      "learning_rate": 1.9643003095012058e-05,
-      "loss": 0.4247,
+      "epoch": 0.5329015162667452,
+      "grad_norm": 0.48302106579477866,
+      "learning_rate": 1.9635096877337045e-05,
+      "loss": 0.4577,
       "step": 3620
     },
     {
-      "epoch": 0.5311355311355311,
-      "grad_norm": 0.3846967561165999,
-      "learning_rate": 1.9640742228663793e-05,
-      "loss": 0.4532,
+      "epoch": 0.5336375680847931,
+      "grad_norm": 0.37366996425582205,
+      "learning_rate": 1.9632800881814146e-05,
+      "loss": 0.4453,
       "step": 3625
     },
     {
-      "epoch": 0.5318681318681319,
-      "grad_norm": 0.39222900221330453,
-      "learning_rate": 1.963847435678666e-05,
-      "loss": 0.4707,
+      "epoch": 0.5343736199028412,
+      "grad_norm": 0.38055230387120575,
+      "learning_rate": 1.9630497820736128e-05,
+      "loss": 0.4683,
       "step": 3630
     },
     {
-      "epoch": 0.5326007326007326,
-      "grad_norm": 0.39753060827257874,
-      "learning_rate": 1.963619948102863e-05,
-      "loss": 0.4667,
+      "epoch": 0.5351096717208892,
+      "grad_norm": 0.4032298844501685,
+      "learning_rate": 1.9628187695792257e-05,
+      "loss": 0.4763,
       "step": 3635
     },
     {
-      "epoch": 0.5333333333333333,
-      "grad_norm": 0.39453610064769346,
-      "learning_rate": 1.9633917603042757e-05,
-      "loss": 0.4495,
+      "epoch": 0.5358457235389371,
+      "grad_norm": 0.39467956429798245,
+      "learning_rate": 1.9625870508676982e-05,
+      "loss": 0.4417,
       "step": 3640
     },
     {
-      "epoch": 0.5340659340659341,
-      "grad_norm": 0.42353928664626905,
-      "learning_rate": 1.9631628724487185e-05,
-      "loss": 0.5042,
+      "epoch": 0.5365817753569851,
+      "grad_norm": 0.4039574603614267,
+      "learning_rate": 1.9623546261089942e-05,
+      "loss": 0.471,
       "step": 3645
     },
     {
-      "epoch": 0.5347985347985348,
-      "grad_norm": 0.4759710558557821,
-      "learning_rate": 1.9629332847025154e-05,
-      "loss": 0.4788,
+      "epoch": 0.5373178271750331,
+      "grad_norm": 0.4038268554128066,
+      "learning_rate": 1.9621214954735944e-05,
+      "loss": 0.4745,
       "step": 3650
     },
     {
-      "epoch": 0.5355311355311355,
-      "grad_norm": 0.4128618181070763,
-      "learning_rate": 1.9627029972324983e-05,
-      "loss": 0.4623,
+      "epoch": 0.5380538789930811,
+      "grad_norm": 0.4225586718667442,
+      "learning_rate": 1.9618876591324973e-05,
+      "loss": 0.4424,
       "step": 3655
     },
     {
-      "epoch": 0.5362637362637362,
-      "grad_norm": 0.4047288745230385,
-      "learning_rate": 1.962472010206007e-05,
-      "loss": 0.4401,
+      "epoch": 0.5387899308111291,
+      "grad_norm": 0.39746799943722094,
+      "learning_rate": 1.9616531172572198e-05,
+      "loss": 0.4522,
       "step": 3660
     },
     {
-      "epoch": 0.536996336996337,
-      "grad_norm": 0.4116700375902194,
-      "learning_rate": 1.9622403237908907e-05,
-      "loss": 0.4547,
+      "epoch": 0.5395259826291771,
+      "grad_norm": 0.411921110031647,
+      "learning_rate": 1.9614178700197955e-05,
+      "loss": 0.454,
       "step": 3665
     },
     {
-      "epoch": 0.5377289377289377,
-      "grad_norm": 0.39702761823883265,
-      "learning_rate": 1.9620079381555062e-05,
-      "loss": 0.4664,
+      "epoch": 0.5402620344472251,
+      "grad_norm": 0.39706878989697264,
+      "learning_rate": 1.961181917592776e-05,
+      "loss": 0.4643,
       "step": 3670
     },
     {
-      "epoch": 0.5384615384615384,
-      "grad_norm": 0.4006365566597177,
-      "learning_rate": 1.9617748534687188e-05,
-      "loss": 0.4483,
+      "epoch": 0.5409980862652731,
+      "grad_norm": 0.39859880762184235,
+      "learning_rate": 1.9609452601492293e-05,
+      "loss": 0.4578,
       "step": 3675
     },
     {
-      "epoch": 0.5391941391941392,
-      "grad_norm": 0.4053964771147014,
-      "learning_rate": 1.961541069899901e-05,
-      "loss": 0.461,
+      "epoch": 0.541734138083321,
+      "grad_norm": 0.3925765914722752,
+      "learning_rate": 1.9607078978627417e-05,
+      "loss": 0.4457,
       "step": 3680
     },
     {
-      "epoch": 0.5399267399267399,
-      "grad_norm": 0.6423049012045474,
-      "learning_rate": 1.961306587618934e-05,
-      "loss": 0.495,
+      "epoch": 0.5424701899013691,
+      "grad_norm": 0.3973018128767815,
+      "learning_rate": 1.9604698309074154e-05,
+      "loss": 0.4636,
       "step": 3685
     },
     {
-      "epoch": 0.5406593406593406,
-      "grad_norm": 0.3910992104613754,
-      "learning_rate": 1.961071406796206e-05,
-      "loss": 0.4634,
+      "epoch": 0.543206241719417,
+      "grad_norm": 0.41295053325560216,
+      "learning_rate": 1.9602310594578705e-05,
+      "loss": 0.4673,
       "step": 3690
     },
     {
-      "epoch": 0.5413919413919414,
-      "grad_norm": 0.41479888215520155,
-      "learning_rate": 1.9608355276026137e-05,
-      "loss": 0.464,
+      "epoch": 0.543942293537465,
+      "grad_norm": 0.40261677890369485,
+      "learning_rate": 1.9599915836892424e-05,
+      "loss": 0.4554,
       "step": 3695
     },
     {
-      "epoch": 0.5421245421245421,
-      "grad_norm": 0.38381801157498935,
-      "learning_rate": 1.9605989502095607e-05,
-      "loss": 0.458,
+      "epoch": 0.5446783453555131,
+      "grad_norm": 0.4162673811273489,
+      "learning_rate": 1.959751403777185e-05,
+      "loss": 0.4709,
       "step": 3700
     },
     {
-      "epoch": 0.5428571428571428,
-      "grad_norm": 0.3726198460602332,
-      "learning_rate": 1.9603616747889577e-05,
-      "loss": 0.4591,
+      "epoch": 0.545414397173561,
+      "grad_norm": 0.3955559642902411,
+      "learning_rate": 1.9595105198978666e-05,
+      "loss": 0.4588,
       "step": 3705
     },
     {
-      "epoch": 0.5435897435897435,
-      "grad_norm": 0.409146983712297,
-      "learning_rate": 1.960123701513223e-05,
+      "epoch": 0.546150448991609,
+      "grad_norm": 0.4015043828053614,
+      "learning_rate": 1.959268932227974e-05,
       "loss": 0.4692,
       "step": 3710
     },
     {
-      "epoch": 0.5443223443223443,
-      "grad_norm": 0.40176825120030335,
-      "learning_rate": 1.9598850305552824e-05,
-      "loss": 0.4717,
+      "epoch": 0.546886500809657,
+      "grad_norm": 0.3852610487294835,
+      "learning_rate": 1.959026640944708e-05,
+      "loss": 0.4506,
       "step": 3715
     },
     {
-      "epoch": 0.545054945054945,
-      "grad_norm": 0.38283812772967896,
-      "learning_rate": 1.9596456620885675e-05,
-      "loss": 0.4486,
+      "epoch": 0.547622552627705,
+      "grad_norm": 0.4528320271943359,
+      "learning_rate": 1.9587836462257878e-05,
+      "loss": 0.4545,
       "step": 3720
     },
     {
-      "epoch": 0.5457875457875457,
-      "grad_norm": 0.41477686180698137,
-      "learning_rate": 1.9594055962870178e-05,
-      "loss": 0.4645,
+      "epoch": 0.548358604445753,
+      "grad_norm": 0.3882142221048134,
+      "learning_rate": 1.9585399482494468e-05,
+      "loss": 0.464,
       "step": 3725
     },
     {
-      "epoch": 0.5465201465201465,
-      "grad_norm": 0.39579938529403774,
-      "learning_rate": 1.9591648333250793e-05,
-      "loss": 0.4583,
+      "epoch": 0.5490946562638009,
+      "grad_norm": 0.4367497179561277,
+      "learning_rate": 1.9582955471944345e-05,
+      "loss": 0.4678,
       "step": 3730
     },
     {
-      "epoch": 0.5472527472527473,
-      "grad_norm": 0.38826954570377586,
-      "learning_rate": 1.9589233733777045e-05,
-      "loss": 0.463,
+      "epoch": 0.549830708081849,
+      "grad_norm": 0.3975678237393734,
+      "learning_rate": 1.958050443240017e-05,
+      "loss": 0.4541,
       "step": 3735
     },
     {
-      "epoch": 0.547985347985348,
-      "grad_norm": 0.39398242670127886,
-      "learning_rate": 1.958681216620352e-05,
-      "loss": 0.4574,
+      "epoch": 0.550566759899897,
+      "grad_norm": 0.40424599351055596,
+      "learning_rate": 1.957804636565975e-05,
+      "loss": 0.4536,
       "step": 3740
     },
     {
-      "epoch": 0.5487179487179488,
-      "grad_norm": 0.41501647675382003,
-      "learning_rate": 1.9584383632289877e-05,
-      "loss": 0.4666,
+      "epoch": 0.5513028117179449,
+      "grad_norm": 0.40267105164143574,
+      "learning_rate": 1.9575581273526055e-05,
+      "loss": 0.4648,
       "step": 3745
     },
     {
-      "epoch": 0.5494505494505495,
-      "grad_norm": 1.564754790229026,
-      "learning_rate": 1.9581948133800822e-05,
-      "loss": 0.4761,
+      "epoch": 0.552038863535993,
+      "grad_norm": 0.4116144060323841,
+      "learning_rate": 1.95731091578072e-05,
+      "loss": 0.4536,
       "step": 3750
     },
     {
-      "epoch": 0.5501831501831502,
-      "grad_norm": 0.4085577720422966,
-      "learning_rate": 1.9579505672506143e-05,
-      "loss": 0.4731,
+      "epoch": 0.5527749153540409,
+      "grad_norm": 0.3931753386822875,
+      "learning_rate": 1.9570630020316455e-05,
+      "loss": 0.4718,
       "step": 3755
     },
     {
-      "epoch": 0.550915750915751,
-      "grad_norm": 0.3965996616045536,
-      "learning_rate": 1.9577056250180667e-05,
-      "loss": 0.4641,
+      "epoch": 0.5535109671720889,
+      "grad_norm": 0.391232231865654,
+      "learning_rate": 1.956814386287224e-05,
+      "loss": 0.4694,
       "step": 3760
     },
     {
-      "epoch": 0.5516483516483517,
-      "grad_norm": 0.38306772647374016,
-      "learning_rate": 1.9574599868604292e-05,
-      "loss": 0.4632,
+      "epoch": 0.554247018990137,
+      "grad_norm": 0.391889438154536,
+      "learning_rate": 1.956565068729813e-05,
+      "loss": 0.4624,
       "step": 3765
     },
     {
-      "epoch": 0.5523809523809524,
-      "grad_norm": 0.39692838594704005,
-      "learning_rate": 1.957213652956197e-05,
-      "loss": 0.4682,
+      "epoch": 0.5549830708081849,
+      "grad_norm": 0.38870877866166037,
+      "learning_rate": 1.9563150495422834e-05,
+      "loss": 0.4646,
       "step": 3770
     },
     {
-      "epoch": 0.5531135531135531,
-      "grad_norm": 0.4139888092877681,
-      "learning_rate": 1.95696662348437e-05,
-      "loss": 0.4565,
+      "epoch": 0.5557191226262329,
+      "grad_norm": 0.40984342682564273,
+      "learning_rate": 1.956064328908022e-05,
+      "loss": 0.4616,
       "step": 3775
     },
     {
-      "epoch": 0.5538461538461539,
-      "grad_norm": 0.4038218231406527,
-      "learning_rate": 1.9567188986244552e-05,
-      "loss": 0.4542,
+      "epoch": 0.5564551744442808,
+      "grad_norm": 0.4094591347657577,
+      "learning_rate": 1.9558129070109294e-05,
+      "loss": 0.4668,
       "step": 3780
     },
     {
-      "epoch": 0.5545787545787546,
-      "grad_norm": 0.4183645918853025,
-      "learning_rate": 1.9564704785564634e-05,
-      "loss": 0.4671,
+      "epoch": 0.5571912262623289,
+      "grad_norm": 0.4107014394366221,
+      "learning_rate": 1.9555607840354217e-05,
+      "loss": 0.4706,
       "step": 3785
     },
     {
-      "epoch": 0.5553113553113553,
-      "grad_norm": 0.3880015641068925,
-      "learning_rate": 1.9562213634609118e-05,
-      "loss": 0.4651,
+      "epoch": 0.5579272780803769,
+      "grad_norm": 0.41844199725210224,
+      "learning_rate": 1.9553079601664274e-05,
+      "loss": 0.4521,
       "step": 3790
     },
     {
-      "epoch": 0.5560439560439561,
-      "grad_norm": 0.406983465349988,
-      "learning_rate": 1.9559715535188216e-05,
-      "loss": 0.4378,
+      "epoch": 0.5586633298984248,
+      "grad_norm": 0.3708737602656749,
+      "learning_rate": 1.9550544355893902e-05,
+      "loss": 0.449,
       "step": 3795
     },
     {
-      "epoch": 0.5567765567765568,
-      "grad_norm": 0.3789384861043702,
-      "learning_rate": 1.955721048911719e-05,
-      "loss": 0.4609,
+      "epoch": 0.5593993817164729,
+      "grad_norm": 0.3912874868887559,
+      "learning_rate": 1.954800210490268e-05,
+      "loss": 0.447,
       "step": 3800
     },
     {
-      "epoch": 0.5575091575091575,
-      "grad_norm": 0.3746455880435221,
-      "learning_rate": 1.9554698498216362e-05,
-      "loss": 0.4552,
+      "epoch": 0.5601354335345208,
+      "grad_norm": 0.3922403999244812,
+      "learning_rate": 1.9545452850555323e-05,
+      "loss": 0.4547,
       "step": 3805
     },
     {
-      "epoch": 0.5582417582417583,
-      "grad_norm": 0.4021508409555428,
-      "learning_rate": 1.9552179564311088e-05,
-      "loss": 0.4597,
+      "epoch": 0.5608714853525688,
+      "grad_norm": 0.4726372690020183,
+      "learning_rate": 1.9542896594721678e-05,
+      "loss": 0.4688,
       "step": 3810
     },
     {
-      "epoch": 0.558974358974359,
-      "grad_norm": 0.41079276918661684,
-      "learning_rate": 1.954965368923177e-05,
-      "loss": 0.4669,
+      "epoch": 0.5616075371706168,
+      "grad_norm": 0.3912213717737321,
+      "learning_rate": 1.9540333339276726e-05,
+      "loss": 0.4432,
       "step": 3815
     },
     {
-      "epoch": 0.5597069597069597,
-      "grad_norm": 0.4101630906550437,
-      "learning_rate": 1.9547120874813862e-05,
-      "loss": 0.4678,
+      "epoch": 0.5623435889886648,
+      "grad_norm": 0.3772805758009394,
+      "learning_rate": 1.9537763086100598e-05,
+      "loss": 0.4585,
       "step": 3820
     },
     {
-      "epoch": 0.5604395604395604,
-      "grad_norm": 0.3612610892397642,
-      "learning_rate": 1.9544581122897854e-05,
-      "loss": 0.4554,
+      "epoch": 0.5630796408067128,
+      "grad_norm": 0.38255828351165105,
+      "learning_rate": 1.953518583707854e-05,
+      "loss": 0.4326,
       "step": 3825
     },
     {
-      "epoch": 0.5611721611721612,
-      "grad_norm": 0.40697790168410697,
-      "learning_rate": 1.9542034435329276e-05,
-      "loss": 0.4623,
+      "epoch": 0.5638156926247608,
+      "grad_norm": 0.4029940798806384,
+      "learning_rate": 1.9532601594100936e-05,
+      "loss": 0.4619,
       "step": 3830
     },
     {
-      "epoch": 0.5619047619047619,
-      "grad_norm": 0.4282245049901043,
-      "learning_rate": 1.9539480813958703e-05,
-      "loss": 0.4513,
+      "epoch": 0.5645517444428088,
+      "grad_norm": 0.406308443969887,
+      "learning_rate": 1.9530010359063306e-05,
+      "loss": 0.4533,
       "step": 3835
     },
     {
-      "epoch": 0.5626373626373626,
-      "grad_norm": 0.3954324592961608,
-      "learning_rate": 1.953692026064175e-05,
-      "loss": 0.4499,
+      "epoch": 0.5652877962608568,
+      "grad_norm": 0.39335542730985024,
+      "learning_rate": 1.9527412133866285e-05,
+      "loss": 0.4503,
       "step": 3840
     },
     {
-      "epoch": 0.5633699633699634,
-      "grad_norm": 0.40670314241955646,
-      "learning_rate": 1.9534352777239055e-05,
-      "loss": 0.4541,
+      "epoch": 0.5660238480789047,
+      "grad_norm": 0.38634711158644086,
+      "learning_rate": 1.9524806920415646e-05,
+      "loss": 0.4512,
       "step": 3845
     },
     {
-      "epoch": 0.5641025641025641,
-      "grad_norm": 0.4161193541654707,
-      "learning_rate": 1.9531778365616315e-05,
-      "loss": 0.4721,
+      "epoch": 0.5667598998969527,
+      "grad_norm": 0.4022091151221169,
+      "learning_rate": 1.952219472062229e-05,
+      "loss": 0.4538,
       "step": 3850
     },
     {
-      "epoch": 0.5648351648351648,
-      "grad_norm": 0.411357018786779,
-      "learning_rate": 1.952919702764424e-05,
-      "loss": 0.4613,
+      "epoch": 0.5674959517150008,
+      "grad_norm": 0.4061506497101422,
+      "learning_rate": 1.951957553640223e-05,
+      "loss": 0.4561,
       "step": 3855
     },
     {
-      "epoch": 0.5655677655677656,
-      "grad_norm": 0.3839854092932068,
-      "learning_rate": 1.952660876519858e-05,
-      "loss": 0.4631,
+      "epoch": 0.5682320035330487,
+      "grad_norm": 0.392747642013472,
+      "learning_rate": 1.9516949369676613e-05,
+      "loss": 0.456,
       "step": 3860
     },
     {
-      "epoch": 0.5663003663003663,
-      "grad_norm": 0.4085926286619489,
-      "learning_rate": 1.952401358016012e-05,
-      "loss": 0.4626,
+      "epoch": 0.5689680553510967,
+      "grad_norm": 0.3917136099359976,
+      "learning_rate": 1.9514316222371703e-05,
+      "loss": 0.4628,
       "step": 3865
     },
     {
-      "epoch": 0.567032967032967,
-      "grad_norm": 0.44557607925036696,
-      "learning_rate": 1.9521411474414683e-05,
-      "loss": 0.4656,
+      "epoch": 0.5697041071691447,
+      "grad_norm": 0.38856963846046494,
+      "learning_rate": 1.9511676096418884e-05,
+      "loss": 0.4402,
       "step": 3870
     },
     {
-      "epoch": 0.5677655677655677,
-      "grad_norm": 0.45987074599106403,
-      "learning_rate": 1.95188024498531e-05,
-      "loss": 0.4495,
+      "epoch": 0.5704401589871927,
+      "grad_norm": 0.37840817197647325,
+      "learning_rate": 1.950902899375466e-05,
+      "loss": 0.447,
       "step": 3875
     },
     {
-      "epoch": 0.5684981684981685,
-      "grad_norm": 0.43349497128485376,
-      "learning_rate": 1.951618650837124e-05,
-      "loss": 0.4531,
+      "epoch": 0.5711762108052407,
+      "grad_norm": 0.4117374694793402,
+      "learning_rate": 1.9506374916320655e-05,
+      "loss": 0.4486,
       "step": 3880
     },
     {
-      "epoch": 0.5692307692307692,
-      "grad_norm": 0.3976789014010414,
-      "learning_rate": 1.9513563651870015e-05,
-      "loss": 0.462,
+      "epoch": 0.5719122626232886,
+      "grad_norm": 0.3961584937171518,
+      "learning_rate": 1.9503713866063606e-05,
+      "loss": 0.4724,
       "step": 3885
     },
     {
-      "epoch": 0.5699633699633699,
-      "grad_norm": 0.407893568851839,
-      "learning_rate": 1.951093388225533e-05,
-      "loss": 0.4486,
+      "epoch": 0.5726483144413367,
+      "grad_norm": 0.39437169614191947,
+      "learning_rate": 1.950104584493536e-05,
+      "loss": 0.4427,
       "step": 3890
     },
     {
-      "epoch": 0.5706959706959707,
-      "grad_norm": 0.3879247914118072,
-      "learning_rate": 1.950829720143814e-05,
-      "loss": 0.4664,
+      "epoch": 0.5733843662593847,
+      "grad_norm": 0.38806710304574166,
+      "learning_rate": 1.9498370854892882e-05,
+      "loss": 0.4675,
       "step": 3895
     },
     {
-      "epoch": 0.5714285714285714,
-      "grad_norm": 0.39537459394340074,
-      "learning_rate": 1.95056536113344e-05,
-      "loss": 0.4664,
+      "epoch": 0.5741204180774326,
+      "grad_norm": 0.40014580513422415,
+      "learning_rate": 1.9495688897898252e-05,
+      "loss": 0.4606,
       "step": 3900
     },
     {
-      "epoch": 0.5721611721611721,
-      "grad_norm": 0.39366888693656604,
-      "learning_rate": 1.9503003113865112e-05,
-      "loss": 0.4516,
+      "epoch": 0.5748564698954807,
+      "grad_norm": 0.3966916281178557,
+      "learning_rate": 1.9492999975918655e-05,
+      "loss": 0.4704,
       "step": 3905
     },
     {
-      "epoch": 0.5728937728937729,
-      "grad_norm": 0.3782777435360429,
-      "learning_rate": 1.9500345710956278e-05,
-      "loss": 0.4397,
+      "epoch": 0.5755925217135286,
+      "grad_norm": 0.3888609942340231,
+      "learning_rate": 1.9490304090926388e-05,
+      "loss": 0.4663,
       "step": 3910
     },
     {
-      "epoch": 0.5736263736263736,
-      "grad_norm": 0.4016868473347594,
-      "learning_rate": 1.949768140453892e-05,
-      "loss": 0.4573,
+      "epoch": 0.5763285735315766,
+      "grad_norm": 0.39360525777740146,
+      "learning_rate": 1.948760124489885e-05,
+      "loss": 0.4682,
       "step": 3915
     },
     {
-      "epoch": 0.5743589743589743,
-      "grad_norm": 0.40596335251444,
-      "learning_rate": 1.9495010196549082e-05,
-      "loss": 0.4649,
+      "epoch": 0.5770646253496247,
+      "grad_norm": 0.37978437058213976,
+      "learning_rate": 1.948489143981855e-05,
+      "loss": 0.4394,
       "step": 3920
     },
     {
-      "epoch": 0.575091575091575,
-      "grad_norm": 0.39015529793982856,
-      "learning_rate": 1.9492332088927818e-05,
-      "loss": 0.4516,
+      "epoch": 0.5778006771676726,
+      "grad_norm": 0.4076077124842079,
+      "learning_rate": 1.9482174677673104e-05,
+      "loss": 0.4319,
       "step": 3925
     },
     {
-      "epoch": 0.5758241758241758,
-      "grad_norm": 0.38703627226223575,
-      "learning_rate": 1.94896470836212e-05,
-      "loss": 0.4557,
+      "epoch": 0.5785367289857206,
+      "grad_norm": 0.4150972595842849,
+      "learning_rate": 1.947945096045522e-05,
+      "loss": 0.4487,
       "step": 3930
     },
     {
-      "epoch": 0.5765567765567765,
-      "grad_norm": 0.39090512141444317,
-      "learning_rate": 1.948695518258031e-05,
-      "loss": 0.4496,
+      "epoch": 0.5792727808037685,
+      "grad_norm": 0.4588906803973213,
+      "learning_rate": 1.9476720290162724e-05,
+      "loss": 0.4616,
       "step": 3935
     },
     {
-      "epoch": 0.5772893772893772,
-      "grad_norm": 0.3959545873598666,
-      "learning_rate": 1.948425638776124e-05,
-      "loss": 0.4554,
+      "epoch": 0.5800088326218166,
+      "grad_norm": 0.41074938388442134,
+      "learning_rate": 1.947398266879853e-05,
+      "loss": 0.4666,
       "step": 3940
     },
     {
-      "epoch": 0.578021978021978,
-      "grad_norm": 0.394645558625364,
-      "learning_rate": 1.9481550701125095e-05,
-      "loss": 0.4622,
+      "epoch": 0.5807448844398646,
+      "grad_norm": 0.3933742489177997,
+      "learning_rate": 1.947123809837065e-05,
+      "loss": 0.4609,
       "step": 3945
     },
     {
-      "epoch": 0.5787545787545788,
-      "grad_norm": 0.38045088173518693,
-      "learning_rate": 1.9478838124637982e-05,
-      "loss": 0.4321,
+      "epoch": 0.5814809362579125,
+      "grad_norm": 0.3909008868409509,
+      "learning_rate": 1.9468486580892204e-05,
+      "loss": 0.4537,
       "step": 3950
     },
     {
-      "epoch": 0.5794871794871795,
-      "grad_norm": 0.417808747673523,
-      "learning_rate": 1.9476118660271023e-05,
-      "loss": 0.4543,
+      "epoch": 0.5822169880759606,
+      "grad_norm": 0.3974581298450175,
+      "learning_rate": 1.9465728118381388e-05,
+      "loss": 0.462,
       "step": 3955
     },
     {
-      "epoch": 0.5802197802197803,
-      "grad_norm": 0.40359896669430806,
-      "learning_rate": 1.9473392310000337e-05,
-      "loss": 0.4615,
+      "epoch": 0.5829530398940085,
+      "grad_norm": 0.43181893833837104,
+      "learning_rate": 1.9462962712861518e-05,
+      "loss": 0.4716,
       "step": 3960
     },
     {
-      "epoch": 0.580952380952381,
-      "grad_norm": 0.4001322368419974,
-      "learning_rate": 1.947065907580705e-05,
-      "loss": 0.4521,
+      "epoch": 0.5836890917120565,
+      "grad_norm": 0.39145520327042055,
+      "learning_rate": 1.946019036636098e-05,
+      "loss": 0.4915,
       "step": 3965
     },
     {
-      "epoch": 0.5816849816849817,
-      "grad_norm": 0.40688228006534505,
-      "learning_rate": 1.9467918959677294e-05,
-      "loss": 0.4522,
+      "epoch": 0.5844251435301046,
+      "grad_norm": 0.4547161522085445,
+      "learning_rate": 1.9457411080913267e-05,
+      "loss": 0.473,
       "step": 3970
     },
     {
-      "epoch": 0.5824175824175825,
-      "grad_norm": 0.4085036595801529,
-      "learning_rate": 1.9465171963602193e-05,
-      "loss": 0.4577,
+      "epoch": 0.5851611953481525,
+      "grad_norm": 0.37696073979755373,
+      "learning_rate": 1.945462485855695e-05,
+      "loss": 0.4642,
       "step": 3975
     },
     {
-      "epoch": 0.5831501831501832,
-      "grad_norm": 0.3675283160404854,
-      "learning_rate": 1.9462418089577876e-05,
-      "loss": 0.449,
+      "epoch": 0.5858972471662005,
+      "grad_norm": 0.3850517375475203,
+      "learning_rate": 1.945183170133569e-05,
+      "loss": 0.4446,
       "step": 3980
     },
     {
-      "epoch": 0.5838827838827839,
-      "grad_norm": 0.39340033926463114,
-      "learning_rate": 1.9459657339605475e-05,
-      "loss": 0.442,
+      "epoch": 0.5866332989842485,
+      "grad_norm": 0.3896736319504672,
+      "learning_rate": 1.9449031611298245e-05,
+      "loss": 0.4613,
       "step": 3985
     },
     {
-      "epoch": 0.5846153846153846,
-      "grad_norm": 0.38981845508151175,
-      "learning_rate": 1.9456889715691107e-05,
-      "loss": 0.4502,
+      "epoch": 0.5873693508022965,
+      "grad_norm": 0.37559385194345996,
+      "learning_rate": 1.9446224590498447e-05,
+      "loss": 0.4565,
       "step": 3990
     },
     {
-      "epoch": 0.5853479853479854,
-      "grad_norm": 0.39502082453513393,
-      "learning_rate": 1.9454115219845895e-05,
-      "loss": 0.4892,
+      "epoch": 0.5881054026203445,
+      "grad_norm": 0.4316997268846546,
+      "learning_rate": 1.9443410640995213e-05,
+      "loss": 0.45,
       "step": 3995
     },
     {
-      "epoch": 0.5860805860805861,
-      "grad_norm": 0.38910933952745375,
-      "learning_rate": 1.9451333854085945e-05,
-      "loss": 0.4575,
+      "epoch": 0.5888414544383924,
+      "grad_norm": 0.4020080816861019,
+      "learning_rate": 1.9440589764852552e-05,
+      "loss": 0.457,
       "step": 4000
     },
     {
-      "epoch": 0.5868131868131868,
-      "grad_norm": 0.3872505317049606,
-      "learning_rate": 1.9448545620432364e-05,
-      "loss": 0.4572,
+      "epoch": 0.5895775062564405,
+      "grad_norm": 0.3968685516640876,
+      "learning_rate": 1.943776196413954e-05,
+      "loss": 0.4549,
       "step": 4005
     },
     {
-      "epoch": 0.5875457875457876,
-      "grad_norm": 0.4154623026662341,
-      "learning_rate": 1.9445750520911245e-05,
-      "loss": 0.4364,
+      "epoch": 0.5903135580744885,
+      "grad_norm": 0.4383342608575663,
+      "learning_rate": 1.9434927240930336e-05,
+      "loss": 0.4515,
       "step": 4010
     },
     {
-      "epoch": 0.5882783882783883,
-      "grad_norm": 0.40700084808831655,
-      "learning_rate": 1.944294855755367e-05,
-      "loss": 0.4646,
+      "epoch": 0.5910496098925364,
+      "grad_norm": 0.39025368242219294,
+      "learning_rate": 1.9432085597304184e-05,
+      "loss": 0.4563,
       "step": 4015
     },
     {
-      "epoch": 0.589010989010989,
-      "grad_norm": 0.38914455931333786,
-      "learning_rate": 1.9440139732395712e-05,
-      "loss": 0.4498,
+      "epoch": 0.5917856617105844,
+      "grad_norm": 0.38303033158729466,
+      "learning_rate": 1.9429237035345396e-05,
+      "loss": 0.4507,
       "step": 4020
     },
     {
-      "epoch": 0.5897435897435898,
-      "grad_norm": 0.3927240714457741,
-      "learning_rate": 1.9437324047478426e-05,
-      "loss": 0.4706,
+      "epoch": 0.5925217135286324,
+      "grad_norm": 0.3715174410319025,
+      "learning_rate": 1.9426381557143364e-05,
+      "loss": 0.4637,
       "step": 4025
     },
     {
-      "epoch": 0.5904761904761905,
-      "grad_norm": 0.38701060865579207,
-      "learning_rate": 1.943450150484786e-05,
-      "loss": 0.4518,
+      "epoch": 0.5932577653466804,
+      "grad_norm": 0.425662209279403,
+      "learning_rate": 1.9423519164792547e-05,
+      "loss": 0.4546,
       "step": 4030
     },
     {
-      "epoch": 0.5912087912087912,
-      "grad_norm": 0.37099875040863006,
-      "learning_rate": 1.9431672106555033e-05,
-      "loss": 0.4522,
+      "epoch": 0.5939938171647284,
+      "grad_norm": 0.3821996726394133,
+      "learning_rate": 1.942064986039248e-05,
+      "loss": 0.4449,
       "step": 4035
     },
     {
-      "epoch": 0.591941391941392,
-      "grad_norm": 0.3931898441591413,
-      "learning_rate": 1.942883585465595e-05,
-      "loss": 0.4649,
+      "epoch": 0.5947298689827764,
+      "grad_norm": 0.39339832871267766,
+      "learning_rate": 1.941777364604777e-05,
+      "loss": 0.4543,
       "step": 4040
     },
     {
-      "epoch": 0.5926739926739927,
-      "grad_norm": 0.390981673455716,
-      "learning_rate": 1.9425992751211605e-05,
-      "loss": 0.4564,
+      "epoch": 0.5954659208008244,
+      "grad_norm": 0.40259414666017407,
+      "learning_rate": 1.9414890523868084e-05,
+      "loss": 0.4654,
       "step": 4045
     },
     {
-      "epoch": 0.5934065934065934,
-      "grad_norm": 0.3851165832154698,
-      "learning_rate": 1.9423142798287958e-05,
-      "loss": 0.4556,
+      "epoch": 0.5962019726188724,
+      "grad_norm": 0.3880847307330742,
+      "learning_rate": 1.9412000495968165e-05,
+      "loss": 0.4596,
       "step": 4050
     },
     {
-      "epoch": 0.5941391941391941,
-      "grad_norm": 0.6688175508134159,
-      "learning_rate": 1.9420285997955953e-05,
-      "loss": 0.4733,
+      "epoch": 0.5969380244369203,
+      "grad_norm": 0.398812390374236,
+      "learning_rate": 1.9409103564467813e-05,
+      "loss": 0.4561,
       "step": 4055
     },
     {
-      "epoch": 0.5948717948717949,
-      "grad_norm": 0.38458780122196795,
-      "learning_rate": 1.941742235229151e-05,
-      "loss": 0.4545,
+      "epoch": 0.5976740762549684,
+      "grad_norm": 0.4026502125088563,
+      "learning_rate": 1.9406199731491898e-05,
+      "loss": 0.4623,
       "step": 4060
     },
     {
-      "epoch": 0.5956043956043956,
-      "grad_norm": 0.4032597652712211,
-      "learning_rate": 1.9414551863375524e-05,
-      "loss": 0.4498,
+      "epoch": 0.5984101280730163,
+      "grad_norm": 0.39108455961406485,
+      "learning_rate": 1.9403288999170353e-05,
+      "loss": 0.4559,
       "step": 4065
     },
     {
-      "epoch": 0.5963369963369963,
-      "grad_norm": 0.45669759618264094,
-      "learning_rate": 1.941167453329386e-05,
-      "loss": 0.4591,
+      "epoch": 0.5991461798910643,
+      "grad_norm": 0.3927115580987639,
+      "learning_rate": 1.9400371369638164e-05,
+      "loss": 0.4596,
       "step": 4070
     },
     {
-      "epoch": 0.5970695970695971,
-      "grad_norm": 0.4024185340241739,
-      "learning_rate": 1.9408790364137345e-05,
-      "loss": 0.4642,
+      "epoch": 0.5998822317091124,
+      "grad_norm": 0.38912467663104383,
+      "learning_rate": 1.9397446845035384e-05,
+      "loss": 0.4533,
       "step": 4075
     },
     {
-      "epoch": 0.5978021978021978,
-      "grad_norm": 0.3828519017237699,
-      "learning_rate": 1.94058993580018e-05,
-      "loss": 0.4671,
+      "epoch": 0.6006182835271603,
+      "grad_norm": 0.3971049468366541,
+      "learning_rate": 1.9394515427507117e-05,
+      "loss": 0.4434,
       "step": 4080
     },
     {
-      "epoch": 0.5985347985347985,
-      "grad_norm": 0.4034508466964574,
-      "learning_rate": 1.9403001516987992e-05,
-      "loss": 0.4558,
+      "epoch": 0.6013543353452083,
+      "grad_norm": 0.48869481454910163,
+      "learning_rate": 1.939157711920353e-05,
+      "loss": 0.4528,
       "step": 4085
     },
     {
-      "epoch": 0.5992673992673992,
-      "grad_norm": 0.38983586585639907,
-      "learning_rate": 1.9400096843201665e-05,
-      "loss": 0.4697,
+      "epoch": 0.6020903871632562,
+      "grad_norm": 0.39438450042318346,
+      "learning_rate": 1.9388631922279835e-05,
+      "loss": 0.4719,
       "step": 4090
     },
     {
-      "epoch": 0.6,
-      "grad_norm": 0.38934409647363133,
-      "learning_rate": 1.939718533875352e-05,
-      "loss": 0.4674,
+      "epoch": 0.6028264389813043,
+      "grad_norm": 0.3889958357897293,
+      "learning_rate": 1.9385679838896304e-05,
+      "loss": 0.4533,
       "step": 4095
     },
     {
-      "epoch": 0.6007326007326007,
-      "grad_norm": 0.40769464285287194,
-      "learning_rate": 1.9394267005759236e-05,
-      "loss": 0.4743,
+      "epoch": 0.6035624907993523,
+      "grad_norm": 0.4065933272164289,
+      "learning_rate": 1.938272087121826e-05,
+      "loss": 0.4566,
       "step": 4100
     },
     {
-      "epoch": 0.6014652014652014,
-      "grad_norm": 0.418201033254751,
-      "learning_rate": 1.9391341846339435e-05,
-      "loss": 0.4559,
+      "epoch": 0.6042985426174002,
+      "grad_norm": 0.40036695766534475,
+      "learning_rate": 1.9379755021416077e-05,
+      "loss": 0.4684,
       "step": 4105
     },
     {
-      "epoch": 0.6021978021978022,
-      "grad_norm": 0.39055369127735,
-      "learning_rate": 1.9388409862619716e-05,
-      "loss": 0.4668,
+      "epoch": 0.6050345944354483,
+      "grad_norm": 0.4185967221049071,
+      "learning_rate": 1.9376782291665165e-05,
+      "loss": 0.4687,
       "step": 4110
     },
     {
-      "epoch": 0.6029304029304029,
-      "grad_norm": 0.39385123346625744,
-      "learning_rate": 1.938547105673063e-05,
-      "loss": 0.4493,
+      "epoch": 0.6057706462534962,
+      "grad_norm": 0.3647349786411818,
+      "learning_rate": 1.9373802684145997e-05,
+      "loss": 0.43,
       "step": 4115
     },
     {
-      "epoch": 0.6036630036630036,
-      "grad_norm": 0.3778931973342424,
-      "learning_rate": 1.9382525430807684e-05,
-      "loss": 0.4612,
+      "epoch": 0.6065066980715442,
+      "grad_norm": 0.37151788676087305,
+      "learning_rate": 1.9370816201044074e-05,
+      "loss": 0.438,
       "step": 4120
     },
     {
-      "epoch": 0.6043956043956044,
-      "grad_norm": 0.3653423338067703,
-      "learning_rate": 1.9379572986991342e-05,
-      "loss": 0.4517,
+      "epoch": 0.6072427498895923,
+      "grad_norm": 0.398271370044487,
+      "learning_rate": 1.9367822844549963e-05,
+      "loss": 0.4749,
       "step": 4125
     },
     {
-      "epoch": 0.6051282051282051,
-      "grad_norm": 0.3823440864145797,
-      "learning_rate": 1.9376613727427025e-05,
-      "loss": 0.4806,
+      "epoch": 0.6079788017076402,
+      "grad_norm": 0.3891198356257357,
+      "learning_rate": 1.9364822616859243e-05,
+      "loss": 0.4516,
       "step": 4130
     },
     {
-      "epoch": 0.6058608058608058,
-      "grad_norm": 0.3942364943467247,
-      "learning_rate": 1.93736476542651e-05,
-      "loss": 0.4692,
+      "epoch": 0.6087148535256882,
+      "grad_norm": 0.39378864796931595,
+      "learning_rate": 1.936181552017256e-05,
+      "loss": 0.4797,
       "step": 4135
     },
     {
-      "epoch": 0.6065934065934065,
-      "grad_norm": 0.413109206537163,
-      "learning_rate": 1.93706747696609e-05,
-      "loss": 0.4471,
+      "epoch": 0.6094509053437362,
+      "grad_norm": 0.3933386067929528,
+      "learning_rate": 1.935880155669558e-05,
+      "loss": 0.4619,
       "step": 4140
     },
     {
-      "epoch": 0.6073260073260073,
-      "grad_norm": 0.39856129006035784,
-      "learning_rate": 1.936769507577468e-05,
-      "loss": 0.4596,
+      "epoch": 0.6101869571617842,
+      "grad_norm": 0.37594319073788085,
+      "learning_rate": 1.935578072863902e-05,
+      "loss": 0.4587,
       "step": 4145
     },
     {
-      "epoch": 0.608058608058608,
-      "grad_norm": 0.41314992268721973,
-      "learning_rate": 1.936470857477168e-05,
-      "loss": 0.4552,
+      "epoch": 0.6109230089798322,
+      "grad_norm": 0.40009248916342544,
+      "learning_rate": 1.935275303821861e-05,
+      "loss": 0.4795,
       "step": 4150
     },
     {
-      "epoch": 0.6087912087912087,
-      "grad_norm": 0.3914732298327829,
-      "learning_rate": 1.9361715268822056e-05,
-      "loss": 0.4686,
+      "epoch": 0.6116590607978801,
+      "grad_norm": 0.3877757563012712,
+      "learning_rate": 1.9349718487655145e-05,
+      "loss": 0.4585,
       "step": 4155
     },
     {
-      "epoch": 0.6095238095238096,
-      "grad_norm": 0.3889002217098044,
-      "learning_rate": 1.935871516010092e-05,
-      "loss": 0.4724,
+      "epoch": 0.6123951126159282,
+      "grad_norm": 0.39613251424470064,
+      "learning_rate": 1.934667707917443e-05,
+      "loss": 0.449,
       "step": 4160
     },
     {
-      "epoch": 0.6102564102564103,
-      "grad_norm": 0.38348773145064013,
-      "learning_rate": 1.9355708250788326e-05,
-      "loss": 0.444,
+      "epoch": 0.6131311644339762,
+      "grad_norm": 0.4474668269916825,
+      "learning_rate": 1.9343628815007294e-05,
+      "loss": 0.457,
       "step": 4165
     },
     {
-      "epoch": 0.610989010989011,
-      "grad_norm": 0.45372994439130565,
-      "learning_rate": 1.9352694543069278e-05,
-      "loss": 0.4467,
+      "epoch": 0.6138672162520241,
+      "grad_norm": 0.398639026676931,
+      "learning_rate": 1.9340573697389624e-05,
+      "loss": 0.4688,
       "step": 4170
     },
     {
-      "epoch": 0.6117216117216118,
-      "grad_norm": 0.4063317369558111,
-      "learning_rate": 1.9349674039133706e-05,
-      "loss": 0.4899,
+      "epoch": 0.6146032680700722,
+      "grad_norm": 0.39381851068614493,
+      "learning_rate": 1.9337511728562304e-05,
+      "loss": 0.4627,
       "step": 4175
     },
     {
-      "epoch": 0.6124542124542125,
-      "grad_norm": 0.38693500581456275,
-      "learning_rate": 1.9346646741176487e-05,
-      "loss": 0.4607,
+      "epoch": 0.6153393198881201,
+      "grad_norm": 0.3847963629778539,
+      "learning_rate": 1.9334442910771252e-05,
+      "loss": 0.4456,
       "step": 4180
     },
     {
-      "epoch": 0.6131868131868132,
-      "grad_norm": 0.4298047483385306,
-      "learning_rate": 1.9343612651397435e-05,
-      "loss": 0.458,
+      "epoch": 0.6160753717061681,
+      "grad_norm": 0.3906474233560638,
+      "learning_rate": 1.9331367246267425e-05,
+      "loss": 0.4397,
       "step": 4185
     },
     {
-      "epoch": 0.613919413919414,
-      "grad_norm": 0.39996348193380593,
-      "learning_rate": 1.9340571772001295e-05,
-      "loss": 0.458,
+      "epoch": 0.6168114235242161,
+      "grad_norm": 0.40814516018326935,
+      "learning_rate": 1.932828473730678e-05,
+      "loss": 0.4547,
       "step": 4190
     },
     {
-      "epoch": 0.6146520146520147,
-      "grad_norm": 0.4093122753079218,
-      "learning_rate": 1.933752410519775e-05,
-      "loss": 0.4482,
+      "epoch": 0.6175474753422641,
+      "grad_norm": 0.39056453129841734,
+      "learning_rate": 1.9325195386150305e-05,
+      "loss": 0.4419,
       "step": 4195
     },
     {
-      "epoch": 0.6153846153846154,
-      "grad_norm": 0.38555333230727856,
-      "learning_rate": 1.9334469653201412e-05,
-      "loss": 0.4591,
+      "epoch": 0.6182835271603121,
+      "grad_norm": 0.3708542016030073,
+      "learning_rate": 1.932209919506401e-05,
+      "loss": 0.4402,
       "step": 4200
     },
     {
-      "epoch": 0.6161172161172161,
-      "grad_norm": 0.37088920956413135,
-      "learning_rate": 1.9331408418231826e-05,
-      "loss": 0.4534,
+      "epoch": 0.6190195789783601,
+      "grad_norm": 0.3906520277972606,
+      "learning_rate": 1.9318996166318915e-05,
+      "loss": 0.4549,
       "step": 4205
     },
     {
-      "epoch": 0.6168498168498169,
-      "grad_norm": 0.4050551280386644,
-      "learning_rate": 1.9328340402513467e-05,
-      "loss": 0.4523,
+      "epoch": 0.619755630796408,
+      "grad_norm": 0.3873093062714683,
+      "learning_rate": 1.9315886302191056e-05,
+      "loss": 0.4529,
       "step": 4210
     },
     {
-      "epoch": 0.6175824175824176,
-      "grad_norm": 0.44848146017548013,
-      "learning_rate": 1.9325265608275736e-05,
-      "loss": 0.4624,
+      "epoch": 0.6204916826144561,
+      "grad_norm": 0.3724665558860097,
+      "learning_rate": 1.9312769604961486e-05,
+      "loss": 0.4708,
       "step": 4215
     },
     {
-      "epoch": 0.6183150183150183,
-      "grad_norm": 0.3832437196495983,
-      "learning_rate": 1.932218403775295e-05,
-      "loss": 0.4483,
+      "epoch": 0.621227734432504,
+      "grad_norm": 0.3789009595800291,
+      "learning_rate": 1.930964607691627e-05,
+      "loss": 0.4558,
       "step": 4220
     },
     {
-      "epoch": 0.6190476190476191,
-      "grad_norm": 0.38659377362158687,
-      "learning_rate": 1.9319095693184367e-05,
-      "loss": 0.4601,
+      "epoch": 0.621963786250552,
+      "grad_norm": 0.37528776730234276,
+      "learning_rate": 1.9306515720346485e-05,
+      "loss": 0.4578,
       "step": 4225
     },
     {
-      "epoch": 0.6197802197802198,
-      "grad_norm": 0.38946115154622146,
-      "learning_rate": 1.931600057681416e-05,
-      "loss": 0.4665,
+      "epoch": 0.6226998380686001,
+      "grad_norm": 0.39111547258192264,
+      "learning_rate": 1.93033785375482e-05,
+      "loss": 0.4421,
       "step": 4230
     },
     {
-      "epoch": 0.6205128205128205,
-      "grad_norm": 0.396423053727006,
-      "learning_rate": 1.9312898690891416e-05,
-      "loss": 0.4485,
+      "epoch": 0.623435889886648,
+      "grad_norm": 0.38703708176094,
+      "learning_rate": 1.9300234530822522e-05,
+      "loss": 0.4691,
       "step": 4235
     },
     {
-      "epoch": 0.6212454212454213,
-      "grad_norm": 0.36984615870705156,
-      "learning_rate": 1.9309790037670152e-05,
-      "loss": 0.4681,
+      "epoch": 0.624171941704696,
+      "grad_norm": 0.40072906015264564,
+      "learning_rate": 1.929708370247554e-05,
+      "loss": 0.4407,
       "step": 4240
     },
     {
-      "epoch": 0.621978021978022,
-      "grad_norm": 0.3906259863054413,
-      "learning_rate": 1.9306674619409297e-05,
-      "loss": 0.4562,
+      "epoch": 0.624907993522744,
+      "grad_norm": 0.4031722160372937,
+      "learning_rate": 1.9293926054818346e-05,
+      "loss": 0.4685,
       "step": 4245
     },
     {
-      "epoch": 0.6227106227106227,
-      "grad_norm": 0.3777272852941437,
-      "learning_rate": 1.9303552438372698e-05,
-      "loss": 0.4423,
+      "epoch": 0.625644045340792,
+      "grad_norm": 0.39123399596163366,
+      "learning_rate": 1.9290761590167047e-05,
+      "loss": 0.4409,
       "step": 4250
     },
     {
-      "epoch": 0.6234432234432234,
-      "grad_norm": 0.3685244570964051,
-      "learning_rate": 1.9300423496829112e-05,
-      "loss": 0.4373,
+      "epoch": 0.62638009715884,
+      "grad_norm": 0.3931115302109413,
+      "learning_rate": 1.9287590310842742e-05,
+      "loss": 0.4546,
       "step": 4255
     },
     {
-      "epoch": 0.6241758241758242,
-      "grad_norm": 0.3942685902577395,
-      "learning_rate": 1.929728779705221e-05,
-      "loss": 0.4655,
+      "epoch": 0.6271161489768879,
+      "grad_norm": 0.38239123144484477,
+      "learning_rate": 1.9284412219171527e-05,
+      "loss": 0.451,
       "step": 4260
     },
     {
-      "epoch": 0.6249084249084249,
-      "grad_norm": 0.3903998305825669,
-      "learning_rate": 1.929414534132058e-05,
-      "loss": 0.4307,
+      "epoch": 0.627852200794936,
+      "grad_norm": 0.38417626777381236,
+      "learning_rate": 1.9281227317484505e-05,
+      "loss": 0.4561,
       "step": 4265
     },
     {
-      "epoch": 0.6256410256410256,
-      "grad_norm": 0.40182735650557905,
-      "learning_rate": 1.9290996131917712e-05,
-      "loss": 0.4333,
+      "epoch": 0.6285882526129839,
+      "grad_norm": 0.3815448864691956,
+      "learning_rate": 1.9278035608117757e-05,
+      "loss": 0.4703,
       "step": 4270
     },
     {
-      "epoch": 0.6263736263736264,
-      "grad_norm": 0.40889635313254835,
-      "learning_rate": 1.9287840171132007e-05,
-      "loss": 0.4669,
+      "epoch": 0.6293243044310319,
+      "grad_norm": 0.3826401045233026,
+      "learning_rate": 1.9274837093412377e-05,
+      "loss": 0.4466,
       "step": 4275
     },
     {
-      "epoch": 0.6271062271062271,
-      "grad_norm": 0.38070505593169707,
-      "learning_rate": 1.9284677461256774e-05,
-      "loss": 0.4553,
+      "epoch": 0.63006035624908,
+      "grad_norm": 0.408795179267246,
+      "learning_rate": 1.9271631775714435e-05,
+      "loss": 0.4383,
       "step": 4280
     },
     {
-      "epoch": 0.6278388278388278,
-      "grad_norm": 0.38871611338253864,
-      "learning_rate": 1.928150800459022e-05,
-      "loss": 0.44,
+      "epoch": 0.6307964080671279,
+      "grad_norm": 0.3738144718546381,
+      "learning_rate": 1.9268419657374998e-05,
+      "loss": 0.4628,
       "step": 4285
     },
     {
-      "epoch": 0.6285714285714286,
-      "grad_norm": 0.3996118603513095,
-      "learning_rate": 1.9278331803435456e-05,
-      "loss": 0.4554,
+      "epoch": 0.6315324598851759,
+      "grad_norm": 0.41880972530769106,
+      "learning_rate": 1.926520074075012e-05,
+      "loss": 0.4519,
       "step": 4290
     },
     {
-      "epoch": 0.6293040293040293,
-      "grad_norm": 0.39601564233233194,
-      "learning_rate": 1.92751488601005e-05,
-      "loss": 0.4575,
+      "epoch": 0.632268511703224,
+      "grad_norm": 0.3929317274722913,
+      "learning_rate": 1.9261975028200845e-05,
+      "loss": 0.4562,
       "step": 4295
     },
     {
-      "epoch": 0.63003663003663,
-      "grad_norm": 0.3861021385250994,
-      "learning_rate": 1.9271959176898266e-05,
-      "loss": 0.4486,
+      "epoch": 0.6330045635212719,
+      "grad_norm": 0.3941719954305297,
+      "learning_rate": 1.9258742522093193e-05,
+      "loss": 0.4535,
       "step": 4300
     },
     {
-      "epoch": 0.6307692307692307,
-      "grad_norm": 0.3889731285104663,
-      "learning_rate": 1.9268762756146563e-05,
-      "loss": 0.4657,
+      "epoch": 0.6337406153393199,
+      "grad_norm": 0.3962349391567529,
+      "learning_rate": 1.9255503224798174e-05,
+      "loss": 0.4411,
       "step": 4305
     },
     {
-      "epoch": 0.6315018315018315,
-      "grad_norm": 0.3694210604845663,
-      "learning_rate": 1.9265559600168097e-05,
-      "loss": 0.4494,
+      "epoch": 0.6344766671573678,
+      "grad_norm": 0.39061925912957995,
+      "learning_rate": 1.925225713869178e-05,
+      "loss": 0.4616,
       "step": 4310
     },
     {
-      "epoch": 0.6322344322344322,
-      "grad_norm": 0.39217980261947694,
-      "learning_rate": 1.926234971129047e-05,
-      "loss": 0.45,
+      "epoch": 0.6352127189754159,
+      "grad_norm": 0.39295510353538005,
+      "learning_rate": 1.9249004266154973e-05,
+      "loss": 0.4517,
       "step": 4315
     },
     {
-      "epoch": 0.6329670329670329,
-      "grad_norm": 0.3806373972050949,
-      "learning_rate": 1.9259133091846175e-05,
-      "loss": 0.4551,
+      "epoch": 0.6359487707934639,
+      "grad_norm": 0.3833489356645117,
+      "learning_rate": 1.924574460957371e-05,
+      "loss": 0.4549,
       "step": 4320
     },
     {
-      "epoch": 0.6336996336996337,
-      "grad_norm": 0.39546782249074924,
-      "learning_rate": 1.92559097441726e-05,
-      "loss": 0.4525,
+      "epoch": 0.6366848226115118,
+      "grad_norm": 0.40517556279328376,
+      "learning_rate": 1.9242478171338903e-05,
+      "loss": 0.4553,
       "step": 4325
     },
     {
-      "epoch": 0.6344322344322344,
-      "grad_norm": 0.377373667713864,
-      "learning_rate": 1.9252679670612012e-05,
-      "loss": 0.4503,
+      "epoch": 0.6374208744295599,
+      "grad_norm": 0.3922899929303947,
+      "learning_rate": 1.9239204953846456e-05,
+      "loss": 0.465,
       "step": 4330
     },
     {
-      "epoch": 0.6351648351648351,
-      "grad_norm": 0.38167574564074974,
-      "learning_rate": 1.924944287351158e-05,
-      "loss": 0.4367,
+      "epoch": 0.6381569262476078,
+      "grad_norm": 0.395268937479294,
+      "learning_rate": 1.9235924959497237e-05,
+      "loss": 0.4562,
       "step": 4335
     },
     {
-      "epoch": 0.6358974358974359,
-      "grad_norm": 0.39229244686000464,
-      "learning_rate": 1.9246199355223344e-05,
-      "loss": 0.4479,
+      "epoch": 0.6388929780656558,
+      "grad_norm": 0.3809884756259666,
+      "learning_rate": 1.923263819069709e-05,
+      "loss": 0.4498,
       "step": 4340
     },
     {
-      "epoch": 0.6366300366300366,
-      "grad_norm": 0.39043896265781963,
-      "learning_rate": 1.924294911810424e-05,
-      "loss": 0.4542,
+      "epoch": 0.6396290298837038,
+      "grad_norm": 0.37874360253029615,
+      "learning_rate": 1.922934464985682e-05,
+      "loss": 0.4539,
       "step": 4345
     },
     {
-      "epoch": 0.6373626373626373,
-      "grad_norm": 0.38448022559450346,
-      "learning_rate": 1.923969216451607e-05,
-      "loss": 0.4502,
+      "epoch": 0.6403650817017518,
+      "grad_norm": 0.38874389786317726,
+      "learning_rate": 1.92260443393922e-05,
+      "loss": 0.4511,
       "step": 4350
     },
     {
-      "epoch": 0.638095238095238,
-      "grad_norm": 0.3836800759490605,
-      "learning_rate": 1.9236428496825543e-05,
-      "loss": 0.4598,
+      "epoch": 0.6411011335197998,
+      "grad_norm": 0.39098938560232144,
+      "learning_rate": 1.922273726172398e-05,
+      "loss": 0.4489,
       "step": 4355
     },
     {
-      "epoch": 0.6388278388278388,
-      "grad_norm": 0.37757456425964253,
-      "learning_rate": 1.9233158117404224e-05,
-      "loss": 0.4539,
+      "epoch": 0.6418371853378478,
+      "grad_norm": 0.4129270007904453,
+      "learning_rate": 1.921942341927786e-05,
+      "loss": 0.473,
       "step": 4360
     },
     {
-      "epoch": 0.6395604395604395,
-      "grad_norm": 0.3903909062744032,
-      "learning_rate": 1.922988102862856e-05,
-      "loss": 0.4756,
+      "epoch": 0.6425732371558958,
+      "grad_norm": 0.3942484438457897,
+      "learning_rate": 1.921610281448451e-05,
+      "loss": 0.4671,
       "step": 4365
     },
     {
-      "epoch": 0.6402930402930402,
-      "grad_norm": 0.4178460038790184,
-      "learning_rate": 1.9226597232879876e-05,
-      "loss": 0.4507,
+      "epoch": 0.6433092889739438,
+      "grad_norm": 0.37943256505154993,
+      "learning_rate": 1.921277544977956e-05,
+      "loss": 0.4634,
       "step": 4370
     },
     {
-      "epoch": 0.6410256410256411,
-      "grad_norm": 0.35841063994877587,
-      "learning_rate": 1.9223306732544375e-05,
-      "loss": 0.4541,
+      "epoch": 0.6440453407919917,
+      "grad_norm": 0.36557974071879495,
+      "learning_rate": 1.9209441327603587e-05,
+      "loss": 0.4491,
       "step": 4375
     },
     {
-      "epoch": 0.6417582417582418,
-      "grad_norm": 0.41456546802807315,
-      "learning_rate": 1.9220009530013123e-05,
-      "loss": 0.4571,
+      "epoch": 0.6447813926100397,
+      "grad_norm": 0.3834531941887774,
+      "learning_rate": 1.9206100450402142e-05,
+      "loss": 0.4412,
       "step": 4380
     },
     {
-      "epoch": 0.6424908424908425,
-      "grad_norm": 0.4001438732473654,
-      "learning_rate": 1.921670562768207e-05,
-      "loss": 0.4538,
+      "epoch": 0.6455174444280878,
+      "grad_norm": 0.3739752373908016,
+      "learning_rate": 1.920275282062572e-05,
+      "loss": 0.451,
       "step": 4385
     },
     {
-      "epoch": 0.6432234432234433,
-      "grad_norm": 0.4124870743139893,
-      "learning_rate": 1.9213395027952014e-05,
-      "loss": 0.4454,
+      "epoch": 0.6462534962461357,
+      "grad_norm": 0.3862654437042714,
+      "learning_rate": 1.9199398440729775e-05,
+      "loss": 0.4472,
       "step": 4390
     },
     {
-      "epoch": 0.643956043956044,
-      "grad_norm": 0.39453924510346333,
-      "learning_rate": 1.9210077733228634e-05,
-      "loss": 0.4392,
+      "epoch": 0.6469895480641837,
+      "grad_norm": 0.3798904378427613,
+      "learning_rate": 1.91960373131747e-05,
+      "loss": 0.4631,
       "step": 4395
     },
     {
-      "epoch": 0.6446886446886447,
-      "grad_norm": 0.3869409287281537,
-      "learning_rate": 1.9206753745922474e-05,
-      "loss": 0.4557,
+      "epoch": 0.6477255998822317,
+      "grad_norm": 0.37053600436508793,
+      "learning_rate": 1.919266944042585e-05,
+      "loss": 0.4446,
       "step": 4400
     },
     {
-      "epoch": 0.6454212454212455,
-      "grad_norm": 0.38039297534887273,
-      "learning_rate": 1.920342306844894e-05,
-      "loss": 0.47,
+      "epoch": 0.6484616517002797,
+      "grad_norm": 0.38483557594087386,
+      "learning_rate": 1.9189294824953522e-05,
+      "loss": 0.444,
       "step": 4405
     },
     {
-      "epoch": 0.6461538461538462,
-      "grad_norm": 0.39272104310666833,
-      "learning_rate": 1.920008570322829e-05,
-      "loss": 0.4389,
+      "epoch": 0.6491977035183277,
+      "grad_norm": 0.3774301947302271,
+      "learning_rate": 1.918591346923296e-05,
+      "loss": 0.4772,
       "step": 4410
     },
     {
-      "epoch": 0.6468864468864469,
-      "grad_norm": 0.38105753403110915,
-      "learning_rate": 1.9196741652685656e-05,
-      "loss": 0.4542,
+      "epoch": 0.6499337553363757,
+      "grad_norm": 0.38992859254977136,
+      "learning_rate": 1.9182525375744357e-05,
+      "loss": 0.4553,
       "step": 4415
     },
     {
-      "epoch": 0.6476190476190476,
-      "grad_norm": 0.40105121340517647,
-      "learning_rate": 1.9193390919251018e-05,
-      "loss": 0.4625,
+      "epoch": 0.6506698071544237,
+      "grad_norm": 0.3858816731551356,
+      "learning_rate": 1.917913054697284e-05,
+      "loss": 0.4428,
       "step": 4420
     },
     {
-      "epoch": 0.6483516483516484,
-      "grad_norm": 0.370705169564649,
-      "learning_rate": 1.9190033505359217e-05,
-      "loss": 0.4632,
+      "epoch": 0.6514058589724716,
+      "grad_norm": 0.39217801204631847,
+      "learning_rate": 1.9175728985408475e-05,
+      "loss": 0.4639,
       "step": 4425
     },
     {
-      "epoch": 0.6490842490842491,
-      "grad_norm": 0.3862392288450474,
-      "learning_rate": 1.9186669413449947e-05,
-      "loss": 0.4605,
+      "epoch": 0.6521419107905196,
+      "grad_norm": 0.36854030701391716,
+      "learning_rate": 1.9172320693546275e-05,
+      "loss": 0.4514,
       "step": 4430
     },
     {
-      "epoch": 0.6498168498168498,
-      "grad_norm": 0.3923304956651707,
-      "learning_rate": 1.918329864596775e-05,
-      "loss": 0.4465,
+      "epoch": 0.6528779626085677,
+      "grad_norm": 0.38816202609188655,
+      "learning_rate": 1.916890567388618e-05,
+      "loss": 0.4523,
       "step": 4435
     },
     {
-      "epoch": 0.6505494505494506,
-      "grad_norm": 0.40276713300420847,
-      "learning_rate": 1.917992120536203e-05,
-      "loss": 0.4757,
+      "epoch": 0.6536140144266156,
+      "grad_norm": 0.3924972362295436,
+      "learning_rate": 1.9165483928933076e-05,
+      "loss": 0.4303,
       "step": 4440
     },
     {
-      "epoch": 0.6512820512820513,
-      "grad_norm": 0.3884293748786176,
-      "learning_rate": 1.917653709408703e-05,
-      "loss": 0.4562,
+      "epoch": 0.6543500662446636,
+      "grad_norm": 0.3977940095412845,
+      "learning_rate": 1.9162055461196773e-05,
+      "loss": 0.4334,
       "step": 4445
     },
     {
-      "epoch": 0.652014652014652,
-      "grad_norm": 0.4057062403750789,
-      "learning_rate": 1.9173146314601847e-05,
-      "loss": 0.4601,
+      "epoch": 0.6550861180627117,
+      "grad_norm": 0.3966482233204859,
+      "learning_rate": 1.9158620273192013e-05,
+      "loss": 0.4636,
       "step": 4450
     },
     {
-      "epoch": 0.6527472527472528,
-      "grad_norm": 0.40246283304255126,
-      "learning_rate": 1.9169748869370417e-05,
-      "loss": 0.4515,
+      "epoch": 0.6558221698807596,
+      "grad_norm": 0.3776812728523491,
+      "learning_rate": 1.9155178367438477e-05,
+      "loss": 0.4441,
       "step": 4455
     },
     {
-      "epoch": 0.6534798534798535,
-      "grad_norm": 0.3843492378662538,
-      "learning_rate": 1.916634476086152e-05,
-      "loss": 0.463,
+      "epoch": 0.6565582216988076,
+      "grad_norm": 0.3943576820565779,
+      "learning_rate": 1.9151729746460755e-05,
+      "loss": 0.4471,
       "step": 4460
     },
     {
-      "epoch": 0.6542124542124542,
-      "grad_norm": 0.3684455758527936,
-      "learning_rate": 1.9162933991548792e-05,
-      "loss": 0.4606,
+      "epoch": 0.6572942735168555,
+      "grad_norm": 0.3984824508339129,
+      "learning_rate": 1.9148274412788384e-05,
+      "loss": 0.4491,
       "step": 4465
     },
     {
-      "epoch": 0.654945054945055,
-      "grad_norm": 0.4051332136256459,
-      "learning_rate": 1.9159516563910686e-05,
-      "loss": 0.4693,
+      "epoch": 0.6580303253349036,
+      "grad_norm": 0.379015524155088,
+      "learning_rate": 1.91448123689558e-05,
+      "loss": 0.4516,
       "step": 4470
     },
     {
-      "epoch": 0.6556776556776557,
-      "grad_norm": 0.38075983769129534,
-      "learning_rate": 1.915609248043051e-05,
-      "loss": 0.4528,
+      "epoch": 0.6587663771529516,
+      "grad_norm": 0.41906622965058576,
+      "learning_rate": 1.914134361750239e-05,
+      "loss": 0.4687,
       "step": 4475
     },
     {
-      "epoch": 0.6564102564102564,
-      "grad_norm": 0.3936521233161132,
-      "learning_rate": 1.9152661743596405e-05,
-      "loss": 0.4568,
+      "epoch": 0.6595024289709995,
+      "grad_norm": 0.36877550481815546,
+      "learning_rate": 1.9137868160972436e-05,
+      "loss": 0.4464,
       "step": 4480
     },
     {
-      "epoch": 0.6571428571428571,
-      "grad_norm": 0.4033142175594591,
-      "learning_rate": 1.9149224355901345e-05,
-      "loss": 0.4684,
+      "epoch": 0.6602384807890476,
+      "grad_norm": 0.40075594059466546,
+      "learning_rate": 1.913438600191515e-05,
+      "loss": 0.4645,
       "step": 4485
     },
     {
-      "epoch": 0.6578754578754579,
-      "grad_norm": 0.39258181662510266,
-      "learning_rate": 1.9145780319843138e-05,
-      "loss": 0.4537,
+      "epoch": 0.6609745326070955,
+      "grad_norm": 0.392668878233667,
+      "learning_rate": 1.9130897142884658e-05,
+      "loss": 0.4593,
       "step": 4490
     },
     {
-      "epoch": 0.6586080586080586,
-      "grad_norm": 0.38314748902028634,
-      "learning_rate": 1.9142329637924423e-05,
-      "loss": 0.4463,
+      "epoch": 0.6617105844251435,
+      "grad_norm": 0.3872265344456865,
+      "learning_rate": 1.912740158644e-05,
+      "loss": 0.4638,
       "step": 4495
     },
     {
-      "epoch": 0.6593406593406593,
-      "grad_norm": 0.37835249978141144,
-      "learning_rate": 1.9138872312652665e-05,
-      "loss": 0.4578,
+      "epoch": 0.6624466362431916,
+      "grad_norm": 0.37492574952816243,
+      "learning_rate": 1.9123899335145132e-05,
+      "loss": 0.4392,
       "step": 4500
     },
     {
-      "epoch": 0.6600732600732601,
-      "grad_norm": 0.39233915642483885,
-      "learning_rate": 1.913540834654016e-05,
-      "loss": 0.4425,
+      "epoch": 0.6631826880612395,
+      "grad_norm": 0.37314352488191793,
+      "learning_rate": 1.912039039156891e-05,
+      "loss": 0.4497,
       "step": 4505
     },
     {
-      "epoch": 0.6608058608058608,
-      "grad_norm": 0.3948802796808528,
-      "learning_rate": 1.9131937742104034e-05,
-      "loss": 0.4459,
+      "epoch": 0.6639187398792875,
+      "grad_norm": 0.3740219032255518,
+      "learning_rate": 1.9116874758285113e-05,
+      "loss": 0.4524,
       "step": 4510
     },
     {
-      "epoch": 0.6615384615384615,
-      "grad_norm": 0.3837580929885009,
-      "learning_rate": 1.9128460501866227e-05,
-      "loss": 0.4741,
+      "epoch": 0.6646547916973355,
+      "grad_norm": 0.38093178849398496,
+      "learning_rate": 1.9113352437872416e-05,
+      "loss": 0.4559,
       "step": 4515
     },
     {
-      "epoch": 0.6622710622710622,
-      "grad_norm": 0.39146826558693193,
-      "learning_rate": 1.9124976628353504e-05,
-      "loss": 0.4475,
+      "epoch": 0.6653908435153835,
+      "grad_norm": 0.3942264775237251,
+      "learning_rate": 1.9109823432914407e-05,
+      "loss": 0.465,
       "step": 4520
     },
     {
-      "epoch": 0.663003663003663,
-      "grad_norm": 0.414172032887318,
-      "learning_rate": 1.9121486124097457e-05,
-      "loss": 0.4505,
+      "epoch": 0.6661268953334315,
+      "grad_norm": 0.3826711034606891,
+      "learning_rate": 1.9106287745999567e-05,
+      "loss": 0.4605,
       "step": 4525
     },
     {
-      "epoch": 0.6637362637362637,
-      "grad_norm": 0.3925713507074972,
-      "learning_rate": 1.9117988991634488e-05,
-      "loss": 0.4607,
+      "epoch": 0.6668629471514794,
+      "grad_norm": 0.39725631489634533,
+      "learning_rate": 1.910274537972129e-05,
+      "loss": 0.4551,
       "step": 4530
     },
     {
-      "epoch": 0.6644688644688644,
-      "grad_norm": 0.3568564070162253,
-      "learning_rate": 1.911448523350582e-05,
-      "loss": 0.4512,
+      "epoch": 0.6675989989695275,
+      "grad_norm": 0.38788306742870987,
+      "learning_rate": 1.9099196336677864e-05,
+      "loss": 0.4511,
       "step": 4535
     },
     {
-      "epoch": 0.6652014652014652,
-      "grad_norm": 0.37946337855975426,
-      "learning_rate": 1.911097485225749e-05,
-      "loss": 0.4367,
+      "epoch": 0.6683350507875755,
+      "grad_norm": 0.35713665308184506,
+      "learning_rate": 1.9095640619472467e-05,
+      "loss": 0.451,
       "step": 4540
     },
     {
-      "epoch": 0.6659340659340659,
-      "grad_norm": 0.39191564372249904,
-      "learning_rate": 1.9107457850440343e-05,
-      "loss": 0.447,
+      "epoch": 0.6690711026056234,
+      "grad_norm": 0.36416995161535204,
+      "learning_rate": 1.9092078230713184e-05,
+      "loss": 0.4494,
       "step": 4545
     },
     {
-      "epoch": 0.6666666666666666,
-      "grad_norm": 0.38309908021603784,
-      "learning_rate": 1.910393423061004e-05,
-      "loss": 0.4556,
+      "epoch": 0.6698071544236714,
+      "grad_norm": 0.40150764715503834,
+      "learning_rate": 1.9088509173012985e-05,
+      "loss": 0.4615,
       "step": 4550
     },
     {
-      "epoch": 0.6673992673992674,
-      "grad_norm": 0.3651262663209095,
-      "learning_rate": 1.9100403995327057e-05,
-      "loss": 0.4427,
+      "epoch": 0.6705432062417194,
+      "grad_norm": 0.39907922416888963,
+      "learning_rate": 1.9084933448989734e-05,
+      "loss": 0.4589,
       "step": 4555
     },
     {
-      "epoch": 0.6681318681318681,
-      "grad_norm": 0.384790548297463,
-      "learning_rate": 1.909686714715666e-05,
-      "loss": 0.453,
+      "epoch": 0.6712792580597674,
+      "grad_norm": 0.3794968612658755,
+      "learning_rate": 1.9081351061266194e-05,
+      "loss": 0.4602,
       "step": 4560
     },
     {
-      "epoch": 0.6688644688644688,
-      "grad_norm": 0.36691825995705835,
-      "learning_rate": 1.9093323688668935e-05,
-      "loss": 0.4522,
+      "epoch": 0.6720153098778154,
+      "grad_norm": 0.3958971089527187,
+      "learning_rate": 1.9077762012470004e-05,
+      "loss": 0.4479,
       "step": 4565
     },
     {
-      "epoch": 0.6695970695970695,
-      "grad_norm": 0.3708935734293259,
-      "learning_rate": 1.908977362243877e-05,
-      "loss": 0.4443,
+      "epoch": 0.6727513616958634,
+      "grad_norm": 0.38697737531582765,
+      "learning_rate": 1.9074166305233684e-05,
+      "loss": 0.4464,
       "step": 4570
     },
     {
-      "epoch": 0.6703296703296703,
-      "grad_norm": 0.3674641558501077,
-      "learning_rate": 1.9086216951045846e-05,
-      "loss": 0.4473,
+      "epoch": 0.6734874135139114,
+      "grad_norm": 0.38627229075882585,
+      "learning_rate": 1.9070563942194658e-05,
+      "loss": 0.456,
       "step": 4575
     },
     {
-      "epoch": 0.671062271062271,
-      "grad_norm": 0.37713691526144866,
-      "learning_rate": 1.908265367707465e-05,
-      "loss": 0.4443,
+      "epoch": 0.6742234653319593,
+      "grad_norm": 0.3809169910379719,
+      "learning_rate": 1.906695492599522e-05,
+      "loss": 0.4553,
       "step": 4580
     },
     {
-      "epoch": 0.6717948717948717,
-      "grad_norm": 0.3893239948236134,
-      "learning_rate": 1.907908380311447e-05,
-      "loss": 0.4497,
+      "epoch": 0.6749595171500073,
+      "grad_norm": 0.3926000334263371,
+      "learning_rate": 1.9063339259282532e-05,
+      "loss": 0.4414,
       "step": 4585
     },
     {
-      "epoch": 0.6725274725274726,
-      "grad_norm": 0.3866933308377136,
-      "learning_rate": 1.9075507331759376e-05,
-      "loss": 0.471,
+      "epoch": 0.6756955689680554,
+      "grad_norm": 0.3938132421731681,
+      "learning_rate": 1.905971694470866e-05,
+      "loss": 0.4565,
       "step": 4590
     },
     {
-      "epoch": 0.6732600732600733,
-      "grad_norm": 0.3842129913868808,
-      "learning_rate": 1.9071924265608253e-05,
-      "loss": 0.4486,
+      "epoch": 0.6764316207861033,
+      "grad_norm": 0.40668102910719195,
+      "learning_rate": 1.9056087984930524e-05,
+      "loss": 0.471,
       "step": 4595
     },
     {
-      "epoch": 0.673992673992674,
-      "grad_norm": 0.3914051257219504,
-      "learning_rate": 1.9068334607264765e-05,
-      "loss": 0.4695,
+      "epoch": 0.6771676726041513,
+      "grad_norm": 0.3866271529330411,
+      "learning_rate": 1.905245238260993e-05,
+      "loss": 0.454,
       "step": 4600
     },
     {
-      "epoch": 0.6747252747252748,
-      "grad_norm": 0.36762571486296486,
-      "learning_rate": 1.906473835933736e-05,
-      "loss": 0.442,
+      "epoch": 0.6779037244221994,
+      "grad_norm": 0.3765934301591741,
+      "learning_rate": 1.9048810140413553e-05,
+      "loss": 0.4386,
       "step": 4605
     },
     {
-      "epoch": 0.6754578754578755,
-      "grad_norm": 0.38947108354293425,
-      "learning_rate": 1.9061135524439296e-05,
-      "loss": 0.4547,
+      "epoch": 0.6786397762402473,
+      "grad_norm": 0.3909238186923048,
+      "learning_rate": 1.9045161261012937e-05,
+      "loss": 0.4577,
       "step": 4610
     },
     {
-      "epoch": 0.6761904761904762,
-      "grad_norm": 0.3727955645617599,
-      "learning_rate": 1.9057526105188593e-05,
-      "loss": 0.453,
+      "epoch": 0.6793758280582953,
+      "grad_norm": 0.3560298704169209,
+      "learning_rate": 1.9041505747084497e-05,
+      "loss": 0.4633,
       "step": 4615
     },
     {
-      "epoch": 0.676923076923077,
-      "grad_norm": 0.38259607470463275,
-      "learning_rate": 1.9053910104208073e-05,
-      "loss": 0.4576,
+      "epoch": 0.6801118798763433,
+      "grad_norm": 0.37333945690798426,
+      "learning_rate": 1.9037843601309513e-05,
+      "loss": 0.4483,
       "step": 4620
     },
     {
-      "epoch": 0.6776556776556777,
-      "grad_norm": 0.3720365889474071,
-      "learning_rate": 1.905028752412533e-05,
-      "loss": 0.451,
+      "epoch": 0.6808479316943913,
+      "grad_norm": 0.3749886090724038,
+      "learning_rate": 1.9034174826374126e-05,
+      "loss": 0.4494,
       "step": 4625
     },
     {
-      "epoch": 0.6783882783882784,
-      "grad_norm": 0.3827944226686561,
-      "learning_rate": 1.9046658367572743e-05,
-      "loss": 0.4516,
+      "epoch": 0.6815839835124393,
+      "grad_norm": 0.3815012969349644,
+      "learning_rate": 1.9030499424969346e-05,
+      "loss": 0.4434,
       "step": 4630
     },
     {
-      "epoch": 0.6791208791208792,
-      "grad_norm": 0.39885198041841713,
-      "learning_rate": 1.9043022637187478e-05,
-      "loss": 0.4703,
+      "epoch": 0.6823200353304872,
+      "grad_norm": 0.41318111288780823,
+      "learning_rate": 1.902681739979104e-05,
+      "loss": 0.4528,
       "step": 4635
     },
     {
-      "epoch": 0.6798534798534799,
-      "grad_norm": 0.4834737339082414,
-      "learning_rate": 1.903938033561146e-05,
-      "loss": 0.453,
+      "epoch": 0.6830560871485353,
+      "grad_norm": 0.37845100927533615,
+      "learning_rate": 1.902312875353993e-05,
+      "loss": 0.464,
       "step": 4640
     },
     {
-      "epoch": 0.6805860805860806,
-      "grad_norm": 0.40334535967342927,
-      "learning_rate": 1.9035731465491404e-05,
-      "loss": 0.4645,
+      "epoch": 0.6837921389665832,
+      "grad_norm": 0.37294320606052356,
+      "learning_rate": 1.9019433488921603e-05,
+      "loss": 0.4513,
       "step": 4645
     },
     {
-      "epoch": 0.6813186813186813,
-      "grad_norm": 0.3811030815586793,
-      "learning_rate": 1.9032076029478794e-05,
-      "loss": 0.4396,
+      "epoch": 0.6845281907846312,
+      "grad_norm": 0.4047075759083381,
+      "learning_rate": 1.9015731608646495e-05,
+      "loss": 0.4606,
       "step": 4650
     },
     {
-      "epoch": 0.6820512820512821,
-      "grad_norm": 0.37187644492402144,
-      "learning_rate": 1.9028414030229885e-05,
-      "loss": 0.4481,
+      "epoch": 0.6852642426026793,
+      "grad_norm": 0.37858929794206153,
+      "learning_rate": 1.9012023115429892e-05,
+      "loss": 0.4436,
       "step": 4655
     },
     {
-      "epoch": 0.6827838827838828,
-      "grad_norm": 0.45571558554350805,
-      "learning_rate": 1.902474547040569e-05,
-      "loss": 0.464,
+      "epoch": 0.6860002944207272,
+      "grad_norm": 0.5284243788863681,
+      "learning_rate": 1.9008308011991934e-05,
+      "loss": 0.4651,
       "step": 4660
     },
     {
-      "epoch": 0.6835164835164835,
-      "grad_norm": 0.3898194135816126,
-      "learning_rate": 1.9021070352672018e-05,
-      "loss": 0.4326,
+      "epoch": 0.6867363462387752,
+      "grad_norm": 0.3578598682305859,
+      "learning_rate": 1.900458630105761e-05,
+      "loss": 0.4583,
       "step": 4665
     },
     {
-      "epoch": 0.6842490842490843,
-      "grad_norm": 0.373343816023147,
-      "learning_rate": 1.901738867969941e-05,
-      "loss": 0.4446,
+      "epoch": 0.6874723980568233,
+      "grad_norm": 0.36339367791316657,
+      "learning_rate": 1.9000857985356757e-05,
+      "loss": 0.4299,
       "step": 4670
     },
     {
-      "epoch": 0.684981684981685,
-      "grad_norm": 0.3967216092671176,
-      "learning_rate": 1.9013700454163194e-05,
-      "loss": 0.4621,
+      "epoch": 0.6882084498748712,
+      "grad_norm": 0.41633037540042483,
+      "learning_rate": 1.8997123067624052e-05,
+      "loss": 0.4713,
       "step": 4675
     },
     {
-      "epoch": 0.6857142857142857,
-      "grad_norm": 0.3821556837245336,
-      "learning_rate": 1.9010005678743445e-05,
-      "loss": 0.4521,
+      "epoch": 0.6889445016929192,
+      "grad_norm": 0.3710006501995171,
+      "learning_rate": 1.8993381550599017e-05,
+      "loss": 0.4741,
       "step": 4680
     },
     {
-      "epoch": 0.6864468864468865,
-      "grad_norm": 0.36149493353457285,
-      "learning_rate": 1.9006304356125007e-05,
-      "loss": 0.4532,
+      "epoch": 0.6896805535109671,
+      "grad_norm": 0.38508491341111795,
+      "learning_rate": 1.898963343702601e-05,
+      "loss": 0.4435,
       "step": 4685
     },
     {
-      "epoch": 0.6871794871794872,
-      "grad_norm": 0.3916146499177383,
-      "learning_rate": 1.9002596488997476e-05,
-      "loss": 0.469,
+      "epoch": 0.6904166053290152,
+      "grad_norm": 0.3681664224415523,
+      "learning_rate": 1.8985878729654242e-05,
+      "loss": 0.4392,
       "step": 4690
     },
     {
-      "epoch": 0.6879120879120879,
-      "grad_norm": 0.3697458703049274,
-      "learning_rate": 1.8998882080055208e-05,
-      "loss": 0.4514,
+      "epoch": 0.6911526571470632,
+      "grad_norm": 0.36978501750409004,
+      "learning_rate": 1.898211743123774e-05,
+      "loss": 0.4449,
       "step": 4695
     },
     {
-      "epoch": 0.6886446886446886,
-      "grad_norm": 0.3756463360666905,
-      "learning_rate": 1.899516113199731e-05,
-      "loss": 0.4385,
+      "epoch": 0.6918887089651111,
+      "grad_norm": 0.3981217967566912,
