diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,10759 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 15310,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0006531678641410843,
+      "grad_norm": 50.13682207454293,
+      "learning_rate": 1.948051948051948e-06,
+      "loss": 1.429,
+      "step": 10
+    },
+    {
+      "epoch": 0.0013063357282821686,
+      "grad_norm": 6.483749028520749,
+      "learning_rate": 3.896103896103896e-06,
+      "loss": 1.4295,
+      "step": 20
+    },
+    {
+      "epoch": 0.001959503592423253,
+      "grad_norm": 11.258923869500972,
+      "learning_rate": 5.844155844155845e-06,
+      "loss": 1.4339,
+      "step": 30
+    },
+    {
+      "epoch": 0.002612671456564337,
+      "grad_norm": 6.7183749781782875,
+      "learning_rate": 7.792207792207792e-06,
+      "loss": 1.4304,
+      "step": 40
+    },
+    {
+      "epoch": 0.0032658393207054214,
+      "grad_norm": 11.099196437583155,
+      "learning_rate": 9.74025974025974e-06,
+      "loss": 1.3604,
+      "step": 50
+    },
+    {
+      "epoch": 0.003919007184846506,
+      "grad_norm": 15.685328846269847,
+      "learning_rate": 1.168831168831169e-05,
+      "loss": 1.3311,
+      "step": 60
+    },
+    {
+      "epoch": 0.0045721750489875895,
+      "grad_norm": 14.152624074300729,
+      "learning_rate": 1.3636363636363637e-05,
+      "loss": 1.1945,
+      "step": 70
+    },
+    {
+      "epoch": 0.005225342913128674,
+      "grad_norm": 9.464879217329338,
+      "learning_rate": 1.5584415584415583e-05,
+      "loss": 1.0984,
+      "step": 80
+    },
+    {
+      "epoch": 0.005878510777269758,
+      "grad_norm": 4.346270152613842,
+      "learning_rate": 1.753246753246753e-05,
+      "loss": 1.0648,
+      "step": 90
+    },
+    {
+      "epoch": 0.006531678641410843,
+      "grad_norm": 7.209147883203971,
+      "learning_rate": 1.948051948051948e-05,
+      "loss": 1.0054,
+      "step": 100
+    },
+    {
+      "epoch": 0.007184846505551927,
+      "grad_norm": 7.838309905384678,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 0.9712,
+      "step": 110
+    },
+    {
+      "epoch": 0.007838014369693011,
+      "grad_norm": 3.9193862374214126,
+      "learning_rate": 2.337662337662338e-05,
+      "loss": 0.9499,
+      "step": 120
+    },
+    {
+      "epoch": 0.008491182233834096,
+      "grad_norm": 8.131745340798473,
+      "learning_rate": 2.5324675324675325e-05,
+      "loss": 0.8816,
+      "step": 130
+    },
+    {
+      "epoch": 0.009144350097975179,
+      "grad_norm": 10.711063220109187,
+      "learning_rate": 2.7272727272727273e-05,
+      "loss": 0.8879,
+      "step": 140
+    },
+    {
+      "epoch": 0.009797517962116264,
+      "grad_norm": 8.374944885983533,
+      "learning_rate": 2.922077922077922e-05,
+      "loss": 0.8724,
+      "step": 150
+    },
+    {
+      "epoch": 0.010450685826257348,
+      "grad_norm": 14.67427097049884,
+      "learning_rate": 2.9999988399031158e-05,
+      "loss": 0.845,
+      "step": 160
+    },
+    {
+      "epoch": 0.011103853690398433,
+      "grad_norm": 8.900791570201601,
+      "learning_rate": 2.9999917504286554e-05,
+      "loss": 0.8545,
+      "step": 170
+    },
+    {
+      "epoch": 0.011757021554539516,
+      "grad_norm": 4.858043034268298,
+      "learning_rate": 2.999978216008427e-05,
+      "loss": 0.8396,
+      "step": 180
+    },
+    {
+      "epoch": 0.012410189418680601,
+      "grad_norm": 13.951574224432301,
+      "learning_rate": 2.999958236700584e-05,
+      "loss": 0.8306,
+      "step": 190
+    },
+    {
+      "epoch": 0.013063357282821686,
+      "grad_norm": 9.887416710966955,
+      "learning_rate": 2.99993181259097e-05,
+      "loss": 0.8278,
+      "step": 200
+    },
+    {
+      "epoch": 0.013716525146962769,
+      "grad_norm": 3.113643304504475,
+      "learning_rate": 2.9998989437931214e-05,
+      "loss": 0.8318,
+      "step": 210
+    },
+    {
+      "epoch": 0.014369693011103853,
+      "grad_norm": 3.547332330871949,
+      "learning_rate": 2.999859630448263e-05,
+      "loss": 0.831,
+      "step": 220
+    },
+    {
+      "epoch": 0.015022860875244938,
+      "grad_norm": 8.422255470301957,
+      "learning_rate": 2.999813872725311e-05,
+      "loss": 0.8438,
+      "step": 230
+    },
+    {
+      "epoch": 0.015676028739386023,
+      "grad_norm": 966.0900579138572,
+      "learning_rate": 2.9997616708208702e-05,
+      "loss": 0.7792,
+      "step": 240
+    },
+    {
+      "epoch": 0.016329196603527107,
+      "grad_norm": 9.933693210645407,
+      "learning_rate": 2.9997030249592345e-05,
+      "loss": 0.8272,
+      "step": 250
+    },
+    {
+      "epoch": 0.016982364467668192,
+      "grad_norm": 4.180403660535027,
+      "learning_rate": 2.9996379353923846e-05,
+      "loss": 0.8057,
+      "step": 260
+    },
+    {
+      "epoch": 0.017635532331809273,
+      "grad_norm": 3.9202496933723787,
+      "learning_rate": 2.999566402399988e-05,
+      "loss": 0.8032,
+      "step": 270
+    },
+    {
+      "epoch": 0.018288700195950358,
+      "grad_norm": 7.743246766273075,
+      "learning_rate": 2.9994884262893974e-05,
+      "loss": 0.8045,
+      "step": 280
+    },
+    {
+      "epoch": 0.018941868060091443,
+      "grad_norm": 3.6560068946289226,
+      "learning_rate": 2.9994040073956487e-05,
+      "loss": 0.8272,
+      "step": 290
+    },
+    {
+      "epoch": 0.019595035924232528,
+      "grad_norm": 3.92164288932011,
+      "learning_rate": 2.9993131460814615e-05,
+      "loss": 0.8225,
+      "step": 300
+    },
+    {
+      "epoch": 0.020248203788373612,
+      "grad_norm": 8.75806817966988,
+      "learning_rate": 2.9992158427372346e-05,
+      "loss": 0.8036,
+      "step": 310
+    },
+    {
+      "epoch": 0.020901371652514697,
+      "grad_norm": 4.701880610395596,
+      "learning_rate": 2.999112097781047e-05,
+      "loss": 0.7996,
+      "step": 320
+    },
+    {
+      "epoch": 0.02155453951665578,
+      "grad_norm": 7.380549266606561,
+      "learning_rate": 2.9990019116586555e-05,
+      "loss": 0.8004,
+      "step": 330
+    },
+    {
+      "epoch": 0.022207707380796866,
+      "grad_norm": 5.255075090708445,
+      "learning_rate": 2.998885284843491e-05,
+      "loss": 0.8035,
+      "step": 340
+    },
+    {
+      "epoch": 0.022860875244937948,
+      "grad_norm": 10.529978870685822,
+      "learning_rate": 2.9987622178366593e-05,
+      "loss": 0.7879,
+      "step": 350
+    },
+    {
+      "epoch": 0.023514043109079032,
+      "grad_norm": 25.688231817434417,
+      "learning_rate": 2.998632711166936e-05,
+      "loss": 0.8008,
+      "step": 360
+    },
+    {
+      "epoch": 0.024167210973220117,
+      "grad_norm": 5.948674658618657,
+      "learning_rate": 2.998496765390767e-05,
+      "loss": 0.791,
+      "step": 370
+    },
+    {
+      "epoch": 0.024820378837361202,
+      "grad_norm": 13.793404614679359,
+      "learning_rate": 2.998354381092264e-05,
+      "loss": 0.8319,
+      "step": 380
+    },
+    {
+      "epoch": 0.025473546701502287,
+      "grad_norm": 4.171433100352656,
+      "learning_rate": 2.9982055588832035e-05,
+      "loss": 0.8031,
+      "step": 390
+    },
+    {
+      "epoch": 0.02612671456564337,
+      "grad_norm": 10.096197156428968,
+      "learning_rate": 2.9980502994030224e-05,
+      "loss": 0.7615,
+      "step": 400
+    },
+    {
+      "epoch": 0.026779882429784456,
+      "grad_norm": 30.522248705094206,
+      "learning_rate": 2.9978886033188174e-05,
+      "loss": 0.7885,
+      "step": 410
+    },
+    {
+      "epoch": 0.027433050293925537,
+      "grad_norm": 6.45218649775459,
+      "learning_rate": 2.997720471325341e-05,
+      "loss": 0.776,
+      "step": 420
+    },
+    {
+      "epoch": 0.028086218158066622,
+      "grad_norm": 4.498444394274958,
+      "learning_rate": 2.9975459041449976e-05,
+      "loss": 0.8379,
+      "step": 430
+    },
+    {
+      "epoch": 0.028739386022207707,
+      "grad_norm": 2.914112407676264,
+      "learning_rate": 2.997364902527842e-05,
+      "loss": 0.8202,
+      "step": 440
+    },
+    {
+      "epoch": 0.02939255388634879,
+      "grad_norm": 3.68499903809575,
+      "learning_rate": 2.997177467251576e-05,
+      "loss": 0.79,
+      "step": 450
+    },
+    {
+      "epoch": 0.030045721750489876,
+      "grad_norm": 3.434793273613202,
+      "learning_rate": 2.996983599121544e-05,
+      "loss": 0.8031,
+      "step": 460
+    },
+    {
+      "epoch": 0.03069888961463096,
+      "grad_norm": 10.108430844869435,
+      "learning_rate": 2.99678329897073e-05,
+      "loss": 0.7911,
+      "step": 470
+    },
+    {
+      "epoch": 0.031352057478772045,
+      "grad_norm": 10.811615130428738,
+      "learning_rate": 2.996576567659755e-05,
+      "loss": 0.7627,
+      "step": 480
+    },
+    {
+      "epoch": 0.03200522534291313,
+      "grad_norm": 5.292945867397493,
+      "learning_rate": 2.9963634060768714e-05,
+      "loss": 0.7886,
+      "step": 490
+    },
+    {
+      "epoch": 0.032658393207054215,
+      "grad_norm": 5.789433805885046,
+      "learning_rate": 2.996143815137961e-05,
+      "loss": 0.7936,
+      "step": 500
+    },
+    {
+      "epoch": 0.033311561071195296,
+      "grad_norm": 4.780827921269806,
+      "learning_rate": 2.9959177957865304e-05,
+      "loss": 0.7361,
+      "step": 510
+    },
+    {
+      "epoch": 0.033964728935336384,
+      "grad_norm": 3.19844524641641,
+      "learning_rate": 2.9956853489937063e-05,
+      "loss": 0.7536,
+      "step": 520
+    },
+    {
+      "epoch": 0.034617896799477466,
+      "grad_norm": 58.45926650107535,