+      "learning_rate": 1.897834954453538e-05,
+      "loss": 0.4601,
       "step": 4700
     },
     {
-      "epoch": 0.6893772893772894,
-      "grad_norm": 0.373606856904118,
-      "learning_rate": 1.899143364752764e-05,
-      "loss": 0.4665,
+      "epoch": 0.6926247607831592,
+      "grad_norm": 0.36252916603801644,
+      "learning_rate": 1.8974575072310865e-05,
+      "loss": 0.4535,
       "step": 4705
     },
     {
-      "epoch": 0.6901098901098901,
-      "grad_norm": 0.3898021195128183,
-      "learning_rate": 1.898769962935481e-05,
-      "loss": 0.4418,
+      "epoch": 0.6933608126012071,
+      "grad_norm": 0.3863137853908461,
+      "learning_rate": 1.8970794017332732e-05,
+      "loss": 0.4546,
       "step": 4710
     },
     {
-      "epoch": 0.6908424908424908,
-      "grad_norm": 0.35755569340730087,
-      "learning_rate": 1.898395908019217e-05,
-      "loss": 0.4443,
+      "epoch": 0.6940968644192551,
+      "grad_norm": 0.42458899935521194,
+      "learning_rate": 1.896700638237434e-05,
+      "loss": 0.4636,
       "step": 4715
     },
     {
-      "epoch": 0.6915750915750916,
-      "grad_norm": 0.3615767437972971,
-      "learning_rate": 1.8980212002757832e-05,
-      "loss": 0.4562,
+      "epoch": 0.6948329162373031,
+      "grad_norm": 0.407244809353056,
+      "learning_rate": 1.896321217021388e-05,
+      "loss": 0.4563,
       "step": 4720
     },
     {
-      "epoch": 0.6923076923076923,
-      "grad_norm": 0.4006984323153798,
-      "learning_rate": 1.8976458399774637e-05,
-      "loss": 0.4211,
+      "epoch": 0.6955689680553511,
+      "grad_norm": 0.38883704310300615,
+      "learning_rate": 1.8959411383634368e-05,
+      "loss": 0.4553,
       "step": 4725
     },
     {
-      "epoch": 0.693040293040293,
-      "grad_norm": 0.42919622224646925,
-      "learning_rate": 1.897269827397017e-05,
-      "loss": 0.4516,
+      "epoch": 0.6963050198733991,
+      "grad_norm": 0.38064156287812523,
+      "learning_rate": 1.8955604025423636e-05,
+      "loss": 0.473,
       "step": 4730
     },
     {
-      "epoch": 0.6937728937728938,
-      "grad_norm": 0.3778206577833697,
-      "learning_rate": 1.8968931628076762e-05,
-      "loss": 0.4626,
+      "epoch": 0.697041071691447,
+      "grad_norm": 0.42671362604664703,
+      "learning_rate": 1.8951790098374342e-05,
+      "loss": 0.4629,
       "step": 4735
     },
     {
-      "epoch": 0.6945054945054945,
-      "grad_norm": 0.39098805861244557,
-      "learning_rate": 1.8965158464831482e-05,
-      "loss": 0.4572,
+      "epoch": 0.6977771235094951,
+      "grad_norm": 0.387449450624925,
+      "learning_rate": 1.894796960528396e-05,
+      "loss": 0.4436,
       "step": 4740
     },
     {
-      "epoch": 0.6952380952380952,
-      "grad_norm": 0.38792694325380767,
-      "learning_rate": 1.8961378786976127e-05,
-      "loss": 0.4484,
+      "epoch": 0.6985131753275431,
+      "grad_norm": 0.46819007252561434,
+      "learning_rate": 1.8944142548954784e-05,
+      "loss": 0.4379,
       "step": 4745
     },
     {
-      "epoch": 0.6959706959706959,
-      "grad_norm": 0.38773365524372716,
-      "learning_rate": 1.8957592597257233e-05,
-      "loss": 0.4427,
+      "epoch": 0.699249227145591,
+      "grad_norm": 0.37059307285585424,
+      "learning_rate": 1.8940308932193908e-05,
+      "loss": 0.4432,
       "step": 4750
     },
     {
-      "epoch": 0.6967032967032967,
-      "grad_norm": 0.40388591163772436,
-      "learning_rate": 1.8953799898426074e-05,
-      "loss": 0.4581,
+      "epoch": 0.699985278963639,
+      "grad_norm": 0.37315949296564255,
+      "learning_rate": 1.893646875781326e-05,
+      "loss": 0.4499,
       "step": 4755
     },
     {
-      "epoch": 0.6974358974358974,
-      "grad_norm": 0.4246963886620943,
-      "learning_rate": 1.895000069323864e-05,
-      "loss": 0.4629,
+      "epoch": 0.7007213307816871,
+      "grad_norm": 0.3998305870136644,
+      "learning_rate": 1.8932622028629557e-05,
+      "loss": 0.4654,
       "step": 4760
     },
     {
-      "epoch": 0.6981684981684981,
-      "grad_norm": 0.3704772638447488,
-      "learning_rate": 1.8946194984455656e-05,
-      "loss": 0.4554,
+      "epoch": 0.701457382599735,
+      "grad_norm": 0.38356851430459643,
+      "learning_rate": 1.892876874746434e-05,
+      "loss": 0.4544,
       "step": 4765
     },
     {
-      "epoch": 0.6989010989010989,
-      "grad_norm": 0.38894583359585044,
-      "learning_rate": 1.894238277484258e-05,
-      "loss": 0.4508,
+      "epoch": 0.702193434417783,
+      "grad_norm": 0.37100455436607055,
+      "learning_rate": 1.8924908917143946e-05,
+      "loss": 0.4521,
       "step": 4770
     },
     {
-      "epoch": 0.6996336996336996,
-      "grad_norm": 0.39049511016031907,
-      "learning_rate": 1.8938564067169584e-05,
-      "loss": 0.4565,
+      "epoch": 0.702929486235831,
+      "grad_norm": 0.38343900260974045,
+      "learning_rate": 1.8921042540499515e-05,
+      "loss": 0.4502,
       "step": 4775
     },
     {
-      "epoch": 0.7003663003663003,
-      "grad_norm": 0.38434378064854174,
-      "learning_rate": 1.893473886421156e-05,
-      "loss": 0.4452,
+      "epoch": 0.703665538053879,
+      "grad_norm": 0.3995468827419847,
+      "learning_rate": 1.8917169620367003e-05,
+      "loss": 0.4711,
       "step": 4780
     },
     {
-      "epoch": 0.701098901098901,
-      "grad_norm": 0.3807732418526466,
-      "learning_rate": 1.8930907168748134e-05,
-      "loss": 0.453,
+      "epoch": 0.704401589871927,
+      "grad_norm": 0.37961563979427065,
+      "learning_rate": 1.8913290159587143e-05,
+      "loss": 0.456,
       "step": 4785
     },
     {
-      "epoch": 0.7018315018315018,
-      "grad_norm": 0.38048565611635704,
-      "learning_rate": 1.8927068983563633e-05,
-      "loss": 0.4445,
+      "epoch": 0.705137641689975,
+      "grad_norm": 0.36862889514517616,
+      "learning_rate": 1.8909404161005486e-05,
+      "loss": 0.4436,
       "step": 4790
     },
     {
-      "epoch": 0.7025641025641025,
-      "grad_norm": 0.3985839982100088,
-      "learning_rate": 1.892322431144711e-05,
-      "loss": 0.4435,
+      "epoch": 0.705873693508023,
+      "grad_norm": 0.37699874981750026,
+      "learning_rate": 1.8905511627472365e-05,
+      "loss": 0.4587,
       "step": 4795
     },
     {
-      "epoch": 0.7032967032967034,
-      "grad_norm": 0.3701985887177081,
-      "learning_rate": 1.891937315519233e-05,
-      "loss": 0.4335,
+      "epoch": 0.7066097453260709,
+      "grad_norm": 0.3908195086062283,
+      "learning_rate": 1.8901612561842915e-05,
+      "loss": 0.4513,
       "step": 4800
     },
     {
-      "epoch": 0.7040293040293041,
-      "grad_norm": 0.37760096364078544,
-      "learning_rate": 1.891551551759777e-05,
-      "loss": 0.4399,
+      "epoch": 0.7073457971441189,
+      "grad_norm": 0.3665398737233992,
+      "learning_rate": 1.8897706966977058e-05,
+      "loss": 0.4518,
       "step": 4805
     },
     {
-      "epoch": 0.7047619047619048,
-      "grad_norm": 0.4047001591947668,
-      "learning_rate": 1.8911651401466617e-05,
-      "loss": 0.4585,
+      "epoch": 0.708081848962167,
+      "grad_norm": 0.3609870619410624,
+      "learning_rate": 1.88937948457395e-05,
+      "loss": 0.4411,
       "step": 4810
     },
     {
-      "epoch": 0.7054945054945055,
-      "grad_norm": 0.35180736741582225,
-      "learning_rate": 1.8907780809606757e-05,
-      "loss": 0.4374,
+      "epoch": 0.7088179007802149,
+      "grad_norm": 0.38895035996514304,
+      "learning_rate": 1.8889876200999744e-05,
+      "loss": 0.4551,
       "step": 4815
     },
     {
-      "epoch": 0.7062271062271063,
-      "grad_norm": 0.3965065752973784,
-      "learning_rate": 1.8903903744830797e-05,
-      "loss": 0.4712,
+      "epoch": 0.7095539525982629,
+      "grad_norm": 0.35603852102578454,
+      "learning_rate": 1.8885951035632075e-05,
+      "loss": 0.4648,
       "step": 4820
     },
     {
-      "epoch": 0.706959706959707,
-      "grad_norm": 0.4056638023244465,
-      "learning_rate": 1.8900020209956035e-05,
-      "loss": 0.4517,
+      "epoch": 0.710290004416311,
+      "grad_norm": 0.42671666362664806,
+      "learning_rate": 1.8882019352515555e-05,
+      "loss": 0.4887,
       "step": 4825
     },
     {
-      "epoch": 0.7076923076923077,
-      "grad_norm": 0.37618097111880616,
-      "learning_rate": 1.889613020780448e-05,
-      "loss": 0.4591,
+      "epoch": 0.7110260562343589,
+      "grad_norm": 0.3516709921941014,
+      "learning_rate": 1.8878081154534035e-05,
+      "loss": 0.4473,
       "step": 4830
     },
     {
-      "epoch": 0.7084249084249085,
-      "grad_norm": 0.36476485229020567,
-      "learning_rate": 1.8892233741202834e-05,
-      "loss": 0.4576,
+      "epoch": 0.7117621080524069,
+      "grad_norm": 0.37682190495104745,
+      "learning_rate": 1.887413644457614e-05,
+      "loss": 0.4448,
       "step": 4835
     },
     {
-      "epoch": 0.7091575091575092,
-      "grad_norm": 0.41114552994910686,
-      "learning_rate": 1.8888330812982493e-05,
-      "loss": 0.4514,
+      "epoch": 0.7124981598704548,
+      "grad_norm": 0.3734482896287629,
+      "learning_rate": 1.887018522553527e-05,
+      "loss": 0.4576,
       "step": 4840
     },
     {
-      "epoch": 0.7098901098901099,
-      "grad_norm": 0.3938823356135875,
-      "learning_rate": 1.888442142597956e-05,
-      "loss": 0.4489,
+      "epoch": 0.7132342116885029,
+      "grad_norm": 0.36700338444436587,
+      "learning_rate": 1.8866227500309598e-05,
+      "loss": 0.4463,
       "step": 4845
     },
     {
-      "epoch": 0.7106227106227107,
-      "grad_norm": 0.3616177240649825,
-      "learning_rate": 1.888050558303482e-05,
-      "loss": 0.4365,
+      "epoch": 0.7139702635065509,
+      "grad_norm": 0.37609698891657933,
+      "learning_rate": 1.886226327180208e-05,
+      "loss": 0.4568,
       "step": 4850
     },
     {
-      "epoch": 0.7113553113553114,
-      "grad_norm": 0.5594879931276083,
-      "learning_rate": 1.8876583286993758e-05,
-      "loss": 0.4475,
+      "epoch": 0.7147063153245988,
+      "grad_norm": 0.3708912459699521,
+      "learning_rate": 1.8858292542920435e-05,
+      "loss": 0.4551,
       "step": 4855
     },
     {
-      "epoch": 0.7120879120879121,
-      "grad_norm": 0.3803372257837885,
-      "learning_rate": 1.887265454070655e-05,
-      "loss": 0.4488,
+      "epoch": 0.7154423671426469,
+      "grad_norm": 0.3523567619488166,
+      "learning_rate": 1.8854315316577142e-05,
+      "loss": 0.4385,
       "step": 4860
     },
     {
-      "epoch": 0.7128205128205128,
-      "grad_norm": 0.3842306891124848,
-      "learning_rate": 1.886871934702804e-05,
-      "loss": 0.4498,
+      "epoch": 0.7161784189606948,
+      "grad_norm": 0.39825711110012857,
+      "learning_rate": 1.885033159568946e-05,
+      "loss": 0.4661,
       "step": 4865
     },
     {
-      "epoch": 0.7135531135531136,
-      "grad_norm": 0.37564182357370485,
-      "learning_rate": 1.886477770881778e-05,
-      "loss": 0.4516,
+      "epoch": 0.7169144707787428,
+      "grad_norm": 0.37288272822218954,
+      "learning_rate": 1.8846341383179398e-05,
+      "loss": 0.4517,
       "step": 4870
     },
     {
-      "epoch": 0.7142857142857143,
-      "grad_norm": 0.3664584796382565,
-      "learning_rate": 1.8860829628939988e-05,
-      "loss": 0.4355,
+      "epoch": 0.7176505225967909,
+      "grad_norm": 0.4250134249290028,
+      "learning_rate": 1.8842344681973742e-05,
+      "loss": 0.4719,
       "step": 4875
     },
     {
-      "epoch": 0.715018315018315,
-      "grad_norm": 0.37290235571696095,
-      "learning_rate": 1.885687511026358e-05,
-      "loss": 0.4469,
+      "epoch": 0.7183865744148388,
+      "grad_norm": 0.36864907984912687,
+      "learning_rate": 1.8838341495004023e-05,
+      "loss": 0.4427,
       "step": 4880
     },
     {
-      "epoch": 0.7157509157509158,
-      "grad_norm": 0.4064105840730172,
-      "learning_rate": 1.8852914155662133e-05,
-      "loss": 0.4382,
+      "epoch": 0.7191226262328868,
+      "grad_norm": 0.39950862586857455,
+      "learning_rate": 1.883433182520654e-05,
+      "loss": 0.4451,
       "step": 4885
     },
     {
-      "epoch": 0.7164835164835165,
-      "grad_norm": 0.3842875164436108,
-      "learning_rate": 1.8848946768013915e-05,
-      "loss": 0.4408,
+      "epoch": 0.7198586780509347,
+      "grad_norm": 0.38114084933695436,
+      "learning_rate": 1.883031567552234e-05,
+      "loss": 0.4481,
       "step": 4890
     },
     {
-      "epoch": 0.7172161172161172,
-      "grad_norm": 0.38081609641953207,
-      "learning_rate": 1.8844972950201862e-05,
-      "loss": 0.4593,
+      "epoch": 0.7205947298689828,
+      "grad_norm": 0.4040185481089537,
+      "learning_rate": 1.8826293048897224e-05,
+      "loss": 0.4328,
       "step": 4895
     },
     {
-      "epoch": 0.717948717948718,
-      "grad_norm": 0.47127324509102547,
-      "learning_rate": 1.884099270511358e-05,
-      "loss": 0.4521,
+      "epoch": 0.7213307816870308,
+      "grad_norm": 0.3656004947367902,
+      "learning_rate": 1.8822263948281746e-05,
+      "loss": 0.4485,
       "step": 4900
     },
     {
-      "epoch": 0.7186813186813187,
-      "grad_norm": 0.37652291233531765,
-      "learning_rate": 1.883700603564135e-05,
-      "loss": 0.4587,
+      "epoch": 0.7220668335050787,
+      "grad_norm": 0.3553643062230974,
+      "learning_rate": 1.881822837663121e-05,
+      "loss": 0.4431,
       "step": 4905
     },
     {
-      "epoch": 0.7194139194139194,
-      "grad_norm": 0.3615301757266598,
-      "learning_rate": 1.883301294468212e-05,
-      "loss": 0.4501,
+      "epoch": 0.7228028853231268,
+      "grad_norm": 0.39429759042069384,
+      "learning_rate": 1.8814186336905663e-05,
+      "loss": 0.461,
       "step": 4910
     },
     {
-      "epoch": 0.7201465201465201,
-      "grad_norm": 0.36022573004294006,
-      "learning_rate": 1.8829013435137506e-05,
-      "loss": 0.4558,
+      "epoch": 0.7235389371411748,
+      "grad_norm": 0.42350656930755187,
+      "learning_rate": 1.8810137832069896e-05,
+      "loss": 0.4621,
       "step": 4915
     },
     {
-      "epoch": 0.7208791208791209,
-      "grad_norm": 0.35055657248154937,
-      "learning_rate": 1.882500750991378e-05,
-      "loss": 0.4582,
+      "epoch": 0.7242749889592227,
+      "grad_norm": 0.37015861640127484,
+      "learning_rate": 1.880608286509344e-05,
+      "loss": 0.4409,
       "step": 4920
     },
     {
-      "epoch": 0.7216117216117216,
-      "grad_norm": 0.3552183693090387,
-      "learning_rate": 1.882099517192189e-05,
-      "loss": 0.4413,
+      "epoch": 0.7250110407772707,
+      "grad_norm": 0.3759334193858358,
+      "learning_rate": 1.880202143895058e-05,
+      "loss": 0.4443,
       "step": 4925
     },
     {
-      "epoch": 0.7223443223443223,
-      "grad_norm": 0.3674515573775624,
-      "learning_rate": 1.881697642407743e-05,
-      "loss": 0.4502,
+      "epoch": 0.7257470925953187,
+      "grad_norm": 0.3833862965964792,
+      "learning_rate": 1.8797953556620315e-05,
+      "loss": 0.4413,
       "step": 4930
     },
     {
-      "epoch": 0.7230769230769231,
-      "grad_norm": 0.375113043902126,
-      "learning_rate": 1.8812951269300658e-05,
-      "loss": 0.4593,
+      "epoch": 0.7264831444133667,
+      "grad_norm": 0.3772971808673379,
+      "learning_rate": 1.8793879221086403e-05,
+      "loss": 0.4629,
       "step": 4935
     },
     {
-      "epoch": 0.7238095238095238,
-      "grad_norm": 0.4099788031048028,
-      "learning_rate": 1.8808919710516495e-05,
-      "loss": 0.4375,
+      "epoch": 0.7272191962314147,
+      "grad_norm": 0.3874088731230548,
+      "learning_rate": 1.8789798435337322e-05,
+      "loss": 0.471,
       "step": 4940
     },
     {
-      "epoch": 0.7245421245421245,
-      "grad_norm": 0.3691481226512916,
-      "learning_rate": 1.8804881750654504e-05,
-      "loss": 0.4314,
+      "epoch": 0.7279552480494627,
+      "grad_norm": 0.3756942626441261,
+      "learning_rate": 1.878571120236628e-05,
+      "loss": 0.4357,
       "step": 4945
     },
     {
-      "epoch": 0.7252747252747253,
-      "grad_norm": 0.36505639763282793,
-      "learning_rate": 1.88008373926489e-05,
-      "loss": 0.4424,
+      "epoch": 0.7286912998675107,
+      "grad_norm": 0.39184131451221743,
+      "learning_rate": 1.878161752517122e-05,
+      "loss": 0.4451,
       "step": 4950
     },
     {
-      "epoch": 0.726007326007326,
-      "grad_norm": 0.40079042557698097,
-      "learning_rate": 1.879678663943856e-05,
-      "loss": 0.4488,
+      "epoch": 0.7294273516855586,
+      "grad_norm": 0.3765062835007046,
+      "learning_rate": 1.8777517406754817e-05,
+      "loss": 0.4571,
       "step": 4955
     },
     {
-      "epoch": 0.7267399267399267,
-      "grad_norm": 0.35847162619593637,
-      "learning_rate": 1.8792729493966985e-05,
-      "loss": 0.4495,
+      "epoch": 0.7301634035036066,
+      "grad_norm": 0.38876516628659347,
+      "learning_rate": 1.8773410850124454e-05,
+      "loss": 0.4579,
       "step": 4960
     },
     {
-      "epoch": 0.7274725274725274,
-      "grad_norm": 0.3927400030593109,
-      "learning_rate": 1.8788665959182345e-05,
-      "loss": 0.438,
+      "epoch": 0.7308994553216547,
+      "grad_norm": 0.37049204158950866,
+      "learning_rate": 1.876929785829225e-05,
+      "loss": 0.449,
       "step": 4965
     },
     {
-      "epoch": 0.7282051282051282,
-      "grad_norm": 0.3767857839509493,
-      "learning_rate": 1.878459603803744e-05,
-      "loss": 0.4504,
+      "epoch": 0.7316355071397026,
+      "grad_norm": 0.36971484242293473,
+      "learning_rate": 1.876517843427504e-05,
+      "loss": 0.4539,
       "step": 4970
     },
     {
-      "epoch": 0.7289377289377289,
-      "grad_norm": 0.3853596665238205,
-      "learning_rate": 1.8780519733489715e-05,
-      "loss": 0.4373,
+      "epoch": 0.7323715589577506,
+      "grad_norm": 0.39244436955036255,
+      "learning_rate": 1.8761052581094378e-05,
+      "loss": 0.4658,
       "step": 4975
     },
     {
-      "epoch": 0.7296703296703296,
-      "grad_norm": 0.3689570683299289,
-      "learning_rate": 1.877643704850125e-05,
-      "loss": 0.4542,
+      "epoch": 0.7331076107757987,
+      "grad_norm": 0.40362849889587354,
+      "learning_rate": 1.875692030177653e-05,
+      "loss": 0.4564,
       "step": 4980
     },
     {
-      "epoch": 0.7304029304029304,
-      "grad_norm": 0.37575696161804173,
-      "learning_rate": 1.877234798603876e-05,
-      "loss": 0.473,
+      "epoch": 0.7338436625938466,
+      "grad_norm": 0.4008458908739168,
+      "learning_rate": 1.8752781599352476e-05,
+      "loss": 0.45,
       "step": 4985
     },
     {
-      "epoch": 0.7311355311355311,
-      "grad_norm": 0.3725794967777069,
-      "learning_rate": 1.876825254907361e-05,
-      "loss": 0.4725,
+      "epoch": 0.7345797144118946,
+      "grad_norm": 0.40270604908220076,
+      "learning_rate": 1.8748636476857913e-05,
+      "loss": 0.446,
       "step": 4990
     },
     {
-      "epoch": 0.7318681318681318,
-      "grad_norm": 0.4074849563026,
-      "learning_rate": 1.8764150740581772e-05,
-      "loss": 0.4524,
+      "epoch": 0.7353157662299425,
+      "grad_norm": 0.42604148951230386,
+      "learning_rate": 1.874448493733325e-05,
+      "loss": 0.4706,
       "step": 4995
     },
     {
-      "epoch": 0.7326007326007326,
-      "grad_norm": 0.37640767192398394,
-      "learning_rate": 1.8760042563543862e-05,
-      "loss": 0.4671,
+      "epoch": 0.7360518180479906,
+      "grad_norm": 0.41486119749955624,
+      "learning_rate": 1.874032698382358e-05,
+      "loss": 0.4368,
       "step": 5000
     },
     {
-      "epoch": 0.7333333333333333,
-      "grad_norm": 0.3778473707730308,
-      "learning_rate": 1.8755928020945134e-05,
-      "loss": 0.4414,
+      "epoch": 0.7367878698660386,
+      "grad_norm": 0.37245519856096193,
+      "learning_rate": 1.873616261937873e-05,
+      "loss": 0.4455,
       "step": 5005
     },
     {
-      "epoch": 0.734065934065934,
-      "grad_norm": 0.3817578090582604,
-      "learning_rate": 1.875180711577544e-05,
-      "loss": 0.4695,
+      "epoch": 0.7375239216840865,
+      "grad_norm": 0.3818563871959328,
+      "learning_rate": 1.8731991847053213e-05,
+      "loss": 0.4618,
       "step": 5010
     },
     {
-      "epoch": 0.7347985347985349,
-      "grad_norm": 0.36071974765757886,
-      "learning_rate": 1.8747679851029282e-05,
-      "loss": 0.462,
+      "epoch": 0.7382599735021346,
+      "grad_norm": 0.380589725275835,
+      "learning_rate": 1.8727814669906244e-05,
+      "loss": 0.4334,
       "step": 5015
     },
     {
-      "epoch": 0.7355311355311356,
-      "grad_norm": 0.368934043679171,
-      "learning_rate": 1.874354622970577e-05,
-      "loss": 0.4552,
+      "epoch": 0.7389960253201825,
+      "grad_norm": 0.347189931778005,
+      "learning_rate": 1.8723631091001736e-05,
+      "loss": 0.4538,
       "step": 5020
     },
     {
-      "epoch": 0.7362637362637363,
-      "grad_norm": 0.36533841497706593,
-      "learning_rate": 1.8739406254808637e-05,
-      "loss": 0.4576,
+      "epoch": 0.7397320771382305,
+      "grad_norm": 0.3898337158360767,
+      "learning_rate": 1.8719441113408304e-05,
+      "loss": 0.4525,
       "step": 5025
     },
     {
-      "epoch": 0.736996336996337,
-      "grad_norm": 0.36771301086535146,
-      "learning_rate": 1.873525992934623e-05,
-      "loss": 0.4612,
+      "epoch": 0.7404681289562786,
+      "grad_norm": 0.38327819805222985,
+      "learning_rate": 1.8715244740199247e-05,
+      "loss": 0.4497,
       "step": 5030
     },
     {
-      "epoch": 0.7377289377289378,
-      "grad_norm": 0.3791449694508501,
-      "learning_rate": 1.873110725633151e-05,
-      "loss": 0.4517,
+      "epoch": 0.7412041807743265,
+      "grad_norm": 0.3945510025537431,
+      "learning_rate": 1.871104197445256e-05,
+      "loss": 0.4618,
       "step": 5035
     },
     {
-      "epoch": 0.7384615384615385,
-      "grad_norm": 0.3727427986764495,
-      "learning_rate": 1.8726948238782057e-05,
-      "loss": 0.4383,
+      "epoch": 0.7419402325923745,
+      "grad_norm": 0.3806548660495654,
+      "learning_rate": 1.8706832819250923e-05,
+      "loss": 0.4563,
       "step": 5040
     },
     {
-      "epoch": 0.7391941391941392,
-      "grad_norm": 0.3781575792567377,
-      "learning_rate": 1.8722782879720058e-05,
-      "loss": 0.4623,
+      "epoch": 0.7426762844104225,
+      "grad_norm": 0.36540110870625,
+      "learning_rate": 1.8702617277681713e-05,
+      "loss": 0.4511,
       "step": 5045
     },
     {
-      "epoch": 0.73992673992674,
-      "grad_norm": 0.359064891528205,
-      "learning_rate": 1.8718611182172298e-05,
-      "loss": 0.449,
+      "epoch": 0.7434123362284705,
+      "grad_norm": 0.4287468254008677,
+      "learning_rate": 1.8698395352836982e-05,
+      "loss": 0.4437,
       "step": 5050
     },
     {
-      "epoch": 0.7406593406593407,
-      "grad_norm": 0.44843334665346996,
-      "learning_rate": 1.871443314917019e-05,
-      "loss": 0.4493,
+      "epoch": 0.7441483880465185,
+      "grad_norm": 0.3913820437377021,
+      "learning_rate": 1.869416704781346e-05,
+      "loss": 0.4685,
       "step": 5055
     },
     {
-      "epoch": 0.7413919413919414,
-      "grad_norm": 0.3957264108268295,
-      "learning_rate": 1.871024878374973e-05,
-      "loss": 0.4561,
+      "epoch": 0.7448844398645664,
+      "grad_norm": 0.39129385276647827,
+      "learning_rate": 1.868993236571257e-05,
+      "loss": 0.45,
       "step": 5060
     },
     {
-      "epoch": 0.7421245421245422,
-      "grad_norm": 0.37828618552992577,
-      "learning_rate": 1.8706058088951523e-05,
-      "loss": 0.4557,
+      "epoch": 0.7456204916826145,
+      "grad_norm": 0.3673533380962962,
+      "learning_rate": 1.86856913096404e-05,
+      "loss": 0.4959,
       "step": 5065
     },
     {
-      "epoch": 0.7428571428571429,
-      "grad_norm": 0.4035863638705251,
-      "learning_rate": 1.8701861067820777e-05,
-      "loss": 0.447,
+      "epoch": 0.7463565435006625,
+      "grad_norm": 0.4092195675059342,
+      "learning_rate": 1.8681443882707727e-05,
+      "loss": 0.4521,
       "step": 5070
     },
     {
-      "epoch": 0.7435897435897436,
-      "grad_norm": 0.3750533796195528,
-      "learning_rate": 1.8697657723407297e-05,
-      "loss": 0.4583,
+      "epoch": 0.7470925953187104,
+      "grad_norm": 0.39590745577112624,
+      "learning_rate": 1.867719008802998e-05,
+      "loss": 0.4514,
       "step": 5075
     },
     {
-      "epoch": 0.7443223443223443,
-      "grad_norm": 0.37723043217670793,
-      "learning_rate": 1.869344805876547e-05,
-      "loss": 0.451,
+      "epoch": 0.7478286471367585,
+      "grad_norm": 0.38196263810623365,
+      "learning_rate": 1.867292992872728e-05,
+      "loss": 0.4511,
       "step": 5080
     },
     {
-      "epoch": 0.7450549450549451,
-      "grad_norm": 0.39718945678044637,
-      "learning_rate": 1.8689232076954293e-05,
-      "loss": 0.4639,
+      "epoch": 0.7485646989548064,
+      "grad_norm": 0.4043595516480884,
+      "learning_rate": 1.866866340792441e-05,
+      "loss": 0.459,
       "step": 5085
     },
     {
-      "epoch": 0.7457875457875458,
-      "grad_norm": 0.3791498431905776,
-      "learning_rate": 1.8685009781037344e-05,
-      "loss": 0.4927,
+      "epoch": 0.7493007507728544,
+      "grad_norm": 0.402427574093436,
+      "learning_rate": 1.866439052875081e-05,
+      "loss": 0.4419,
       "step": 5090
     },
     {
-      "epoch": 0.7465201465201465,
-      "grad_norm": 0.3799852737272029,
-      "learning_rate": 1.8680781174082788e-05,
-      "loss": 0.4501,
+      "epoch": 0.7500368025909024,
+      "grad_norm": 0.3612962663270505,
+      "learning_rate": 1.8660111294340596e-05,
+      "loss": 0.4448,
       "step": 5095
     },
     {
-      "epoch": 0.7472527472527473,
-      "grad_norm": 0.36068963179554164,
-      "learning_rate": 1.8676546259163383e-05,
-      "loss": 0.4516,
+      "epoch": 0.7507728544089504,
+      "grad_norm": 0.3855820456686503,
+      "learning_rate": 1.8655825707832535e-05,
+      "loss": 0.4501,
       "step": 5100
     },
     {
-      "epoch": 0.747985347985348,
-      "grad_norm": 0.35788361997095947,
-      "learning_rate": 1.8672305039356465e-05,
-      "loss": 0.4522,
+      "epoch": 0.7515089062269984,
+      "grad_norm": 0.37589164340013,
+      "learning_rate": 1.8651533772370062e-05,
+      "loss": 0.4479,
       "step": 5105
     },
     {
-      "epoch": 0.7487179487179487,
-      "grad_norm": 0.3433115907360614,
-      "learning_rate": 1.866805751774395e-05,
-      "loss": 0.4414,
+      "epoch": 0.7522449580450463,
+      "grad_norm": 0.38967281407068854,
+      "learning_rate": 1.8647235491101265e-05,
+      "loss": 0.4383,
       "step": 5110
     },
     {
-      "epoch": 0.7494505494505495,
-      "grad_norm": 0.4014853165182206,
-      "learning_rate": 1.866380369741234e-05,
-      "loss": 0.4405,
+      "epoch": 0.7529810098630944,
+      "grad_norm": 0.3836742905405019,
+      "learning_rate": 1.8642930867178887e-05,
+      "loss": 0.4441,
       "step": 5115
     },
     {
-      "epoch": 0.7501831501831502,
-      "grad_norm": 0.38277765018604176,
-      "learning_rate": 1.865954358145271e-05,
-      "loss": 0.45,
+      "epoch": 0.7537170616811424,
+      "grad_norm": 0.3730181494757174,
+      "learning_rate": 1.8638619903760316e-05,
+      "loss": 0.4478,
       "step": 5120
     },
     {
-      "epoch": 0.7509157509157509,
-      "grad_norm": 0.3801569195926829,
-      "learning_rate": 1.865527717296071e-05,
-      "loss": 0.4372,
+      "epoch": 0.7544531134991903,
+      "grad_norm": 0.3925730698628966,
+      "learning_rate": 1.8634302604007608e-05,
+      "loss": 0.448,
       "step": 5125
     },
     {
-      "epoch": 0.7516483516483516,
-      "grad_norm": 0.3755925269128549,
-      "learning_rate": 1.8651004475036564e-05,
-      "loss": 0.4533,
+      "epoch": 0.7551891653172383,
+      "grad_norm": 0.37571986510259725,
+      "learning_rate": 1.8629978971087447e-05,
+      "loss": 0.4543,
       "step": 5130
     },
     {
-      "epoch": 0.7523809523809524,
-      "grad_norm": 0.3611593354999632,
-      "learning_rate": 1.864672549078506e-05,
-      "loss": 0.4687,
+      "epoch": 0.7559252171352864,
+      "grad_norm": 0.3941350023240938,
+      "learning_rate": 1.862564900817118e-05,
+      "loss": 0.4657,
       "step": 5135
     },
     {
-      "epoch": 0.7531135531135531,
-      "grad_norm": 0.36469816725356635,
-      "learning_rate": 1.8642440223315567e-05,
-      "loss": 0.4404,
+      "epoch": 0.7566612689533343,
+      "grad_norm": 0.3930544565143214,
+      "learning_rate": 1.862131271843478e-05,
+      "loss": 0.4459,
       "step": 5140
     },
     {
-      "epoch": 0.7538461538461538,
-      "grad_norm": 0.36037064742670877,
-      "learning_rate": 1.8638148675742003e-05,
-      "loss": 0.4422,
+      "epoch": 0.7573973207713823,
+      "grad_norm": 0.3871592158191773,
+      "learning_rate": 1.8616970105058868e-05,
+      "loss": 0.4655,
       "step": 5145
     },
     {
-      "epoch": 0.7545787545787546,
-      "grad_norm": 0.37843169049672115,
-      "learning_rate": 1.8633850851182864e-05,
-      "loss": 0.4359,
+      "epoch": 0.7581333725894303,
+      "grad_norm": 0.3686584661352915,
+      "learning_rate": 1.8612621171228706e-05,
+      "loss": 0.4463,
       "step": 5150
     },
     {
-      "epoch": 0.7553113553113553,
-      "grad_norm": 0.369373339711034,
-      "learning_rate": 1.86295467527612e-05,
-      "loss": 0.4388,
+      "epoch": 0.7588694244074783,
+      "grad_norm": 0.37756209216050757,
+      "learning_rate": 1.860826592013419e-05,
+      "loss": 0.4412,
       "step": 5155
     },
     {
-      "epoch": 0.756043956043956,
-      "grad_norm": 0.36148931406656815,
-      "learning_rate": 1.8625236383604618e-05,
-      "loss": 0.4331,
+      "epoch": 0.7596054762255263,
+      "grad_norm": 0.41869147081399966,
+      "learning_rate": 1.860390435496986e-05,
+      "loss": 0.4518,
       "step": 5160
     },
     {
-      "epoch": 0.7567765567765568,
-      "grad_norm": 0.3593933995850175,
-      "learning_rate": 1.8620919746845292e-05,
-      "loss": 0.4437,
+      "epoch": 0.7603415280435742,
+      "grad_norm": 0.36431124316624397,
+      "learning_rate": 1.8599536478934857e-05,
+      "loss": 0.4488,
       "step": 5165
     },
     {
-      "epoch": 0.7575091575091575,
-      "grad_norm": 0.3778438132302129,
-      "learning_rate": 1.861659684561994e-05,
-      "loss": 0.458,
+      "epoch": 0.7610775798616223,
+      "grad_norm": 0.45365567304602106,
+      "learning_rate": 1.8595162295232985e-05,
+      "loss": 0.4367,
       "step": 5170
     },
     {
-      "epoch": 0.7582417582417582,
-      "grad_norm": 0.36715635699796345,
-      "learning_rate": 1.8612267683069833e-05,
-      "loss": 0.4391,
+      "epoch": 0.7618136316796702,
+      "grad_norm": 0.37125941809574287,
+      "learning_rate": 1.8590781807072653e-05,
+      "loss": 0.4437,
       "step": 5175
     },
     {
-      "epoch": 0.7589743589743589,
-      "grad_norm": 0.38118576475659105,
-      "learning_rate": 1.8607932262340797e-05,
-      "loss": 0.4476,
+      "epoch": 0.7625496834977182,
+      "grad_norm": 0.3670272715328077,
+      "learning_rate": 1.8586395017666907e-05,
+      "loss": 0.4559,
       "step": 5180
     },
     {
-      "epoch": 0.7597069597069597,
-      "grad_norm": 0.36684997930037677,
-      "learning_rate": 1.8603590586583203e-05,
-      "loss": 0.4356,
+      "epoch": 0.7632857353157663,
+      "grad_norm": 0.3603956656614328,
+      "learning_rate": 1.8582001930233408e-05,
+      "loss": 0.4459,
       "step": 5185
     },
     {
-      "epoch": 0.7604395604395604,
-      "grad_norm": 0.3761906469625915,
-      "learning_rate": 1.859924265895197e-05,
-      "loss": 0.5002,
+      "epoch": 0.7640217871338142,
+      "grad_norm": 0.3684551799200026,
+      "learning_rate": 1.8577602547994432e-05,
+      "loss": 0.4339,
       "step": 5190
     },
     {
-      "epoch": 0.7611721611721611,
-      "grad_norm": 0.4942240977546921,
-      "learning_rate": 1.8594888482606553e-05,
-      "loss": 0.4955,
+      "epoch": 0.7647578389518622,
+      "grad_norm": 0.34139189157603933,
+      "learning_rate": 1.857319687417689e-05,
+      "loss": 0.4515,
       "step": 5195
     },
     {
-      "epoch": 0.7619047619047619,
-      "grad_norm": 0.3877521828288198,
-      "learning_rate": 1.8590528060710955e-05,
-      "loss": 0.4414,
+      "epoch": 0.7654938907699103,
+      "grad_norm": 0.3628155667104919,
+      "learning_rate": 1.8568784912012287e-05,
+      "loss": 0.4527,
       "step": 5200
     },
     {
-      "epoch": 0.7626373626373626,
-      "grad_norm": 0.3887164371745051,
-      "learning_rate": 1.858616139643371e-05,
-      "loss": 0.4483,
+      "epoch": 0.7662299425879582,
+      "grad_norm": 0.3856449838688728,
+      "learning_rate": 1.856436666473675e-05,
+      "loss": 0.4607,
       "step": 5205
     },
     {
-      "epoch": 0.7633699633699633,
-      "grad_norm": 0.370497863473674,
-      "learning_rate": 1.85817884929479e-05,
-      "loss": 0.4555,
+      "epoch": 0.7669659944060062,
+      "grad_norm": 0.3608153117696857,
+      "learning_rate": 1.855994213559102e-05,
+      "loss": 0.4362,
       "step": 5210
     },
     {
-      "epoch": 0.764102564102564,
-      "grad_norm": 0.37618089203333643,
-      "learning_rate": 1.8577409353431125e-05,
-      "loss": 0.4419,
+      "epoch": 0.7677020462240541,
+      "grad_norm": 0.40070515487703645,
+      "learning_rate": 1.855551132782044e-05,
+      "loss": 0.449,
       "step": 5215
     },
     {
-      "epoch": 0.7648351648351648,
-      "grad_norm": 0.37049761322207986,
-      "learning_rate": 1.857302398106553e-05,
-      "loss": 0.4486,
+      "epoch": 0.7684380980421022,
+      "grad_norm": 0.3677756171763214,
+      "learning_rate": 1.8551074244674955e-05,
+      "loss": 0.4561,
       "step": 5220
     },
     {
-      "epoch": 0.7655677655677655,
-      "grad_norm": 0.3700544252440904,
-      "learning_rate": 1.856863237903778e-05,
-      "loss": 0.4559,
+      "epoch": 0.7691741498601502,
+      "grad_norm": 0.3781817453838708,
+      "learning_rate": 1.8546630889409123e-05,
+      "loss": 0.4517,
       "step": 5225
     },
     {
-      "epoch": 0.7663003663003664,
-      "grad_norm": 0.388488061559954,
-      "learning_rate": 1.856423455053907e-05,
-      "loss": 0.4328,
+      "epoch": 0.7699102016781981,
+      "grad_norm": 0.3627035431665186,
+      "learning_rate": 1.8542181265282094e-05,
+      "loss": 0.4436,
       "step": 5230
     },
     {
-      "epoch": 0.7670329670329671,
-      "grad_norm": 0.3653284188280524,
-      "learning_rate": 1.8559830498765126e-05,
-      "loss": 0.4335,
+      "epoch": 0.7706462534962462,
+      "grad_norm": 0.37592957274051353,
+      "learning_rate": 1.853772537555762e-05,
+      "loss": 0.4564,
       "step": 5235
     },
     {
-      "epoch": 0.7677655677655678,
-      "grad_norm": 0.35741335636323357,
-      "learning_rate": 1.8555420226916186e-05,
-      "loss": 0.4351,
+      "epoch": 0.7713823053142941,
+      "grad_norm": 0.3885122459222397,
+      "learning_rate": 1.8533263223504046e-05,
+      "loss": 0.4371,
       "step": 5240
     },
     {
-      "epoch": 0.7684981684981685,
-      "grad_norm": 0.38870008021616226,
-      "learning_rate": 1.8551003738197017e-05,
-      "loss": 0.4258,
+      "epoch": 0.7721183571323421,
+      "grad_norm": 0.36563992082775965,
+      "learning_rate": 1.8528794812394316e-05,
+      "loss": 0.4597,
       "step": 5245
     },
     {
-      "epoch": 0.7692307692307693,
-      "grad_norm": 0.35448825136665196,
-      "learning_rate": 1.8546581035816894e-05,
-      "loss": 0.4682,
+      "epoch": 0.7728544089503901,
+      "grad_norm": 0.3711770613797605,
+      "learning_rate": 1.852432014550596e-05,
+      "loss": 0.4513,
       "step": 5250
     },
     {
-      "epoch": 0.76996336996337,
-      "grad_norm": 0.3647680958110189,
-      "learning_rate": 1.854215212298961e-05,
-      "loss": 0.4454,
+      "epoch": 0.7735904607684381,
+      "grad_norm": 0.3914413582610382,
+      "learning_rate": 1.8519839226121092e-05,
+      "loss": 0.4517,
       "step": 5255
     },
     {
-      "epoch": 0.7706959706959707,
-      "grad_norm": 0.3699915302758118,
-      "learning_rate": 1.853771700293348e-05,
-      "loss": 0.4744,
+      "epoch": 0.7743265125864861,
+      "grad_norm": 0.35992934113945746,
+      "learning_rate": 1.8515352057526423e-05,
+      "loss": 0.4344,
       "step": 5260
     },
     {
-      "epoch": 0.7714285714285715,
-      "grad_norm": 0.40320021190841804,
-      "learning_rate": 1.8533275678871316e-05,
-      "loss": 0.475,
+      "epoch": 0.775062564404534,
+      "grad_norm": 0.3764103775263548,
+      "learning_rate": 1.8510858643013245e-05,
+      "loss": 0.4399,
       "step": 5265
     },
     {
-      "epoch": 0.7721611721611722,
-      "grad_norm": 0.39419311147330366,
-      "learning_rate": 1.8528828154030446e-05,
-      "loss": 0.4567,
+      "epoch": 0.7757986162225821,
+      "grad_norm": 0.3511428578042135,
+      "learning_rate": 1.8506358985877427e-05,
+      "loss": 0.4523,
       "step": 5270
     },
     {
-      "epoch": 0.7728937728937729,
-      "grad_norm": 0.3716502661782266,
-      "learning_rate": 1.8524374431642705e-05,
-      "loss": 0.4585,
+      "epoch": 0.7765346680406301,
+      "grad_norm": 0.3972820359769529,
+      "learning_rate": 1.8501853089419413e-05,
+      "loss": 0.4543,
       "step": 5275
     },
     {
-      "epoch": 0.7736263736263737,
-      "grad_norm": 0.39443663328176837,
-      "learning_rate": 1.8519914514944428e-05,
-      "loss": 0.4466,
+      "epoch": 0.777270719858678,
+      "grad_norm": 0.3756548072147889,
+      "learning_rate": 1.849734095694424e-05,
+      "loss": 0.4699,
       "step": 5280
     },
     {
-      "epoch": 0.7743589743589744,
-      "grad_norm": 0.39711950445167926,
-      "learning_rate": 1.8515448407176453e-05,
-      "loss": 0.4628,
+      "epoch": 0.778006771676726,
+      "grad_norm": 0.3747943458424141,
+      "learning_rate": 1.8492822591761497e-05,
+      "loss": 0.4367,
       "step": 5285
     },
     {
-      "epoch": 0.7750915750915751,
-      "grad_norm": 0.384051731059323,
-      "learning_rate": 1.8510976111584113e-05,
-      "loss": 0.436,
+      "epoch": 0.7787428234947741,
+      "grad_norm": 0.37093375853708366,
+      "learning_rate": 1.8488297997185362e-05,
+      "loss": 0.4661,
       "step": 5290
     },
     {
-      "epoch": 0.7758241758241758,
-      "grad_norm": 0.38157072979128603,
-      "learning_rate": 1.8506497631417243e-05,
-      "loss": 0.482,
+      "epoch": 0.779478875312822,
+      "grad_norm": 0.38644659694052486,
+      "learning_rate": 1.8483767176534576e-05,
+      "loss": 0.4555,
       "step": 5295
     },
     {
-      "epoch": 0.7765567765567766,
-      "grad_norm": 0.3870879852096543,
-      "learning_rate": 1.8502012969930167e-05,
-      "loss": 0.4362,
+      "epoch": 0.78021492713087,
+      "grad_norm": 0.38387569939452526,
+      "learning_rate": 1.8479230133132447e-05,
+      "loss": 0.447,
       "step": 5300
     },
     {
-      "epoch": 0.7772893772893773,
-      "grad_norm": 0.353015417387719,
-      "learning_rate": 1.8497522130381705e-05,
-      "loss": 0.4504,
+      "epoch": 0.780950978948918,
+      "grad_norm": 0.376852133988765,
+      "learning_rate": 1.8474686870306847e-05,
+      "loss": 0.4349,
       "step": 5305
     },
     {
-      "epoch": 0.778021978021978,
-      "grad_norm": 0.36233329639826845,
-      "learning_rate": 1.8493025116035164e-05,
-      "loss": 0.4502,
+      "epoch": 0.781687030766966,
+      "grad_norm": 0.38169512123303556,
+      "learning_rate": 1.8470137391390207e-05,
+      "loss": 0.4414,
       "step": 5310
     },
     {
-      "epoch": 0.7787545787545788,
-      "grad_norm": 0.3568028276113107,
-      "learning_rate": 1.848852193015834e-05,
-      "loss": 0.4433,
+      "epoch": 0.782423082585014,
+      "grad_norm": 0.3689428108576067,
+      "learning_rate": 1.8465581699719524e-05,
+      "loss": 0.4636,
       "step": 5315
     },
     {
-      "epoch": 0.7794871794871795,
-      "grad_norm": 0.3601597074511503,
-      "learning_rate": 1.8484012576023506e-05,
-      "loss": 0.4318,
+      "epoch": 0.783159134403062,
+      "grad_norm": 0.371240196462308,
+      "learning_rate": 1.846101979863635e-05,
+      "loss": 0.4482,
       "step": 5320
     },
     {
-      "epoch": 0.7802197802197802,
-      "grad_norm": 0.3641022279761321,
-      "learning_rate": 1.847949705690743e-05,
-      "loss": 0.4501,
+      "epoch": 0.78389518622111,
+      "grad_norm": 0.39432719446589,
+      "learning_rate": 1.8456451691486786e-05,
+      "loss": 0.4498,
       "step": 5325
     },
     {
-      "epoch": 0.780952380952381,
-      "grad_norm": 0.3467781441532753,
-      "learning_rate": 1.8474975376091353e-05,
-      "loss": 0.4453,
+      "epoch": 0.7846312380391579,
+      "grad_norm": 0.3657710193307759,
+      "learning_rate": 1.845187738162149e-05,
+      "loss": 0.4494,
       "step": 5330
     },
     {
-      "epoch": 0.7816849816849817,
-      "grad_norm": 0.3858985111673107,
-      "learning_rate": 1.847044753686099e-05,
-      "loss": 0.445,
+      "epoch": 0.7853672898572059,
+      "grad_norm": 0.3680275296634426,
+      "learning_rate": 1.8447296872395673e-05,
+      "loss": 0.4501,
       "step": 5335
     },
     {
-      "epoch": 0.7824175824175824,
-      "grad_norm": 0.3807176887377528,
-      "learning_rate": 1.8465913542506535e-05,
-      "loss": 0.4391,
+      "epoch": 0.786103341675254,
+      "grad_norm": 0.37786387115319886,
+      "learning_rate": 1.844271016716908e-05,
+      "loss": 0.4612,
       "step": 5340
     },
     {
-      "epoch": 0.7831501831501831,
-      "grad_norm": 0.35993120225254266,
-      "learning_rate": 1.8461373396322656e-05,
-      "loss": 0.455,
+      "epoch": 0.7868393934933019,
+      "grad_norm": 0.39300182510769294,
+      "learning_rate": 1.8438117269306016e-05,
+      "loss": 0.4625,
       "step": 5345
     },
     {
-      "epoch": 0.7838827838827839,
-      "grad_norm": 0.3579613273358578,
-      "learning_rate": 1.845682710160849e-05,
-      "loss": 0.4485,
+      "epoch": 0.7875754453113499,
+      "grad_norm": 0.3571373651113558,
+      "learning_rate": 1.843351818217532e-05,
+      "loss": 0.4491,
       "step": 5350
     },
     {
-      "epoch": 0.7846153846153846,
-      "grad_norm": 0.7095086087647203,
-      "learning_rate": 1.845227466166764e-05,
-      "loss": 0.4382,
+      "epoch": 0.788311497129398,
+      "grad_norm": 0.35989435857796004,
+      "learning_rate": 1.8428912909150378e-05,
+      "loss": 0.4383,
       "step": 5355
     },
     {
-      "epoch": 0.7853479853479853,
-      "grad_norm": 0.3672460689680624,
-      "learning_rate": 1.8447716079808173e-05,
-      "loss": 0.4448,
+      "epoch": 0.7890475489474459,
+      "grad_norm": 0.36017249210285296,
+      "learning_rate": 1.84243014536091e-05,
+      "loss": 0.4561,
       "step": 5360
     },
     {
-      "epoch": 0.7860805860805861,
-      "grad_norm": 0.37486097935123897,
-      "learning_rate": 1.8443151359342628e-05,
-      "loss": 0.4482,
+      "epoch": 0.7897836007654939,
+      "grad_norm": 0.3599029744954481,
+      "learning_rate": 1.8419683818933943e-05,
+      "loss": 0.4512,
       "step": 5365
     },
     {
-      "epoch": 0.7868131868131868,
-      "grad_norm": 0.36079331013575355,
-      "learning_rate": 1.843858050358799e-05,
-      "loss": 0.4352,
+      "epoch": 0.7905196525835418,
+      "grad_norm": 0.3731630654712982,
+      "learning_rate": 1.8415060008511888e-05,
+      "loss": 0.4525,
       "step": 5370
     },
     {
-      "epoch": 0.7875457875457875,
-      "grad_norm": 0.39482594783223374,
-      "learning_rate": 1.8434003515865723e-05,
-      "loss": 0.4567,
+      "epoch": 0.7912557044015899,
+      "grad_norm": 0.3700359157670555,
+      "learning_rate": 1.8410430025734455e-05,
+      "loss": 0.4587,
       "step": 5375
     },
     {
-      "epoch": 0.7882783882783883,
-      "grad_norm": 0.3999742065119079,
-      "learning_rate": 1.8429420399501726e-05,
-      "loss": 0.4613,
+      "epoch": 0.7919917562196379,
+      "grad_norm": 0.40329273308198865,
+      "learning_rate": 1.840579387399769e-05,
+      "loss": 0.4299,
       "step": 5380
     },
     {
-      "epoch": 0.789010989010989,
-      "grad_norm": 0.3838801617736062,
-      "learning_rate": 1.8424831157826367e-05,
-      "loss": 0.4642,
+      "epoch": 0.7927278080376858,
+      "grad_norm": 0.3557122428580658,
+      "learning_rate": 1.8401151556702147e-05,
+      "loss": 0.4356,
       "step": 5385
     },
     {
-      "epoch": 0.7897435897435897,
-      "grad_norm": 0.3614566587995,
-      "learning_rate": 1.842023579417445e-05,
-      "loss": 0.4446,
+      "epoch": 0.7934638598557339,
+      "grad_norm": 0.3785726951882142,
+      "learning_rate": 1.8396503077252922e-05,
+      "loss": 0.4405,
       "step": 5390
     },
     {
-      "epoch": 0.7904761904761904,
-      "grad_norm": 0.3858603380855061,
-      "learning_rate": 1.841563431188524e-05,
-      "loss": 0.4496,
+      "epoch": 0.7941999116737818,
+      "grad_norm": 0.3869038189914558,
+      "learning_rate": 1.8391848439059627e-05,
+      "loss": 0.4445,
       "step": 5395
     },
     {
-      "epoch": 0.7912087912087912,
-      "grad_norm": 0.3598539639497055,
-      "learning_rate": 1.8411026714302442e-05,
-      "loss": 0.4609,
+      "epoch": 0.7949359634918298,
+      "grad_norm": 0.3539505443058832,
+      "learning_rate": 1.838718764553639e-05,
+      "loss": 0.4293,
       "step": 5400
     },
     {
-      "epoch": 0.7919413919413919,
-      "grad_norm": 0.3938895475003292,
-      "learning_rate": 1.840641300477421e-05,
-      "loss": 0.439,
+      "epoch": 0.7956720153098779,
+      "grad_norm": 0.40425072185625943,
+      "learning_rate": 1.8382520700101842e-05,
+      "loss": 0.4567,
       "step": 5405
     },
     {
-      "epoch": 0.7926739926739926,
-      "grad_norm": 0.6150085736480287,
-      "learning_rate": 1.8401793186653137e-05,
-      "loss": 0.4785,
+      "epoch": 0.7964080671279258,
+      "grad_norm": 0.3711515438134564,
+      "learning_rate": 1.837784760617915e-05,
+      "loss": 0.437,
       "step": 5410
     },
     {
-      "epoch": 0.7934065934065934,
-      "grad_norm": 0.35701126359919005,
-      "learning_rate": 1.8397167263296248e-05,
-      "loss": 0.453,
+      "epoch": 0.7971441189459738,
+      "grad_norm": 0.3818939300370364,
+      "learning_rate": 1.8373168367195967e-05,
+      "loss": 0.4459,
       "step": 5415
     },
     {
-      "epoch": 0.7941391941391941,
-      "grad_norm": 0.38688320595265163,
-      "learning_rate": 1.8392535238065016e-05,
-      "loss": 0.4546,
+      "epoch": 0.7978801707640217,
+      "grad_norm": 0.365958939061682,
+      "learning_rate": 1.8368482986584474e-05,
+      "loss": 0.4463,
       "step": 5420
     },
     {
-      "epoch": 0.7948717948717948,
-      "grad_norm": 0.37797055568968796,
-      "learning_rate": 1.8387897114325337e-05,
-      "loss": 0.4566,
+      "epoch": 0.7986162225820698,
+      "grad_norm": 0.34264355403477026,
+      "learning_rate": 1.836379146778134e-05,
+      "loss": 0.4365,
       "step": 5425
     },
     {
-      "epoch": 0.7956043956043956,
-      "grad_norm": 0.3709536305983092,
-      "learning_rate": 1.838325289544755e-05,
-      "loss": 0.4515,
+      "epoch": 0.7993522744001178,
+      "grad_norm": 0.3827506711198045,
+      "learning_rate": 1.8359093814227743e-05,
+      "loss": 0.4464,
       "step": 5430
     },
     {
-      "epoch": 0.7963369963369963,
-      "grad_norm": 0.3974020225169862,
-      "learning_rate": 1.837860258480641e-05,
-      "loss": 0.4595,
+      "epoch": 0.8000883262181657,
+      "grad_norm": 0.3737144115110153,
+      "learning_rate": 1.8354390029369367e-05,
+      "loss": 0.4439,
       "step": 5435
     },
     {
-      "epoch": 0.7970695970695971,
-      "grad_norm": 0.3727483028673619,
-      "learning_rate": 1.837394618578111e-05,
-      "loss": 0.4514,
+      "epoch": 0.8008243780362138,
+      "grad_norm": 0.3660125324278533,
+      "learning_rate": 1.8349680116656383e-05,
+      "loss": 0.4347,
       "step": 5440
     },
     {
-      "epoch": 0.7978021978021979,
-      "grad_norm": 0.36935258408678506,
-      "learning_rate": 1.8369283701755264e-05,
-      "loss": 0.4473,
+      "epoch": 0.8015604298542618,
+      "grad_norm": 0.35499618388758564,
+      "learning_rate": 1.834496407954346e-05,
+      "loss": 0.4455,
       "step": 5445
     },
     {
-      "epoch": 0.7985347985347986,
-      "grad_norm": 0.3899646603873487,
-      "learning_rate": 1.8364615136116902e-05,
-      "loss": 0.4619,
+      "epoch": 0.8022964816723097,
+      "grad_norm": 0.391762324948392,
+      "learning_rate": 1.8340241921489762e-05,
+      "loss": 0.4508,
       "step": 5450
     },
     {
-      "epoch": 0.7992673992673993,
-      "grad_norm": 0.3938468777203018,
-      "learning_rate": 1.8359940492258482e-05,
-      "loss": 0.465,
+      "epoch": 0.8030325334903577,
+      "grad_norm": 0.3675686249642425,
+      "learning_rate": 1.833551364595894e-05,
+      "loss": 0.4541,
       "step": 5455
     },
     {
-      "epoch": 0.8,
-      "grad_norm": 0.39489444569801624,
-      "learning_rate": 1.835525977357687e-05,
-      "loss": 0.4531,
+      "epoch": 0.8037685853084057,
+      "grad_norm": 0.35284461393648797,
+      "learning_rate": 1.8330779256419133e-05,
+      "loss": 0.4223,
       "step": 5460
     },
     {
-      "epoch": 0.8007326007326008,
-      "grad_norm": 0.36553764851582793,
-      "learning_rate": 1.8350572983473353e-05,
-      "loss": 0.4537,
+      "epoch": 0.8045046371264537,
+      "grad_norm": 0.3572095074816016,
+      "learning_rate": 1.8326038756342967e-05,
+      "loss": 0.4287,
       "step": 5465
     },
     {
-      "epoch": 0.8014652014652015,
-      "grad_norm": 0.35314847287217677,
-      "learning_rate": 1.834588012535363e-05,
-      "loss": 0.4433,
+      "epoch": 0.8052406889445017,
+      "grad_norm": 0.36797577285222993,
+      "learning_rate": 1.8321292149207542e-05,
+      "loss": 0.4338,
       "step": 5470
     },
     {
-      "epoch": 0.8021978021978022,
-      "grad_norm": 0.35878611079441036,
-      "learning_rate": 1.8341181202627806e-05,
-      "loss": 0.4475,
+      "epoch": 0.8059767407625497,
+      "grad_norm": 0.3992223730063507,
+      "learning_rate": 1.831653943849445e-05,
+      "loss": 0.4508,
       "step": 5475
     },
     {
-      "epoch": 0.802930402930403,
-      "grad_norm": 0.3539347328291622,
-      "learning_rate": 1.833647621871039e-05,
-      "loss": 0.4386,
+      "epoch": 0.8067127925805977,
+      "grad_norm": 0.35710109084837255,
+      "learning_rate": 1.8311780627689746e-05,
+      "loss": 0.442,
       "step": 5480
     },
     {
-      "epoch": 0.8036630036630037,
-      "grad_norm": 0.37458522095393504,
-      "learning_rate": 1.83317651770203e-05,
-      "loss": 0.4488,
+      "epoch": 0.8074488443986456,
+      "grad_norm": 0.36857684404559116,
+      "learning_rate": 1.830701572028397e-05,
+      "loss": 0.4393,
       "step": 5485
     },
     {
-      "epoch": 0.8043956043956044,
-      "grad_norm": 0.36047244047702937,
-      "learning_rate": 1.832704808098086e-05,
-      "loss": 0.4472,
+      "epoch": 0.8081848962166936,
+      "grad_norm": 0.365337760645847,
+      "learning_rate": 1.8302244719772128e-05,
+      "loss": 0.4584,
       "step": 5490
     },
     {
-      "epoch": 0.8051282051282052,
-      "grad_norm": 0.3681356297525558,
-      "learning_rate": 1.8322324934019783e-05,
-      "loss": 0.4329,
+      "epoch": 0.8089209480347417,
+      "grad_norm": 0.36272299441325834,
+      "learning_rate": 1.82974676296537e-05,
+      "loss": 0.4414,
       "step": 5495
     },
     {
-      "epoch": 0.8058608058608059,
-      "grad_norm": 0.35249822895051003,
-      "learning_rate": 1.831759573956919e-05,
-      "loss": 0.4422,
+      "epoch": 0.8096569998527896,
+      "grad_norm": 0.3808089020887524,
+      "learning_rate": 1.829268445343263e-05,
+      "loss": 0.4494,
       "step": 5500
     },
     {
-      "epoch": 0.8065934065934066,
-      "grad_norm": 0.3787470928657058,
-      "learning_rate": 1.831286050106559e-05,
-      "loss": 0.4525,
+      "epoch": 0.8103930516708376,
+      "grad_norm": 0.36515311233950665,
+      "learning_rate": 1.828789519461732e-05,
+      "loss": 0.4304,
       "step": 5505
     },
     {
-      "epoch": 0.8073260073260073,
-      "grad_norm": 0.3614213832013432,
-      "learning_rate": 1.8308119221949883e-05,
-      "loss": 0.4449,
+      "epoch": 0.8111291034888857,
+      "grad_norm": 0.3883250866338634,
+      "learning_rate": 1.8283099856720644e-05,
+      "loss": 0.4531,
       "step": 5510
     },
     {
-      "epoch": 0.8080586080586081,
-      "grad_norm": 0.3622051651540198,
-      "learning_rate": 1.830337190566736e-05,
-      "loss": 0.4237,
+      "epoch": 0.8118651553069336,
+      "grad_norm": 0.36668979698642856,
+      "learning_rate": 1.8278298443259933e-05,
+      "loss": 0.4543,
       "step": 5515
     },
     {
-      "epoch": 0.8087912087912088,
-      "grad_norm": 0.3590555015654966,
-      "learning_rate": 1.8298618555667702e-05,
-      "loss": 0.4327,
+      "epoch": 0.8126012071249816,
+      "grad_norm": 0.36618403443345937,
+      "learning_rate": 1.827349095775697e-05,
+      "loss": 0.4531,
       "step": 5520
     },
     {
-      "epoch": 0.8095238095238095,
-      "grad_norm": 0.37340106282476715,
-      "learning_rate": 1.8293859175404967e-05,
-      "loss": 0.4384,
+      "epoch": 0.8133372589430296,
+      "grad_norm": 0.347384132693151,
+      "learning_rate": 1.8268677403737987e-05,
+      "loss": 0.4284,
       "step": 5525
     },
     {
-      "epoch": 0.8102564102564103,
-      "grad_norm": 0.36540095124697075,
-      "learning_rate": 1.828909376833761e-05,
-      "loss": 0.438,
+      "epoch": 0.8140733107610776,
+      "grad_norm": 0.35052747224037406,
+      "learning_rate": 1.826385778473368e-05,
+      "loss": 0.4692,
       "step": 5530
     },
     {
-      "epoch": 0.810989010989011,
-      "grad_norm": 0.3656833313182153,
-      "learning_rate": 1.8284322337928447e-05,
-      "loss": 0.4369,
+      "epoch": 0.8148093625791256,
+      "grad_norm": 0.3775231856628442,
+      "learning_rate": 1.8259032104279192e-05,
+      "loss": 0.4519,
       "step": 5535
     },
     {
-      "epoch": 0.8117216117216117,
-      "grad_norm": 0.38878197123203745,
-      "learning_rate": 1.827954488764468e-05,
-      "loss": 0.4432,
+      "epoch": 0.8155454143971735,
+      "grad_norm": 0.38341860673581035,
+      "learning_rate": 1.8254200365914095e-05,
+      "loss": 0.4513,
       "step": 5540
     },
     {
-      "epoch": 0.8124542124542125,
-      "grad_norm": 0.40991346099163334,
-      "learning_rate": 1.8274761420957885e-05,
-      "loss": 0.4442,
+      "epoch": 0.8162814662152216,
+      "grad_norm": 0.3757091428549333,
+      "learning_rate": 1.8249362573182426e-05,
+      "loss": 0.4421,
       "step": 5545
     },
     {
-      "epoch": 0.8131868131868132,
-      "grad_norm": 0.37087513786427795,
-      "learning_rate": 1.8269971941344015e-05,
-      "loss": 0.4441,
+      "epoch": 0.8170175180332695,
+      "grad_norm": 0.37270354415776735,
+      "learning_rate": 1.8244518729632653e-05,
+      "loss": 0.4304,
       "step": 5550
     },
     {
-      "epoch": 0.8139194139194139,
-      "grad_norm": 0.36966756582618115,
-      "learning_rate": 1.8265176452283376e-05,
-      "loss": 0.4529,
+      "epoch": 0.8177535698513175,
+      "grad_norm": 0.3523588981852036,
+      "learning_rate": 1.823966883881768e-05,
+      "loss": 0.421,
       "step": 5555
     },
     {
-      "epoch": 0.8146520146520146,
-      "grad_norm": 0.36065429278785466,
-      "learning_rate": 1.8260374957260663e-05,
-      "loss": 0.452,
+      "epoch": 0.8184896216693656,
+      "grad_norm": 0.36890743395894915,
+      "learning_rate": 1.8234812904294848e-05,
+      "loss": 0.4385,
       "step": 5560
     },
     {
-      "epoch": 0.8153846153846154,
-      "grad_norm": 0.3755528851797531,
-      "learning_rate": 1.8255567459764915e-05,
-      "loss": 0.4495,
+      "epoch": 0.8192256734874135,
+      "grad_norm": 0.3679958989076562,
+      "learning_rate": 1.8229950929625937e-05,
+      "loss": 0.4527,
       "step": 5565
     },
     {
-      "epoch": 0.8161172161172161,
-      "grad_norm": 0.38585828003647493,
-      "learning_rate": 1.825075396328955e-05,
-      "loss": 0.4563,
+      "epoch": 0.8199617253054615,
+      "grad_norm": 0.3757095310963074,
+      "learning_rate": 1.822508291837715e-05,
+      "loss": 0.4502,
       "step": 5570
     },
     {
-      "epoch": 0.8168498168498168,
-      "grad_norm": 0.38564318808684317,
-      "learning_rate": 1.8245934471332328e-05,
-      "loss": 0.4674,
+      "epoch": 0.8206977771235094,
+      "grad_norm": 0.36416547211663713,
+      "learning_rate": 1.8220208874119124e-05,
+      "loss": 0.4401,
       "step": 5575
     },
     {
-      "epoch": 0.8175824175824176,
-      "grad_norm": 0.3852320217786936,
-      "learning_rate": 1.8241108987395383e-05,
-      "loss": 0.4577,
+      "epoch": 0.8214338289415575,
+      "grad_norm": 0.38246500140389866,
+      "learning_rate": 1.821532880042691e-05,
+      "loss": 0.4553,
       "step": 5580
     },
     {
-      "epoch": 0.8183150183150183,
-      "grad_norm": 0.36745376533563195,
-      "learning_rate": 1.8236277514985187e-05,
-      "loss": 0.4577,
+      "epoch": 0.8221698807596055,
+      "grad_norm": 0.37685866397775736,
+      "learning_rate": 1.8210442700879993e-05,
+      "loss": 0.4506,
       "step": 5585
     },
     {
-      "epoch": 0.819047619047619,
-      "grad_norm": 0.35681374230355783,
-      "learning_rate": 1.8231440057612578e-05,
-      "loss": 0.4525,
+      "epoch": 0.8229059325776534,
+      "grad_norm": 0.36557315146939534,
+      "learning_rate": 1.8205550579062277e-05,
+      "loss": 0.4456,
       "step": 5590
     },
     {
-      "epoch": 0.8197802197802198,
-      "grad_norm": 0.3858709344287839,
-      "learning_rate": 1.8226596618792735e-05,
-      "loss": 0.4462,
+      "epoch": 0.8236419843957015,
+      "grad_norm": 0.361921077104969,
+      "learning_rate": 1.820065243856208e-05,
+      "loss": 0.4566,
       "step": 5595
     },
     {
-      "epoch": 0.8205128205128205,
-      "grad_norm": 0.3713342510413519,
-      "learning_rate": 1.8221747202045182e-05,
-      "loss": 0.4614,
+      "epoch": 0.8243780362137495,
+      "grad_norm": 0.3875910863569535,
+      "learning_rate": 1.8195748282972136e-05,
+      "loss": 0.4402,
       "step": 5600
     },
     {
-      "epoch": 0.8212454212454212,
-      "grad_norm": 0.37547437011409723,
-      "learning_rate": 1.821689181089379e-05,
-      "loss": 0.4526,
+      "epoch": 0.8251140880317974,
+      "grad_norm": 0.37048417394215916,
+      "learning_rate": 1.819083811588958e-05,
+      "loss": 0.4537,
       "step": 5605
     },
     {
-      "epoch": 0.8219780219780219,
-      "grad_norm": 0.3618467967252856,
-      "learning_rate": 1.8212030448866774e-05,
-      "loss": 0.4319,
+      "epoch": 0.8258501398498455,
+      "grad_norm": 0.35012768784241244,
+      "learning_rate": 1.8185921940915977e-05,
+      "loss": 0.4365,
       "step": 5610
     },
     {
-      "epoch": 0.8227106227106227,
-      "grad_norm": 0.3448704961367217,
-      "learning_rate": 1.820716311949668e-05,
-      "loss": 0.4887,
+      "epoch": 0.8265861916678934,
+      "grad_norm": 0.3705116290320156,
+      "learning_rate": 1.8180999761657286e-05,
+      "loss": 0.437,
       "step": 5615
     },
     {
-      "epoch": 0.8234432234432234,
-      "grad_norm": 0.3567342227035512,
-      "learning_rate": 1.8202289826320396e-05,
-      "loss": 0.447,
+      "epoch": 0.8273222434859414,
+      "grad_norm": 0.3587912899651877,
+      "learning_rate": 1.8176071581723872e-05,
+      "loss": 0.4319,
       "step": 5620
     },
     {
-      "epoch": 0.8241758241758241,
-      "grad_norm": 0.3590067657015829,
-      "learning_rate": 1.819741057287915e-05,
-      "loss": 0.4389,
+      "epoch": 0.8280582953039894,
+      "grad_norm": 0.3838217342869182,
+      "learning_rate": 1.8171137404730497e-05,
+      "loss": 0.4501,
       "step": 5625
     },
     {
-      "epoch": 0.8249084249084249,
-      "grad_norm": 0.38231454074130955,
-      "learning_rate": 1.8192525362718486e-05,
-      "loss": 0.4415,
+      "epoch": 0.8287943471220374,
+      "grad_norm": 0.3715639137108274,
+      "learning_rate": 1.816619723429633e-05,
+      "loss": 0.4485,
       "step": 5630
     },
     {
-      "epoch": 0.8256410256410256,
-      "grad_norm": 0.3746456485110328,
-      "learning_rate": 1.8187634199388284e-05,
-      "loss": 0.4397,
+      "epoch": 0.8295303989400854,
+      "grad_norm": 0.3483529315906707,
+      "learning_rate": 1.8161251074044938e-05,
+      "loss": 0.4464,
       "step": 5635
     },
     {
-      "epoch": 0.8263736263736263,
-      "grad_norm": 0.36611002508904406,
-      "learning_rate": 1.8182737086442757e-05,
-      "loss": 0.45,
+      "epoch": 0.8302664507581333,
+      "grad_norm": 0.36112338980486547,
+      "learning_rate": 1.8156298927604265e-05,
+      "loss": 0.4251,
       "step": 5640
     },
     {
-      "epoch": 0.827106227106227,
-      "grad_norm": 0.37341336111571827,
-      "learning_rate": 1.817783402744043e-05,
-      "loss": 0.492,
+      "epoch": 0.8310025025761814,
+      "grad_norm": 0.35503588938112196,
+      "learning_rate": 1.8151340798606664e-05,
+      "loss": 0.4565,
       "step": 5645
     },
     {
-      "epoch": 0.8278388278388278,
-      "grad_norm": 0.7970515407489609,
-      "learning_rate": 1.8172925025944153e-05,
-      "loss": 0.4431,
+      "epoch": 0.8317385543942294,
+      "grad_norm": 0.35872582895347865,
+      "learning_rate": 1.8146376690688867e-05,
+      "loss": 0.452,
       "step": 5650
     },
     {
-      "epoch": 0.8285714285714286,
-      "grad_norm": 0.36273365672650953,
-      "learning_rate": 1.8168010085521092e-05,
-      "loss": 0.4464,
+      "epoch": 0.8324746062122773,
+      "grad_norm": 0.39098541892175553,
+      "learning_rate": 1.8141406607491994e-05,
+      "loss": 0.4332,
       "step": 5655
     },
     {
-      "epoch": 0.8293040293040294,
-      "grad_norm": 0.35503852594146296,
-      "learning_rate": 1.8163089209742738e-05,
-      "loss": 0.4419,
+      "epoch": 0.8332106580303253,
+      "grad_norm": 0.34949566666207704,
+      "learning_rate": 1.8136430552661548e-05,
+      "loss": 0.4243,
       "step": 5660
     },
     {
-      "epoch": 0.8300366300366301,
-      "grad_norm": 0.35280501209823245,
-      "learning_rate": 1.8158162402184885e-05,
-      "loss": 0.4412,
+      "epoch": 0.8339467098483734,
+      "grad_norm": 0.3674696893778593,
+      "learning_rate": 1.8131448529847408e-05,
+      "loss": 0.4335,
       "step": 5665
     },
     {
-      "epoch": 0.8307692307692308,
-      "grad_norm": 0.381689499406052,
-      "learning_rate": 1.8153229666427637e-05,
-      "loss": 0.4331,
+      "epoch": 0.8346827616664213,
+      "grad_norm": 0.35916143870324896,
+      "learning_rate": 1.812646054270384e-05,
+      "loss": 0.4565,
       "step": 5670
     },
     {
-      "epoch": 0.8315018315018315,
-      "grad_norm": 0.3563507854326389,
-      "learning_rate": 1.8148291006055414e-05,
-      "loss": 0.4367,
+      "epoch": 0.8354188134844693,
+      "grad_norm": 0.38713495753289473,
+      "learning_rate": 1.812146659488947e-05,
+      "loss": 0.4559,
       "step": 5675
     },
     {
-      "epoch": 0.8322344322344323,
-      "grad_norm": 0.39339940600097917,
-      "learning_rate": 1.8143346424656934e-05,
-      "loss": 0.4355,
+      "epoch": 0.8361548653025173,
+      "grad_norm": 0.35110284160129296,
+      "learning_rate": 1.8116466690067305e-05,
+      "loss": 0.4377,
       "step": 5680
     },
     {
-      "epoch": 0.832967032967033,
-      "grad_norm": 0.3647447079242811,
-      "learning_rate": 1.8138395925825218e-05,
-      "loss": 0.4456,
+      "epoch": 0.8368909171205653,
+      "grad_norm": 0.3776798856757189,
+      "learning_rate": 1.8111460831904734e-05,
+      "loss": 0.4369,
       "step": 5685
     },
     {
-      "epoch": 0.8336996336996337,
-      "grad_norm": 0.3630697500736815,
-      "learning_rate": 1.8133439513157593e-05,
-      "loss": 0.4477,
+      "epoch": 0.8376269689386133,
+      "grad_norm": 0.36307720486779066,
+      "learning_rate": 1.8106449024073487e-05,
+      "loss": 0.4657,
       "step": 5690
     },
     {
-      "epoch": 0.8344322344322345,
-      "grad_norm": 0.3509428195025076,
-      "learning_rate": 1.8128477190255677e-05,
-      "loss": 0.4365,
+      "epoch": 0.8383630207566612,
+      "grad_norm": 0.37569502905454044,
+      "learning_rate": 1.8101431270249678e-05,
+      "loss": 0.4414,
       "step": 5695
     },
     {
-      "epoch": 0.8351648351648352,
-      "grad_norm": 0.37631260789384546,
-      "learning_rate": 1.812350896072538e-05,
-      "loss": 0.4362,
+      "epoch": 0.8390990725747093,
+      "grad_norm": 0.38106182035348096,
+      "learning_rate": 1.8096407574113766e-05,
+      "loss": 0.4506,
       "step": 5700
     },
     {
-      "epoch": 0.8358974358974359,
-      "grad_norm": 0.34402078961448856,
-      "learning_rate": 1.8118534828176922e-05,
-      "loss": 0.4538,
+      "epoch": 0.8398351243927572,
+      "grad_norm": 0.37332006831666176,
+      "learning_rate": 1.8091377939350585e-05,
+      "loss": 0.4356,
       "step": 5705
     },
     {
-      "epoch": 0.8366300366300367,
-      "grad_norm": 0.36987338499672917,
-      "learning_rate": 1.8113554796224787e-05,
-      "loss": 0.4626,
+      "epoch": 0.8405711762108052,
+      "grad_norm": 0.38142312594246786,
+      "learning_rate": 1.808634236964932e-05,
+      "loss": 0.4533,
       "step": 5710
     },
     {
-      "epoch": 0.8373626373626374,
-      "grad_norm": 0.37382664447171465,
-      "learning_rate": 1.8108568868487767e-05,
-      "loss": 0.4375,
+      "epoch": 0.8413072280288533,
+      "grad_norm": 0.35027498208673846,
+      "learning_rate": 1.80813008687035e-05,
+      "loss": 0.4288,
       "step": 5715
     },
     {
-      "epoch": 0.8380952380952381,
-      "grad_norm": 0.49701574346598654,
-      "learning_rate": 1.810357704858892e-05,
-      "loss": 0.4584,
+      "epoch": 0.8420432798469012,
+      "grad_norm": 0.36118089061149933,
+      "learning_rate": 1.8076253440211015e-05,
+      "loss": 0.463,
       "step": 5720
     },
     {
-      "epoch": 0.8388278388278388,
-      "grad_norm": 0.3946556829023878,
-      "learning_rate": 1.8098579340155603e-05,
-      "loss": 0.4536,
+      "epoch": 0.8427793316649492,
+      "grad_norm": 0.3479218004425318,
+      "learning_rate": 1.8071200087874096e-05,
+      "loss": 0.4247,
       "step": 5725
     },
     {
-      "epoch": 0.8395604395604396,
-      "grad_norm": 0.37085792154551794,
-      "learning_rate": 1.8093575746819443e-05,
-      "loss": 0.4339,
+      "epoch": 0.8435153834829971,
+      "grad_norm": 0.3647558131692801,
+      "learning_rate": 1.8066140815399325e-05,
+      "loss": 0.4499,
       "step": 5730
     },
     {
-      "epoch": 0.8402930402930403,
-      "grad_norm": 0.3996995534274054,
-      "learning_rate": 1.8088566272216344e-05,
-      "loss": 0.4593,
+      "epoch": 0.8442514353010452,
+      "grad_norm": 0.36710243654144514,
+      "learning_rate": 1.8061075626497623e-05,
+      "loss": 0.4456,
       "step": 5735
     },
     {
-      "epoch": 0.841025641025641,
-      "grad_norm": 0.36609811623203165,
-      "learning_rate": 1.8083550919986483e-05,
-      "loss": 0.4393,
+      "epoch": 0.8449874871190932,
+      "grad_norm": 0.37048760818431076,
+      "learning_rate": 1.8056004524884256e-05,
+      "loss": 0.4583,
       "step": 5740
     },
     {
-      "epoch": 0.8417582417582418,
-      "grad_norm": 0.36578815772035955,
-      "learning_rate": 1.8078529693774308e-05,
-      "loss": 0.4886,
+      "epoch": 0.8457235389371411,
+      "grad_norm": 2.407465284178379,
+      "learning_rate": 1.8050927514278813e-05,
+      "loss": 0.446,
       "step": 5745
     },
     {
-      "epoch": 0.8424908424908425,
-      "grad_norm": 0.3943288970973504,
-      "learning_rate": 1.807350259722854e-05,
-      "loss": 0.4413,
+      "epoch": 0.8464595907551892,
+      "grad_norm": 0.39853603765662204,
+      "learning_rate": 1.8045844598405233e-05,
+      "loss": 0.4483,
       "step": 5750
     },
     {
-      "epoch": 0.8432234432234432,
-      "grad_norm": 0.3404081783087885,
-      "learning_rate": 1.806846963400216e-05,
-      "loss": 0.4443,
+      "epoch": 0.8471956425732372,
+      "grad_norm": 0.3913222133525507,
+      "learning_rate": 1.8040755780991776e-05,
+      "loss": 0.4388,
       "step": 5755
     },
     {
-      "epoch": 0.843956043956044,
-      "grad_norm": 0.35096481604460117,
-      "learning_rate": 1.8063430807752414e-05,
-      "loss": 0.4433,
+      "epoch": 0.8479316943912851,
+      "grad_norm": 0.35525574871247056,
+      "learning_rate": 1.803566106577104e-05,
+      "loss": 0.4328,
       "step": 5760
     },
     {
-      "epoch": 0.8446886446886447,
-      "grad_norm": 0.38710053606474926,
-      "learning_rate": 1.805838612214081e-05,
-      "loss": 0.4396,
+      "epoch": 0.8486677462093332,
+      "grad_norm": 0.3620075978660273,
+      "learning_rate": 1.8030560456479933e-05,
+      "loss": 0.4395,
       "step": 5765
     },
     {
-      "epoch": 0.8454212454212454,
-      "grad_norm": 0.3651629348457347,
-      "learning_rate": 1.8053335580833112e-05,
-      "loss": 0.4475,
+      "epoch": 0.8494037980273811,
+      "grad_norm": 0.3518378202004742,
+      "learning_rate": 1.8025453956859707e-05,
+      "loss": 0.4345,
       "step": 5770
     },
     {
-      "epoch": 0.8461538461538461,
-      "grad_norm": 0.3987845742126223,
-      "learning_rate": 1.804827918749934e-05,
-      "loss": 0.4409,
+      "epoch": 0.8501398498454291,
+      "grad_norm": 0.3810142137831866,
+      "learning_rate": 1.802034157065592e-05,
+      "loss": 0.4457,
       "step": 5775
     },
     {
-      "epoch": 0.8468864468864469,
-      "grad_norm": 0.37633061700939063,
-      "learning_rate": 1.8043216945813764e-05,
-      "loss": 0.4495,
+      "epoch": 0.8508759016634772,
+      "grad_norm": 0.3417713865113361,
+      "learning_rate": 1.801522330161845e-05,
+      "loss": 0.4427,
       "step": 5780
     },
     {
-      "epoch": 0.8476190476190476,