+      "learning_rate": 2.9954464757582325e-05,
+      "loss": 0.7816,
+      "step": 530
+    },
+    {
+      "epoch": 0.03527106466361855,
+      "grad_norm": 7.287791706613491,
+      "learning_rate": 2.995201177106464e-05,
+      "loss": 0.7597,
+      "step": 540
+    },
+    {
+      "epoch": 0.035924232527759635,
+      "grad_norm": 5.025544527071579,
+      "learning_rate": 2.9949494540923645e-05,
+      "loss": 0.7975,
+      "step": 550
+    },
+    {
+      "epoch": 0.036577400391900716,
+      "grad_norm": 16.859946084801752,
+      "learning_rate": 2.9946913077975016e-05,
+      "loss": 0.752,
+      "step": 560
+    },
+    {
+      "epoch": 0.037230568256041804,
+      "grad_norm": 9.540096992303516,
+      "learning_rate": 2.9944267393310395e-05,
+      "loss": 0.7701,
+      "step": 570
+    },
+    {
+      "epoch": 0.037883736120182886,
+      "grad_norm": 3.905591008866653,
+      "learning_rate": 2.994155749829738e-05,
+      "loss": 0.7681,
+      "step": 580
+    },
+    {
+      "epoch": 0.038536903984323974,
+      "grad_norm": 5.130639113979418,
+      "learning_rate": 2.993878340457945e-05,
+      "loss": 0.7719,
+      "step": 590
+    },
+    {
+      "epoch": 0.039190071848465055,
+      "grad_norm": 6.976901500022458,
+      "learning_rate": 2.9935945124075926e-05,
+      "loss": 0.8087,
+      "step": 600
+    },
+    {
+      "epoch": 0.039843239712606136,
+      "grad_norm": 3.2874593928044766,
+      "learning_rate": 2.9933042668981924e-05,
+      "loss": 0.7503,
+      "step": 610
+    },
+    {
+      "epoch": 0.040496407576747225,
+      "grad_norm": 6.152264074384876,
+      "learning_rate": 2.993007605176828e-05,
+      "loss": 0.7865,
+      "step": 620
+    },
+    {
+      "epoch": 0.041149575440888306,
+      "grad_norm": 4.05841892368891,
+      "learning_rate": 2.9927045285181526e-05,
+      "loss": 0.8067,
+      "step": 630
+    },
+    {
+      "epoch": 0.041802743305029394,
+      "grad_norm": 8.56078728023433,
+      "learning_rate": 2.992395038224382e-05,
+      "loss": 0.7806,
+      "step": 640
+    },
+    {
+      "epoch": 0.042455911169170475,
+      "grad_norm": 4.239768109368226,
+      "learning_rate": 2.992079135625289e-05,
+      "loss": 0.7711,
+      "step": 650
+    },
+    {
+      "epoch": 0.04310907903331156,
+      "grad_norm": 8.718034174803794,
+      "learning_rate": 2.9917568220781976e-05,
+      "loss": 0.7526,
+      "step": 660
+    },
+    {
+      "epoch": 0.043762246897452645,
+      "grad_norm": 6.8733904254859635,
+      "learning_rate": 2.9914280989679778e-05,
+      "loss": 0.7822,
+      "step": 670
+    },
+    {
+      "epoch": 0.04441541476159373,
+      "grad_norm": 4.682363111426855,
+      "learning_rate": 2.9910929677070387e-05,
+      "loss": 0.7724,
+      "step": 680
+    },
+    {
+      "epoch": 0.045068582625734814,
+      "grad_norm": 6.781911228833026,
+      "learning_rate": 2.9907514297353243e-05,
+      "loss": 0.7858,
+      "step": 690
+    },
+    {
+      "epoch": 0.045721750489875895,
+      "grad_norm": 4.18015613246061,
+      "learning_rate": 2.9904034865203037e-05,
+      "loss": 0.7328,
+      "step": 700
+    },
+    {
+      "epoch": 0.046374918354016983,
+      "grad_norm": 8.73866715173814,
+      "learning_rate": 2.99004913955697e-05,
+      "loss": 0.7489,
+      "step": 710
+    },
+    {
+      "epoch": 0.047028086218158065,
+      "grad_norm": 2.1158891476329287,
+      "learning_rate": 2.9896883903678288e-05,
+      "loss": 0.7838,
+      "step": 720
+    },
+    {
+      "epoch": 0.04768125408229915,
+      "grad_norm": 4.40400562888192,
+      "learning_rate": 2.9893212405028946e-05,
+      "loss": 0.7756,
+      "step": 730
+    },
+    {
+      "epoch": 0.048334421946440234,
+      "grad_norm": 3.2733836021629332,
+      "learning_rate": 2.9889476915396834e-05,
+      "loss": 0.8157,
+      "step": 740
+    },
+    {
+      "epoch": 0.04898758981058132,
+      "grad_norm": 9.691053563211746,
+      "learning_rate": 2.9885677450832064e-05,
+      "loss": 0.7459,
+      "step": 750
+    },
+    {
+      "epoch": 0.049640757674722404,
+      "grad_norm": 5.514889762655123,
+      "learning_rate": 2.9881814027659618e-05,
+      "loss": 0.7647,
+      "step": 760
+    },
+    {
+      "epoch": 0.050293925538863485,
+      "grad_norm": 5.300888116668702,
+      "learning_rate": 2.9877886662479287e-05,
+      "loss": 0.7747,
+      "step": 770
+    },
+    {
+      "epoch": 0.05094709340300457,
+      "grad_norm": 5.100283311940675,
+      "learning_rate": 2.98738953721656e-05,
+      "loss": 0.8026,
+      "step": 780
+    },
+    {
+      "epoch": 0.051600261267145654,
+      "grad_norm": 3.765640195443926,
+      "learning_rate": 2.986984017386776e-05,
+      "loss": 0.7805,
+      "step": 790
+    },
+    {
+      "epoch": 0.05225342913128674,
+      "grad_norm": 8.129135253261586,
+      "learning_rate": 2.986572108500954e-05,
+      "loss": 0.7925,
+      "step": 800
+    },
+    {
+      "epoch": 0.052906596995427824,
+      "grad_norm": 6.170634084606177,
+      "learning_rate": 2.9861538123289246e-05,
+      "loss": 0.7931,
+      "step": 810
+    },
+    {
+      "epoch": 0.05355976485956891,
+      "grad_norm": 8.434072292757108,
+      "learning_rate": 2.9857291306679617e-05,
+      "loss": 0.8074,
+      "step": 820
+    },
+    {
+      "epoch": 0.05421293272370999,
+      "grad_norm": 3.0182177214560464,
+      "learning_rate": 2.985298065342776e-05,
+      "loss": 0.7815,
+      "step": 830
+    },
+    {
+      "epoch": 0.054866100587851074,
+      "grad_norm": 4.281090058172293,
+      "learning_rate": 2.984860618205505e-05,
+      "loss": 0.7628,
+      "step": 840
+    },
+    {
+      "epoch": 0.05551926845199216,
+      "grad_norm": 7.669782613486172,
+      "learning_rate": 2.9844167911357088e-05,
+      "loss": 0.8085,
+      "step": 850
+    },
+    {
+      "epoch": 0.056172436316133244,
+      "grad_norm": 10.278577834140307,
+      "learning_rate": 2.983966586040358e-05,
+      "loss": 0.7776,
+      "step": 860
+    },
+    {
+      "epoch": 0.05682560418027433,
+      "grad_norm": 5.472825455483349,
+      "learning_rate": 2.9835100048538293e-05,
+      "loss": 0.7852,
+      "step": 870
+    },
+    {
+      "epoch": 0.05747877204441541,
+      "grad_norm": 18.571831717674197,
+      "learning_rate": 2.9830470495378928e-05,
+      "loss": 0.7604,
+      "step": 880
+    },
+    {
+      "epoch": 0.0581319399085565,
+      "grad_norm": 45.51617747690787,
+      "learning_rate": 2.9825777220817087e-05,
+      "loss": 0.7829,
+      "step": 890
+    },
+    {
+      "epoch": 0.05878510777269758,
+      "grad_norm": 21.44711612975514,
+      "learning_rate": 2.9821020245018137e-05,
+      "loss": 0.7728,
+      "step": 900
+    },
+    {
+      "epoch": 0.05943827563683867,
+      "grad_norm": 2.5662956039754348,
+      "learning_rate": 2.981619958842116e-05,
+      "loss": 0.8107,
+      "step": 910
+    },
+    {
+      "epoch": 0.06009144350097975,
+      "grad_norm": 6.927752270821966,
+      "learning_rate": 2.9811315271738854e-05,
+      "loss": 0.7859,
+      "step": 920
+    },
+    {
+      "epoch": 0.06074461136512083,
+      "grad_norm": 4.418111864600093,
+      "learning_rate": 2.9806367315957434e-05,
+      "loss": 0.7631,
+      "step": 930
+    },
+    {
+      "epoch": 0.06139777922926192,
+      "grad_norm": 27.811599133935445,
+      "learning_rate": 2.980135574233656e-05,
+      "loss": 0.7858,
+      "step": 940
+    },
+    {
+      "epoch": 0.062050947093403,
+      "grad_norm": 1.5924527741860024,
+      "learning_rate": 2.979628057240923e-05,
+      "loss": 0.7261,
+      "step": 950
+    },
+    {
+      "epoch": 0.06270411495754409,
+      "grad_norm": 2.631126701746248,
+      "learning_rate": 2.9791141827981684e-05,
+      "loss": 0.7734,
+      "step": 960
+    },
+    {
+      "epoch": 0.06335728282168518,
+      "grad_norm": 9.827368484273071,
+      "learning_rate": 2.9785939531133343e-05,
+      "loss": 0.7732,
+      "step": 970
+    },
+    {
+      "epoch": 0.06401045068582625,
+      "grad_norm": 3.482548940825404,
+      "learning_rate": 2.978067370421667e-05,
+      "loss": 0.762,
+      "step": 980
+    },
+    {
+      "epoch": 0.06466361854996734,
+      "grad_norm": 4.301384574294822,
+      "learning_rate": 2.9775344369857102e-05,
+      "loss": 0.7824,
+      "step": 990
+    },
+    {
+      "epoch": 0.06531678641410843,
+      "grad_norm": 6.880269476026657,
+      "learning_rate": 2.976995155095295e-05,
+      "loss": 0.7596,
+      "step": 1000
+    },
+    {
+      "epoch": 0.0659699542782495,
+      "grad_norm": 4.441018871903774,
+      "learning_rate": 2.9764495270675286e-05,
+      "loss": 0.7105,
+      "step": 1010
+    },
+    {
+      "epoch": 0.06662312214239059,
+      "grad_norm": 5.327517477621752,
+      "learning_rate": 2.975897555246786e-05,
+      "loss": 0.8219,
+      "step": 1020
+    },
+    {
+      "epoch": 0.06727629000653168,
+      "grad_norm": 6.659492383519359,
+      "learning_rate": 2.9753392420047e-05,
+      "loss": 0.7628,
+      "step": 1030
+    },
+    {
+      "epoch": 0.06792945787067277,
+      "grad_norm": 3.460680812160345,
+      "learning_rate": 2.9747745897401487e-05,
+      "loss": 0.7538,
+      "step": 1040
+    },
+    {
+      "epoch": 0.06858262573481384,
+      "grad_norm": 5.411519563956954,
+      "learning_rate": 2.9742036008792472e-05,
+      "loss": 0.7394,
+      "step": 1050
+    },
+    {
+      "epoch": 0.06923579359895493,
+      "grad_norm": 18.5454870281513,
+      "learning_rate": 2.9736262778753382e-05,