-      "grad_norm": 0.37078702743335046,
-      "learning_rate": 1.803814885945491e-05,
-      "loss": 0.45,
+      "epoch": 0.8516119534815251,
+      "grad_norm": 0.3434090925556998,
+      "learning_rate": 1.8010099153501496e-05,
+      "loss": 0.4376,
       "step": 5785
     },
     {
-      "epoch": 0.8483516483516483,
-      "grad_norm": 0.36461446476347853,
-      "learning_rate": 1.8033074932105542e-05,
-      "loss": 0.4506,
+      "epoch": 0.8523480052995731,
+      "grad_norm": 0.37073730240412717,
+      "learning_rate": 1.8004969130063562e-05,
+      "loss": 0.4445,
       "step": 5790
     },
     {
-      "epoch": 0.8490842490842491,
-      "grad_norm": 0.3777899347748043,
-      "learning_rate": 1.8027995167452678e-05,
-      "loss": 0.4647,
+      "epoch": 0.853084057117621,
+      "grad_norm": 0.3706926685948443,
+      "learning_rate": 1.7999833235067466e-05,
+      "loss": 0.4364,
       "step": 5795
     },
     {
-      "epoch": 0.8498168498168498,
-      "grad_norm": 0.36394643705921864,
-      "learning_rate": 1.802290956918757e-05,
-      "loss": 0.4509,
+      "epoch": 0.8538201089356691,
+      "grad_norm": 0.35952848302399276,
+      "learning_rate": 1.799469147228033e-05,
+      "loss": 0.4594,
       "step": 5800
     },
     {
-      "epoch": 0.8505494505494505,
-      "grad_norm": 0.3685824062385075,
-      "learning_rate": 1.8017818141005712e-05,
-      "loss": 0.4452,
+      "epoch": 0.8545561607537171,
+      "grad_norm": 0.366251188214727,
+      "learning_rate": 1.7989543845473585e-05,
+      "loss": 0.4568,
       "step": 5805
     },
     {
-      "epoch": 0.8512820512820513,
-      "grad_norm": 0.37407351999618116,
-      "learning_rate": 1.8012720886606834e-05,
-      "loss": 0.429,
+      "epoch": 0.855292212571765,
+      "grad_norm": 0.37387266078880926,
+      "learning_rate": 1.7984390358422957e-05,
+      "loss": 0.4522,
       "step": 5810
     },
     {
-      "epoch": 0.852014652014652,
-      "grad_norm": 0.35006965931360773,
-      "learning_rate": 1.80076178096949e-05,
-      "loss": 0.4525,
+      "epoch": 0.856028264389813,
+      "grad_norm": 0.35727442551212146,
+      "learning_rate": 1.7979231014908475e-05,
+      "loss": 0.4547,
       "step": 5815
     },
     {
-      "epoch": 0.8527472527472527,
-      "grad_norm": 0.35092174273004206,
-      "learning_rate": 1.800250891397811e-05,
-      "loss": 0.4406,
+      "epoch": 0.8567643162078611,
+      "grad_norm": 0.35751741935540254,
+      "learning_rate": 1.7974065818714456e-05,
+      "loss": 0.4556,
       "step": 5820
     },
     {
-      "epoch": 0.8534798534798534,
-      "grad_norm": 0.37239973812022864,
-      "learning_rate": 1.799739420316888e-05,
-      "loss": 0.4521,
+      "epoch": 0.857500368025909,
+      "grad_norm": 0.3548710093691114,
+      "learning_rate": 1.7968894773629516e-05,
+      "loss": 0.435,
       "step": 5825
     },
     {
-      "epoch": 0.8542124542124542,
-      "grad_norm": 0.385803878666077,
-      "learning_rate": 1.7992273680983864e-05,
-      "loss": 0.441,
+      "epoch": 0.858236419843957,
+      "grad_norm": 0.34775308478188827,
+      "learning_rate": 1.796371788344657e-05,
+      "loss": 0.4367,
       "step": 5830
     },
     {
-      "epoch": 0.8549450549450549,
-      "grad_norm": 0.3685372405718134,
-      "learning_rate": 1.7987147351143933e-05,
-      "loss": 0.4475,
+      "epoch": 0.858972471662005,
+      "grad_norm": 0.3811514664148062,
+      "learning_rate": 1.7958535151962803e-05,
+      "loss": 0.4521,
       "step": 5835
     },
     {
-      "epoch": 0.8556776556776556,
-      "grad_norm": 0.37185126386807804,
-      "learning_rate": 1.7982015217374176e-05,
-      "loss": 0.4393,
+      "epoch": 0.859708523480053,
+      "grad_norm": 0.3669226247199922,
+      "learning_rate": 1.795334658297969e-05,
+      "loss": 0.4397,
       "step": 5840
     },
     {
-      "epoch": 0.8564102564102564,
-      "grad_norm": 0.3473113321126282,
-      "learning_rate": 1.7976877283403906e-05,
-      "loss": 0.4145,
+      "epoch": 0.860444575298101,
+      "grad_norm": 0.35854880100158387,
+      "learning_rate": 1.7948152180303e-05,
+      "loss": 0.4238,
       "step": 5845
     },
     {
-      "epoch": 0.8571428571428571,
-      "grad_norm": 0.36999708682756344,
-      "learning_rate": 1.7971733552966645e-05,
-      "loss": 0.4394,
+      "epoch": 0.861180627116149,
+      "grad_norm": 0.3654599811899801,
+      "learning_rate": 1.7942951947742763e-05,
+      "loss": 0.4304,
       "step": 5850
     },
     {
-      "epoch": 0.8578754578754578,
-      "grad_norm": 0.3793725160321135,
-      "learning_rate": 1.7966584029800135e-05,
-      "loss": 0.4557,
+      "epoch": 0.861916678934197,
+      "grad_norm": 0.34496532912792455,
+      "learning_rate": 1.7937745889113292e-05,
+      "loss": 0.4309,
       "step": 5855
     },
     {
-      "epoch": 0.8586080586080586,
-      "grad_norm": 0.35404015757458934,
-      "learning_rate": 1.796142871764631e-05,
-      "loss": 0.4283,
+      "epoch": 0.8626527307522449,
+      "grad_norm": 0.37889428636763145,
+      "learning_rate": 1.7932534008233182e-05,
+      "loss": 0.4609,
       "step": 5860
     },
     {
-      "epoch": 0.8593406593406593,
-      "grad_norm": 0.356545090974131,
-      "learning_rate": 1.7956267620251334e-05,
-      "loss": 0.4399,
+      "epoch": 0.863388782570293,
+      "grad_norm": 0.38209494491819235,
+      "learning_rate": 1.7927316308925283e-05,
+      "loss": 0.4531,
       "step": 5865
     },
     {
-      "epoch": 0.8600732600732601,
-      "grad_norm": 0.3533846397534374,
-      "learning_rate": 1.7951100741365555e-05,
-      "loss": 0.4363,
+      "epoch": 0.864124834388341,
+      "grad_norm": 0.36509393103751375,
+      "learning_rate": 1.7922092795016732e-05,
+      "loss": 0.4378,
       "step": 5870
     },
     {
-      "epoch": 0.8608058608058609,
-      "grad_norm": 0.3751400459972393,
-      "learning_rate": 1.7945928084743533e-05,
-      "loss": 0.4481,
+      "epoch": 0.8648608862063889,
+      "grad_norm": 0.34771616466055044,
+      "learning_rate": 1.791686347033891e-05,
+      "loss": 0.4419,
       "step": 5875
     },
     {
-      "epoch": 0.8615384615384616,
-      "grad_norm": 0.3559244611269117,
-      "learning_rate": 1.7940749654144026e-05,
-      "loss": 0.4361,
+      "epoch": 0.8655969380244369,
+      "grad_norm": 0.3674692070761661,
+      "learning_rate": 1.791162833872747e-05,
+      "loss": 0.4398,
       "step": 5880
     },
     {
-      "epoch": 0.8622710622710623,
-      "grad_norm": 0.3758775698301192,
-      "learning_rate": 1.7935565453329983e-05,
-      "loss": 0.4554,
+      "epoch": 0.8663329898424849,
+      "grad_norm": 0.3520396685118552,
+      "learning_rate": 1.790638740402233e-05,
+      "loss": 0.4416,
       "step": 5885
     },
     {
-      "epoch": 0.863003663003663,
-      "grad_norm": 0.6880224103063644,
-      "learning_rate": 1.793037548606855e-05,
-      "loss": 0.449,
+      "epoch": 0.8670690416605329,
+      "grad_norm": 0.3719988966714447,
+      "learning_rate": 1.7901140670067656e-05,
+      "loss": 0.4554,
       "step": 5890
     },
     {
-      "epoch": 0.8637362637362638,
-      "grad_norm": 0.3539214046988438,
-      "learning_rate": 1.792517975613106e-05,
-      "loss": 0.4345,
+      "epoch": 0.8678050934785809,
+      "grad_norm": 0.3690570201357628,
+      "learning_rate": 1.789588814071187e-05,
+      "loss": 0.4331,
       "step": 5895
     },
     {
-      "epoch": 0.8644688644688645,
-      "grad_norm": 0.35642889205036643,
-      "learning_rate": 1.7919978267293038e-05,
-      "loss": 0.4458,
+      "epoch": 0.8685411452966288,
+      "grad_norm": 0.34978248618101204,
+      "learning_rate": 1.789062981980765e-05,
+      "loss": 0.439,
       "step": 5900
     },
     {
-      "epoch": 0.8652014652014652,
-      "grad_norm": 0.3942719622252805,
-      "learning_rate": 1.7914771023334186e-05,
-      "loss": 0.4469,
+      "epoch": 0.8692771971146769,
+      "grad_norm": 0.36624291745278736,
+      "learning_rate": 1.788536571121191e-05,
+      "loss": 0.4323,
       "step": 5905
     },
     {
-      "epoch": 0.865934065934066,
-      "grad_norm": 0.3912235103465084,
-      "learning_rate": 1.7909558028038397e-05,
-      "loss": 0.4382,
+      "epoch": 0.8700132489327249,
+      "grad_norm": 0.46229638344586804,
+      "learning_rate": 1.788009581878582e-05,
+      "loss": 0.4477,
       "step": 5910
     },
     {
-      "epoch": 0.8666666666666667,
-      "grad_norm": 0.36516670112938615,
-      "learning_rate": 1.7904339285193738e-05,
-      "loss": 0.4473,
+      "epoch": 0.8707493007507728,
+      "grad_norm": 0.37396607227419815,
+      "learning_rate": 1.787482014639479e-05,
+      "loss": 0.4393,
       "step": 5915
     },
     {
-      "epoch": 0.8673992673992674,
-      "grad_norm": 0.3551416896853839,
-      "learning_rate": 1.7899114798592455e-05,
-      "loss": 0.4524,
+      "epoch": 0.8714853525688209,
+      "grad_norm": 0.37122982449214265,
+      "learning_rate": 1.7869538697908466e-05,
+      "loss": 0.4402,
       "step": 5920
     },
     {
-      "epoch": 0.8681318681318682,
-      "grad_norm": 0.3642603079549456,
-      "learning_rate": 1.7893884572030963e-05,
-      "loss": 0.4514,
+      "epoch": 0.8722214043868688,
+      "grad_norm": 0.37120657056831224,
+      "learning_rate": 1.7864251477200738e-05,
+      "loss": 0.4437,
       "step": 5925
     },
     {
-      "epoch": 0.8688644688644689,
-      "grad_norm": 0.36531254956832565,
-      "learning_rate": 1.7888648609309855e-05,
-      "loss": 0.4402,
+      "epoch": 0.8729574562049168,
+      "grad_norm": 0.3585818869382761,
+      "learning_rate": 1.7858958488149724e-05,
+      "loss": 0.4368,
       "step": 5930
     },
     {
-      "epoch": 0.8695970695970696,
-      "grad_norm": 0.3614156415868774,
-      "learning_rate": 1.7883406914233887e-05,
-      "loss": 0.4391,
+      "epoch": 0.8736935080229649,
+      "grad_norm": 0.3698141558742395,
+      "learning_rate": 1.7853659734637766e-05,
+      "loss": 0.4413,
       "step": 5935
     },
     {
-      "epoch": 0.8703296703296703,
-      "grad_norm": 0.3504318199203323,
-      "learning_rate": 1.7878159490611985e-05,
-      "loss": 0.4274,
+      "epoch": 0.8744295598410128,
+      "grad_norm": 0.3855876831267745,
+      "learning_rate": 1.7848355220551456e-05,
+      "loss": 0.4387,
       "step": 5940
     },
     {
-      "epoch": 0.8710622710622711,
-      "grad_norm": 0.37194708447240693,
-      "learning_rate": 1.787290634225723e-05,
-      "loss": 0.4446,
+      "epoch": 0.8751656116590608,
+      "grad_norm": 0.36621342556791076,
+      "learning_rate": 1.7843044949781583e-05,
+      "loss": 0.4479,
       "step": 5945
     },
     {
-      "epoch": 0.8717948717948718,
-      "grad_norm": 0.3806751249182061,
-      "learning_rate": 1.7867647472986867e-05,
-      "loss": 0.4355,
+      "epoch": 0.8759016634771087,
+      "grad_norm": 0.3424070085839823,
+      "learning_rate": 1.7837728926223186e-05,
+      "loss": 0.4428,
       "step": 5950
     },
     {
-      "epoch": 0.8725274725274725,
-      "grad_norm": 0.3858153470499486,
-      "learning_rate": 1.7862382886622306e-05,
-      "loss": 0.4309,
+      "epoch": 0.8766377152951568,
+      "grad_norm": 0.35449364017215884,
+      "learning_rate": 1.7832407153775502e-05,
+      "loss": 0.4522,
       "step": 5955
     },
     {
-      "epoch": 0.8732600732600733,
-      "grad_norm": 0.3539393871318372,
-      "learning_rate": 1.78571125869891e-05,
-      "loss": 0.4282,
+      "epoch": 0.8773737671132048,
+      "grad_norm": 0.3525405003036111,
+      "learning_rate": 1.7827079636342e-05,
+      "loss": 0.4311,
       "step": 5960
     },
     {
-      "epoch": 0.873992673992674,
-      "grad_norm": 0.3698727253718363,
-      "learning_rate": 1.785183657791696e-05,
-      "loss": 0.438,
+      "epoch": 0.8781098189312527,
+      "grad_norm": 0.35645386727273587,
+      "learning_rate": 1.782174637783035e-05,
+      "loss": 0.4411,
       "step": 5965
     },
     {
-      "epoch": 0.8747252747252747,
-      "grad_norm": 0.36033087121528645,
-      "learning_rate": 1.784655486323974e-05,
-      "loss": 0.4463,
+      "epoch": 0.8788458707493008,
+      "grad_norm": 0.3666268050359117,
+      "learning_rate": 1.7816407382152443e-05,
+      "loss": 0.447,
       "step": 5970
     },
     {
-      "epoch": 0.8754578754578755,
-      "grad_norm": 0.3696278035088527,
-      "learning_rate": 1.7841267446795444e-05,
-      "loss": 0.4459,
+      "epoch": 0.8795819225673488,
+      "grad_norm": 0.37250456874223414,
+      "learning_rate": 1.7811062653224374e-05,
+      "loss": 0.4655,
       "step": 5975
     },
     {
-      "epoch": 0.8761904761904762,
-      "grad_norm": 0.3641746049449239,
-      "learning_rate": 1.7835974332426224e-05,
-      "loss": 0.4212,
+      "epoch": 0.8803179743853967,
+      "grad_norm": 0.36092929099158944,
+      "learning_rate": 1.7805712194966443e-05,
+      "loss": 0.4566,
       "step": 5980
     },
     {
-      "epoch": 0.8769230769230769,
-      "grad_norm": 0.3496867702032214,
-      "learning_rate": 1.783067552397836e-05,
-      "loss": 0.4407,
+      "epoch": 0.8810540262034448,
+      "grad_norm": 0.33567737532624775,
+      "learning_rate": 1.7800356011303156e-05,
+      "loss": 0.4474,
       "step": 5985
     },
     {
-      "epoch": 0.8776556776556776,
-      "grad_norm": 0.3614859564979733,
-      "learning_rate": 1.782537102530228e-05,
-      "loss": 0.4522,
+      "epoch": 0.8817900780214927,
+      "grad_norm": 0.3653761820052591,
+      "learning_rate": 1.7794994106163212e-05,
+      "loss": 0.4493,
       "step": 5990
     },
     {
-      "epoch": 0.8783882783882784,
-      "grad_norm": 0.36421283641217894,
-      "learning_rate": 1.782006084025255e-05,
-      "loss": 0.4285,
+      "epoch": 0.8825261298395407,
+      "grad_norm": 0.35656753170206057,
+      "learning_rate": 1.778962648347951e-05,
+      "loss": 0.4358,
       "step": 5995
     },
     {
-      "epoch": 0.8791208791208791,
-      "grad_norm": 0.419513631468471,
-      "learning_rate": 1.7814744972687857e-05,
-      "loss": 0.4455,
+      "epoch": 0.8832621816575887,
+      "grad_norm": 0.37457391578364474,
+      "learning_rate": 1.7784253147189146e-05,
+      "loss": 0.4259,
       "step": 6000
     },
     {
-      "epoch": 0.8798534798534798,
-      "grad_norm": 0.354330410359772,
-      "learning_rate": 1.7809423426471018e-05,
-      "loss": 0.4656,
+      "epoch": 0.8839982334756367,
+      "grad_norm": 0.3566126547130828,
+      "learning_rate": 1.7778874101233406e-05,
+      "loss": 0.4465,
       "step": 6005
     },
     {
-      "epoch": 0.8805860805860806,
-      "grad_norm": 0.35900884915326187,
-      "learning_rate": 1.7804096205468987e-05,
-      "loss": 0.4473,
+      "epoch": 0.8847342852936847,
+      "grad_norm": 0.34727232006464764,
+      "learning_rate": 1.7773489349557756e-05,
+      "loss": 0.4375,
       "step": 6010
     },
     {
-      "epoch": 0.8813186813186813,
-      "grad_norm": 0.36804505720762803,
-      "learning_rate": 1.7798763313552832e-05,
-      "loss": 0.4381,
+      "epoch": 0.8854703371117326,
+      "grad_norm": 0.3486190773511903,
+      "learning_rate": 1.7768098896111854e-05,
+      "loss": 0.446,
       "step": 6015
     },
     {
-      "epoch": 0.882051282051282,
-      "grad_norm": 0.36697586215792644,
-      "learning_rate": 1.779342475459775e-05,
-      "loss": 0.4416,
+      "epoch": 0.8862063889297807,
+      "grad_norm": 0.3655089745027463,
+      "learning_rate": 1.7762702744849546e-05,
+      "loss": 0.4559,
       "step": 6020
     },
     {
-      "epoch": 0.8827838827838828,
-      "grad_norm": 0.3616003666688806,
-      "learning_rate": 1.7788080532483043e-05,
-      "loss": 0.4402,
+      "epoch": 0.8869424407478287,
+      "grad_norm": 0.3878902418363454,
+      "learning_rate": 1.7757300899728842e-05,
+      "loss": 0.4387,
       "step": 6025
     },
     {
-      "epoch": 0.8835164835164835,
-      "grad_norm": 0.3625678232884831,
-      "learning_rate": 1.7782730651092143e-05,
-      "loss": 0.4272,
+      "epoch": 0.8876784925658766,
+      "grad_norm": 0.3715951742442913,
+      "learning_rate": 1.775189336471194e-05,
+      "loss": 0.4461,
       "step": 6030
     },
     {
-      "epoch": 0.8842490842490842,
-      "grad_norm": 0.3525429739793186,
-      "learning_rate": 1.7777375114312584e-05,
-      "loss": 0.4376,
+      "epoch": 0.8884145443839246,
+      "grad_norm": 0.352923429841573,
+      "learning_rate": 1.774648014376521e-05,
+      "loss": 0.4571,
       "step": 6035
     },
     {
-      "epoch": 0.884981684981685,
-      "grad_norm": 0.37246107416507285,
-      "learning_rate": 1.7772013926036016e-05,
-      "loss": 0.4341,
+      "epoch": 0.8891505962019726,
+      "grad_norm": 0.37039532769096317,
+      "learning_rate": 1.774106124085919e-05,
+      "loss": 0.4529,
       "step": 6040
     },
     {
-      "epoch": 0.8857142857142857,
-      "grad_norm": 0.34048150719409837,
-      "learning_rate": 1.7766647090158193e-05,
-      "loss": 0.4323,
+      "epoch": 0.8898866480200206,
+      "grad_norm": 0.5768260851908226,
+      "learning_rate": 1.773563665996859e-05,
+      "loss": 0.4527,
       "step": 6045
     },
     {
-      "epoch": 0.8864468864468864,
-      "grad_norm": 0.37398024115120954,
-      "learning_rate": 1.7761274610578968e-05,
-      "loss": 0.4434,
+      "epoch": 0.8906226998380686,
+      "grad_norm": 0.4870960586341211,
+      "learning_rate": 1.7730206405072272e-05,
+      "loss": 0.4468,
       "step": 6050
     },
     {
-      "epoch": 0.8871794871794871,
-      "grad_norm": 0.35873272657760785,
-      "learning_rate": 1.7755896491202303e-05,
-      "loss": 0.4494,
+      "epoch": 0.8913587516561166,
+      "grad_norm": 0.36762519408457883,
+      "learning_rate": 1.772477048015328e-05,
+      "loss": 0.4568,
       "step": 6055
     },
     {
-      "epoch": 0.8879120879120879,
-      "grad_norm": 0.35741385020399524,
-      "learning_rate": 1.7750512735936257e-05,
-      "loss": 0.4421,
+      "epoch": 0.8920948034741646,
+      "grad_norm": 0.3637046517058373,
+      "learning_rate": 1.7719328889198798e-05,
+      "loss": 0.433,
       "step": 6060
     },
     {
-      "epoch": 0.8886446886446886,
-      "grad_norm": 0.3656854547321221,
-      "learning_rate": 1.774512334869298e-05,
-      "loss": 0.4566,
+      "epoch": 0.8928308552922126,
+      "grad_norm": 0.3714092685023693,
+      "learning_rate": 1.7713881636200177e-05,
+      "loss": 0.4553,
       "step": 6065
     },
     {
-      "epoch": 0.8893772893772893,
-      "grad_norm": 0.4183610536274334,
-      "learning_rate": 1.7739728333388718e-05,
-      "loss": 0.4478,
+      "epoch": 0.8935669071102605,
+      "grad_norm": 0.3560384199032731,
+      "learning_rate": 1.7708428725152917e-05,
+      "loss": 0.4411,
       "step": 6070
     },
     {
-      "epoch": 0.8901098901098901,
-      "grad_norm": 0.3429267484869677,
-      "learning_rate": 1.7734327693943805e-05,
-      "loss": 0.4319,
+      "epoch": 0.8943029589283086,
+      "grad_norm": 0.3570093335300102,
+      "learning_rate": 1.7702970160056675e-05,
+      "loss": 0.4531,
       "step": 6075
     },
     {
-      "epoch": 0.8908424908424909,
-      "grad_norm": 0.36437477716150785,
-      "learning_rate": 1.7728921434282662e-05,
-      "loss": 0.4519,
+      "epoch": 0.8950390107463565,
+      "grad_norm": 0.36570811442375184,
+      "learning_rate": 1.769750594491524e-05,
+      "loss": 0.458,
       "step": 6080
     },
     {
-      "epoch": 0.8915750915750916,
-      "grad_norm": 0.3620889390157865,
-      "learning_rate": 1.7723509558333797e-05,
-      "loss": 0.4305,
+      "epoch": 0.8957750625644045,
+      "grad_norm": 0.3625464342620715,
+      "learning_rate": 1.769203608373656e-05,
+      "loss": 0.4533,
       "step": 6085
     },
     {
-      "epoch": 0.8923076923076924,
-      "grad_norm": 0.36916306992865705,
-      "learning_rate": 1.771809207002979e-05,
-      "loss": 0.4648,
+      "epoch": 0.8965111143824526,
+      "grad_norm": 0.34712557080290696,
+      "learning_rate": 1.7686560580532718e-05,
+      "loss": 0.4398,
       "step": 6090
     },
     {
-      "epoch": 0.8930402930402931,
-      "grad_norm": 0.36276907808748354,
-      "learning_rate": 1.7712668973307314e-05,
-      "loss": 0.4446,
+      "epoch": 0.8972471662005005,
+      "grad_norm": 0.3434525247932784,
+      "learning_rate": 1.768107943931994e-05,
+      "loss": 0.4451,
       "step": 6095
     },
     {
-      "epoch": 0.8937728937728938,
-      "grad_norm": 0.3713824748145057,
-      "learning_rate": 1.7707240272107103e-05,
-      "loss": 0.4376,
+      "epoch": 0.8979832180185485,
+      "grad_norm": 0.49155308252521906,
+      "learning_rate": 1.7675592664118576e-05,
+      "loss": 0.4711,
       "step": 6100
     },
     {
-      "epoch": 0.8945054945054945,
-      "grad_norm": 0.3853780727628513,
-      "learning_rate": 1.7701805970373973e-05,
-      "loss": 0.4524,
+      "epoch": 0.8987192698365964,
+      "grad_norm": 2.8143041060928846,
+      "learning_rate": 1.7670100258953122e-05,
+      "loss": 0.5056,
       "step": 6105
     },
     {
-      "epoch": 0.8952380952380953,
-      "grad_norm": 0.37390910022738727,
-      "learning_rate": 1.7696366072056807e-05,
-      "loss": 0.4338,
+      "epoch": 0.8994553216546445,
+      "grad_norm": 0.37123973097542096,
+      "learning_rate": 1.7664602227852197e-05,
+      "loss": 0.4635,
       "step": 6110
     },
     {
-      "epoch": 0.895970695970696,
-      "grad_norm": 0.36797190937530155,
-      "learning_rate": 1.7690920581108554e-05,
-      "loss": 0.4503,
+      "epoch": 0.9001913734726925,
+      "grad_norm": 0.3388106319630038,
+      "learning_rate": 1.7659098574848545e-05,
+      "loss": 0.4525,
       "step": 6115
     },
     {
-      "epoch": 0.8967032967032967,
-      "grad_norm": 0.3600886980839345,
-      "learning_rate": 1.7685469501486223e-05,
-      "loss": 0.4346,
+      "epoch": 0.9009274252907404,
+      "grad_norm": 0.3229785070032047,
+      "learning_rate": 1.7653589303979037e-05,
+      "loss": 0.4424,
       "step": 6120
     },
     {
-      "epoch": 0.8974358974358975,
-      "grad_norm": 0.36132169150003796,
-      "learning_rate": 1.768001283715089e-05,
-      "loss": 0.4362,
+      "epoch": 0.9016634771087885,
+      "grad_norm": 0.35178580499584966,
+      "learning_rate": 1.7648074419284667e-05,
+      "loss": 0.4564,
       "step": 6125
     },
     {
-      "epoch": 0.8981684981684982,
-      "grad_norm": 0.3594715208945071,
-      "learning_rate": 1.7674550592067697e-05,
-      "loss": 0.45,
+      "epoch": 0.9023995289268365,
+      "grad_norm": 0.38055354355902876,
+      "learning_rate": 1.7642553924810536e-05,
+      "loss": 0.4374,
       "step": 6130
     },
     {
-      "epoch": 0.8989010989010989,
-      "grad_norm": 0.35705956611102607,
-      "learning_rate": 1.7669082770205816e-05,
-      "loss": 0.4471,
+      "epoch": 0.9031355807448844,
+      "grad_norm": 0.33434831557500305,
+      "learning_rate": 1.7637027824605874e-05,
+      "loss": 0.4462,
       "step": 6135
     },
     {
-      "epoch": 0.8996336996336997,
-      "grad_norm": 0.3651317182837795,
-      "learning_rate": 1.7663609375538497e-05,
-      "loss": 0.444,
+      "epoch": 0.9038716325629325,
+      "grad_norm": 0.4707904364370907,
+      "learning_rate": 1.7631496122724013e-05,
+      "loss": 0.4366,
       "step": 6140
     },
     {
-      "epoch": 0.9003663003663004,
-      "grad_norm": 0.3460089635371655,
-      "learning_rate": 1.7658130412043028e-05,
-      "loss": 0.4356,
+      "epoch": 0.9046076843809804,
+      "grad_norm": 0.36720105019429955,
+      "learning_rate": 1.7625958823222392e-05,
+      "loss": 0.4316,
       "step": 6145
     },
     {
-      "epoch": 0.9010989010989011,
-      "grad_norm": 0.49053288325000116,
-      "learning_rate": 1.7652645883700743e-05,
-      "loss": 0.459,
+      "epoch": 0.9053437361990284,
+      "grad_norm": 0.3957096565472825,
+      "learning_rate": 1.7620415930162566e-05,
+      "loss": 0.4609,
       "step": 6150
     },
     {
-      "epoch": 0.9018315018315018,
-      "grad_norm": 0.35973716200334555,
-      "learning_rate": 1.7647155794497023e-05,
-      "loss": 0.4436,
+      "epoch": 0.9060797880170764,
+      "grad_norm": 0.39061760884331537,
+      "learning_rate": 1.761486744761019e-05,
+      "loss": 0.4433,
       "step": 6155
     },
     {
-      "epoch": 0.9025641025641026,
-      "grad_norm": 0.36711220324788274,
-      "learning_rate": 1.7641660148421292e-05,
-      "loss": 0.4347,
+      "epoch": 0.9068158398351244,
+      "grad_norm": 0.38308404334046636,
+      "learning_rate": 1.7609313379635003e-05,
+      "loss": 0.4329,
       "step": 6160
     },
     {
-      "epoch": 0.9032967032967033,
-      "grad_norm": 0.47210235359387503,
-      "learning_rate": 1.7636158949467007e-05,
-      "loss": 0.4442,
+      "epoch": 0.9075518916531724,
+      "grad_norm": 0.36070664454033097,
+      "learning_rate": 1.760375373031087e-05,
+      "loss": 0.4334,
       "step": 6165
     },
     {
-      "epoch": 0.904029304029304,
-      "grad_norm": 0.3572689800679001,
-      "learning_rate": 1.763065220163166e-05,
-      "loss": 0.4505,
+      "epoch": 0.9082879434712203,
+      "grad_norm": 0.36561843004243144,
+      "learning_rate": 1.759818850371572e-05,
+      "loss": 0.439,
       "step": 6170
     },
     {
-      "epoch": 0.9047619047619048,
-      "grad_norm": 0.3736991210791609,
-      "learning_rate": 1.762513990891678e-05,
-      "loss": 0.4316,
+      "epoch": 0.9090239952892684,
+      "grad_norm": 0.359896754316887,
+      "learning_rate": 1.759261770393159e-05,
+      "loss": 0.4495,
       "step": 6175
     },
     {
-      "epoch": 0.9054945054945055,
-      "grad_norm": 0.36974271768950734,
-      "learning_rate": 1.7619622075327917e-05,
-      "loss": 0.4536,
+      "epoch": 0.9097600471073164,
+      "grad_norm": 0.34522120477333906,
+      "learning_rate": 1.758704133504461e-05,
+      "loss": 0.4499,
       "step": 6180
     },
     {
-      "epoch": 0.9062271062271062,
-      "grad_norm": 0.336076224398836,
-      "learning_rate": 1.7614098704874658e-05,
-      "loss": 0.4858,
+      "epoch": 0.9104960989253643,
+      "grad_norm": 0.3552054415454232,
+      "learning_rate": 1.758145940114497e-05,
+      "loss": 0.4433,
       "step": 6185
     },
     {
-      "epoch": 0.906959706959707,
-      "grad_norm": 0.3711009134399822,
-      "learning_rate": 1.7608569801570607e-05,
-      "loss": 0.444,
+      "epoch": 0.9112321507434124,
+      "grad_norm": 0.372895555177369,
+      "learning_rate": 1.7575871906326974e-05,
+      "loss": 0.4366,
       "step": 6190
     },
     {
-      "epoch": 0.9076923076923077,
-      "grad_norm": 0.3657491601918828,
-      "learning_rate": 1.7603035369433386e-05,
-      "loss": 0.4394,
+      "epoch": 0.9119682025614603,
+      "grad_norm": 0.6292986406740444,
+      "learning_rate": 1.7570278854688983e-05,
+      "loss": 0.4475,
       "step": 6195
     },
     {
-      "epoch": 0.9084249084249084,
-      "grad_norm": 0.3514255523290082,
-      "learning_rate": 1.759749541248464e-05,
-      "loss": 0.432,
+      "epoch": 0.9127042543795083,
+      "grad_norm": 0.3782237374359399,
+      "learning_rate": 1.7564680250333435e-05,
+      "loss": 0.4481,
       "step": 6200
     },
     {
-      "epoch": 0.9091575091575091,
-      "grad_norm": 0.3508779729073267,
-      "learning_rate": 1.7591949934750033e-05,
-      "loss": 0.4447,
+      "epoch": 0.9134403061975563,
+      "grad_norm": 0.3645107698776769,
+      "learning_rate": 1.755907609736685e-05,
+      "loss": 0.4952,
       "step": 6205
     },
     {
-      "epoch": 0.9098901098901099,
-      "grad_norm": 0.3510627678095866,
-      "learning_rate": 1.7586398940259222e-05,
-      "loss": 0.426,
+      "epoch": 0.9141763580156043,
+      "grad_norm": 0.36528595341648235,
+      "learning_rate": 1.7553466399899817e-05,
+      "loss": 0.4499,
       "step": 6210
     },
     {
-      "epoch": 0.9106227106227106,
-      "grad_norm": 0.33746241125074666,
-      "learning_rate": 1.7580842433045897e-05,
-      "loss": 0.4478,
+      "epoch": 0.9149124098336523,
+      "grad_norm": 0.47759719786471405,
+      "learning_rate": 1.7547851162046982e-05,
+      "loss": 0.4303,
       "step": 6215
     },
     {
-      "epoch": 0.9113553113553113,
-      "grad_norm": 0.3604438714051711,
-      "learning_rate": 1.7575280417147734e-05,
-      "loss": 0.4372,
+      "epoch": 0.9156484616517003,
+      "grad_norm": 0.3346974109016065,
+      "learning_rate": 1.7542230387927065e-05,
+      "loss": 0.4355,
       "step": 6220
     },
     {
-      "epoch": 0.9120879120879121,
-      "grad_norm": 0.3519089567901582,
-      "learning_rate": 1.7569712896606426e-05,
-      "loss": 0.4426,
+      "epoch": 0.9163845134697483,
+      "grad_norm": 1.6476739983025133,
+      "learning_rate": 1.753660408166284e-05,
+      "loss": 0.4423,
       "step": 6225
     },
     {
-      "epoch": 0.9128205128205128,
-      "grad_norm": 0.3685181626075242,
-      "learning_rate": 1.7564139875467657e-05,
-      "loss": 0.437,
+      "epoch": 0.9171205652877963,
+      "grad_norm": 0.3536202967982217,
+      "learning_rate": 1.7530972247381145e-05,
+      "loss": 0.4266,
       "step": 6230
     },
     {
-      "epoch": 0.9135531135531135,
-      "grad_norm": 0.36047567895159466,
-      "learning_rate": 1.755856135778111e-05,
-      "loss": 0.4396,
+      "epoch": 0.9178566171058442,
+      "grad_norm": 0.35739307239263124,
+      "learning_rate": 1.752533488921287e-05,
+      "loss": 0.4424,
       "step": 6235
     },
     {
-      "epoch": 0.9142857142857143,
-      "grad_norm": 0.35269599269066276,
-      "learning_rate": 1.7552977347600465e-05,
-      "loss": 0.4579,
+      "epoch": 0.9185926689238922,
+      "grad_norm": 0.35407192958711053,
+      "learning_rate": 1.751969201129295e-05,
+      "loss": 0.4564,
       "step": 6240
     },
     {
-      "epoch": 0.915018315018315,
-      "grad_norm": 0.4721634093341208,
-      "learning_rate": 1.754738784898339e-05,
-      "loss": 0.4388,
+      "epoch": 0.9193287207419403,
+      "grad_norm": 0.3549391504865819,
+      "learning_rate": 1.7514043617760383e-05,
+      "loss": 0.4394,
       "step": 6245
     },
     {
-      "epoch": 0.9157509157509157,
-      "grad_norm": 0.385978071550493,
-      "learning_rate": 1.7541792865991544e-05,
-      "loss": 0.4478,
+      "epoch": 0.9200647725599882,
+      "grad_norm": 0.3613180296609404,
+      "learning_rate": 1.7508389712758194e-05,
+      "loss": 0.4514,
       "step": 6250
     },
     {
-      "epoch": 0.9164835164835164,
-      "grad_norm": 0.34983217522519433,
-      "learning_rate": 1.753619240269057e-05,
-      "loss": 0.4222,
+      "epoch": 0.9208008243780362,
+      "grad_norm": 0.36808098019046603,
+      "learning_rate": 1.750273030043347e-05,
+      "loss": 0.4476,
       "step": 6255
     },
     {
-      "epoch": 0.9172161172161172,
-      "grad_norm": 0.36751855200567896,
-      "learning_rate": 1.7530586463150093e-05,
-      "loss": 0.4456,
+      "epoch": 0.9215368761960842,
+      "grad_norm": 0.36013451859199064,
+      "learning_rate": 1.749706538493733e-05,
+      "loss": 0.447,
       "step": 6260
     },
     {
-      "epoch": 0.9179487179487179,
-      "grad_norm": 0.37216248620528397,
-      "learning_rate": 1.7524975051443717e-05,
-      "loss": 0.4409,
+      "epoch": 0.9222729280141322,
+      "grad_norm": 0.4846557052431126,
+      "learning_rate": 1.749139497042492e-05,
+      "loss": 0.455,
       "step": 6265
     },
     {
-      "epoch": 0.9186813186813186,
-      "grad_norm": 0.36424445832211716,
-      "learning_rate": 1.751935817164902e-05,
-      "loss": 0.4348,
+      "epoch": 0.9230089798321802,
+      "grad_norm": 0.3824131604475871,
+      "learning_rate": 1.7485719061055432e-05,
+      "loss": 0.4624,
       "step": 6270
     },
     {
-      "epoch": 0.9194139194139194,
-      "grad_norm": 0.36065981663403024,
-      "learning_rate": 1.7513735827847563e-05,
-      "loss": 0.4386,
+      "epoch": 0.9237450316502281,
+      "grad_norm": 0.35124039668306284,
+      "learning_rate": 1.7480037660992088e-05,
+      "loss": 0.436,
       "step": 6275
     },
     {
-      "epoch": 0.9201465201465201,
-      "grad_norm": 0.3625708534458232,
-      "learning_rate": 1.750810802412487e-05,
-      "loss": 0.4654,
+      "epoch": 0.9244810834682762,
+      "grad_norm": 0.346051781465662,
+      "learning_rate": 1.7474350774402125e-05,
+      "loss": 0.4325,
       "step": 6280
     },
     {
-      "epoch": 0.9208791208791208,
-      "grad_norm": 0.3599274400943971,
-      "learning_rate": 1.7502474764570423e-05,
-      "loss": 0.4391,
+      "epoch": 0.9252171352863242,
+      "grad_norm": 0.3634607060672718,
+      "learning_rate": 1.746865840545682e-05,
+      "loss": 0.4764,
       "step": 6285
     },
     {
-      "epoch": 0.9216117216117216,
-      "grad_norm": 0.364927090667394,
-      "learning_rate": 1.749683605327769e-05,
-      "loss": 0.4437,
+      "epoch": 0.9259531871043721,
+      "grad_norm": 0.3484249956967804,
+      "learning_rate": 1.746296055833146e-05,
+      "loss": 0.439,
       "step": 6290
     },
     {
-      "epoch": 0.9223443223443224,
-      "grad_norm": 0.37376849612086793,
-      "learning_rate": 1.7491191894344086e-05,
-      "loss": 0.4497,
+      "epoch": 0.9266892389224202,
+      "grad_norm": 0.34103080377541267,
+      "learning_rate": 1.7457257237205355e-05,
+      "loss": 0.4325,
       "step": 6295
     },
     {
-      "epoch": 0.9230769230769231,
-      "grad_norm": 0.36168022937777394,
-      "learning_rate": 1.7485542291870988e-05,
-      "loss": 0.4409,
+      "epoch": 0.9274252907404681,
+      "grad_norm": 0.3728186219181222,
+      "learning_rate": 1.7451548446261828e-05,
+      "loss": 0.4516,
       "step": 6300
     },
     {
-      "epoch": 0.9238095238095239,
-      "grad_norm": 0.36368160179195114,
-      "learning_rate": 1.747988724996373e-05,
-      "loss": 0.4426,
+      "epoch": 0.9281613425585161,
+      "grad_norm": 0.3749731536453726,
+      "learning_rate": 1.744583418968822e-05,
+      "loss": 0.4212,
       "step": 6305
     },
     {
-      "epoch": 0.9245421245421246,
-      "grad_norm": 0.3729270678007294,
-      "learning_rate": 1.7474226772731594e-05,
-      "loss": 0.4612,
+      "epoch": 0.9288973943765642,
+      "grad_norm": 0.3508576954304995,
+      "learning_rate": 1.744011447167587e-05,
+      "loss": 0.4329,
       "step": 6310
     },
     {
-      "epoch": 0.9252747252747253,
-      "grad_norm": 0.35789516182360515,
-      "learning_rate": 1.746856086428782e-05,
-      "loss": 0.4604,
+      "epoch": 0.9296334461946121,
+      "grad_norm": 0.5724180355430332,
+      "learning_rate": 1.7434389296420132e-05,
+      "loss": 0.4329,
       "step": 6315
     },
     {
-      "epoch": 0.926007326007326,
-      "grad_norm": 0.35201255632556694,
-      "learning_rate": 1.746288952874958e-05,
-      "loss": 0.4607,
+      "epoch": 0.9303694980126601,
+      "grad_norm": 0.3665694394257063,
+      "learning_rate": 1.742865866812036e-05,
+      "loss": 0.4531,
       "step": 6320
     },
     {
-      "epoch": 0.9267399267399268,
-      "grad_norm": 0.38693859829795285,
-      "learning_rate": 1.7457212770238005e-05,
-      "loss": 0.4535,
+      "epoch": 0.931105549830708,
+      "grad_norm": 0.36417954650458134,
+      "learning_rate": 1.742292259097991e-05,
+      "loss": 0.4351,
       "step": 6325
     },
     {
-      "epoch": 0.9274725274725275,
-      "grad_norm": 0.34439330211437136,
-      "learning_rate": 1.7451530592878166e-05,
-      "loss": 0.4412,
+      "epoch": 0.9318416016487561,
+      "grad_norm": 0.3624096794766889,
+      "learning_rate": 1.7417181069206134e-05,
+      "loss": 0.4214,
       "step": 6330
     },
     {
-      "epoch": 0.9282051282051282,
-      "grad_norm": 0.35655682713389775,
-      "learning_rate": 1.7445843000799056e-05,
-      "loss": 0.4414,
+      "epoch": 0.9325776534668041,
+      "grad_norm": 0.36998020696000744,
+      "learning_rate": 1.741143410701037e-05,
+      "loss": 0.4451,
       "step": 6335
     },
     {
-      "epoch": 0.928937728937729,
-      "grad_norm": 0.3901791934309999,
-      "learning_rate": 1.7440149998133623e-05,
-      "loss": 0.4534,
+      "epoch": 0.933313705284852,
+      "grad_norm": 0.3668824985527801,
+      "learning_rate": 1.7405681708607956e-05,
+      "loss": 0.4415,
       "step": 6340
     },
     {
-      "epoch": 0.9296703296703297,
-      "grad_norm": 0.3410383184097131,
-      "learning_rate": 1.743445158901873e-05,
-      "loss": 0.437,
+      "epoch": 0.9340497571029001,
+      "grad_norm": 0.3519589134807377,
+      "learning_rate": 1.739992387821821e-05,
+      "loss": 0.4305,
       "step": 6345
     },
     {
-      "epoch": 0.9304029304029304,
-      "grad_norm": 0.3735459297517335,
-      "learning_rate": 1.7428747777595175e-05,
-      "loss": 0.433,
+      "epoch": 0.934785808920948,
+      "grad_norm": 0.3662117555408821,
+      "learning_rate": 1.7394160620064446e-05,
+      "loss": 0.4366,
       "step": 6350
     },
     {
-      "epoch": 0.9311355311355312,
-      "grad_norm": 0.37303056296747494,
-      "learning_rate": 1.742303856800769e-05,
-      "loss": 0.4388,
+      "epoch": 0.935521860738996,
+      "grad_norm": 0.3644631455233325,
+      "learning_rate": 1.738839193837394e-05,
+      "loss": 0.4658,
       "step": 6355
     },
     {
-      "epoch": 0.9318681318681319,
-      "grad_norm": 0.38304861292937553,
-      "learning_rate": 1.741732396440491e-05,
-      "loss": 0.4468,
+      "epoch": 0.936257912557044,
+      "grad_norm": 0.3771542118502433,
+      "learning_rate": 1.738261783737797e-05,
+      "loss": 0.4504,
       "step": 6360
     },
     {
-      "epoch": 0.9326007326007326,
-      "grad_norm": 0.3710683430964102,
-      "learning_rate": 1.741160397093941e-05,
-      "loss": 0.4478,
+      "epoch": 0.936993964375092,
+      "grad_norm": 0.35699812366864364,
+      "learning_rate": 1.7376838321311767e-05,
+      "loss": 0.442,
       "step": 6365
     },
     {
-      "epoch": 0.9333333333333333,
-      "grad_norm": 0.34469952745375604,
-      "learning_rate": 1.7405878591767668e-05,
-      "loss": 0.4127,
+      "epoch": 0.93773001619314,
+      "grad_norm": 0.3365994856786233,
+      "learning_rate": 1.737105339441455e-05,
+      "loss": 0.4323,
       "step": 6370
     },
     {
-      "epoch": 0.9340659340659341,
-      "grad_norm": 0.3608003661132213,
-      "learning_rate": 1.7400147831050084e-05,
-      "loss": 0.4265,
+      "epoch": 0.938466068011188,
+      "grad_norm": 0.35267502505215387,
+      "learning_rate": 1.73652630609295e-05,
+      "loss": 0.4276,
       "step": 6375
     },
     {
-      "epoch": 0.9347985347985348,
-      "grad_norm": 0.3591532950471118,
-      "learning_rate": 1.7394411692950967e-05,
-      "loss": 0.4352,
+      "epoch": 0.939202119829236,
+      "grad_norm": 0.39998013176522534,
+      "learning_rate": 1.735946732510376e-05,
+      "loss": 0.4396,
       "step": 6380
     },
     {
-      "epoch": 0.9355311355311355,
-      "grad_norm": 0.3815730058848865,
-      "learning_rate": 1.738867018163852e-05,
-      "loss": 0.4379,
+      "epoch": 0.939938171647284,
+      "grad_norm": 0.35133524270179606,
+      "learning_rate": 1.7353666191188438e-05,
+      "loss": 0.4415,
       "step": 6385
     },
     {
-      "epoch": 0.9362637362637363,
-      "grad_norm": 0.3781376412513171,
-      "learning_rate": 1.7382923301284875e-05,
-      "loss": 0.4255,
+      "epoch": 0.9406742234653319,
+      "grad_norm": 0.35417661046495785,
+      "learning_rate": 1.7347859663438605e-05,
+      "loss": 0.4385,
       "step": 6390
     },
     {
-      "epoch": 0.936996336996337,
-      "grad_norm": 0.36822030608402645,
-      "learning_rate": 1.7377171056066043e-05,
-      "loss": 0.439,
+      "epoch": 0.94141027528338,
+      "grad_norm": 0.34602419575372995,
+      "learning_rate": 1.734204774611329e-05,
+      "loss": 0.4552,
       "step": 6395
     },
     {
-      "epoch": 0.9377289377289377,
-      "grad_norm": 0.387825877238993,
-      "learning_rate": 1.737141345016195e-05,
-      "loss": 0.444,
+      "epoch": 0.942146327101428,
+      "grad_norm": 0.3525285571832158,
+      "learning_rate": 1.7336230443475467e-05,
+      "loss": 0.4278,
       "step": 6400
     },
     {
-      "epoch": 0.9384615384615385,
-      "grad_norm": 0.3695779787399637,
-      "learning_rate": 1.7365650487756406e-05,
-      "loss": 0.4453,
+      "epoch": 0.9428823789194759,
+      "grad_norm": 0.366003333949412,
+      "learning_rate": 1.733040775979207e-05,
+      "loss": 0.4381,
       "step": 6405
     },
     {
-      "epoch": 0.9391941391941392,
-      "grad_norm": 0.34677697199272484,
-      "learning_rate": 1.735988217303712e-05,
-      "loss": 0.4258,
+      "epoch": 0.9436184307375239,
+      "grad_norm": 0.3655463281507159,
+      "learning_rate": 1.732457969933397e-05,
+      "loss": 0.4405,
       "step": 6410
     },
     {
-      "epoch": 0.9399267399267399,
-      "grad_norm": 0.33535546429344404,
-      "learning_rate": 1.7354108510195685e-05,
-      "loss": 0.4327,
+      "epoch": 0.9443544825555719,
+      "grad_norm": 0.35214648952212846,
+      "learning_rate": 1.731874626637598e-05,
+      "loss": 0.4543,
       "step": 6415
     },
     {
-      "epoch": 0.9406593406593406,
-      "grad_norm": 0.359277392298737,
-      "learning_rate": 1.734832950342759e-05,
-      "loss": 0.4441,
+      "epoch": 0.9450905343736199,
+      "grad_norm": 0.36696287785978987,
+      "learning_rate": 1.7312907465196872e-05,
+      "loss": 0.4315,
       "step": 6420
     },
     {
-      "epoch": 0.9413919413919414,
-      "grad_norm": 0.3491643509615073,
-      "learning_rate": 1.7342545156932188e-05,
-      "loss": 0.4391,
+      "epoch": 0.9458265861916679,
+      "grad_norm": 0.38460952863517545,
+      "learning_rate": 1.7307063300079336e-05,
+      "loss": 0.4474,
       "step": 6425
     },
     {
-      "epoch": 0.9421245421245421,
-      "grad_norm": 0.34818401780870933,
-      "learning_rate": 1.7336755474912732e-05,
-      "loss": 0.4348,
+      "epoch": 0.9465626380097159,
+      "grad_norm": 0.35064441815279995,
+      "learning_rate": 1.7301213775310007e-05,
+      "loss": 0.4446,
       "step": 6430
     },
     {
-      "epoch": 0.9428571428571428,
-      "grad_norm": 0.3431591990728147,
-      "learning_rate": 1.7330960461576345e-05,
-      "loss": 0.4241,
+      "epoch": 0.9472986898277639,
+      "grad_norm": 0.3799412772308449,
+      "learning_rate": 1.7295358895179447e-05,
+      "loss": 0.46,
       "step": 6435
     },
     {
-      "epoch": 0.9435897435897436,
-      "grad_norm": 0.3560754743461474,
-      "learning_rate": 1.7325160121134025e-05,
-      "loss": 0.4398,
+      "epoch": 0.9480347416458119,
+      "grad_norm": 0.34400248191431165,
+      "learning_rate": 1.7289498663982143e-05,
+      "loss": 0.4343,
       "step": 6440
     },
     {
-      "epoch": 0.9443223443223443,
-      "grad_norm": 0.36686489644856285,
-      "learning_rate": 1.7319354457800637e-05,
-      "loss": 0.4458,
+      "epoch": 0.9487707934638598,
+      "grad_norm": 0.3567780544127945,
+      "learning_rate": 1.728363308601652e-05,
+      "loss": 0.4529,
       "step": 6445
     },
     {
-      "epoch": 0.945054945054945,
-      "grad_norm": 0.36003566234090917,
-      "learning_rate": 1.7313543475794914e-05,
-      "loss": 0.4561,
+      "epoch": 0.9495068452819079,
+      "grad_norm": 0.387213551748438,
+      "learning_rate": 1.7277762165584907e-05,
+      "loss": 0.4467,
       "step": 6450
     },
     {
-      "epoch": 0.9457875457875458,
-      "grad_norm": 0.34456401606560105,
-      "learning_rate": 1.730772717933946e-05,
-      "loss": 0.4685,
+      "epoch": 0.9502428970999558,
+      "grad_norm": 0.37209564706427456,
+      "learning_rate": 1.7271885906993565e-05,
+      "loss": 0.4399,
       "step": 6455
     },
     {
-      "epoch": 0.9465201465201465,
-      "grad_norm": 0.46881033387419574,
-      "learning_rate": 1.730190557266074e-05,
-      "loss": 0.4587,
+      "epoch": 0.9509789489180038,
+      "grad_norm": 0.35848848144937395,
+      "learning_rate": 1.7266004314552673e-05,
+      "loss": 0.4631,
       "step": 6460
     },
     {
-      "epoch": 0.9472527472527472,
-      "grad_norm": 0.3399485977148579,
-      "learning_rate": 1.7296078659989067e-05,
-      "loss": 0.4359,
+      "epoch": 0.9517150007360519,
+      "grad_norm": 0.3489687393519528,
+      "learning_rate": 1.72601173925763e-05,
+      "loss": 0.4417,
       "step": 6465
     },
     {
-      "epoch": 0.947985347985348,
-      "grad_norm": 0.37908412933430996,
-      "learning_rate": 1.7290246445558624e-05,
-      "loss": 0.4463,
+      "epoch": 0.9524510525540998,
+      "grad_norm": 0.37541684163596567,
+      "learning_rate": 1.7254225145382448e-05,
+      "loss": 0.4514,
       "step": 6470
     },
     {
-      "epoch": 0.9487179487179487,
-      "grad_norm": 0.3679538773079892,
-      "learning_rate": 1.728440893360744e-05,
-      "loss": 0.4367,
+      "epoch": 0.9531871043721478,
+      "grad_norm": 0.3472589809228936,
+      "learning_rate": 1.7248327577293018e-05,
+      "loss": 0.4271,
       "step": 6475
     },
     {
-      "epoch": 0.9494505494505494,
-      "grad_norm": 0.35266979897054596,
-      "learning_rate": 1.727856612837739e-05,
-      "loss": 0.4351,
+      "epoch": 0.9539231561901957,
+      "grad_norm": 0.3550590943995679,
+      "learning_rate": 1.724242469263381e-05,
+      "loss": 0.4226,
       "step": 6480
     },
     {
-      "epoch": 0.9501831501831501,
-      "grad_norm": 0.35482267331253475,
-      "learning_rate": 1.7272718034114202e-05,
-      "loss": 0.4299,
+      "epoch": 0.9546592080082438,
+      "grad_norm": 0.362827565511125,
+      "learning_rate": 1.7236516495734527e-05,
+      "loss": 0.4293,
       "step": 6485
     },
     {
-      "epoch": 0.9509157509157509,
-      "grad_norm": 0.3552781405115571,
-      "learning_rate": 1.7266864655067445e-05,
-      "loss": 0.4537,
+      "epoch": 0.9553952598262918,
+      "grad_norm": 0.3540006386075476,
+      "learning_rate": 1.7230602990928767e-05,
+      "loss": 0.4231,
       "step": 6490
     },
     {
-      "epoch": 0.9516483516483516,
-      "grad_norm": 0.3801275450482461,
-      "learning_rate": 1.726100599549052e-05,
-      "loss": 0.4445,
+      "epoch": 0.9561313116443397,
+      "grad_norm": 0.3569997769418889,
+      "learning_rate": 1.722468418255402e-05,
+      "loss": 0.425,
       "step": 6495
     },
     {
-      "epoch": 0.9523809523809523,
-      "grad_norm": 0.3671191261228335,
-      "learning_rate": 1.7255142059640683e-05,
-      "loss": 0.4416,
+      "epoch": 0.9568673634623878,
+      "grad_norm": 0.3691091286255934,
+      "learning_rate": 1.7218760074951668e-05,
+      "loss": 0.4494,
       "step": 6500
     },
     {
-      "epoch": 0.9531135531135531,
-      "grad_norm": 0.35384501680106467,
-      "learning_rate": 1.724927285177901e-05,
-      "loss": 0.4361,
+      "epoch": 0.9576034152804357,
+      "grad_norm": 0.3606401671759627,
+      "learning_rate": 1.7212830672466983e-05,
+      "loss": 0.4432,
       "step": 6505
     },
     {
-      "epoch": 0.9538461538461539,
-      "grad_norm": 0.3931706899709501,
-      "learning_rate": 1.7243398376170408e-05,
-      "loss": 0.4518,
+      "epoch": 0.9583394670984837,
+      "grad_norm": 0.35153856178393444,
+      "learning_rate": 1.7206895979449116e-05,
+      "loss": 0.4439,
       "step": 6510
     },
     {
-      "epoch": 0.9545787545787546,
-      "grad_norm": 0.3691599143921214,
-      "learning_rate": 1.7237518637083622e-05,
-      "loss": 0.4317,
+      "epoch": 0.9590755189165318,
+      "grad_norm": 0.3481970728848475,
+      "learning_rate": 1.7200956000251098e-05,
+      "loss": 0.4328,
       "step": 6515
     },
     {
-      "epoch": 0.9553113553113554,
-      "grad_norm": 0.3758014577879604,
-      "learning_rate": 1.7231633638791208e-05,
-      "loss": 0.4507,
+      "epoch": 0.9598115707345797,
+      "grad_norm": 0.36545719654945297,
+      "learning_rate": 1.719501073922984e-05,
+      "loss": 0.4199,
       "step": 6520
     },
     {
-      "epoch": 0.9560439560439561,
-      "grad_norm": 0.35354209070515097,
-      "learning_rate": 1.722574338556956e-05,
-      "loss": 0.4442,
+      "epoch": 0.9605476225526277,
+      "grad_norm": 0.3585760396318208,
+      "learning_rate": 1.718906020074613e-05,
+      "loss": 0.4445,
       "step": 6525
     },
     {
-      "epoch": 0.9567765567765568,
-      "grad_norm": 0.35157129831484946,
-      "learning_rate": 1.721984788169887e-05,
-      "loss": 0.4575,
+      "epoch": 0.9612836743706757,
+      "grad_norm": 0.3516536066245614,
+      "learning_rate": 1.718310438916462e-05,
+      "loss": 0.4289,
       "step": 6530
     },
     {
-      "epoch": 0.9575091575091575,
-      "grad_norm": 0.3846250090546446,
-      "learning_rate": 1.7213947131463172e-05,
-      "loss": 0.4451,
+      "epoch": 0.9620197261887237,
+      "grad_norm": 0.35861427654949146,
+      "learning_rate": 1.7177143308853837e-05,
+      "loss": 0.4423,
       "step": 6535
     },
     {
-      "epoch": 0.9582417582417583,
-      "grad_norm": 1.284228046597185,
-      "learning_rate": 1.7208041139150285e-05,
-      "loss": 0.4341,
+      "epoch": 0.9627557780067717,
+      "grad_norm": 0.36958938982697254,
+      "learning_rate": 1.7171176964186162e-05,
+      "loss": 0.4219,
       "step": 6540
     },
     {
-      "epoch": 0.958974358974359,
-      "grad_norm": 0.3788015557626036,
-      "learning_rate": 1.720212990905185e-05,
-      "loss": 0.4212,
+      "epoch": 0.9634918298248196,
+      "grad_norm": 0.337681353132239,
+      "learning_rate": 1.7165205359537852e-05,
+      "loss": 0.423,
       "step": 6545
     },
     {
-      "epoch": 0.9597069597069597,
-      "grad_norm": 0.3779066676468523,
-      "learning_rate": 1.719621344546332e-05,
-      "loss": 0.4436,
+      "epoch": 0.9642278816428677,
+      "grad_norm": 0.3681033444413669,
+      "learning_rate": 1.715922849928901e-05,
+      "loss": 0.4416,
       "step": 6550
     },
     {
-      "epoch": 0.9604395604395605,
-      "grad_norm": 0.35572791641424206,
-      "learning_rate": 1.7190291752683942e-05,
-      "loss": 0.4473,
+      "epoch": 0.9649639334609157,
+      "grad_norm": 0.35824920695173623,
+      "learning_rate": 1.71532463878236e-05,
+      "loss": 0.4514,
       "step": 6555
     },
     {
-      "epoch": 0.9611721611721612,
-      "grad_norm": 0.37632881645512,
-      "learning_rate": 1.7184364835016768e-05,
-      "loss": 0.4437,
+      "epoch": 0.9656999852789636,
+      "grad_norm": 0.35027880844473847,
+      "learning_rate": 1.7147259029529434e-05,
+      "loss": 0.4348,
       "step": 6560
     },
     {
-      "epoch": 0.9619047619047619,
-      "grad_norm": 0.3463323921208554,
-      "learning_rate": 1.7178432696768634e-05,
-      "loss": 0.4343,
+      "epoch": 0.9664360370970116,
+      "grad_norm": 0.35007600047694154,
+      "learning_rate": 1.7141266428798173e-05,
+      "loss": 0.4628,
       "step": 6565
     },
     {
-      "epoch": 0.9626373626373627,
-      "grad_norm": 0.34930030130427925,
-      "learning_rate": 1.717249534225019e-05,
-      "loss": 0.4431,
+      "epoch": 0.9671720889150596,
+      "grad_norm": 0.3424031971108177,
+      "learning_rate": 1.7135268590025327e-05,
+      "loss": 0.4378,
       "step": 6570
     },
     {
-      "epoch": 0.9633699633699634,
-      "grad_norm": 0.3579662908578417,
-      "learning_rate": 1.716655277577586e-05,
-      "loss": 0.4344,
+      "epoch": 0.9679081407331076,
+      "grad_norm": 0.3480667011906856,
+      "learning_rate": 1.712926551761025e-05,
+      "loss": 0.4362,
       "step": 6575
     },
     {
-      "epoch": 0.9641025641025641,
-      "grad_norm": 0.34469614850564645,
-      "learning_rate": 1.7160605001663867e-05,
-      "loss": 0.4349,
+      "epoch": 0.9686441925511556,
+      "grad_norm": 0.3492989838682148,
+      "learning_rate": 1.7123257215956127e-05,
+      "loss": 0.4317,
       "step": 6580
     },
     {
-      "epoch": 0.9648351648351648,
-      "grad_norm": 0.3358036188767999,
-      "learning_rate": 1.715465202423621e-05,
-      "loss": 0.4476,
+      "epoch": 0.9693802443692036,
+      "grad_norm": 0.3647848112736977,
+      "learning_rate": 1.711724368946998e-05,
+      "loss": 0.4379,
       "step": 6585
     },
     {
-      "epoch": 0.9655677655677656,
-      "grad_norm": 0.37389608845541517,
-      "learning_rate": 1.714869384781867e-05,
-      "loss": 0.4569,
+      "epoch": 0.9701162961872516,
+      "grad_norm": 0.3647649308439903,
+      "learning_rate": 1.7111224942562663e-05,
+      "loss": 0.4493,
       "step": 6590
     },
     {
-      "epoch": 0.9663003663003663,
-      "grad_norm": 0.3532112957351536,
-      "learning_rate": 1.714273047674081e-05,
-      "loss": 0.4298,
+      "epoch": 0.9708523480052996,
+      "grad_norm": 0.3753953789277182,
+      "learning_rate": 1.7105200979648874e-05,
+      "loss": 0.4381,
       "step": 6595
     },
     {
-      "epoch": 0.967032967032967,
-      "grad_norm": 0.3615175520371604,
-      "learning_rate": 1.713676191533596e-05,
-      "loss": 0.438,
+      "epoch": 0.9715883998233475,
+      "grad_norm": 0.37606826388414516,
+      "learning_rate": 1.709917180514711e-05,
+      "loss": 0.4624,
       "step": 6600
     },
     {
-      "epoch": 0.9677655677655678,
-      "grad_norm": 0.45637238587490664,
-      "learning_rate": 1.7130788167941235e-05,
-      "loss": 0.4735,
+      "epoch": 0.9723244516413956,
+      "grad_norm": 0.3737809109027871,
+      "learning_rate": 1.7093137423479714e-05,
+      "loss": 0.4232,
       "step": 6605
     },
     {
-      "epoch": 0.9684981684981685,
-      "grad_norm": 0.37272041568938485,
-      "learning_rate": 1.7124809238897508e-05,
-      "loss": 0.4399,
+      "epoch": 0.9730605034594435,
+      "grad_norm": 0.3780741713516723,
+      "learning_rate": 1.708709783907284e-05,
+      "loss": 0.4386,
       "step": 6610
     },
     {
-      "epoch": 0.9692307692307692,
-      "grad_norm": 0.37066324869804335,
-      "learning_rate": 1.7118825132549416e-05,
-      "loss": 0.448,
+      "epoch": 0.9737965552774915,
+      "grad_norm": 0.3686313080726544,
+      "learning_rate": 1.7081053056356452e-05,
+      "loss": 0.4561,
       "step": 6615
     },
     {
-      "epoch": 0.96996336996337,
-      "grad_norm": 0.38415151064760394,
-      "learning_rate": 1.711283585324536e-05,
-      "loss": 0.4382,
+      "epoch": 0.9745326070955396,
+      "grad_norm": 0.3557694193587092,
+      "learning_rate": 1.7075003079764337e-05,
+      "loss": 0.4234,
       "step": 6620
     },
     {
-      "epoch": 0.9706959706959707,
-      "grad_norm": 0.38204200328668275,
-      "learning_rate": 1.710684140533751e-05,
-      "loss": 0.455,
+      "epoch": 0.9752686589135875,
+      "grad_norm": 0.4071086693129615,
+      "learning_rate": 1.7068947913734085e-05,
+      "loss": 0.4395,
       "step": 6625
     },
     {
-      "epoch": 0.9714285714285714,
-      "grad_norm": 0.3676552434938297,
-      "learning_rate": 1.7100841793181773e-05,
-      "loss": 0.4404,
+      "epoch": 0.9760047107316355,
+      "grad_norm": 0.3453854466174774,
+      "learning_rate": 1.7062887562707097e-05,
+      "loss": 0.4232,
       "step": 6630
     },
     {
-      "epoch": 0.9721611721611721,
-      "grad_norm": 0.36719276204679707,
-      "learning_rate": 1.709483702113783e-05,
-      "loss": 0.4375,
+      "epoch": 0.9767407625496835,
+      "grad_norm": 0.3697484920484151,
+      "learning_rate": 1.7056822031128572e-05,
+      "loss": 0.4353,
       "step": 6635
     },
     {
-      "epoch": 0.9728937728937729,
-      "grad_norm": 0.5109971002491437,
-      "learning_rate": 1.7088827093569092e-05,
-      "loss": 0.481,
+      "epoch": 0.9774768143677315,
+      "grad_norm": 0.3654491247280484,
+      "learning_rate": 1.7050751323447515e-05,
+      "loss": 0.4331,
       "step": 6640
     },
     {
-      "epoch": 0.9736263736263736,
-      "grad_norm": 0.38455542337125864,
-      "learning_rate": 1.7082812014842732e-05,
+      "epoch": 0.9782128661857795,
+      "grad_norm": 0.36223079918316325,
+      "learning_rate": 1.7044675444116726e-05,
       "loss": 0.441,
       "step": 6645
     },
     {
-      "epoch": 0.9743589743589743,
-      "grad_norm": 0.35342114488002463,
-      "learning_rate": 1.7076791789329655e-05,
-      "loss": 0.4305,
+      "epoch": 0.9789489180038274,
+      "grad_norm": 0.36604867602552976,
+      "learning_rate": 1.703859439759279e-05,
+      "loss": 0.4442,
       "step": 6650
     },
     {
-      "epoch": 0.9750915750915751,
-      "grad_norm": 0.3581437314772134,
-      "learning_rate": 1.707076642140451e-05,
-      "loss": 0.4464,
+      "epoch": 0.9796849698218755,
+      "grad_norm": 0.4501107912031965,
+      "learning_rate": 1.703250818833609e-05,
+      "loss": 0.4406,
       "step": 6655
     },
     {
-      "epoch": 0.9758241758241758,
-      "grad_norm": 0.3615948884476211,
-      "learning_rate": 1.7064735915445683e-05,
-      "loss": 0.4365,
+      "epoch": 0.9804210216399234,
+      "grad_norm": 0.3657745148199479,
+      "learning_rate": 1.7026416820810797e-05,
+      "loss": 0.4385,
       "step": 6660
     },
     {
-      "epoch": 0.9765567765567765,
-      "grad_norm": 0.3411342605948674,
-      "learning_rate": 1.7058700275835293e-05,
-      "loss": 0.431,
+      "epoch": 0.9811570734579714,
+      "grad_norm": 0.37544557601913253,
+      "learning_rate": 1.7020320299484868e-05,
+      "loss": 0.4357,
       "step": 6665
     },
     {
-      "epoch": 0.9772893772893773,
-      "grad_norm": 0.37749787772603005,
-      "learning_rate": 1.705265950695919e-05,
-      "loss": 0.4436,
+      "epoch": 0.9818931252760195,
+      "grad_norm": 0.3729086931283695,
+      "learning_rate": 1.7014218628830026e-05,
+      "loss": 0.4457,
       "step": 6670
     },
     {
-      "epoch": 0.978021978021978,
-      "grad_norm": 0.3569827914060955,
-      "learning_rate": 1.7046613613206954e-05,
-      "loss": 0.4497,
+      "epoch": 0.9826291770940674,
+      "grad_norm": 0.3547966965289486,
+      "learning_rate": 1.7008111813321786e-05,
+      "loss": 0.4493,
       "step": 6675
     },
     {
-      "epoch": 0.9787545787545787,
-      "grad_norm": 0.3563775667772839,
-      "learning_rate": 1.7040562598971885e-05,
-      "loss": 0.4228,
+      "epoch": 0.9833652289121154,
+      "grad_norm": 0.35360612224713783,
+      "learning_rate": 1.700199985743943e-05,
+      "loss": 0.4332,
       "step": 6680
     },
     {
-      "epoch": 0.9794871794871794,
-      "grad_norm": 0.36042392380094623,
-      "learning_rate": 1.7034506468651e-05,
-      "loss": 0.4409,
+      "epoch": 0.9841012807301635,
+      "grad_norm": 0.34229050450837223,
+      "learning_rate": 1.699588276566601e-05,
+      "loss": 0.4352,
       "step": 6685
     },
     {
-      "epoch": 0.9802197802197802,
-      "grad_norm": 0.33415031185831234,
-      "learning_rate": 1.702844522664504e-05,
-      "loss": 0.4275,
+      "epoch": 0.9848373325482114,
+      "grad_norm": 0.36038163944276935,
+      "learning_rate": 1.6989760542488345e-05,
+      "loss": 0.4547,
       "step": 6690
     },
     {
-      "epoch": 0.9809523809523809,
-      "grad_norm": 0.33092743255070933,
-      "learning_rate": 1.7022378877358466e-05,
-      "loss": 0.4433,
+      "epoch": 0.9855733843662594,
+      "grad_norm": 0.37582589633438,
+      "learning_rate": 1.6983633192397023e-05,
+      "loss": 0.4421,
       "step": 6695
     },
     {
-      "epoch": 0.9816849816849816,
-      "grad_norm": 0.3751336919353944,
-      "learning_rate": 1.7016307425199437e-05,
-      "loss": 0.4477,
+      "epoch": 0.9863094361843073,
+      "grad_norm": 0.3752879091825568,
+      "learning_rate": 1.697750071988639e-05,
+      "loss": 0.4251,
       "step": 6700
     },
     {
-      "epoch": 0.9824175824175824,
-      "grad_norm": 0.35752258164803286,
-      "learning_rate": 1.7010230874579828e-05,
-      "loss": 0.4206,
+      "epoch": 0.9870454880023554,
+      "grad_norm": 0.3629269641916705,
+      "learning_rate": 1.697136312945455e-05,
+      "loss": 0.4491,
       "step": 6705
     },
     {
-      "epoch": 0.9831501831501831,
-      "grad_norm": 0.37406972717380194,
-      "learning_rate": 1.7004149229915218e-05,
-      "loss": 0.4414,
+      "epoch": 0.9877815398204034,
+      "grad_norm": 0.3601063193255718,
+      "learning_rate": 1.696522042560335e-05,
+      "loss": 0.4332,
       "step": 6710
     },
     {
-      "epoch": 0.9838827838827838,
-      "grad_norm": 0.35195564329231965,
-      "learning_rate": 1.699806249562488e-05,
-      "loss": 0.4389,
+      "epoch": 0.9885175916384513,
+      "grad_norm": 0.35379971824952605,
+      "learning_rate": 1.6959072612838402e-05,
+      "loss": 0.444,
       "step": 6715
     },
     {
-      "epoch": 0.9846153846153847,
-      "grad_norm": 0.3386954263403885,
-      "learning_rate": 1.6991970676131803e-05,
-      "loss": 0.4472,
+      "epoch": 0.9892536434564994,
+      "grad_norm": 0.35869165263922875,
+      "learning_rate": 1.695291969566906e-05,
+      "loss": 0.4257,
       "step": 6720
     },
     {
-      "epoch": 0.9853479853479854,
-      "grad_norm": 0.34771883733053655,
-      "learning_rate": 1.698587377586266e-05,
-      "loss": 0.4538,
+      "epoch": 0.9899896952745473,
+      "grad_norm": 0.43804822187478154,
+      "learning_rate": 1.6946761678608427e-05,
+      "loss": 0.4496,
       "step": 6725
     },
     {
-      "epoch": 0.9860805860805861,
-      "grad_norm": 0.3721375026399225,
-      "learning_rate": 1.6979771799247804e-05,
-      "loss": 0.4472,
+      "epoch": 0.9907257470925953,
+      "grad_norm": 0.34786445151583756,
+      "learning_rate": 1.6940598566173332e-05,
+      "loss": 0.4334,
       "step": 6730
     },
     {
-      "epoch": 0.9868131868131869,
-      "grad_norm": 0.36602925797704333,
-      "learning_rate": 1.6973664750721298e-05,
-      "loss": 0.4414,
+      "epoch": 0.9914617989106433,
+      "grad_norm": 0.3821109055666706,
+      "learning_rate": 1.6934430362884358e-05,
+      "loss": 0.4436,
       "step": 6735
     },
     {
-      "epoch": 0.9875457875457876,
-      "grad_norm": 0.35450346411272515,
-      "learning_rate": 1.696755263472088e-05,
-      "loss": 0.4396,
+      "epoch": 0.9921978507286913,
+      "grad_norm": 0.3760672110821931,
+      "learning_rate": 1.6928257073265813e-05,
+      "loss": 0.444,
       "step": 6740
     },
     {
-      "epoch": 0.9882783882783883,
-      "grad_norm": 0.36493321114968236,
-      "learning_rate": 1.6961435455687977e-05,
-      "loss": 0.4343,
+      "epoch": 0.9929339025467393,
+      "grad_norm": 0.3590919223385004,
+      "learning_rate": 1.6922078701845735e-05,
+      "loss": 0.4385,
       "step": 6745
     },
     {
-      "epoch": 0.989010989010989,
-      "grad_norm": 0.34588020475657766,
-      "learning_rate": 1.6955313218067683e-05,
-      "loss": 0.4273,
+      "epoch": 0.9936699543647873,
+      "grad_norm": 0.36928170979574315,
+      "learning_rate": 1.6915895253155903e-05,
+      "loss": 0.4352,
       "step": 6750
     },
     {
-      "epoch": 0.9897435897435898,
-      "grad_norm": 0.37991788360791556,
-      "learning_rate": 1.694918592630878e-05,
-      "loss": 0.445,
+      "epoch": 0.9944060061828353,
+      "grad_norm": 0.35372962548149794,
+      "learning_rate": 1.69097067317318e-05,
+      "loss": 0.4386,
       "step": 6755
     },
     {
-      "epoch": 0.9904761904761905,
-      "grad_norm": 0.3396032374471012,
-      "learning_rate": 1.6943053584863713e-05,
-      "loss": 0.4228,
+      "epoch": 0.9951420580008833,
+      "grad_norm": 0.36209291653208947,
+      "learning_rate": 1.690351314211264e-05,
+      "loss": 0.4344,
       "step": 6760
     },
     {
-      "epoch": 0.9912087912087912,
-      "grad_norm": 0.36592387012699884,
-      "learning_rate": 1.6936916198188607e-05,
-      "loss": 0.4428,
+      "epoch": 0.9958781098189312,
+      "grad_norm": 0.34743601508889976,
+      "learning_rate": 1.689731448884136e-05,
+      "loss": 0.4347,
       "step": 6765
     },
     {
-      "epoch": 0.991941391941392,
-      "grad_norm": 0.34982183026324387,
-      "learning_rate": 1.693077377074324e-05,
-      "loss": 0.4442,
+      "epoch": 0.9966141616369792,
+      "grad_norm": 0.34687773366384644,
+      "learning_rate": 1.6891110776464595e-05,
+      "loss": 0.4233,
       "step": 6770
     },
     {
-      "epoch": 0.9926739926739927,
-      "grad_norm": 0.3522221002127369,
-      "learning_rate": 1.6924626306991064e-05,
-      "loss": 0.4311,
+      "epoch": 0.9973502134550273,
+      "grad_norm": 0.40739649568714137,
+      "learning_rate": 1.6884902009532712e-05,
+      "loss": 0.4392,
       "step": 6775
     },
     {
-      "epoch": 0.9934065934065934,
-      "grad_norm": 0.35422036214466257,
-      "learning_rate": 1.691847381139919e-05,
-      "loss": 0.4357,
+      "epoch": 0.9980862652730752,
+      "grad_norm": 0.36589283999767597,
+      "learning_rate": 1.6878688192599766e-05,
+      "loss": 0.4301,
       "step": 6780
     },
     {
-      "epoch": 0.9941391941391942,
-      "grad_norm": 0.35280747117561606,
-      "learning_rate": 1.6912316288438376e-05,
-      "loss": 0.43,
+      "epoch": 0.9988223170911232,
+      "grad_norm": 0.3461679332252419,
+      "learning_rate": 1.6872469330223533e-05,
+      "loss": 0.441,
       "step": 6785
     },
     {
-      "epoch": 0.9948717948717949,
-      "grad_norm": 0.3421279801067773,
-      "learning_rate": 1.6906153742583043e-05,
-      "loss": 0.4276,
+      "epoch": 0.9995583689091712,
+      "grad_norm": 0.3767409034985814,
+      "learning_rate": 1.6866245426965474e-05,
+      "loss": 0.4304,
       "step": 6790
     },
     {
-      "epoch": 0.9956043956043956,
-      "grad_norm": 0.33916936148692983,
-      "learning_rate": 1.689998617831126e-05,
-      "loss": 0.4316,
+      "epoch": 1.0,
+      "eval_loss": 0.43871477246284485,
+      "eval_runtime": 1.5515,
+      "eval_samples_per_second": 15.469,
+      "eval_steps_per_second": 0.645,
+      "step": 6793
+    },
+    {
+      "epoch": 1.0002944207272193,
+      "grad_norm": 0.3866995374176673,
+      "learning_rate": 1.6860016487390755e-05,
+      "loss": 0.393,
       "step": 6795
     },
     {
-      "epoch": 0.9963369963369964,
-      "grad_norm": 0.3739503000291955,
-      "learning_rate": 1.6893813600104733e-05,
-      "loss": 0.4415,
+      "epoch": 1.0010304725452672,
+      "grad_norm": 0.3999949768145471,
+      "learning_rate": 1.685378251606824e-05,
+      "loss": 0.3513,
       "step": 6800
     },
     {
-      "epoch": 0.9970695970695971,
-      "grad_norm": 0.3674210126216178,
-      "learning_rate": 1.688763601244883e-05,
-      "loss": 0.4369,
+      "epoch": 1.0017665243633151,
+      "grad_norm": 0.39338192031228963,
+      "learning_rate": 1.6847543517570472e-05,
+      "loss": 0.3479,
       "step": 6805
     },
     {
-      "epoch": 0.9978021978021978,
-      "grad_norm": 0.3474202293916676,
-      "learning_rate": 1.688145341983254e-05,
-      "loss": 0.456,
+      "epoch": 1.002502576181363,
+      "grad_norm": 0.383222857954703,
+      "learning_rate": 1.6841299496473696e-05,
+      "loss": 0.3409,
       "step": 6810
     },
     {
-      "epoch": 0.9985347985347985,
-      "grad_norm": 0.36861289532532876,
-      "learning_rate": 1.6875265826748506e-05,
-      "loss": 0.4436,
+      "epoch": 1.0032386279994112,
+      "grad_norm": 0.38779269601567395,
+      "learning_rate": 1.6835050457357833e-05,
+      "loss": 0.3432,
       "step": 6815
     },
     {
-      "epoch": 0.9992673992673993,
-      "grad_norm": 0.3574161164208663,
-      "learning_rate": 1.6869073237692987e-05,
-      "loss": 0.451,
+      "epoch": 1.0039746798174591,
+      "grad_norm": 0.37810640687408825,
+      "learning_rate": 1.682879640480648e-05,
+      "loss": 0.3231,
       "step": 6820
     },
     {
-      "epoch": 1.0,
-      "grad_norm": 0.3408889832195101,
-      "learning_rate": 1.686287565716589e-05,
-      "loss": 0.4316,
-      "step": 6825
-    },
-    {
-      "epoch": 1.0,
-      "eval_loss": 0.40975621342658997,
-      "eval_runtime": 1.5969,
-      "eval_samples_per_second": 15.03,
-      "eval_steps_per_second": 0.626,
+      "epoch": 1.004710731635507,
+      "grad_norm": 0.3651737100583659,
+      "learning_rate": 1.6822537343406925e-05,
+      "loss": 0.3216,
       "step": 6825
     },
     {
-      "epoch": 1.0007326007326007,
-      "grad_norm": 0.36201863726007955,
-      "learning_rate": 1.685667308967073e-05,
-      "loss": 0.3413,
+      "epoch": 1.0054467834535552,
+      "grad_norm": 0.36337384098568465,
+      "learning_rate": 1.6816273277750116e-05,
+      "loss": 0.3303,
       "step": 6830
     },
     {
-      "epoch": 1.0014652014652015,
-      "grad_norm": 0.3902424360566943,
-      "learning_rate": 1.6850465539714667e-05,
-      "loss": 0.3322,
+      "epoch": 1.0061828352716031,
+      "grad_norm": 0.4087666983005035,
+      "learning_rate": 1.6810004212430677e-05,
+      "loss": 0.3304,
       "step": 6835
     },
     {
-      "epoch": 1.0021978021978022,
-      "grad_norm": 0.3975415065836676,
-      "learning_rate": 1.6844253011808468e-05,
-      "loss": 0.3273,
+      "epoch": 1.006918887089651,
+      "grad_norm": 0.3807588186952967,
+      "learning_rate": 1.6803730152046905e-05,
+      "loss": 0.3217,
       "step": 6840
     },
     {
-      "epoch": 1.002930402930403,
-      "grad_norm": 0.4061898685848133,
-      "learning_rate": 1.6838035510466516e-05,
-      "loss": 0.3322,
+      "epoch": 1.0076549389076992,
+      "grad_norm": 0.3759058072774958,
+      "learning_rate": 1.6797451101200748e-05,
+      "loss": 0.3243,
       "step": 6845
     },
     {
-      "epoch": 1.0036630036630036,
-      "grad_norm": 0.3593019021745724,
-      "learning_rate": 1.6831813040206817e-05,
-      "loss": 0.3217,
+      "epoch": 1.0083909907257471,
+      "grad_norm": 0.3861754610930233,
+      "learning_rate": 1.6791167064497828e-05,
+      "loss": 0.3207,
       "step": 6850
     },
     {
-      "epoch": 1.0043956043956044,
-      "grad_norm": 0.36837764560120567,
-      "learning_rate": 1.6825585605550982e-05,
-      "loss": 0.339,
+      "epoch": 1.009127042543795,
+      "grad_norm": 0.3622151306663367,
+      "learning_rate": 1.6784878046547414e-05,
+      "loss": 0.3311,
       "step": 6855
     },
     {
-      "epoch": 1.005128205128205,
-      "grad_norm": 0.3867234105785456,