+      "loss": 0.7538,
+      "step": 1060
+    },
+    {
+      "epoch": 0.06988896146309602,
+      "grad_norm": 3.491031463918346,
+      "learning_rate": 2.9730426232089786e-05,
+      "loss": 0.7784,
+      "step": 1070
+    },
+    {
+      "epoch": 0.0705421293272371,
+      "grad_norm": 5.651477971498172,
+      "learning_rate": 2.9724526393879303e-05,
+      "loss": 0.7284,
+      "step": 1080
+    },
+    {
+      "epoch": 0.07119529719137818,
+      "grad_norm": 6.328779021951344,
+      "learning_rate": 2.9718563289471506e-05,
+      "loss": 0.7461,
+      "step": 1090
+    },
+    {
+      "epoch": 0.07184846505551927,
+      "grad_norm": 6.698982150067563,
+      "learning_rate": 2.9712536944487777e-05,
+      "loss": 0.7337,
+      "step": 1100
+    },
+    {
+      "epoch": 0.07250163291966036,
+      "grad_norm": 8.045318038088014,
+      "learning_rate": 2.970644738482125e-05,
+      "loss": 0.7653,
+      "step": 1110
+    },
+    {
+      "epoch": 0.07315480078380143,
+      "grad_norm": 7.165574308433186,
+      "learning_rate": 2.9700294636636652e-05,
+      "loss": 0.7282,
+      "step": 1120
+    },
+    {
+      "epoch": 0.07380796864794252,
+      "grad_norm": 28.043634475051775,
+      "learning_rate": 2.9694078726370218e-05,
+      "loss": 0.7691,
+      "step": 1130
+    },
+    {
+      "epoch": 0.07446113651208361,
+      "grad_norm": 3.415324032603739,
+      "learning_rate": 2.9687799680729552e-05,
+      "loss": 0.7767,
+      "step": 1140
+    },
+    {
+      "epoch": 0.07511430437622468,
+      "grad_norm": 5.9188102098763995,
+      "learning_rate": 2.9681457526693553e-05,
+      "loss": 0.7984,
+      "step": 1150
+    },
+    {
+      "epoch": 0.07576747224036577,
+      "grad_norm": 12.190132678391851,
+      "learning_rate": 2.9675052291512262e-05,
+      "loss": 0.7411,
+      "step": 1160
+    },
+    {
+      "epoch": 0.07642064010450686,
+      "grad_norm": 3.8362362001431265,
+      "learning_rate": 2.966858400270676e-05,
+      "loss": 0.7234,
+      "step": 1170
+    },
+    {
+      "epoch": 0.07707380796864795,
+      "grad_norm": 7.046244861654371,
+      "learning_rate": 2.966205268806904e-05,
+      "loss": 0.7348,
+      "step": 1180
+    },
+    {
+      "epoch": 0.07772697583278902,
+      "grad_norm": 3.5760388018830613,
+      "learning_rate": 2.9655458375661913e-05,
+      "loss": 0.7434,
+      "step": 1190
+    },
+    {
+      "epoch": 0.07838014369693011,
+      "grad_norm": 4.7886859191892714,
+      "learning_rate": 2.9648801093818846e-05,
+      "loss": 0.7574,
+      "step": 1200
+    },
+    {
+      "epoch": 0.0790333115610712,
+      "grad_norm": 3.8124034431612475,
+      "learning_rate": 2.964208087114389e-05,
+      "loss": 0.7775,
+      "step": 1210
+    },
+    {
+      "epoch": 0.07968647942521227,
+      "grad_norm": 4.986098227783862,
+      "learning_rate": 2.9635297736511505e-05,
+      "loss": 0.7407,
+      "step": 1220
+    },
+    {
+      "epoch": 0.08033964728935336,
+      "grad_norm": 5.391683192672441,
+      "learning_rate": 2.962845171906648e-05,
+      "loss": 0.7119,
+      "step": 1230
+    },
+    {
+      "epoch": 0.08099281515349445,
+      "grad_norm": 5.190092794658212,
+      "learning_rate": 2.9621542848223787e-05,
+      "loss": 0.7643,
+      "step": 1240
+    },
+    {
+      "epoch": 0.08164598301763554,
+      "grad_norm": 17.85573580184784,
+      "learning_rate": 2.961457115366845e-05,
+      "loss": 0.771,
+      "step": 1250
+    },
+    {
+      "epoch": 0.08229915088177661,
+      "grad_norm": 3.112489760048809,
+      "learning_rate": 2.960753666535543e-05,
+      "loss": 0.7856,
+      "step": 1260
+    },
+    {
+      "epoch": 0.0829523187459177,
+      "grad_norm": 8.048692812727564,
+      "learning_rate": 2.9600439413509496e-05,
+      "loss": 0.7928,
+      "step": 1270
+    },
+    {
+      "epoch": 0.08360548661005879,
+      "grad_norm": 6.309739519597919,
+      "learning_rate": 2.9593279428625078e-05,
+      "loss": 0.7604,
+      "step": 1280
+    },
+    {
+      "epoch": 0.08425865447419988,
+      "grad_norm": 6.314599622778636,
+      "learning_rate": 2.958605674146615e-05,
+      "loss": 0.7681,
+      "step": 1290
+    },
+    {
+      "epoch": 0.08491182233834095,
+      "grad_norm": 6.952462496966706,
+      "learning_rate": 2.9578771383066117e-05,
+      "loss": 0.7606,
+      "step": 1300
+    },
+    {
+      "epoch": 0.08556499020248204,
+      "grad_norm": 5.972111738893218,
+      "learning_rate": 2.9571423384727632e-05,
+      "loss": 0.7344,
+      "step": 1310
+    },
+    {
+      "epoch": 0.08621815806662313,
+      "grad_norm": 6.369021435943961,
+      "learning_rate": 2.9564012778022506e-05,
+      "loss": 0.7873,
+      "step": 1320
+    },
+    {
+      "epoch": 0.0868713259307642,
+      "grad_norm": 8.110777158554084,
+      "learning_rate": 2.955653959479155e-05,
+      "loss": 0.7646,
+      "step": 1330
+    },
+    {
+      "epoch": 0.08752449379490529,
+      "grad_norm": 4.019702548129581,
+      "learning_rate": 2.9549003867144453e-05,
+      "loss": 0.7408,
+      "step": 1340
+    },
+    {
+      "epoch": 0.08817766165904638,
+      "grad_norm": 13.973111625931505,
+      "learning_rate": 2.9541405627459627e-05,
+      "loss": 0.6958,
+      "step": 1350
+    },
+    {
+      "epoch": 0.08883082952318747,
+      "grad_norm": 2.718868998411597,
+      "learning_rate": 2.9533744908384074e-05,
+      "loss": 0.7668,
+      "step": 1360
+    },
+    {
+      "epoch": 0.08948399738732854,
+      "grad_norm": 2.8832063069963643,
+      "learning_rate": 2.9526021742833267e-05,
+      "loss": 0.7744,
+      "step": 1370
+    },
+    {
+      "epoch": 0.09013716525146963,
+      "grad_norm": 2.802666831448759,
+      "learning_rate": 2.951823616399097e-05,
+      "loss": 0.7265,
+      "step": 1380
+    },
+    {
+      "epoch": 0.09079033311561072,
+      "grad_norm": 7.0402793044257965,
+      "learning_rate": 2.9510388205309123e-05,
+      "loss": 0.7341,
+      "step": 1390
+    },
+    {
+      "epoch": 0.09144350097975179,
+      "grad_norm": 3.9078944292597333,
+      "learning_rate": 2.9502477900507694e-05,
+      "loss": 0.7243,
+      "step": 1400
+    },
+    {
+      "epoch": 0.09209666884389288,
+      "grad_norm": 3.570504815662888,
+      "learning_rate": 2.949450528357452e-05,
+      "loss": 0.7714,
+      "step": 1410
+    },
+    {
+      "epoch": 0.09274983670803397,
+      "grad_norm": 3.261257189226981,
+      "learning_rate": 2.9486470388765183e-05,
+      "loss": 0.784,
+      "step": 1420
+    },
+    {
+      "epoch": 0.09340300457217506,
+      "grad_norm": 8.416258345667368,
+      "learning_rate": 2.9478373250602844e-05,
+      "loss": 0.7665,
+      "step": 1430
+    },
+    {
+      "epoch": 0.09405617243631613,
+      "grad_norm": 4.735214914280431,
+      "learning_rate": 2.9470213903878108e-05,
+      "loss": 0.725,
+      "step": 1440
+    },
+    {
+      "epoch": 0.09470934030045722,
+      "grad_norm": 4.961243146957466,
+      "learning_rate": 2.946199238364887e-05,
+      "loss": 0.7589,
+      "step": 1450
+    },
+    {
+      "epoch": 0.0953625081645983,
+      "grad_norm": 4.324872388776254,
+      "learning_rate": 2.9453708725240154e-05,
+      "loss": 0.7854,
+      "step": 1460
+    },
+    {
+      "epoch": 0.09601567602873938,
+      "grad_norm": 4.738763468609272,
+      "learning_rate": 2.944536296424397e-05,
+      "loss": 0.7405,
+      "step": 1470
+    },
+    {
+      "epoch": 0.09666884389288047,
+      "grad_norm": 22.601514566359473,
+      "learning_rate": 2.943695513651918e-05,
+      "loss": 0.7413,
+      "step": 1480
+    },
+    {
+      "epoch": 0.09732201175702156,
+      "grad_norm": 75.19621793286049,
+      "learning_rate": 2.9428485278191295e-05,
+      "loss": 0.7446,
+      "step": 1490
+    },
+    {
+      "epoch": 0.09797517962116264,
+      "grad_norm": 5.707450334330352,
+      "learning_rate": 2.941995342565238e-05,
+      "loss": 0.7351,
+      "step": 1500
+    },
+    {
+      "epoch": 0.09862834748530372,
+      "grad_norm": 5.128650388705621,
+      "learning_rate": 2.941135961556085e-05,
+      "loss": 0.759,
+      "step": 1510
+    },
+    {
+      "epoch": 0.09928151534944481,
+      "grad_norm": 8.469700383423309,
+      "learning_rate": 2.940270388484134e-05,
+      "loss": 0.7141,
+      "step": 1520
+    },
+    {
+      "epoch": 0.0999346832135859,
+      "grad_norm": 5.139404894150258,
+      "learning_rate": 2.939398627068452e-05,
+      "loss": 0.7814,
+      "step": 1530
+    },
+    {
+      "epoch": 0.10058785107772697,
+      "grad_norm": 12.839584083356975,
+      "learning_rate": 2.938520681054697e-05,
+      "loss": 0.7216,
+      "step": 1540
+    },
+    {
+      "epoch": 0.10124101894186806,
+      "grad_norm": 12.828728051827055,
+      "learning_rate": 2.9376365542150997e-05,
+      "loss": 0.7255,
+      "step": 1550
+    },
+    {
+      "epoch": 0.10189418680600915,
+      "grad_norm": 3.271346778822822,
+      "learning_rate": 2.9367462503484467e-05,
+      "loss": 0.7404,
+      "step": 1560
+    },
+    {
+      "epoch": 0.10254735467015023,
+      "grad_norm": 5.95331499091922,
+      "learning_rate": 2.935849773280066e-05,
+      "loss": 0.7715,
+      "step": 1570
+    },
+    {
+      "epoch": 0.10320052253429131,
+      "grad_norm": 6.876843846760272,
+      "learning_rate": 2.9349471268618096e-05,
+      "loss": 0.7226,
+      "step": 1580
+    },
+    {
+      "epoch": 0.1038536903984324,
+      "grad_norm": 4.813657150858963,
+      "learning_rate": 2.9340383149720373e-05,
+      "loss": 0.7405,
+      "step": 1590