-      "learning_rate": 1.6819353211024227e-05,
-      "loss": 0.3418,
+      "epoch": 1.0098630943618432,
+      "grad_norm": 0.3649932755248596,
+      "learning_rate": 1.6778584051962434e-05,
+      "loss": 0.3389,
       "step": 6860
     },
     {
-      "epoch": 1.0058608058608058,
-      "grad_norm": 0.3812144593341558,
-      "learning_rate": 1.681311586115538e-05,
-      "loss": 0.3458,
+      "epoch": 1.010599146179891,
+      "grad_norm": 0.3766699822581901,
+      "learning_rate": 1.6772285085359465e-05,
+      "loss": 0.3313,
       "step": 6865
     },
     {
-      "epoch": 1.0065934065934066,
-      "grad_norm": 0.37054904889141077,
-      "learning_rate": 1.680687356047686e-05,
-      "loss": 0.3429,
+      "epoch": 1.011335197997939,
+      "grad_norm": 0.3562643210167549,
+      "learning_rate": 1.676598115135873e-05,
+      "loss": 0.3243,
       "step": 6870
     },
     {
-      "epoch": 1.0073260073260073,
-      "grad_norm": 0.3598095442970735,
-      "learning_rate": 1.6800626313524696e-05,
-      "loss": 0.3561,
+      "epoch": 1.012071249815987,
+      "grad_norm": 0.3625978112414053,
+      "learning_rate": 1.67596722545841e-05,
+      "loss": 0.3311,
       "step": 6875
     },
     {
-      "epoch": 1.008058608058608,
-      "grad_norm": 0.3675041920360404,
-      "learning_rate": 1.67943741248385e-05,
-      "loss": 0.3415,
+      "epoch": 1.012807301634035,
+      "grad_norm": 0.36347877354101427,
+      "learning_rate": 1.6753358399663073e-05,
+      "loss": 0.3365,
       "step": 6880
     },
     {
-      "epoch": 1.0087912087912088,
-      "grad_norm": 0.38062489277406575,
-      "learning_rate": 1.6788116998961476e-05,
-      "loss": 0.3247,
+      "epoch": 1.013543353452083,
+      "grad_norm": 0.3837021753325475,
+      "learning_rate": 1.6747039591226802e-05,
+      "loss": 0.3441,
       "step": 6885
     },
     {
-      "epoch": 1.0095238095238095,
-      "grad_norm": 0.3605291982310506,
-      "learning_rate": 1.6781854940440424e-05,
-      "loss": 0.3345,
+      "epoch": 1.014279405270131,
+      "grad_norm": 0.3695457828638176,
+      "learning_rate": 1.6740715833910064e-05,
+      "loss": 0.3215,
       "step": 6890
     },
     {
-      "epoch": 1.0102564102564102,
-      "grad_norm": 0.36236834595661455,
-      "learning_rate": 1.677558795382572e-05,
-      "loss": 0.3682,
+      "epoch": 1.015015457088179,
+      "grad_norm": 0.3702978633218317,
+      "learning_rate": 1.6734387132351268e-05,
+      "loss": 0.3375,
       "step": 6895
     },
     {
-      "epoch": 1.010989010989011,
-      "grad_norm": 0.3690521144299775,
-      "learning_rate": 1.6769316043671327e-05,
-      "loss": 0.3386,
+      "epoch": 1.015751508906227,
+      "grad_norm": 0.3845715172792146,
+      "learning_rate": 1.672805349119244e-05,
+      "loss": 0.3265,
       "step": 6900
     },
     {
-      "epoch": 1.0117216117216117,
-      "grad_norm": 0.3760777588079089,
-      "learning_rate": 1.6763039214534782e-05,
-      "loss": 0.3295,
+      "epoch": 1.016487560724275,
+      "grad_norm": 0.37701638039133323,
+      "learning_rate": 1.6721714915079242e-05,
+      "loss": 0.3315,
       "step": 6905
     },
     {
-      "epoch": 1.0124542124542124,
-      "grad_norm": 0.3726828659631818,
-      "learning_rate": 1.6756757470977194e-05,
-      "loss": 0.3262,
+      "epoch": 1.017223612542323,
+      "grad_norm": 0.3533591923899028,
+      "learning_rate": 1.671537140866095e-05,
+      "loss": 0.3297,
       "step": 6910
     },
     {
-      "epoch": 1.0131868131868131,
-      "grad_norm": 0.3905666821718478,
-      "learning_rate": 1.675047081756325e-05,
-      "loss": 0.3227,
+      "epoch": 1.017959664360371,
+      "grad_norm": 0.3531464949026519,
+      "learning_rate": 1.6709022976590457e-05,
+      "loss": 0.3259,
       "step": 6915
     },
     {
-      "epoch": 1.0139194139194139,
-      "grad_norm": 0.3830327834647385,
-      "learning_rate": 1.67441792588612e-05,
-      "loss": 0.3326,
+      "epoch": 1.018695716178419,
+      "grad_norm": 0.3625200138845915,
+      "learning_rate": 1.6702669623524273e-05,
+      "loss": 0.3274,
       "step": 6920
     },
     {
-      "epoch": 1.0146520146520146,
-      "grad_norm": 0.3571195382115666,
-      "learning_rate": 1.673788279944286e-05,
-      "loss": 0.3364,
+      "epoch": 1.0194317679964668,
+      "grad_norm": 0.36846966065311787,
+      "learning_rate": 1.669631135412251e-05,
+      "loss": 0.3248,
       "step": 6925
     },
     {
-      "epoch": 1.0153846153846153,
-      "grad_norm": 0.3576188937698423,
-      "learning_rate": 1.6731581443883603e-05,
-      "loss": 0.3409,
+      "epoch": 1.020167819814515,
+      "grad_norm": 0.37776175075594376,
+      "learning_rate": 1.668994817304889e-05,
+      "loss": 0.3457,
       "step": 6930
     },
     {
-      "epoch": 1.016117216117216,
-      "grad_norm": 0.3643609147540888,
-      "learning_rate": 1.672527519676237e-05,
-      "loss": 0.3341,
+      "epoch": 1.020903871632563,
+      "grad_norm": 0.36513036919289815,
+      "learning_rate": 1.668358008497074e-05,
+      "loss": 0.3293,
       "step": 6935
     },
     {
-      "epoch": 1.0168498168498168,
-      "grad_norm": 0.3801775662889062,
-      "learning_rate": 1.6718964062661646e-05,
-      "loss": 0.3394,
+      "epoch": 1.0216399234506108,
+      "grad_norm": 0.35371926801796266,
+      "learning_rate": 1.667720709455898e-05,
+      "loss": 0.3136,
       "step": 6940
     },
     {
-      "epoch": 1.0175824175824175,
-      "grad_norm": 0.3639872213399938,
-      "learning_rate": 1.6712648046167478e-05,
-      "loss": 0.3377,
+      "epoch": 1.022375975268659,
+      "grad_norm": 0.3837532723126793,
+      "learning_rate": 1.667082920648813e-05,
+      "loss": 0.326,
       "step": 6945
     },
     {
-      "epoch": 1.0183150183150182,
-      "grad_norm": 0.3782266895702019,
-      "learning_rate": 1.670632715186944e-05,
-      "loss": 0.3282,
+      "epoch": 1.023112027086707,
+      "grad_norm": 0.3716385640915241,
+      "learning_rate": 1.6664446425436306e-05,
+      "loss": 0.3298,
       "step": 6950
     },
     {
-      "epoch": 1.019047619047619,
-      "grad_norm": 0.8201464850800466,
-      "learning_rate": 1.6700001384360677e-05,
-      "loss": 0.3354,
+      "epoch": 1.0238480789047548,
+      "grad_norm": 0.3503289333931796,
+      "learning_rate": 1.66580587560852e-05,
+      "loss": 0.3159,
       "step": 6955
     },
     {
-      "epoch": 1.0197802197802197,
-      "grad_norm": 0.37112401011560336,
-      "learning_rate": 1.669367074823786e-05,
-      "loss": 0.3224,
+      "epoch": 1.024584130722803,
+      "grad_norm": 0.36948330992076095,
+      "learning_rate": 1.6651666203120112e-05,
+      "loss": 0.3372,
       "step": 6960
     },
     {
-      "epoch": 1.0205128205128204,
-      "grad_norm": 0.3846686023445629,
-      "learning_rate": 1.66873352481012e-05,
-      "loss": 0.3255,
+      "epoch": 1.0253201825408509,
+      "grad_norm": 0.3954984454681354,
+      "learning_rate": 1.66452687712299e-05,
+      "loss": 0.327,
       "step": 6965
     },
     {
-      "epoch": 1.0212454212454212,
-      "grad_norm": 0.37024088061424254,
-      "learning_rate": 1.6680994888554437e-05,
-      "loss": 0.3265,
+      "epoch": 1.0260562343588988,
+      "grad_norm": 0.37768644504929283,
+      "learning_rate": 1.663886646510701e-05,
+      "loss": 0.3237,
       "step": 6970
     },
     {
-      "epoch": 1.021978021978022,
-      "grad_norm": 0.35699115192148506,
-      "learning_rate": 1.6674649674204858e-05,
-      "loss": 0.3403,
+      "epoch": 1.026792286176947,
+      "grad_norm": 0.366095485562135,
+      "learning_rate": 1.663245928944747e-05,
+      "loss": 0.3032,
       "step": 6975
     },
     {
-      "epoch": 1.0227106227106226,
-      "grad_norm": 0.3658745623537328,
-      "learning_rate": 1.6668299609663265e-05,
-      "loss": 0.317,
+      "epoch": 1.0275283379949949,
+      "grad_norm": 0.363572513558161,
+      "learning_rate": 1.6626047248950866e-05,
+      "loss": 0.3252,
       "step": 6980
     },
     {
-      "epoch": 1.0234432234432234,
-      "grad_norm": 0.37798107793091024,
-      "learning_rate": 1.666194469954398e-05,
-      "loss": 0.3222,
+      "epoch": 1.0282643898130428,
+      "grad_norm": 0.3844984292096295,
+      "learning_rate": 1.6619630348320376e-05,
+      "loss": 0.3348,
       "step": 6985
     },
     {
-      "epoch": 1.024175824175824,
-      "grad_norm": 0.36710683342586753,
-      "learning_rate": 1.6655584948464868e-05,
-      "loss": 0.3384,
+      "epoch": 1.0290004416310907,
+      "grad_norm": 0.3632802823565491,
+      "learning_rate": 1.661320859226271e-05,
+      "loss": 0.3207,
       "step": 6990
     },
     {
-      "epoch": 1.0249084249084248,
-      "grad_norm": 0.35108828497233013,
-      "learning_rate": 1.6649220361047283e-05,
-      "loss": 0.3262,
+      "epoch": 1.0297364934491389,
+      "grad_norm": 0.3704916233754107,
+      "learning_rate": 1.6606781985488163e-05,
+      "loss": 0.3228,
       "step": 6995
     },
     {
-      "epoch": 1.0256410256410255,
-      "grad_norm": 0.38319064202144587,
-      "learning_rate": 1.6642850941916117e-05,
-      "loss": 0.3264,
+      "epoch": 1.0304725452671868,
+      "grad_norm": 0.3540067165419581,
+      "learning_rate": 1.6600350532710585e-05,
+      "loss": 0.3215,
       "step": 7000
     },
     {
-      "epoch": 1.0263736263736263,
-      "grad_norm": 0.36976692043937126,
-      "learning_rate": 1.6636476695699757e-05,
-      "loss": 0.3469,
+      "epoch": 1.0312085970852347,
+      "grad_norm": 0.3809778134546721,
+      "learning_rate": 1.659391423864737e-05,
+      "loss": 0.3188,
       "step": 7005
     },
     {
-      "epoch": 1.027106227106227,
-      "grad_norm": 0.3863933852480613,
-      "learning_rate": 1.663009762703011e-05,
-      "loss": 0.3195,
+      "epoch": 1.0319446489032829,
+      "grad_norm": 0.3583524428586189,
+      "learning_rate": 1.658747310801948e-05,
+      "loss": 0.3106,
       "step": 7010
     },
     {
-      "epoch": 1.0278388278388277,
-      "grad_norm": 0.381641946588195,
-      "learning_rate": 1.6623713740542576e-05,
-      "loss": 0.3285,
+      "epoch": 1.0326807007213308,
+      "grad_norm": 0.37977996018229804,
+      "learning_rate": 1.65810271455514e-05,
+      "loss": 0.3244,
       "step": 7015
     },
     {
-      "epoch": 1.0285714285714285,
-      "grad_norm": 0.39398956094879073,
-      "learning_rate": 1.661732504087606e-05,
-      "loss": 0.3258,
+      "epoch": 1.0334167525393787,
+      "grad_norm": 0.37847548505124745,
+      "learning_rate": 1.657457635597119e-05,
+      "loss": 0.3304,
       "step": 7020
     },
     {
-      "epoch": 1.0293040293040292,
-      "grad_norm": 0.39078417256772674,
-      "learning_rate": 1.6610931532672978e-05,
-      "loss": 0.3478,
+      "epoch": 1.0341528043574268,
+      "grad_norm": 0.37455421483361456,
+      "learning_rate": 1.6568120744010423e-05,
+      "loss": 0.3354,
       "step": 7025
     },
     {
-      "epoch": 1.03003663003663,
-      "grad_norm": 0.3625107322517052,
-      "learning_rate": 1.660453322057922e-05,
-      "loss": 0.317,
+      "epoch": 1.0348888561754748,
+      "grad_norm": 0.3739794280581571,
+      "learning_rate": 1.6561660314404222e-05,
+      "loss": 0.3361,
       "step": 7030
     },
     {
-      "epoch": 1.0307692307692307,
-      "grad_norm": 0.36369646810472994,
-      "learning_rate": 1.6598130109244172e-05,
-      "loss": 0.3274,
+      "epoch": 1.0356249079935227,
+      "grad_norm": 0.3540033001276769,
+      "learning_rate": 1.6555195071891247e-05,
+      "loss": 0.3195,
       "step": 7035
     },
     {
-      "epoch": 1.0315018315018314,
-      "grad_norm": 0.37324379752665476,
-      "learning_rate": 1.6591722203320714e-05,
-      "loss": 0.3222,
+      "epoch": 1.0363609598115708,
+      "grad_norm": 0.3815730252997327,
+      "learning_rate": 1.6548725021213677e-05,
+      "loss": 0.3355,
       "step": 7040
     },
     {
-      "epoch": 1.0322344322344323,
-      "grad_norm": 0.36774576270196524,
-      "learning_rate": 1.6585309507465208e-05,
-      "loss": 0.3396,
+      "epoch": 1.0370970116296188,
+      "grad_norm": 0.36803983074933216,
+      "learning_rate": 1.654225016711723e-05,
+      "loss": 0.3457,
       "step": 7045
     },
     {
-      "epoch": 1.032967032967033,
-      "grad_norm": 0.37855913620930015,
-      "learning_rate": 1.657889202633749e-05,
-      "loss": 0.3272,
+      "epoch": 1.0378330634476667,
+      "grad_norm": 0.37669783307078764,
+      "learning_rate": 1.6535770514351138e-05,
+      "loss": 0.332,
       "step": 7050
     },
     {
-      "epoch": 1.0336996336996338,
-      "grad_norm": 0.3949421463343061,
-      "learning_rate": 1.6572469764600883e-05,
-      "loss": 0.3334,
+      "epoch": 1.0385691152657146,
+      "grad_norm": 0.3572094760337152,
+      "learning_rate": 1.6529286067668157e-05,
+      "loss": 0.3194,
       "step": 7055
     },
     {
-      "epoch": 1.0344322344322345,
-      "grad_norm": 0.3815264146033247,
-      "learning_rate": 1.6566042726922176e-05,
-      "loss": 0.3176,
+      "epoch": 1.0393051670837627,
+      "grad_norm": 0.366098538014259,
+      "learning_rate": 1.6522796831824558e-05,
+      "loss": 0.3228,
       "step": 7060
     },
     {
-      "epoch": 1.0351648351648353,
-      "grad_norm": 0.37446850601118586,
-      "learning_rate": 1.6559610917971635e-05,
-      "loss": 0.3204,
+      "epoch": 1.0400412189018107,
+      "grad_norm": 0.38345492926038116,
+      "learning_rate": 1.6516302811580128e-05,
+      "loss": 0.3327,
       "step": 7065
     },
     {
-      "epoch": 1.035897435897436,
-      "grad_norm": 0.3564938395181809,
-      "learning_rate": 1.6553174342422984e-05,
-      "loss": 0.3261,
+      "epoch": 1.0407772707198586,
+      "grad_norm": 0.3623147985855647,
+      "learning_rate": 1.6509804011698155e-05,
+      "loss": 0.3186,
       "step": 7070
     },
     {
-      "epoch": 1.0366300366300367,
-      "grad_norm": 0.38035681846336156,
-      "learning_rate": 1.6546733004953424e-05,
-      "loss": 0.3347,
+      "epoch": 1.0415133225379067,
+      "grad_norm": 0.3731387729661672,
+      "learning_rate": 1.6503300436945444e-05,
+      "loss": 0.3291,
       "step": 7075
     },
     {
-      "epoch": 1.0373626373626375,
-      "grad_norm": 0.3708597688900623,
-      "learning_rate": 1.6540286910243604e-05,
-      "loss": 0.3287,
+      "epoch": 1.0422493743559547,
+      "grad_norm": 0.3664054074993125,
+      "learning_rate": 1.649679209209229e-05,
+      "loss": 0.33,
       "step": 7080
     },
     {
-      "epoch": 1.0380952380952382,
-      "grad_norm": 0.36445235144132193,
-      "learning_rate": 1.6533836062977634e-05,
-      "loss": 0.3145,
+      "epoch": 1.0429854261740026,
+      "grad_norm": 0.3679294654290743,
+      "learning_rate": 1.64902789819125e-05,
+      "loss": 0.3302,
       "step": 7085
     },
     {
-      "epoch": 1.038827838827839,
-      "grad_norm": 0.3619995557101858,
-      "learning_rate": 1.6527380467843076e-05,
-      "loss": 0.3327,
+      "epoch": 1.0437214779920507,
+      "grad_norm": 0.36296276512830733,
+      "learning_rate": 1.6483761111183367e-05,
+      "loss": 0.325,
       "step": 7090
     },
     {
-      "epoch": 1.0395604395604396,
-      "grad_norm": 0.36870240909186675,
-      "learning_rate": 1.6520920129530953e-05,
-      "loss": 0.337,
+      "epoch": 1.0444575298100987,
+      "grad_norm": 0.3660734529889537,
+      "learning_rate": 1.647723848468568e-05,
+      "loss": 0.3199,
       "step": 7095
     },
     {
-      "epoch": 1.0402930402930404,
-      "grad_norm": 0.35909337971119526,
-      "learning_rate": 1.651445505273572e-05,
-      "loss": 0.3432,
+      "epoch": 1.0451935816281466,
+      "grad_norm": 0.3485985572971079,
+      "learning_rate": 1.6470711107203713e-05,
+      "loss": 0.3226,
       "step": 7100
     },
     {
-      "epoch": 1.041025641025641,
-      "grad_norm": 0.38235517642065997,
-      "learning_rate": 1.650798524215528e-05,
-      "loss": 0.3372,
+      "epoch": 1.0459296334461947,
+      "grad_norm": 0.37937783027479494,
+      "learning_rate": 1.6464178983525226e-05,
+      "loss": 0.339,
       "step": 7105
     },
     {
-      "epoch": 1.0417582417582418,
-      "grad_norm": 0.39971357321203715,
-      "learning_rate": 1.6501510702490975e-05,
-      "loss": 0.3263,
+      "epoch": 1.0466656852642426,
+      "grad_norm": 0.36644636641304684,
+      "learning_rate": 1.6457642118441462e-05,
+      "loss": 0.346,
       "step": 7110
     },
     {
-      "epoch": 1.0424908424908426,
-      "grad_norm": 0.38484932847901626,
-      "learning_rate": 1.649503143844759e-05,
-      "loss": 0.3226,
+      "epoch": 1.0474017370822906,
+      "grad_norm": 0.36447140404101613,
+      "learning_rate": 1.6451100516747143e-05,
+      "loss": 0.3236,
       "step": 7115
     },
     {
-      "epoch": 1.0432234432234433,
-      "grad_norm": 0.3875184534367614,
-      "learning_rate": 1.648854745473334e-05,
-      "loss": 0.3359,
+      "epoch": 1.0481377889003385,
+      "grad_norm": 0.34816755648439046,
+      "learning_rate": 1.644455418324046e-05,
+      "loss": 0.3168,
       "step": 7120
     },
     {
-      "epoch": 1.043956043956044,
-      "grad_norm": 0.38058217345183687,
-      "learning_rate": 1.6482058756059865e-05,
-      "loss": 0.3226,
+      "epoch": 1.0488738407183866,
+      "grad_norm": 0.37863055036846316,
+      "learning_rate": 1.643800312272308e-05,
+      "loss": 0.3404,
       "step": 7125
     },
     {
-      "epoch": 1.0446886446886448,
-      "grad_norm": 0.37616687124736103,
-      "learning_rate": 1.647556534714223e-05,
-      "loss": 0.3398,
+      "epoch": 1.0496098925364346,
+      "grad_norm": 0.37362205464081283,
+      "learning_rate": 1.643144734000013e-05,
+      "loss": 0.3126,
       "step": 7130
     },
     {
-      "epoch": 1.0454212454212455,
-      "grad_norm": 0.37108574407021766,
-      "learning_rate": 1.6469067232698935e-05,
-      "loss": 0.3367,
+      "epoch": 1.0503459443544825,
+      "grad_norm": 0.3546035403542768,
+      "learning_rate": 1.642488683988022e-05,
+      "loss": 0.329,
       "step": 7135
     },
     {
-      "epoch": 1.0461538461538462,
-      "grad_norm": 0.34837614061109273,
-      "learning_rate": 1.6462564417451882e-05,
-      "loss": 0.3222,
+      "epoch": 1.0510819961725306,
+      "grad_norm": 0.36941811259031215,
+      "learning_rate": 1.6418321627175387e-05,
+      "loss": 0.3359,
       "step": 7140
     },
     {
-      "epoch": 1.046886446886447,
-      "grad_norm": 0.3690043003938606,
-      "learning_rate": 1.6456056906126407e-05,
-      "loss": 0.3331,
+      "epoch": 1.0518180479905785,
+      "grad_norm": 0.3532587579319103,
+      "learning_rate": 1.641175170670115e-05,
+      "loss": 0.3203,
       "step": 7145
     },
     {
-      "epoch": 1.0476190476190477,
-      "grad_norm": 0.3717115798790444,
-      "learning_rate": 1.6449544703451246e-05,
-      "loss": 0.3234,
+      "epoch": 1.0525540998086265,
+      "grad_norm": 0.3822778060319189,
+      "learning_rate": 1.640517708327648e-05,
+      "loss": 0.3263,
       "step": 7150
     },
     {
-      "epoch": 1.0483516483516484,
-      "grad_norm": 0.3634923181962294,
-      "learning_rate": 1.6443027814158544e-05,
-      "loss": 0.3307,
+      "epoch": 1.0532901516266746,
+      "grad_norm": 0.3678203451952706,
+      "learning_rate": 1.6398597761723788e-05,
+      "loss": 0.3219,
       "step": 7155
     },
     {
-      "epoch": 1.0490842490842491,
-      "grad_norm": 0.34368299374778816,
-      "learning_rate": 1.643650624298386e-05,
-      "loss": 0.3214,
+      "epoch": 1.0540262034447225,
+      "grad_norm": 0.37407041493293564,
+      "learning_rate": 1.6392013746868933e-05,
+      "loss": 0.3239,
       "step": 7160
     },
     {
-      "epoch": 1.0498168498168499,
-      "grad_norm": 0.35863896650208915,
-      "learning_rate": 1.6429979994666148e-05,
-      "loss": 0.3292,
+      "epoch": 1.0547622552627705,
+      "grad_norm": 0.3559475388496358,
+      "learning_rate": 1.6385425043541223e-05,
+      "loss": 0.3397,
       "step": 7165
     },
     {
-      "epoch": 1.0505494505494506,
-      "grad_norm": 0.3671958170838681,
-      "learning_rate": 1.6423449073947765e-05,
-      "loss": 0.3101,
+      "epoch": 1.0554983070808186,
+      "grad_norm": 0.3681921421056088,
+      "learning_rate": 1.63788316565734e-05,
+      "loss": 0.3392,
       "step": 7170
     },
     {
-      "epoch": 1.0512820512820513,
-      "grad_norm": 0.3866998396533175,
-      "learning_rate": 1.6416913485574455e-05,
-      "loss": 0.3184,
+      "epoch": 1.0562343588988665,
+      "grad_norm": 0.3710832881752622,
+      "learning_rate": 1.6372233590801638e-05,
+      "loss": 0.3381,
       "step": 7175
     },
     {
-      "epoch": 1.052014652014652,
-      "grad_norm": 0.35278906085633155,
-      "learning_rate": 1.6410373234295365e-05,
-      "loss": 0.334,
+      "epoch": 1.0569704107169144,
+      "grad_norm": 0.3581905728893183,
+      "learning_rate": 1.636563085106555e-05,
+      "loss": 0.3349,
       "step": 7180
     },
     {
-      "epoch": 1.0527472527472528,
-      "grad_norm": 0.3608851491793124,
-      "learning_rate": 1.6403828324863028e-05,
-      "loss": 0.3328,
+      "epoch": 1.0577064625349624,
+      "grad_norm": 0.36005171786885787,
+      "learning_rate": 1.635902344220817e-05,
+      "loss": 0.3152,
       "step": 7185
     },
     {
-      "epoch": 1.0534798534798535,
-      "grad_norm": 0.3889648649155586,
-      "learning_rate": 1.6397278762033353e-05,
-      "loss": 0.3297,
+      "epoch": 1.0584425143530105,
+      "grad_norm": 0.36169934059746134,
+      "learning_rate": 1.6352411369075968e-05,
+      "loss": 0.3272,
       "step": 7190
     },
     {
-      "epoch": 1.0542124542124542,
-      "grad_norm": 0.3631199609889286,
-      "learning_rate": 1.639072455056564e-05,
-      "loss": 0.3256,
+      "epoch": 1.0591785661710584,
+      "grad_norm": 0.36241019282764586,
+      "learning_rate": 1.634579463651882e-05,
+      "loss": 0.3221,
       "step": 7195
     },
     {
-      "epoch": 1.054945054945055,
-      "grad_norm": 0.35555239975058406,
-      "learning_rate": 1.638416569522256e-05,
-      "loss": 0.3192,
+      "epoch": 1.0599146179891064,
+      "grad_norm": 0.3663937100330741,
+      "learning_rate": 1.6339173249390038e-05,
+      "loss": 0.3244,
       "step": 7200
     },
     {
-      "epoch": 1.0556776556776557,
-      "grad_norm": 0.38583850949236403,
-      "learning_rate": 1.637760220077017e-05,
-      "loss": 0.3284,
+      "epoch": 1.0606506698071545,
+      "grad_norm": 0.3684450024490567,
+      "learning_rate": 1.6332547212546328e-05,
+      "loss": 0.3291,
       "step": 7205
     },
     {
-      "epoch": 1.0564102564102564,
-      "grad_norm": 0.37213001830232645,
-      "learning_rate": 1.637103407197788e-05,
-      "loss": 0.3332,
+      "epoch": 1.0613867216252024,
+      "grad_norm": 0.3850360448314454,
+      "learning_rate": 1.632591653084782e-05,
+      "loss": 0.3135,
       "step": 7210
     },
     {
-      "epoch": 1.0571428571428572,
-      "grad_norm": 0.38617384644077335,
-      "learning_rate": 1.6364461313618484e-05,
-      "loss": 0.3161,
+      "epoch": 1.0621227734432503,
+      "grad_norm": 0.35248505793040924,
+      "learning_rate": 1.6319281209158045e-05,
+      "loss": 0.3238,
       "step": 7215
     },
     {
-      "epoch": 1.057875457875458,
-      "grad_norm": 0.37460027814008917,
-      "learning_rate": 1.6357883930468134e-05,
-      "loss": 0.3311,
+      "epoch": 1.0628588252612985,
+      "grad_norm": 0.38223933242711317,
+      "learning_rate": 1.6312641252343945e-05,
+      "loss": 0.3424,
       "step": 7220
     },
     {
-      "epoch": 1.0586080586080586,
-      "grad_norm": 0.37347735128227266,
-      "learning_rate": 1.635130192730634e-05,
-      "loss": 0.3198,
+      "epoch": 1.0635948770793464,
+      "grad_norm": 0.3658573144825204,
+      "learning_rate": 1.6305996665275854e-05,
+      "loss": 0.3288,
       "step": 7225
     },
     {
-      "epoch": 1.0593406593406594,
-      "grad_norm": 0.36952088656401244,
-      "learning_rate": 1.6344715308915976e-05,
-      "loss": 0.3307,
+      "epoch": 1.0643309288973943,
+      "grad_norm": 0.37273113745613967,
+      "learning_rate": 1.62993474528275e-05,
+      "loss": 0.3492,
       "step": 7230
     },
     {
-      "epoch": 1.06007326007326,
-      "grad_norm": 0.3555251241647386,
-      "learning_rate": 1.6338124080083256e-05,
-      "loss": 0.3226,
+      "epoch": 1.0650669807154425,
+      "grad_norm": 0.3602935228828251,
+      "learning_rate": 1.6292693619876016e-05,
+      "loss": 0.3241,
       "step": 7235
     },
     {
-      "epoch": 1.0608058608058608,
-      "grad_norm": 0.3916542016656616,
-      "learning_rate": 1.633152824559776e-05,
-      "loss": 0.3293,
+      "epoch": 1.0658030325334904,
+      "grad_norm": 0.3857924962744493,
+      "learning_rate": 1.628603517130191e-05,
+      "loss": 0.3372,
       "step": 7240
     },
     {
-      "epoch": 1.0615384615384615,
-      "grad_norm": 0.3698139759757373,
-      "learning_rate": 1.6324927810252412e-05,
-      "loss": 0.3266,
+      "epoch": 1.0665390843515383,
+      "grad_norm": 0.3573069234170001,
+      "learning_rate": 1.6279372111989084e-05,
+      "loss": 0.3128,
       "step": 7245
     },
     {
-      "epoch": 1.0622710622710623,
-      "grad_norm": 0.37141954324269505,
-      "learning_rate": 1.6318322778843467e-05,
-      "loss": 0.3335,
+      "epoch": 1.0672751361695862,
+      "grad_norm": 0.3751310313442755,
+      "learning_rate": 1.627270444682482e-05,
+      "loss": 0.3361,
       "step": 7250
     },
     {
-      "epoch": 1.063003663003663,
-      "grad_norm": 0.37461222790810195,
-      "learning_rate": 1.6311713156170535e-05,
-      "loss": 0.3145,
+      "epoch": 1.0680111879876344,
+      "grad_norm": 0.37863537501379835,
+      "learning_rate": 1.626603218069978e-05,
+      "loss": 0.3183,
       "step": 7255
     },
     {
-      "epoch": 1.0637362637362637,
-      "grad_norm": 0.3922862206834006,
-      "learning_rate": 1.6305098947036552e-05,
-      "loss": 0.3332,
+      "epoch": 1.0687472398056823,
+      "grad_norm": 0.3670003362762075,
+      "learning_rate": 1.6259355318507996e-05,
+      "loss": 0.318,
       "step": 7260
     },
     {
-      "epoch": 1.0644688644688645,
-      "grad_norm": 0.36960020871177257,
-      "learning_rate": 1.629848015624779e-05,
-      "loss": 0.3369,
+      "epoch": 1.0694832916237302,
+      "grad_norm": 0.3750423003028345,
+      "learning_rate": 1.6252673865146876e-05,
+      "loss": 0.3297,
       "step": 7265
     },
     {
-      "epoch": 1.0652014652014652,
-      "grad_norm": 0.37500279281500304,
-      "learning_rate": 1.6291856788613853e-05,
-      "loss": 0.3425,
+      "epoch": 1.0702193434417784,
+      "grad_norm": 0.37475485693836696,
+      "learning_rate": 1.6245987825517193e-05,
+      "loss": 0.3257,
       "step": 7270
     },
     {
-      "epoch": 1.065934065934066,
-      "grad_norm": 0.36412353671363584,
-      "learning_rate": 1.628522884894767e-05,
-      "loss": 0.3185,
+      "epoch": 1.0709553952598263,
+      "grad_norm": 0.3704294227241169,
+      "learning_rate": 1.6239297204523082e-05,
+      "loss": 0.3121,
       "step": 7275
     },
     {
-      "epoch": 1.0666666666666667,
-      "grad_norm": 0.38656051240449313,
-      "learning_rate": 1.6278596342065483e-05,
-      "loss": 0.3304,
+      "epoch": 1.0716914470778742,
+      "grad_norm": 0.3776907071057796,
+      "learning_rate": 1.623260200707205e-05,
+      "loss": 0.3252,
       "step": 7280
     },
     {
-      "epoch": 1.0673992673992674,
-      "grad_norm": 0.37517022440229425,
-      "learning_rate": 1.627195927278687e-05,
-      "loss": 0.3421,
+      "epoch": 1.0724274988959224,
+      "grad_norm": 0.37423229911933753,
+      "learning_rate": 1.622590223807494e-05,
+      "loss": 0.3276,
       "step": 7285
     },
     {
-      "epoch": 1.0681318681318681,
-      "grad_norm": 0.3536741231559894,
-      "learning_rate": 1.626531764593471e-05,
-      "loss": 0.333,
+      "epoch": 1.0731635507139703,
+      "grad_norm": 0.37166084869277755,
+      "learning_rate": 1.6219197902445965e-05,
+      "loss": 0.3284,
       "step": 7290
     },
     {
-      "epoch": 1.0688644688644688,
-      "grad_norm": 0.36764122405217287,
-      "learning_rate": 1.62586714663352e-05,
-      "loss": 0.3177,
+      "epoch": 1.0738996025320182,
+      "grad_norm": 0.3642307755417687,
+      "learning_rate": 1.6212489005102687e-05,
+      "loss": 0.3216,
       "step": 7295
     },
     {
-      "epoch": 1.0695970695970696,
-      "grad_norm": 0.3714716959836512,
-      "learning_rate": 1.6252020738817845e-05,
-      "loss": 0.3369,
+      "epoch": 1.0746356543500664,
+      "grad_norm": 0.37194292127247996,
+      "learning_rate": 1.6205775550966006e-05,
+      "loss": 0.3218,
       "step": 7300
     },
     {
-      "epoch": 1.0703296703296703,
-      "grad_norm": 0.38888598011485115,
-      "learning_rate": 1.624536546821545e-05,
-      "loss": 0.3207,
+      "epoch": 1.0753717061681143,
+      "grad_norm": 0.366829209386239,
+      "learning_rate": 1.6199057544960166e-05,
+      "loss": 0.3303,
       "step": 7305
     },
     {
-      "epoch": 1.071062271062271,
-      "grad_norm": 0.3785750277463407,
-      "learning_rate": 1.6238705659364125e-05,
-      "loss": 0.32,
+      "epoch": 1.0761077579861622,
+      "grad_norm": 0.35373260814593444,
+      "learning_rate": 1.6192334992012754e-05,
+      "loss": 0.3352,
       "step": 7310
     },
     {
-      "epoch": 1.0717948717948718,
-      "grad_norm": 0.3875034283018185,
-      "learning_rate": 1.6232041317103283e-05,
-      "loss": 0.3267,
+      "epoch": 1.0768438098042101,
+      "grad_norm": 0.36101873658612077,
+      "learning_rate": 1.618560789705469e-05,
+      "loss": 0.3468,
       "step": 7315
     },
     {
-      "epoch": 1.0725274725274725,
-      "grad_norm": 0.38196225424286717,
-      "learning_rate": 1.6225372446275624e-05,
-      "loss": 0.3351,
+      "epoch": 1.0775798616222583,
+      "grad_norm": 0.3730530720901144,
+      "learning_rate": 1.617887626502022e-05,
+      "loss": 0.326,
       "step": 7320
     },
     {
-      "epoch": 1.0732600732600732,
-      "grad_norm": 0.3624491647078481,
-      "learning_rate": 1.621869905172714e-05,
-      "loss": 0.3335,
+      "epoch": 1.0783159134403062,
+      "grad_norm": 0.35642291924912095,
+      "learning_rate": 1.6172140100846933e-05,
+      "loss": 0.3263,
       "step": 7325
     },
     {
-      "epoch": 1.073992673992674,
-      "grad_norm": 0.5550075430236053,
-      "learning_rate": 1.6212021138307108e-05,
-      "loss": 0.3257,
+      "epoch": 1.0790519652583541,
+      "grad_norm": 0.36746353036862045,
+      "learning_rate": 1.6165399409475723e-05,
+      "loss": 0.33,
       "step": 7330
     },
     {
-      "epoch": 1.0747252747252747,
-      "grad_norm": 0.37399371992603847,
-      "learning_rate": 1.6205338710868094e-05,
-      "loss": 0.3381,
+      "epoch": 1.0797880170764023,
+      "grad_norm": 0.3724282305050785,
+      "learning_rate": 1.615865419585082e-05,
+      "loss": 0.3176,
       "step": 7335
     },
     {
-      "epoch": 1.0754578754578754,
-      "grad_norm": 0.3548002458460286,
-      "learning_rate": 1.6198651774265947e-05,
-      "loss": 0.3186,
+      "epoch": 1.0805240688944502,
+      "grad_norm": 0.3696224933488861,
+      "learning_rate": 1.6151904464919764e-05,
+      "loss": 0.3151,
       "step": 7340
     },
     {
-      "epoch": 1.0761904761904761,
-      "grad_norm": 0.37323170647942955,
-      "learning_rate": 1.619196033335978e-05,
-      "loss": 0.3307,
+      "epoch": 1.081260120712498,
+      "grad_norm": 0.37514572425232984,
+      "learning_rate": 1.614515022163341e-05,
+      "loss": 0.3218,
       "step": 7345
     },
     {
-      "epoch": 1.0769230769230769,
-      "grad_norm": 0.36708288742710005,
-      "learning_rate": 1.618526439301199e-05,
-      "loss": 0.3248,
+      "epoch": 1.0819961725305463,
+      "grad_norm": 0.36528477402573506,
+      "learning_rate": 1.6138391470945917e-05,
+      "loss": 0.3247,
       "step": 7350
     },
     {
-      "epoch": 1.0776556776556776,
-      "grad_norm": 0.36098653159048294,
-      "learning_rate": 1.6178563958088237e-05,
-      "loss": 0.3236,
+      "epoch": 1.0827322243485942,
+      "grad_norm": 0.37119579299867844,
+      "learning_rate": 1.613162821781476e-05,
+      "loss": 0.319,
       "step": 7355
     },
     {
-      "epoch": 1.0783882783882783,
-      "grad_norm": 0.36894116539216226,
-      "learning_rate": 1.6171859033457457e-05,
-      "loss": 0.3333,
+      "epoch": 1.083468276166642,
+      "grad_norm": 0.3572730238206756,
+      "learning_rate": 1.6124860467200717e-05,
+      "loss": 0.3169,
       "step": 7360
     },
     {
-      "epoch": 1.079120879120879,
-      "grad_norm": 0.38834742229919167,
-      "learning_rate": 1.6165149623991835e-05,
-      "loss": 0.3166,
+      "epoch": 1.08420432798469,
+      "grad_norm": 0.3870278421078157,
+      "learning_rate": 1.611808822406785e-05,
+      "loss": 0.3217,
       "step": 7365
     },
     {
-      "epoch": 1.0798534798534798,
-      "grad_norm": 0.3587364240448985,
-      "learning_rate": 1.6158435734566823e-05,
-      "loss": 0.3339,
+      "epoch": 1.0849403798027382,
+      "grad_norm": 0.37792978078405554,
+      "learning_rate": 1.611131149338353e-05,
+      "loss": 0.3376,
       "step": 7370
     },
     {
-      "epoch": 1.0805860805860805,
-      "grad_norm": 0.3635888549564469,
-      "learning_rate": 1.6151717370061125e-05,
-      "loss": 0.3229,
+      "epoch": 1.085676431620786,
+      "grad_norm": 0.37105444025462786,
+      "learning_rate": 1.6104530280118417e-05,
+      "loss": 0.337,
       "step": 7375
     },
     {
-      "epoch": 1.0813186813186813,
-      "grad_norm": 0.3731882016211775,
-      "learning_rate": 1.6144994535356704e-05,
-      "loss": 0.3134,
+      "epoch": 1.086412483438834,
+      "grad_norm": 0.38318384864267696,
+      "learning_rate": 1.6097744589246453e-05,
+      "loss": 0.3382,
       "step": 7380
     },
     {
-      "epoch": 1.082051282051282,
-      "grad_norm": 0.35711735047435655,
-      "learning_rate": 1.6138267235338764e-05,
-      "loss": 0.3269,
+      "epoch": 1.0871485352568822,
+      "grad_norm": 0.3702880421771977,
+      "learning_rate": 1.6090954425744875e-05,
+      "loss": 0.3282,
       "step": 7385
     },
     {
-      "epoch": 1.0827838827838827,
-      "grad_norm": 0.383226470412408,
-      "learning_rate": 1.6131535474895753e-05,
-      "loss": 0.3268,
+      "epoch": 1.08788458707493,
+      "grad_norm": 0.36901695935820905,
+      "learning_rate": 1.6084159794594193e-05,
+      "loss": 0.3264,
       "step": 7390
     },
     {
-      "epoch": 1.0835164835164834,
-      "grad_norm": 0.37425974187830674,
-      "learning_rate": 1.6124799258919367e-05,
-      "loss": 0.3405,
+      "epoch": 1.088620638892978,
+      "grad_norm": 0.3558368550213881,
+      "learning_rate": 1.6077360700778194e-05,
+      "loss": 0.3275,
       "step": 7395
     },
     {
-      "epoch": 1.0842490842490842,
-      "grad_norm": 0.3647449028203382,
-      "learning_rate": 1.611805859230453e-05,
-      "loss": 0.3152,
+      "epoch": 1.0893566907110261,
+      "grad_norm": 0.3604908209621351,
+      "learning_rate": 1.6070557149283933e-05,
+      "loss": 0.3297,
       "step": 7400
     },
     {
-      "epoch": 1.084981684981685,
-      "grad_norm": 0.38305661847604566,
-      "learning_rate": 1.6111313479949415e-05,
-      "loss": 0.3234,
+      "epoch": 1.090092742529074,
+      "grad_norm": 0.37617018641738653,
+      "learning_rate": 1.6063749145101756e-05,
+      "loss": 0.3315,
       "step": 7405
     },
     {
-      "epoch": 1.0857142857142856,
-      "grad_norm": 0.3850155020309742,
-      "learning_rate": 1.6104563926755407e-05,
-      "loss": 0.3167,
+      "epoch": 1.090828794347122,
+      "grad_norm": 0.381591827635015,
+      "learning_rate": 1.6056936693225246e-05,
+      "loss": 0.3243,
       "step": 7410
     },
     {
-      "epoch": 1.0864468864468864,
-      "grad_norm": 0.3896911044662533,
-      "learning_rate": 1.6097809937627134e-05,
-      "loss": 0.3416,
+      "epoch": 1.09156484616517,
+      "grad_norm": 0.36013113925771034,
+      "learning_rate": 1.6050119798651275e-05,
+      "loss": 0.3293,
       "step": 7415
     },
     {
-      "epoch": 1.087179487179487,
-      "grad_norm": 0.3808117650518625,
-      "learning_rate": 1.609105151747244e-05,
-      "loss": 0.3407,
+      "epoch": 1.092300897983218,
+      "grad_norm": 0.3493409370534532,
+      "learning_rate": 1.604329846637995e-05,
+      "loss": 0.3337,
       "step": 7420
     },
     {
-      "epoch": 1.0879120879120878,
-      "grad_norm": 0.39129833508135087,
-      "learning_rate": 1.6084288671202384e-05,
-      "loss": 0.3279,
+      "epoch": 1.093036949801266,
+      "grad_norm": 0.3956493470643981,
+      "learning_rate": 1.603647270141465e-05,
+      "loss": 0.3411,
       "step": 7425
     },
     {
-      "epoch": 1.0886446886446886,
-      "grad_norm": 0.3619807548374353,
-      "learning_rate": 1.607752140373125e-05,
-      "loss": 0.3235,
+      "epoch": 1.093773001619314,
+      "grad_norm": 0.3656847275893169,
+      "learning_rate": 1.6029642508762e-05,
+      "loss": 0.3384,
       "step": 7430
     },
     {
-      "epoch": 1.0893772893772893,
-      "grad_norm": 0.3596232789121412,
-      "learning_rate": 1.6070749719976534e-05,
-      "loss": 0.3291,
+      "epoch": 1.094509053437362,
+      "grad_norm": 0.36577327273484317,
+      "learning_rate": 1.602280789343187e-05,
+      "loss": 0.3582,
       "step": 7435
     },
     {
-      "epoch": 1.09010989010989,
-      "grad_norm": 0.37294768245070653,
-      "learning_rate": 1.6063973624858936e-05,
-      "loss": 0.3357,
+      "epoch": 1.09524510525541,
+      "grad_norm": 0.353669292274733,
+      "learning_rate": 1.6015968860437382e-05,
+      "loss": 0.3335,
       "step": 7440
     },
     {
-      "epoch": 1.0908424908424907,
-      "grad_norm": 0.36847948298689304,
-      "learning_rate": 1.6057193123302363e-05,
-      "loss": 0.3393,
+      "epoch": 1.095981157073458,
+      "grad_norm": 0.3702265509160007,
+      "learning_rate": 1.600912541479489e-05,
+      "loss": 0.3298,
       "step": 7445
     },
     {
-      "epoch": 1.0915750915750915,
-      "grad_norm": 0.3692013692711026,
-      "learning_rate": 1.605040822023393e-05,
-      "loss": 0.3222,
+      "epoch": 1.096717208891506,
+      "grad_norm": 0.3573253785340228,
+      "learning_rate": 1.600227756152398e-05,
+      "loss": 0.3147,
       "step": 7450
     },
     {
-      "epoch": 1.0923076923076924,
-      "grad_norm": 0.365490580175772,
-      "learning_rate": 1.6043618920583945e-05,
-      "loss": 0.3219,
+      "epoch": 1.097453260709554,
+      "grad_norm": 0.3638891137918482,
+      "learning_rate": 1.5995425305647486e-05,
+      "loss": 0.3207,
       "step": 7455
     },
     {
-      "epoch": 1.0930402930402932,
-      "grad_norm": 0.37989741813217187,
-      "learning_rate": 1.60368252292859e-05,
-      "loss": 0.3158,
+      "epoch": 1.0981893125276019,
+      "grad_norm": 0.39218922642738785,
+      "learning_rate": 1.5988568652191463e-05,
+      "loss": 0.3235,
       "step": 7460
     },
     {
-      "epoch": 1.0937728937728939,
-      "grad_norm": 0.3857019049089393,
-      "learning_rate": 1.6030027151276504e-05,
-      "loss": 0.3491,
+      "epoch": 1.09892536434565,
+      "grad_norm": 0.3823963695714701,
+      "learning_rate": 1.598170760618519e-05,
+      "loss": 0.3283,
       "step": 7465
     },
     {
-      "epoch": 1.0945054945054946,
-      "grad_norm": 0.4380529333414578,
-      "learning_rate": 1.6023224691495636e-05,
-      "loss": 0.3324,
+      "epoch": 1.099661416163698,
+      "grad_norm": 0.3641134525190089,
+      "learning_rate": 1.5974842172661168e-05,
+      "loss": 0.3286,
       "step": 7470
     },
     {
-      "epoch": 1.0952380952380953,
-      "grad_norm": 0.3686451050298967,
-      "learning_rate": 1.601641785488636e-05,
-      "loss": 0.3419,
+      "epoch": 1.1003974679817459,
+      "grad_norm": 0.3701467819083142,
+      "learning_rate": 1.5967972356655125e-05,
+      "loss": 0.3274,
       "step": 7475
     },
     {
-      "epoch": 1.095970695970696,
-      "grad_norm": 0.3760684142547285,
-      "learning_rate": 1.600960664639492e-05,
-      "loss": 0.3384,
+      "epoch": 1.1011335197997938,
+      "grad_norm": 0.36222762668218655,
+      "learning_rate": 1.596109816320599e-05,
+      "loss": 0.3291,
       "step": 7480
     },
     {
-      "epoch": 1.0967032967032968,
-      "grad_norm": 0.3767875855720967,
-      "learning_rate": 1.6002791070970743e-05,
-      "loss": 0.3295,
+      "epoch": 1.101869571617842,
+      "grad_norm": 0.35521764335109396,
+      "learning_rate": 1.5954219597355907e-05,
+      "loss": 0.3258,
       "step": 7485
     },
     {
-      "epoch": 1.0974358974358975,
-      "grad_norm": 0.3845254569319136,
-      "learning_rate": 1.599597113356643e-05,
-      "loss": 0.336,
+      "epoch": 1.1026056234358899,
+      "grad_norm": 0.34725491118799207,
+      "learning_rate": 1.594733666415024e-05,
+      "loss": 0.3262,
       "step": 7490
     },
     {
-      "epoch": 1.0981684981684983,
-      "grad_norm": 0.3679990125041362,
-      "learning_rate": 1.5989146839137745e-05,
-      "loss": 0.3371,
+      "epoch": 1.1033416752539378,
+      "grad_norm": 0.3597769806477287,
+      "learning_rate": 1.594044936863754e-05,
+      "loss": 0.3198,
       "step": 7495
     },
     {
-      "epoch": 1.098901098901099,
-      "grad_norm": 0.36193018957769496,
-      "learning_rate": 1.598231819264361e-05,
-      "loss": 0.3274,
+      "epoch": 1.104077727071986,
+      "grad_norm": 0.3731792537772293,
+      "learning_rate": 1.5933557715869562e-05,
+      "loss": 0.3313,
       "step": 7500
     },
     {
-      "epoch": 1.0996336996336997,
-      "grad_norm": 0.36564077411332907,
-      "learning_rate": 1.5975485199046134e-05,
-      "loss": 0.3277,
+      "epoch": 1.1048137788900338,
+      "grad_norm": 0.37604908190706937,
+      "learning_rate": 1.5926661710901262e-05,
+      "loss": 0.3397,
       "step": 7505
     },
     {
-      "epoch": 1.1003663003663005,
-      "grad_norm": 0.3792791798482742,
-      "learning_rate": 1.596864786331057e-05,
-      "loss": 0.3357,
+      "epoch": 1.1055498307080818,
+      "grad_norm": 0.375418592273806,
+      "learning_rate": 1.5919761358790783e-05,
+      "loss": 0.318,
       "step": 7510
     },
     {
-      "epoch": 1.1010989010989012,
-      "grad_norm": 0.37858832146394067,
-      "learning_rate": 1.596180619040532e-05,
-      "loss": 0.3382,
+      "epoch": 1.10628588252613,
+      "grad_norm": 0.3731120352461751,
+      "learning_rate": 1.5912856664599456e-05,
+      "loss": 0.3307,
       "step": 7515
     },
     {
-      "epoch": 1.101831501831502,
-      "grad_norm": 0.38131450688598406,
-      "learning_rate": 1.5954960185301945e-05,
-      "loss": 0.3332,
+      "epoch": 1.1070219343441778,
+      "grad_norm": 0.3456206801012039,
+      "learning_rate": 1.590594763339181e-05,
+      "loss": 0.3273,
       "step": 7520
     },
     {
-      "epoch": 1.1025641025641026,
-      "grad_norm": 0.3795639108042152,
-      "learning_rate": 1.594810985297516e-05,
-      "loss": 0.3407,
+      "epoch": 1.1077579861622258,
+      "grad_norm": 0.38499107405021915,
+      "learning_rate": 1.5899034270235526e-05,
+      "loss": 0.338,
       "step": 7525
     },
     {
-      "epoch": 1.1032967032967034,
-      "grad_norm": 0.38294554175489987,
-      "learning_rate": 1.5941255198402817e-05,
-      "loss": 0.3409,
+      "epoch": 1.108494037980274,
+      "grad_norm": 0.3773814368821201,
+      "learning_rate": 1.5892116580201503e-05,
+      "loss": 0.336,
       "step": 7530
     },
     {
-      "epoch": 1.104029304029304,
-      "grad_norm": 0.36047873074214287,
-      "learning_rate": 1.5934396226565908e-05,
-      "loss": 0.3274,
+      "epoch": 1.1092300897983218,
+      "grad_norm": 0.3802879200646699,
+      "learning_rate": 1.588519456836377e-05,
+      "loss": 0.3422,
       "step": 7535
     },
     {
-      "epoch": 1.1047619047619048,
-      "grad_norm": 0.37183620442862136,
-      "learning_rate": 1.5927532942448563e-05,
-      "loss": 0.3346,
+      "epoch": 1.1099661416163698,
+      "grad_norm": 0.34333379544800796,
+      "learning_rate": 1.5878268239799567e-05,
+      "loss": 0.3341,
       "step": 7540
     },
     {
-      "epoch": 1.1054945054945056,
-      "grad_norm": 0.366742474177052,
-      "learning_rate": 1.592066535103805e-05,
-      "loss": 0.3349,
+      "epoch": 1.1107021934344177,
+      "grad_norm": 0.3826944545514652,
+      "learning_rate": 1.5871337599589273e-05,
+      "loss": 0.3371,
       "step": 7545
     },
     {
-      "epoch": 1.1062271062271063,
-      "grad_norm": 0.3825427448250812,
-      "learning_rate": 1.5913793457324765e-05,
-      "loss": 0.3382,
+      "epoch": 1.1114382452524658,
+      "grad_norm": 0.38749648865209124,
+      "learning_rate": 1.5864402652816433e-05,
+      "loss": 0.342,
       "step": 7550
     },
     {
-      "epoch": 1.106959706959707,
-      "grad_norm": 0.3630028811395133,
-      "learning_rate": 1.5906917266302224e-05,
-      "loss": 0.3194,
+      "epoch": 1.1121742970705137,
+      "grad_norm": 0.37327300355640614,
+      "learning_rate": 1.5857463404567758e-05,
+      "loss": 0.3208,
       "step": 7555
     },
     {
-      "epoch": 1.1076923076923078,
-      "grad_norm": 0.36450569088762086,
-      "learning_rate": 1.5900036782967074e-05,
-      "loss": 0.3388,
+      "epoch": 1.1129103488885617,
+      "grad_norm": 0.35851734984850764,
+      "learning_rate": 1.5850519859933115e-05,
+      "loss": 0.329,
       "step": 7560
     },
     {
-      "epoch": 1.1084249084249085,
-      "grad_norm": 0.36750525189012,
-      "learning_rate": 1.589315201231908e-05,
-      "loss": 0.3202,
+      "epoch": 1.1136464007066098,
+      "grad_norm": 0.370965625314894,
+      "learning_rate": 1.584357202400552e-05,
+      "loss": 0.3267,
       "step": 7565
     },
     {
-      "epoch": 1.1091575091575092,
-      "grad_norm": 0.36010730121886997,
-      "learning_rate": 1.5886262959361114e-05,
-      "loss": 0.3168,
+      "epoch": 1.1143824525246577,
+      "grad_norm": 0.3617068011157726,
+      "learning_rate": 1.5836619901881128e-05,
+      "loss": 0.3392,
       "step": 7570
     },
     {
-      "epoch": 1.10989010989011,
-      "grad_norm": 0.37408685972832806,
-      "learning_rate": 1.5879369629099167e-05,
-      "loss": 0.3288,
+      "epoch": 1.1151185043427057,
+      "grad_norm": 0.3536373085796306,
+      "learning_rate": 1.5829663498659247e-05,
+      "loss": 0.3321,
       "step": 7575
     },
     {
-      "epoch": 1.1106227106227107,
-      "grad_norm": 0.3458106728692704,
-      "learning_rate": 1.587247202654235e-05,
-      "loss": 0.3253,
+      "epoch": 1.1158545561607538,
+      "grad_norm": 0.35162096268803267,
+      "learning_rate": 1.582270281944233e-05,
+      "loss": 0.3308,
       "step": 7580
     },
     {
-      "epoch": 1.1113553113553114,
-      "grad_norm": 0.36639603817817384,
-      "learning_rate": 1.5865570156702848e-05,
-      "loss": 0.3323,
+      "epoch": 1.1165906079788017,
+      "grad_norm": 0.37021379314830466,
+      "learning_rate": 1.581573786933595e-05,
+      "loss": 0.3359,
       "step": 7585
     },
     {
-      "epoch": 1.1120879120879121,
-      "grad_norm": 0.3955687450980298,
-      "learning_rate": 1.5858664024595973e-05,
-      "loss": 0.3369,
+      "epoch": 1.1173266597968496,
+      "grad_norm": 0.3595300119663517,
+      "learning_rate": 1.5808768653448835e-05,
+      "loss": 0.339,
       "step": 7590
     },
     {
-      "epoch": 1.1128205128205129,
-      "grad_norm": 0.3611858582920874,
-      "learning_rate": 1.5851753635240122e-05,
-      "loss": 0.3192,
+      "epoch": 1.1180627116148978,
+      "grad_norm": 0.3571647240421222,
+      "learning_rate": 1.5801795176892817e-05,
+      "loss": 0.3296,
       "step": 7595
     },
     {
-      "epoch": 1.1135531135531136,
-      "grad_norm": 0.3765189228908552,
-      "learning_rate": 1.58448389936568e-05,
-      "loss": 0.3587,
+      "epoch": 1.1187987634329457,
+      "grad_norm": 0.3741042476940163,
+      "learning_rate": 1.5794817444782875e-05,
+      "loss": 0.3171,
       "step": 7600
     },
     {
-      "epoch": 1.1142857142857143,
-      "grad_norm": 0.3686849977742123,
-      "learning_rate": 1.5837920104870578e-05,
-      "loss": 0.3258,
+      "epoch": 1.1195348152509936,
+      "grad_norm": 0.3654862897816044,
+      "learning_rate": 1.578783546223709e-05,
+      "loss": 0.3317,
       "step": 7605
     },
     {
-      "epoch": 1.115018315018315,
-      "grad_norm": 0.3522453696751579,
-      "learning_rate": 1.583099697390913e-05,
-      "loss": 0.3221,
+      "epoch": 1.1202708670690416,
+      "grad_norm": 0.36602886343286756,
+      "learning_rate": 1.5780849234376683e-05,
+      "loss": 0.3293,
       "step": 7610
     },
     {
-      "epoch": 1.1157509157509158,
-      "grad_norm": 0.35684852493198504,
-      "learning_rate": 1.5824069605803214e-05,
-      "loss": 0.3275,
+      "epoch": 1.1210069188870897,
+      "grad_norm": 0.36170190489700843,
+      "learning_rate": 1.5773858766325968e-05,
+      "loss": 0.3318,
       "step": 7615
     },
     {
-      "epoch": 1.1164835164835165,
-      "grad_norm": 0.36542984023806907,
-      "learning_rate": 1.5817138005586655e-05,
-      "loss": 0.3294,
+      "epoch": 1.1217429707051376,
+      "grad_norm": 0.3660675851852243,
+      "learning_rate": 1.5766864063212376e-05,
+      "loss": 0.3309,
       "step": 7620
     },
     {
-      "epoch": 1.1172161172161172,
-      "grad_norm": 0.35939595810284075,
-      "learning_rate": 1.5810202178296368e-05,
-      "loss": 0.3209,
+      "epoch": 1.1224790225231855,
+      "grad_norm": 0.3702399812281134,
+      "learning_rate": 1.5759865130166456e-05,
+      "loss": 0.3277,
       "step": 7625
     },
     {
-      "epoch": 1.117948717948718,
-      "grad_norm": 0.37967256270515004,
-      "learning_rate": 1.5803262128972328e-05,
-      "loss": 0.3334,
+      "epoch": 1.1232150743412337,
+      "grad_norm": 0.3650387017992945,
+      "learning_rate": 1.575286197232184e-05,
+      "loss": 0.3318,
       "step": 7630
     },
     {
-      "epoch": 1.1186813186813187,
-      "grad_norm": 0.38638494038204896,
-      "learning_rate": 1.579631786265758e-05,
-      "loss": 0.3286,
+      "epoch": 1.1239511261592816,
+      "grad_norm": 0.3718314594470514,
+      "learning_rate": 1.5745854594815266e-05,
+      "loss": 0.3229,
       "step": 7635
     },
     {
-      "epoch": 1.1194139194139194,
-      "grad_norm": 0.3697407078462657,
-      "learning_rate": 1.5789369384398243e-05,
-      "loss": 0.3212,
+      "epoch": 1.1246871779773295,
+      "grad_norm": 0.39133302673669706,
+      "learning_rate": 1.5738843002786583e-05,
+      "loss": 0.3324,
       "step": 7640
     },
     {
-      "epoch": 1.1201465201465202,
-      "grad_norm": 0.38392638083705544,
-      "learning_rate": 1.578241669924348e-05,
-      "loss": 0.3331,
+      "epoch": 1.1254232297953777,
+      "grad_norm": 0.35355252114799085,
+      "learning_rate": 1.5731827201378703e-05,
+      "loss": 0.3319,
       "step": 7645
     },
     {
-      "epoch": 1.120879120879121,
-      "grad_norm": 0.3524484748280436,
-      "learning_rate": 1.577545981224552e-05,
-      "loss": 0.3266,
+      "epoch": 1.1261592816134256,
+      "grad_norm": 0.35336190013193636,
+      "learning_rate": 1.5724807195737654e-05,
+      "loss": 0.3182,
       "step": 7650
     },
     {
-      "epoch": 1.1216117216117216,
-      "grad_norm": 0.36401648760373023,
-      "learning_rate": 1.5768498728459652e-05,
-      "loss": 0.3368,
+      "epoch": 1.1268953334314735,
+      "grad_norm": 0.34575606792533403,
+      "learning_rate": 1.571778299101252e-05,
+      "loss": 0.3281,
       "step": 7655
     },
     {
-      "epoch": 1.1223443223443224,
-      "grad_norm": 0.37450744884160897,
-      "learning_rate": 1.5761533452944202e-05,
-      "loss": 0.3419,
+      "epoch": 1.1276313852495217,
+      "grad_norm": 0.37899990266269695,
+      "learning_rate": 1.571075459235549e-05,
+      "loss": 0.3371,
       "step": 7660
     },
     {
-      "epoch": 1.123076923076923,
-      "grad_norm": 0.3544280061502412,
-      "learning_rate": 1.5754563990760553e-05,
-      "loss": 0.33,
+      "epoch": 1.1283674370675696,
+      "grad_norm": 0.4079749309408199,
+      "learning_rate": 1.5703722004921814e-05,
+      "loss": 0.3258,
       "step": 7665
     },
     {
-      "epoch": 1.1238095238095238,
-      "grad_norm": 0.3717329727145123,
-      "learning_rate": 1.574759034697312e-05,
-      "loss": 0.3198,
+      "epoch": 1.1291034888856175,
+      "grad_norm": 0.3752359611414991,
+      "learning_rate": 1.5696685233869822e-05,
+      "loss": 0.3408,
       "step": 7670
     },
     {
-      "epoch": 1.1245421245421245,
-      "grad_norm": 0.36191640761951016,
-      "learning_rate": 1.5740612526649365e-05,
-      "loss": 0.3062,
+      "epoch": 1.1298395407036654,
+      "grad_norm": 0.3535172272002126,
+      "learning_rate": 1.5689644284360907e-05,
+      "loss": 0.3177,
       "step": 7675
     },
     {
-      "epoch": 1.1252747252747253,
-      "grad_norm": 0.3753666793046584,
-      "learning_rate": 1.573363053485978e-05,
-      "loss": 0.3243,
+      "epoch": 1.1305755925217136,
+      "grad_norm": 0.3578109901016605,
+      "learning_rate": 1.5682599161559526e-05,
+      "loss": 0.3284,
       "step": 7680
     },
     {
-      "epoch": 1.126007326007326,
-      "grad_norm": 0.3681268297340232,
-      "learning_rate": 1.572664437667789e-05,
-      "loss": 0.3201,
+      "epoch": 1.1313116443397615,
+      "grad_norm": 0.36881161064106277,
+      "learning_rate": 1.5675549870633203e-05,
+      "loss": 0.3294,
       "step": 7685
     },
     {
-      "epoch": 1.1267399267399267,
-      "grad_norm": 0.35574977395336677,
-      "learning_rate": 1.5719654057180252e-05,
-      "loss": 0.3249,
+      "epoch": 1.1320476961578094,
+      "grad_norm": 0.3769556399428797,
+      "learning_rate": 1.5668496416752516e-05,
+      "loss": 0.3231,
       "step": 7690
     },
     {
-      "epoch": 1.1274725274725275,
-      "grad_norm": 0.3605750556824243,
-      "learning_rate": 1.571265958144643e-05,
-      "loss": 0.3259,
+      "epoch": 1.1327837479758576,
+      "grad_norm": 0.37732308424714966,
+      "learning_rate": 1.56614388050911e-05,
+      "loss": 0.3397,
       "step": 7695
     },
     {
-      "epoch": 1.1282051282051282,
-      "grad_norm": 0.3794265246396994,
-      "learning_rate": 1.5705660954559035e-05,
-      "loss": 0.3415,
+      "epoch": 1.1335197997939055,
+      "grad_norm": 0.3786106463303893,
+      "learning_rate": 1.5654377040825634e-05,
+      "loss": 0.3448,
       "step": 7700
     },
     {
-      "epoch": 1.128937728937729,
-      "grad_norm": 0.3889151141065052,
-      "learning_rate": 1.569865818160367e-05,
-      "loss": 0.3201,
+      "epoch": 1.1342558516119534,
+      "grad_norm": 0.36074186899464955,
+      "learning_rate": 1.5647311129135846e-05,
+      "loss": 0.3256,
       "step": 7705
     },
     {
-      "epoch": 1.1296703296703297,
-      "grad_norm": 0.3520533857170965,
-      "learning_rate": 1.5691651267668968e-05,
-      "loss": 0.3272,
+      "epoch": 1.1349919034300016,
+      "grad_norm": 0.35720108673609746,
+      "learning_rate": 1.5640241075204507e-05,
+      "loss": 0.3234,
       "step": 7710
     },
     {
-      "epoch": 1.1304029304029304,
-      "grad_norm": 0.3754701024332263,
-      "learning_rate": 1.5684640217846555e-05,
-      "loss": 0.3368,
+      "epoch": 1.1357279552480495,
+      "grad_norm": 0.36684266672294585,
+      "learning_rate": 1.5633166884217424e-05,
+      "loss": 0.3346,
       "step": 7715
     },
     {
-      "epoch": 1.1311355311355311,
-      "grad_norm": 0.3470744278797411,
-      "learning_rate": 1.5677625037231075e-05,
-      "loss": 0.3256,
+      "epoch": 1.1364640070660974,
+      "grad_norm": 0.36578036204933856,
+      "learning_rate": 1.562608856136344e-05,
+      "loss": 0.3282,
       "step": 7720
     },
     {
-      "epoch": 1.1318681318681318,
-      "grad_norm": 0.3816451249663348,
-      "learning_rate": 1.5670605730920172e-05,
+      "epoch": 1.1372000588841455,
+      "grad_norm": 0.3664420736827633,
+      "learning_rate": 1.561900611183442e-05,
       "loss": 0.3202,
       "step": 7725
     },
     {
-      "epoch": 1.1326007326007326,
-      "grad_norm": 0.38362504353673,
-      "learning_rate": 1.5663582304014485e-05,
-      "loss": 0.3312,
+      "epoch": 1.1379361107021935,
+      "grad_norm": 0.36621600662848175,
+      "learning_rate": 1.561191954082528e-05,
+      "loss": 0.3218,
       "step": 7730
     },
     {
-      "epoch": 1.1333333333333333,
-      "grad_norm": 3.0992510017108925,
-      "learning_rate": 1.565655476161765e-05,
-      "loss": 0.3367,
+      "epoch": 1.1386721625202414,
+      "grad_norm": 0.3728211225397672,
+      "learning_rate": 1.5604828853533932e-05,
+      "loss": 0.3605,
       "step": 7735
     },
     {
-      "epoch": 1.134065934065934,
-      "grad_norm": 0.3783866557689988,
-      "learning_rate": 1.5649523108836283e-05,
-      "loss": 0.3281,
+      "epoch": 1.1394082143382893,
+      "grad_norm": 0.3607580410927885,
+      "learning_rate": 1.559773405516132e-05,
+      "loss": 0.3176,
       "step": 7740
     },
     {
-      "epoch": 1.1347985347985348,
-      "grad_norm": 0.37579886304900423,
-      "learning_rate": 1.5642487350780005e-05,
-      "loss": 0.3268,
+      "epoch": 1.1401442661563375,
+      "grad_norm": 0.3381911753649782,
+      "learning_rate": 1.5590635150911403e-05,
+      "loss": 0.3296,
       "step": 7745
     },
     {
-      "epoch": 1.1355311355311355,
-      "grad_norm": 0.38265023290453665,
-      "learning_rate": 1.5635447492561403e-05,
-      "loss": 0.3346,
+      "epoch": 1.1408803179743854,
+      "grad_norm": 0.36632901662190737,
+      "learning_rate": 1.5583532145991148e-05,
+      "loss": 0.326,
       "step": 7750
     },
     {
-      "epoch": 1.1362637362637362,
-      "grad_norm": 0.3737657873990553,
-      "learning_rate": 1.5628403539296058e-05,
-      "loss": 0.3476,
+      "epoch": 1.1416163697924333,
+      "grad_norm": 0.3710051357140936,
+      "learning_rate": 1.557642504561054e-05,
+      "loss": 0.335,
       "step": 7755
     },
     {
-      "epoch": 1.136996336996337,
-      "grad_norm": 0.38456256719679427,
-      "learning_rate": 1.5621355496102518e-05,
-      "loss": 0.3336,
+      "epoch": 1.1423524216104814,
+      "grad_norm": 0.3785712866752345,
+      "learning_rate": 1.5569313854982547e-05,
+      "loss": 0.3195,
       "step": 7760
     },
     {
-      "epoch": 1.1377289377289377,
-      "grad_norm": 0.3773294586268947,
-      "learning_rate": 1.5614303368102296e-05,
-      "loss": 0.3202,
+      "epoch": 1.1430884734285294,
+      "grad_norm": 0.3711928974743087,
+      "learning_rate": 1.5562198579323166e-05,
+      "loss": 0.3298,
       "step": 7765
     },
     {
-      "epoch": 1.1384615384615384,
-      "grad_norm": 0.3431922523735431,
-      "learning_rate": 1.5607247160419892e-05,
-      "loss": 0.3344,
+      "epoch": 1.1438245252465773,
+      "grad_norm": 0.3945063867288393,
+      "learning_rate": 1.555507922385136e-05,
+      "loss": 0.3303,
       "step": 7770
     },
     {
-      "epoch": 1.1391941391941391,
-      "grad_norm": 0.3641852418008754,
-      "learning_rate": 1.5600186878182765e-05,
-      "loss": 0.3271,
+      "epoch": 1.1445605770646254,
+      "grad_norm": 0.37004262858780046,
+      "learning_rate": 1.5547955793789108e-05,
+      "loss": 0.3285,
       "step": 7775
     },
     {
-      "epoch": 1.1399267399267399,
-      "grad_norm": 0.3589745389126589,
-      "learning_rate": 1.5593122526521316e-05,
-      "loss": 0.339,
+      "epoch": 1.1452966288826734,
+      "grad_norm": 0.367078837440859,
+      "learning_rate": 1.5540828294361366e-05,
+      "loss": 0.3154,
       "step": 7780
     },
     {
-      "epoch": 1.1406593406593406,
-      "grad_norm": 0.3837329528546989,
-      "learning_rate": 1.5586054110568924e-05,
-      "loss": 0.3295,
+      "epoch": 1.1460326807007213,
+      "grad_norm": 0.3805392937153245,
+      "learning_rate": 1.553369673079608e-05,
+      "loss": 0.3393,
       "step": 7785
     },
     {
-      "epoch": 1.1413919413919413,
-      "grad_norm": 0.35521255269951685,
-      "learning_rate": 1.557898163546191e-05,
-      "loss": 0.3275,
+      "epoch": 1.1467687325187694,
+      "grad_norm": 0.3869227223495453,
+      "learning_rate": 1.5526561108324165e-05,
+      "loss": 0.3256,
       "step": 7790
     },
     {
-      "epoch": 1.142124542124542,
-      "grad_norm": 0.3758897169810696,
-      "learning_rate": 1.557190510633956e-05,
-      "loss": 0.3445,
+      "epoch": 1.1475047843368174,
+      "grad_norm": 0.37063069123814935,
+      "learning_rate": 1.5519421432179537e-05,
+      "loss": 0.3546,
       "step": 7795
     },
     {
-      "epoch": 1.1428571428571428,
-      "grad_norm": 0.3668639865507922,
-      "learning_rate": 1.556482452834409e-05,
-      "loss": 0.3405,
+      "epoch": 1.1482408361548653,
+      "grad_norm": 0.397345443614491,
+      "learning_rate": 1.551227770759906e-05,
+      "loss": 0.322,
       "step": 7800
     },
     {
-      "epoch": 1.1435897435897435,
-      "grad_norm": 0.38211289546674637,
-      "learning_rate": 1.555773990662066e-05,
-      "loss": 0.3205,
+      "epoch": 1.1489768879729132,
+      "grad_norm": 0.3616619578212314,
+      "learning_rate": 1.550512993982259e-05,
+      "loss": 0.3245,
       "step": 7805
     },
     {
-      "epoch": 1.1443223443223443,
-      "grad_norm": 0.36502561275991563,
-      "learning_rate": 1.5550651246317377e-05,
-      "loss": 0.3152,
+      "epoch": 1.1497129397909613,
+      "grad_norm": 0.38807546453238934,
+      "learning_rate": 1.5497978134092925e-05,
+      "loss": 0.3321,
       "step": 7810
     },
     {
-      "epoch": 1.145054945054945,
-      "grad_norm": 0.3859358824241421,
-      "learning_rate": 1.554355855258527e-05,
-      "loss": 0.3328,
+      "epoch": 1.1504489916090093,
+      "grad_norm": 0.3802255631829449,
+      "learning_rate": 1.549082229565584e-05,
+      "loss": 0.3431,
       "step": 7815
     },
     {
-      "epoch": 1.1457875457875457,
-      "grad_norm": 0.3674888296797926,
-      "learning_rate": 1.5536461830578315e-05,
-      "loss": 0.3367,
+      "epoch": 1.1511850434270572,
+      "grad_norm": 0.3827544811135087,
+      "learning_rate": 1.5483662429760074e-05,
+      "loss": 0.3217,
       "step": 7820
     },
     {
-      "epoch": 1.1465201465201464,
-      "grad_norm": 0.38519406190838723,
-      "learning_rate": 1.5529361085453404e-05,
-      "loss": 0.335,
+      "epoch": 1.1519210952451053,
+      "grad_norm": 0.3763534138365317,
+      "learning_rate": 1.54764985416573e-05,
+      "loss": 0.3185,
       "step": 7825
     },
     {
-      "epoch": 1.1472527472527472,
-      "grad_norm": 0.37311672813530217,
-      "learning_rate": 1.5522256322370347e-05,
-      "loss": 0.3271,
+      "epoch": 1.1526571470631533,
+      "grad_norm": 0.3613715699731057,
+      "learning_rate": 1.546933063660216e-05,
+      "loss": 0.3546,
       "step": 7830
     },
     {
-      "epoch": 1.147985347985348,
-      "grad_norm": 0.3664044245248359,
-      "learning_rate": 1.5515147546491897e-05,
-      "loss": 0.3371,
+      "epoch": 1.1533931988812012,
+      "grad_norm": 0.39850937018007737,
+      "learning_rate": 1.546215871985223e-05,
+      "loss": 0.3424,
       "step": 7835
     },
     {
-      "epoch": 1.1487179487179486,
-      "grad_norm": 0.3558901582608226,
-      "learning_rate": 1.550803476298369e-05,
-      "loss": 0.3277,
+      "epoch": 1.1541292506992493,
+      "grad_norm": 0.36542223270984864,
+      "learning_rate": 1.5454982796668034e-05,
+      "loss": 0.3348,
       "step": 7840
     },
     {
-      "epoch": 1.1494505494505494,
-      "grad_norm": 0.3644128091998396,
-      "learning_rate": 1.5500917977014306e-05,
-      "loss": 0.3268,
+      "epoch": 1.1548653025172972,
+      "grad_norm": 0.3568104809348479,
+      "learning_rate": 1.5447802872313035e-05,
+      "loss": 0.3194,
       "step": 7845
     },
     {
-      "epoch": 1.15018315018315,
-      "grad_norm": 0.3701432111945302,
-      "learning_rate": 1.549379719375521e-05,
-      "loss": 0.3389,
+      "epoch": 1.1556013543353452,
+      "grad_norm": 0.36075462577489675,
+      "learning_rate": 1.544061895205363e-05,
+      "loss": 0.3203,
       "step": 7850
     },
     {
-      "epoch": 1.1509157509157508,
-      "grad_norm": 0.3794847352445503,
-      "learning_rate": 1.5486672418380786e-05,
-      "loss": 0.3449,
+      "epoch": 1.1563374061533933,
+      "grad_norm": 0.40442991289429586,
+      "learning_rate": 1.5433431041159145e-05,
+      "loss": 0.3348,
       "step": 7855
     },
     {
-      "epoch": 1.1516483516483516,
-      "grad_norm": 0.3492217553790412,
-      "learning_rate": 1.5479543656068313e-05,
-      "loss": 0.3208,
+      "epoch": 1.1570734579714412,
+      "grad_norm": 0.3779860773248329,
+      "learning_rate": 1.5426239144901836e-05,
+      "loss": 0.3187,
       "step": 7860
     },
     {
-      "epoch": 1.1523809523809523,
-      "grad_norm": 0.3736344818072641,
-      "learning_rate": 1.547241091199797e-05,
-      "loss": 0.3327,
+      "epoch": 1.1578095097894892,
+      "grad_norm": 0.38069957354854694,
+      "learning_rate": 1.5419043268556885e-05,
+      "loss": 0.341,
       "step": 7865
     },
     {
-      "epoch": 1.153113553113553,
-      "grad_norm": 0.37726343120982136,
-      "learning_rate": 1.546527419135282e-05,
-      "loss": 0.3342,
+      "epoch": 1.158545561607537,
+      "grad_norm": 0.3533837974365164,
+      "learning_rate": 1.5411843417402385e-05,
+      "loss": 0.3277,
       "step": 7870
     },
     {
-      "epoch": 1.1538461538461537,
-      "grad_norm": 0.3827996368281623,
-      "learning_rate": 1.545813349931883e-05,
-      "loss": 0.3351,
+      "epoch": 1.1592816134255852,
+      "grad_norm": 0.3555697016806076,
+      "learning_rate": 1.5404639596719344e-05,
+      "loss": 0.3273,
       "step": 7875
     },
     {
-      "epoch": 1.1545787545787545,
-      "grad_norm": 0.37145135784737665,
-      "learning_rate": 1.5450988841084844e-05,
-      "loss": 0.3035,
+      "epoch": 1.1600176652436331,
+      "grad_norm": 0.39214145093818986,
+      "learning_rate": 1.5397431811791695e-05,
+      "loss": 0.3322,
       "step": 7880
     },
     {
-      "epoch": 1.1553113553113552,
-      "grad_norm": 0.37420784345115,
-      "learning_rate": 1.5443840221842586e-05,
-      "loss": 0.3376,
+      "epoch": 1.160753717061681,
+      "grad_norm": 0.382712644714872,
+      "learning_rate": 1.5390220067906267e-05,
+      "loss": 0.3311,
       "step": 7885
     },
     {
-      "epoch": 1.156043956043956,
-      "grad_norm": 0.36574848734418397,
-      "learning_rate": 1.543668764678667e-05,
-      "loss": 0.3274,
+      "epoch": 1.1614897688797292,
+      "grad_norm": 0.38411978704744937,
+      "learning_rate": 1.5383004370352792e-05,
+      "loss": 0.3228,
       "step": 7890
     },
     {
-      "epoch": 1.1567765567765567,
-      "grad_norm": 0.3632006021916965,
-      "learning_rate": 1.5429531121114565e-05,
-      "loss": 0.3402,
+      "epoch": 1.1622258206977771,
+      "grad_norm": 0.36607839507668566,
+      "learning_rate": 1.5375784724423914e-05,
+      "loss": 0.3179,
       "step": 7895
     },
     {
-      "epoch": 1.1575091575091574,
-      "grad_norm": 0.38938668112252456,
-      "learning_rate": 1.542237065002663e-05,
-      "loss": 0.3325,
+      "epoch": 1.162961872515825,
+      "grad_norm": 0.35317476761041544,
+      "learning_rate": 1.536856113541516e-05,
+      "loss": 0.3258,
       "step": 7900
     },
     {
-      "epoch": 1.1582417582417581,
-      "grad_norm": 0.3938272738348802,
-      "learning_rate": 1.5415206238726086e-05,
-      "loss": 0.3357,
+      "epoch": 1.1636979243338732,
+      "grad_norm": 0.37769960094645616,
+      "learning_rate": 1.5361333608624955e-05,
+      "loss": 0.322,
       "step": 7905
     },
     {
-      "epoch": 1.1589743589743589,
-      "grad_norm": 0.3702130333111147,
-      "learning_rate": 1.5408037892419012e-05,
-      "loss": 0.3448,
+      "epoch": 1.1644339761519211,
+      "grad_norm": 0.6377854025940484,
+      "learning_rate": 1.5354102149354614e-05,
+      "loss": 0.3388,
       "step": 7910
     },
     {
-      "epoch": 1.1597069597069596,
-      "grad_norm": 0.37480011471437447,
-      "learning_rate": 1.5400865616314343e-05,
-      "loss": 0.3325,
+      "epoch": 1.165170027969969,
+      "grad_norm": 0.3842292521328781,
+      "learning_rate": 1.534686676290833e-05,
+      "loss": 0.339,
       "step": 7915
     },
     {
-      "epoch": 1.1604395604395605,
-      "grad_norm": 0.36020313134221427,
-      "learning_rate": 1.5393689415623883e-05,
-      "loss": 0.3229,
+      "epoch": 1.1659060797880172,
+      "grad_norm": 0.3792690103737219,
+      "learning_rate": 1.5339627454593186e-05,
+      "loss": 0.3266,
       "step": 7920
     },
     {
-      "epoch": 1.1611721611721613,
-      "grad_norm": 0.3621293229694593,
-      "learning_rate": 1.5386509295562278e-05,
-      "loss": 0.3431,
+      "epoch": 1.1666421316060651,
+      "grad_norm": 0.3676587886070995,
+      "learning_rate": 1.533238422971913e-05,
+      "loss": 0.3233,
       "step": 7925
     },
     {
-      "epoch": 1.161904761904762,
-      "grad_norm": 0.35748855267964647,
-      "learning_rate": 1.5379325261347024e-05,
-      "loss": 0.3254,
+      "epoch": 1.167378183424113,
+      "grad_norm": 0.36727558731704385,
+      "learning_rate": 1.532513709359899e-05,
+      "loss": 0.315,
       "step": 7930
     },
     {
-      "epoch": 1.1626373626373627,
-      "grad_norm": 0.3442282617139529,
-      "learning_rate": 1.5372137318198463e-05,
-      "loss": 0.3017,
+      "epoch": 1.168114235242161,
+      "grad_norm": 0.3506297472158483,
+      "learning_rate": 1.5317886051548466e-05,
+      "loss": 0.3274,
       "step": 7935
     },
     {
-      "epoch": 1.1633699633699635,
-      "grad_norm": 0.3787677128934951,
-      "learning_rate": 1.5364945471339773e-05,
-      "loss": 0.3173,
+      "epoch": 1.168850287060209,
+      "grad_norm": 0.3664467108879879,
+      "learning_rate": 1.5310631108886116e-05,
+      "loss": 0.3302,