+    },
+    {
+      "epoch": 0.10450685826257348,
+      "grad_norm": 18.641291309864204,
+      "learning_rate": 2.9331233415155986e-05,
+      "loss": 0.7195,
+      "step": 1600
+    },
+    {
+      "epoch": 0.10516002612671456,
+      "grad_norm": 4.725836680776646,
+      "learning_rate": 2.9322022104238183e-05,
+      "loss": 0.7936,
+      "step": 1610
+    },
+    {
+      "epoch": 0.10581319399085565,
+      "grad_norm": 2.7058577643128507,
+      "learning_rate": 2.9312749256544788e-05,
+      "loss": 0.7124,
+      "step": 1620
+    },
+    {
+      "epoch": 0.10646636185499674,
+      "grad_norm": 2.076137194339343,
+      "learning_rate": 2.9303414911918015e-05,
+      "loss": 0.7315,
+      "step": 1630
+    },
+    {
+      "epoch": 0.10711952971913782,
+      "grad_norm": 3.1215441095992587,
+      "learning_rate": 2.9294019110464318e-05,
+      "loss": 0.7024,
+      "step": 1640
+    },
+    {
+      "epoch": 0.1077726975832789,
+      "grad_norm": 6.738642426914526,
+      "learning_rate": 2.92845618925542e-05,
+      "loss": 0.7719,
+      "step": 1650
+    },
+    {
+      "epoch": 0.10842586544741999,
+      "grad_norm": 61.13561345420421,
+      "learning_rate": 2.9275043298822054e-05,
+      "loss": 0.7832,
+      "step": 1660
+    },
+    {
+      "epoch": 0.10907903331156107,
+      "grad_norm": 5.770332213677231,
+      "learning_rate": 2.9265463370165997e-05,
+      "loss": 0.7037,
+      "step": 1670
+    },
+    {
+      "epoch": 0.10973220117570215,
+      "grad_norm": 4.453443656361965,
+      "learning_rate": 2.9255822147747658e-05,
+      "loss": 0.7521,
+      "step": 1680
+    },
+    {
+      "epoch": 0.11038536903984324,
+      "grad_norm": 3.881967986158134,
+      "learning_rate": 2.924611967299204e-05,
+      "loss": 0.7312,
+      "step": 1690
+    },
+    {
+      "epoch": 0.11103853690398433,
+      "grad_norm": 4.7459812229458604,
+      "learning_rate": 2.9236355987587325e-05,
+      "loss": 0.7439,
+      "step": 1700
+    },
+    {
+      "epoch": 0.11169170476812541,
+      "grad_norm": 3.9650318420593953,
+      "learning_rate": 2.9226531133484685e-05,
+      "loss": 0.7815,
+      "step": 1710
+    },
+    {
+      "epoch": 0.11234487263226649,
+      "grad_norm": 3.287402968914717,
+      "learning_rate": 2.9216645152898125e-05,
+      "loss": 0.7322,
+      "step": 1720
+    },
+    {
+      "epoch": 0.11299804049640758,
+      "grad_norm": 4.898218858819153,
+      "learning_rate": 2.9206698088304276e-05,
+      "loss": 0.6943,
+      "step": 1730
+    },
+    {
+      "epoch": 0.11365120836054866,
+      "grad_norm": 8.742472435174454,
+      "learning_rate": 2.919668998244225e-05,
+      "loss": 0.7678,
+      "step": 1740
+    },
+    {
+      "epoch": 0.11430437622468975,
+      "grad_norm": 11.256465495078574,
+      "learning_rate": 2.9186620878313404e-05,
+      "loss": 0.7589,
+      "step": 1750
+    },
+    {
+      "epoch": 0.11495754408883083,
+      "grad_norm": 4.467024036364152,
+      "learning_rate": 2.9176490819181196e-05,
+      "loss": 0.7555,
+      "step": 1760
+    },
+    {
+      "epoch": 0.11561071195297191,
+      "grad_norm": 7.081866692189798,
+      "learning_rate": 2.9166299848570993e-05,
+      "loss": 0.7166,
+      "step": 1770
+    },
+    {
+      "epoch": 0.116263879817113,
+      "grad_norm": 2.909828362138725,
+      "learning_rate": 2.9156048010269866e-05,
+      "loss": 0.7397,
+      "step": 1780
+    },
+    {
+      "epoch": 0.11691704768125408,
+      "grad_norm": 14.007574484697054,
+      "learning_rate": 2.9145735348326426e-05,
+      "loss": 0.7565,
+      "step": 1790
+    },
+    {
+      "epoch": 0.11757021554539517,
+      "grad_norm": 3.294438333989939,
+      "learning_rate": 2.9135361907050604e-05,
+      "loss": 0.7642,
+      "step": 1800
+    },
+    {
+      "epoch": 0.11822338340953625,
+      "grad_norm": 11.798762007605262,
+      "learning_rate": 2.9124927731013496e-05,
+      "loss": 0.7116,
+      "step": 1810
+    },
+    {
+      "epoch": 0.11887655127367734,
+      "grad_norm": 2.710752541336547,
+      "learning_rate": 2.9114432865047144e-05,
+      "loss": 0.7397,
+      "step": 1820
+    },
+    {
+      "epoch": 0.11952971913781842,
+      "grad_norm": 10.017624011337196,
+      "learning_rate": 2.9103877354244362e-05,
+      "loss": 0.7187,
+      "step": 1830
+    },
+    {
+      "epoch": 0.1201828870019595,
+      "grad_norm": 6.811000553967279,
+      "learning_rate": 2.9093261243958528e-05,
+      "loss": 0.7354,
+      "step": 1840
+    },
+    {
+      "epoch": 0.12083605486610059,
+      "grad_norm": 6.731318583444567,
+      "learning_rate": 2.908258457980339e-05,
+      "loss": 0.7384,
+      "step": 1850
+    },
+    {
+      "epoch": 0.12148922273024167,
+      "grad_norm": 5.388750098539709,