       "step": 7940
     },
     {
-      "epoch": 1.1641025641025642,
-      "grad_norm": 0.36004708114704104,
-      "learning_rate": 1.535774972599698e-05,
-      "loss": 0.3207,
+      "epoch": 1.169586338878257,
+      "grad_norm": 0.35742067793297144,
+      "learning_rate": 1.530337227093336e-05,
+      "loss": 0.3317,
       "step": 7945
     },
     {
-      "epoch": 1.164835164835165,
-      "grad_norm": 0.37144192704319284,
-      "learning_rate": 1.5350550087398924e-05,
-      "loss": 0.3221,
+      "epoch": 1.170322390696305,
+      "grad_norm": 0.3683776726669245,
+      "learning_rate": 1.5296109543014485e-05,
+      "loss": 0.3158,
       "step": 7950
     },
     {
-      "epoch": 1.1655677655677656,
-      "grad_norm": 0.36107154609209374,
-      "learning_rate": 1.5343346560777287e-05,
-      "loss": 0.3289,
+      "epoch": 1.171058442514353,
+      "grad_norm": 0.3518427229639874,
+      "learning_rate": 1.5288842930456612e-05,
+      "loss": 0.336,
       "step": 7955
     },
     {
-      "epoch": 1.1663003663003664,
-      "grad_norm": 0.35726675993109364,
-      "learning_rate": 1.533613915136658e-05,
-      "loss": 0.3127,
+      "epoch": 1.171794494332401,
+      "grad_norm": 0.3487845596887438,
+      "learning_rate": 1.5281572438589733e-05,
+      "loss": 0.3198,
       "step": 7960
     },
     {
-      "epoch": 1.167032967032967,
-      "grad_norm": 0.3451931270804777,
-      "learning_rate": 1.5328927864404125e-05,
-      "loss": 0.3194,
+      "epoch": 1.172530546150449,
+      "grad_norm": 0.379988722680662,
+      "learning_rate": 1.5274298072746667e-05,
+      "loss": 0.3363,
       "step": 7965
     },
     {
-      "epoch": 1.1677655677655678,
-      "grad_norm": 0.3539475521927589,
-      "learning_rate": 1.532171270513007e-05,
-      "loss": 0.3358,
+      "epoch": 1.1732665979684969,
+      "grad_norm": 0.3549929624633907,
+      "learning_rate": 1.5267019838263087e-05,
+      "loss": 0.3217,
       "step": 7970
     },
     {
-      "epoch": 1.1684981684981686,
-      "grad_norm": 0.37785181308734567,
-      "learning_rate": 1.5314493678787367e-05,
-      "loss": 0.3331,
+      "epoch": 1.174002649786545,
+      "grad_norm": 0.35987419358775735,
+      "learning_rate": 1.52597377404775e-05,
+      "loss": 0.3193,
       "step": 7975
     },
     {
-      "epoch": 1.1692307692307693,
-      "grad_norm": 0.35682878347893937,
-      "learning_rate": 1.5307270790621786e-05,
-      "loss": 0.3358,
+      "epoch": 1.174738701604593,
+      "grad_norm": 0.36268787035510847,
+      "learning_rate": 1.525245178473124e-05,
+      "loss": 0.332,
       "step": 7980
     },
     {
-      "epoch": 1.16996336996337,
-      "grad_norm": 0.3528224207274319,
-      "learning_rate": 1.5300044045881904e-05,
-      "loss": 0.3308,
+      "epoch": 1.175474753422641,
+      "grad_norm": 0.35312966399350015,
+      "learning_rate": 1.5245161976368483e-05,
+      "loss": 0.3245,
       "step": 7985
     },
     {
-      "epoch": 1.1706959706959708,
-      "grad_norm": 0.39431002749350935,
-      "learning_rate": 1.5292813449819096e-05,
-      "loss": 0.3289,
+      "epoch": 1.176210805240689,
+      "grad_norm": 0.3695684293336986,
+      "learning_rate": 1.523786832073622e-05,
+      "loss": 0.3304,
       "step": 7990
     },
     {
-      "epoch": 1.1714285714285715,
-      "grad_norm": 0.3814929945950606,
-      "learning_rate": 1.528557900768754e-05,
-      "loss": 0.3426,
+      "epoch": 1.176946857058737,
+      "grad_norm": 0.3627286953491536,
+      "learning_rate": 1.5230570823184275e-05,
+      "loss": 0.3201,
       "step": 7995
     },
     {
-      "epoch": 1.1721611721611722,
-      "grad_norm": 0.366680665409208,
-      "learning_rate": 1.5278340724744204e-05,
-      "loss": 0.3314,
+      "epoch": 1.1776829088767848,
+      "grad_norm": 0.40544249315424846,
+      "learning_rate": 1.5223269489065278e-05,
+      "loss": 0.3462,
       "step": 8000
     },
     {
-      "epoch": 1.172893772893773,
-      "grad_norm": 0.3770267801244452,
-      "learning_rate": 1.5271098606248848e-05,
-      "loss": 0.3462,
+      "epoch": 1.178418960694833,
+      "grad_norm": 0.36346570837263104,
+      "learning_rate": 1.5215964323734676e-05,
+      "loss": 0.3324,
       "step": 8005
     },
     {
-      "epoch": 1.1736263736263737,
-      "grad_norm": 0.3514381440374758,
-      "learning_rate": 1.5263852657464016e-05,
-      "loss": 0.3194,
+      "epoch": 1.179155012512881,
+      "grad_norm": 0.3838103972999597,
+      "learning_rate": 1.5208655332550733e-05,
+      "loss": 0.3408,
       "step": 8010
     },
     {
-      "epoch": 1.1743589743589744,
-      "grad_norm": 0.4070492328211306,
-      "learning_rate": 1.5256602883655048e-05,
-      "loss": 0.3406,
+      "epoch": 1.1798910643309288,
+      "grad_norm": 0.3649917618606179,
+      "learning_rate": 1.5201342520874516e-05,
+      "loss": 0.328,
       "step": 8015
     },
     {
-      "epoch": 1.1750915750915751,
-      "grad_norm": 0.37712251514160994,
-      "learning_rate": 1.524934929009005e-05,
-      "loss": 0.334,
+      "epoch": 1.180627116148977,
+      "grad_norm": 0.350483182694087,
+      "learning_rate": 1.5194025894069889e-05,
+      "loss": 0.3277,
       "step": 8020
     },
     {
-      "epoch": 1.1758241758241759,
-      "grad_norm": 0.36395694590387156,
-      "learning_rate": 1.524209188203991e-05,
-      "loss": 0.3306,
+      "epoch": 1.181363167967025,
+      "grad_norm": 0.3650230449083749,
+      "learning_rate": 1.5186705457503518e-05,
+      "loss": 0.3338,
       "step": 8025
     },
     {
-      "epoch": 1.1765567765567766,
-      "grad_norm": 0.34500068083160973,
-      "learning_rate": 1.5234830664778283e-05,
-      "loss": 0.3239,
+      "epoch": 1.1820992197850728,
+      "grad_norm": 0.3757414247876498,
+      "learning_rate": 1.5179381216544871e-05,
+      "loss": 0.3355,
       "step": 8030
     },
     {
-      "epoch": 1.1772893772893773,
-      "grad_norm": 0.38764441603246896,
-      "learning_rate": 1.5227565643581604e-05,
-      "loss": 0.3516,
+      "epoch": 1.1828352716031207,
+      "grad_norm": 0.3581694313948743,
+      "learning_rate": 1.5172053176566193e-05,
+      "loss": 0.3266,
       "step": 8035
     },
     {
-      "epoch": 1.178021978021978,
-      "grad_norm": 0.37146934837029344,
-      "learning_rate": 1.5220296823729058e-05,
-      "loss": 0.3289,
+      "epoch": 1.1835713234211689,
+      "grad_norm": 0.35127920224525677,
+      "learning_rate": 1.5164721342942525e-05,
+      "loss": 0.3195,
       "step": 8040
     },
     {
-      "epoch": 1.1787545787545788,
-      "grad_norm": 0.37252913742566135,
-      "learning_rate": 1.5213024210502592e-05,
-      "loss": 0.3312,
+      "epoch": 1.1843073752392168,
+      "grad_norm": 0.382706540935447,
+      "learning_rate": 1.5157385721051684e-05,
+      "loss": 0.3187,
       "step": 8045
     },
     {
-      "epoch": 1.1794871794871795,
-      "grad_norm": 0.3724879987768289,
-      "learning_rate": 1.520574780918692e-05,
-      "loss": 0.3237,
+      "epoch": 1.185043427057265,
+      "grad_norm": 0.3557674138377118,
+      "learning_rate": 1.5150046316274279e-05,
+      "loss": 0.3196,
       "step": 8050
     },
     {
-      "epoch": 1.1802197802197802,
-      "grad_norm": 0.36320171408168545,
-      "learning_rate": 1.5198467625069495e-05,
-      "loss": 0.3148,
+      "epoch": 1.1857794788753129,
+      "grad_norm": 0.4068462556776589,
+      "learning_rate": 1.514270313399367e-05,
+      "loss": 0.3488,
       "step": 8055
     },
     {
-      "epoch": 1.180952380952381,
-      "grad_norm": 0.36288113898737523,
-      "learning_rate": 1.5191183663440536e-05,
-      "loss": 0.3197,
+      "epoch": 1.1865155306933608,
+      "grad_norm": 0.3616144646434693,
+      "learning_rate": 1.5135356179596008e-05,
+      "loss": 0.3422,
       "step": 8060
     },
     {
-      "epoch": 1.1816849816849817,
-      "grad_norm": 0.379903173867347,
-      "learning_rate": 1.518389592959299e-05,
-      "loss": 0.3384,
+      "epoch": 1.1872515825114087,
+      "grad_norm": 0.36811862803805906,
+      "learning_rate": 1.5128005458470207e-05,
+      "loss": 0.3158,
       "step": 8065
     },
     {
-      "epoch": 1.1824175824175824,
-      "grad_norm": 0.3616805925974018,
-      "learning_rate": 1.5176604428822553e-05,
-      "loss": 0.3313,
+      "epoch": 1.1879876343294569,
+      "grad_norm": 0.3674274299321005,
+      "learning_rate": 1.5120650976007941e-05,
+      "loss": 0.3316,
       "step": 8070
     },
     {
-      "epoch": 1.1831501831501832,
-      "grad_norm": 0.36957783238072606,
-      "learning_rate": 1.5169309166427658e-05,
-      "loss": 0.311,
+      "epoch": 1.1887236861475048,
+      "grad_norm": 0.3802645780409709,
+      "learning_rate": 1.5113292737603637e-05,
+      "loss": 0.3447,
       "step": 8075
     },
     {
-      "epoch": 1.183882783882784,
-      "grad_norm": 0.3953556623600997,
-      "learning_rate": 1.5162010147709467e-05,
-      "loss": 0.3216,
+      "epoch": 1.1894597379655527,
+      "grad_norm": 0.37954697064242204,
+      "learning_rate": 1.510593074865449e-05,
+      "loss": 0.3324,
       "step": 8080
     },
     {
-      "epoch": 1.1846153846153846,
-      "grad_norm": 0.3656002263760393,
-      "learning_rate": 1.515470737797188e-05,
-      "loss": 0.3314,
+      "epoch": 1.1901957897836009,
+      "grad_norm": 0.35419114349413583,
+      "learning_rate": 1.5098565014560436e-05,
+      "loss": 0.3438,
       "step": 8085
     },
     {
-      "epoch": 1.1853479853479854,
-      "grad_norm": 0.3646806027598352,
-      "learning_rate": 1.5147400862521519e-05,
-      "loss": 0.3153,
+      "epoch": 1.1909318416016488,
+      "grad_norm": 0.3849631664576678,
+      "learning_rate": 1.5091195540724161e-05,
+      "loss": 0.3388,
       "step": 8090
     },
     {
-      "epoch": 1.186080586080586,
-      "grad_norm": 0.37511690500361244,
-      "learning_rate": 1.514009060666772e-05,
-      "loss": 0.3384,
+      "epoch": 1.1916678934196967,
+      "grad_norm": 0.36959082043307084,
+      "learning_rate": 1.5083822332551096e-05,
+      "loss": 0.336,
       "step": 8095
     },
     {
-      "epoch": 1.1868131868131868,
-      "grad_norm": 0.35905818732390354,
-      "learning_rate": 1.5132776615722548e-05,
-      "loss": 0.3175,
+      "epoch": 1.1924039452377446,
+      "grad_norm": 0.3747436998910247,
+      "learning_rate": 1.5076445395449408e-05,
+      "loss": 0.3541,
       "step": 8100
     },
     {
-      "epoch": 1.1875457875457875,
-      "grad_norm": 0.34736547131077833,
-      "learning_rate": 1.512545889500078e-05,
-      "loss": 0.3477,
+      "epoch": 1.1931399970557928,
+      "grad_norm": 0.3676246232971342,
+      "learning_rate": 1.5069064734830004e-05,
+      "loss": 0.3208,
       "step": 8105
     },
     {
-      "epoch": 1.1882783882783883,
-      "grad_norm": 0.38204286817696836,
-      "learning_rate": 1.5118137449819896e-05,
-      "loss": 0.3266,
+      "epoch": 1.1938760488738407,
+      "grad_norm": 0.3732760017534803,
+      "learning_rate": 1.5061680356106512e-05,
+      "loss": 0.3296,
       "step": 8110
     },
     {
-      "epoch": 1.189010989010989,
-      "grad_norm": 0.3695606595203843,
-      "learning_rate": 1.5110812285500092e-05,
-      "loss": 0.3444,
+      "epoch": 1.1946121006918886,
+      "grad_norm": 0.361151747364808,
+      "learning_rate": 1.5054292264695298e-05,
+      "loss": 0.312,
       "step": 8115
     },
     {
-      "epoch": 1.1897435897435897,
-      "grad_norm": 0.3662372991469289,
-      "learning_rate": 1.5103483407364259e-05,
-      "loss": 0.3438,
+      "epoch": 1.1953481525099368,
+      "grad_norm": 0.3525055638444933,
+      "learning_rate": 1.5046900466015449e-05,
+      "loss": 0.331,
       "step": 8120
     },
     {
-      "epoch": 1.1904761904761905,
-      "grad_norm": 0.3800380320948473,
-      "learning_rate": 1.5096150820737993e-05,
-      "loss": 0.3411,
+      "epoch": 1.1960842043279847,
+      "grad_norm": 0.34297585949643866,
+      "learning_rate": 1.5039504965488763e-05,
+      "loss": 0.3256,
       "step": 8125
     },
     {
-      "epoch": 1.1912087912087912,
-      "grad_norm": 0.3681599569688727,
-      "learning_rate": 1.5088814530949578e-05,
-      "loss": 0.3162,
+      "epoch": 1.1968202561460326,
+      "grad_norm": 0.36955829764220444,
+      "learning_rate": 1.5032105768539766e-05,
+      "loss": 0.3303,
       "step": 8130
     },
     {
-      "epoch": 1.191941391941392,
-      "grad_norm": 0.3690148512442379,
-      "learning_rate": 1.5081474543329994e-05,
+      "epoch": 1.1975563079640807,
+      "grad_norm": 0.367012429873709,
+      "learning_rate": 1.5024702880595685e-05,
       "loss": 0.3386,
       "step": 8135
     },
     {
-      "epoch": 1.1926739926739927,
-      "grad_norm": 0.36979493878805475,
-      "learning_rate": 1.507413086321291e-05,
-      "loss": 0.3222,
+      "epoch": 1.1982923597821287,
+      "grad_norm": 0.4441685962696162,
+      "learning_rate": 1.501729630708646e-05,
+      "loss": 0.3397,
       "step": 8140
     },
     {
-      "epoch": 1.1934065934065934,
-      "grad_norm": 0.37053537861558267,
-      "learning_rate": 1.5066783495934666e-05,
-      "loss": 0.3184,
+      "epoch": 1.1990284116001766,
+      "grad_norm": 0.3749987318191806,
+      "learning_rate": 1.5009886053444731e-05,
+      "loss": 0.3199,
       "step": 8145
     },
     {
-      "epoch": 1.1941391941391941,
-      "grad_norm": 0.3616924947670475,
-      "learning_rate": 1.5059432446834304e-05,
-      "loss": 0.3302,
+      "epoch": 1.1997644634182247,
+      "grad_norm": 0.3621790453109267,
+      "learning_rate": 1.5002472125105839e-05,
+      "loss": 0.3239,
       "step": 8150
     },
     {
-      "epoch": 1.1948717948717948,
-      "grad_norm": 0.3609144662763727,
-      "learning_rate": 1.5052077721253513e-05,
-      "loss": 0.3392,
+      "epoch": 1.2005005152362727,
+      "grad_norm": 0.3644631410689124,
+      "learning_rate": 1.4995054527507819e-05,
+      "loss": 0.3427,
       "step": 8155
     },
     {
-      "epoch": 1.1956043956043956,
-      "grad_norm": 0.3883153977065576,
-      "learning_rate": 1.5044719324536678e-05,
-      "loss": 0.3316,
+      "epoch": 1.2012365670543206,
+      "grad_norm": 0.3647555049194253,
+      "learning_rate": 1.4987633266091401e-05,
+      "loss": 0.3372,
       "step": 8160
     },
     {
-      "epoch": 1.1963369963369963,
-      "grad_norm": 0.3839491724986923,
-      "learning_rate": 1.5037357262030833e-05,
-      "loss": 0.3404,
+      "epoch": 1.2019726188723685,
+      "grad_norm": 0.37004960872914333,
+      "learning_rate": 1.4980208346300003e-05,
+      "loss": 0.3447,
       "step": 8165
     },
     {
-      "epoch": 1.197069597069597,
-      "grad_norm": 0.36233751768245326,
-      "learning_rate": 1.5029991539085695e-05,
-      "loss": 0.3174,
+      "epoch": 1.2027086706904166,
+      "grad_norm": 0.35489534411365686,
+      "learning_rate": 1.4972779773579715e-05,
+      "loss": 0.3406,
       "step": 8170
     },
     {
-      "epoch": 1.1978021978021978,
-      "grad_norm": 0.35458010898660675,
-      "learning_rate": 1.502262216105362e-05,
-      "loss": 0.3392,
+      "epoch": 1.2034447225084646,
+      "grad_norm": 0.3746861117742125,
+      "learning_rate": 1.496534755337932e-05,
+      "loss": 0.3282,
       "step": 8175
     },
     {
-      "epoch": 1.1985347985347985,
-      "grad_norm": 0.37002015116368614,
-      "learning_rate": 1.501524913328964e-05,
-      "loss": 0.3449,
+      "epoch": 1.2041807743265125,
+      "grad_norm": 0.3466432584153397,
+      "learning_rate": 1.495791169115027e-05,
+      "loss": 0.3171,
       "step": 8180
     },
     {
-      "epoch": 1.1992673992673992,
-      "grad_norm": 0.3872586755325121,
-      "learning_rate": 1.5007872461151422e-05,
-      "loss": 0.3414,
+      "epoch": 1.2049168261445606,
+      "grad_norm": 0.37221546959629925,
+      "learning_rate": 1.4950472192346692e-05,
+      "loss": 0.3356,
       "step": 8185
     },
     {
-      "epoch": 1.2,
-      "grad_norm": 0.37861235573923263,
-      "learning_rate": 1.5000492149999292e-05,
-      "loss": 0.3351,
+      "epoch": 1.2056528779626086,
+      "grad_norm": 0.35670248460846826,
+      "learning_rate": 1.4943029062425371e-05,
+      "loss": 0.3203,
       "step": 8190
     },
     {
-      "epoch": 1.2007326007326007,
-      "grad_norm": 0.367540897067357,
-      "learning_rate": 1.499310820519622e-05,
-      "loss": 0.3353,
+      "epoch": 1.2063889297806565,
+      "grad_norm": 0.3688758812161541,
+      "learning_rate": 1.4935582306845766e-05,
+      "loss": 0.3268,
       "step": 8195
     },
     {
-      "epoch": 1.2014652014652014,
-      "grad_norm": 0.3724045832054376,
-      "learning_rate": 1.4985720632107806e-05,
-      "loss": 0.3482,
+      "epoch": 1.2071249815987046,
+      "grad_norm": 0.3513532074539236,
+      "learning_rate": 1.4928131931069993e-05,
+      "loss": 0.3247,
       "step": 8200
     },
     {
-      "epoch": 1.2021978021978021,
-      "grad_norm": 0.35784985312179274,
-      "learning_rate": 1.4978329436102299e-05,
-      "loss": 0.311,
+      "epoch": 1.2078610334167525,
+      "grad_norm": 0.36681068074591217,
+      "learning_rate": 1.4920677940562821e-05,
+      "loss": 0.3264,
       "step": 8205
     },
     {
-      "epoch": 1.2029304029304029,
-      "grad_norm": 0.37299163526608586,
-      "learning_rate": 1.4970934622550575e-05,
-      "loss": 0.3265,
+      "epoch": 1.2085970852348005,
+      "grad_norm": 0.36728867213792643,
+      "learning_rate": 1.4913220340791675e-05,
+      "loss": 0.3354,
       "step": 8210
     },
     {
-      "epoch": 1.2036630036630036,
-      "grad_norm": 0.406549166369611,
-      "learning_rate": 1.4963536196826142e-05,
-      "loss": 0.3262,
+      "epoch": 1.2093331370528486,
+      "grad_norm": 0.3734185422872777,
+      "learning_rate": 1.4905759137226613e-05,
+      "loss": 0.3213,
       "step": 8215
     },
     {
-      "epoch": 1.2043956043956043,
-      "grad_norm": 0.37598436424901355,
-      "learning_rate": 1.4956134164305122e-05,
-      "loss": 0.3305,
+      "epoch": 1.2100691888708965,
+      "grad_norm": 0.34587592794620753,
+      "learning_rate": 1.4898294335340362e-05,
+      "loss": 0.3323,
       "step": 8220
     },
     {
-      "epoch": 1.205128205128205,
-      "grad_norm": 0.387205644932967,
-      "learning_rate": 1.4948728530366274e-05,
-      "loss": 0.3303,
+      "epoch": 1.2108052406889445,
+      "grad_norm": 0.391945239905069,
+      "learning_rate": 1.4890825940608258e-05,
+      "loss": 0.3435,
       "step": 8225
     },
     {
-      "epoch": 1.2058608058608058,
-      "grad_norm": 0.45070623988161446,
-      "learning_rate": 1.4941319300390964e-05,
-      "loss": 0.3428,
+      "epoch": 1.2115412925069924,
+      "grad_norm": 0.3769895340415957,
+      "learning_rate": 1.4883353958508297e-05,
+      "loss": 0.3243,
       "step": 8230
     },
     {
-      "epoch": 1.2065934065934065,
-      "grad_norm": 0.3812905313874793,
-      "learning_rate": 1.4933906479763174e-05,
-      "loss": 0.3374,
+      "epoch": 1.2122773443250405,
+      "grad_norm": 0.3656163340344307,
+      "learning_rate": 1.4875878394521095e-05,
+      "loss": 0.3246,
       "step": 8235
     },
     {
-      "epoch": 1.2073260073260073,
-      "grad_norm": 0.3578092009296961,
-      "learning_rate": 1.4926490073869489e-05,
-      "loss": 0.3223,
+      "epoch": 1.2130133961430885,
+      "grad_norm": 0.3833402863688335,
+      "learning_rate": 1.4868399254129897e-05,
+      "loss": 0.3331,
       "step": 8240
     },
     {
-      "epoch": 1.208058608058608,
-      "grad_norm": 0.3560719148641902,
-      "learning_rate": 1.491907008809911e-05,
-      "loss": 0.3478,
+      "epoch": 1.2137494479611364,
+      "grad_norm": 0.36664815475729085,
+      "learning_rate": 1.4860916542820572e-05,
+      "loss": 0.3264,
       "step": 8245
     },
     {
-      "epoch": 1.2087912087912087,
-      "grad_norm": 0.37157832798696866,
-      "learning_rate": 1.491164652784383e-05,
-      "loss": 0.335,
+      "epoch": 1.2144854997791845,
+      "grad_norm": 0.35949597475343587,
+      "learning_rate": 1.4853430266081605e-05,
+      "loss": 0.3183,
       "step": 8250
     },
     {
-      "epoch": 1.2095238095238094,
-      "grad_norm": 0.3739105088939359,
-      "learning_rate": 1.4904219398498048e-05,
-      "loss": 0.3167,
+      "epoch": 1.2152215515972324,
+      "grad_norm": 0.3660859424494218,
+      "learning_rate": 1.4845940429404102e-05,
+      "loss": 0.3263,
       "step": 8255
     },
     {
-      "epoch": 1.2102564102564102,
-      "grad_norm": 0.38130465027508265,
-      "learning_rate": 1.489678870545875e-05,
-      "loss": 0.3272,
+      "epoch": 1.2159576034152804,
+      "grad_norm": 0.38645814628202896,
+      "learning_rate": 1.4838447038281776e-05,
+      "loss": 0.3351,
       "step": 8260
     },
     {
-      "epoch": 1.210989010989011,
-      "grad_norm": 0.36724998187833974,
-      "learning_rate": 1.4889354454125514e-05,
-      "loss": 0.3222,
+      "epoch": 1.2166936552333285,
+      "grad_norm": 0.37364129020267395,
+      "learning_rate": 1.483095009821095e-05,
+      "loss": 0.3368,
       "step": 8265
     },
     {
-      "epoch": 1.2117216117216116,
-      "grad_norm": 0.35581485026719434,
-      "learning_rate": 1.48819166499005e-05,
-      "loss": 0.3272,
+      "epoch": 1.2174297070513764,
+      "grad_norm": 0.3795810400075801,
+      "learning_rate": 1.4823449614690546e-05,
+      "loss": 0.3402,
       "step": 8270
     },
     {
-      "epoch": 1.2124542124542124,
-      "grad_norm": 0.3613045999885797,
-      "learning_rate": 1.4874475298188459e-05,
-      "loss": 0.3314,
+      "epoch": 1.2181657588694244,
+      "grad_norm": 0.37596010996766704,
+      "learning_rate": 1.4815945593222087e-05,
+      "loss": 0.3269,
       "step": 8275
     },
     {
-      "epoch": 1.213186813186813,
-      "grad_norm": 0.3796933374568938,
-      "learning_rate": 1.4867030404396704e-05,
-      "loss": 0.3452,
+      "epoch": 1.2189018106874725,
+      "grad_norm": 0.36523560144358447,
+      "learning_rate": 1.4808438039309694e-05,
+      "loss": 0.3252,
       "step": 8280
     },
     {
-      "epoch": 1.2139194139194138,
-      "grad_norm": 0.37809382765289035,
-      "learning_rate": 1.4859581973935143e-05,
-      "loss": 0.3235,
+      "epoch": 1.2196378625055204,
+      "grad_norm": 0.34706175679498286,
+      "learning_rate": 1.4800926958460074e-05,
+      "loss": 0.3258,
       "step": 8285
     },
     {
-      "epoch": 1.2146520146520146,
-      "grad_norm": 0.375067447560607,
-      "learning_rate": 1.4852130012216233e-05,
-      "loss": 0.3282,
+      "epoch": 1.2203739143235683,
+      "grad_norm": 0.38196664520781104,
+      "learning_rate": 1.4793412356182524e-05,
+      "loss": 0.3464,
       "step": 8290
     },
     {
-      "epoch": 1.2153846153846155,
-      "grad_norm": 0.3677148038994275,
-      "learning_rate": 1.4844674524655012e-05,
-      "loss": 0.3162,
+      "epoch": 1.2211099661416163,
+      "grad_norm": 0.36135280891199084,
+      "learning_rate": 1.4785894237988925e-05,
+      "loss": 0.3427,
       "step": 8295
     },
     {
-      "epoch": 1.2161172161172162,
-      "grad_norm": 0.3544576137509881,
-      "learning_rate": 1.4837215516669075e-05,
-      "loss": 0.3264,
+      "epoch": 1.2218460179596644,
+      "grad_norm": 0.37205099054015695,
+      "learning_rate": 1.4778372609393736e-05,
+      "loss": 0.341,
       "step": 8300
     },
     {
-      "epoch": 1.216849816849817,
-      "grad_norm": 0.3652573674769994,
-      "learning_rate": 1.4829752993678567e-05,
-      "loss": 0.3168,
+      "epoch": 1.2225820697777123,
+      "grad_norm": 0.37013183154571094,
+      "learning_rate": 1.4770847475913987e-05,
+      "loss": 0.3389,
       "step": 8305
     },
     {
-      "epoch": 1.2175824175824177,
-      "grad_norm": 0.3630429672853273,
-      "learning_rate": 1.4822286961106207e-05,
-      "loss": 0.3436,
+      "epoch": 1.2233181215957603,
+      "grad_norm": 0.38064850562116975,
+      "learning_rate": 1.4763318843069283e-05,
+      "loss": 0.3395,
       "step": 8310
     },
     {
-      "epoch": 1.2183150183150184,
-      "grad_norm": 0.367403987412385,
-      "learning_rate": 1.4814817424377243e-05,
-      "loss": 0.3376,
+      "epoch": 1.2240541734138084,
+      "grad_norm": 0.35981229827963146,
+      "learning_rate": 1.4755786716381795e-05,
+      "loss": 0.3375,
       "step": 8315
     },
     {
-      "epoch": 1.2190476190476192,
-      "grad_norm": 0.3483654972378673,
-      "learning_rate": 1.4807344388919482e-05,
-      "loss": 0.3344,
+      "epoch": 1.2247902252318563,
+      "grad_norm": 0.37567356663031276,
+      "learning_rate": 1.4748251101376254e-05,
+      "loss": 0.329,
       "step": 8320
     },
     {
-      "epoch": 1.2197802197802199,
-      "grad_norm": 0.38261252546018976,
-      "learning_rate": 1.4799867860163271e-05,
-      "loss": 0.3254,
+      "epoch": 1.2255262770499042,
+      "grad_norm": 0.3524202293263575,
+      "learning_rate": 1.4740712003579958e-05,
+      "loss": 0.3238,
       "step": 8325
     },
     {
-      "epoch": 1.2205128205128206,
-      "grad_norm": 0.3474567263427877,
-      "learning_rate": 1.4792387843541493e-05,
-      "loss": 0.3348,
+      "epoch": 1.2262623288679524,
+      "grad_norm": 0.383927065818772,
+      "learning_rate": 1.4733169428522747e-05,
+      "loss": 0.3261,
       "step": 8330
     },
     {
-      "epoch": 1.2212454212454213,
-      "grad_norm": 0.37537770891416594,
-      "learning_rate": 1.4784904344489566e-05,
-      "loss": 0.3284,
+      "epoch": 1.2269983806860003,
+      "grad_norm": 0.3599752959297416,
+      "learning_rate": 1.472562338173702e-05,
+      "loss": 0.3303,
       "step": 8335
     },
     {
-      "epoch": 1.221978021978022,
-      "grad_norm": 0.3537658201505351,
-      "learning_rate": 1.477741736844544e-05,
-      "loss": 0.3231,
+      "epoch": 1.2277344325040482,
+      "grad_norm": 0.3628757359151891,
+      "learning_rate": 1.4718073868757725e-05,
+      "loss": 0.317,
       "step": 8340
     },
     {
-      "epoch": 1.2227106227106228,
-      "grad_norm": 0.36084436630424543,
-      "learning_rate": 1.4769926920849594e-05,
-      "loss": 0.3271,
+      "epoch": 1.2284704843220964,
+      "grad_norm": 0.3479433541346978,
+      "learning_rate": 1.4710520895122342e-05,
+      "loss": 0.3229,
       "step": 8345
     },
     {
-      "epoch": 1.2234432234432235,
-      "grad_norm": 0.3599157152830956,
-      "learning_rate": 1.4762433007145023e-05,
-      "loss": 0.3226,
+      "epoch": 1.2292065361401443,
+      "grad_norm": 0.37622766300668603,
+      "learning_rate": 1.47029644663709e-05,
+      "loss": 0.3229,
       "step": 8350
     },
     {
-      "epoch": 1.2241758241758243,
-      "grad_norm": 0.3672901470854127,
-      "learning_rate": 1.4754935632777243e-05,
-      "loss": 0.3362,
+      "epoch": 1.2299425879581922,
+      "grad_norm": 0.3723072755457436,
+      "learning_rate": 1.4695404588045957e-05,
+      "loss": 0.3324,
       "step": 8355
     },
     {
-      "epoch": 1.224908424908425,
-      "grad_norm": 0.3678116918328648,
-      "learning_rate": 1.474743480319429e-05,
-      "loss": 0.32,
+      "epoch": 1.2306786397762401,
+      "grad_norm": 0.36619369272822905,
+      "learning_rate": 1.4687841265692595e-05,
+      "loss": 0.3279,
       "step": 8360
     },
     {
-      "epoch": 1.2256410256410257,
-      "grad_norm": 0.37755682120382705,
-      "learning_rate": 1.4739930523846706e-05,
-      "loss": 0.3328,
+      "epoch": 1.2314146915942883,
+      "grad_norm": 0.3576517937588514,
+      "learning_rate": 1.4680274504858438e-05,
+      "loss": 0.3378,
       "step": 8365
     },
     {
-      "epoch": 1.2263736263736265,
-      "grad_norm": 0.37187743144942614,
-      "learning_rate": 1.4732422800187541e-05,
-      "loss": 0.3203,
+      "epoch": 1.2321507434123362,
+      "grad_norm": 0.35634771515086305,
+      "learning_rate": 1.467270431109362e-05,
+      "loss": 0.3325,
       "step": 8370
     },
     {
-      "epoch": 1.2271062271062272,
-      "grad_norm": 0.3634089522289552,
-      "learning_rate": 1.4724911637672341e-05,
-      "loss": 0.3267,
+      "epoch": 1.2328867952303841,
+      "grad_norm": 0.3718668057674085,
+      "learning_rate": 1.4665130689950796e-05,
+      "loss": 0.3389,
       "step": 8375
     },
     {
-      "epoch": 1.227838827838828,
-      "grad_norm": 0.36006484573969755,
-      "learning_rate": 1.4717397041759167e-05,
-      "loss": 0.355,
+      "epoch": 1.2336228470484323,
+      "grad_norm": 0.3720584699709386,
+      "learning_rate": 1.4657553646985134e-05,
+      "loss": 0.3298,
       "step": 8380
     },
     {
-      "epoch": 1.2285714285714286,
-      "grad_norm": 0.3758528148797375,
-      "learning_rate": 1.4709879017908559e-05,
-      "loss": 0.3235,
+      "epoch": 1.2343588988664802,
+      "grad_norm": 0.3600181242934194,
+      "learning_rate": 1.4649973187754314e-05,
+      "loss": 0.3231,
       "step": 8385
     },
     {
-      "epoch": 1.2293040293040294,
-      "grad_norm": 0.3640560499525934,
-      "learning_rate": 1.470235757158356e-05,
-      "loss": 0.3244,
+      "epoch": 1.2350949506845281,
+      "grad_norm": 0.3660149728251252,
+      "learning_rate": 1.464238931781852e-05,
+      "loss": 0.3259,
       "step": 8390
     },
     {
-      "epoch": 1.23003663003663,
-      "grad_norm": 0.3718633192806409,
-      "learning_rate": 1.4694832708249685e-05,
-      "loss": 0.3085,
+      "epoch": 1.2358310025025763,
+      "grad_norm": 0.35515569774446815,
+      "learning_rate": 1.4634802042740439e-05,
+      "loss": 0.3242,
       "step": 8395
     },
     {
-      "epoch": 1.2307692307692308,
-      "grad_norm": 0.3832121158376519,
-      "learning_rate": 1.4687304433374948e-05,
-      "loss": 0.3344,
+      "epoch": 1.2365670543206242,
+      "grad_norm": 0.3581377400151042,
+      "learning_rate": 1.4627211368085255e-05,
+      "loss": 0.3215,
       "step": 8400
     },
     {
-      "epoch": 1.2315018315018316,
-      "grad_norm": 0.3607369656228018,
-      "learning_rate": 1.4679772752429833e-05,
-      "loss": 0.3209,
+      "epoch": 1.2373031061386721,
+      "grad_norm": 0.366841625067584,
+      "learning_rate": 1.4619617299420646e-05,
+      "loss": 0.328,
       "step": 8405
     },
     {
-      "epoch": 1.2322344322344323,
-      "grad_norm": 0.3752129417665394,
-      "learning_rate": 1.4672237670887304e-05,
-      "loss": 0.332,
+      "epoch": 1.2380391579567203,
+      "grad_norm": 0.3639641969747141,
+      "learning_rate": 1.4612019842316778e-05,
+      "loss": 0.3361,
       "step": 8410
     },
     {
-      "epoch": 1.232967032967033,
-      "grad_norm": 0.3872267779943652,
-      "learning_rate": 1.4664699194222787e-05,
-      "loss": 0.3402,
+      "epoch": 1.2387752097747682,
+      "grad_norm": 0.35529413346033073,
+      "learning_rate": 1.4604419002346306e-05,
+      "loss": 0.3368,
       "step": 8415
     },
     {
-      "epoch": 1.2336996336996338,
-      "grad_norm": 0.3690844057598334,
-      "learning_rate": 1.4657157327914189e-05,
-      "loss": 0.3007,
+      "epoch": 1.239511261592816,
+      "grad_norm": 0.35635443827472735,
+      "learning_rate": 1.4596814785084361e-05,
+      "loss": 0.3309,
       "step": 8420
     },
     {
-      "epoch": 1.2344322344322345,
-      "grad_norm": 0.35874344873262004,
-      "learning_rate": 1.4649612077441867e-05,
-      "loss": 0.3219,
+      "epoch": 1.240247313410864,
+      "grad_norm": 0.35017175827485586,
+      "learning_rate": 1.4589207196108556e-05,
+      "loss": 0.3475,
       "step": 8425
     },
     {
-      "epoch": 1.2351648351648352,
-      "grad_norm": 0.36923537952671387,
-      "learning_rate": 1.4642063448288644e-05,
-      "loss": 0.3264,
+      "epoch": 1.2409833652289122,
+      "grad_norm": 0.3796049352713231,
+      "learning_rate": 1.4581596240998975e-05,
+      "loss": 0.3198,
       "step": 8430
     },
     {
-      "epoch": 1.235897435897436,
-      "grad_norm": 0.3652820014861731,
-      "learning_rate": 1.4634511445939796e-05,
-      "loss": 0.3326,
+      "epoch": 1.24171941704696,
+      "grad_norm": 0.3959085195621158,
+      "learning_rate": 1.457398192533817e-05,
+      "loss": 0.3236,
       "step": 8435
     },
     {
-      "epoch": 1.2366300366300367,
-      "grad_norm": 0.39221916060357304,
-      "learning_rate": 1.4626956075883056e-05,
-      "loss": 0.3139,
+      "epoch": 1.242455468865008,
+      "grad_norm": 0.36997520474040174,
+      "learning_rate": 1.4566364254711159e-05,
+      "loss": 0.3311,
       "step": 8440
     },
     {
-      "epoch": 1.2373626373626374,
-      "grad_norm": 0.3732286402178842,
-      "learning_rate": 1.4619397343608593e-05,
-      "loss": 0.3387,
+      "epoch": 1.2431915206830562,
+      "grad_norm": 0.36217574393869123,
+      "learning_rate": 1.4558743234705421e-05,
+      "loss": 0.3042,
       "step": 8445
     },
     {
-      "epoch": 1.2380952380952381,
-      "grad_norm": 0.37948437560162124,
-      "learning_rate": 1.4611835254609026e-05,
-      "loss": 0.332,
+      "epoch": 1.243927572501104,
+      "grad_norm": 0.36238317738922754,
+      "learning_rate": 1.4551118870910892e-05,
+      "loss": 0.324,
       "step": 8450
     },
     {
-      "epoch": 1.2388278388278389,
-      "grad_norm": 0.3759087075387545,
-      "learning_rate": 1.4604269814379417e-05,
-      "loss": 0.3246,
+      "epoch": 1.244663624319152,
+      "grad_norm": 0.3815961739715229,
+      "learning_rate": 1.4543491168919962e-05,
+      "loss": 0.3424,
       "step": 8455
     },
     {
-      "epoch": 1.2395604395604396,
-      "grad_norm": 0.3904251539355884,
-      "learning_rate": 1.4596701028417252e-05,
-      "loss": 0.347,
+      "epoch": 1.2453996761372002,
+      "grad_norm": 0.36245930207044025,
+      "learning_rate": 1.4535860134327463e-05,
+      "loss": 0.3287,
       "step": 8460
     },
     {
-      "epoch": 1.2402930402930403,
-      "grad_norm": 0.36065774214941765,
-      "learning_rate": 1.4589128902222462e-05,
-      "loss": 0.3233,
+      "epoch": 1.246135727955248,
+      "grad_norm": 0.3806379970593242,
+      "learning_rate": 1.452822577273068e-05,
+      "loss": 0.3449,
       "step": 8465
     },
     {
-      "epoch": 1.241025641025641,
-      "grad_norm": 0.407025265298805,
-      "learning_rate": 1.4581553441297395e-05,
-      "loss": 0.3241,
+      "epoch": 1.246871779773296,
+      "grad_norm": 0.38549050902527776,
+      "learning_rate": 1.4520588089729334e-05,
+      "loss": 0.3446,
       "step": 8470
     },
     {
-      "epoch": 1.2417582417582418,
-      "grad_norm": 0.3788864713097616,
-      "learning_rate": 1.4573974651146826e-05,
-      "loss": 0.3165,
+      "epoch": 1.2476078315913441,
+      "grad_norm": 0.35650246505780847,
+      "learning_rate": 1.4512947090925588e-05,
+      "loss": 0.3251,
       "step": 8475
     },
     {
-      "epoch": 1.2424908424908425,
-      "grad_norm": 0.36013901165566703,
-      "learning_rate": 1.4566392537277943e-05,
-      "loss": 0.316,
+      "epoch": 1.248343883409392,
+      "grad_norm": 0.37933112214965986,
+      "learning_rate": 1.4505302781924022e-05,
+      "loss": 0.3443,
       "step": 8480
     },
     {
-      "epoch": 1.2432234432234432,
-      "grad_norm": 0.36320193457939876,
-      "learning_rate": 1.4558807105200363e-05,
-      "loss": 0.3483,
+      "epoch": 1.24907993522744,
+      "grad_norm": 0.3660288041654388,
+      "learning_rate": 1.449765516833166e-05,
+      "loss": 0.3387,
       "step": 8485
     },
     {
-      "epoch": 1.243956043956044,
-      "grad_norm": 0.35449035301865955,
-      "learning_rate": 1.4551218360426098e-05,
-      "loss": 0.3229,
+      "epoch": 1.249815987045488,
+      "grad_norm": 0.3742166905347593,
+      "learning_rate": 1.4490004255757942e-05,
+      "loss": 0.3279,
       "step": 8490
     },
     {
-      "epoch": 1.2446886446886447,
-      "grad_norm": 0.36942036637958453,
-      "learning_rate": 1.4543626308469582e-05,
-      "loss": 0.3167,
+      "epoch": 1.250552038863536,
+      "grad_norm": 0.36730577795557384,
+      "learning_rate": 1.4482350049814726e-05,
+      "loss": 0.3226,
       "step": 8495
     },
     {
-      "epoch": 1.2454212454212454,
-      "grad_norm": 0.372517896904522,
-      "learning_rate": 1.4536030954847644e-05,
-      "loss": 0.3095,
+      "epoch": 1.251288090681584,
+      "grad_norm": 0.3781036384265569,
+      "learning_rate": 1.4474692556116294e-05,
+      "loss": 0.33,
       "step": 8500
     },
     {
-      "epoch": 1.2461538461538462,
-      "grad_norm": 0.38610477964467854,
-      "learning_rate": 1.4528432305079511e-05,
-      "loss": 0.3331,
+      "epoch": 1.252024142499632,
+      "grad_norm": 0.36531158977157197,
+      "learning_rate": 1.4467031780279336e-05,
+      "loss": 0.3295,
       "step": 8505
     },
     {
-      "epoch": 1.246886446886447,
-      "grad_norm": 0.3657326446656885,
-      "learning_rate": 1.4520830364686809e-05,
-      "loss": 0.3253,
+      "epoch": 1.25276019431768,
+      "grad_norm": 0.35942077111844317,
+      "learning_rate": 1.4459367727922944e-05,
+      "loss": 0.3392,
       "step": 8510
     },
     {
-      "epoch": 1.2476190476190476,
-      "grad_norm": 0.3709210974612211,
-      "learning_rate": 1.4513225139193555e-05,
-      "loss": 0.3195,
+      "epoch": 1.253496246135728,
+      "grad_norm": 0.36649771551562454,
+      "learning_rate": 1.4451700404668621e-05,
+      "loss": 0.3293,
       "step": 8515
     },
     {
-      "epoch": 1.2483516483516484,
-      "grad_norm": 0.3671791379170008,
-      "learning_rate": 1.450561663412615e-05,
-      "loss": 0.3223,
+      "epoch": 1.2542322979537759,
+      "grad_norm": 0.3763748103335829,
+      "learning_rate": 1.4444029816140263e-05,
+      "loss": 0.3243,
       "step": 8520
     },
     {
-      "epoch": 1.249084249084249,
-      "grad_norm": 0.38509386166245346,
-      "learning_rate": 1.4498004855013382e-05,
-      "loss": 0.3476,
+      "epoch": 1.2549683497718238,
+      "grad_norm": 0.36270232876998154,
+      "learning_rate": 1.4436355967964164e-05,
+      "loss": 0.3322,
       "step": 8525
     },
     {
-      "epoch": 1.2498168498168498,
-      "grad_norm": 0.367727118151911,
-      "learning_rate": 1.449038980738641e-05,
-      "loss": 0.3339,
+      "epoch": 1.255704401589872,
+      "grad_norm": 0.36816829320081507,
+      "learning_rate": 1.4428678865769005e-05,
+      "loss": 0.3274,
       "step": 8530
     },
     {
-      "epoch": 1.2505494505494505,
-      "grad_norm": 0.3544379472629194,
-      "learning_rate": 1.4482771496778786e-05,
-      "loss": 0.3196,
+      "epoch": 1.2564404534079199,
+      "grad_norm": 0.38531463244299674,
+      "learning_rate": 1.4420998515185866e-05,
+      "loss": 0.3325,
       "step": 8535
     },
     {
-      "epoch": 1.2512820512820513,
-      "grad_norm": 0.37169888332658535,
-      "learning_rate": 1.4475149928726412e-05,
-      "loss": 0.3275,
+      "epoch": 1.257176505225968,
+      "grad_norm": 0.3484703231115639,
+      "learning_rate": 1.4413314921848196e-05,
+      "loss": 0.3195,
       "step": 8540
     },
     {
-      "epoch": 1.252014652014652,
-      "grad_norm": 0.37627880630537924,
-      "learning_rate": 1.446752510876757e-05,
-      "loss": 0.3251,
+      "epoch": 1.257912557044016,
+      "grad_norm": 0.3592125180448874,
+      "learning_rate": 1.4405628091391826e-05,
+      "loss": 0.3305,
       "step": 8545
     },
     {
-      "epoch": 1.2527472527472527,
-      "grad_norm": 0.36691119104223485,
-      "learning_rate": 1.4459897042442899e-05,
-      "loss": 0.3282,
+      "epoch": 1.2586486088620639,
+      "grad_norm": 0.35839328417008115,
+      "learning_rate": 1.4397938029454963e-05,
+      "loss": 0.3451,
       "step": 8550
     },
     {
-      "epoch": 1.2534798534798535,
-      "grad_norm": 0.36354308528144275,
-      "learning_rate": 1.4452265735295403e-05,
-      "loss": 0.335,
+      "epoch": 1.2593846606801118,
+      "grad_norm": 0.351399175552356,
+      "learning_rate": 1.4390244741678188e-05,
+      "loss": 0.3192,
       "step": 8555
     },
     {
-      "epoch": 1.2542124542124542,
-      "grad_norm": 0.3615088687926003,
-      "learning_rate": 1.4444631192870432e-05,
-      "loss": 0.3288,
+      "epoch": 1.26012071249816,
+      "grad_norm": 0.36756762895463796,
+      "learning_rate": 1.4382548233704442e-05,
+      "loss": 0.3254,
       "step": 8560
     },
     {
-      "epoch": 1.254945054945055,
-      "grad_norm": 0.355044070495033,
-      "learning_rate": 1.4436993420715703e-05,
-      "loss": 0.3175,
+      "epoch": 1.2608567643162079,
+      "grad_norm": 0.3674226336001027,
+      "learning_rate": 1.437484851117903e-05,
+      "loss": 0.3282,
       "step": 8565
     },
     {
-      "epoch": 1.2556776556776557,
-      "grad_norm": 0.3607423844814643,
-      "learning_rate": 1.442935242438126e-05,
-      "loss": 0.3223,
+      "epoch": 1.2615928161342558,
+      "grad_norm": 0.3685558855603122,
+      "learning_rate": 1.4367145579749613e-05,
+      "loss": 0.3405,
       "step": 8570
     },
     {
-      "epoch": 1.2564102564102564,
-      "grad_norm": 0.3724667809650596,
-      "learning_rate": 1.44217082094195e-05,
-      "loss": 0.3366,
+      "epoch": 1.262328867952304,
+      "grad_norm": 0.3926982981546724,
+      "learning_rate": 1.435943944506621e-05,
+      "loss": 0.3424,
       "step": 8575
     },
     {
-      "epoch": 1.2571428571428571,
-      "grad_norm": 0.36369061081177334,
-      "learning_rate": 1.441406078138516e-05,
-      "loss": 0.3297,
+      "epoch": 1.2630649197703518,
+      "grad_norm": 0.3582756586857344,
+      "learning_rate": 1.4351730112781187e-05,
+      "loss": 0.32,
       "step": 8580
     },
     {
-      "epoch": 1.2578754578754578,
-      "grad_norm": 0.37023878295575663,
-      "learning_rate": 1.4406410145835308e-05,
-      "loss": 0.311,
+      "epoch": 1.2638009715883998,
+      "grad_norm": 0.35466591104612183,
+      "learning_rate": 1.434401758854926e-05,
+      "loss": 0.3194,
       "step": 8585
     },
     {
-      "epoch": 1.2586080586080586,
-      "grad_norm": 0.3628455438920992,
-      "learning_rate": 1.4398756308329349e-05,
-      "loss": 0.3337,
+      "epoch": 1.2645370234064477,
+      "grad_norm": 0.35817468850934986,
+      "learning_rate": 1.4336301878027477e-05,
+      "loss": 0.3099,
       "step": 8590
     },
     {
-      "epoch": 1.2593406593406593,
-      "grad_norm": 0.36990808978523027,
-      "learning_rate": 1.4391099274429006e-05,
-      "loss": 0.3281,
+      "epoch": 1.2652730752244958,
+      "grad_norm": 0.36091388941579955,
+      "learning_rate": 1.4328582986875225e-05,
+      "loss": 0.3295,
       "step": 8595
     },
     {
-      "epoch": 1.26007326007326,
-      "grad_norm": 0.35199942728347305,
-      "learning_rate": 1.4383439049698332e-05,
-      "loss": 0.3344,
+      "epoch": 1.2660091270425438,
+      "grad_norm": 0.36965641039069697,
+      "learning_rate": 1.4320860920754234e-05,
+      "loss": 0.3168,
       "step": 8600
     },
     {
-      "epoch": 1.2608058608058608,
-      "grad_norm": 0.35389556178404175,
-      "learning_rate": 1.4375775639703694e-05,
-      "loss": 0.3224,
+      "epoch": 1.266745178860592,
+      "grad_norm": 0.354656763532126,
+      "learning_rate": 1.4313135685328555e-05,
+      "loss": 0.3386,
       "step": 8605
     },
     {
-      "epoch": 1.2615384615384615,
-      "grad_norm": 0.36957955875618836,
-      "learning_rate": 1.4368109050013774e-05,
-      "loss": 0.33,
+      "epoch": 1.2674812306786398,
+      "grad_norm": 0.3554833547625228,
+      "learning_rate": 1.4305407286264559e-05,
+      "loss": 0.3286,
       "step": 8610
     },
     {
-      "epoch": 1.2622710622710622,
-      "grad_norm": 0.3596168698950371,
-      "learning_rate": 1.4360439286199567e-05,
-      "loss": 0.3256,
+      "epoch": 1.2682172824966877,
+      "grad_norm": 0.3745807752955344,
+      "learning_rate": 1.429767572923095e-05,
+      "loss": 0.3295,
       "step": 8615
     },
     {
-      "epoch": 1.263003663003663,
-      "grad_norm": 0.3519913059085024,
-      "learning_rate": 1.4352766353834376e-05,
-      "loss": 0.3305,
+      "epoch": 1.2689533343147357,
+      "grad_norm": 0.3640930985321038,
+      "learning_rate": 1.4289941019898734e-05,
+      "loss": 0.328,
       "step": 8620
     },
     {
-      "epoch": 1.2637362637362637,
-      "grad_norm": 0.36741257893638307,
-      "learning_rate": 1.4345090258493797e-05,
-      "loss": 0.33,
+      "epoch": 1.2696893861327838,
+      "grad_norm": 0.35928982508334845,
+      "learning_rate": 1.4282203163941243e-05,
+      "loss": 0.3262,
       "step": 8625
     },
     {
-      "epoch": 1.2644688644688644,
-      "grad_norm": 0.3677419887989739,
-      "learning_rate": 1.433741100575574e-05,
-      "loss": 0.3334,
+      "epoch": 1.2704254379508317,
+      "grad_norm": 0.3540992145513367,
+      "learning_rate": 1.4274462167034107e-05,
+      "loss": 0.318,
       "step": 8630
     },
     {
-      "epoch": 1.2652014652014651,
-      "grad_norm": 0.35636100129085274,
-      "learning_rate": 1.4329728601200394e-05,
-      "loss": 0.3172,
+      "epoch": 1.2711614897688797,
+      "grad_norm": 0.3780018794421239,
+      "learning_rate": 1.4266718034855263e-05,
+      "loss": 0.3442,
       "step": 8635
     },
     {
-      "epoch": 1.2659340659340659,
-      "grad_norm": 0.3539343371341077,
-      "learning_rate": 1.4322043050410248e-05,
-      "loss": 0.3064,
+      "epoch": 1.2718975415869278,
+      "grad_norm": 0.3837064881062433,
+      "learning_rate": 1.4258970773084953e-05,
+      "loss": 0.341,
       "step": 8640
     },
     {
-      "epoch": 1.2666666666666666,
-      "grad_norm": 0.37701181028549535,
-      "learning_rate": 1.431435435897007e-05,
-      "loss": 0.3214,
+      "epoch": 1.2726335934049757,
+      "grad_norm": 0.37181236065817,
+      "learning_rate": 1.4251220387405704e-05,
+      "loss": 0.3239,
       "step": 8645
     },
     {
-      "epoch": 1.2673992673992673,
-      "grad_norm": 0.36302656770740904,
-      "learning_rate": 1.4306662532466922e-05,
-      "loss": 0.3258,
+      "epoch": 1.2733696452230236,
+      "grad_norm": 0.37514772793321255,
+      "learning_rate": 1.4243466883502344e-05,
+      "loss": 0.3308,
       "step": 8650
     },
     {
-      "epoch": 1.268131868131868,
-      "grad_norm": 0.36850903779226113,
-      "learning_rate": 1.4298967576490129e-05,
-      "loss": 0.3242,
+      "epoch": 1.2741056970410716,
+      "grad_norm": 0.3888885055005383,
+      "learning_rate": 1.4235710267061982e-05,
+      "loss": 0.3354,
       "step": 8655
     },
     {
-      "epoch": 1.2688644688644688,
-      "grad_norm": 0.34623155368688263,
-      "learning_rate": 1.4291269496631302e-05,
-      "loss": 0.3181,
+      "epoch": 1.2748417488591197,
+      "grad_norm": 0.3608557831671145,
+      "learning_rate": 1.422795054377401e-05,
+      "loss": 0.3271,
       "step": 8660
     },
     {
-      "epoch": 1.2695970695970695,
-      "grad_norm": 0.35643796388726406,
-      "learning_rate": 1.4283568298484311e-05,
-      "loss": 0.3291,
+      "epoch": 1.2755778006771676,
+      "grad_norm": 0.3771696834669466,
+      "learning_rate": 1.42201877193301e-05,
+      "loss": 0.3255,
       "step": 8665
     },
     {
-      "epoch": 1.2703296703296703,
-      "grad_norm": 0.3596620954791388,
-      "learning_rate": 1.4275863987645306e-05,
-      "loss": 0.3382,
+      "epoch": 1.2763138524952158,
+      "grad_norm": 0.37443360147557536,
+      "learning_rate": 1.4212421799424209e-05,
+      "loss": 0.32,
       "step": 8670
     },
     {
-      "epoch": 1.271062271062271,
-      "grad_norm": 0.357891218900445,
-      "learning_rate": 1.4268156569712688e-05,
-      "loss": 0.3308,
+      "epoch": 1.2770499043132637,
+      "grad_norm": 0.3675708947898225,
+      "learning_rate": 1.420465278975255e-05,
+      "loss": 0.3441,
       "step": 8675
     },
     {
-      "epoch": 1.2717948717948717,
-      "grad_norm": 0.4040598027313928,
-      "learning_rate": 1.426044605028712e-05,
-      "loss": 0.2991,
+      "epoch": 1.2777859561313116,
+      "grad_norm": 0.3838957751072269,
+      "learning_rate": 1.4196880696013601e-05,
+      "loss": 0.3241,
       "step": 8680
     },
     {
-      "epoch": 1.2725274725274724,
-      "grad_norm": 0.38530342372185855,
-      "learning_rate": 1.4252732434971513e-05,
-      "loss": 0.3206,
+      "epoch": 1.2785220079493596,
+      "grad_norm": 0.3779740176842661,
+      "learning_rate": 1.4189105523908118e-05,
+      "loss": 0.314,
       "step": 8685
     },
     {
-      "epoch": 1.2732600732600732,
-      "grad_norm": 0.36613327828340003,
-      "learning_rate": 1.424501572937104e-05,
-      "loss": 0.3388,
+      "epoch": 1.2792580597674077,
+      "grad_norm": 0.3572228985114576,
+      "learning_rate": 1.4181327279139101e-05,
+      "loss": 0.3325,
       "step": 8690
     },
     {
-      "epoch": 1.273992673992674,
-      "grad_norm": 0.39254721011390975,
-      "learning_rate": 1.4237295939093107e-05,
-      "loss": 0.3247,
+      "epoch": 1.2799941115854556,
+      "grad_norm": 0.35543397823294043,
+      "learning_rate": 1.4173545967411813e-05,
+      "loss": 0.329,
       "step": 8695
     },
     {
-      "epoch": 1.2747252747252746,
-      "grad_norm": 0.39593210630206305,
-      "learning_rate": 1.4229573069747369e-05,
-      "loss": 0.322,
+      "epoch": 1.2807301634035035,
+      "grad_norm": 0.3520570139100687,
+      "learning_rate": 1.416576159443375e-05,
+      "loss": 0.3315,
       "step": 8700
     },
     {
-      "epoch": 1.2754578754578754,
-      "grad_norm": 0.387086930088111,
-      "learning_rate": 1.4221847126945712e-05,
-      "loss": 0.3408,
+      "epoch": 1.2814662152215517,
+      "grad_norm": 0.3646897986733931,
+      "learning_rate": 1.4157974165914679e-05,
+      "loss": 0.3412,
       "step": 8705
     },
     {
-      "epoch": 1.276190476190476,
-      "grad_norm": 0.3651049771678776,
-      "learning_rate": 1.4214118116302263e-05,
-      "loss": 0.3195,
+      "epoch": 1.2822022670395996,
+      "grad_norm": 0.38837148575650376,
+      "learning_rate": 1.4150183687566585e-05,
+      "loss": 0.3327,
       "step": 8710
     },
     {
-      "epoch": 1.2769230769230768,
-      "grad_norm": 0.35826044372401333,
-      "learning_rate": 1.4206386043433373e-05,
-      "loss": 0.3279,
+      "epoch": 1.2829383188576475,
+      "grad_norm": 0.3576885191598486,
+      "learning_rate": 1.4142390165103706e-05,
+      "loss": 0.3433,
       "step": 8715
     },
     {
-      "epoch": 1.2776556776556776,
-      "grad_norm": 0.388816510302223,
-      "learning_rate": 1.4198650913957618e-05,
-      "loss": 0.3298,
+      "epoch": 1.2836743706756955,
+      "grad_norm": 0.3650856510045297,
+      "learning_rate": 1.41345936042425e-05,
+      "loss": 0.3211,
       "step": 8720
     },
     {
-      "epoch": 1.2783882783882783,
-      "grad_norm": 0.37877951613726674,
-      "learning_rate": 1.4190912733495803e-05,
-      "loss": 0.3303,
+      "epoch": 1.2844104224937436,
+      "grad_norm": 0.36091953357015577,
+      "learning_rate": 1.4126794010701667e-05,
+      "loss": 0.3187,
       "step": 8725
     },
     {
-      "epoch": 1.279120879120879,
-      "grad_norm": 0.3533529400655342,
-      "learning_rate": 1.418317150767094e-05,
-      "loss": 0.3466,
+      "epoch": 1.2851464743117915,
+      "grad_norm": 0.38524623893423865,
+      "learning_rate": 1.4118991390202121e-05,
+      "loss": 0.3332,
       "step": 8730
     },
     {
-      "epoch": 1.2798534798534797,
-      "grad_norm": 0.36787801088700656,
-      "learning_rate": 1.4175427242108254e-05,
-      "loss": 0.3423,
+      "epoch": 1.2858825261298397,
+      "grad_norm": 0.36112224267826737,
+      "learning_rate": 1.4111185748466998e-05,
+      "loss": 0.3143,
       "step": 8735
     },
     {
-      "epoch": 1.2805860805860805,
-      "grad_norm": 0.36406972417139055,
-      "learning_rate": 1.416767994243519e-05,
-      "loss": 0.3317,
+      "epoch": 1.2866185779478876,
+      "grad_norm": 0.3662161681057587,
+      "learning_rate": 1.4103377091221652e-05,
+      "loss": 0.3367,
       "step": 8740
     },
     {
-      "epoch": 1.2813186813186812,
-      "grad_norm": 0.3792476507833185,
-      "learning_rate": 1.4159929614281388e-05,
-      "loss": 0.3274,
+      "epoch": 1.2873546297659355,
+      "grad_norm": 0.36892244558442167,
+      "learning_rate": 1.4095565424193653e-05,
+      "loss": 0.3219,
       "step": 8745
     },
     {
-      "epoch": 1.282051282051282,
-      "grad_norm": 0.37446305544770797,
-      "learning_rate": 1.4152176263278685e-05,
-      "loss": 0.3495,
+      "epoch": 1.2880906815839834,
+      "grad_norm": 0.36791046504615804,
+      "learning_rate": 1.4087750753112768e-05,
+      "loss": 0.3265,
       "step": 8750
     },
     {
-      "epoch": 1.2827838827838827,
-      "grad_norm": 0.35389577373111636,
-      "learning_rate": 1.4144419895061128e-05,
-      "loss": 0.3257,
+      "epoch": 1.2888267334020316,
+      "grad_norm": 0.3758473221858236,
+      "learning_rate": 1.4079933083710978e-05,
+      "loss": 0.3285,
       "step": 8755
     },
     {
-      "epoch": 1.2835164835164834,
-      "grad_norm": 0.372566504995886,
-      "learning_rate": 1.4136660515264948e-05,
-      "loss": 0.3322,
+      "epoch": 1.2895627852200795,
+      "grad_norm": 0.36744876180300123,
+      "learning_rate": 1.4072112421722458e-05,
+      "loss": 0.3338,
       "step": 8760
     },
     {
-      "epoch": 1.2842490842490841,
-      "grad_norm": 0.3671150331560627,
-      "learning_rate": 1.4128898129528564e-05,
-      "loss": 0.3229,
+      "epoch": 1.2902988370381274,
+      "grad_norm": 0.35555498644135036,
+      "learning_rate": 1.4064288772883572e-05,
+      "loss": 0.3304,
       "step": 8765
     },
     {
-      "epoch": 1.2849816849816849,
-      "grad_norm": 0.36495466749219146,
-      "learning_rate": 1.4121132743492579e-05,
-      "loss": 0.3362,
+      "epoch": 1.2910348888561756,
+      "grad_norm": 0.34417999944016525,
+      "learning_rate": 1.4056462142932894e-05,
+      "loss": 0.3268,
       "step": 8770
     },
     {
-      "epoch": 1.2857142857142856,
-      "grad_norm": 0.3757007917022563,
-      "learning_rate": 1.4113364362799779e-05,
-      "loss": 0.3197,
+      "epoch": 1.2917709406742235,
+      "grad_norm": 0.361373389659639,
+      "learning_rate": 1.4048632537611165e-05,
+      "loss": 0.3108,
       "step": 8775
     },
     {
-      "epoch": 1.2864468864468863,
-      "grad_norm": 0.35545541798657093,
-      "learning_rate": 1.410559299309513e-05,
-      "loss": 0.3198,
+      "epoch": 1.2925069924922714,
+      "grad_norm": 0.35328174056540423,
+      "learning_rate": 1.4040799962661318e-05,
+      "loss": 0.3265,
       "step": 8780
     },
     {
-      "epoch": 1.287179487179487,
-      "grad_norm": 0.3601142341210226,
-      "learning_rate": 1.4097818640025762e-05,
-      "loss": 0.3268,
+      "epoch": 1.2932430443103193,
+      "grad_norm": 0.35560284177991747,
+      "learning_rate": 1.4032964423828457e-05,
+      "loss": 0.3246,
       "step": 8785
     },
     {
-      "epoch": 1.2879120879120878,
-      "grad_norm": 0.3794549038591166,
-      "learning_rate": 1.4090041309240975e-05,
-      "loss": 0.3386,
+      "epoch": 1.2939790961283675,
+      "grad_norm": 0.3762197826941611,
+      "learning_rate": 1.4025125926859872e-05,
+      "loss": 0.3357,
       "step": 8790
     },
     {
-      "epoch": 1.2886446886446885,
-      "grad_norm": 0.38102768293395706,
-      "learning_rate": 1.408226100639223e-05,
-      "loss": 0.3239,
+      "epoch": 1.2947151479464154,
+      "grad_norm": 0.36908458606723793,
+      "learning_rate": 1.401728447750501e-05,
+      "loss": 0.3417,
       "step": 8795
     },
     {
-      "epoch": 1.2893772893772895,
-      "grad_norm": 0.3491164728930594,
-      "learning_rate": 1.407447773713316e-05,
-      "loss": 0.3404,
+      "epoch": 1.2954511997644635,
+      "grad_norm": 0.34380332096863414,
+      "learning_rate": 1.400944008151549e-05,
+      "loss": 0.3197,
       "step": 8800
     },
     {
-      "epoch": 1.2901098901098902,
-      "grad_norm": 0.3578550888810835,
-      "learning_rate": 1.4066691507119541e-05,
-      "loss": 0.3314,
+      "epoch": 1.2961872515825115,
+      "grad_norm": 0.377858377856986,
+      "learning_rate": 1.4001592744645097e-05,
+      "loss": 0.3259,
       "step": 8805
     },
     {
-      "epoch": 1.290842490842491,
-      "grad_norm": 0.3583668492046273,
-      "learning_rate": 1.40589023220093e-05,
-      "loss": 0.3368,
+      "epoch": 1.2969233034005594,
+      "grad_norm": 0.36625931009980567,
+      "learning_rate": 1.3993742472649761e-05,
+      "loss": 0.3387,
       "step": 8810
     },
     {
-      "epoch": 1.2915750915750916,
-      "grad_norm": 0.35734139851501734,
-      "learning_rate": 1.4051110187462524e-05,
-      "loss": 0.3167,
+      "epoch": 1.2976593552186073,
+      "grad_norm": 0.3673278872501545,
+      "learning_rate": 1.3985889271287574e-05,
+      "loss": 0.3286,
       "step": 8815
     },
     {
-      "epoch": 1.2923076923076924,
-      "grad_norm": 0.37017966348980635,
-      "learning_rate": 1.4043315109141431e-05,
-      "loss": 0.3323,
+      "epoch": 1.2983954070366555,
+      "grad_norm": 0.3724532284800494,
+      "learning_rate": 1.3978033146318776e-05,
+      "loss": 0.3301,
       "step": 8820
     },
     {
-      "epoch": 1.293040293040293,
-      "grad_norm": 0.3658193901198154,
-      "learning_rate": 1.4035517092710383e-05,
-      "loss": 0.3218,
+      "epoch": 1.2991314588547034,
+      "grad_norm": 0.3744687861341638,
+      "learning_rate": 1.3970174103505748e-05,
+      "loss": 0.3464,
       "step": 8825
     },
     {
-      "epoch": 1.2937728937728938,
-      "grad_norm": 0.40716340886974844,
-      "learning_rate": 1.4027716143835873e-05,
-      "loss": 0.347,
+      "epoch": 1.2998675106727513,
+      "grad_norm": 0.3519718326489585,
+      "learning_rate": 1.3962312148613011e-05,
+      "loss": 0.3018,
       "step": 8830
     },
     {
-      "epoch": 1.2945054945054946,
-      "grad_norm": 0.37169502716577363,
-      "learning_rate": 1.4019912268186532e-05,
-      "loss": 0.3336,
+      "epoch": 1.3006035624907994,
+      "grad_norm": 0.3830185797720154,
+      "learning_rate": 1.3954447287407225e-05,
+      "loss": 0.3217,
       "step": 8835
     },
     {
-      "epoch": 1.2952380952380953,
-      "grad_norm": 0.36309543152317264,
-      "learning_rate": 1.401210547143311e-05,
-      "loss": 0.3426,
+      "epoch": 1.3013396143088474,
+      "grad_norm": 0.3570941300178156,
+      "learning_rate": 1.394657952565718e-05,
+      "loss": 0.3144,
       "step": 8840
     },
     {
-      "epoch": 1.295970695970696,
-      "grad_norm": 0.35397933043600555,
-      "learning_rate": 1.4004295759248493e-05,
-      "loss": 0.3342,
+      "epoch": 1.3020756661268953,
+      "grad_norm": 0.36288790463875925,
+      "learning_rate": 1.3938708869133795e-05,
+      "loss": 0.3323,
       "step": 8845
     },
     {
-      "epoch": 1.2967032967032968,
-      "grad_norm": 0.3795439470130737,
-      "learning_rate": 1.3996483137307663e-05,
-      "loss": 0.3337,
+      "epoch": 1.3028117179449432,
+      "grad_norm": 0.3447747037380813,
+      "learning_rate": 1.3930835323610113e-05,
+      "loss": 0.3171,
       "step": 8850
     },
     {
-      "epoch": 1.2974358974358975,
-      "grad_norm": 0.36739244392710657,
-      "learning_rate": 1.3988667611287739e-05,
-      "loss": 0.3414,
+      "epoch": 1.3035477697629914,
+      "grad_norm": 0.36034095883531797,
+      "learning_rate": 1.3922958894861292e-05,
+      "loss": 0.3172,
       "step": 8855
     },
     {
-      "epoch": 1.2981684981684982,
-      "grad_norm": 0.3564963815512248,
-      "learning_rate": 1.3980849186867934e-05,
-      "loss": 0.3352,
+      "epoch": 1.3042838215810393,
+      "grad_norm": 0.36541817309380326,
+      "learning_rate": 1.391507958866461e-05,
+      "loss": 0.3348,
       "step": 8860
     },
     {
-      "epoch": 1.298901098901099,
-      "grad_norm": 0.35018651495652114,
-      "learning_rate": 1.3973027869729578e-05,
-      "loss": 0.3348,
+      "epoch": 1.3050198733990874,
+      "grad_norm": 0.375051385256162,
+      "learning_rate": 1.3907197410799444e-05,
+      "loss": 0.3251,
       "step": 8865
     },
     {
-      "epoch": 1.2996336996336997,
-      "grad_norm": 0.3586857206932244,
-      "learning_rate": 1.3965203665556099e-05,
-      "loss": 0.3436,
+      "epoch": 1.3057559252171353,
+      "grad_norm": 0.381012639470291,
+      "learning_rate": 1.38993123670473e-05,
+      "loss": 0.3333,
       "step": 8870
     },
     {
-      "epoch": 1.3003663003663004,
-      "grad_norm": 0.3744151077440032,
-      "learning_rate": 1.3957376580033021e-05,
-      "loss": 0.3396,
+      "epoch": 1.3064919770351833,
+      "grad_norm": 0.3640874685498063,
+      "learning_rate": 1.389142446319176e-05,
+      "loss": 0.3305,
       "step": 8875
     },
     {
-      "epoch": 1.3010989010989011,
-      "grad_norm": 1.0865154890091215,
-      "learning_rate": 1.3949546618847965e-05,
-      "loss": 0.3311,
+      "epoch": 1.3072280288532312,
+      "grad_norm": 0.367532895857759,
+      "learning_rate": 1.3883533705018521e-05,
+      "loss": 0.3279,
       "step": 8880
     },
     {
-      "epoch": 1.3018315018315019,
-      "grad_norm": 0.3602674332465463,
-      "learning_rate": 1.3941713787690638e-05,
-      "loss": 0.3229,
+      "epoch": 1.3079640806712793,
+      "grad_norm": 0.3626074299799139,
+      "learning_rate": 1.3875640098315371e-05,
+      "loss": 0.3171,
       "step": 8885
     },
     {
-      "epoch": 1.3025641025641026,
-      "grad_norm": 0.3680869993512229,
-      "learning_rate": 1.3933878092252839e-05,
-      "loss": 0.3399,
+      "epoch": 1.3087001324893273,
+      "grad_norm": 0.3616593543054651,
+      "learning_rate": 1.386774364887218e-05,
+      "loss": 0.3243,
       "step": 8890
     },
     {
-      "epoch": 1.3032967032967033,
-      "grad_norm": 0.37199644244728475,
-      "learning_rate": 1.3926039538228438e-05,
-      "loss": 0.345,
+      "epoch": 1.3094361843073752,
+      "grad_norm": 0.36257693911461103,
+      "learning_rate": 1.3859844362480912e-05,
+      "loss": 0.3309,
       "step": 8895
     },
     {
-      "epoch": 1.304029304029304,
-      "grad_norm": 0.35505154096683034,
-      "learning_rate": 1.3918198131313391e-05,
-      "loss": 0.3219,
+      "epoch": 1.3101722361254233,
+      "grad_norm": 0.39690486121273244,
+      "learning_rate": 1.3851942244935608e-05,
+      "loss": 0.3272,
       "step": 8900
     },
     {
-      "epoch": 1.3047619047619048,
-      "grad_norm": 0.3732293335189872,
-      "learning_rate": 1.3910353877205722e-05,
-      "loss": 0.3205,
+      "epoch": 1.3109082879434713,
+      "grad_norm": 0.3742390186875793,
+      "learning_rate": 1.3844037302032382e-05,
+      "loss": 0.3293,
       "step": 8905
     },
     {
-      "epoch": 1.3054945054945055,
-      "grad_norm": 0.38105851569610655,
-      "learning_rate": 1.3902506781605533e-05,
-      "loss": 0.3392,
+      "epoch": 1.3116443397615192,
+      "grad_norm": 0.35758003112894904,
+      "learning_rate": 1.3836129539569427e-05,
+      "loss": 0.3298,
       "step": 8910
     },
     {
-      "epoch": 1.3062271062271062,
-      "grad_norm": 0.3694012986495021,
-      "learning_rate": 1.389465685021498e-05,
-      "loss": 0.3426,
+      "epoch": 1.312380391579567,
+      "grad_norm": 0.36120571548511105,
+      "learning_rate": 1.3828218963347002e-05,
+      "loss": 0.3148,
       "step": 8915
     },
     {
-      "epoch": 1.306959706959707,
-      "grad_norm": 0.3661925729167942,
-      "learning_rate": 1.3886804088738282e-05,
-      "loss": 0.331,
+      "epoch": 1.3131164433976152,
+      "grad_norm": 0.35412896216423156,
+      "learning_rate": 1.3820305579167425e-05,
+      "loss": 0.3267,
       "step": 8920
     },
     {
-      "epoch": 1.3076923076923077,
-      "grad_norm": 0.3730823269174414,
-      "learning_rate": 1.3878948502881716e-05,
-      "loss": 0.35,
+      "epoch": 1.3138524952156632,
+      "grad_norm": 0.37285971521264993,
+      "learning_rate": 1.3812389392835082e-05,
+      "loss": 0.3541,
       "step": 8925
     },
     {
-      "epoch": 1.3084249084249084,
-      "grad_norm": 0.3772251825415492,
-      "learning_rate": 1.3871090098353615e-05,
-      "loss": 0.329,
+      "epoch": 1.3145885470337113,
+      "grad_norm": 0.3717384916161553,
+      "learning_rate": 1.3804470410156407e-05,
+      "loss": 0.3186,
       "step": 8930
     },
     {
-      "epoch": 1.3091575091575092,
-      "grad_norm": 0.3601992251208424,
-      "learning_rate": 1.3863228880864352e-05,
-      "loss": 0.3366,
+      "epoch": 1.3153245988517592,
+      "grad_norm": 0.37521720529271974,
+      "learning_rate": 1.3796548636939888e-05,
+      "loss": 0.3264,
       "step": 8935
     },
     {
-      "epoch": 1.30989010989011,
-      "grad_norm": 0.3840853889143167,
-      "learning_rate": 1.3855364856126354e-05,
-      "loss": 0.3307,
+      "epoch": 1.3160606506698072,
+      "grad_norm": 0.3606370679059983,
+      "learning_rate": 1.3788624078996062e-05,
+      "loss": 0.3194,
       "step": 8940
     },
     {
-      "epoch": 1.3106227106227106,
-      "grad_norm": 0.3634546535899789,
-      "learning_rate": 1.3847498029854078e-05,
-      "loss": 0.335,
+      "epoch": 1.316796702487855,
+      "grad_norm": 0.3663504972186256,
+      "learning_rate": 1.3780696742137508e-05,
+      "loss": 0.3326,
       "step": 8945
     },
     {
-      "epoch": 1.3113553113553114,
-      "grad_norm": 0.3605399864380127,
-      "learning_rate": 1.3839628407764026e-05,
-      "loss": 0.3364,
+      "epoch": 1.317532754305903,
+      "grad_norm": 0.3749521221694766,
+      "learning_rate": 1.3772766632178836e-05,
+      "loss": 0.3353,
       "step": 8950
     },
     {
-      "epoch": 1.312087912087912,
-      "grad_norm": 0.3850927201382173,
-      "learning_rate": 1.3831755995574722e-05,
-      "loss": 0.3273,
+      "epoch": 1.3182688061239511,
+      "grad_norm": 0.35160529477528407,
+      "learning_rate": 1.37648337549367e-05,
+      "loss": 0.3247,
       "step": 8955
     },
     {
-      "epoch": 1.3128205128205128,
-      "grad_norm": 0.3633776638654286,
-      "learning_rate": 1.382388079900673e-05,
-      "loss": 0.3249,
+      "epoch": 1.319004857941999,
+      "grad_norm": 0.36493455882005593,
+      "learning_rate": 1.3756898116229778e-05,
+      "loss": 0.329,
       "step": 8960
     },
     {
-      "epoch": 1.3135531135531135,
-      "grad_norm": 0.3630863466409669,
-      "learning_rate": 1.381600282378262e-05,
-      "loss": 0.3146,
+      "epoch": 1.3197409097600472,
+      "grad_norm": 0.3749110524136941,
+      "learning_rate": 1.3748959721878774e-05,
+      "loss": 0.3394,
       "step": 8965
     },
     {
-      "epoch": 1.3142857142857143,
-      "grad_norm": 0.355632113123853,
-      "learning_rate": 1.3808122075626999e-05,
-      "loss": 0.3364,
+      "epoch": 1.3204769615780951,
+      "grad_norm": 0.376136778529708,
+      "learning_rate": 1.3741018577706415e-05,
+      "loss": 0.3217,
       "step": 8970
     },
     {
-      "epoch": 1.315018315018315,
-      "grad_norm": 0.36949929872541576,
-      "learning_rate": 1.3800238560266475e-05,
-      "loss": 0.3282,
+      "epoch": 1.321213013396143,
+      "grad_norm": 0.38402482862451526,
+      "learning_rate": 1.3733074689537446e-05,
+      "loss": 0.3414,
       "step": 8975
     },
     {
-      "epoch": 1.3157509157509157,
-      "grad_norm": 0.38108448533492045,
-      "learning_rate": 1.3792352283429681e-05,
-      "loss": 0.3373,
+      "epoch": 1.321949065214191,
+      "grad_norm": 0.3700605631799661,
+      "learning_rate": 1.3725128063198622e-05,
+      "loss": 0.3171,
       "step": 8980
     },
     {
-      "epoch": 1.3164835164835165,
-      "grad_norm": 0.37695903994833746,
-      "learning_rate": 1.3784463250847239e-05,
-      "loss": 0.3311,
+      "epoch": 1.3226851170322391,
+      "grad_norm": 0.3678747227565391,
+      "learning_rate": 1.371717870451871e-05,
+      "loss": 0.3287,
       "step": 8985
     },
     {
-      "epoch": 1.3172161172161172,
-      "grad_norm": 0.36201406054495167,
-      "learning_rate": 1.3776571468251787e-05,
-      "loss": 0.3401,
+      "epoch": 1.323421168850287,
+      "grad_norm": 0.37230065112792066,
+      "learning_rate": 1.370922661932847e-05,
+      "loss": 0.3289,
       "step": 8990
     },
     {
-      "epoch": 1.317948717948718,
-      "grad_norm": 0.3643729233817631,
-      "learning_rate": 1.3768676941377957e-05,
-      "loss": 0.3294,
+      "epoch": 1.3241572206683352,
+      "grad_norm": 0.3529879764377384,
+      "learning_rate": 1.3701271813460684e-05,
+      "loss": 0.3388,
       "step": 8995
     },
     {
-      "epoch": 1.3186813186813187,
-      "grad_norm": 0.37217102276035735,
-      "learning_rate": 1.3760779675962376e-05,
-      "loss": 0.3316,
+      "epoch": 1.3248932724863831,
+      "grad_norm": 0.36540336257391304,
+      "learning_rate": 1.3693314292750102e-05,
+      "loss": 0.3178,
       "step": 9000
     },
     {
-      "epoch": 1.3194139194139194,
-      "grad_norm": 0.3791660413316379,
-      "learning_rate": 1.3752879677743663e-05,
-      "loss": 0.3131,
+      "epoch": 1.325629324304431,
+      "grad_norm": 0.376472909085647,
+      "learning_rate": 1.3685354063033485e-05,
+      "loss": 0.3236,
       "step": 9005
     },
     {
-      "epoch": 1.3201465201465201,
-      "grad_norm": 0.39211783715886,
-      "learning_rate": 1.3744976952462414e-05,
-      "loss": 0.3268,
+      "epoch": 1.326365376122479,
+      "grad_norm": 0.37850104947330887,
+      "learning_rate": 1.3677391130149578e-05,
+      "loss": 0.3212,
       "step": 9010
     },
     {
-      "epoch": 1.3208791208791208,
-      "grad_norm": 0.36582428340374257,
-      "learning_rate": 1.3737071505861215e-05,
-      "loss": 0.3477,
+      "epoch": 1.3271014279405269,
+      "grad_norm": 0.35916930303255235,
+      "learning_rate": 1.3669425499939105e-05,
+      "loss": 0.3203,
       "step": 9015
     },
     {
-      "epoch": 1.3216117216117216,
-      "grad_norm": 0.37117014585149033,
-      "learning_rate": 1.3729163343684629e-05,
-      "loss": 0.3313,
+      "epoch": 1.327837479758575,
+      "grad_norm": 0.3608480287386314,
+      "learning_rate": 1.3661457178244764e-05,
+      "loss": 0.3224,
       "step": 9020
     },
     {
-      "epoch": 1.3223443223443223,
-      "grad_norm": 0.36382474665096526,
-      "learning_rate": 1.3721252471679191e-05,
-      "loss": 0.3326,
+      "epoch": 1.328573531576623,
+      "grad_norm": 0.35597939890032604,
+      "learning_rate": 1.3653486170911239e-05,
+      "loss": 0.3335,
       "step": 9025
     },
     {
-      "epoch": 1.323076923076923,
-      "grad_norm": 0.3642539317171026,
-      "learning_rate": 1.37133388955934e-05,
-      "loss": 0.3302,
+      "epoch": 1.329309583394671,
+      "grad_norm": 0.3640202335939862,
+      "learning_rate": 1.3645512483785175e-05,
+      "loss": 0.3305,
       "step": 9030
     },
     {
-      "epoch": 1.3238095238095238,
-      "grad_norm": 0.3628206861452227,
-      "learning_rate": 1.370542262117773e-05,
-      "loss": 0.331,
+      "epoch": 1.330045635212719,
+      "grad_norm": 0.3826170418063535,
+      "learning_rate": 1.3637536122715187e-05,
+      "loss": 0.3212,
       "step": 9035
     },
     {
-      "epoch": 1.3245421245421245,
-      "grad_norm": 0.3880906668746477,
-      "learning_rate": 1.3697503654184608e-05,
-      "loss": 0.3444,
+      "epoch": 1.330781687030767,
+      "grad_norm": 0.38003943131416446,
+      "learning_rate": 1.3629557093551846e-05,
+      "loss": 0.3262,
       "step": 9040
     },
     {
-      "epoch": 1.3252747252747252,
-      "grad_norm": 0.3457486717511427,
-      "learning_rate": 1.3689582000368425e-05,
-      "loss": 0.3345,
+      "epoch": 1.3315177388488149,
+      "grad_norm": 0.3635462646251249,
+      "learning_rate": 1.3621575402147686e-05,
+      "loss": 0.3133,
       "step": 9045
     },
     {
-      "epoch": 1.326007326007326,
-      "grad_norm": 0.3657750441820086,
-      "learning_rate": 1.3681657665485508e-05,
-      "loss": 0.3305,
+      "epoch": 1.332253790666863,
+      "grad_norm": 0.3560170171295613,
+      "learning_rate": 1.361359105435719e-05,
+      "loss": 0.3324,
       "step": 9050
     },
     {
-      "epoch": 1.3267399267399267,
-      "grad_norm": 0.3505090638445965,
-      "learning_rate": 1.3673730655294152e-05,
-      "loss": 0.3203,
+      "epoch": 1.332989842484911,
+      "grad_norm": 0.5920590711538712,
+      "learning_rate": 1.3605604056036793e-05,
+      "loss": 0.3718,
       "step": 9055
     },
     {
-      "epoch": 1.3274725274725274,
-      "grad_norm": 0.375871260070138,
-      "learning_rate": 1.3665800975554587e-05,
-      "loss": 0.3115,
+      "epoch": 1.3337258943029588,
+      "grad_norm": 0.7817575272113128,
+      "learning_rate": 1.3597614413044868e-05,
+      "loss": 0.4103,
       "step": 9060
     },
     {
-      "epoch": 1.3282051282051281,
-      "grad_norm": 0.379502320322331,
-      "learning_rate": 1.3657868632028983e-05,
-      "loss": 0.3357,
+      "epoch": 1.334461946121007,
+      "grad_norm": 0.436224414420997,
+      "learning_rate": 1.3589622131241736e-05,
+      "loss": 0.3421,
       "step": 9065
     },
     {
-      "epoch": 1.3289377289377289,
-      "grad_norm": 0.3615803192854987,
-      "learning_rate": 1.3649933630481444e-05,
-      "loss": 0.3264,
+      "epoch": 1.335197997939055,
+      "grad_norm": 0.3691518001975837,
+      "learning_rate": 1.3581627216489644e-05,
+      "loss": 0.3356,
       "step": 9070
     },
     {
-      "epoch": 1.3296703296703296,
-      "grad_norm": 0.3769021122295167,
-      "learning_rate": 1.3641995976678012e-05,
-      "loss": 0.3378,
+      "epoch": 1.3359340497571028,
+      "grad_norm": 0.36678308623805267,
+      "learning_rate": 1.357362967465278e-05,
+      "loss": 0.3184,
       "step": 9075
     },
     {
-      "epoch": 1.3304029304029303,
-      "grad_norm": 0.3624634399641649,
-      "learning_rate": 1.3634055676386644e-05,
-      "loss": 0.3158,
+      "epoch": 1.3366701015751508,
+      "grad_norm": 0.45367278789041315,
+      "learning_rate": 1.3565629511597252e-05,
+      "loss": 0.3295,
       "step": 9080
     },
     {
-      "epoch": 1.331135531135531,
-      "grad_norm": 0.3597843061920691,
-      "learning_rate": 1.3626112735377235e-05,
-      "loss": 0.3404,
+      "epoch": 1.337406153393199,
+      "grad_norm": 0.38150781862087807,
+      "learning_rate": 1.3557626733191094e-05,
+      "loss": 0.3153,
       "step": 9085
     },
     {
-      "epoch": 1.3318681318681318,
-      "grad_norm": 0.3678244039216215,
-      "learning_rate": 1.3618167159421591e-05,
-      "loss": 0.3288,
+      "epoch": 1.3381422052112468,
+      "grad_norm": 0.37214536870557213,
+      "learning_rate": 1.3549621345304256e-05,
+      "loss": 0.3227,
       "step": 9090
     },
     {
-      "epoch": 1.3326007326007325,
-      "grad_norm": 0.40084734783958803,
-      "learning_rate": 1.3610218954293436e-05,
-      "loss": 0.3235,
+      "epoch": 1.338878257029295,
+      "grad_norm": 0.3493050932086383,
+      "learning_rate": 1.3541613353808604e-05,
+      "loss": 0.3348,
       "step": 9095
     },
     {
-      "epoch": 1.3333333333333333,
-      "grad_norm": 0.3724983949020907,
-      "learning_rate": 1.3602268125768392e-05,
-      "loss": 0.3367,
+      "epoch": 1.339614308847343,
+      "grad_norm": 0.3686246268789654,
+      "learning_rate": 1.3533602764577913e-05,
+      "loss": 0.3368,
       "step": 9100
     },
     {
-      "epoch": 1.334065934065934,
-      "grad_norm": 0.36524272902640703,
-      "learning_rate": 1.3594314679624008e-05,
-      "loss": 0.3149,
+      "epoch": 1.3403503606653908,
+      "grad_norm": 0.3778724437437195,
+      "learning_rate": 1.3525589583487861e-05,
+      "loss": 0.3357,
       "step": 9105
     },
     {
-      "epoch": 1.3347985347985347,
-      "grad_norm": 0.3933348954444407,
-      "learning_rate": 1.3586358621639722e-05,
-      "loss": 0.337,
+      "epoch": 1.3410864124834387,
+      "grad_norm": 0.37353179073977216,
+      "learning_rate": 1.3517573816416034e-05,
+      "loss": 0.3239,
       "step": 9110
     },
     {
-      "epoch": 1.3355311355311354,
-      "grad_norm": 0.3566922658589224,
-      "learning_rate": 1.3578399957596869e-05,
-      "loss": 0.3258,
+      "epoch": 1.3418224643014869,
+      "grad_norm": 0.36758473241192974,
+      "learning_rate": 1.3509555469241911e-05,
+      "loss": 0.3281,
       "step": 9115
     },
     {
-      "epoch": 1.3362637362637364,
-      "grad_norm": 0.3622582797251259,
-      "learning_rate": 1.3570438693278682e-05,
-      "loss": 0.3313,
+      "epoch": 1.3425585161195348,
+      "grad_norm": 0.3688353394560033,
+      "learning_rate": 1.3501534547846859e-05,
+      "loss": 0.3133,
       "step": 9120
     },
     {
-      "epoch": 1.3369963369963371,
-      "grad_norm": 0.35355792278152115,
-      "learning_rate": 1.3562474834470285e-05,
-      "loss": 0.3308,
+      "epoch": 1.3432945679375827,
+      "grad_norm": 0.35636360295775227,
+      "learning_rate": 1.349351105811414e-05,
+      "loss": 0.3236,
       "step": 9125
     },
     {
-      "epoch": 1.3377289377289379,
-      "grad_norm": 0.35095601290160816,
-      "learning_rate": 1.3554508386958685e-05,
-      "loss": 0.3187,
+      "epoch": 1.3440306197556309,
+      "grad_norm": 0.36193316658793956,
+      "learning_rate": 1.348548500592889e-05,
+      "loss": 0.331,
       "step": 9130
     },
     {
-      "epoch": 1.3384615384615386,
-      "grad_norm": 0.366020871166583,
-      "learning_rate": 1.354653935653277e-05,
-      "loss": 0.322,
+      "epoch": 1.3447666715736788,
+      "grad_norm": 0.3531798465100575,
+      "learning_rate": 1.3477456397178137e-05,
+      "loss": 0.3144,
       "step": 9135
     },
     {
-      "epoch": 1.3391941391941393,
-      "grad_norm": 0.3727293424181606,
-      "learning_rate": 1.3538567748983303e-05,
-      "loss": 0.3318,
+      "epoch": 1.3455027233917267,
+      "grad_norm": 0.37679106122190187,
+      "learning_rate": 1.346942523775078e-05,
+      "loss": 0.3257,
       "step": 9140
     },
     {
-      "epoch": 1.33992673992674,
-      "grad_norm": 0.36974071866429414,
-      "learning_rate": 1.3530593570102923e-05,
-      "loss": 0.3221,
+      "epoch": 1.3462387752097746,
+      "grad_norm": 0.3784726502374104,
+      "learning_rate": 1.3461391533537586e-05,
+      "loss": 0.3329,
       "step": 9145
     },
     {
-      "epoch": 1.3406593406593408,
-      "grad_norm": 0.35864610045404177,
-      "learning_rate": 1.3522616825686135e-05,
-      "loss": 0.3133,
+      "epoch": 1.3469748270278228,
+      "grad_norm": 0.3630903210951978,
+      "learning_rate": 1.345335529043119e-05,
+      "loss": 0.3258,
       "step": 9150
     },
     {
-      "epoch": 1.3413919413919415,
-      "grad_norm": 0.36740161583874237,
-      "learning_rate": 1.351463752152931e-05,
-      "loss": 0.3236,
+      "epoch": 1.3477108788458707,
+      "grad_norm": 0.3837233711679743,
+      "learning_rate": 1.3445316514326085e-05,
+      "loss": 0.315,
       "step": 9155
     },
     {
-      "epoch": 1.3421245421245422,
-      "grad_norm": 0.36983404972777706,
-      "learning_rate": 1.3506655663430681e-05,
-      "loss": 0.3278,
+      "epoch": 1.3484469306639189,
+      "grad_norm": 0.3803724064684297,
+      "learning_rate": 1.3437275211118635e-05,
+      "loss": 0.3371,
       "step": 9160
     },
     {
-      "epoch": 1.342857142857143,
-      "grad_norm": 0.379522902841167,
-      "learning_rate": 1.3498671257190332e-05,
-      "loss": 0.3387,
+      "epoch": 1.3491829824819668,
+      "grad_norm": 0.3671361885025908,
+      "learning_rate": 1.3429231386707042e-05,
+      "loss": 0.3256,
       "step": 9165
     },
     {
-      "epoch": 1.3435897435897437,
-      "grad_norm": 0.3636575507838711,
-      "learning_rate": 1.34906843086102e-05,
-      "loss": 0.3218,
+      "epoch": 1.3499190343000147,
+      "grad_norm": 0.36363935603638564,
+      "learning_rate": 1.3421185046991364e-05,
+      "loss": 0.3221,
       "step": 9170
     },
     {
-      "epoch": 1.3443223443223444,
-      "grad_norm": 0.35300522843109694,
-      "learning_rate": 1.3482694823494074e-05,
-      "loss": 0.3238,
+      "epoch": 1.3506550861180626,
+      "grad_norm": 0.36065150620607056,
+      "learning_rate": 1.3413136197873504e-05,
+      "loss": 0.3181,
       "step": 9175
     },
     {
-      "epoch": 1.3450549450549452,
-      "grad_norm": 0.3890464743415906,
-      "learning_rate": 1.347470280764758e-05,
-      "loss": 0.3361,
+      "epoch": 1.3513911379361108,
+      "grad_norm": 0.3693935412323531,
+      "learning_rate": 1.3405084845257213e-05,
+      "loss": 0.314,
       "step": 9180
     },
     {
-      "epoch": 1.345787545787546,
-      "grad_norm": 0.3838027933005435,
-      "learning_rate": 1.3466708266878182e-05,
-      "loss": 0.3332,
+      "epoch": 1.3521271897541587,
+      "grad_norm": 0.3626229080305777,
+      "learning_rate": 1.3397030995048061e-05,
+      "loss": 0.3297,
       "step": 9185
     },
     {
-      "epoch": 1.3465201465201466,
-      "grad_norm": 0.36039931461903624,
-      "learning_rate": 1.345871120699519e-05,
-      "loss": 0.3172,
+      "epoch": 1.3528632415722066,
+      "grad_norm": 0.3634563658225217,
+      "learning_rate": 1.3388974653153466e-05,
+      "loss": 0.3266,
       "step": 9190
     },
     {
-      "epoch": 1.3472527472527474,
-      "grad_norm": 0.3548207792166718,
-      "learning_rate": 1.345071163380973e-05,
-      "loss": 0.3329,
+      "epoch": 1.3535992933902548,
+      "grad_norm": 0.37342942136131024,
+      "learning_rate": 1.3380915825482666e-05,
+      "loss": 0.3391,
       "step": 9195
     },
     {
-      "epoch": 1.347985347985348,
-      "grad_norm": 0.3633084594988163,
-      "learning_rate": 1.3442709553134767e-05,
-      "loss": 0.3209,
+      "epoch": 1.3543353452083027,
+      "grad_norm": 0.3563786317240363,
+      "learning_rate": 1.3372854517946725e-05,
+      "loss": 0.32,
       "step": 9200
     },
     {
-      "epoch": 1.3487179487179488,
-      "grad_norm": 0.3650618479253569,
-      "learning_rate": 1.3434704970785077e-05,
-      "loss": 0.3217,
+      "epoch": 1.3550713970263506,
+      "grad_norm": 0.35602494064860596,
+      "learning_rate": 1.3364790736458522e-05,
+      "loss": 0.3241,
       "step": 9205
     },
     {
-      "epoch": 1.3494505494505495,
-      "grad_norm": 0.34552783167904033,
-      "learning_rate": 1.3426697892577263e-05,
-      "loss": 0.3036,
+      "epoch": 1.3558074488443985,
+      "grad_norm": 0.3567453104136102,
+      "learning_rate": 1.3356724486932756e-05,
+      "loss": 0.3354,
       "step": 9210
     },
     {
-      "epoch": 1.3501831501831503,
-      "grad_norm": 0.3620188951071061,
-      "learning_rate": 1.3418688324329737e-05,
-      "loss": 0.3231,
+      "epoch": 1.3565435006624467,
+      "grad_norm": 0.37979383603228695,
+      "learning_rate": 1.334865577528593e-05,
+      "loss": 0.3301,
       "step": 9215
     },
     {
-      "epoch": 1.350915750915751,
-      "grad_norm": 0.3722775578935979,
-      "learning_rate": 1.3410676271862722e-05,
-      "loss": 0.3295,
+      "epoch": 1.3572795524804946,
+      "grad_norm": 0.3766475839970793,
+      "learning_rate": 1.334058460743636e-05,
+      "loss": 0.3336,
       "step": 9220
     },
     {
-      "epoch": 1.3516483516483517,
-      "grad_norm": 0.3643391038118818,
-      "learning_rate": 1.3402661740998237e-05,
-      "loss": 0.3354,
+      "epoch": 1.3580156042985427,
+      "grad_norm": 0.3715167034687806,
+      "learning_rate": 1.333251098930416e-05,
+      "loss": 0.3346,
       "step": 9225
     },
     {
-      "epoch": 1.3523809523809525,
-      "grad_norm": 0.37717278139689026,
-      "learning_rate": 1.3394644737560123e-05,
-      "loss": 0.3183,
+      "epoch": 1.3587516561165907,
+      "grad_norm": 0.362141433485265,
+      "learning_rate": 1.3324434926811242e-05,
+      "loss": 0.3235,
       "step": 9230
     },
     {
-      "epoch": 1.3531135531135532,
-      "grad_norm": 0.3779918764191859,
-      "learning_rate": 1.3386625267373993e-05,
-      "loss": 0.3258,
+      "epoch": 1.3594877079346386,
+      "grad_norm": 0.3688076294470631,
+      "learning_rate": 1.3316356425881306e-05,
+      "loss": 0.3353,
       "step": 9235
     },
     {
-      "epoch": 1.353846153846154,
-      "grad_norm": 0.37789420055359735,
-      "learning_rate": 1.337860333626728e-05,
-      "loss": 0.3274,
+      "epoch": 1.3602237597526865,
+      "grad_norm": 0.36877605896245913,
+      "learning_rate": 1.330827549243985e-05,
+      "loss": 0.3141,
       "step": 9240
     },
     {
-      "epoch": 1.3545787545787547,
-      "grad_norm": 0.3549763157893029,
-      "learning_rate": 1.3370578950069173e-05,
-      "loss": 0.3239,
+      "epoch": 1.3609598115707346,
+      "grad_norm": 0.35566835653531914,
+      "learning_rate": 1.3300192132414153e-05,
+      "loss": 0.314,
       "step": 9245
     },
     {
-      "epoch": 1.3553113553113554,
-      "grad_norm": 0.36645175308764016,
-      "learning_rate": 1.3362552114610671e-05,
-      "loss": 0.3348,
+      "epoch": 1.3616958633887826,
+      "grad_norm": 0.37602890117093807,
+      "learning_rate": 1.3292106351733265e-05,
+      "loss": 0.3301,
       "step": 9250
     },
     {
-      "epoch": 1.3560439560439561,
-      "grad_norm": 0.3701155894863453,
-      "learning_rate": 1.3354522835724542e-05,
-      "loss": 0.3461,
+      "epoch": 1.3624319152068305,
+      "grad_norm": 0.3762513015610662,
+      "learning_rate": 1.3284018156328022e-05,
+      "loss": 0.3357,
       "step": 9255
     },
     {
-      "epoch": 1.3567765567765568,
-      "grad_norm": 0.3730913997528624,
-      "learning_rate": 1.3346491119245331e-05,
-      "loss": 0.3115,
+      "epoch": 1.3631679670248786,
+      "grad_norm": 0.34985999884758706,
+      "learning_rate": 1.3275927552131027e-05,
+      "loss": 0.3286,
       "step": 9260
     },
     {
-      "epoch": 1.3575091575091576,
-      "grad_norm": 0.36728438979467976,
-      "learning_rate": 1.3338456971009352e-05,
-      "loss": 0.3244,
+      "epoch": 1.3639040188429266,
+      "grad_norm": 0.37783002776214625,
+      "learning_rate": 1.3267834545076652e-05,
+      "loss": 0.3355,
       "step": 9265
     },
     {
-      "epoch": 1.3582417582417583,
-      "grad_norm": 0.3647202374966414,
-      "learning_rate": 1.3330420396854694e-05,
-      "loss": 0.3326,
+      "epoch": 1.3646400706609745,
+      "grad_norm": 0.33845063438904066,
+      "learning_rate": 1.3259739141101026e-05,
+      "loss": 0.3282,
       "step": 9270
     },
     {
-      "epoch": 1.358974358974359,
-      "grad_norm": 0.35244098745828395,
-      "learning_rate": 1.3322381402621197e-05,
-      "loss": 0.3205,
+      "epoch": 1.3653761224790224,
+      "grad_norm": 0.3678656537852835,
+      "learning_rate": 1.3251641346142048e-05,
+      "loss": 0.3401,
       "step": 9275
     },
     {
-      "epoch": 1.3597069597069598,
-      "grad_norm": 0.3810067078097531,
-      "learning_rate": 1.3314339994150467e-05,
-      "loss": 0.3205,
+      "epoch": 1.3661121742970705,
+      "grad_norm": 0.3587484044138878,
+      "learning_rate": 1.3243541166139355e-05,
+      "loss": 0.3289,
       "step": 9280
     },
     {
-      "epoch": 1.3604395604395605,
-      "grad_norm": 0.3678887243014611,
-      "learning_rate": 1.3306296177285867e-05,
-      "loss": 0.3361,
+      "epoch": 1.3668482261151185,
+      "grad_norm": 0.3658949912844781,
+      "learning_rate": 1.3235438607034348e-05,
+      "loss": 0.3249,
       "step": 9285
     },
     {
-      "epoch": 1.3611721611721612,
-      "grad_norm": 0.39245251166398054,
-      "learning_rate": 1.3298249957872502e-05,
-      "loss": 0.3306,
+      "epoch": 1.3675842779331666,
+      "grad_norm": 0.35026397400861287,
+      "learning_rate": 1.3227333674770163e-05,
+      "loss": 0.3175,
       "step": 9290
     },
     {
-      "epoch": 1.361904761904762,
-      "grad_norm": 0.36205216028276305,
-      "learning_rate": 1.3290201341757231e-05,
-      "loss": 0.3227,
+      "epoch": 1.3683203297512145,
+      "grad_norm": 0.34018425443285993,
+      "learning_rate": 1.3219226375291678e-05,
+      "loss": 0.3124,
       "step": 9295
     },
     {
-      "epoch": 1.3626373626373627,
-      "grad_norm": 0.39561693455023267,
-      "learning_rate": 1.3282150334788649e-05,
-      "loss": 0.3243,
+      "epoch": 1.3690563815692625,
+      "grad_norm": 0.37839391469181477,
+      "learning_rate": 1.321111671454551e-05,
+      "loss": 0.3414,
       "step": 9300
     },
     {
-      "epoch": 1.3633699633699634,
-      "grad_norm": 0.3678755847393884,
-      "learning_rate": 1.3274096942817086e-05,
-      "loss": 0.3194,
+      "epoch": 1.3697924333873104,
+      "grad_norm": 0.3619436665518319,
+      "learning_rate": 1.320300469848001e-05,
+      "loss": 0.3278,
       "step": 9305
     },
     {
-      "epoch": 1.3641025641025641,
-      "grad_norm": 0.358546203031558,
-      "learning_rate": 1.3266041171694615e-05,
-      "loss": 0.3163,
+      "epoch": 1.3705284852053585,
+      "grad_norm": 0.361850051949316,
+      "learning_rate": 1.3194890333045254e-05,
+      "loss": 0.3095,
       "step": 9310
     },
     {
-      "epoch": 1.3648351648351649,
-      "grad_norm": 0.3652618532577394,
-      "learning_rate": 1.325798302727503e-05,
-      "loss": 0.3113,
+      "epoch": 1.3712645370234064,
+      "grad_norm": 0.353165857016833,
+      "learning_rate": 1.3186773624193036e-05,
+      "loss": 0.3178,
       "step": 9315
     },
     {
-      "epoch": 1.3655677655677656,
-      "grad_norm": 0.35800691736373186,
-      "learning_rate": 1.324992251541385e-05,
-      "loss": 0.3257,
+      "epoch": 1.3720005888414544,
+      "grad_norm": 0.35284171647302653,
+      "learning_rate": 1.3178654577876882e-05,
+      "loss": 0.32,
       "step": 9320
     },
     {
-      "epoch": 1.3663003663003663,
-      "grad_norm": 0.36688013681817694,
-      "learning_rate": 1.3241859641968318e-05,
-      "loss": 0.3328,
+      "epoch": 1.3727366406595025,
+      "grad_norm": 0.3566799951767712,
+      "learning_rate": 1.3170533200052019e-05,
+      "loss": 0.326,
       "step": 9325
     },
     {
-      "epoch": 1.367032967032967,
-      "grad_norm": 0.354463599391648,
-      "learning_rate": 1.3233794412797387e-05,
-      "loss": 0.3205,
+      "epoch": 1.3734726924775504,
+      "grad_norm": 0.38476022515812874,
+      "learning_rate": 1.316240949667539e-05,
+      "loss": 0.3296,
       "step": 9330
     },
     {
-      "epoch": 1.3677655677655678,
-      "grad_norm": 0.37314536883933214,
-      "learning_rate": 1.3225726833761729e-05,
-      "loss": 0.3243,
+      "epoch": 1.3742087442955984,
+      "grad_norm": 0.3545994591146522,
+      "learning_rate": 1.3154283473705645e-05,
+      "loss": 0.3226,
       "step": 9335
     },
     {
-      "epoch": 1.3684981684981685,
-      "grad_norm": 0.34344786208460487,
-      "learning_rate": 1.321765691072372e-05,
-      "loss": 0.3095,
+      "epoch": 1.3749447961136463,
+      "grad_norm": 0.3630231695401911,
+      "learning_rate": 1.3146155137103135e-05,
+      "loss": 0.3332,
       "step": 9340
     },
     {
-      "epoch": 1.3692307692307693,
-      "grad_norm": 0.3672887626205064,
-      "learning_rate": 1.3209584649547438e-05,
-      "loss": 0.3353,
+      "epoch": 1.3756808479316944,
+      "grad_norm": 0.36868872546961123,
+      "learning_rate": 1.3138024492829902e-05,
+      "loss": 0.329,
       "step": 9345
     },
     {
-      "epoch": 1.36996336996337,
-      "grad_norm": 0.3999571631089137,
-      "learning_rate": 1.3201510056098665e-05,
-      "loss": 0.3308,
+      "epoch": 1.3764168997497424,
+      "grad_norm": 0.3751309059411541,
+      "learning_rate": 1.3129891546849688e-05,
+      "loss": 0.3432,
       "step": 9350
     },
     {
-      "epoch": 1.3706959706959707,
-      "grad_norm": 0.34257353112977285,
-      "learning_rate": 1.3193433136244868e-05,
-      "loss": 0.3123,
+      "epoch": 1.3771529515677905,
+      "grad_norm": 0.3615824589106786,
+      "learning_rate": 1.3121756305127925e-05,
+      "loss": 0.3221,
       "step": 9355
     },
     {
-      "epoch": 1.3714285714285714,
-      "grad_norm": 0.34737676485578445,
-      "learning_rate": 1.3185353895855215e-05,
-      "loss": 0.3301,
+      "epoch": 1.3778890033858384,
+      "grad_norm": 0.3516015293803255,
+      "learning_rate": 1.311361877363172e-05,
+      "loss": 0.3167,
       "step": 9360
     },
     {
-      "epoch": 1.3721611721611722,
-      "grad_norm": 0.36982798486097596,
-      "learning_rate": 1.317727234080056e-05,
-      "loss": 0.3308,
+      "epoch": 1.3786250552038863,
+      "grad_norm": 0.37834677729887595,
+      "learning_rate": 1.3105478958329865e-05,
+      "loss": 0.3322,
       "step": 9365
     },
     {
-      "epoch": 1.372893772893773,
-      "grad_norm": 0.3669270882321095,
-      "learning_rate": 1.3169188476953422e-05,
-      "loss": 0.3305,
+      "epoch": 1.3793611070219343,
+      "grad_norm": 0.3626664899177328,
+      "learning_rate": 1.3097336865192825e-05,
+      "loss": 0.3223,
       "step": 9370
     },
     {
-      "epoch": 1.3736263736263736,
-      "grad_norm": 0.3628003702591798,
-      "learning_rate": 1.316110231018803e-05,
-      "loss": 0.3301,
+      "epoch": 1.3800971588399824,
+      "grad_norm": 0.3518899413107858,
+      "learning_rate": 1.3089192500192742e-05,
+      "loss": 0.3163,
       "step": 9375
     },
     {
-      "epoch": 1.3743589743589744,
-      "grad_norm": 0.35373465603923576,
-      "learning_rate": 1.3153013846380249e-05,
-      "loss": 0.3159,
+      "epoch": 1.3808332106580303,
+      "grad_norm": 0.3674765109654592,
+      "learning_rate": 1.3081045869303417e-05,
+      "loss": 0.3383,
       "step": 9380
     },
     {
-      "epoch": 1.375091575091575,
-      "grad_norm": 0.3655259797331987,
-      "learning_rate": 1.314492309140764e-05,
-      "loss": 0.3282,
+      "epoch": 1.3815692624760783,
+      "grad_norm": 0.3416972537434523,
+      "learning_rate": 1.3072896978500312e-05,
+      "loss": 0.3086,
       "step": 9385
     },
     {
-      "epoch": 1.3758241758241758,
-      "grad_norm": 0.3539664614226634,
-      "learning_rate": 1.313683005114942e-05,
-      "loss": 0.3144,
+      "epoch": 1.3823053142941264,
+      "grad_norm": 0.3464971046445158,
+      "learning_rate": 1.3064745833760554e-05,
+      "loss": 0.3049,
       "step": 9390
     },
     {
-      "epoch": 1.3765567765567766,
-      "grad_norm": 0.3828454934600868,
-      "learning_rate": 1.3128734731486465e-05,
-      "loss": 0.3183,
+      "epoch": 1.3830413661121743,
+      "grad_norm": 0.3616307252212219,
+      "learning_rate": 1.3056592441062921e-05,
+      "loss": 0.3348,
       "step": 9395
     },
     {
-      "epoch": 1.3772893772893773,
-      "grad_norm": 0.3621509618726607,
-      "learning_rate": 1.3120637138301309e-05,
-      "loss": 0.3311,
+      "epoch": 1.3837774179302222,
+      "grad_norm": 0.3729590232939161,
+      "learning_rate": 1.3048436806387831e-05,
+      "loss": 0.3232,
       "step": 9400
     },
     {
-      "epoch": 1.378021978021978,
-      "grad_norm": 0.40709700114637165,
-      "learning_rate": 1.3112537277478139e-05,
-      "loss": 0.3392,
+      "epoch": 1.3845134697482702,
+      "grad_norm": 0.3755404182669986,
+      "learning_rate": 1.304027893571736e-05,
+      "loss": 0.3491,
       "step": 9405
     },
     {
-      "epoch": 1.3787545787545787,
-      "grad_norm": 0.35733461333741195,
-      "learning_rate": 1.3104435154902789e-05,
-      "loss": 0.3015,
+      "epoch": 1.3852495215663183,
+      "grad_norm": 0.35654169900498894,
+      "learning_rate": 1.3032118835035216e-05,
+      "loss": 0.3201,
       "step": 9410
     },
     {
-      "epoch": 1.3794871794871795,
-      "grad_norm": 0.3683668380476496,
-      "learning_rate": 1.3096330776462738e-05,
-      "loss": 0.3485,
+      "epoch": 1.3859855733843662,
+      "grad_norm": 0.3630148565795268,
+      "learning_rate": 1.3023956510326747e-05,
+      "loss": 0.3159,
       "step": 9415
     },
     {
-      "epoch": 1.3802197802197802,
-      "grad_norm": 0.37440441580436834,
-      "learning_rate": 1.3088224148047102e-05,
-      "loss": 0.3179,
+      "epoch": 1.3867216252024144,
+      "grad_norm": 0.3806107009949009,
+      "learning_rate": 1.301579196757893e-05,
+      "loss": 0.3427,
       "step": 9420
     },
     {
-      "epoch": 1.380952380952381,
-      "grad_norm": 0.3761069786909644,
-      "learning_rate": 1.3080115275546633e-05,
-      "loss": 0.3231,
+      "epoch": 1.3874576770204623,
+      "grad_norm": 0.3609297145972763,
+      "learning_rate": 1.3007625212780362e-05,
+      "loss": 0.3359,
       "step": 9425
     },
     {
-      "epoch": 1.3816849816849817,
-      "grad_norm": 0.36812148125041294,
-      "learning_rate": 1.3072004164853708e-05,
-      "loss": 0.3318,
+      "epoch": 1.3881937288385102,
+      "grad_norm": 0.3647762309908433,
+      "learning_rate": 1.2999456251921274e-05,
+      "loss": 0.328,
       "step": 9430
     },
     {
-      "epoch": 1.3824175824175824,
-      "grad_norm": 0.3680198040443312,
-      "learning_rate": 1.3063890821862343e-05,
-      "loss": 0.3173,
+      "epoch": 1.3889297806565581,
+      "grad_norm": 0.3635213304603404,
+      "learning_rate": 1.2991285090993508e-05,
+      "loss": 0.3469,
       "step": 9435
     },
     {
-      "epoch": 1.3831501831501831,
-      "grad_norm": 0.364173471028859,
-      "learning_rate": 1.3055775252468168e-05,
-      "loss": 0.3143,
+      "epoch": 1.3896658324746063,
+      "grad_norm": 0.35263350449891173,
+      "learning_rate": 1.2983111735990526e-05,
+      "loss": 0.3262,
       "step": 9440
     },
     {
-      "epoch": 1.3838827838827839,
-      "grad_norm": 0.35694349814381804,
-      "learning_rate": 1.3047657462568429e-05,
-      "loss": 0.3134,
+      "epoch": 1.3904018842926542,
+      "grad_norm": 0.3582007215649553,
+      "learning_rate": 1.297493619290739e-05,
+      "loss": 0.3223,
       "step": 9445
     },
     {
-      "epoch": 1.3846153846153846,
-      "grad_norm": 0.3857759577482839,
-      "learning_rate": 1.303953745806199e-05,
-      "loss": 0.3222,
+      "epoch": 1.3911379361107021,
+      "grad_norm": 0.3575063738974865,
+      "learning_rate": 1.2966758467740778e-05,
+      "loss": 0.3301,
       "step": 9450
     },
     {
-      "epoch": 1.3853479853479853,
-      "grad_norm": 0.3592364706404115,
-      "learning_rate": 1.3031415244849322e-05,
-      "loss": 0.3198,
+      "epoch": 1.3918739879287503,
+      "grad_norm": 0.3737456120836393,
+      "learning_rate": 1.2958578566488962e-05,
+      "loss": 0.3179,
       "step": 9455
     },
     {
-      "epoch": 1.386080586080586,
-      "grad_norm": 0.36463687509292364,
-      "learning_rate": 1.3023290828832505e-05,
-      "loss": 0.3349,
+      "epoch": 1.3926100397467982,
+      "grad_norm": 0.3627947568637503,
+      "learning_rate": 1.2950396495151806e-05,
+      "loss": 0.3371,
       "step": 9460
     },
     {
-      "epoch": 1.3868131868131868,
-      "grad_norm": 0.3523885736969607,
-      "learning_rate": 1.301516421591521e-05,
-      "loss": 0.3116,
+      "epoch": 1.3933460915648461,
+      "grad_norm": 0.34610506809506375,
+      "learning_rate": 1.294221225973078e-05,
+      "loss": 0.3178,
       "step": 9465
     },
     {
-      "epoch": 1.3875457875457875,
-      "grad_norm": 0.3672763338810888,
-      "learning_rate": 1.3007035412002716e-05,
-      "loss": 0.3394,
+      "epoch": 1.394082143382894,
+      "grad_norm": 0.3622671936332881,
+      "learning_rate": 1.2934025866228921e-05,
+      "loss": 0.3259,
       "step": 9470
     },
     {
-      "epoch": 1.3882783882783882,
-      "grad_norm": 0.3663609919844645,
-      "learning_rate": 1.2998904423001884e-05,
-      "loss": 0.3307,
+      "epoch": 1.3948181952009422,
+      "grad_norm": 0.3654711313494927,
+      "learning_rate": 1.2925837320650863e-05,
+      "loss": 0.332,
       "step": 9475
     },
     {
-      "epoch": 1.389010989010989,
-      "grad_norm": 0.3800146791680141,
-      "learning_rate": 1.2990771254821173e-05,
-      "loss": 0.3343,
+      "epoch": 1.3955542470189901,
+      "grad_norm": 0.37062897243648746,
+      "learning_rate": 1.291764662900282e-05,
+      "loss": 0.3256,
       "step": 9480
     },
     {
-      "epoch": 1.3897435897435897,
-      "grad_norm": 0.34233438244524395,
-      "learning_rate": 1.2982635913370615e-05,
-      "loss": 0.3184,
+      "epoch": 1.3962902988370383,
+      "grad_norm": 0.3616799422926142,
+      "learning_rate": 1.2909453797292573e-05,
+      "loss": 0.3303,
       "step": 9485
     },
     {
-      "epoch": 1.3904761904761904,
-      "grad_norm": 0.3650168981731367,
-      "learning_rate": 1.2974498404561833e-05,
-      "loss": 0.3288,
+      "epoch": 1.3970263506550862,
+      "grad_norm": 0.3625431047824287,
+      "learning_rate": 1.2901258831529476e-05,
+      "loss": 0.335,
       "step": 9490
     },
     {
-      "epoch": 1.3912087912087912,
-      "grad_norm": 0.3736883637980164,
-      "learning_rate": 1.2966358734308009e-05,
-      "loss": 0.3344,
+      "epoch": 1.397762402473134,
+      "grad_norm": 0.37848657053974294,
+      "learning_rate": 1.2893061737724443e-05,
+      "loss": 0.3237,
       "step": 9495
     },
     {
-      "epoch": 1.3919413919413919,
-      "grad_norm": 0.37266422942715227,
-      "learning_rate": 1.2958216908523913e-05,
-      "loss": 0.3287,
+      "epoch": 1.398498454291182,
+      "grad_norm": 0.360329755944578,
+      "learning_rate": 1.2884862521889962e-05,
+      "loss": 0.332,
       "step": 9500
     },
     {
-      "epoch": 1.3926739926739926,
-      "grad_norm": 0.3735090036989961,
-      "learning_rate": 1.295007293312587e-05,
-      "loss": 0.3412,
+      "epoch": 1.3992345061092302,
+      "grad_norm": 0.35846538165300607,
+      "learning_rate": 1.2876661190040065e-05,
+      "loss": 0.336,
       "step": 9505
     },
     {
-      "epoch": 1.3934065934065933,
-      "grad_norm": 0.3736261624860495,
-      "learning_rate": 1.2941926814031771e-05,
-      "loss": 0.3181,
+      "epoch": 1.399970557927278,
+      "grad_norm": 0.36634166720874467,
+      "learning_rate": 1.2868457748190338e-05,
+      "loss": 0.3153,
       "step": 9510
     },
     {
-      "epoch": 1.394139194139194,
-      "grad_norm": 0.36911044210076915,
-      "learning_rate": 1.293377855716106e-05,
-      "loss": 0.3184,
+      "epoch": 1.400706609745326,
+      "grad_norm": 0.346140825339715,
+      "learning_rate": 1.286025220235792e-05,
+      "loss": 0.3131,
       "step": 9515
     },
     {
-      "epoch": 1.3948717948717948,
-      "grad_norm": 0.371489730667187,
-      "learning_rate": 1.2925628168434745e-05,
-      "loss": 0.3283,
+      "epoch": 1.4014426615633742,
+      "grad_norm": 0.3649084286169258,
+      "learning_rate": 1.2852044558561485e-05,
+      "loss": 0.3186,
       "step": 9520
     },
     {
-      "epoch": 1.3956043956043955,
-      "grad_norm": 0.3572826907489989,
-      "learning_rate": 1.2917475653775375e-05,
-      "loss": 0.3316,
+      "epoch": 1.402178713381422,
+      "grad_norm": 0.3647161259847614,
+      "learning_rate": 1.2843834822821254e-05,
+      "loss": 0.3212,
       "step": 9525
     },
     {
-      "epoch": 1.3963369963369963,
-      "grad_norm": 0.38473967712910584,
-      "learning_rate": 1.2909321019107042e-05,
-      "loss": 0.3387,
+      "epoch": 1.40291476519947,
+      "grad_norm": 0.36547776557727774,
+      "learning_rate": 1.2835623001158982e-05,
+      "loss": 0.3197,
       "step": 9530
     },
     {
-      "epoch": 1.397069597069597,
-      "grad_norm": 0.3652353261820711,
-      "learning_rate": 1.2901164270355384e-05,
-      "loss": 0.3136,
+      "epoch": 1.403650817017518,
+      "grad_norm": 0.4406783824401941,
+      "learning_rate": 1.2827409099597952e-05,
+      "loss": 0.3346,
       "step": 9535
     },
     {
-      "epoch": 1.3978021978021977,
-      "grad_norm": 0.3686381795113925,
-      "learning_rate": 1.2893005413447575e-05,
-      "loss": 0.3273,
+      "epoch": 1.404386868835566,
+      "grad_norm": 0.3804708880600225,
+      "learning_rate": 1.2819193124162963e-05,
+      "loss": 0.332,
       "step": 9540
     },
     {
-      "epoch": 1.3985347985347985,
-      "grad_norm": 0.3600860568488306,
-      "learning_rate": 1.2884844454312312e-05,
-      "loss": 0.3263,
+      "epoch": 1.405122920653614,
+      "grad_norm": 0.4583885112627424,
+      "learning_rate": 1.2810975080880352e-05,
+      "loss": 0.3277,
       "step": 9545
     },
     {
-      "epoch": 1.3992673992673992,
-      "grad_norm": 0.3519233136105934,
-      "learning_rate": 1.2876681398879837e-05,
-      "loss": 0.3165,
+      "epoch": 1.4058589724716621,
+      "grad_norm": 0.35210293086921174,
+      "learning_rate": 1.2802754975777962e-05,
+      "loss": 0.3205,
       "step": 9550
     },
     {
-      "epoch": 1.4,
-      "grad_norm": 0.38029208605876735,
-      "learning_rate": 1.2868516253081897e-05,
-      "loss": 0.3349,
+      "epoch": 1.40659502428971,
+      "grad_norm": 0.36636420152330823,
+      "learning_rate": 1.2794532814885152e-05,
+      "loss": 0.3308,
       "step": 9555
     },
     {
-      "epoch": 1.4007326007326006,
-      "grad_norm": 0.35691788504252764,
-      "learning_rate": 1.2860349022851772e-05,
-      "loss": 0.3092,
+      "epoch": 1.407331076107758,
+      "grad_norm": 0.3594038505788453,
+      "learning_rate": 1.2786308604232786e-05,
+      "loss": 0.333,
       "step": 9560
     },
     {
-      "epoch": 1.4014652014652014,
-      "grad_norm": 0.35239929231196526,
-      "learning_rate": 1.2852179714124241e-05,
-      "loss": 0.3188,
+      "epoch": 1.408067127925806,
+      "grad_norm": 0.37265979886466455,
+      "learning_rate": 1.2778082349853238e-05,
+      "loss": 0.3215,
       "step": 9565
     },
     {
-      "epoch": 1.402197802197802,
-      "grad_norm": 0.37262611721225447,
-      "learning_rate": 1.2844008332835616e-05,
-      "loss": 0.3381,
+      "epoch": 1.408803179743854,
+      "grad_norm": 0.35821379029717326,
+      "learning_rate": 1.2769854057780371e-05,
+      "loss": 0.3161,
       "step": 9570
     },
     {
-      "epoch": 1.4029304029304028,
-      "grad_norm": 0.36208087825524055,
-      "learning_rate": 1.2835834884923695e-05,
-      "loss": 0.335,
+      "epoch": 1.409539231561902,
+      "grad_norm": 0.37051766835051625,
+      "learning_rate": 1.2761623734049548e-05,
+      "loss": 0.3221,
       "step": 9575
     },
     {
-      "epoch": 1.4036630036630036,
-      "grad_norm": 0.37367278329464243,
-      "learning_rate": 1.2827659376327779e-05,
-      "loss": 0.3214,
+      "epoch": 1.41027528337995,
+      "grad_norm": 0.36328167945139916,
+      "learning_rate": 1.2753391384697629e-05,
+      "loss": 0.3315,
       "step": 9580
     },
     {
-      "epoch": 1.4043956043956043,
-      "grad_norm": 0.3695960378064843,
-      "learning_rate": 1.2819481812988683e-05,
-      "loss": 0.3241,
+      "epoch": 1.411011335197998,
+      "grad_norm": 0.3736414650967738,
+      "learning_rate": 1.2745157015762946e-05,
+      "loss": 0.3418,
       "step": 9585
     },
     {
-      "epoch": 1.405128205128205,
-      "grad_norm": 0.38738496590910493,
-      "learning_rate": 1.2811302200848699e-05,
-      "loss": 0.337,
+      "epoch": 1.411747387016046,
+      "grad_norm": 0.36379868859809367,
+      "learning_rate": 1.2736920633285323e-05,
+      "loss": 0.341,
       "step": 9590
     },
     {
-      "epoch": 1.4058608058608058,
-      "grad_norm": 0.3791099520469478,
-      "learning_rate": 1.2803120545851613e-05,
-      "loss": 0.3412,
+      "epoch": 1.4124834388340939,
+      "grad_norm": 0.39212028191698084,
+      "learning_rate": 1.2728682243306062e-05,
+      "loss": 0.3246,
       "step": 9595
     },
     {
-      "epoch": 1.4065934065934065,
-      "grad_norm": 0.35992251660877783,
-      "learning_rate": 1.2794936853942695e-05,
-      "loss": 0.3317,
+      "epoch": 1.4132194906521418,
+      "grad_norm": 0.3675636161944304,
+      "learning_rate": 1.2720441851867921e-05,
+      "loss": 0.3236,
       "step": 9600
     },
     {
-      "epoch": 1.4073260073260072,
-      "grad_norm": 0.37344607639813765,
-      "learning_rate": 1.2786751131068698e-05,
-      "loss": 0.3252,
+      "epoch": 1.41395554247019,
+      "grad_norm": 0.35409789751463777,
+      "learning_rate": 1.2712199465015148e-05,
+      "loss": 0.324,
       "step": 9605
     },
     {
-      "epoch": 1.408058608058608,
-      "grad_norm": 0.36916880428194454,
-      "learning_rate": 1.2778563383177845e-05,
-      "loss": 0.3097,
+      "epoch": 1.4146915942882379,
+      "grad_norm": 0.3601135565256549,
+      "learning_rate": 1.2703955088793436e-05,
+      "loss": 0.3295,
       "step": 9610
     },
     {
-      "epoch": 1.4087912087912087,
-      "grad_norm": 0.3651252085756624,
-      "learning_rate": 1.277037361621984e-05,
-      "loss": 0.327,
+      "epoch": 1.415427646106286,
+      "grad_norm": 0.3532116865072582,
+      "learning_rate": 1.2695708729249954e-05,
+      "loss": 0.3299,
       "step": 9615
     },
     {
-      "epoch": 1.4095238095238094,
-      "grad_norm": 0.3686790573730101,
-      "learning_rate": 1.2762181836145839e-05,
-      "loss": 0.3319,
+      "epoch": 1.416163697924334,
+      "grad_norm": 0.3832317141267909,
+      "learning_rate": 1.2687460392433313e-05,
+      "loss": 0.3323,
       "step": 9620
     },
     {
-      "epoch": 1.4102564102564101,
-      "grad_norm": 0.388258250297523,
-      "learning_rate": 1.275398804890848e-05,
-      "loss": 0.3412,
+      "epoch": 1.4168997497423819,
+      "grad_norm": 0.37985665667603924,
+      "learning_rate": 1.267921008439358e-05,
+      "loss": 0.3252,
       "step": 9625
     },
     {
-      "epoch": 1.4109890109890109,
-      "grad_norm": 0.3849595586698791,
-      "learning_rate": 1.274579226046184e-05,
-      "loss": 0.3451,
+      "epoch": 1.4176358015604298,
+      "grad_norm": 0.36948719326893054,
+      "learning_rate": 1.2670957811182268e-05,
+      "loss": 0.3112,
       "step": 9630
     },
     {
-      "epoch": 1.4117216117216116,
-      "grad_norm": 0.3592011858806382,
-      "learning_rate": 1.273759447676147e-05,
-      "loss": 0.3392,
+      "epoch": 1.4183718533784777,
+      "grad_norm": 0.3546223816404936,
+      "learning_rate": 1.2662703578852326e-05,
+      "loss": 0.3249,
       "step": 9635
     },
     {
-      "epoch": 1.4124542124542123,
-      "grad_norm": 0.37786836832104254,
-      "learning_rate": 1.2729394703764349e-05,
-      "loss": 0.3272,
+      "epoch": 1.4191079051965259,
+      "grad_norm": 0.3673572397597473,
+      "learning_rate": 1.2654447393458146e-05,
+      "loss": 0.3419,
       "step": 9640
     },
     {
-      "epoch": 1.413186813186813,
-      "grad_norm": 0.3712268682424444,
-      "learning_rate": 1.2721192947428922e-05,
-      "loss": 0.335,
+      "epoch": 1.4198439570145738,
+      "grad_norm": 0.35982448182968724,
+      "learning_rate": 1.2646189261055552e-05,
+      "loss": 0.3336,
       "step": 9645
     },
     {
-      "epoch": 1.4139194139194138,
-      "grad_norm": 0.3694596181035189,
-      "learning_rate": 1.271298921371506e-05,
-      "loss": 0.3257,
+      "epoch": 1.420580008832622,
+      "grad_norm": 0.3479702237193538,
+      "learning_rate": 1.2637929187701794e-05,
+      "loss": 0.3279,
       "step": 9650
     },
     {
-      "epoch": 1.4146520146520147,
-      "grad_norm": 0.36434138856836823,
-      "learning_rate": 1.2704783508584085e-05,
-      "loss": 0.3319,
+      "epoch": 1.4213160606506698,
+      "grad_norm": 0.36333010166494983,
+      "learning_rate": 1.2629667179455545e-05,
+      "loss": 0.3303,
       "step": 9655
     },
     {
-      "epoch": 1.4153846153846155,
-      "grad_norm": 0.3620604578743319,
-      "learning_rate": 1.2696575837998734e-05,
-      "loss": 0.3287,
+      "epoch": 1.4220521124687178,
+      "grad_norm": 0.36443666861560564,
+      "learning_rate": 1.26214032423769e-05,
+      "loss": 0.336,
       "step": 9660
     },
     {
-      "epoch": 1.4161172161172162,
-      "grad_norm": 0.36846199251115214,
-      "learning_rate": 1.2688366207923192e-05,
-      "loss": 0.3161,
+      "epoch": 1.4227881642867657,
+      "grad_norm": 0.37994334126913654,
+      "learning_rate": 1.2613137382527368e-05,
+      "loss": 0.3308,
       "step": 9665
     },
     {
-      "epoch": 1.416849816849817,
-      "grad_norm": 0.3709853962372448,
-      "learning_rate": 1.2680154624323049e-05,
-      "loss": 0.3246,
+      "epoch": 1.4235242161048138,
+      "grad_norm": 0.3550709018491184,
+      "learning_rate": 1.2604869605969863e-05,
+      "loss": 0.3189,
       "step": 9670
     },
     {
-      "epoch": 1.4175824175824177,
-      "grad_norm": 0.35401331608077374,
-      "learning_rate": 1.267194109316533e-05,
-      "loss": 0.3162,
+      "epoch": 1.4242602679228618,
+      "grad_norm": 0.3807182018327525,
+      "learning_rate": 1.2596599918768714e-05,
+      "loss": 0.3164,
       "step": 9675
     },
     {
-      "epoch": 1.4183150183150184,
-      "grad_norm": 0.37944078709014817,
-      "learning_rate": 1.2663725620418466e-05,
-      "loss": 0.3289,
+      "epoch": 1.4249963197409097,
+      "grad_norm": 0.3535627540152235,
+      "learning_rate": 1.258832832698965e-05,
+      "loss": 0.3331,
       "step": 9680
     },
     {
-      "epoch": 1.4190476190476191,
-      "grad_norm": 0.36152080018956056,
-      "learning_rate": 1.2655508212052299e-05,
-      "loss": 0.328,
+      "epoch": 1.4257323715589578,
+      "grad_norm": 0.3597186046691988,
+      "learning_rate": 1.2580054836699787e-05,
+      "loss": 0.3357,
       "step": 9685
     },
     {
-      "epoch": 1.4197802197802198,
-      "grad_norm": 0.363601312531346,
-      "learning_rate": 1.2647288874038081e-05,
-      "loss": 0.3382,
+      "epoch": 1.4264684233770057,
+      "grad_norm": 0.3607898725088784,
+      "learning_rate": 1.2571779453967645e-05,
+      "loss": 0.3177,
       "step": 9690
     },
     {
-      "epoch": 1.4205128205128206,
-      "grad_norm": 0.3519597170541447,
-      "learning_rate": 1.263906761234847e-05,
-      "loss": 0.3103,
+      "epoch": 1.4272044751950537,
+      "grad_norm": 0.3685107178438591,
+      "learning_rate": 1.2563502184863123e-05,
+      "loss": 0.3111,
       "step": 9695
     },
     {
-      "epoch": 1.4212454212454213,
-      "grad_norm": 0.36322466129170256,
-      "learning_rate": 1.2630844432957511e-05,
-      "loss": 0.3382,
+      "epoch": 1.4279405270131016,
+      "grad_norm": 0.3685121524762989,
+      "learning_rate": 1.2555223035457513e-05,
+      "loss": 0.312,
       "step": 9700
     },
     {
-      "epoch": 1.421978021978022,
-      "grad_norm": 0.3533094718983132,
-      "learning_rate": 1.2622619341840651e-05,
-      "loss": 0.3299,
+      "epoch": 1.4286765788311497,
+      "grad_norm": 0.36777856634479505,
+      "learning_rate": 1.2546942011823482e-05,
+      "loss": 0.3203,
       "step": 9705
     },
     {
-      "epoch": 1.4227106227106228,
-      "grad_norm": 0.3634175150959428,
-      "learning_rate": 1.2614392344974726e-05,
-      "loss": 0.3213,
+      "epoch": 1.4294126306491977,
+      "grad_norm": 0.3763432254595332,
+      "learning_rate": 1.2538659120035065e-05,
+      "loss": 0.3351,
       "step": 9710
     },
     {
-      "epoch": 1.4234432234432235,
-      "grad_norm": 0.3777985192244422,
-      "learning_rate": 1.2606163448337951e-05,
-      "loss": 0.3427,
+      "epoch": 1.4301486824672458,
+      "grad_norm": 0.3560562401177806,
+      "learning_rate": 1.2530374366167681e-05,
+      "loss": 0.3154,
       "step": 9715
     },
     {
-      "epoch": 1.4241758241758242,
-      "grad_norm": 0.3678173360075112,
-      "learning_rate": 1.2597932657909925e-05,
-      "loss": 0.3216,
+      "epoch": 1.4308847342852937,
+      "grad_norm": 0.3808190872790811,
+      "learning_rate": 1.2522087756298105e-05,
+      "loss": 0.3383,
       "step": 9720
     },
     {
-      "epoch": 1.424908424908425,
-      "grad_norm": 0.3852857078850646,
-      "learning_rate": 1.2589699979671625e-05,
-      "loss": 0.3381,
+      "epoch": 1.4316207861033416,
+      "grad_norm": 0.352390935554495,
+      "learning_rate": 1.251379929650447e-05,
+      "loss": 0.3163,
       "step": 9725
     },
     {
-      "epoch": 1.4256410256410257,
-      "grad_norm": 0.4059792809134705,
-      "learning_rate": 1.2581465419605401e-05,
-      "loss": 0.3288,
+      "epoch": 1.4323568379213896,
+      "grad_norm": 0.34418188179796083,
+      "learning_rate": 1.2505508992866277e-05,
+      "loss": 0.3261,
       "step": 9730
     },
     {
-      "epoch": 1.4263736263736264,
-      "grad_norm": 0.3791798416926858,
-      "learning_rate": 1.2573228983694959e-05,
-      "loss": 0.326,
+      "epoch": 1.4330928897394377,
+      "grad_norm": 0.369413162101418,
+      "learning_rate": 1.2497216851464376e-05,
+      "loss": 0.3386,
       "step": 9735
     },
     {
-      "epoch": 1.4271062271062271,
-      "grad_norm": 0.36026587693992806,
-      "learning_rate": 1.2564990677925384e-05,
-      "loss": 0.3318,
+      "epoch": 1.4338289415574856,
+      "grad_norm": 0.35288828172615627,
+      "learning_rate": 1.2488922878380956e-05,
+      "loss": 0.3185,
       "step": 9740
     },
     {
-      "epoch": 1.4278388278388279,
-      "grad_norm": 0.38410791336784683,
-      "learning_rate": 1.255675050828311e-05,
-      "loss": 0.3293,
+      "epoch": 1.4345649933755336,
+      "grad_norm": 0.37710524998769407,
+      "learning_rate": 1.2480627079699559e-05,
+      "loss": 0.334,
       "step": 9745
     },
     {
-      "epoch": 1.4285714285714286,
-      "grad_norm": 0.36705079163522697,
-      "learning_rate": 1.2548508480755928e-05,
-      "loss": 0.3297,
+      "epoch": 1.4353010451935817,
+      "grad_norm": 0.3688631183905549,
+      "learning_rate": 1.2472329461505066e-05,
+      "loss": 0.3308,
       "step": 9750
     },
     {
-      "epoch": 1.4293040293040293,
-      "grad_norm": 0.3476239543435942,
-      "learning_rate": 1.2540264601332974e-05,
-      "loss": 0.3279,
+      "epoch": 1.4360370970116296,
+      "grad_norm": 0.34165518712974824,
+      "learning_rate": 1.246403002988369e-05,
+      "loss": 0.3364,
       "step": 9755
     },
     {
-      "epoch": 1.43003663003663,
-      "grad_norm": 0.37436542488898616,
-      "learning_rate": 1.253201887600474e-05,
-      "loss": 0.3279,
+      "epoch": 1.4367731488296775,
+      "grad_norm": 0.35624705422598835,
+      "learning_rate": 1.2455728790922969e-05,
+      "loss": 0.3209,
       "step": 9760
     },
     {
-      "epoch": 1.4307692307692308,
-      "grad_norm": 0.3742352733144454,
-      "learning_rate": 1.2523771310763055e-05,
-      "loss": 0.3277,
+      "epoch": 1.4375092006477255,
+      "grad_norm": 0.3843382314486735,
+      "learning_rate": 1.2447425750711781e-05,
+      "loss": 0.309,
       "step": 9765
     },
     {
-      "epoch": 1.4315018315018315,
-      "grad_norm": 0.35372661192560517,
-      "learning_rate": 1.2515521911601081e-05,
-      "loss": 0.3234,
+      "epoch": 1.4382452524657736,
+      "grad_norm": 0.3679391034259437,
+      "learning_rate": 1.243912091534031e-05,
+      "loss": 0.332,
       "step": 9770
     },
     {
-      "epoch": 1.4322344322344323,
-      "grad_norm": 0.38120300969208226,
-      "learning_rate": 1.2507270684513311e-05,
-      "loss": 0.3128,
+      "epoch": 1.4389813042838215,
+      "grad_norm": 0.3775044558879287,
+      "learning_rate": 1.243081429090006e-05,
+      "loss": 0.336,
       "step": 9775
     },
     {
-      "epoch": 1.432967032967033,
-      "grad_norm": 0.35550449926136707,
-      "learning_rate": 1.2499017635495578e-05,
-      "loss": 0.3324,
+      "epoch": 1.4397173561018697,
+      "grad_norm": 0.37351475069537915,
+      "learning_rate": 1.2422505883483855e-05,
+      "loss": 0.3163,
       "step": 9780
     },
     {
-      "epoch": 1.4336996336996337,
-      "grad_norm": 0.35052414795080994,
-      "learning_rate": 1.2490762770545027e-05,
-      "loss": 0.3276,
+      "epoch": 1.4404534079199176,
+      "grad_norm": 0.3722635180490597,
+      "learning_rate": 1.241419569918582e-05,
+      "loss": 0.3441,
       "step": 9785
     },
     {
-      "epoch": 1.4344322344322344,
-      "grad_norm": 0.3670233096104474,
-      "learning_rate": 1.2482506095660131e-05,
-      "loss": 0.3224,
+      "epoch": 1.4411894597379655,
+      "grad_norm": 0.37797484378951485,
+      "learning_rate": 1.2405883744101388e-05,
+      "loss": 0.3157,
       "step": 9790
     },
     {
-      "epoch": 1.4351648351648352,
-      "grad_norm": 0.35942624267497264,
-      "learning_rate": 1.2474247616840668e-05,
-      "loss": 0.3272,
+      "epoch": 1.4419255115560135,
+      "grad_norm": 0.9221811180744349,
+      "learning_rate": 1.2397570024327284e-05,
+      "loss": 0.3344,
       "step": 9795
     },
     {
-      "epoch": 1.435897435897436,
-      "grad_norm": 0.35216775529264543,
-      "learning_rate": 1.2465987340087742e-05,
-      "loss": 0.3138,
+      "epoch": 1.4426615633740616,
+      "grad_norm": 0.34544402258408435,
+      "learning_rate": 1.2389254545961533e-05,
+      "loss": 0.3256,
       "step": 9800
     },
     {
-      "epoch": 1.4366300366300366,
-      "grad_norm": 0.36405057285971654,
-      "learning_rate": 1.2457725271403747e-05,
-      "loss": 0.3329,
+      "epoch": 1.4433976151921095,
+      "grad_norm": 0.3550167591332494,
+      "learning_rate": 1.238093731510345e-05,
+      "loss": 0.3299,
       "step": 9805
     },
     {
-      "epoch": 1.4373626373626374,
-      "grad_norm": 0.3468316475716749,
-      "learning_rate": 1.2449461416792392e-05,
-      "loss": 0.3241,
+      "epoch": 1.4441336670101574,
+      "grad_norm": 0.36703674325087776,
+      "learning_rate": 1.237261833785363e-05,
+      "loss": 0.3194,
       "step": 9810
     },
     {
-      "epoch": 1.438095238095238,
-      "grad_norm": 0.36416323875613926,
-      "learning_rate": 1.2441195782258673e-05,
-      "loss": 0.311,
+      "epoch": 1.4448697188282056,
+      "grad_norm": 0.3558987265422059,
+      "learning_rate": 1.2364297620313957e-05,
+      "loss": 0.317,
       "step": 9815
     },
     {
-      "epoch": 1.4388278388278388,
-      "grad_norm": 0.35905622215077654,
-      "learning_rate": 1.243292837380889e-05,
-      "loss": 0.3239,
+      "epoch": 1.4456057706462535,
+      "grad_norm": 0.35348406119176673,
+      "learning_rate": 1.2355975168587585e-05,
+      "loss": 0.3191,
       "step": 9820
     },
     {
-      "epoch": 1.4395604395604396,
-      "grad_norm": 0.39314554037862787,
-      "learning_rate": 1.2424659197450623e-05,
-      "loss": 0.3261,
+      "epoch": 1.4463418224643014,
+      "grad_norm": 0.36008976029499745,
+      "learning_rate": 1.234765098877894e-05,
+      "loss": 0.3418,
       "step": 9825
     },
     {
-      "epoch": 1.4402930402930403,
-      "grad_norm": 0.3664159966632821,
-      "learning_rate": 1.2416388259192747e-05,
-      "loss": 0.3254,
+      "epoch": 1.4470778742823494,
+      "grad_norm": 0.36495364878620473,
+      "learning_rate": 1.2339325086993722e-05,
+      "loss": 0.3273,
       "step": 9830
     },
     {
-      "epoch": 1.441025641025641,
-      "grad_norm": 0.3671811945041504,
-      "learning_rate": 1.2408115565045405e-05,
-      "loss": 0.3306,
+      "epoch": 1.4478139261003975,
+      "grad_norm": 0.35149843081280185,
+      "learning_rate": 1.2330997469338887e-05,
+      "loss": 0.3056,
       "step": 9835
     },
     {
-      "epoch": 1.4417582417582417,
-      "grad_norm": 0.3703868523472143,
-      "learning_rate": 1.2399841121020022e-05,
-      "loss": 0.3276,
+      "epoch": 1.4485499779184454,
+      "grad_norm": 0.36363605789573816,
+      "learning_rate": 1.2322668141922653e-05,
+      "loss": 0.3259,
       "step": 9840
     },
     {
-      "epoch": 1.4424908424908425,
-      "grad_norm": 0.37727072105652976,
-      "learning_rate": 1.2391564933129296e-05,
-      "loss": 0.3271,
+      "epoch": 1.4492860297364936,
+      "grad_norm": 0.36226765876901146,
+      "learning_rate": 1.2314337110854493e-05,
+      "loss": 0.314,
       "step": 9845
     },
     {
-      "epoch": 1.4432234432234432,
-      "grad_norm": 0.3779886259459971,
-      "learning_rate": 1.2383287007387193e-05,
-      "loss": 0.3092,
+      "epoch": 1.4500220815545415,
+      "grad_norm": 0.37791775400535604,
+      "learning_rate": 1.2306004382245125e-05,
+      "loss": 0.3387,
       "step": 9850
     },
     {
-      "epoch": 1.443956043956044,
-      "grad_norm": 0.35423291980474786,
-      "learning_rate": 1.2375007349808935e-05,
-      "loss": 0.3211,
+      "epoch": 1.4507581333725894,
+      "grad_norm": 0.3644828069004087,
+      "learning_rate": 1.2297669962206517e-05,
+      "loss": 0.3324,
       "step": 9855
     },
     {
-      "epoch": 1.4446886446886447,
-      "grad_norm": 0.3750162854977784,
-      "learning_rate": 1.236672596641101e-05,
-      "loss": 0.3358,
+      "epoch": 1.4514941851906373,
+      "grad_norm": 0.35087905074213327,
+      "learning_rate": 1.228933385685188e-05,
+      "loss": 0.3092,
       "step": 9860
     },
     {
-      "epoch": 1.4454212454212454,
-      "grad_norm": 0.387055357114979,
-      "learning_rate": 1.2358442863211155e-05,
-      "loss": 0.3373,
+      "epoch": 1.4522302370086855,
+      "grad_norm": 0.36174562446762215,
+      "learning_rate": 1.2280996072295648e-05,
+      "loss": 0.3264,
       "step": 9865
     },
     {
-      "epoch": 1.4461538461538461,
-      "grad_norm": 0.35012916522582016,
-      "learning_rate": 1.235015804622836e-05,
-      "loss": 0.3273,
+      "epoch": 1.4529662888267334,
+      "grad_norm": 0.3826809189587954,
+      "learning_rate": 1.2272656614653502e-05,
+      "loss": 0.338,
       "step": 9870
     },
     {
-      "epoch": 1.4468864468864469,
-      "grad_norm": 0.3733919351729767,
-      "learning_rate": 1.2341871521482863e-05,
-      "loss": 0.3296,
+      "epoch": 1.4537023406447813,
+      "grad_norm": 0.5562019271309896,
+      "learning_rate": 1.226431549004234e-05,
+      "loss": 0.3317,
       "step": 9875
     },
     {
-      "epoch": 1.4476190476190476,
-      "grad_norm": 0.348525810007892,
-      "learning_rate": 1.2333583294996129e-05,
-      "loss": 0.3179,
+      "epoch": 1.4544383924628295,
+      "grad_norm": 0.38202217470226957,
+      "learning_rate": 1.225597270458029e-05,
+      "loss": 0.3298,
       "step": 9880
     },
     {
-      "epoch": 1.4483516483516483,
-      "grad_norm": 0.3438451533587944,
-      "learning_rate": 1.2325293372790879e-05,
-      "loss": 0.3176,
+      "epoch": 1.4551744442808774,
+      "grad_norm": 0.36762194453903313,
+      "learning_rate": 1.2247628264386695e-05,
+      "loss": 0.3194,
       "step": 9885
     },
     {
-      "epoch": 1.449084249084249,
-      "grad_norm": 0.3474758651582091,
-      "learning_rate": 1.231700176089105e-05,
-      "loss": 0.3203,
+      "epoch": 1.4559104960989253,
+      "grad_norm": 0.36538088919194556,
+      "learning_rate": 1.2239282175582112e-05,
+      "loss": 0.3241,
       "step": 9890
     },
     {
-      "epoch": 1.4498168498168498,
-      "grad_norm": 0.35908142035083745,
-      "learning_rate": 1.2308708465321818e-05,
-      "loss": 0.3265,
+      "epoch": 1.4566465479169732,
+      "grad_norm": 0.35099984200093015,
+      "learning_rate": 1.2230934444288303e-05,
+      "loss": 0.3361,
       "step": 9895
     },
     {
-      "epoch": 1.4505494505494505,
-      "grad_norm": 0.3529094470709392,
-      "learning_rate": 1.230041349210958e-05,
-      "loss": 0.3182,
+      "epoch": 1.4573825997350214,
+      "grad_norm": 0.34827385648775994,
+      "learning_rate": 1.222258507662824e-05,
+      "loss": 0.3122,
       "step": 9900
     },
     {
-      "epoch": 1.4512820512820512,
-      "grad_norm": 0.3507377986859153,
-      "learning_rate": 1.2292116847281945e-05,
-      "loss": 0.3296,
+      "epoch": 1.4581186515530693,
+      "grad_norm": 0.3771645798567479,
+      "learning_rate": 1.2214234078726095e-05,
+      "loss": 0.3314,
       "step": 9905
     },
     {
-      "epoch": 1.452014652014652,
-      "grad_norm": 0.3617787600137893,
-      "learning_rate": 1.2283818536867742e-05,
-      "loss": 0.3261,
+      "epoch": 1.4588547033711174,
+      "grad_norm": 0.35445021978032787,
+      "learning_rate": 1.2205881456707228e-05,
+      "loss": 0.3303,
       "step": 9910
     },
     {
-      "epoch": 1.4527472527472527,
-      "grad_norm": 0.36655083849655967,
-      "learning_rate": 1.2275518566897014e-05,
-      "loss": 0.3394,
+      "epoch": 1.4595907551891654,
+      "grad_norm": 0.3522443625084909,
+      "learning_rate": 1.2197527216698211e-05,
+      "loss": 0.3166,
       "step": 9915
     },
     {
-      "epoch": 1.4534798534798534,
-      "grad_norm": 0.37343934762016906,
-      "learning_rate": 1.2267216943401004e-05,
-      "loss": 0.3316,
+      "epoch": 1.4603268070072133,
+      "grad_norm": 0.35827516974628426,
+      "learning_rate": 1.2189171364826774e-05,
+      "loss": 0.3409,
       "step": 9920
     },
     {
-      "epoch": 1.4542124542124542,
-      "grad_norm": 0.359956083144425,
-      "learning_rate": 1.2258913672412161e-05,
-      "loss": 0.3188,
+      "epoch": 1.4610628588252612,
+      "grad_norm": 0.36352057787727604,
+      "learning_rate": 1.2180813907221854e-05,
+      "loss": 0.3325,
       "step": 9925
     },
     {
-      "epoch": 1.4549450549450549,
-      "grad_norm": 0.36959838363749536,
-      "learning_rate": 1.225060875996412e-05,
-      "loss": 0.3269,
+      "epoch": 1.4617989106433094,
+      "grad_norm": 0.3457932075927401,
+      "learning_rate": 1.217245485001355e-05,
+      "loss": 0.3122,
       "step": 9930
     },
     {
-      "epoch": 1.4556776556776556,
-      "grad_norm": 0.3547512039117925,
-      "learning_rate": 1.224230221209173e-05,
-      "loss": 0.3192,
+      "epoch": 1.4625349624613573,
+      "grad_norm": 0.35550973394070706,
+      "learning_rate": 1.216409419933314e-05,
+      "loss": 0.3322,
       "step": 9935
     },
     {
-      "epoch": 1.4564102564102563,
-      "grad_norm": 0.36457371437656955,
-      "learning_rate": 1.223399403483101e-05,
-      "loss": 0.3252,
+      "epoch": 1.4632710142794052,
+      "grad_norm": 0.3528983985628642,
+      "learning_rate": 1.2155731961313078e-05,
+      "loss": 0.325,
       "step": 9940
     },
     {
-      "epoch": 1.457142857142857,
-      "grad_norm": 0.36923966954367354,
-      "learning_rate": 1.2225684234219165e-05,
-      "loss": 0.3252,
+      "epoch": 1.4640070660974533,
+      "grad_norm": 0.3771517536740423,
+      "learning_rate": 1.2147368142086962e-05,
+      "loss": 0.3223,
       "step": 9945
     },
     {
-      "epoch": 1.4578754578754578,
-      "grad_norm": 0.348627234105607,
-      "learning_rate": 1.221737281629459e-05,
-      "loss": 0.3119,
+      "epoch": 1.4647431179155013,
+      "grad_norm": 0.3646297373366747,
+      "learning_rate": 1.2139002747789572e-05,
+      "loss": 0.321,
       "step": 9950
     },
     {
-      "epoch": 1.4586080586080585,
-      "grad_norm": 0.37792018293515667,
-      "learning_rate": 1.220905978709685e-05,
-      "loss": 0.3262,
+      "epoch": 1.4654791697335492,
+      "grad_norm": 0.37253843366455597,
+      "learning_rate": 1.2130635784556832e-05,
+      "loss": 0.3234,
       "step": 9955
     },
     {
-      "epoch": 1.4593406593406593,
-      "grad_norm": 0.35387054623983993,
-      "learning_rate": 1.220074515266667e-05,
-      "loss": 0.3206,
+      "epoch": 1.4662152215515971,
+      "grad_norm": 0.36076725034414125,
+      "learning_rate": 1.212226725852582e-05,
+      "loss": 0.3249,
       "step": 9960
     },
     {
-      "epoch": 1.46007326007326,
-      "grad_norm": 0.3686030934970665,
-      "learning_rate": 1.2192428919045964e-05,
-      "loss": 0.3249,
+      "epoch": 1.4669512733696453,
+      "grad_norm": 0.39489509622684243,
+      "learning_rate": 1.211389717583476e-05,
+      "loss": 0.3214,
       "step": 9965
     },
     {
-      "epoch": 1.4608058608058607,
-      "grad_norm": 0.3594201922264351,
-      "learning_rate": 1.2184111092277784e-05,
-      "loss": 0.3146,
+      "epoch": 1.4676873251876932,
+      "grad_norm": 0.3692945107321283,
+      "learning_rate": 1.210552554262301e-05,
+      "loss": 0.3184,
       "step": 9970
     },
     {
-      "epoch": 1.4615384615384617,
-      "grad_norm": 0.37403194478495244,
-      "learning_rate": 1.2175791678406356e-05,
-      "loss": 0.3334,
+      "epoch": 1.4684233770057413,
+      "grad_norm": 0.37403882553149215,
+      "learning_rate": 1.2097152365031083e-05,
+      "loss": 0.3147,
       "step": 9975
     },
     {
-      "epoch": 1.4622710622710624,
-      "grad_norm": 0.35752027957136523,
-      "learning_rate": 1.2167470683477054e-05,
-      "loss": 0.3306,
+      "epoch": 1.4691594288237892,
+      "grad_norm": 0.34774471877863267,
+      "learning_rate": 1.2088777649200607e-05,
+      "loss": 0.3232,
       "step": 9980
     },
     {
-      "epoch": 1.4630036630036631,
-      "grad_norm": 0.3860281228719006,
-      "learning_rate": 1.2159148113536401e-05,
-      "loss": 0.3259,
+      "epoch": 1.4698954806418372,
+      "grad_norm": 0.3561958651402395,
+      "learning_rate": 1.2080401401274347e-05,
+      "loss": 0.3158,
       "step": 9985
     },
     {
-      "epoch": 1.4637362637362639,
-      "grad_norm": 0.34715910320392346,
-      "learning_rate": 1.2150823974632066e-05,
-      "loss": 0.3316,
+      "epoch": 1.470631532459885,
+      "grad_norm": 0.37059627371950304,
+      "learning_rate": 1.207202362739619e-05,
+      "loss": 0.3231,
       "step": 9990
     },
     {
-      "epoch": 1.4644688644688646,
-      "grad_norm": 0.36121488522937034,
-      "learning_rate": 1.2142498272812853e-05,
-      "loss": 0.3297,
+      "epoch": 1.4713675842779332,
+      "grad_norm": 0.35692510403444117,
+      "learning_rate": 1.2063644333711142e-05,
+      "loss": 0.3245,
       "step": 9995
     },
     {
-      "epoch": 1.4652014652014653,
-      "grad_norm": 0.36155581049444635,
-      "learning_rate": 1.2134171014128708e-05,
-      "loss": 0.3389,
+      "epoch": 1.4721036360959812,
+      "grad_norm": 0.3485071439118818,
+      "learning_rate": 1.2055263526365327e-05,
+      "loss": 0.3224,
       "step": 10000
     },
     {
-      "epoch": 1.465934065934066,
-      "grad_norm": 0.3623554102698869,
-      "learning_rate": 1.2125842204630705e-05,
-      "loss": 0.3353,
+      "epoch": 1.472839687914029,
+      "grad_norm": 0.37337489308107585,
+      "learning_rate": 1.204688121150597e-05,
+      "loss": 0.3404,
       "step": 10005
     },
     {
-      "epoch": 1.4666666666666668,
-      "grad_norm": 0.36221237189765315,
-      "learning_rate": 1.2117511850371047e-05,
-      "loss": 0.3146,
+      "epoch": 1.4735757397320772,
+      "grad_norm": 0.35341948300371445,
+      "learning_rate": 1.2038497395281415e-05,
+      "loss": 0.3229,
       "step": 10010
     },
     {
-      "epoch": 1.4673992673992675,
-      "grad_norm": 0.3588595459686551,
-      "learning_rate": 1.2109179957403051e-05,
-      "loss": 0.3261,
+      "epoch": 1.4743117915501252,
+      "grad_norm": 0.3578055509928498,
+      "learning_rate": 1.2030112083841096e-05,
+      "loss": 0.3159,
       "step": 10015
     },
     {
-      "epoch": 1.4681318681318682,
-      "grad_norm": 0.3764291592050611,
-      "learning_rate": 1.2100846531781172e-05,
-      "loss": 0.3303,
+      "epoch": 1.475047843368173,
+      "grad_norm": 0.3465916189627656,
+      "learning_rate": 1.2021725283335552e-05,
+      "loss": 0.3123,
       "step": 10020
     },
     {
-      "epoch": 1.468864468864469,
-      "grad_norm": 0.3731595124575503,
-      "learning_rate": 1.2092511579560956e-05,
-      "loss": 0.3208,
+      "epoch": 1.475783895186221,
+      "grad_norm": 0.3630730241253402,
+      "learning_rate": 1.2013336999916407e-05,
+      "loss": 0.3246,
       "step": 10025
     },
     {
-      "epoch": 1.4695970695970697,
-      "grad_norm": 0.34902848189581964,
-      "learning_rate": 1.2084175106799073e-05,
-      "loss": 0.3145,
+      "epoch": 1.4765199470042691,
+      "grad_norm": 0.3507765218242383,
+      "learning_rate": 1.200494723973638e-05,
+      "loss": 0.3311,
       "step": 10030
     },
     {
-      "epoch": 1.4703296703296704,
-      "grad_norm": 0.3591207789465635,
-      "learning_rate": 1.2075837119553286e-05,
-      "loss": 0.3149,
+      "epoch": 1.477255998822317,
+      "grad_norm": 0.35638794461805806,
+      "learning_rate": 1.1996556008949263e-05,
+      "loss": 0.3264,
       "step": 10035
     },
     {
-      "epoch": 1.4710622710622712,
-      "grad_norm": 0.3774858195106769,
-      "learning_rate": 1.2067497623882476e-05,
-      "loss": 0.3227,
+      "epoch": 1.4779920506403652,
+      "grad_norm": 0.3621182798155676,
+      "learning_rate": 1.1988163313709938e-05,
+      "loss": 0.3483,
       "step": 10040
     },
     {
-      "epoch": 1.471794871794872,
-      "grad_norm": 0.35215313715073393,
-      "learning_rate": 1.20591566258466e-05,
-      "loss": 0.3077,
+      "epoch": 1.4787281024584131,
+      "grad_norm": 0.37484358727447487,
+      "learning_rate": 1.1979769160174352e-05,
+      "loss": 0.3208,
       "step": 10045
     },
     {
-      "epoch": 1.4725274725274726,
-      "grad_norm": 0.3768631505032113,
-      "learning_rate": 1.2050814131506727e-05,
-      "loss": 0.3262,
+      "epoch": 1.479464154276461,
+      "grad_norm": 0.3603445775138095,
+      "learning_rate": 1.197137355449953e-05,
+      "loss": 0.3139,
       "step": 10050
     },
     {
-      "epoch": 1.4732600732600734,
-      "grad_norm": 0.34814112261314933,
-      "learning_rate": 1.2042470146924993e-05,
-      "loss": 0.3226,
+      "epoch": 1.480200206094509,
+      "grad_norm": 0.3664768734757381,
+      "learning_rate": 1.1962976502843555e-05,
+      "loss": 0.3339,
       "step": 10055
     },
     {
-      "epoch": 1.473992673992674,
-      "grad_norm": 0.3863813275545428,
-      "learning_rate": 1.2034124678164631e-05,
-      "loss": 0.3213,
+      "epoch": 1.4809362579125571,
+      "grad_norm": 0.3638454991526735,
+      "learning_rate": 1.1954578011365577e-05,
+      "loss": 0.3257,
       "step": 10060
     },
     {
-      "epoch": 1.4747252747252748,
-      "grad_norm": 0.3892196789186328,
-      "learning_rate": 1.2025777731289947e-05,
-      "loss": 0.3129,
+      "epoch": 1.481672309730605,
+      "grad_norm": 0.36466330838741023,
+      "learning_rate": 1.1946178086225797e-05,
+      "loss": 0.3286,
       "step": 10065
     },
     {
-      "epoch": 1.4754578754578755,
-      "grad_norm": 0.36067816661970625,
-      "learning_rate": 1.2017429312366327e-05,
-      "loss": 0.3261,
+      "epoch": 1.482408361548653,
+      "grad_norm": 0.3703936862028979,
+      "learning_rate": 1.1937776733585465e-05,
+      "loss": 0.325,
       "step": 10070
     },
     {
-      "epoch": 1.4761904761904763,
-      "grad_norm": 0.3706845896472646,
-      "learning_rate": 1.2009079427460216e-05,
-      "loss": 0.3119,
+      "epoch": 1.483144413366701,
+      "grad_norm": 0.3667536628158024,
+      "learning_rate": 1.1929373959606884e-05,
+      "loss": 0.319,
       "step": 10075
     },
     {
-      "epoch": 1.476923076923077,
-      "grad_norm": 0.357211144171938,
-      "learning_rate": 1.2000728082639133e-05,
-      "loss": 0.3283,
+      "epoch": 1.483880465184749,
+      "grad_norm": 0.36572450136518264,
+      "learning_rate": 1.1920969770453402e-05,
+      "loss": 0.3151,
       "step": 10080
     },
     {
-      "epoch": 1.4776556776556777,
-      "grad_norm": 0.36128512918836525,
-      "learning_rate": 1.1992375283971652e-05,
-      "loss": 0.3201,
+      "epoch": 1.484616517002797,
+      "grad_norm": 0.3759356836534542,
+      "learning_rate": 1.1912564172289397e-05,
+      "loss": 0.3436,
       "step": 10085
     },
     {
-      "epoch": 1.4783882783882785,
-      "grad_norm": 0.3527024476733989,
-      "learning_rate": 1.1984021037527408e-05,
-      "loss": 0.321,
+      "epoch": 1.4853525688208449,
+      "grad_norm": 0.35598313718531005,
+      "learning_rate": 1.1904157171280282e-05,
+      "loss": 0.3398,
       "step": 10090
     },
     {
-      "epoch": 1.4791208791208792,
-      "grad_norm": 0.36190414974375074,
-      "learning_rate": 1.197566534937709e-05,
-      "loss": 0.3294,
+      "epoch": 1.486088620638893,
+      "grad_norm": 0.3669169951482803,
+      "learning_rate": 1.1895748773592506e-05,
+      "loss": 0.3231,
       "step": 10095
     },
     {
-      "epoch": 1.47985347985348,
-      "grad_norm": 0.3527905520818129,
-      "learning_rate": 1.196730822559243e-05,
-      "loss": 0.331,
+      "epoch": 1.486824672456941,
+      "grad_norm": 0.3693018739850583,
+      "learning_rate": 1.1887338985393537e-05,
+      "loss": 0.3247,
       "step": 10100
     },
     {
-      "epoch": 1.4805860805860807,
-      "grad_norm": 0.35543633549908826,
-      "learning_rate": 1.1958949672246198e-05,
-      "loss": 0.3281,
+      "epoch": 1.487560724274989,
+      "grad_norm": 0.3610580507913168,
+      "learning_rate": 1.1878927812851862e-05,
+      "loss": 0.315,
       "step": 10105
     },
     {
-      "epoch": 1.4813186813186814,
-      "grad_norm": 0.3602609678076822,
-      "learning_rate": 1.195058969541222e-05,
-      "loss": 0.3135,
+      "epoch": 1.488296776093037,
+      "grad_norm": 0.43074367823109083,
+      "learning_rate": 1.1870515262136989e-05,
+      "loss": 0.3277,
       "step": 10110
     },
     {
-      "epoch": 1.4820512820512821,
-      "grad_norm": 0.37628993017748585,
-      "learning_rate": 1.1942228301165335e-05,
-      "loss": 0.323,
+      "epoch": 1.489032827911085,
+      "grad_norm": 0.3617920059705095,
+      "learning_rate": 1.1862101339419421e-05,
+      "loss": 0.3224,
       "step": 10115
     },
     {
-      "epoch": 1.4827838827838828,
-      "grad_norm": 0.3746875216036666,
-      "learning_rate": 1.1933865495581429e-05,
-      "loss": 0.3159,
+      "epoch": 1.4897688797291329,
+      "grad_norm": 0.34565184093631734,
+      "learning_rate": 1.1853686050870691e-05,
+      "loss": 0.3241,
       "step": 10120
     },
     {
-      "epoch": 1.4835164835164836,
-      "grad_norm": 0.3644924663906216,
-      "learning_rate": 1.1925501284737409e-05,
-      "loss": 0.3191,
+      "epoch": 1.490504931547181,
+      "grad_norm": 0.38547787369437103,
+      "learning_rate": 1.184526940266332e-05,
+      "loss": 0.3419,
       "step": 10125
     },
     {
-      "epoch": 1.4842490842490843,
-      "grad_norm": 0.3549440442907127,
-      "learning_rate": 1.1917135674711193e-05,
-      "loss": 0.3124,
+      "epoch": 1.491240983365229,
+      "grad_norm": 0.36804240650848136,
+      "learning_rate": 1.1836851400970824e-05,
+      "loss": 0.3168,
       "step": 10130
     },
     {
-      "epoch": 1.484981684981685,
-      "grad_norm": 0.37470622439881085,
-      "learning_rate": 1.190876867158173e-05,
-      "loss": 0.3266,
+      "epoch": 1.4919770351832768,
+      "grad_norm": 0.36688440785220466,
+      "learning_rate": 1.182843205196772e-05,
+      "loss": 0.3271,
       "step": 10135
     },
     {
-      "epoch": 1.4857142857142858,
-      "grad_norm": 0.39389565715663216,
-      "learning_rate": 1.1900400281428974e-05,
-      "loss": 0.3133,
+      "epoch": 1.492713087001325,
+      "grad_norm": 0.3526309047769882,
+      "learning_rate": 1.1820011361829509e-05,
+      "loss": 0.3174,
       "step": 10140
     },
     {
-      "epoch": 1.4864468864468865,
-      "grad_norm": 0.3668611786885184,
-      "learning_rate": 1.1892030510333893e-05,
-      "loss": 0.3131,
+      "epoch": 1.493449138819373,
+      "grad_norm": 0.36672637373054146,
+      "learning_rate": 1.1811589336732673e-05,
+      "loss": 0.3179,
       "step": 10145
     },
     {
-      "epoch": 1.4871794871794872,
-      "grad_norm": 0.3626705673304234,
-      "learning_rate": 1.1883659364378441e-05,
-      "loss": 0.3241,
+      "epoch": 1.4941851906374208,
+      "grad_norm": 0.3623402155215344,
+      "learning_rate": 1.180316598285468e-05,
+      "loss": 0.3374,
       "step": 10150
     },
     {
-      "epoch": 1.487912087912088,
-      "grad_norm": 0.34833211232219496,
-      "learning_rate": 1.1875286849645598e-05,
-      "loss": 0.3223,
+      "epoch": 1.4949212424554688,
+      "grad_norm": 0.3630440450286294,
+      "learning_rate": 1.1794741306373966e-05,
+      "loss": 0.3381,
       "step": 10155
     },
     {
-      "epoch": 1.4886446886446887,
-      "grad_norm": 0.3657456508591803,
-      "learning_rate": 1.1866912972219315e-05,
-      "loss": 0.3333,
+      "epoch": 1.495657294273517,
+      "grad_norm": 0.37018899419751444,
+      "learning_rate": 1.1786315313469943e-05,
+      "loss": 0.3314,
       "step": 10160
     },
     {
-      "epoch": 1.4893772893772894,
-      "grad_norm": 0.387084111689358,
-      "learning_rate": 1.1858537738184548e-05,
-      "loss": 0.3275,
+      "epoch": 1.4963933460915648,
+      "grad_norm": 0.36149997093325387,
+      "learning_rate": 1.177788801032298e-05,
+      "loss": 0.317,
       "step": 10165
     },
     {
-      "epoch": 1.4901098901098901,
-      "grad_norm": 0.3740119251070165,
-      "learning_rate": 1.1850161153627227e-05,
-      "loss": 0.3221,
+      "epoch": 1.497129397909613,
+      "grad_norm": 0.38896520393181283,
+      "learning_rate": 1.1769459403114419e-05,
+      "loss": 0.331,
       "step": 10170
     },
     {
-      "epoch": 1.4908424908424909,
-      "grad_norm": 0.36085846264377625,
-      "learning_rate": 1.1841783224634279e-05,
-      "loss": 0.3286,
+      "epoch": 1.497865449727661,
+      "grad_norm": 0.3735773166816534,
+      "learning_rate": 1.1761029498026547e-05,
+      "loss": 0.3357,
       "step": 10175
     },
     {
-      "epoch": 1.4915750915750916,
-      "grad_norm": 0.36478035016967936,
-      "learning_rate": 1.1833403957293588e-05,
-      "loss": 0.3178,
+      "epoch": 1.4986015015457088,
+      "grad_norm": 0.3682398591528469,
+      "learning_rate": 1.175259830124261e-05,
+      "loss": 0.3322,
       "step": 10180
     },
     {
-      "epoch": 1.4923076923076923,
-      "grad_norm": 0.3668196694426494,
-      "learning_rate": 1.1825023357694032e-05,
-      "loss": 0.31,
+      "epoch": 1.4993375533637567,
+      "grad_norm": 0.34970333168068707,
+      "learning_rate": 1.1744165818946803e-05,
+      "loss": 0.3387,
       "step": 10185
     },
     {
-      "epoch": 1.493040293040293,
-      "grad_norm": 0.3793473590675217,
-      "learning_rate": 1.1816641431925437e-05,
-      "loss": 0.3271,
+      "epoch": 1.5000736051818047,
+      "grad_norm": 0.37456208444503025,
+      "learning_rate": 1.1735732057324258e-05,
+      "loss": 0.3443,
       "step": 10190
     },
     {
-      "epoch": 1.4937728937728938,
-      "grad_norm": 0.34993742755389273,
-      "learning_rate": 1.1808258186078609e-05,
-      "loss": 0.3182,
+      "epoch": 1.5008096569998528,
+      "grad_norm": 0.37699410156485463,
+      "learning_rate": 1.1727297022561043e-05,
+      "loss": 0.3228,
       "step": 10195
     },
     {
-      "epoch": 1.4945054945054945,
-      "grad_norm": 0.35281709590537846,
-      "learning_rate": 1.17998736262453e-05,
-      "loss": 0.3214,
+      "epoch": 1.5015457088179007,
+      "grad_norm": 0.37620292003651296,
+      "learning_rate": 1.1718860720844174e-05,
+      "loss": 0.3341,
       "step": 10200
     },
     {
-      "epoch": 1.4952380952380953,
-      "grad_norm": 0.3625611653357865,
-      "learning_rate": 1.1791487758518231e-05,
-      "loss": 0.3265,
+      "epoch": 1.5022817606359489,
+      "grad_norm": 0.37180888214117386,
+      "learning_rate": 1.1710423158361576e-05,
+      "loss": 0.3216,
       "step": 10205
     },
     {
-      "epoch": 1.495970695970696,
-      "grad_norm": 0.3756459611364907,
-      "learning_rate": 1.1783100588991057e-05,
-      "loss": 0.3397,
+      "epoch": 1.5030178124539968,
+      "grad_norm": 0.38653617755091274,
+      "learning_rate": 1.170198434130212e-05,
+      "loss": 0.3271,
       "step": 10210
     },
     {
-      "epoch": 1.4967032967032967,
-      "grad_norm": 0.35708722996847425,
-      "learning_rate": 1.1774712123758392e-05,
-      "loss": 0.324,
+      "epoch": 1.5037538642720447,
+      "grad_norm": 0.3794914965002413,
+      "learning_rate": 1.1693544275855578e-05,
+      "loss": 0.3245,
       "step": 10215
     },
     {
-      "epoch": 1.4974358974358974,
-      "grad_norm": 0.3573300867905406,
-      "learning_rate": 1.1766322368915786e-05,
-      "loss": 0.3216,
+      "epoch": 1.5044899160900926,
+      "grad_norm": 0.37085292438323103,
+      "learning_rate": 1.1685102968212653e-05,
+      "loss": 0.3232,
       "step": 10220
     },
     {
-      "epoch": 1.4981684981684982,
-      "grad_norm": 0.3592396734716313,
-      "learning_rate": 1.1757931330559728e-05,
-      "loss": 0.3183,
+      "epoch": 1.5052259679081408,
+      "grad_norm": 0.3499641864552821,
+      "learning_rate": 1.1676660424564951e-05,
+      "loss": 0.3121,
       "step": 10225
     },
     {
-      "epoch": 1.498901098901099,
-      "grad_norm": 0.34595230332393084,
-      "learning_rate": 1.1749539014787637e-05,
-      "loss": 0.3145,
+      "epoch": 1.5059620197261887,
+      "grad_norm": 0.3788203274765936,
+      "learning_rate": 1.1668216651104987e-05,
+      "loss": 0.3285,
       "step": 10230
     },
     {
-      "epoch": 1.4996336996336996,
-      "grad_norm": 0.36411036031952054,
-      "learning_rate": 1.1741145427697862e-05,
-      "loss": 0.332,
+      "epoch": 1.5066980715442368,
+      "grad_norm": 0.3806845182429137,
+      "learning_rate": 1.1659771654026175e-05,
+      "loss": 0.3333,
       "step": 10235
     },
     {
-      "epoch": 1.5003663003663004,
-      "grad_norm": 0.35389093482467654,
-      "learning_rate": 1.1732750575389678e-05,
-      "loss": 0.3181,
+      "epoch": 1.5074341233622848,
+      "grad_norm": 0.3728816337005067,
+      "learning_rate": 1.1651325439522831e-05,
+      "loss": 0.3367,
       "step": 10240
     },
     {
-      "epoch": 1.501098901098901,
-      "grad_norm": 0.35385316323976235,
-      "learning_rate": 1.1724354463963274e-05,
-      "loss": 0.3214,
+      "epoch": 1.5081701751803327,
+      "grad_norm": 0.35231974719445514,
+      "learning_rate": 1.1642878013790162e-05,
+      "loss": 0.3085,
       "step": 10245
     },
     {
-      "epoch": 1.5018315018315018,
-      "grad_norm": 0.36749115264217064,
-      "learning_rate": 1.1715957099519761e-05,
-      "loss": 0.3498,
+      "epoch": 1.5089062269983806,
+      "grad_norm": 0.3686547819457083,
+      "learning_rate": 1.163442938302426e-05,
+      "loss": 0.335,
       "step": 10250
     },
     {
-      "epoch": 1.5025641025641026,
-      "grad_norm": 0.3638081390632744,
-      "learning_rate": 1.1707558488161155e-05,
-      "loss": 0.3298,
+      "epoch": 1.5096422788164285,
+      "grad_norm": 0.37844065365899404,
+      "learning_rate": 1.1625979553422106e-05,
+      "loss": 0.3256,
       "step": 10255
     },
     {
-      "epoch": 1.5032967032967033,
-      "grad_norm": 0.3783219086875028,
-      "learning_rate": 1.1699158635990375e-05,
-      "loss": 0.3369,
+      "epoch": 1.5103783306344767,
+      "grad_norm": 0.3582203092302988,
+      "learning_rate": 1.1617528531181563e-05,
+      "loss": 0.3275,
       "step": 10260
     },
     {
-      "epoch": 1.504029304029304,
-      "grad_norm": 0.3793939099463113,
-      "learning_rate": 1.1690757549111254e-05,
-      "loss": 0.3296,
+      "epoch": 1.5111143824525246,
+      "grad_norm": 0.3607991359308337,
+      "learning_rate": 1.1609076322501359e-05,
+      "loss": 0.3444,
       "step": 10265
     },
     {
-      "epoch": 1.5047619047619047,
-      "grad_norm": 0.34984401356687794,
-      "learning_rate": 1.1682355233628511e-05,
-      "loss": 0.3137,
+      "epoch": 1.5118504342705728,
+      "grad_norm": 0.37645404960000367,
+      "learning_rate": 1.1600622933581099e-05,
+      "loss": 0.3276,
       "step": 10270
     },
     {
-      "epoch": 1.5054945054945055,
-      "grad_norm": 0.34412254714811025,
-      "learning_rate": 1.1673951695647761e-05,
-      "loss": 0.3172,
+      "epoch": 1.5125864860886207,
+      "grad_norm": 0.34736576685366,
+      "learning_rate": 1.1592168370621251e-05,
+      "loss": 0.3297,
       "step": 10275
     },
     {
-      "epoch": 1.5062271062271062,
-      "grad_norm": 0.37108875475539005,
-      "learning_rate": 1.1665546941275508e-05,
-      "loss": 0.3163,
+      "epoch": 1.5133225379066686,
+      "grad_norm": 0.3648936267082656,
+      "learning_rate": 1.1583712639823145e-05,
+      "loss": 0.327,
       "step": 10280
     },
     {
-      "epoch": 1.506959706959707,
-      "grad_norm": 0.33443059117591795,
-      "learning_rate": 1.1657140976619139e-05,
-      "loss": 0.3133,
+      "epoch": 1.5140585897247165,
+      "grad_norm": 0.36467424229661827,
+      "learning_rate": 1.1575255747388973e-05,
+      "loss": 0.3132,
       "step": 10285
     },
     {
-      "epoch": 1.5076923076923077,
-      "grad_norm": 0.3696466136196227,
-      "learning_rate": 1.1648733807786924e-05,
-      "loss": 0.3106,
+      "epoch": 1.5147946415427647,
+      "grad_norm": 0.35051553540759905,
+      "learning_rate": 1.1566797699521767e-05,
+      "loss": 0.3119,
       "step": 10290
     },
     {
-      "epoch": 1.5084249084249084,
-      "grad_norm": 0.36664578097064954,
-      "learning_rate": 1.1640325440888e-05,
-      "loss": 0.3304,
+      "epoch": 1.5155306933608126,
+      "grad_norm": 0.36720314239688623,
+      "learning_rate": 1.155833850242542e-05,
+      "loss": 0.3301,
       "step": 10295
     },
     {
-      "epoch": 1.5091575091575091,
-      "grad_norm": 0.36115273815475946,
-      "learning_rate": 1.1631915882032385e-05,
-      "loss": 0.3339,
+      "epoch": 1.5162667451788607,
+      "grad_norm": 0.35964944120373893,
+      "learning_rate": 1.1549878162304658e-05,
+      "loss": 0.3122,
       "step": 10300
     },
     {
-      "epoch": 1.5098901098901099,
-      "grad_norm": 0.3505883530659083,
-      "learning_rate": 1.1623505137330953e-05,
-      "loss": 0.316,
+      "epoch": 1.5170027969969087,
+      "grad_norm": 0.36040990796822525,
+      "learning_rate": 1.154141668536505e-05,
+      "loss": 0.3341,
       "step": 10305
     },
     {
-      "epoch": 1.5106227106227106,
-      "grad_norm": 0.34948126102936344,
-      "learning_rate": 1.1615093212895447e-05,
-      "loss": 0.327,
+      "epoch": 1.5177388488149566,
+      "grad_norm": 0.37975850258227956,
+      "learning_rate": 1.1532954077812994e-05,
+      "loss": 0.3231,
       "step": 10310
     },
     {
-      "epoch": 1.5113553113553113,
-      "grad_norm": 0.35469293234832766,
-      "learning_rate": 1.1606680114838466e-05,
-      "loss": 0.313,
+      "epoch": 1.5184749006330045,
+      "grad_norm": 0.34035083714004977,
+      "learning_rate": 1.1524490345855725e-05,
+      "loss": 0.3199,
       "step": 10315
     },
     {
-      "epoch": 1.512087912087912,
-      "grad_norm": 0.3609675556063852,
-      "learning_rate": 1.1598265849273461e-05,
-      "loss": 0.323,
+      "epoch": 1.5192109524510524,
+      "grad_norm": 0.3499332403239346,
+      "learning_rate": 1.15160254957013e-05,
+      "loss": 0.3138,
       "step": 10320
     },
     {
-      "epoch": 1.5128205128205128,
-      "grad_norm": 0.36927530181563994,
-      "learning_rate": 1.1589850422314724e-05,
-      "loss": 0.3216,
+      "epoch": 1.5199470042691006,
+      "grad_norm": 0.3477176687159296,
+      "learning_rate": 1.1507559533558592e-05,
+      "loss": 0.3071,
       "step": 10325
     },
     {
-      "epoch": 1.5135531135531135,
-      "grad_norm": 0.36973297374522507,
-      "learning_rate": 1.1581433840077407e-05,
-      "loss": 0.3197,
+      "epoch": 1.5206830560871485,
+      "grad_norm": 0.3490031242582119,
+      "learning_rate": 1.1499092465637292e-05,
+      "loss": 0.3187,
       "step": 10330
     },
     {
-      "epoch": 1.5142857142857142,
-      "grad_norm": 0.3539413717678107,
-      "learning_rate": 1.1573016108677488e-05,
-      "loss": 0.3254,
+      "epoch": 1.5214191079051966,
+      "grad_norm": 0.371079624204397,
+      "learning_rate": 1.1490624298147903e-05,
+      "loss": 0.3224,
       "step": 10335
     },
     {
-      "epoch": 1.515018315018315,
-      "grad_norm": 0.3629618588684632,
-      "learning_rate": 1.1564597234231784e-05,
-      "loss": 0.3162,
+      "epoch": 1.5221551597232446,
+      "grad_norm": 0.35838804760470716,
+      "learning_rate": 1.1482155037301737e-05,
+      "loss": 0.3389,
       "step": 10340
     },
     {
-      "epoch": 1.5157509157509157,
-      "grad_norm": 0.36103904069581366,
-      "learning_rate": 1.1556177222857943e-05,
-      "loss": 0.3421,
+      "epoch": 1.5228912115412925,
+      "grad_norm": 0.3681521204329716,
+      "learning_rate": 1.1473684689310902e-05,
+      "loss": 0.3233,
       "step": 10345
     },
     {
-      "epoch": 1.5164835164835164,
-      "grad_norm": 0.36482620207020305,
-      "learning_rate": 1.1547756080674443e-05,
-      "loss": 0.3322,
+      "epoch": 1.5236272633593404,
+      "grad_norm": 0.3662058095887824,
+      "learning_rate": 1.1465213260388306e-05,
+      "loss": 0.3177,
       "step": 10350
     },
     {
-      "epoch": 1.5172161172161172,
-      "grad_norm": 0.37832423909240065,
-      "learning_rate": 1.1539333813800574e-05,
-      "loss": 0.3312,
+      "epoch": 1.5243633151773883,
+      "grad_norm": 0.3589300711209559,
+      "learning_rate": 1.1456740756747652e-05,
+      "loss": 0.3107,
       "step": 10355
     },
     {
-      "epoch": 1.5179487179487179,
-      "grad_norm": 0.363637168839375,
-      "learning_rate": 1.1530910428356456e-05,
-      "loss": 0.3137,
+      "epoch": 1.5250993669954365,
+      "grad_norm": 0.363259271248277,
+      "learning_rate": 1.1448267184603434e-05,
+      "loss": 0.3197,
       "step": 10360
     },
     {
-      "epoch": 1.5186813186813186,
-      "grad_norm": 0.36076878236286214,
-      "learning_rate": 1.1522485930463008e-05,
-      "loss": 0.3249,
+      "epoch": 1.5258354188134846,
+      "grad_norm": 0.3616125954800502,
+      "learning_rate": 1.143979255017092e-05,
+      "loss": 0.2999,
       "step": 10365
     },
     {
-      "epoch": 1.5194139194139193,
-      "grad_norm": 0.3530507898066571,
-      "learning_rate": 1.151406032624197e-05,
-      "loss": 0.3205,
+      "epoch": 1.5265714706315325,
+      "grad_norm": 0.3831938753206401,
+      "learning_rate": 1.1431316859666161e-05,
+      "loss": 0.3276,
       "step": 10370
     },
     {
-      "epoch": 1.52014652014652,
-      "grad_norm": 0.3627256562251993,
-      "learning_rate": 1.1505633621815874e-05,
-      "loss": 0.3267,
+      "epoch": 1.5273075224495805,
+      "grad_norm": 0.3828513312969688,
+      "learning_rate": 1.142284011930599e-05,
+      "loss": 0.3434,
       "step": 10375
     },
     {
-      "epoch": 1.5208791208791208,
-      "grad_norm": 0.3826960538358999,
-      "learning_rate": 1.149720582330807e-05,
-      "loss": 0.3289,
+      "epoch": 1.5280435742676284,
+      "grad_norm": 0.36639232431420626,
+      "learning_rate": 1.1414362335307999e-05,
+      "loss": 0.3253,
       "step": 10380
     },
     {
-      "epoch": 1.5216117216117215,
-      "grad_norm": 0.363760880261881,
-      "learning_rate": 1.1488776936842677e-05,
-      "loss": 0.3188,
+      "epoch": 1.5287796260856763,
+      "grad_norm": 0.34079896688551425,
+      "learning_rate": 1.1405883513890548e-05,
+      "loss": 0.3182,
       "step": 10385
     },
     {
-      "epoch": 1.5223443223443223,
-      "grad_norm": 0.3441238001593689,
-      "learning_rate": 1.1480346968544625e-05,
-      "loss": 0.3149,
+      "epoch": 1.5295156779037244,
+      "grad_norm": 0.368570821282134,
+      "learning_rate": 1.139740366127277e-05,
+      "loss": 0.33,
       "step": 10390
     },
     {
-      "epoch": 1.523076923076923,
-      "grad_norm": 0.36243279548096724,
-      "learning_rate": 1.1471915924539618e-05,
-      "loss": 0.3367,
+      "epoch": 1.5302517297217724,
+      "grad_norm": 0.552368359972773,
+      "learning_rate": 1.1388922783674541e-05,
+      "loss": 0.3125,
       "step": 10395
     },
     {
-      "epoch": 1.5238095238095237,
-      "grad_norm": 0.35259506480546055,
-      "learning_rate": 1.1463483810954156e-05,
-      "loss": 0.3104,
+      "epoch": 1.5309877815398205,
+      "grad_norm": 0.3678382956172097,
+      "learning_rate": 1.138044088731649e-05,
+      "loss": 0.3217,
       "step": 10400
     },
     {
-      "epoch": 1.5245421245421245,
-      "grad_norm": 0.3586045539694791,
-      "learning_rate": 1.1455050633915505e-05,
-      "loss": 0.3142,
+      "epoch": 1.5317238333578684,
+      "grad_norm": 0.3755751095061701,
+      "learning_rate": 1.137195797842e-05,
+      "loss": 0.3378,
       "step": 10405
     },
     {
-      "epoch": 1.5252747252747252,
-      "grad_norm": 0.35840004554520566,
-      "learning_rate": 1.14466163995517e-05,
-      "loss": 0.3004,
+      "epoch": 1.5324598851759164,
+      "grad_norm": 0.36753548836349775,
+      "learning_rate": 1.1363474063207195e-05,
+      "loss": 0.3196,
       "step": 10410
     },
     {
-      "epoch": 1.526007326007326,
-      "grad_norm": 0.36577144977124004,
-      "learning_rate": 1.1438181113991558e-05,
-      "loss": 0.3227,
+      "epoch": 1.5331959369939643,
+      "grad_norm": 0.36975794388840644,
+      "learning_rate": 1.1354989147900925e-05,
+      "loss": 0.3343,
       "step": 10415
     },
     {
-      "epoch": 1.5267399267399266,
-      "grad_norm": 0.3639942415883003,
-      "learning_rate": 1.1429744783364649e-05,
-      "loss": 0.3328,
+      "epoch": 1.5339319888120122,
+      "grad_norm": 0.3505498342913042,
+      "learning_rate": 1.1346503238724793e-05,
+      "loss": 0.3246,
       "step": 10420
     },
     {
-      "epoch": 1.5274725274725274,
-      "grad_norm": 0.3847428994647183,
-      "learning_rate": 1.1421307413801303e-05,
-      "loss": 0.3347,
+      "epoch": 1.5346680406300603,
+      "grad_norm": 0.3589513304964758,
+      "learning_rate": 1.1338016341903117e-05,
+      "loss": 0.325,
       "step": 10425
     },
     {
-      "epoch": 1.528205128205128,
-      "grad_norm": 0.3785174667722064,
-      "learning_rate": 1.1412869011432613e-05,
-      "loss": 0.3242,
+      "epoch": 1.5354040924481085,
+      "grad_norm": 1.4329025323902576,
+      "learning_rate": 1.1329528463660947e-05,
+      "loss": 0.3158,
       "step": 10430
     },
     {
-      "epoch": 1.5289377289377288,
-      "grad_norm": 0.3585718814838943,
-      "learning_rate": 1.1404429582390413e-05,
-      "loss": 0.3083,
+      "epoch": 1.5361401442661564,
+      "grad_norm": 0.365909611883646,
+      "learning_rate": 1.1321039610224047e-05,
+      "loss": 0.3306,
       "step": 10435
     },
     {
-      "epoch": 1.5296703296703296,
-      "grad_norm": 0.3543393918113152,
-      "learning_rate": 1.1395989132807287e-05,
-      "loss": 0.3113,
+      "epoch": 1.5368761960842043,
+      "grad_norm": 0.36666338483023514,
+      "learning_rate": 1.1312549787818898e-05,
+      "loss": 0.3283,
       "step": 10440
     },
     {
-      "epoch": 1.5304029304029303,
-      "grad_norm": 0.3679860017169941,
-      "learning_rate": 1.138754766881657e-05,
-      "loss": 0.3252,
+      "epoch": 1.5376122479022523,
+      "grad_norm": 0.3634710371258079,
+      "learning_rate": 1.1304059002672696e-05,
+      "loss": 0.3163,
       "step": 10445
     },
     {
-      "epoch": 1.531135531135531,
-      "grad_norm": 0.3493482479821301,
-      "learning_rate": 1.1379105196552311e-05,
-      "loss": 0.3156,
+      "epoch": 1.5383482997203002,
+      "grad_norm": 0.365589652938516,
+      "learning_rate": 1.129556726101334e-05,
+      "loss": 0.325,
       "step": 10450
     },
     {
-      "epoch": 1.5318681318681318,
-      "grad_norm": 0.36896222763653935,
-      "learning_rate": 1.1370661722149314e-05,
-      "loss": 0.3365,
+      "epoch": 1.5390843515383483,
+      "grad_norm": 0.3835416876507009,
+      "learning_rate": 1.1287074569069427e-05,
+      "loss": 0.3265,
       "step": 10455
     },
     {
-      "epoch": 1.5326007326007325,
-      "grad_norm": 0.36606842295285735,
-      "learning_rate": 1.1362217251743099e-05,
-      "loss": 0.3247,
+      "epoch": 1.5398204033563963,
+      "grad_norm": 0.3592191576539692,
+      "learning_rate": 1.1278580933070257e-05,
+      "loss": 0.3203,
       "step": 10460
     },
     {
-      "epoch": 1.5333333333333332,
-      "grad_norm": 0.3634722621767758,
-      "learning_rate": 1.135377179146992e-05,
-      "loss": 0.3263,
+      "epoch": 1.5405564551744444,
+      "grad_norm": 0.35570709486871005,
+      "learning_rate": 1.1270086359245822e-05,
+      "loss": 0.3083,
       "step": 10465
     },
     {
-      "epoch": 1.534065934065934,
-      "grad_norm": 0.36243529766864707,
-      "learning_rate": 1.1345325347466738e-05,
-      "loss": 0.3228,
+      "epoch": 1.5412925069924923,
+      "grad_norm": 0.3834732048865809,
+      "learning_rate": 1.1261590853826795e-05,
+      "loss": 0.34,
       "step": 10470
     },
     {
-      "epoch": 1.5347985347985347,
-      "grad_norm": 0.3510096037048702,
-      "learning_rate": 1.1336877925871237e-05,
-      "loss": 0.3231,
+      "epoch": 1.5420285588105402,
+      "grad_norm": 0.36468683586721007,
+      "learning_rate": 1.1253094423044542e-05,
+      "loss": 0.3217,
       "step": 10475
     },
     {
-      "epoch": 1.5355311355311354,
-      "grad_norm": 0.3569480778437148,
-      "learning_rate": 1.1328429532821806e-05,
-      "loss": 0.325,
+      "epoch": 1.5427646106285882,
+      "grad_norm": 0.3717550741983341,
+      "learning_rate": 1.1244597073131101e-05,
+      "loss": 0.3334,
       "step": 10480
     },
     {
-      "epoch": 1.5362637362637361,
-      "grad_norm": 0.37322684386375793,
-      "learning_rate": 1.1319980174457546e-05,
-      "loss": 0.3202,
+      "epoch": 1.543500662446636,
+      "grad_norm": 0.3607937943081953,
+      "learning_rate": 1.1236098810319182e-05,
+      "loss": 0.3187,
       "step": 10485
     },
     {
-      "epoch": 1.5369963369963369,
-      "grad_norm": 0.36709767538456534,
-      "learning_rate": 1.1311529856918258e-05,
-      "loss": 0.3284,
+      "epoch": 1.5442367142646842,
+      "grad_norm": 0.3638012987883055,
+      "learning_rate": 1.1227599640842178e-05,
+      "loss": 0.3322,
       "step": 10490
     },
     {
-      "epoch": 1.5377289377289376,
-      "grad_norm": 0.36067118105211804,
-      "learning_rate": 1.1303078586344434e-05,
-      "loss": 0.3175,
+      "epoch": 1.5449727660827324,
+      "grad_norm": 0.3513757646645374,
+      "learning_rate": 1.1219099570934131e-05,
+      "loss": 0.3109,
       "step": 10495
     },
     {
-      "epoch": 1.5384615384615383,
-      "grad_norm": 0.3632107295175718,
-      "learning_rate": 1.1294626368877266e-05,
-      "loss": 0.3204,
+      "epoch": 1.5457088179007803,
+      "grad_norm": 0.37690237543820104,
+      "learning_rate": 1.121059860682975e-05,
+      "loss": 0.3328,
       "step": 10500
     },
     {
-      "epoch": 1.539194139194139,
-      "grad_norm": 0.3652132794095445,
-      "learning_rate": 1.128617321065863e-05,
-      "loss": 0.319,
+      "epoch": 1.5464448697188282,
+      "grad_norm": 0.36143736312535396,
+      "learning_rate": 1.1202096754764403e-05,
+      "loss": 0.3239,
       "step": 10505
     },
     {
-      "epoch": 1.5399267399267398,
-      "grad_norm": 0.38779027451907705,
-      "learning_rate": 1.1277719117831088e-05,
-      "loss": 0.3406,
+      "epoch": 1.5471809215368761,
+      "grad_norm": 0.36181670331350985,
+      "learning_rate": 1.1193594020974105e-05,
+      "loss": 0.3115,
       "step": 10510
     },
     {
-      "epoch": 1.5406593406593405,
-      "grad_norm": 0.3884935873103975,
-      "learning_rate": 1.126926409653788e-05,
-      "loss": 0.3294,
+      "epoch": 1.547916973354924,
+      "grad_norm": 0.3702153858630142,
+      "learning_rate": 1.1185090411695519e-05,
+      "loss": 0.3188,
       "step": 10515
     },
     {
-      "epoch": 1.5413919413919412,
-      "grad_norm": 0.3455480192963863,
-      "learning_rate": 1.1260808152922916e-05,
-      "loss": 0.3074,
+      "epoch": 1.5486530251729722,
+      "grad_norm": 0.3581307229029183,
+      "learning_rate": 1.1176585933165949e-05,
+      "loss": 0.3245,
       "step": 10520
     },
     {
-      "epoch": 1.542124542124542,
-      "grad_norm": 0.3502506088120491,
-      "learning_rate": 1.1252351293130785e-05,
-      "loss": 0.3221,
+      "epoch": 1.5493890769910201,
+      "grad_norm": 0.3682623206651115,
+      "learning_rate": 1.1168080591623343e-05,
+      "loss": 0.3251,
       "step": 10525
     },
     {
-      "epoch": 1.5428571428571427,
-      "grad_norm": 0.3527800077691515,
-      "learning_rate": 1.1243893523306735e-05,
-      "loss": 0.327,
+      "epoch": 1.5501251288090683,
+      "grad_norm": 0.35429114443960585,
+      "learning_rate": 1.1159574393306272e-05,
+      "loss": 0.3269,
       "step": 10530
     },
     {
-      "epoch": 1.5435897435897434,
-      "grad_norm": 0.34779222936948845,
-      "learning_rate": 1.1235434849596683e-05,
-      "loss": 0.3221,
+      "epoch": 1.5508611806271162,
+      "grad_norm": 0.3661458881367873,
+      "learning_rate": 1.1151067344453941e-05,
+      "loss": 0.3362,
       "step": 10535
     },
     {
-      "epoch": 1.5443223443223442,
-      "grad_norm": 0.3581405926693141,
-      "learning_rate": 1.1226975278147191e-05,
-      "loss": 0.3249,
+      "epoch": 1.5515972324451641,
+      "grad_norm": 0.3401509268807909,
+      "learning_rate": 1.1142559451306183e-05,
+      "loss": 0.3248,
       "step": 10540
     },
     {
-      "epoch": 1.545054945054945,
-      "grad_norm": 0.3652384063913515,
-      "learning_rate": 1.1218514815105483e-05,
-      "loss": 0.3362,
+      "epoch": 1.552333284263212,
+      "grad_norm": 0.3473392948542666,
+      "learning_rate": 1.1134050720103438e-05,
+      "loss": 0.3285,
       "step": 10545
     },
     {
-      "epoch": 1.5457875457875456,
-      "grad_norm": 0.3718363796824984,
-      "learning_rate": 1.1210053466619427e-05,
-      "loss": 0.3251,
+      "epoch": 1.55306933608126,
+      "grad_norm": 0.36994117748891175,
+      "learning_rate": 1.112554115708677e-05,
+      "loss": 0.3128,
       "step": 10550
     },
     {
-      "epoch": 1.5465201465201464,
-      "grad_norm": 0.3617586012853975,
-      "learning_rate": 1.1201591238837536e-05,
-      "loss": 0.321,
+      "epoch": 1.553805387899308,
+      "grad_norm": 0.33668187505090125,
+      "learning_rate": 1.1117030768497854e-05,
+      "loss": 0.3254,
       "step": 10555
     },
     {
-      "epoch": 1.5472527472527473,
-      "grad_norm": 0.35963645658759585,
-      "learning_rate": 1.1193128137908962e-05,
-      "loss": 0.3317,
+      "epoch": 1.5545414397173563,
+      "grad_norm": 0.3618387095405319,
+      "learning_rate": 1.110851956057897e-05,
+      "loss": 0.3316,
       "step": 10560
     },
     {
-      "epoch": 1.547985347985348,
-      "grad_norm": 0.37752854864033075,
-      "learning_rate": 1.1184664169983486e-05,
-      "loss": 0.3276,
+      "epoch": 1.5552774915354042,
+      "grad_norm": 0.3491498712906266,
+      "learning_rate": 1.110000753957299e-05,
+      "loss": 0.3369,
       "step": 10565
     },
     {
-      "epoch": 1.5487179487179488,
-      "grad_norm": 0.38661437404420046,
-      "learning_rate": 1.117619934121153e-05,
-      "loss": 0.3261,
+      "epoch": 1.556013543353452,
+      "grad_norm": 0.3495879911555467,
+      "learning_rate": 1.1091494711723395e-05,
+      "loss": 0.3171,
       "step": 10570
     },
     {
-      "epoch": 1.5494505494505495,
-      "grad_norm": 0.36685370857951194,
-      "learning_rate": 1.1167733657744131e-05,
-      "loss": 0.3209,
+      "epoch": 1.5567495951715,
+      "grad_norm": 0.3907826414201101,
+      "learning_rate": 1.1082981083274252e-05,
+      "loss": 0.3196,
       "step": 10575
     },
     {
-      "epoch": 1.5501831501831502,
-      "grad_norm": 0.3686733422250125,
-      "learning_rate": 1.1159267125732954e-05,
-      "loss": 0.3192,
+      "epoch": 1.557485646989548,
+      "grad_norm": 0.36974060030759315,
+      "learning_rate": 1.107446666047021e-05,
+      "loss": 0.322,
       "step": 10580
     },
     {
-      "epoch": 1.550915750915751,
-      "grad_norm": 0.35997746733424546,
-      "learning_rate": 1.1150799751330273e-05,
-      "loss": 0.304,
+      "epoch": 1.558221698807596,
+      "grad_norm": 0.364638380907563,
+      "learning_rate": 1.1065951449556509e-05,
+      "loss": 0.3177,
       "step": 10585
     },
     {
-      "epoch": 1.5516483516483517,
-      "grad_norm": 0.35912020225512303,
-      "learning_rate": 1.1142331540688982e-05,
-      "loss": 0.3182,
+      "epoch": 1.558957750625644,
+      "grad_norm": 0.36830616107237507,
+      "learning_rate": 1.1057435456778967e-05,
+      "loss": 0.3169,
       "step": 10590
     },
     {
-      "epoch": 1.5523809523809524,
-      "grad_norm": 0.3704574449036329,
-      "learning_rate": 1.113386249996258e-05,
-      "loss": 0.3055,
+      "epoch": 1.5596938024436922,
+      "grad_norm": 0.36250697632861917,
+      "learning_rate": 1.1048918688383967e-05,
+      "loss": 0.3305,
       "step": 10595
     },
     {
-      "epoch": 1.5531135531135531,
-      "grad_norm": 0.37596344315036007,
-      "learning_rate": 1.1125392635305171e-05,
-      "loss": 0.325,
+      "epoch": 1.56042985426174,
+      "grad_norm": 0.35473255475644067,
+      "learning_rate": 1.104040115061847e-05,
+      "loss": 0.3196,
       "step": 10600
     },
     {
-      "epoch": 1.5538461538461539,
-      "grad_norm": 0.37007140282096485,
-      "learning_rate": 1.1116921952871451e-05,
-      "loss": 0.3304,
+      "epoch": 1.561165906079788,
+      "grad_norm": 0.3697377467377593,
+      "learning_rate": 1.1031882849729996e-05,
+      "loss": 0.3196,
       "step": 10605
     },
     {
-      "epoch": 1.5545787545787546,
-      "grad_norm": 0.3542567239055139,
-      "learning_rate": 1.110845045881672e-05,
-      "loss": 0.3483,
+      "epoch": 1.561901957897836,
+      "grad_norm": 0.35367404218811516,
+      "learning_rate": 1.1023363791966628e-05,
+      "loss": 0.3191,
       "step": 10610
     },
     {
-      "epoch": 1.5553113553113553,
-      "grad_norm": 0.3704260386299412,
-      "learning_rate": 1.1099978159296853e-05,
-      "loss": 0.3354,
+      "epoch": 1.5626380097158838,
+      "grad_norm": 0.37424352447772774,
+      "learning_rate": 1.1014843983577002e-05,
+      "loss": 0.3441,
       "step": 10615
     },
     {
-      "epoch": 1.556043956043956,
-      "grad_norm": 0.35617120318620293,
-      "learning_rate": 1.1091505060468334e-05,
-      "loss": 0.3145,
+      "epoch": 1.563374061533932,
+      "grad_norm": 0.3507694301126513,
+      "learning_rate": 1.1006323430810304e-05,
+      "loss": 0.3193,
       "step": 10620
     },
     {
-      "epoch": 1.5567765567765568,
-      "grad_norm": 0.3630933646411497,
-      "learning_rate": 1.1083031168488201e-05,
-      "loss": 0.3126,
+      "epoch": 1.5641101133519801,
+      "grad_norm": 0.3507120881822479,
+      "learning_rate": 1.0997802139916267e-05,
+      "loss": 0.3262,
       "step": 10625
     },
     {
-      "epoch": 1.5575091575091575,
-      "grad_norm": 0.3635654024668279,
-      "learning_rate": 1.1074556489514088e-05,
-      "loss": 0.3287,
+      "epoch": 1.564846165170028,
+      "grad_norm": 0.360651879708966,
+      "learning_rate": 1.0989280117145168e-05,
+      "loss": 0.3273,
       "step": 10630
     },
     {
-      "epoch": 1.5582417582417583,
-      "grad_norm": 0.36926283154091916,
-      "learning_rate": 1.106608102970419e-05,
-      "loss": 0.3311,
+      "epoch": 1.565582216988076,
+      "grad_norm": 0.35962526178526283,
+      "learning_rate": 1.0980757368747819e-05,
+      "loss": 0.3247,
       "step": 10635
     },
     {
-      "epoch": 1.558974358974359,
-      "grad_norm": 0.38017762750849377,
-      "learning_rate": 1.1057604795217275e-05,
-      "loss": 0.3226,
+      "epoch": 1.566318268806124,
+      "grad_norm": 0.3567122421354603,
+      "learning_rate": 1.097223390097556e-05,
+      "loss": 0.3214,
       "step": 10640
     },
     {
-      "epoch": 1.5597069597069597,
-      "grad_norm": 0.37296679726598,
-      "learning_rate": 1.1049127792212678e-05,
-      "loss": 0.3332,
+      "epoch": 1.5670543206241718,
+      "grad_norm": 0.3707989706920427,
+      "learning_rate": 1.0963709720080262e-05,
+      "loss": 0.3221,
       "step": 10645
     },
     {
-      "epoch": 1.5604395604395604,
-      "grad_norm": 0.37025087485170594,
-      "learning_rate": 1.1040650026850274e-05,
-      "loss": 0.325,
+      "epoch": 1.56779037244222,
+      "grad_norm": 0.3639122505804451,
+      "learning_rate": 1.0955184832314323e-05,
+      "loss": 0.3276,
       "step": 10650
     },
     {
-      "epoch": 1.5611721611721612,
-      "grad_norm": 0.3599046109588884,
-      "learning_rate": 1.1032171505290515e-05,
-      "loss": 0.3284,
+      "epoch": 1.568526424260268,
+      "grad_norm": 0.3603234990700586,
+      "learning_rate": 1.0946659243930653e-05,
+      "loss": 0.3027,
       "step": 10655
     },
     {
-      "epoch": 1.561904761904762,
-      "grad_norm": 0.3713491859565942,
-      "learning_rate": 1.102369223369439e-05,
-      "loss": 0.3191,
+      "epoch": 1.569262476078316,
+      "grad_norm": 0.36826062208996563,
+      "learning_rate": 1.0938132961182682e-05,
+      "loss": 0.3132,
       "step": 10660
     },
     {
-      "epoch": 1.5626373626373626,
-      "grad_norm": 0.3977929765942424,
-      "learning_rate": 1.1015212218223433e-05,
-      "loss": 0.3326,
+      "epoch": 1.569998527896364,
+      "grad_norm": 0.3550575067788097,
+      "learning_rate": 1.0929605990324345e-05,
+      "loss": 0.3158,
       "step": 10665
     },
     {
-      "epoch": 1.5633699633699634,
-      "grad_norm": 0.38423993709303933,
-      "learning_rate": 1.100673146503972e-05,
-      "loss": 0.3239,
+      "epoch": 1.5707345797144119,
+      "grad_norm": 0.3482416489950832,
+      "learning_rate": 1.0921078337610083e-05,
+      "loss": 0.3088,
       "step": 10670
     },
     {
-      "epoch": 1.564102564102564,
-      "grad_norm": 0.36288425856840323,
-      "learning_rate": 1.0998249980305867e-05,
-      "loss": 0.3256,
+      "epoch": 1.5714706315324598,
+      "grad_norm": 0.35761848210115393,
+      "learning_rate": 1.0912550009294834e-05,
+      "loss": 0.3368,
       "step": 10675
     },
     {
-      "epoch": 1.5648351648351648,
-      "grad_norm": 0.350555719189037,
-      "learning_rate": 1.0989767770185017e-05,
-      "loss": 0.3096,
+      "epoch": 1.5722066833505077,
+      "grad_norm": 0.3763713827683073,
+      "learning_rate": 1.0904021011634033e-05,
+      "loss": 0.3423,
       "step": 10680
     },
     {
-      "epoch": 1.5655677655677656,
-      "grad_norm": 0.38318092351307165,
-      "learning_rate": 1.098128484084084e-05,
-      "loss": 0.3324,
+      "epoch": 1.5729427351685559,
+      "grad_norm": 0.37829590035732397,
+      "learning_rate": 1.0895491350883612e-05,
+      "loss": 0.3208,
       "step": 10685
     },
     {
-      "epoch": 1.5663003663003663,
-      "grad_norm": 0.3618784359736923,
-      "learning_rate": 1.0972801198437533e-05,
-      "loss": 0.3229,
+      "epoch": 1.573678786986604,
+      "grad_norm": 0.36432735184879916,
+      "learning_rate": 1.0886961033299985e-05,
+      "loss": 0.3178,
       "step": 10690
     },
     {
-      "epoch": 1.567032967032967,
-      "grad_norm": 0.37990979190549173,
-      "learning_rate": 1.0964316849139806e-05,
-      "loss": 0.3229,
+      "epoch": 1.574414838804652,
+      "grad_norm": 0.37211183953152666,
+      "learning_rate": 1.0878430065140045e-05,
+      "loss": 0.3202,
       "step": 10695
     },
     {
-      "epoch": 1.5677655677655677,
-      "grad_norm": 0.3640985661929933,
-      "learning_rate": 1.0955831799112886e-05,
-      "loss": 0.3176,
+      "epoch": 1.5751508906226999,
+      "grad_norm": 0.3667957151162588,
+      "learning_rate": 1.0869898452661167e-05,
+      "loss": 0.3222,
       "step": 10700
     },
     {
-      "epoch": 1.5684981684981685,
-      "grad_norm": 0.37439906801966494,
-      "learning_rate": 1.0947346054522508e-05,
-      "loss": 0.3175,
+      "epoch": 1.5758869424407478,
+      "grad_norm": 0.35726128301731563,
+      "learning_rate": 1.0861366202121193e-05,
+      "loss": 0.3261,
       "step": 10705
     },
     {
-      "epoch": 1.5692307692307692,
-      "grad_norm": 0.3669987106800802,
-      "learning_rate": 1.093885962153491e-05,
-      "loss": 0.3235,
+      "epoch": 1.5766229942587957,
+      "grad_norm": 0.35733711708395866,
+      "learning_rate": 1.0852833319778442e-05,
+      "loss": 0.3171,
       "step": 10710
     },
     {
-      "epoch": 1.56996336996337,
-      "grad_norm": 0.3559820665989296,
-      "learning_rate": 1.0930372506316837e-05,
-      "loss": 0.33,
+      "epoch": 1.5773590460768439,
+      "grad_norm": 0.363055551154169,
+      "learning_rate": 1.0844299811891686e-05,
+      "loss": 0.3165,
       "step": 10715
     },
     {
-      "epoch": 1.5706959706959707,
-      "grad_norm": 0.35190310727973406,
-      "learning_rate": 1.0921884715035518e-05,
-      "loss": 0.3228,
+      "epoch": 1.5780950978948918,
+      "grad_norm": 0.35831292835177575,
+      "learning_rate": 1.0835765684720159e-05,
+      "loss": 0.322,
       "step": 10720
     },
     {
-      "epoch": 1.5714285714285714,
-      "grad_norm": 0.36359573484072494,
-      "learning_rate": 1.0913396253858685e-05,
-      "loss": 0.3341,
+      "epoch": 1.57883114971294,
+      "grad_norm": 0.36235266345268446,
+      "learning_rate": 1.0827230944523555e-05,
+      "loss": 0.3174,
       "step": 10725
     },
     {
-      "epoch": 1.5721611721611721,
-      "grad_norm": 0.355697692121416,
-      "learning_rate": 1.0904907128954551e-05,
-      "loss": 0.3383,
+      "epoch": 1.5795672015309878,
+      "grad_norm": 0.35646086170906893,
+      "learning_rate": 1.0818695597562012e-05,
+      "loss": 0.3119,
       "step": 10730
     },
     {
-      "epoch": 1.5728937728937729,
-      "grad_norm": 0.3461701669379614,
-      "learning_rate": 1.0896417346491811e-05,
-      "loss": 0.3188,
+      "epoch": 1.5803032533490358,
+      "grad_norm": 0.3698149547592337,
+      "learning_rate": 1.0810159650096116e-05,
+      "loss": 0.3219,
       "step": 10735
     },
     {
-      "epoch": 1.5736263736263736,
-      "grad_norm": 0.366987635423063,
-      "learning_rate": 1.088792691263964e-05,
-      "loss": 0.3222,
+      "epoch": 1.5810393051670837,
+      "grad_norm": 0.3475767326719413,
+      "learning_rate": 1.080162310838689e-05,
+      "loss": 0.3173,
       "step": 10740
     },
     {
-      "epoch": 1.5743589743589743,
-      "grad_norm": 0.37928410154460174,
-      "learning_rate": 1.0879435833567683e-05,
-      "loss": 0.3234,
+      "epoch": 1.5817753569851316,
+      "grad_norm": 0.3468615818674762,
+      "learning_rate": 1.0793085978695797e-05,
+      "loss": 0.307,
       "step": 10745
     },
     {
-      "epoch": 1.575091575091575,
-      "grad_norm": 0.37405645256523873,
-      "learning_rate": 1.0870944115446062e-05,
-      "loss": 0.3238,
+      "epoch": 1.5825114088031798,
+      "grad_norm": 0.35105289247848803,
+      "learning_rate": 1.0784548267284727e-05,
+      "loss": 0.3435,
       "step": 10750
     },
     {
-      "epoch": 1.5758241758241758,
-      "grad_norm": 0.3683261076337646,
-      "learning_rate": 1.0862451764445352e-05,
-      "loss": 0.3336,
+      "epoch": 1.583247460621228,
+      "grad_norm": 0.3506312110225094,
+      "learning_rate": 1.0776009980415997e-05,
+      "loss": 0.3219,
       "step": 10755
     },
     {
-      "epoch": 1.5765567765567765,
-      "grad_norm": 0.3634826374067524,
-      "learning_rate": 1.0853958786736592e-05,
-      "loss": 0.3251,
+      "epoch": 1.5839835124392758,
+      "grad_norm": 0.3785318098001804,
+      "learning_rate": 1.0767471124352355e-05,
+      "loss": 0.3143,
       "step": 10760
     },
     {
-      "epoch": 1.5772893772893772,
-      "grad_norm": 0.3500778650743875,
-      "learning_rate": 1.0845465188491284e-05,
-      "loss": 0.3223,
+      "epoch": 1.5847195642573237,
+      "grad_norm": 0.3814559123042075,
+      "learning_rate": 1.0758931705356952e-05,
+      "loss": 0.3272,
       "step": 10765
     },
     {
-      "epoch": 1.578021978021978,
-      "grad_norm": 0.35777802621337773,
-      "learning_rate": 1.0836970975881369e-05,
-      "loss": 0.3079,
+      "epoch": 1.5854556160753717,
+      "grad_norm": 0.3698078447907398,
+      "learning_rate": 1.0750391729693363e-05,
+      "loss": 0.3219,
       "step": 10770
     },
     {
-      "epoch": 1.578754578754579,
-      "grad_norm": 0.35417650772038833,
-      "learning_rate": 1.0828476155079248e-05,
-      "loss": 0.3199,
+      "epoch": 1.5861916678934196,
+      "grad_norm": 0.36454633834510386,
+      "learning_rate": 1.0741851203625569e-05,
+      "loss": 0.328,
       "step": 10775
     },
     {
-      "epoch": 1.5794871794871796,
-      "grad_norm": 0.36711234633272344,
-      "learning_rate": 1.0819980732257747e-05,
-      "loss": 0.3236,
+      "epoch": 1.5869277197114677,
+      "grad_norm": 0.37238642644164754,
+      "learning_rate": 1.0733310133417952e-05,
+      "loss": 0.3151,
       "step": 10780
     },
     {
-      "epoch": 1.5802197802197804,
-      "grad_norm": 0.36720371790919215,
-      "learning_rate": 1.0811484713590143e-05,
-      "loss": 0.3178,
+      "epoch": 1.5876637715295157,
+      "grad_norm": 0.3681230149728814,
+      "learning_rate": 1.0724768525335287e-05,
+      "loss": 0.3211,
       "step": 10785
     },
     {
-      "epoch": 1.580952380952381,
-      "grad_norm": 0.3586111875426069,
-      "learning_rate": 1.0802988105250139e-05,
-      "loss": 0.3243,
+      "epoch": 1.5883998233475638,
+      "grad_norm": 0.3672618834917921,
+      "learning_rate": 1.0716226385642768e-05,
+      "loss": 0.321,
       "step": 10790
     },
     {
-      "epoch": 1.5816849816849818,
-      "grad_norm": 0.34797918189175675,
-      "learning_rate": 1.0794490913411878e-05,
-      "loss": 0.3262,
+      "epoch": 1.5891358751656117,
+      "grad_norm": 0.34443770378247834,
+      "learning_rate": 1.0707683720605946e-05,
+      "loss": 0.3365,
       "step": 10795
     },
     {
-      "epoch": 1.5824175824175826,
-      "grad_norm": 0.35506683705733666,
-      "learning_rate": 1.0785993144249903e-05,
-      "loss": 0.3164,
+      "epoch": 1.5898719269836596,
+      "grad_norm": 0.3385417936764532,
+      "learning_rate": 1.069914053649078e-05,
+      "loss": 0.3161,
       "step": 10800
     },
     {
-      "epoch": 1.5831501831501833,
-      "grad_norm": 0.35529140738617393,
-      "learning_rate": 1.0777494803939204e-05,
-      "loss": 0.3323,
+      "epoch": 1.5906079788017076,
+      "grad_norm": 0.35463070768443467,
+      "learning_rate": 1.0690596839563602e-05,
+      "loss": 0.3255,
       "step": 10805
     },
     {
-      "epoch": 1.583882783882784,
-      "grad_norm": 0.3655902249344232,
-      "learning_rate": 1.0768995898655168e-05,
-      "loss": 0.3194,
+      "epoch": 1.5913440306197555,
+      "grad_norm": 0.3784813860449901,
+      "learning_rate": 1.0682052636091122e-05,
+      "loss": 0.3339,
       "step": 10810
     },
     {
-      "epoch": 1.5846153846153848,
-      "grad_norm": 0.3427492588722873,
-      "learning_rate": 1.0760496434573602e-05,
-      "loss": 0.3196,
+      "epoch": 1.5920800824378036,
+      "grad_norm": 0.3684324877590342,
+      "learning_rate": 1.0673507932340422e-05,
+      "loss": 0.3293,
       "step": 10815
     },
     {
-      "epoch": 1.5853479853479855,
-      "grad_norm": 0.3714957174420384,
-      "learning_rate": 1.0751996417870711e-05,
-      "loss": 0.3213,
+      "epoch": 1.5928161342558516,
+      "grad_norm": 0.35754642686646815,
+      "learning_rate": 1.0664962734578944e-05,
+      "loss": 0.3191,
       "step": 10820
     },
     {
-      "epoch": 1.5860805860805862,
-      "grad_norm": 0.35428875402177135,
-      "learning_rate": 1.074349585472311e-05,
-      "loss": 0.3092,
+      "epoch": 1.5935521860738997,
+      "grad_norm": 0.3633164982949058,
+      "learning_rate": 1.0656417049074504e-05,
+      "loss": 0.317,
       "step": 10825
     },
     {
-      "epoch": 1.586813186813187,
-      "grad_norm": 0.36841388820423904,
-      "learning_rate": 1.0734994751307799e-05,
-      "loss": 0.325,
+      "epoch": 1.5942882378919476,
+      "grad_norm": 0.36734435701073914,
+      "learning_rate": 1.0647870882095266e-05,
+      "loss": 0.3268,
       "step": 10830
     },
     {
-      "epoch": 1.5875457875457877,
-      "grad_norm": 0.3614957303732827,
-      "learning_rate": 1.0726493113802187e-05,
-      "loss": 0.3294,
+      "epoch": 1.5950242897099955,
+      "grad_norm": 0.3591129856402927,
+      "learning_rate": 1.0639324239909753e-05,
+      "loss": 0.3177,
       "step": 10835
     },
     {
-      "epoch": 1.5882783882783884,
-      "grad_norm": 0.3386499493874705,
-      "learning_rate": 1.0717990948384062e-05,
-      "loss": 0.3201,
+      "epoch": 1.5957603415280435,
+      "grad_norm": 0.3540197079602139,
+      "learning_rate": 1.0630777128786839e-05,
+      "loss": 0.3265,
       "step": 10840
     },
     {
-      "epoch": 1.5890109890109891,
-      "grad_norm": 0.3520618352837802,
-      "learning_rate": 1.0709488261231592e-05,
-      "loss": 0.3342,
+      "epoch": 1.5964963933460916,
+      "grad_norm": 0.35862077364823264,
+      "learning_rate": 1.0622229554995727e-05,
+      "loss": 0.3252,
       "step": 10845
     },
     {
-      "epoch": 1.5897435897435899,
-      "grad_norm": 0.36867953482519555,
-      "learning_rate": 1.0700985058523328e-05,
-      "loss": 0.3228,
+      "epoch": 1.5972324451641395,
+      "grad_norm": 0.3712442507726749,
+      "learning_rate": 1.0613681524805977e-05,
+      "loss": 0.3315,
       "step": 10850
     },
     {
-      "epoch": 1.5904761904761906,
-      "grad_norm": 0.35508710260508825,
-      "learning_rate": 1.0692481346438198e-05,
-      "loss": 0.3119,
+      "epoch": 1.5979684969821877,
+      "grad_norm": 0.36313447074643235,
+      "learning_rate": 1.0605133044487473e-05,
+      "loss": 0.3316,
       "step": 10855
     },
     {
-      "epoch": 1.5912087912087913,
-      "grad_norm": 0.35303838750743477,
-      "learning_rate": 1.0683977131155503e-05,
-      "loss": 0.3146,
+      "epoch": 1.5987045488002356,
+      "grad_norm": 0.3685281765703767,
+      "learning_rate": 1.0596584120310436e-05,
+      "loss": 0.3098,
       "step": 10860
     },
     {
-      "epoch": 1.591941391941392,
-      "grad_norm": 0.3618356902258964,
-      "learning_rate": 1.0675472418854894e-05,
-      "loss": 0.3088,
+      "epoch": 1.5994406006182835,
+      "grad_norm": 0.37036564665213156,
+      "learning_rate": 1.0588034758545406e-05,
+      "loss": 0.3189,
       "step": 10865
     },
     {
-      "epoch": 1.5926739926739928,
-      "grad_norm": 0.36667750846484176,
-      "learning_rate": 1.0666967215716403e-05,
-      "loss": 0.3309,
+      "epoch": 1.6001766524363314,
+      "grad_norm": 0.3455809354941265,
+      "learning_rate": 1.0579484965463251e-05,
+      "loss": 0.3185,
       "step": 10870
     },
     {
-      "epoch": 1.5934065934065935,
-      "grad_norm": 0.3592496842622154,
-      "learning_rate": 1.06584615279204e-05,
-      "loss": 0.3298,
+      "epoch": 1.6009127042543794,
+      "grad_norm": 0.3689604249317098,
+      "learning_rate": 1.057093474733515e-05,
+      "loss": 0.3385,
       "step": 10875
     },
     {
-      "epoch": 1.5941391941391942,
-      "grad_norm": 0.3710415634025953,
-      "learning_rate": 1.0649955361647629e-05,
-      "loss": 0.3237,
+      "epoch": 1.6016487560724275,
+      "grad_norm": 0.34771801563235544,
+      "learning_rate": 1.0562384110432595e-05,
+      "loss": 0.3111,
       "step": 10880
     },
     {
-      "epoch": 1.594871794871795,
-      "grad_norm": 0.3735755916958751,
-      "learning_rate": 1.0641448723079161e-05,
-      "loss": 0.3398,
+      "epoch": 1.6023848078904754,
+      "grad_norm": 0.3631979360473968,
+      "learning_rate": 1.0553833061027388e-05,
+      "loss": 0.318,
       "step": 10885
     },
     {
-      "epoch": 1.5956043956043957,
-      "grad_norm": 0.35845605896563393,
-      "learning_rate": 1.0632941618396422e-05,
-      "loss": 0.3265,
+      "epoch": 1.6031208597085236,
+      "grad_norm": 0.36583268761609733,
+      "learning_rate": 1.0545281605391624e-05,
+      "loss": 0.3399,
       "step": 10890
     },
     {
-      "epoch": 1.5963369963369964,
-      "grad_norm": 0.36511863236781306,
-      "learning_rate": 1.062443405378117e-05,
-      "loss": 0.3239,
+      "epoch": 1.6038569115265715,
+      "grad_norm": 0.35643244914326294,
+      "learning_rate": 1.053672974979771e-05,
+      "loss": 0.3158,
       "step": 10895
     },
     {
-      "epoch": 1.5970695970695972,
-      "grad_norm": 0.38137540806176407,
-      "learning_rate": 1.0615926035415505e-05,
-      "loss": 0.3323,
+      "epoch": 1.6045929633446194,
+      "grad_norm": 0.3802049703572021,
+      "learning_rate": 1.0528177500518337e-05,
+      "loss": 0.3368,
       "step": 10900
     },
     {
-      "epoch": 1.597802197802198,
-      "grad_norm": 0.35783473945763244,
-      "learning_rate": 1.060741756948185e-05,
-      "loss": 0.3308,
+      "epoch": 1.6053290151626673,
+      "grad_norm": 0.3465421836301589,
+      "learning_rate": 1.0519624863826487e-05,
+      "loss": 0.3178,
       "step": 10905
     },
     {
-      "epoch": 1.5985347985347986,
-      "grad_norm": 0.35899576905804875,
-      "learning_rate": 1.0598908662162954e-05,
-      "loss": 0.3167,
+      "epoch": 1.6060650669807155,
+      "grad_norm": 0.3566612836576884,
+      "learning_rate": 1.051107184599543e-05,
+      "loss": 0.3159,
       "step": 10910
     },
     {
-      "epoch": 1.5992673992673994,
-      "grad_norm": 0.36913210827949633,
-      "learning_rate": 1.059039931964189e-05,
-      "loss": 0.3191,
+      "epoch": 1.6068011187987634,
+      "grad_norm": 0.3684066675373982,
+      "learning_rate": 1.0502518453298705e-05,
+      "loss": 0.3256,
       "step": 10915
     },
     {
-      "epoch": 1.6,
-      "grad_norm": 0.3673374725819731,
-      "learning_rate": 1.0581889548102045e-05,
-      "loss": 0.3283,
+      "epoch": 1.6075371706168116,
+      "grad_norm": 0.3537753053942633,
+      "learning_rate": 1.0493964692010139e-05,
+      "loss": 0.3221,
       "step": 10920
     },
     {
-      "epoch": 1.6007326007326008,
-      "grad_norm": 0.37097335997045566,
-      "learning_rate": 1.057337935372712e-05,
-      "loss": 0.3326,
+      "epoch": 1.6082732224348595,
+      "grad_norm": 0.36413617741163884,
+      "learning_rate": 1.048541056840382e-05,
+      "loss": 0.3277,
       "step": 10925
     },
     {
-      "epoch": 1.6014652014652015,
-      "grad_norm": 0.3733652663620874,
-      "learning_rate": 1.0564868742701118e-05,
-      "loss": 0.3274,
+      "epoch": 1.6090092742529074,
+      "grad_norm": 0.36908699563989056,
+      "learning_rate": 1.0476856088754109e-05,
+      "loss": 0.3351,
       "step": 10930
     },
     {
-      "epoch": 1.6021978021978023,
-      "grad_norm": 0.3743462647789397,
-      "learning_rate": 1.055635772120835e-05,
-      "loss": 0.3409,
+      "epoch": 1.6097453260709553,
+      "grad_norm": 0.3591005393616249,
+      "learning_rate": 1.0468301259335619e-05,
+      "loss": 0.3285,
       "step": 10935
     },
     {
-      "epoch": 1.602930402930403,
-      "grad_norm": 0.36994655865314047,
-      "learning_rate": 1.0547846295433423e-05,
-      "loss": 0.3234,
+      "epoch": 1.6104813778890033,
+      "grad_norm": 0.3699031485739882,
+      "learning_rate": 1.0459746086423226e-05,
+      "loss": 0.3164,
       "step": 10940
     },
     {
-      "epoch": 1.6036630036630037,
-      "grad_norm": 0.3535248128927632,
-      "learning_rate": 1.053933447156124e-05,
-      "loss": 0.3093,
+      "epoch": 1.6112174297070514,
+      "grad_norm": 0.36334337911899667,
+      "learning_rate": 1.045119057629206e-05,
+      "loss": 0.3154,
       "step": 10945
     },
     {
-      "epoch": 1.6043956043956045,
-      "grad_norm": 0.35583013158139354,
-      "learning_rate": 1.0530822255776992e-05,
-      "loss": 0.3262,
+      "epoch": 1.6119534815250993,
+      "grad_norm": 0.36368224707182034,
+      "learning_rate": 1.0442634735217492e-05,
+      "loss": 0.3204,
       "step": 10950
     },
     {
-      "epoch": 1.6051282051282052,
-      "grad_norm": 0.35764098107248543,
-      "learning_rate": 1.052230965426615e-05,
-      "loss": 0.3192,
+      "epoch": 1.6126895333431475,
+      "grad_norm": 0.35331692575170015,
+      "learning_rate": 1.0434078569475137e-05,
+      "loss": 0.3268,
       "step": 10955
     },
     {
-      "epoch": 1.605860805860806,
-      "grad_norm": 0.3542715998233082,
-      "learning_rate": 1.0513796673214473e-05,
-      "loss": 0.3235,
+      "epoch": 1.6134255851611954,
+      "grad_norm": 0.3697883544764186,
+      "learning_rate": 1.0425522085340854e-05,
+      "loss": 0.3255,
       "step": 10960
     },
     {
-      "epoch": 1.6065934065934067,
-      "grad_norm": 0.35516004274014784,
-      "learning_rate": 1.0505283318807993e-05,
-      "loss": 0.3112,
+      "epoch": 1.6141616369792433,
+      "grad_norm": 0.3391213438029729,
+      "learning_rate": 1.0416965289090732e-05,
+      "loss": 0.3195,
       "step": 10965
     },
     {
-      "epoch": 1.6073260073260074,
-      "grad_norm": 0.3525396292954829,
-      "learning_rate": 1.049676959723301e-05,
-      "loss": 0.3214,
+      "epoch": 1.6148976887972912,
+      "grad_norm": 0.3717901742788422,
+      "learning_rate": 1.0408408187001084e-05,
+      "loss": 0.3165,
       "step": 10970
     },
     {
-      "epoch": 1.6080586080586081,
-      "grad_norm": 0.3612216554197001,
-      "learning_rate": 1.0488255514676099e-05,
-      "loss": 0.3283,
+      "epoch": 1.6156337406153392,
+      "grad_norm": 0.370841723274499,
+      "learning_rate": 1.0399850785348456e-05,
+      "loss": 0.337,
       "step": 10975
     },
     {
-      "epoch": 1.6087912087912088,
-      "grad_norm": 0.3504714435860718,
-      "learning_rate": 1.0479741077324082e-05,
-      "loss": 0.3266,
+      "epoch": 1.6163697924333873,
+      "grad_norm": 0.3506437728714445,
+      "learning_rate": 1.0391293090409606e-05,
+      "loss": 0.3102,
       "step": 10980
     },
     {
-      "epoch": 1.6095238095238096,
-      "grad_norm": 0.3771640077062345,
-      "learning_rate": 1.0471226291364059e-05,
-      "loss": 0.3344,
+      "epoch": 1.6171058442514354,
+      "grad_norm": 0.3857407088614722,
+      "learning_rate": 1.0382735108461512e-05,
+      "loss": 0.3198,
       "step": 10985
     },
     {
-      "epoch": 1.6102564102564103,
-      "grad_norm": 0.3400328640071071,
-      "learning_rate": 1.0462711162983368e-05,
-      "loss": 0.3151,
+      "epoch": 1.6178418960694834,
+      "grad_norm": 0.357091301270163,
+      "learning_rate": 1.0374176845781359e-05,
+      "loss": 0.3204,
       "step": 10990
     },
     {
-      "epoch": 1.610989010989011,
-      "grad_norm": 0.3829867285536634,
-      "learning_rate": 1.04541956983696e-05,
-      "loss": 0.325,
+      "epoch": 1.6185779478875313,
+      "grad_norm": 0.36391568727062185,
+      "learning_rate": 1.0365618308646543e-05,
+      "loss": 0.3136,
       "step": 10995
     },
     {
-      "epoch": 1.6117216117216118,
-      "grad_norm": 0.3672879257109648,
-      "learning_rate": 1.0445679903710593e-05,
-      "loss": 0.3189,
+      "epoch": 1.6193139997055792,
+      "grad_norm": 0.35090557524644195,
+      "learning_rate": 1.0357059503334659e-05,
+      "loss": 0.3043,
       "step": 11000
     },
     {
-      "epoch": 1.6124542124542125,
-      "grad_norm": 0.35467707509195023,
-      "learning_rate": 1.0437163785194422e-05,
-      "loss": 0.3244,
+      "epoch": 1.6200500515236271,
+      "grad_norm": 0.3617778129695343,
+      "learning_rate": 1.0348500436123496e-05,
+      "loss": 0.3351,
       "step": 11005
     },
     {
-      "epoch": 1.6131868131868132,
-      "grad_norm": 0.38129661853960295,
-      "learning_rate": 1.04286473490094e-05,
-      "loss": 0.3303,
+      "epoch": 1.6207861033416753,
+      "grad_norm": 0.3707477885752319,
+      "learning_rate": 1.0339941113291039e-05,
+      "loss": 0.3114,
       "step": 11010
     },
     {
-      "epoch": 1.613919413919414,
-      "grad_norm": 0.3733812647570801,
-      "learning_rate": 1.0420130601344069e-05,
-      "loss": 0.3224,
+      "epoch": 1.6215221551597232,
+      "grad_norm": 0.3677822612090754,
+      "learning_rate": 1.0331381541115455e-05,
+      "loss": 0.3162,
       "step": 11015
     },
     {
-      "epoch": 1.6146520146520147,
-      "grad_norm": 0.36123883288541436,
-      "learning_rate": 1.0411613548387193e-05,
-      "loss": 0.319,
+      "epoch": 1.6222582069777713,
+      "grad_norm": 0.38415185428291143,
+      "learning_rate": 1.0322821725875099e-05,
+      "loss": 0.3334,
       "step": 11020
     },
     {
-      "epoch": 1.6153846153846154,
-      "grad_norm": 0.351822350363969,
-      "learning_rate": 1.0403096196327771e-05,
-      "loss": 0.321,
+      "epoch": 1.6229942587958193,
+      "grad_norm": 0.3525281549471282,
+      "learning_rate": 1.0314261673848503e-05,
+      "loss": 0.3184,
       "step": 11025
     },
     {
-      "epoch": 1.6161172161172161,
-      "grad_norm": 0.3526604466203164,
-      "learning_rate": 1.0394578551355001e-05,
-      "loss": 0.3244,
+      "epoch": 1.6237303106138672,
+      "grad_norm": 0.3595350913903759,
+      "learning_rate": 1.0305701391314373e-05,
+      "loss": 0.3199,
       "step": 11030
     },
     {
-      "epoch": 1.6168498168498169,
-      "grad_norm": 0.3832048490662678,
-      "learning_rate": 1.038606061965831e-05,
-      "loss": 0.3294,
+      "epoch": 1.6244663624319151,
+      "grad_norm": 0.3758754292921578,
+      "learning_rate": 1.0297140884551583e-05,
+      "loss": 0.3135,
       "step": 11035
     },
     {
-      "epoch": 1.6175824175824176,
-      "grad_norm": 0.35259499703016156,
-      "learning_rate": 1.0377542407427325e-05,
-      "loss": 0.2995,
+      "epoch": 1.625202414249963,
+      "grad_norm": 0.35343584921681986,
+      "learning_rate": 1.0288580159839175e-05,
+      "loss": 0.3146,
       "step": 11040
     },
     {
-      "epoch": 1.6183150183150183,
-      "grad_norm": 0.3586698360311384,
-      "learning_rate": 1.036902392085188e-05,
-      "loss": 0.3324,
+      "epoch": 1.6259384660680112,
+      "grad_norm": 0.36261730003499415,
+      "learning_rate": 1.0280019223456343e-05,
+      "loss": 0.3325,
       "step": 11045
     },
     {
-      "epoch": 1.619047619047619,
-      "grad_norm": 0.35723597592106465,
-      "learning_rate": 1.0360505166122001e-05,
-      "loss": 0.3226,
+      "epoch": 1.6266745178860593,
+      "grad_norm": 0.3518738009250834,
+      "learning_rate": 1.0271458081682441e-05,
+      "loss": 0.3217,
       "step": 11050
     },
     {
-      "epoch": 1.6197802197802198,
-      "grad_norm": 0.350480501809775,
-      "learning_rate": 1.0351986149427923e-05,
-      "loss": 0.305,
+      "epoch": 1.6274105697041072,
+      "grad_norm": 0.3515017849940598,
+      "learning_rate": 1.026289674079698e-05,
+      "loss": 0.3209,
       "step": 11055
     },
     {
-      "epoch": 1.6205128205128205,
-      "grad_norm": 0.34894633948528614,
-      "learning_rate": 1.034346687696006e-05,
-      "loss": 0.3261,
+      "epoch": 1.6281466215221552,
+      "grad_norm": 0.3836613497178749,
+      "learning_rate": 1.0254335207079605e-05,
+      "loss": 0.3406,
       "step": 11060
     },
     {
-      "epoch": 1.6212454212454213,
-      "grad_norm": 0.35853179807194685,
-      "learning_rate": 1.0334947354909016e-05,
-      "loss": 0.3166,
+      "epoch": 1.628882673340203,
+      "grad_norm": 0.3532132535375887,
+      "learning_rate": 1.0245773486810112e-05,
+      "loss": 0.3071,
       "step": 11065
     },
     {
-      "epoch": 1.621978021978022,
-      "grad_norm": 0.35003878951060013,
-      "learning_rate": 1.0326427589465572e-05,
-      "loss": 0.3309,
+      "epoch": 1.629618725158251,
+      "grad_norm": 0.3797202524827566,
+      "learning_rate": 1.0237211586268426e-05,
+      "loss": 0.3177,
       "step": 11070
     },
     {
-      "epoch": 1.6227106227106227,
-      "grad_norm": 0.3544484958483273,
-      "learning_rate": 1.0317907586820693e-05,
-      "loss": 0.324,
+      "epoch": 1.6303547769762992,
+      "grad_norm": 0.36629482001705427,
+      "learning_rate": 1.0228649511734609e-05,
+      "loss": 0.3172,
       "step": 11075
     },
     {
-      "epoch": 1.6234432234432234,
-      "grad_norm": 0.36531716273820747,
-      "learning_rate": 1.030938735316551e-05,
-      "loss": 0.332,
+      "epoch": 1.631090828794347,
+      "grad_norm": 0.37781312306936443,
+      "learning_rate": 1.0220087269488855e-05,
+      "loss": 0.328,
       "step": 11080
     },
     {
-      "epoch": 1.6241758241758242,
-      "grad_norm": 0.35930407251291396,
-      "learning_rate": 1.0300866894691327e-05,
-      "loss": 0.3254,
+      "epoch": 1.6318268806123952,
+      "grad_norm": 0.35776610699652955,
+      "learning_rate": 1.021152486581147e-05,
+      "loss": 0.3179,
       "step": 11085
     },
     {
-      "epoch": 1.624908424908425,
-      "grad_norm": 0.38384908214237906,
-      "learning_rate": 1.029234621758961e-05,
-      "loss": 0.3275,
+      "epoch": 1.6325629324304431,
+      "grad_norm": 0.3763015931361535,
+      "learning_rate": 1.0202962306982888e-05,
+      "loss": 0.3226,
       "step": 11090
     },
     {
-      "epoch": 1.6256410256410256,
-      "grad_norm": 0.38532367580929405,
-      "learning_rate": 1.0283825328051979e-05,
-      "loss": 0.3287,
+      "epoch": 1.633298984248491,
+      "grad_norm": 0.38514429091010766,
+      "learning_rate": 1.0194399599283651e-05,
+      "loss": 0.3371,
       "step": 11095
     },
     {
-      "epoch": 1.6263736263736264,
-      "grad_norm": 0.3773210361204794,
-      "learning_rate": 1.0275304232270218e-05,
-      "loss": 0.3135,
+      "epoch": 1.634035036066539,
+      "grad_norm": 0.35793241978316737,
+      "learning_rate": 1.0185836748994415e-05,
+      "loss": 0.3277,
       "step": 11100
     },
     {
-      "epoch": 1.627106227106227,
-      "grad_norm": 0.3657496818091057,
-      "learning_rate": 1.0266782936436245e-05,
-      "loss": 0.324,
+      "epoch": 1.634771087884587,
+      "grad_norm": 0.37497368524438673,
+      "learning_rate": 1.0177273762395934e-05,
+      "loss": 0.3288,
       "step": 11105
     },
     {
-      "epoch": 1.6278388278388278,
-      "grad_norm": 0.36895175386064044,
-      "learning_rate": 1.0258261446742146e-05,
-      "loss": 0.3255,
+      "epoch": 1.635507139702635,
+      "grad_norm": 0.35607809608183405,
+      "learning_rate": 1.0168710645769068e-05,
+      "loss": 0.3252,
       "step": 11110
     },
     {
-      "epoch": 1.6285714285714286,
-      "grad_norm": 0.3531853261695217,
-      "learning_rate": 1.0249739769380127e-05,
-      "loss": 0.3182,
+      "epoch": 1.6362431915206832,
+      "grad_norm": 0.35851208555279285,
+      "learning_rate": 1.0160147405394773e-05,
+      "loss": 0.3214,
       "step": 11115
     },
     {
-      "epoch": 1.6293040293040293,
-      "grad_norm": 0.3513094472672362,
-      "learning_rate": 1.0241217910542538e-05,
-      "loss": 0.3184,
+      "epoch": 1.6369792433387311,
+      "grad_norm": 0.3658319036685307,
+      "learning_rate": 1.015158404755409e-05,
+      "loss": 0.3244,
       "step": 11120
     },
     {
-      "epoch": 1.63003663003663,
-      "grad_norm": 0.37444752388476754,
-      "learning_rate": 1.0232695876421867e-05,
-      "loss": 0.3297,
+      "epoch": 1.637715295156779,
+      "grad_norm": 0.3558989585624267,
+      "learning_rate": 1.0143020578528147e-05,
+      "loss": 0.3233,
       "step": 11125
     },
     {
-      "epoch": 1.6307692307692307,
-      "grad_norm": 0.3627686960119489,
-      "learning_rate": 1.0224173673210725e-05,
-      "loss": 0.3254,
+      "epoch": 1.638451346974827,
+      "grad_norm": 0.36495324263581885,
+      "learning_rate": 1.0134457004598159e-05,
+      "loss": 0.3301,
       "step": 11130
     },
     {
-      "epoch": 1.6315018315018315,
-      "grad_norm": 0.36297662351495313,
-      "learning_rate": 1.0215651307101834e-05,
-      "loss": 0.3248,
+      "epoch": 1.639187398792875,
+      "grad_norm": 0.35825485258982076,
+      "learning_rate": 1.012589333204542e-05,
+      "loss": 0.3179,
       "step": 11135
     },
     {
-      "epoch": 1.6322344322344322,
-      "grad_norm": 0.3851208137308293,
-      "learning_rate": 1.0207128784288056e-05,
-      "loss": 0.3271,
+      "epoch": 1.639923450610923,
+      "grad_norm": 0.3660277705088314,
+      "learning_rate": 1.0117329567151283e-05,
+      "loss": 0.3117,
       "step": 11140
     },
     {
-      "epoch": 1.632967032967033,
-      "grad_norm": 0.4672639910485386,
-      "learning_rate": 1.0198606110962352e-05,
-      "loss": 0.3215,
+      "epoch": 1.640659502428971,
+      "grad_norm": 0.3729825922231736,
+      "learning_rate": 1.0108765716197175e-05,
+      "loss": 0.3321,
       "step": 11145
     },
     {
-      "epoch": 1.6336996336996337,
-      "grad_norm": 0.3567162738256589,
-      "learning_rate": 1.0190083293317799e-05,
-      "loss": 0.3106,
+      "epoch": 1.641395554247019,
+      "grad_norm": 0.36488120158917253,
+      "learning_rate": 1.0100201785464595e-05,
+      "loss": 0.3306,
       "step": 11150
     },
     {
-      "epoch": 1.6344322344322344,
-      "grad_norm": 0.3577431612902422,
-      "learning_rate": 1.018156033754757e-05,
-      "loss": 0.3283,
+      "epoch": 1.642131606065067,
+      "grad_norm": 0.3659224948954743,
+      "learning_rate": 1.0091637781235089e-05,
+      "loss": 0.3207,
       "step": 11155
     },
     {
-      "epoch": 1.6351648351648351,
-      "grad_norm": 0.35993573230953435,
-      "learning_rate": 1.017303724984495e-05,
-      "loss": 0.3169,
+      "epoch": 1.642867657883115,
+      "grad_norm": 0.3560749614077936,
+      "learning_rate": 1.0083073709790263e-05,
+      "loss": 0.3166,
       "step": 11160
     },
     {
-      "epoch": 1.6358974358974359,
-      "grad_norm": 0.3653056370504911,
-      "learning_rate": 1.0164514036403312e-05,
-      "loss": 0.3401,
+      "epoch": 1.6436037097011629,
+      "grad_norm": 0.3676288149442803,
+      "learning_rate": 1.0074509577411773e-05,
+      "loss": 0.3174,
       "step": 11165
     },
     {
-      "epoch": 1.6366300366300366,
-      "grad_norm": 0.3643747070588311,
-      "learning_rate": 1.015599070341613e-05,
-      "loss": 0.3242,
+      "epoch": 1.6443397615192108,
+      "grad_norm": 0.3643760202709763,
+      "learning_rate": 1.0065945390381313e-05,
+      "loss": 0.3288,
       "step": 11170
     },
     {
-      "epoch": 1.6373626373626373,
-      "grad_norm": 0.34649245210939217,
-      "learning_rate": 1.0147467257076944e-05,
-      "loss": 0.3183,
+      "epoch": 1.645075813337259,
+      "grad_norm": 0.35903035116306165,
+      "learning_rate": 1.0057381154980623e-05,
+      "loss": 0.3159,
       "step": 11175
     },
     {
-      "epoch": 1.638095238095238,
-      "grad_norm": 0.3457145517754876,
-      "learning_rate": 1.0138943703579403e-05,
-      "loss": 0.3211,
+      "epoch": 1.645811865155307,
+      "grad_norm": 0.3590465882106527,
+      "learning_rate": 1.0048816877491478e-05,
+      "loss": 0.3139,
       "step": 11180
     },
     {
-      "epoch": 1.6388278388278388,
-      "grad_norm": 0.3605086861180557,
-      "learning_rate": 1.0130420049117213e-05,
-      "loss": 0.3178,
+      "epoch": 1.646547916973355,
+      "grad_norm": 0.36736710168489134,
+      "learning_rate": 1.0040252564195685e-05,
+      "loss": 0.3206,
       "step": 11185
     },
     {
-      "epoch": 1.6395604395604395,
-      "grad_norm": 0.3679034058115278,
-      "learning_rate": 1.012189629988417e-05,
-      "loss": 0.3193,
+      "epoch": 1.647283968791403,
+      "grad_norm": 0.33976068327768527,
+      "learning_rate": 1.0031688221375067e-05,
+      "loss": 0.32,
       "step": 11190
     },
     {
-      "epoch": 1.6402930402930402,
-      "grad_norm": 0.3634000193943487,
-      "learning_rate": 1.0113372462074119e-05,
-      "loss": 0.3269,
+      "epoch": 1.6480200206094509,
+      "grad_norm": 0.3484346612302931,
+      "learning_rate": 1.0023123855311485e-05,
+      "loss": 0.3116,
       "step": 11195
     },
     {
-      "epoch": 1.641025641025641,
-      "grad_norm": 0.3445737908775674,
-      "learning_rate": 1.0104848541880988e-05,
-      "loss": 0.3223,
+      "epoch": 1.6487560724274988,
+      "grad_norm": 0.3669113171294673,
+      "learning_rate": 1.0014559472286805e-05,
+      "loss": 0.3225,
       "step": 11200
     },
     {
-      "epoch": 1.6417582417582417,
-      "grad_norm": 0.35776440734834947,
-      "learning_rate": 1.0096324545498753e-05,
-      "loss": 0.3097,
+      "epoch": 1.649492124245547,
+      "grad_norm": 0.3733938014122494,
+      "learning_rate": 1.000599507858291e-05,
+      "loss": 0.3289,
       "step": 11205
     },
     {
-      "epoch": 1.6424908424908424,
-      "grad_norm": 0.3642975159317143,
-      "learning_rate": 1.0087800479121456e-05,
-      "loss": 0.3107,
+      "epoch": 1.6502281760635948,
+      "grad_norm": 0.3550753000278096,
+      "learning_rate": 9.997430680481686e-06,
+      "loss": 0.3129,
       "step": 11210
     },
     {
-      "epoch": 1.6432234432234432,
-      "grad_norm": 0.38435340333841345,
-      "learning_rate": 1.0079276348943176e-05,
-      "loss": 0.3457,
+      "epoch": 1.650964227881643,
+      "grad_norm": 0.37822252297886577,
+      "learning_rate": 9.988866284265037e-06,
+      "loss": 0.3245,
       "step": 11215
     },
     {
-      "epoch": 1.6439560439560439,
-      "grad_norm": 0.3575008550302539,
-      "learning_rate": 1.0070752161158054e-05,
-      "loss": 0.3337,
+      "epoch": 1.651700279699691,
+      "grad_norm": 0.3598989594453636,
+      "learning_rate": 9.980301896214844e-06,
+      "loss": 0.3312,
       "step": 11220
     },
     {
-      "epoch": 1.6446886446886446,
-      "grad_norm": 0.35737103483565313,
-      "learning_rate": 1.006222792196026e-05,
-      "loss": 0.338,
+      "epoch": 1.6524363315177388,
+      "grad_norm": 0.34117953852475663,
+      "learning_rate": 9.971737522612997e-06,
+      "loss": 0.3218,
       "step": 11225
     },
     {
-      "epoch": 1.6454212454212453,
-      "grad_norm": 0.3607550861529369,
-      "learning_rate": 1.0053703637544008e-05,
-      "loss": 0.3266,
+      "epoch": 1.6531723833357868,
+      "grad_norm": 0.37219646669441314,
+      "learning_rate": 9.96317316974137e-06,
+      "loss": 0.3175,
       "step": 11230
     },
     {
-      "epoch": 1.646153846153846,
-      "grad_norm": 0.37471399109586223,
-      "learning_rate": 1.0045179314103541e-05,
-      "loss": 0.3145,
+      "epoch": 1.6539084351538347,
+      "grad_norm": 0.3597108993691857,
+      "learning_rate": 9.954608843881829e-06,
+      "loss": 0.3149,
       "step": 11235
     },
     {
-      "epoch": 1.6468864468864468,
-      "grad_norm": 0.36592659676528055,
-      "learning_rate": 1.0036654957833136e-05,
-      "loss": 0.3399,
+      "epoch": 1.6546444869718828,
+      "grad_norm": 0.37151719154104074,
+      "learning_rate": 9.946044551316201e-06,
+      "loss": 0.3215,
       "step": 11240
     },
     {
-      "epoch": 1.6476190476190475,
-      "grad_norm": 0.3733831180052479,
-      "learning_rate": 1.0028130574927086e-05,
-      "loss": 0.321,
+      "epoch": 1.655380538789931,
+      "grad_norm": 0.35416901973604353,
+      "learning_rate": 9.937480298326315e-06,
+      "loss": 0.3143,
       "step": 11245
     },
     {
-      "epoch": 1.6483516483516483,
-      "grad_norm": 0.37609573288460046,
-      "learning_rate": 1.0019606171579713e-05,
-      "loss": 0.3293,
+      "epoch": 1.656116590607979,
+      "grad_norm": 0.35593571254728973,
+      "learning_rate": 9.928916091193954e-06,
+      "loss": 0.3294,
       "step": 11250
     },
     {
-      "epoch": 1.649084249084249,
-      "grad_norm": 0.38631610438090525,
-      "learning_rate": 1.0011081753985344e-05,
-      "loss": 0.3228,
+      "epoch": 1.6568526424260268,
+      "grad_norm": 0.3555883291894213,
+      "learning_rate": 9.920351936200866e-06,
+      "loss": 0.3098,
       "step": 11255
     },
     {
-      "epoch": 1.6498168498168497,
-      "grad_norm": 0.3571573245803005,
-      "learning_rate": 1.0002557328338321e-05,
-      "loss": 0.3079,
+      "epoch": 1.6575886942440747,
+      "grad_norm": 0.3698675521263941,
+      "learning_rate": 9.911787839628776e-06,
+      "loss": 0.3161,
       "step": 11260
     },
     {
-      "epoch": 1.6505494505494505,
-      "grad_norm": 0.36012963207083964,
-      "learning_rate": 9.994032900832996e-06,
-      "loss": 0.3235,
+      "epoch": 1.6583247460621227,
+      "grad_norm": 0.364093115448152,
+      "learning_rate": 9.90322380775935e-06,
+      "loss": 0.3279,
       "step": 11265
     },
     {
-      "epoch": 1.6512820512820512,
-      "grad_norm": 0.37324023344599405,
-      "learning_rate": 9.98550847766371e-06,
-      "loss": 0.3291,
+      "epoch": 1.6590607978801708,
+      "grad_norm": 0.3577712807114756,
+      "learning_rate": 9.89465984687421e-06,
+      "loss": 0.3234,
       "step": 11270
     },
     {
-      "epoch": 1.652014652014652,
-      "grad_norm": 0.3582050090756073,
-      "learning_rate": 9.976984065024821e-06,
-      "loss": 0.3362,
+      "epoch": 1.6597968496982187,
+      "grad_norm": 0.35390192193669356,
+      "learning_rate": 9.886095963254932e-06,
+      "loss": 0.3301,
       "step": 11275
     },
     {
-      "epoch": 1.6527472527472526,
-      "grad_norm": 0.35320476631232933,
-      "learning_rate": 9.968459669110655e-06,
-      "loss": 0.3302,
+      "epoch": 1.6605329015162669,
+      "grad_norm": 0.36574374597043474,
+      "learning_rate": 9.87753216318303e-06,
+      "loss": 0.3271,
       "step": 11280
     },
     {
-      "epoch": 1.6534798534798534,
-      "grad_norm": 0.3689275784664794,
-      "learning_rate": 9.959935296115547e-06,
-      "loss": 0.3321,
+      "epoch": 1.6612689533343148,
+      "grad_norm": 0.36397686094226955,
+      "learning_rate": 9.868968452939963e-06,
+      "loss": 0.3117,
       "step": 11285
     },
     {
-      "epoch": 1.654212454212454,
-      "grad_norm": 0.37880888788076,
-      "learning_rate": 9.951410952233802e-06,
-      "loss": 0.3334,
+      "epoch": 1.6620050051523627,
+      "grad_norm": 0.35990205667438896,
+      "learning_rate": 9.860404838807113e-06,
+      "loss": 0.3092,
       "step": 11290
     },
     {
-      "epoch": 1.6549450549450548,
-      "grad_norm": 0.3660165873618415,
-      "learning_rate": 9.942886643659709e-06,
-      "loss": 0.3048,
+      "epoch": 1.6627410569704106,
+      "grad_norm": 0.36417756157379905,
+      "learning_rate": 9.851841327065806e-06,
+      "loss": 0.32,
       "step": 11295
     },
     {
-      "epoch": 1.6556776556776556,
-      "grad_norm": 0.36704002452407064,
-      "learning_rate": 9.934362376587533e-06,
-      "loss": 0.3183,
+      "epoch": 1.6634771087884586,
+      "grad_norm": 0.3614972664039947,
+      "learning_rate": 9.843277923997278e-06,
+      "loss": 0.3246,
       "step": 11300
     },
     {
-      "epoch": 1.6564102564102563,
-      "grad_norm": 0.3660284088454685,
-      "learning_rate": 9.925838157211502e-06,
-      "loss": 0.3083,
+      "epoch": 1.6642131606065067,
+      "grad_norm": 0.35576121965472,
+      "learning_rate": 9.834714635882698e-06,
+      "loss": 0.2991,
       "step": 11305
     },
     {
-      "epoch": 1.657142857142857,
-      "grad_norm": 0.3471567412573336,
-      "learning_rate": 9.91731399172582e-06,
-      "loss": 0.3153,
+      "epoch": 1.6649492124245548,
+      "grad_norm": 0.367200520412062,
+      "learning_rate": 9.826151469003141e-06,
+      "loss": 0.323,
       "step": 11310
     },
     {
-      "epoch": 1.6578754578754578,
-      "grad_norm": 0.35843231218545296,
-      "learning_rate": 9.908789886324643e-06,
-      "loss": 0.3179,
+      "epoch": 1.6656852642426028,
+      "grad_norm": 0.3612485358302765,
+      "learning_rate": 9.817588429639596e-06,
+      "loss": 0.3166,
       "step": 11315
     },
     {
-      "epoch": 1.6586080586080585,
-      "grad_norm": 0.3597872512866094,
-      "learning_rate": 9.90026584720208e-06,
-      "loss": 0.3227,
+      "epoch": 1.6664213160606507,
+      "grad_norm": 0.35158091722786156,
+      "learning_rate": 9.809025524072968e-06,
+      "loss": 0.3067,
       "step": 11320
     },
     {
-      "epoch": 1.6593406593406592,
-      "grad_norm": 0.379192260668644,
-      "learning_rate": 9.891741880552209e-06,
-      "loss": 0.3258,
+      "epoch": 1.6671573678786986,
+      "grad_norm": 0.3592537087503226,
+      "learning_rate": 9.800462758584046e-06,
+      "loss": 0.3319,
       "step": 11325
     },
     {
-      "epoch": 1.66007326007326,
-      "grad_norm": 0.36784732716595225,
-      "learning_rate": 9.883217992569036e-06,
-      "loss": 0.3348,
+      "epoch": 1.6678934196967465,
+      "grad_norm": 0.3697049960418154,
+      "learning_rate": 9.791900139453531e-06,
+      "loss": 0.3181,
       "step": 11330
     },
     {
-      "epoch": 1.6608058608058607,
-      "grad_norm": 0.36793894286584244,
-      "learning_rate": 9.874694189446519e-06,
-      "loss": 0.3231,
+      "epoch": 1.6686294715147947,
+      "grad_norm": 0.359087187359359,
+      "learning_rate": 9.783337672962008e-06,
+      "loss": 0.3198,
       "step": 11335
     },
     {
-      "epoch": 1.6615384615384614,
-      "grad_norm": 0.35319494338805274,
-      "learning_rate": 9.866170477378557e-06,
-      "loss": 0.3331,
+      "epoch": 1.6693655233328426,
+      "grad_norm": 0.3534372970274228,
+      "learning_rate": 9.774775365389954e-06,
+      "loss": 0.3156,
       "step": 11340
     },
     {
-      "epoch": 1.6622710622710621,
-      "grad_norm": 0.37349245702611267,
-      "learning_rate": 9.857646862558974e-06,
-      "loss": 0.3294,
+      "epoch": 1.6701015751508907,
+      "grad_norm": 0.3692204619726876,
+      "learning_rate": 9.766213223017736e-06,
+      "loss": 0.3211,
       "step": 11345
     },
     {
-      "epoch": 1.6630036630036629,
-      "grad_norm": 0.34602637357840227,
-      "learning_rate": 9.849123351181535e-06,
-      "loss": 0.3079,
+      "epoch": 1.6708376269689387,
+      "grad_norm": 0.37216382490969285,
+      "learning_rate": 9.757651252125583e-06,
+      "loss": 0.3217,
       "step": 11350
     },
     {
-      "epoch": 1.6637362637362636,
-      "grad_norm": 0.37056173134429615,
-      "learning_rate": 9.840599949439921e-06,
-      "loss": 0.3333,
+      "epoch": 1.6715736787869866,
+      "grad_norm": 0.3561463345746021,
+      "learning_rate": 9.749089458993616e-06,
+      "loss": 0.3233,
       "step": 11355
     },
     {
-      "epoch": 1.6644688644688643,
-      "grad_norm": 0.3509466617996749,
-      "learning_rate": 9.832076663527731e-06,
-      "loss": 0.319,
+      "epoch": 1.6723097306050345,
+      "grad_norm": 0.4247699534265859,
+      "learning_rate": 9.740527849901809e-06,
+      "loss": 0.3292,
       "step": 11360
     },
     {
-      "epoch": 1.665201465201465,
-      "grad_norm": 0.35369028529684854,
-      "learning_rate": 9.823553499638493e-06,
-      "loss": 0.3227,
+      "epoch": 1.6730457824230824,
+      "grad_norm": 0.35419797220640364,
+      "learning_rate": 9.731966431130017e-06,
+      "loss": 0.3118,
       "step": 11365
     },
     {
-      "epoch": 1.6659340659340658,
-      "grad_norm": 0.37531209868540477,
-      "learning_rate": 9.815030463965626e-06,
-      "loss": 0.3276,
+      "epoch": 1.6737818342411306,
+      "grad_norm": 0.3577280841595848,
+      "learning_rate": 9.723405208957942e-06,
+      "loss": 0.329,
       "step": 11370
     },
     {
-      "epoch": 1.6666666666666665,
-      "grad_norm": 0.37552137751925635,
-      "learning_rate": 9.806507562702484e-06,
-      "loss": 0.3317,
+      "epoch": 1.6745178860591787,
+      "grad_norm": 0.3760532483194485,
+      "learning_rate": 9.714844189665149e-06,
+      "loss": 0.3161,
       "step": 11375
     },
     {
-      "epoch": 1.6673992673992672,
-      "grad_norm": 0.36847104061062425,
-      "learning_rate": 9.797984802042294e-06,
-      "loss": 0.3159,
+      "epoch": 1.6752539378772267,
+      "grad_norm": 0.3707340009696498,
+      "learning_rate": 9.70628337953106e-06,
+      "loss": 0.3023,
       "step": 11380
     },
     {
-      "epoch": 1.668131868131868,
-      "grad_norm": 0.36028399439737113,
-      "learning_rate": 9.789462188178194e-06,
-      "loss": 0.3226,
+      "epoch": 1.6759899896952746,
+      "grad_norm": 0.3692362425405184,
+      "learning_rate": 9.697722784834927e-06,
+      "loss": 0.3222,
       "step": 11385
     },
     {
-      "epoch": 1.6688644688644687,
-      "grad_norm": 0.34575748084355296,
-      "learning_rate": 9.780939727303224e-06,
-      "loss": 0.313,
+      "epoch": 1.6767260415133225,
+      "grad_norm": 0.34301558419759154,
+      "learning_rate": 9.68916241185586e-06,
+      "loss": 0.3168,
       "step": 11390
     },
     {
-      "epoch": 1.6695970695970694,
-      "grad_norm": 0.3810415750955356,
-      "learning_rate": 9.772417425610298e-06,
-      "loss": 0.3335,
+      "epoch": 1.6774620933313704,
+      "grad_norm": 0.3836449066934659,
+      "learning_rate": 9.680602266872798e-06,
+      "loss": 0.3241,
       "step": 11395
     },
     {
-      "epoch": 1.6703296703296702,
-      "grad_norm": 0.35729218845572247,
-      "learning_rate": 9.763895289292219e-06,
-      "loss": 0.3223,
+      "epoch": 1.6781981451494186,
+      "grad_norm": 0.37376075700344913,
+      "learning_rate": 9.672042356164513e-06,
+      "loss": 0.3178,
       "step": 11400
     },
     {
-      "epoch": 1.671062271062271,
-      "grad_norm": 0.3572872776478421,
-      "learning_rate": 9.755373324541676e-06,
-      "loss": 0.3153,
+      "epoch": 1.6789341969674665,
+      "grad_norm": 0.3480720853755703,
+      "learning_rate": 9.663482686009608e-06,
+      "loss": 0.313,
       "step": 11405
     },
     {
-      "epoch": 1.6717948717948716,
-      "grad_norm": 0.3520264191162397,
-      "learning_rate": 9.746851537551226e-06,
-      "loss": 0.3242,
+      "epoch": 1.6796702487855146,
+      "grad_norm": 0.3760967779127571,
+      "learning_rate": 9.654923262686505e-06,
+      "loss": 0.3235,
       "step": 11410
     },
     {
-      "epoch": 1.6725274725274726,
-      "grad_norm": 0.3658144165650994,
-      "learning_rate": 9.7383299345133e-06,
-      "loss": 0.3229,
+      "epoch": 1.6804063006035626,
+      "grad_norm": 0.3630609784506012,
+      "learning_rate": 9.646364092473456e-06,
+      "loss": 0.321,
       "step": 11415
     },
     {
-      "epoch": 1.6732600732600733,
-      "grad_norm": 0.36332179684457677,
-      "learning_rate": 9.729808521620195e-06,
-      "loss": 0.3305,
+      "epoch": 1.6811423524216105,
+      "grad_norm": 0.3430999427290298,
+      "learning_rate": 9.637805181648512e-06,
+      "loss": 0.3109,
       "step": 11420
     },
     {
-      "epoch": 1.673992673992674,
-      "grad_norm": 0.3533756945313288,
-      "learning_rate": 9.721287305064063e-06,
-      "loss": 0.3139,
+      "epoch": 1.6818784042396584,
+      "grad_norm": 0.3676861284324845,
+      "learning_rate": 9.62924653648955e-06,
+      "loss": 0.326,
       "step": 11425
     },
     {
-      "epoch": 1.6747252747252748,
-      "grad_norm": 0.36557240811520275,
-      "learning_rate": 9.712766291036934e-06,
-      "loss": 0.3201,
+      "epoch": 1.6826144560577063,
+      "grad_norm": 0.3723415131808058,
+      "learning_rate": 9.620688163274233e-06,
+      "loss": 0.3223,
       "step": 11430
     },
     {
-      "epoch": 1.6754578754578755,
-      "grad_norm": 0.38135701407153655,
-      "learning_rate": 9.704245485730662e-06,
-      "loss": 0.3178,
+      "epoch": 1.6833505078757545,
+      "grad_norm": 0.35744480513891475,
+      "learning_rate": 9.612130068280045e-06,
+      "loss": 0.3028,
       "step": 11435
     },
     {
-      "epoch": 1.6761904761904762,
-      "grad_norm": 0.3528386651605164,
-      "learning_rate": 9.695724895336971e-06,
-      "loss": 0.3067,
+      "epoch": 1.6840865596938026,
+      "grad_norm": 0.3729092247273533,
+      "learning_rate": 9.603572257784248e-06,
+      "loss": 0.3317,
       "step": 11440
     },
     {
-      "epoch": 1.676923076923077,
-      "grad_norm": 0.35714768527223584,
-      "learning_rate": 9.687204526047425e-06,
-      "loss": 0.3179,
+      "epoch": 1.6848226115118505,
+      "grad_norm": 0.37357385427756795,
+      "learning_rate": 9.595014738063904e-06,
+      "loss": 0.3238,
       "step": 11445
     },
     {
-      "epoch": 1.6776556776556777,
-      "grad_norm": 0.3691697304427781,
-      "learning_rate": 9.678684384053416e-06,
-      "loss": 0.3324,
+      "epoch": 1.6855586633298985,
+      "grad_norm": 0.3558762064092378,
+      "learning_rate": 9.58645751539587e-06,
+      "loss": 0.3176,
       "step": 11450
     },
     {
-      "epoch": 1.6783882783882784,
-      "grad_norm": 0.3557419470671424,
-      "learning_rate": 9.670164475546185e-06,
-      "loss": 0.3157,
+      "epoch": 1.6862947151479464,
+      "grad_norm": 0.37981691762110004,
+      "learning_rate": 9.577900596056763e-06,
+      "loss": 0.3126,
       "step": 11455
     },
     {
-      "epoch": 1.6791208791208792,
-      "grad_norm": 0.37132815959014553,
-      "learning_rate": 9.661644806716794e-06,
-      "loss": 0.3169,
+      "epoch": 1.6870307669659943,
+      "grad_norm": 0.3601599487447567,
+      "learning_rate": 9.569343986323003e-06,
+      "loss": 0.3193,
       "step": 11460
     },
     {
-      "epoch": 1.6798534798534799,
-      "grad_norm": 0.37700426652777763,
-      "learning_rate": 9.653125383756136e-06,
-      "loss": 0.3254,
+      "epoch": 1.6877668187840424,
+      "grad_norm": 0.36503126469877706,
+      "learning_rate": 9.56078769247076e-06,
+      "loss": 0.3191,
       "step": 11465
     },
     {
-      "epoch": 1.6805860805860806,
-      "grad_norm": 0.3802221839669021,
-      "learning_rate": 9.644606212854928e-06,
-      "loss": 0.3231,
+      "epoch": 1.6885028706020904,
+      "grad_norm": 0.3614226010658769,
+      "learning_rate": 9.552231720775993e-06,
+      "loss": 0.3183,
       "step": 11470
     },
     {
-      "epoch": 1.6813186813186813,
-      "grad_norm": 0.3492871322651001,
-      "learning_rate": 9.636087300203688e-06,
-      "loss": 0.3193,
+      "epoch": 1.6892389224201385,
+      "grad_norm": 0.3640655776601726,
+      "learning_rate": 9.543676077514405e-06,
+      "loss": 0.3237,
       "step": 11475
     },
     {
-      "epoch": 1.682051282051282,
-      "grad_norm": 0.3491380835719865,
-      "learning_rate": 9.627568651992769e-06,
-      "loss": 0.3227,
+      "epoch": 1.6899749742381864,
+      "grad_norm": 0.3737650398752546,
+      "learning_rate": 9.535120768961474e-06,
+      "loss": 0.323,
       "step": 11480
     },
     {
-      "epoch": 1.6827838827838828,
-      "grad_norm": 0.35610491273786476,
-      "learning_rate": 9.619050274412314e-06,
-      "loss": 0.3249,
+      "epoch": 1.6907110260562344,
+      "grad_norm": 0.3737683092309896,
+      "learning_rate": 9.526565801392427e-06,
+      "loss": 0.3286,
       "step": 11485
     },
     {
-      "epoch": 1.6835164835164835,
-      "grad_norm": 0.3630010097767952,
-      "learning_rate": 9.610532173652277e-06,
-      "loss": 0.334,
+      "epoch": 1.6914470778742823,
+      "grad_norm": 0.36946542307505453,
+      "learning_rate": 9.518011181082232e-06,
+      "loss": 0.3197,
       "step": 11490
     },
     {
-      "epoch": 1.6842490842490843,
-      "grad_norm": 0.35617526499782054,
-      "learning_rate": 9.602014355902411e-06,
-      "loss": 0.3163,
+      "epoch": 1.6921831296923302,
+      "grad_norm": 0.3662432745684035,
+      "learning_rate": 9.509456914305621e-06,
+      "loss": 0.3277,
       "step": 11495
     },
     {
-      "epoch": 1.684981684981685,
-      "grad_norm": 0.33270492209004626,
-      "learning_rate": 9.59349682735226e-06,
-      "loss": 0.3038,
+      "epoch": 1.6929191815103783,
+      "grad_norm": 0.3539749203414942,
+      "learning_rate": 9.500903007337046e-06,
+      "loss": 0.3221,
       "step": 11500
     },
     {
-      "epoch": 1.6857142857142857,
-      "grad_norm": 0.36685906208659197,
-      "learning_rate": 9.584979594191164e-06,
-      "loss": 0.3267,
+      "epoch": 1.6936552333284263,
+      "grad_norm": 0.3700640257217791,
+      "learning_rate": 9.492349466450713e-06,
+      "loss": 0.3112,
       "step": 11505
     },
     {
-      "epoch": 1.6864468864468865,
-      "grad_norm": 0.37653572563578436,
-      "learning_rate": 9.576462662608242e-06,
-      "loss": 0.331,
+      "epoch": 1.6943912851464744,
+      "grad_norm": 0.35620461331761233,
+      "learning_rate": 9.483796297920545e-06,
+      "loss": 0.3243,
       "step": 11510
     },
     {
-      "epoch": 1.6871794871794872,
-      "grad_norm": 0.35231959569304594,
-      "learning_rate": 9.567946038792392e-06,
-      "loss": 0.3092,
+      "epoch": 1.6951273369645223,
+      "grad_norm": 0.3786066512503005,
+      "learning_rate": 9.475243508020205e-06,
+      "loss": 0.3219,
       "step": 11515
     },
     {
-      "epoch": 1.687912087912088,
-      "grad_norm": 0.3654106238710355,
-      "learning_rate": 9.559429728932298e-06,
-      "loss": 0.3281,
+      "epoch": 1.6958633887825703,
+      "grad_norm": 0.3624362374642801,
+      "learning_rate": 9.466691103023068e-06,
+      "loss": 0.3197,
       "step": 11520
     },
     {
-      "epoch": 1.6886446886446886,
-      "grad_norm": 0.3808746833599318,
-      "learning_rate": 9.550913739216408e-06,
-      "loss": 0.3379,
+      "epoch": 1.6965994406006182,
+      "grad_norm": 0.3764621400962587,
+      "learning_rate": 9.45813908920223e-06,
+      "loss": 0.3162,
       "step": 11525
     },
     {
-      "epoch": 1.6893772893772894,
-      "grad_norm": 0.339177410019425,
-      "learning_rate": 9.542398075832944e-06,
-      "loss": 0.3176,
+      "epoch": 1.6973354924186663,
+      "grad_norm": 0.36089901406092384,
+      "learning_rate": 9.449587472830506e-06,
+      "loss": 0.3214,
       "step": 11530
     },
     {
-      "epoch": 1.69010989010989,
-      "grad_norm": 0.35545930670267695,
-      "learning_rate": 9.533882744969881e-06,
-      "loss": 0.3143,
+      "epoch": 1.6980715442367142,
+      "grad_norm": 0.3597727779102053,
+      "learning_rate": 9.441036260180403e-06,
+      "loss": 0.3209,
       "step": 11535
     },
     {
-      "epoch": 1.6908424908424908,
-      "grad_norm": 0.37729381669541545,
-      "learning_rate": 9.525367752814959e-06,
-      "loss": 0.3224,
+      "epoch": 1.6988075960547624,
+      "grad_norm": 0.3668632424412183,
+      "learning_rate": 9.432485457524156e-06,
+      "loss": 0.3002,
       "step": 11540
     },
     {
-      "epoch": 1.6915750915750916,
-      "grad_norm": 0.36357090958989985,
-      "learning_rate": 9.51685310555567e-06,
-      "loss": 0.3332,
+      "epoch": 1.6995436478728103,
+      "grad_norm": 0.3534486717975531,
+      "learning_rate": 9.423935071133675e-06,
+      "loss": 0.3119,
       "step": 11545
     },
     {
-      "epoch": 1.6923076923076923,
-      "grad_norm": 0.3669470447992764,
-      "learning_rate": 9.508338809379262e-06,
-      "loss": 0.3251,
+      "epoch": 1.7002796996908582,
+      "grad_norm": 0.3607190146159087,
+      "learning_rate": 9.415385107280579e-06,
+      "loss": 0.314,
       "step": 11550
     },
     {
-      "epoch": 1.693040293040293,
-      "grad_norm": 0.3537300132509414,
-      "learning_rate": 9.499824870472713e-06,
-      "loss": 0.3171,
+      "epoch": 1.7010157515089062,
+      "grad_norm": 0.3801212229524694,
+      "learning_rate": 9.406835572236181e-06,
+      "loss": 0.3162,
       "step": 11555
     },
     {
-      "epoch": 1.6937728937728938,
-      "grad_norm": 0.3485674140211883,
-      "learning_rate": 9.49131129502276e-06,
-      "loss": 0.3083,
+      "epoch": 1.701751803326954,
+      "grad_norm": 0.3582045818399246,
+      "learning_rate": 9.398286472271463e-06,
+      "loss": 0.3376,
       "step": 11560
     },
     {
-      "epoch": 1.6945054945054945,
-      "grad_norm": 0.36353765013873424,
-      "learning_rate": 9.482798089215857e-06,
-      "loss": 0.3313,
+      "epoch": 1.7024878551450022,
+      "grad_norm": 0.36865534925799104,
+      "learning_rate": 9.389737813657104e-06,
+      "loss": 0.3104,
       "step": 11565
     },
     {
-      "epoch": 1.6952380952380952,
-      "grad_norm": 0.3638997172794509,
-      "learning_rate": 9.474285259238207e-06,
-      "loss": 0.332,
+      "epoch": 1.7032239069630501,
+      "grad_norm": 0.37979748480494235,
+      "learning_rate": 9.381189602663446e-06,
+      "loss": 0.3168,
       "step": 11570
     },
     {
-      "epoch": 1.695970695970696,
-      "grad_norm": 0.3671892924647035,
-      "learning_rate": 9.465772811275731e-06,
-      "loss": 0.3239,
+      "epoch": 1.7039599587810983,
+      "grad_norm": 0.35547574158266904,
+      "learning_rate": 9.372641845560517e-06,
+      "loss": 0.3324,
       "step": 11575
     },
     {
-      "epoch": 1.6967032967032967,
-      "grad_norm": 0.3664783726017456,
-      "learning_rate": 9.457260751514062e-06,
-      "loss": 0.3307,
+      "epoch": 1.7046960105991462,
+      "grad_norm": 0.3735632819775801,
+      "learning_rate": 9.364094548617994e-06,
+      "loss": 0.3366,
       "step": 11580
     },
     {
-      "epoch": 1.6974358974358974,
-      "grad_norm": 0.3614158901243449,
-      "learning_rate": 9.448749086138577e-06,
-      "loss": 0.3066,
+      "epoch": 1.7054320624171941,
+      "grad_norm": 0.3633804896754463,
+      "learning_rate": 9.355547718105235e-06,
+      "loss": 0.3151,
       "step": 11585
     },
     {
-      "epoch": 1.6981684981684981,
-      "grad_norm": 0.36630132335180743,
-      "learning_rate": 9.44023782133434e-06,
-      "loss": 0.3172,
+      "epoch": 1.706168114235242,
+      "grad_norm": 0.36624531828629564,
+      "learning_rate": 9.347001360291248e-06,
+      "loss": 0.3193,
       "step": 11590
     },
     {
-      "epoch": 1.6989010989010989,
-      "grad_norm": 0.37453361569912347,
-      "learning_rate": 9.431726963286145e-06,
-      "loss": 0.3434,
+      "epoch": 1.7069041660532902,
+      "grad_norm": 0.3642397698832851,
+      "learning_rate": 9.338455481444687e-06,
+      "loss": 0.316,
       "step": 11595
     },
     {
-      "epoch": 1.6996336996336996,
-      "grad_norm": 0.3910068476298356,
-      "learning_rate": 9.423216518178474e-06,
-      "loss": 0.3227,
+      "epoch": 1.7076402178713381,
+      "grad_norm": 0.3544364277571668,
+      "learning_rate": 9.32991008783387e-06,
+      "loss": 0.3257,
       "step": 11600
     },
     {
-      "epoch": 1.7003663003663003,
-      "grad_norm": 0.355405058117633,
-      "learning_rate": 9.414706492195516e-06,
-      "loss": 0.3099,
+      "epoch": 1.7083762696893863,
+      "grad_norm": 0.360528483261268,
+      "learning_rate": 9.321365185726742e-06,
+      "loss": 0.3219,
       "step": 11605
     },
     {
-      "epoch": 1.701098901098901,
-      "grad_norm": 0.36483327355080997,
-      "learning_rate": 9.406196891521156e-06,
-      "loss": 0.324,
+      "epoch": 1.7091123215074342,
+      "grad_norm": 0.3602159048395578,
+      "learning_rate": 9.312820781390904e-06,
+      "loss": 0.3211,
       "step": 11610
     },
     {
-      "epoch": 1.7018315018315018,
-      "grad_norm": 0.3502827754750937,
-      "learning_rate": 9.397687722338972e-06,
-      "loss": 0.3161,
+      "epoch": 1.7098483733254821,
+      "grad_norm": 0.3634016704012471,
+      "learning_rate": 9.304276881093581e-06,
+      "loss": 0.3243,
       "step": 11615
     },
     {
-      "epoch": 1.7025641025641025,
-      "grad_norm": 0.3652283535458358,
-      "learning_rate": 9.38917899083222e-06,
-      "loss": 0.3171,
+      "epoch": 1.71058442514353,
+      "grad_norm": 0.3721367208511192,
+      "learning_rate": 9.295733491101627e-06,
+      "loss": 0.3302,
       "step": 11620
     },
     {
-      "epoch": 1.7032967032967035,
-      "grad_norm": 0.3592422843651675,
-      "learning_rate": 9.38067070318385e-06,
-      "loss": 0.3255,
+      "epoch": 1.711320476961578,
+      "grad_norm": 0.37734953264996884,
+      "learning_rate": 9.287190617681535e-06,
+      "loss": 0.3252,
       "step": 11625
     },
     {
-      "epoch": 1.7040293040293042,
-      "grad_norm": 0.36443238950693096,
-      "learning_rate": 9.372162865576475e-06,
-      "loss": 0.3262,
+      "epoch": 1.712056528779626,
+      "grad_norm": 0.3745177210233044,
+      "learning_rate": 9.278648267099401e-06,
+      "loss": 0.326,
       "step": 11630
     },
     {
-      "epoch": 1.704761904761905,
-      "grad_norm": 0.3560247837436095,
-      "learning_rate": 9.363655484192398e-06,
-      "loss": 0.3309,
+      "epoch": 1.712792580597674,
+      "grad_norm": 0.344599327114937,
+      "learning_rate": 9.270106445620955e-06,
+      "loss": 0.3032,
       "step": 11635
     },
     {
-      "epoch": 1.7054945054945057,
-      "grad_norm": 0.3625416445524746,
-      "learning_rate": 9.355148565213575e-06,
-      "loss": 0.3143,
+      "epoch": 1.7135286324157222,
+      "grad_norm": 0.36728269631926297,
+      "learning_rate": 9.26156515951152e-06,
+      "loss": 0.3106,
       "step": 11640
     },
     {
-      "epoch": 1.7062271062271064,
-      "grad_norm": 0.35517830591439836,
-      "learning_rate": 9.346642114821633e-06,
-      "loss": 0.3113,
+      "epoch": 1.71426468423377,
+      "grad_norm": 0.38568932462723304,
+      "learning_rate": 9.253024415036048e-06,
+      "loss": 0.3314,
       "step": 11645
     },
     {
-      "epoch": 1.7069597069597071,
-      "grad_norm": 0.3638641870666348,
-      "learning_rate": 9.33813613919786e-06,
-      "loss": 0.3084,
+      "epoch": 1.715000736051818,
+      "grad_norm": 0.35816029310772535,
+      "learning_rate": 9.244484218459076e-06,
+      "loss": 0.3153,
       "step": 11650
     },
     {
-      "epoch": 1.7076923076923078,
-      "grad_norm": 0.37051615835261886,
-      "learning_rate": 9.329630644523194e-06,
-      "loss": 0.3158,
+      "epoch": 1.715736787869866,
+      "grad_norm": 0.3675921115941209,
+      "learning_rate": 9.235944576044749e-06,
+      "loss": 0.3292,
       "step": 11655
     },
     {
-      "epoch": 1.7084249084249086,
-      "grad_norm": 0.34965271769287465,
-      "learning_rate": 9.32112563697823e-06,
-      "loss": 0.3383,
+      "epoch": 1.7164728396879139,
+      "grad_norm": 0.3681446675015252,
+      "learning_rate": 9.227405494056801e-06,
+      "loss": 0.3168,
       "step": 11660
     },
     {
-      "epoch": 1.7091575091575093,
-      "grad_norm": 0.38475360882844284,
-      "learning_rate": 9.312621122743204e-06,
-      "loss": 0.3254,
+      "epoch": 1.717208891505962,
+      "grad_norm": 0.37302842671111597,
+      "learning_rate": 9.218866978758552e-06,
+      "loss": 0.306,
       "step": 11665
     },
     {
-      "epoch": 1.70989010989011,
-      "grad_norm": 0.387561531309359,
-      "learning_rate": 9.304117107997993e-06,
-      "loss": 0.3261,
+      "epoch": 1.7179449433240102,
+      "grad_norm": 0.36980108376078813,
+      "learning_rate": 9.21032903641292e-06,
+      "loss": 0.3295,
       "step": 11670
     },
     {
-      "epoch": 1.7106227106227108,
-      "grad_norm": 0.3603233156524001,
-      "learning_rate": 9.295613598922118e-06,
-      "loss": 0.3316,
+      "epoch": 1.718680995142058,
+      "grad_norm": 0.38557337432235433,
+      "learning_rate": 9.201791673282382e-06,
+      "loss": 0.3162,
       "step": 11675
     },
     {
-      "epoch": 1.7113553113553115,
-      "grad_norm": 0.36202320390050996,
-      "learning_rate": 9.287110601694725e-06,
-      "loss": 0.3071,
+      "epoch": 1.719417046960106,
+      "grad_norm": 0.3481759959979618,
+      "learning_rate": 9.193254895629008e-06,
+      "loss": 0.3244,
       "step": 11680
     },
     {
-      "epoch": 1.7120879120879122,
-      "grad_norm": 0.37107975876343996,
-      "learning_rate": 9.278608122494583e-06,
-      "loss": 0.3062,
+      "epoch": 1.720153098778154,
+      "grad_norm": 0.3866550958650616,
+      "learning_rate": 9.184718709714434e-06,
+      "loss": 0.3291,
       "step": 11685
     },
     {
-      "epoch": 1.712820512820513,
-      "grad_norm": 0.3485488826316779,
-      "learning_rate": 9.270106167500104e-06,
-      "loss": 0.321,
+      "epoch": 1.7208891505962018,
+      "grad_norm": 0.3419138818206259,
+      "learning_rate": 9.176183121799855e-06,
+      "loss": 0.3125,
       "step": 11690
     },
     {
-      "epoch": 1.7135531135531137,
-      "grad_norm": 0.3438649627021723,
-      "learning_rate": 9.2616047428893e-06,
-      "loss": 0.334,
+      "epoch": 1.72162520241425,
+      "grad_norm": 0.37414788069335114,
+      "learning_rate": 9.167648138146034e-06,
+      "loss": 0.3473,
       "step": 11695
     },
     {
-      "epoch": 1.7142857142857144,
-      "grad_norm": 0.3601423798069909,
-      "learning_rate": 9.253103854839809e-06,
-      "loss": 0.3275,
+      "epoch": 1.722361254232298,
+      "grad_norm": 0.3630243483220065,
+      "learning_rate": 9.15911376501329e-06,
+      "loss": 0.3323,
       "step": 11700
     },
     {
-      "epoch": 1.7150183150183151,
-      "grad_norm": 0.4273422718384637,
-      "learning_rate": 9.244603509528872e-06,
-      "loss": 0.309,
+      "epoch": 1.723097306050346,
+      "grad_norm": 0.37503200070205966,
+      "learning_rate": 9.150580008661493e-06,
+      "loss": 0.3326,
       "step": 11705
     },
     {
-      "epoch": 1.7157509157509159,
-      "grad_norm": 0.35416457301507553,
-      "learning_rate": 9.236103713133338e-06,
-      "loss": 0.3335,
+      "epoch": 1.723833357868394,
+      "grad_norm": 0.34626640175721035,
+      "learning_rate": 9.142046875350057e-06,
+      "loss": 0.3097,
       "step": 11710
     },
     {
-      "epoch": 1.7164835164835166,
-      "grad_norm": 0.36121807433007413,
-      "learning_rate": 9.227604471829661e-06,
-      "loss": 0.3182,
+      "epoch": 1.724569409686442,
+      "grad_norm": 0.36312723593020796,
+      "learning_rate": 9.133514371337946e-06,
+      "loss": 0.317,
       "step": 11715
     },
     {
-      "epoch": 1.7172161172161173,
-      "grad_norm": 0.37011667956356065,
-      "learning_rate": 9.219105791793887e-06,
-      "loss": 0.3089,
+      "epoch": 1.7253054615044898,
+      "grad_norm": 0.3606522624501329,
+      "learning_rate": 9.124982502883665e-06,
+      "loss": 0.3124,
       "step": 11720
     },
     {
-      "epoch": 1.717948717948718,
-      "grad_norm": 0.35744838489747943,
-      "learning_rate": 9.210607679201655e-06,
-      "loss": 0.3202,
+      "epoch": 1.7260415133225377,
+      "grad_norm": 0.3533629263205742,
+      "learning_rate": 9.116451276245236e-06,
+      "loss": 0.3093,
       "step": 11725
     },
     {
-      "epoch": 1.7186813186813188,
-      "grad_norm": 0.3618705083407859,
-      "learning_rate": 9.202110140228195e-06,
-      "loss": 0.3204,
+      "epoch": 1.726777565140586,
+      "grad_norm": 0.4016205710576624,
+      "learning_rate": 9.107920697680233e-06,
+      "loss": 0.3313,
       "step": 11730
     },
     {
-      "epoch": 1.7194139194139195,
-      "grad_norm": 0.3767219311852911,
-      "learning_rate": 9.193613181048307e-06,
-      "loss": 0.3156,
+      "epoch": 1.727513616958634,
+      "grad_norm": 0.3519027384206317,
+      "learning_rate": 9.099390773445732e-06,
+      "loss": 0.3138,
       "step": 11735
     },
     {
-      "epoch": 1.7201465201465203,
-      "grad_norm": 0.36440152463634456,
-      "learning_rate": 9.185116807836396e-06,
-      "loss": 0.3236,
+      "epoch": 1.728249668776682,
+      "grad_norm": 0.3380942173230528,
+      "learning_rate": 9.090861509798346e-06,
+      "loss": 0.3183,
       "step": 11740
     },
     {
-      "epoch": 1.720879120879121,
-      "grad_norm": 0.34897844104693576,
-      "learning_rate": 9.17662102676641e-06,
-      "loss": 0.325,
+      "epoch": 1.7289857205947299,
+      "grad_norm": 0.3718209014692894,
+      "learning_rate": 9.082332912994197e-06,
+      "loss": 0.344,
       "step": 11745
     },
     {
-      "epoch": 1.7216117216117217,
-      "grad_norm": 0.3859538342940784,
-      "learning_rate": 9.168125844011894e-06,
-      "loss": 0.3365,
+      "epoch": 1.7297217724127778,
+      "grad_norm": 0.36591863526661805,
+      "learning_rate": 9.073804989288916e-06,
+      "loss": 0.3246,
       "step": 11750
     },
     {
-      "epoch": 1.7223443223443224,
-      "grad_norm": 0.38341833728345387,
-      "learning_rate": 9.15963126574594e-06,
-      "loss": 0.3243,
+      "epoch": 1.7304578242308257,
+      "grad_norm": 0.36798432127102254,
+      "learning_rate": 9.065277744937649e-06,
+      "loss": 0.3235,
       "step": 11755
     },
     {
-      "epoch": 1.7230769230769232,
-      "grad_norm": 0.3816267087520552,
-      "learning_rate": 9.151137298141202e-06,
-      "loss": 0.3296,
+      "epoch": 1.7311938760488739,
+      "grad_norm": 0.3667674320692904,
+      "learning_rate": 9.056751186195027e-06,
+      "loss": 0.3178,
       "step": 11760
     },
     {
-      "epoch": 1.723809523809524,
-      "grad_norm": 0.36212006407993647,
-      "learning_rate": 9.142643947369908e-06,
-      "loss": 0.3214,
+      "epoch": 1.7319299278669218,
+      "grad_norm": 0.3576490276811153,
+      "learning_rate": 9.0482253193152e-06,
+      "loss": 0.3211,
       "step": 11765
     },
     {
-      "epoch": 1.7245421245421246,
-      "grad_norm": 0.3664742867450757,
-      "learning_rate": 9.134151219603814e-06,
-      "loss": 0.3064,
+      "epoch": 1.73266597968497,
+      "grad_norm": 0.3536654300568951,
+      "learning_rate": 9.03970015055179e-06,
+      "loss": 0.3158,
       "step": 11770
     },
     {
-      "epoch": 1.7252747252747254,
-      "grad_norm": 0.3692896138491992,
-      "learning_rate": 9.125659121014238e-06,
-      "loss": 0.3225,
+      "epoch": 1.7334020315030179,
+      "grad_norm": 0.37498095373415735,
+      "learning_rate": 9.031175686157922e-06,
+      "loss": 0.325,
       "step": 11775
     },
     {
-      "epoch": 1.726007326007326,
-      "grad_norm": 0.34267349535358754,
-      "learning_rate": 9.11716765777204e-06,
-      "loss": 0.3242,
+      "epoch": 1.7341380833210658,
+      "grad_norm": 0.3617610370844727,
+      "learning_rate": 9.02265193238619e-06,
+      "loss": 0.3227,
       "step": 11780
     },
     {
-      "epoch": 1.7267399267399268,
-      "grad_norm": 0.3637841308342026,
-      "learning_rate": 9.108676836047607e-06,
-      "loss": 0.3089,
+      "epoch": 1.7348741351391137,
+      "grad_norm": 0.37322089530807573,
+      "learning_rate": 9.014128895488687e-06,
+      "loss": 0.3258,
       "step": 11785
     },
     {
-      "epoch": 1.7274725274725276,
-      "grad_norm": 0.3642777765178664,
-      "learning_rate": 9.10018666201088e-06,
-      "loss": 0.3132,
+      "epoch": 1.7356101869571616,
+      "grad_norm": 0.38040694565479116,
+      "learning_rate": 9.005606581716963e-06,
+      "loss": 0.3224,
       "step": 11790
     },
     {
-      "epoch": 1.7282051282051283,
-      "grad_norm": 0.35229523365361093,
-      "learning_rate": 9.09169714183131e-06,
-      "loss": 0.3117,
+      "epoch": 1.7363462387752098,
+      "grad_norm": 0.3656320697872988,
+      "learning_rate": 8.997084997322037e-06,
+      "loss": 0.2979,
       "step": 11795
     },
     {
-      "epoch": 1.728937728937729,
-      "grad_norm": 0.3826853747273903,
-      "learning_rate": 9.083208281677876e-06,
-      "loss": 0.3255,
+      "epoch": 1.737082290593258,
+      "grad_norm": 0.37893430632244535,
+      "learning_rate": 8.988564148554407e-06,
+      "loss": 0.3419,
       "step": 11800
     },
     {
-      "epoch": 1.7296703296703297,
-      "grad_norm": 0.3543829606636349,
-      "learning_rate": 9.074720087719092e-06,
-      "loss": 0.3141,
+      "epoch": 1.7378183424113058,
+      "grad_norm": 0.3341585420064362,
+      "learning_rate": 8.980044041664017e-06,
+      "loss": 0.3145,
       "step": 11805
     },
     {
-      "epoch": 1.7304029304029305,
-      "grad_norm": 0.3430715485759192,
-      "learning_rate": 9.06623256612297e-06,
-      "loss": 0.3104,
+      "epoch": 1.7385543942293538,
+      "grad_norm": 0.3738781275149642,
+      "learning_rate": 8.97152468290028e-06,
+      "loss": 0.3229,
       "step": 11810
     },
     {
-      "epoch": 1.7311355311355312,
-      "grad_norm": 0.3593795927418556,
-      "learning_rate": 9.057745723057047e-06,
-      "loss": 0.3098,
+      "epoch": 1.7392904460474017,
+      "grad_norm": 0.3634285542240195,
+      "learning_rate": 8.963006078512045e-06,
+      "loss": 0.3193,
       "step": 11815
     },
     {
-      "epoch": 1.731868131868132,
-      "grad_norm": 0.3613407364550061,
-      "learning_rate": 9.049259564688358e-06,
-      "loss": 0.3202,
+      "epoch": 1.7400264978654496,
+      "grad_norm": 0.3602021798762396,
+      "learning_rate": 8.954488234747623e-06,
+      "loss": 0.3019,
       "step": 11820
     },
     {
-      "epoch": 1.7326007326007327,
-      "grad_norm": 0.37682799720557586,
-      "learning_rate": 9.04077409718344e-06,
-      "loss": 0.3301,
+      "epoch": 1.7407625496834978,
+      "grad_norm": 0.3588648373572842,
+      "learning_rate": 8.945971157854759e-06,
+      "loss": 0.3155,
       "step": 11825
     },
     {
-      "epoch": 1.7333333333333334,
-      "grad_norm": 0.3620679460262053,
-      "learning_rate": 9.03228932670834e-06,
-      "loss": 0.3247,
+      "epoch": 1.7414986015015457,
+      "grad_norm": 0.3709253063204643,
+      "learning_rate": 8.937454854080634e-06,
+      "loss": 0.3135,
       "step": 11830
     },
     {
-      "epoch": 1.7340659340659341,
-      "grad_norm": 0.35903382088234326,
-      "learning_rate": 9.023805259428584e-06,
-      "loss": 0.3267,
+      "epoch": 1.7422346533195938,
+      "grad_norm": 0.3418584837920466,
+      "learning_rate": 8.928939329671869e-06,
+      "loss": 0.314,
       "step": 11835
     },
     {
-      "epoch": 1.7347985347985349,
-      "grad_norm": 0.3696872323845864,
-      "learning_rate": 9.01532190150919e-06,
-      "loss": 0.3238,
+      "epoch": 1.7429707051376417,
+      "grad_norm": 0.35383193781994043,
+      "learning_rate": 8.920424590874501e-06,
+      "loss": 0.323,
       "step": 11840
     },
     {
-      "epoch": 1.7355311355311356,
-      "grad_norm": 0.3625054588114784,
-      "learning_rate": 9.006839259114671e-06,
-      "loss": 0.3114,
+      "epoch": 1.7437067569556897,
+      "grad_norm": 0.3814977669249137,
+      "learning_rate": 8.911910643934009e-06,
+      "loss": 0.3269,
       "step": 11845
     },
     {
-      "epoch": 1.7362637362637363,
-      "grad_norm": 0.3959261330186957,
-      "learning_rate": 8.998357338409008e-06,
-      "loss": 0.3218,
+      "epoch": 1.7444428087737376,
+      "grad_norm": 0.3562890684910219,
+      "learning_rate": 8.90339749509527e-06,
+      "loss": 0.3197,
       "step": 11850
     },
     {
-      "epoch": 1.736996336996337,
-      "grad_norm": 0.3848786684433087,
-      "learning_rate": 8.989876145555664e-06,
-      "loss": 0.3185,
+      "epoch": 1.7451788605917855,
+      "grad_norm": 0.35410266884530084,
+      "learning_rate": 8.894885150602593e-06,
+      "loss": 0.3293,
       "step": 11855
     },
     {
-      "epoch": 1.7377289377289378,
-      "grad_norm": 0.36761605390162744,
-      "learning_rate": 8.98139568671757e-06,
-      "loss": 0.3149,
+      "epoch": 1.7459149124098337,
+      "grad_norm": 0.3562446742834736,
+      "learning_rate": 8.886373616699693e-06,
+      "loss": 0.3191,
       "step": 11860
     },
     {
-      "epoch": 1.7384615384615385,
-      "grad_norm": 0.3669698126387999,
-      "learning_rate": 8.972915968057126e-06,
-      "loss": 0.3141,
+      "epoch": 1.7466509642278818,
+      "grad_norm": 0.3520153963443219,
+      "learning_rate": 8.877862899629683e-06,
+      "loss": 0.3114,
       "step": 11865
     },
     {
-      "epoch": 1.7391941391941392,
-      "grad_norm": 0.3589977908981079,
-      "learning_rate": 8.964436995736192e-06,
-      "loss": 0.3278,
+      "epoch": 1.7473870160459297,
+      "grad_norm": 0.34841949956418056,
+      "learning_rate": 8.869353005635083e-06,
+      "loss": 0.3235,
       "step": 11870
     },
     {
-      "epoch": 1.73992673992674,
-      "grad_norm": 0.36174481925983926,
-      "learning_rate": 8.955958775916085e-06,
-      "loss": 0.3191,
+      "epoch": 1.7481230678639776,
+      "grad_norm": 0.3717679710206202,
+      "learning_rate": 8.860843940957808e-06,
+      "loss": 0.3228,
       "step": 11875
     },
     {
-      "epoch": 1.7406593406593407,
-      "grad_norm": 0.3637791217098147,
-      "learning_rate": 8.947481314757584e-06,
-      "loss": 0.3245,
+      "epoch": 1.7488591196820256,
+      "grad_norm": 0.36003777702495904,
+      "learning_rate": 8.852335711839163e-06,
+      "loss": 0.3264,
       "step": 11880
     },
     {
-      "epoch": 1.7413919413919414,
-      "grad_norm": 0.35930403580444054,
-      "learning_rate": 8.939004618420904e-06,
-      "loss": 0.324,
+      "epoch": 1.7495951715000735,
+      "grad_norm": 0.37197546550778643,
+      "learning_rate": 8.843828324519848e-06,
+      "loss": 0.3265,
       "step": 11885
     },
     {
-      "epoch": 1.7421245421245422,
-      "grad_norm": 0.3613932801966608,
-      "learning_rate": 8.930528693065705e-06,
-      "loss": 0.3003,
+      "epoch": 1.7503312233181216,
+      "grad_norm": 0.3618537335555118,
+      "learning_rate": 8.835321785239933e-06,
+      "loss": 0.318,
       "step": 11890
     },
     {
-      "epoch": 1.7428571428571429,
-      "grad_norm": 0.35673925437381887,
-      "learning_rate": 8.922053544851104e-06,
-      "loss": 0.3261,
+      "epoch": 1.7510672751361696,
+      "grad_norm": 0.3588009092977853,
+      "learning_rate": 8.826816100238877e-06,
+      "loss": 0.3242,
       "step": 11895
     },
     {
-      "epoch": 1.7435897435897436,
-      "grad_norm": 0.3696486051991411,
-      "learning_rate": 8.913579179935632e-06,
-      "loss": 0.3016,
+      "epoch": 1.7518033269542177,
+      "grad_norm": 0.3517401747717276,
+      "learning_rate": 8.818311275755503e-06,
+      "loss": 0.327,
       "step": 11900
     },
     {
-      "epoch": 1.7443223443223443,
-      "grad_norm": 0.369575499643484,
-      "learning_rate": 8.905105604477259e-06,
-      "loss": 0.3193,
+      "epoch": 1.7525393787722656,
+      "grad_norm": 0.35204195451593884,
+      "learning_rate": 8.809807318028014e-06,
+      "loss": 0.3181,
       "step": 11905
     },
     {
-      "epoch": 1.745054945054945,
-      "grad_norm": 0.35873594091365585,
-      "learning_rate": 8.896632824633388e-06,
-      "loss": 0.3131,
+      "epoch": 1.7532754305903135,
+      "grad_norm": 0.3741922261189321,
+      "learning_rate": 8.801304233293966e-06,
+      "loss": 0.3193,
       "step": 11910
     },
     {
-      "epoch": 1.7457875457875458,
-      "grad_norm": 0.36175872599773284,
-      "learning_rate": 8.888160846560831e-06,
-      "loss": 0.3375,
+      "epoch": 1.7540114824083615,
+      "grad_norm": 0.3420225028684218,
+      "learning_rate": 8.792802027790277e-06,
+      "loss": 0.3098,
       "step": 11915
     },
     {
-      "epoch": 1.7465201465201465,
-      "grad_norm": 0.3535551741500669,
-      "learning_rate": 8.879689676415831e-06,
-      "loss": 0.3223,
+      "epoch": 1.7547475342264094,
+      "grad_norm": 0.3584657902915247,
+      "learning_rate": 8.784300707753233e-06,
+      "loss": 0.3108,
       "step": 11920
     },
     {
-      "epoch": 1.7472527472527473,
-      "grad_norm": 0.3587953450284838,
-      "learning_rate": 8.871219320354035e-06,
-      "loss": 0.3126,
+      "epoch": 1.7554835860444575,
+      "grad_norm": 0.3626945700051613,
+      "learning_rate": 8.775800279418448e-06,
+      "loss": 0.3323,
       "step": 11925
     },
     {
-      "epoch": 1.747985347985348,
-      "grad_norm": 0.3728671882691343,
-      "learning_rate": 8.862749784530496e-06,
-      "loss": 0.336,
+      "epoch": 1.7562196378625057,
+      "grad_norm": 0.359525665805917,
+      "learning_rate": 8.767300749020903e-06,
+      "loss": 0.3166,
       "step": 11930
     },
     {
-      "epoch": 1.7487179487179487,
-      "grad_norm": 0.352639816244091,
-      "learning_rate": 8.854281075099682e-06,
-      "loss": 0.3137,
+      "epoch": 1.7569556896805536,
+      "grad_norm": 0.3824743319272481,
+      "learning_rate": 8.758802122794905e-06,
+      "loss": 0.3222,
       "step": 11935
     },
     {
-      "epoch": 1.7494505494505495,
-      "grad_norm": 0.3629683606087605,
-      "learning_rate": 8.845813198215447e-06,
-      "loss": 0.3089,
+      "epoch": 1.7576917414986015,
+      "grad_norm": 0.36741441504023997,
+      "learning_rate": 8.750304406974112e-06,
+      "loss": 0.3238,
       "step": 11940
     },
     {
-      "epoch": 1.7501831501831502,
-      "grad_norm": 0.3629504487027256,
-      "learning_rate": 8.837346160031058e-06,
-      "loss": 0.3218,
+      "epoch": 1.7584277933166494,
+      "grad_norm": 0.41256846537416036,
+      "learning_rate": 8.741807607791499e-06,
+      "loss": 0.323,
       "step": 11945
     },
     {
-      "epoch": 1.750915750915751,
-      "grad_norm": 0.3700814012190313,
-      "learning_rate": 8.828879966699154e-06,
-      "loss": 0.3123,
+      "epoch": 1.7591638451346974,
+      "grad_norm": 0.3730241749912984,
+      "learning_rate": 8.733311731479381e-06,
+      "loss": 0.3208,
       "step": 11950
     },
     {
-      "epoch": 1.7516483516483516,
-      "grad_norm": 0.3498093037693907,
-      "learning_rate": 8.820414624371764e-06,
-      "loss": 0.3057,
+      "epoch": 1.7598998969527455,
+      "grad_norm": 0.36336133028175893,
+      "learning_rate": 8.72481678426939e-06,
+      "loss": 0.3129,
       "step": 11955
     },
     {
-      "epoch": 1.7523809523809524,
-      "grad_norm": 0.3419278360394549,
-      "learning_rate": 8.811950139200312e-06,
-      "loss": 0.3249,
+      "epoch": 1.7606359487707934,
+      "grad_norm": 0.36831929539521435,
+      "learning_rate": 8.716322772392475e-06,
+      "loss": 0.3295,
       "step": 11960
     },
     {
-      "epoch": 1.753113553113553,
-      "grad_norm": 0.36547061947732895,
-      "learning_rate": 8.803486517335582e-06,
-      "loss": 0.3092,
+      "epoch": 1.7613720005888416,
+      "grad_norm": 0.34778430647164144,
+      "learning_rate": 8.707829702078909e-06,
+      "loss": 0.3094,
       "step": 11965
     },
     {
-      "epoch": 1.7538461538461538,
-      "grad_norm": 0.3796888728609283,
-      "learning_rate": 8.79502376492774e-06,
-      "loss": 0.3167,
+      "epoch": 1.7621080524068895,
+      "grad_norm": 0.3512160561283258,
+      "learning_rate": 8.699337579558256e-06,
+      "loss": 0.3168,
       "step": 11970
     },
     {
-      "epoch": 1.7545787545787546,
-      "grad_norm": 0.35939786416676966,
-      "learning_rate": 8.786561888126322e-06,
-      "loss": 0.3181,
+      "epoch": 1.7628441042249374,
+      "grad_norm": 0.36281175964620516,
+      "learning_rate": 8.690846411059406e-06,
+      "loss": 0.3068,
       "step": 11975
     },
     {
-      "epoch": 1.7553113553113553,
-      "grad_norm": 0.3746500651417673,
-      "learning_rate": 8.778100893080218e-06,
-      "loss": 0.3123,
+      "epoch": 1.7635801560429853,
+      "grad_norm": 0.3557135443955918,
+      "learning_rate": 8.682356202810533e-06,
+      "loss": 0.322,
       "step": 11980
     },
     {
-      "epoch": 1.756043956043956,
-      "grad_norm": 0.3988165668389479,
-      "learning_rate": 8.76964078593769e-06,
-      "loss": 0.3289,
+      "epoch": 1.7643162078610333,
+      "grad_norm": 0.3862860265518121,
+      "learning_rate": 8.673866961039113e-06,
+      "loss": 0.326,
       "step": 11985
     },
     {
-      "epoch": 1.7567765567765568,
-      "grad_norm": 0.378046991149681,
-      "learning_rate": 8.761181572846346e-06,
-      "loss": 0.3226,
+      "epoch": 1.7650522596790814,
+      "grad_norm": 0.36122095809919136,
+      "learning_rate": 8.665378691971921e-06,
+      "loss": 0.3188,
       "step": 11990
     },
     {
-      "epoch": 1.7575091575091575,
-      "grad_norm": 0.3573243416450657,
-      "learning_rate": 8.752723259953139e-06,
-      "loss": 0.2911,
+      "epoch": 1.7657883114971296,
+      "grad_norm": 0.38796255344158376,
+      "learning_rate": 8.656891401835e-06,
+      "loss": 0.31,
       "step": 11995
     },
     {
-      "epoch": 1.7582417582417582,
-      "grad_norm": 0.3500972735399683,
-      "learning_rate": 8.744265853404388e-06,
-      "loss": 0.3132,
+      "epoch": 1.7665243633151775,
+      "grad_norm": 0.3826508403570477,
+      "learning_rate": 8.648405096853693e-06,
+      "loss": 0.3252,
       "step": 12000
     },
     {
-      "epoch": 1.758974358974359,
-      "grad_norm": 0.3707317801737707,
-      "learning_rate": 8.735809359345728e-06,
-      "loss": 0.3241,
+      "epoch": 1.7672604151332254,
+      "grad_norm": 0.3579974076454023,
+      "learning_rate": 8.639919783252607e-06,
+      "loss": 0.3112,
       "step": 12005
     },
     {
-      "epoch": 1.7597069597069597,
-      "grad_norm": 0.38081215733081497,
-      "learning_rate": 8.727353783922157e-06,
-      "loss": 0.3282,
+      "epoch": 1.7679964669512733,
+      "grad_norm": 0.3760365883238129,
+      "learning_rate": 8.631435467255633e-06,
+      "loss": 0.3143,
       "step": 12010
     },
     {
-      "epoch": 1.7604395604395604,
-      "grad_norm": 0.3799190748276812,
-      "learning_rate": 8.718899133277982e-06,
-      "loss": 0.3206,
+      "epoch": 1.7687325187693212,
+      "grad_norm": 0.36629558996385087,
+      "learning_rate": 8.622952155085916e-06,
+      "loss": 0.3162,
       "step": 12015
     },
     {
-      "epoch": 1.7611721611721611,
-      "grad_norm": 0.3898396143724759,
-      "learning_rate": 8.710445413556849e-06,
-      "loss": 0.3107,
+      "epoch": 1.7694685705873694,
+      "grad_norm": 0.3675236452603656,
+      "learning_rate": 8.614469852965882e-06,
+      "loss": 0.3236,
       "step": 12020
     },
     {
-      "epoch": 1.7619047619047619,
-      "grad_norm": 0.35235777614661096,
-      "learning_rate": 8.701992630901734e-06,
-      "loss": 0.3001,
+      "epoch": 1.7702046224054173,
+      "grad_norm": 0.38324532629745844,
+      "learning_rate": 8.605988567117209e-06,
+      "loss": 0.3252,
       "step": 12025
     },
     {
-      "epoch": 1.7626373626373626,
-      "grad_norm": 0.36150924715042293,
-      "learning_rate": 8.693540791454918e-06,
-      "loss": 0.3128,
+      "epoch": 1.7709406742234655,
+      "grad_norm": 0.35252116350935897,
+      "learning_rate": 8.59750830376082e-06,
+      "loss": 0.314,
       "step": 12030
     },
     {
-      "epoch": 1.7633699633699633,
-      "grad_norm": 0.36502636344229555,
-      "learning_rate": 8.685089901358e-06,
-      "loss": 0.3133,
+      "epoch": 1.7716767260415134,
+      "grad_norm": 0.36930180379288086,
+      "learning_rate": 8.589029069116906e-06,
+      "loss": 0.3265,
       "step": 12035
     },
     {
-      "epoch": 1.764102564102564,
-      "grad_norm": 0.3763387916193856,
-      "learning_rate": 8.676639966751906e-06,
-      "loss": 0.3213,
+      "epoch": 1.7724127778595613,
+      "grad_norm": 0.35890515558951985,
+      "learning_rate": 8.580550869404885e-06,
+      "loss": 0.3079,
       "step": 12040
     },
     {
-      "epoch": 1.7648351648351648,
-      "grad_norm": 0.3657571774478381,
-      "learning_rate": 8.668190993776839e-06,
-      "loss": 0.3071,
+      "epoch": 1.7731488296776092,
+      "grad_norm": 0.3532222003117065,
+      "learning_rate": 8.57207371084343e-06,
+      "loss": 0.3176,
       "step": 12045
     },
     {
-      "epoch": 1.7655677655677655,
-      "grad_norm": 0.36736547172165485,
-      "learning_rate": 8.65974298857233e-06,
-      "loss": 0.3133,
+      "epoch": 1.7738848814956572,
+      "grad_norm": 0.38119816410804086,
+      "learning_rate": 8.563597599650443e-06,
+      "loss": 0.3038,
       "step": 12050
     },
     {
-      "epoch": 1.7663003663003662,
-      "grad_norm": 0.35536369325449246,
-      "learning_rate": 8.65129595727719e-06,
-      "loss": 0.3114,
+      "epoch": 1.7746209333137053,
+      "grad_norm": 0.37005561562251305,
+      "learning_rate": 8.55512254204306e-06,
+      "loss": 0.3213,
       "step": 12055
     },
     {
-      "epoch": 1.767032967032967,
-      "grad_norm": 0.353251646692839,
-      "learning_rate": 8.642849906029524e-06,
-      "loss": 0.3229,
+      "epoch": 1.7753569851317534,
+      "grad_norm": 0.3584610558736543,
+      "learning_rate": 8.546648544237646e-06,
+      "loss": 0.3338,
       "step": 12060
     },
     {
-      "epoch": 1.7677655677655677,
-      "grad_norm": 0.35091542948003573,
-      "learning_rate": 8.634404840966735e-06,
-      "loss": 0.316,
+      "epoch": 1.7760930369498014,
+      "grad_norm": 0.3626234503128022,
+      "learning_rate": 8.538175612449782e-06,
+      "loss": 0.315,
       "step": 12065
     },
     {
-      "epoch": 1.7684981684981684,
-      "grad_norm": 0.3632197737658276,
-      "learning_rate": 8.625960768225497e-06,
-      "loss": 0.3146,
+      "epoch": 1.7768290887678493,
+      "grad_norm": 0.3488058930300539,
+      "learning_rate": 8.529703752894279e-06,
+      "loss": 0.3115,
       "step": 12070
     },
     {
-      "epoch": 1.7692307692307692,
-      "grad_norm": 0.3711999092792661,
-      "learning_rate": 8.617517693941774e-06,
-      "loss": 0.3076,
+      "epoch": 1.7775651405858972,
+      "grad_norm": 0.37069769076875203,
+      "learning_rate": 8.521232971785144e-06,
+      "loss": 0.3269,
       "step": 12075
     },
     {
-      "epoch": 1.76996336996337,
-      "grad_norm": 0.3717352113676794,
-      "learning_rate": 8.609075624250797e-06,
-      "loss": 0.3061,
+      "epoch": 1.7783011924039451,
+      "grad_norm": 0.3370641432737643,
+      "learning_rate": 8.512763275335613e-06,
+      "loss": 0.32,
       "step": 12080
     },
     {
-      "epoch": 1.7706959706959706,
-      "grad_norm": 0.3466586631315521,
-      "learning_rate": 8.600634565287068e-06,
-      "loss": 0.3139,
+      "epoch": 1.7790372442219933,
+      "grad_norm": 0.36673057430849193,
+      "learning_rate": 8.504294669758112e-06,
+      "loss": 0.3164,
       "step": 12085
     },
     {
-      "epoch": 1.7714285714285714,
-      "grad_norm": 0.3707281491997957,
-      "learning_rate": 8.592194523184358e-06,
-      "loss": 0.3212,
+      "epoch": 1.7797732960400412,
+      "grad_norm": 0.35165694757146687,
+      "learning_rate": 8.49582716126427e-06,
+      "loss": 0.3221,
       "step": 12090
     },
     {
-      "epoch": 1.772161172161172,
-      "grad_norm": 0.378197663128552,
-      "learning_rate": 8.583755504075691e-06,
-      "loss": 0.3215,
+      "epoch": 1.7805093478580893,
+      "grad_norm": 0.3538281398166754,
+      "learning_rate": 8.487360756064919e-06,
+      "loss": 0.3084,
       "step": 12095
     },
     {
-      "epoch": 1.7728937728937728,
-      "grad_norm": 0.362223401095478,
-      "learning_rate": 8.575317514093361e-06,
-      "loss": 0.3154,
+      "epoch": 1.7812453996761373,
+      "grad_norm": 0.36598785212209156,
+      "learning_rate": 8.478895460370066e-06,
+      "loss": 0.3116,
       "step": 12100
     },
     {
-      "epoch": 1.7736263736263735,
-      "grad_norm": 0.37039390221107055,
-      "learning_rate": 8.566880559368904e-06,
-      "loss": 0.3177,
+      "epoch": 1.7819814514941852,
+      "grad_norm": 0.3924230807528141,
+      "learning_rate": 8.470431280388923e-06,
+      "loss": 0.3189,
       "step": 12105
     },
     {
-      "epoch": 1.7743589743589743,
-      "grad_norm": 0.3493911690566796,
-      "learning_rate": 8.5584446460331e-06,
-      "loss": 0.3005,
+      "epoch": 1.782717503312233,
+      "grad_norm": 0.3708815885384361,
+      "learning_rate": 8.461968222329867e-06,
+      "loss": 0.3193,
       "step": 12110
     },
     {
-      "epoch": 1.775091575091575,
-      "grad_norm": 0.3624882466553709,
-      "learning_rate": 8.550009780215989e-06,
-      "loss": 0.3241,
+      "epoch": 1.783453555130281,
+      "grad_norm": 0.362426670862962,
+      "learning_rate": 8.453506292400467e-06,
+      "loss": 0.3257,
       "step": 12115
     },
     {
-      "epoch": 1.7758241758241757,
-      "grad_norm": 0.3483831840108606,
-      "learning_rate": 8.541575968046833e-06,
-      "loss": 0.3129,
+      "epoch": 1.7841896069483292,
+      "grad_norm": 0.378441819659184,
+      "learning_rate": 8.445045496807449e-06,
+      "loss": 0.3245,
       "step": 12120
     },
     {
-      "epoch": 1.7765567765567765,
-      "grad_norm": 0.35044635499766985,
-      "learning_rate": 8.533143215654133e-06,
-      "loss": 0.3081,
+      "epoch": 1.784925658766377,
+      "grad_norm": 0.36617807113448303,
+      "learning_rate": 8.436585841756722e-06,
+      "loss": 0.3237,
       "step": 12125
     },
     {
-      "epoch": 1.7772893772893772,
-      "grad_norm": 0.35353472094648314,
-      "learning_rate": 8.524711529165626e-06,
-      "loss": 0.3267,
+      "epoch": 1.7856617105844252,
+      "grad_norm": 0.34036036676664333,
+      "learning_rate": 8.428127333453348e-06,
+      "loss": 0.3131,
       "step": 12130
     },
     {
-      "epoch": 1.778021978021978,
-      "grad_norm": 0.342653777236269,
-      "learning_rate": 8.516280914708266e-06,
-      "loss": 0.3096,
+      "epoch": 1.7863977624024732,
+      "grad_norm": 0.3542528845990886,
+      "learning_rate": 8.419669978101548e-06,
+      "loss": 0.3072,
       "step": 12135
     },
     {
-      "epoch": 1.7787545787545787,
-      "grad_norm": 0.37301413986706805,
-      "learning_rate": 8.507851378408237e-06,
-      "loss": 0.3192,
+      "epoch": 1.787133814220521,
+      "grad_norm": 0.37214659955252083,
+      "learning_rate": 8.411213781904708e-06,
+      "loss": 0.3093,
       "step": 12140
     },
     {
-      "epoch": 1.7794871794871794,
-      "grad_norm": 0.377140992048411,
-      "learning_rate": 8.499422926390934e-06,
-      "loss": 0.2989,
+      "epoch": 1.787869866038569,
+      "grad_norm": 0.3636326726321218,
+      "learning_rate": 8.402758751065348e-06,
+      "loss": 0.3148,
       "step": 12145
     },
     {
-      "epoch": 1.7802197802197801,
-      "grad_norm": 0.36740624146539635,
-      "learning_rate": 8.490995564780956e-06,
-      "loss": 0.32,
+      "epoch": 1.7886059178566172,
+      "grad_norm": 0.36761872027165726,
+      "learning_rate": 8.394304891785146e-06,
+      "loss": 0.3099,
       "step": 12150
     },
     {
-      "epoch": 1.7809523809523808,
-      "grad_norm": 0.37427727435317387,
-      "learning_rate": 8.482569299702133e-06,
-      "loss": 0.3299,
+      "epoch": 1.789341969674665,
+      "grad_norm": 0.36991402909033944,
+      "learning_rate": 8.38585221026492e-06,
+      "loss": 0.3007,
       "step": 12155
     },
     {
-      "epoch": 1.7816849816849816,
-      "grad_norm": 0.34890657708463696,
-      "learning_rate": 8.474144137277467e-06,
-      "loss": 0.3224,
+      "epoch": 1.7900780214927132,
+      "grad_norm": 0.35490913152173836,
+      "learning_rate": 8.377400712704615e-06,
+      "loss": 0.303,
       "step": 12160
     },
     {
-      "epoch": 1.7824175824175823,
-      "grad_norm": 0.3627739903636609,
-      "learning_rate": 8.46572008362919e-06,
-      "loss": 0.3171,
+      "epoch": 1.7908140733107611,
+      "grad_norm": 0.37416797974475163,
+      "learning_rate": 8.368950405303313e-06,
+      "loss": 0.3185,
       "step": 12165
     },
     {
-      "epoch": 1.783150183150183,
-      "grad_norm": 0.3520214269699448,
-      "learning_rate": 8.457297144878707e-06,
-      "loss": 0.2979,
+      "epoch": 1.791550125128809,
+      "grad_norm": 0.3690968491397019,
+      "learning_rate": 8.360501294259226e-06,
+      "loss": 0.3245,
       "step": 12170
     },
     {
-      "epoch": 1.7838827838827838,
-      "grad_norm": 0.3284681340561394,
-      "learning_rate": 8.448875327146616e-06,
-      "loss": 0.3259,
+      "epoch": 1.792286176946857,
+      "grad_norm": 0.3777750945068548,
+      "learning_rate": 8.352053385769684e-06,
+      "loss": 0.322,
       "step": 12175
     },
     {
-      "epoch": 1.7846153846153845,
-      "grad_norm": 0.37039478931971476,
-      "learning_rate": 8.440454636552711e-06,
-      "loss": 0.3213,
+      "epoch": 1.793022228764905,
+      "grad_norm": 0.3522470081324157,
+      "learning_rate": 8.343606686031134e-06,
+      "loss": 0.3299,
       "step": 12180
     },
     {
-      "epoch": 1.7853479853479852,
-      "grad_norm": 0.35992801895165194,
-      "learning_rate": 8.432035079215955e-06,
-      "loss": 0.3109,
+      "epoch": 1.793758280582953,
+      "grad_norm": 0.34710180852617534,
+      "learning_rate": 8.335161201239139e-06,
+      "loss": 0.3135,
       "step": 12185
     },
     {
-      "epoch": 1.786080586080586,
-      "grad_norm": 0.3678153677791269,
-      "learning_rate": 8.423616661254492e-06,
-      "loss": 0.3083,
+      "epoch": 1.794494332401001,
+      "grad_norm": 0.38013564054672294,
+      "learning_rate": 8.326716937588377e-06,
+      "loss": 0.3006,
       "step": 12190
     },
     {
-      "epoch": 1.7868131868131867,
-      "grad_norm": 0.36959737300887124,
-      "learning_rate": 8.415199388785644e-06,
-      "loss": 0.3233,
+      "epoch": 1.7952303842190491,
+      "grad_norm": 0.36667626878510645,
+      "learning_rate": 8.318273901272615e-06,
+      "loss": 0.317,
       "step": 12195
     },
     {
-      "epoch": 1.7875457875457874,
-      "grad_norm": 0.33852252301682484,
-      "learning_rate": 8.406783267925885e-06,
-      "loss": 0.3119,
+      "epoch": 1.795966436037097,
+      "grad_norm": 0.34718088134615965,
+      "learning_rate": 8.309832098484738e-06,
+      "loss": 0.3026,
       "step": 12200
     },
     {
-      "epoch": 1.7882783882783881,
-      "grad_norm": 0.35094676778190215,
-      "learning_rate": 8.398368304790876e-06,
-      "loss": 0.3237,
+      "epoch": 1.796702487855145,
+      "grad_norm": 0.3598725298892135,
+      "learning_rate": 8.301391535416707e-06,
+      "loss": 0.3124,
       "step": 12205
     },
     {
-      "epoch": 1.7890109890109889,
-      "grad_norm": 0.35305394428075426,
-      "learning_rate": 8.38995450549541e-06,
-      "loss": 0.3293,
+      "epoch": 1.797438539673193,
+      "grad_norm": 0.3589121979620242,
+      "learning_rate": 8.292952218259593e-06,
+      "loss": 0.3137,
       "step": 12210
     },
     {
-      "epoch": 1.7897435897435896,
-      "grad_norm": 0.3681227248765676,
-      "learning_rate": 8.381541876153452e-06,
-      "loss": 0.3181,
+      "epoch": 1.798174591491241,
+      "grad_norm": 0.356684416404285,
+      "learning_rate": 8.284514153203539e-06,
+      "loss": 0.3191,
       "step": 12215
     },
     {
-      "epoch": 1.7904761904761903,
-      "grad_norm": 0.3733700036678804,
-      "learning_rate": 8.373130422878112e-06,
-      "loss": 0.3153,
+      "epoch": 1.798910643309289,
+      "grad_norm": 0.3739630741271687,
+      "learning_rate": 8.27607734643777e-06,
+      "loss": 0.3224,
       "step": 12220
     },
     {
-      "epoch": 1.791208791208791,
-      "grad_norm": 0.36975361792702527,
-      "learning_rate": 8.364720151781646e-06,
-      "loss": 0.3324,
+      "epoch": 1.799646695127337,
+      "grad_norm": 0.3618864468446184,
+      "learning_rate": 8.267641804150604e-06,
+      "loss": 0.325,
       "step": 12225
-    },
-    {
-      "epoch": 1.7919413919413918,
-      "grad_norm": 0.3763464818047654,
-      "learning_rate": 8.35631106897545e-06,
-      "loss": 0.3307,
-      "step": 12230
-    },
-    {
-      "epoch": 1.7926739926739925,
-      "grad_norm": 0.36270037639646124,
-      "learning_rate": 8.347903180570059e-06,
-      "loss": 0.3166,
-      "step": 12235
-    },
-    {
-      "epoch": 1.7934065934065933,
-      "grad_norm": 0.37520757617123063,
-      "learning_rate": 8.339496492675131e-06,
-      "loss": 0.3126,
-      "step": 12240
-    },
-    {
-      "epoch": 1.794139194139194,
-      "grad_norm": 0.371521615358303,
-      "learning_rate": 8.331091011399465e-06,
-      "loss": 0.3183,
-      "step": 12245
-    },
-    {
-      "epoch": 1.7948717948717947,
-      "grad_norm": 0.37233301877276737,
-      "learning_rate": 8.32268674285098e-06,
-      "loss": 0.3308,
-      "step": 12250
-    },
-    {
-      "epoch": 1.7956043956043954,
-      "grad_norm": 0.3527802399491386,
-      "learning_rate": 8.314283693136698e-06,
-      "loss": 0.2959,
-      "step": 12255
-    },
-    {
-      "epoch": 1.7963369963369962,
-      "grad_norm": 0.36589424525701286,
-      "learning_rate": 8.305881868362781e-06,
-      "loss": 0.3092,
-      "step": 12260
-    },
-    {
-      "epoch": 1.7970695970695971,
-      "grad_norm": 0.37079766823485905,
-      "learning_rate": 8.29748127463448e-06,
-      "loss": 0.3257,
-      "step": 12265
-    },
-    {
-      "epoch": 1.7978021978021979,
-      "grad_norm": 0.36814513089002854,
-      "learning_rate": 8.289081918056163e-06,
-      "loss": 0.3201,
-      "step": 12270
-    },
-    {
-      "epoch": 1.7985347985347986,
-      "grad_norm": 0.39298767808091684,
-      "learning_rate": 8.280683804731294e-06,
-      "loss": 0.3186,
-      "step": 12275
-    },
-    {
-      "epoch": 1.7992673992673993,
-      "grad_norm": 0.36487639427802976,
-      "learning_rate": 8.27228694076243e-06,
-      "loss": 0.3142,
-      "step": 12280
-    },
-    {
-      "epoch": 1.8,
-      "grad_norm": 0.35845251340553863,
-      "learning_rate": 8.263891332251234e-06,
-      "loss": 0.3308,
-      "step": 12285
     }
   ],
   "logging_steps": 5,
-  "max_steps": 20475,
+  "max_steps": 20379,
   "num_input_tokens_seen": 0,
   "num_train_epochs": 3,
-  "save_steps": 2048,
+  "save_steps": 2038,
   "stateful_callbacks": {
     "TrainerControl": {
       "args": {
@@ -17240,7 +17156,7 @@
       "attributes": {}
     }
   },
-  "total_flos": 1.0331011254583296e+16,
+  "total_flos": 1.0280566863691776e+16,
   "train_batch_size": 1,
   "trial_name": null,
   "trial_params": null