+      "learning_rate": 2.9071847407652892e-05,
+      "loss": 0.7755,
+      "step": 1860
+    },
+    {
+      "epoch": 0.12214239059438275,
+      "grad_norm": 5.4169301286838625,
+      "learning_rate": 2.9061049773640943e-05,
+      "loss": 0.7103,
+      "step": 1870
+    },
+    {
+      "epoch": 0.12279555845852384,
+      "grad_norm": 4.42587627635462,
+      "learning_rate": 2.905019172416124e-05,
+      "loss": 0.7438,
+      "step": 1880
+    },
+    {
+      "epoch": 0.12344872632266493,
+      "grad_norm": 31.054664897790516,
+      "learning_rate": 2.903927330586707e-05,
+      "loss": 0.7478,
+      "step": 1890
+    },
+    {
+      "epoch": 0.124101894186806,
+      "grad_norm": 7.972291406323152,
+      "learning_rate": 2.9028294565671097e-05,
+      "loss": 0.7421,
+      "step": 1900
+    },
+    {
+      "epoch": 0.1247550620509471,
+      "grad_norm": 7.274125571643045,
+      "learning_rate": 2.9017255550745174e-05,
+      "loss": 0.7979,
+      "step": 1910
+    },
+    {
+      "epoch": 0.12540822991508818,
+      "grad_norm": 2.8814760330477016,
+      "learning_rate": 2.900615630852013e-05,
+      "loss": 0.7396,
+      "step": 1920
+    },
+    {
+      "epoch": 0.12606139777922926,
+      "grad_norm": 4.905355189661924,
+      "learning_rate": 2.8994996886685567e-05,
+      "loss": 0.7547,
+      "step": 1930
+    },
+    {
+      "epoch": 0.12671456564337036,
+      "grad_norm": 9.505898100329482,
+      "learning_rate": 2.8983777333189662e-05,
+      "loss": 0.7304,
+      "step": 1940
+    },
+    {
+      "epoch": 0.12736773350751143,
+      "grad_norm": 3.323201006898466,
+      "learning_rate": 2.8972497696238954e-05,
+      "loss": 0.697,
+      "step": 1950
+    },
+    {
+      "epoch": 0.1280209013716525,
+      "grad_norm": 29.662689056629922,
+      "learning_rate": 2.8961158024298148e-05,
+      "loss": 0.7271,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1286740692357936,
+      "grad_norm": 7.41746219554662,
+      "learning_rate": 2.894975836608989e-05,
+      "loss": 0.7538,
+      "step": 1970
+    },
+    {
+      "epoch": 0.12932723709993468,
+      "grad_norm": 4.4427785506369775,
+      "learning_rate": 2.8938298770594568e-05,
+      "loss": 0.7547,
+      "step": 1980
+    },
+    {
+      "epoch": 0.12998040496407576,
+      "grad_norm": 7.768812174099432,
+      "learning_rate": 2.89267792870501e-05,
+      "loss": 0.7507,
+      "step": 1990
+    },
+    {
+      "epoch": 0.13063357282821686,
+      "grad_norm": 6.238069973821311,
+      "learning_rate": 2.891519996495172e-05,
+      "loss": 0.7371,
+      "step": 2000
+    },
+    {
+      "epoch": 0.13128674069235793,
+      "grad_norm": 3.5939260688441923,
+      "learning_rate": 2.8903560854051777e-05,
+      "loss": 0.7411,
+      "step": 2010
+    },
+    {
+      "epoch": 0.131939908556499,
+      "grad_norm": 14.321069542353115,
+      "learning_rate": 2.8891862004359495e-05,
+      "loss": 0.7161,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1325930764206401,
+      "grad_norm": 3.8064655011185975,
+      "learning_rate": 2.8880103466140798e-05,
+      "loss": 0.753,
+      "step": 2030
+    },
+    {
+      "epoch": 0.13324624428478118,
+      "grad_norm": 3.1071775713204413,
+      "learning_rate": 2.8868285289918044e-05,
+      "loss": 0.7919,
+      "step": 2040
+    },
+    {
+      "epoch": 0.1338994121489223,
+      "grad_norm": 7.793511836887719,
+      "learning_rate": 2.885640752646986e-05,
+      "loss": 0.7162,
+      "step": 2050
+    },
+    {
+      "epoch": 0.13455258001306336,
+      "grad_norm": 4.22301736393872,
+      "learning_rate": 2.8844470226830882e-05,
+      "loss": 0.7529,
+      "step": 2060
+    },
+    {
+      "epoch": 0.13520574787720444,
+      "grad_norm": 9.346993486220331,
+      "learning_rate": 2.883247344229156e-05,
+      "loss": 0.7674,
+      "step": 2070
+    },
+    {
+      "epoch": 0.13585891574134554,
+      "grad_norm": 15.933630619441285,
+      "learning_rate": 2.882041722439793e-05,
+      "loss": 0.7193,
+      "step": 2080
+    },
+    {
+      "epoch": 0.1365120836054866,
+      "grad_norm": 4.190682866936528,
+      "learning_rate": 2.880830162495138e-05,
+      "loss": 0.7673,
+      "step": 2090
+    },
+    {
+      "epoch": 0.13716525146962769,
+      "grad_norm": 7.037534897279013,
+      "learning_rate": 2.8796126696008465e-05,
+      "loss": 0.7254,
+      "step": 2100
+    },
+    {
+      "epoch": 0.1378184193337688,
+      "grad_norm": 2.6255399965993043,
+      "learning_rate": 2.8783892489880636e-05,
+      "loss": 0.7119,
+      "step": 2110
+    },
+    {
+      "epoch": 0.13847158719790986,
+      "grad_norm": 42.27865008405173,
+      "learning_rate": 2.8771599059134048e-05,
+      "loss": 0.7424,
+      "step": 2120
+    },
+    {
+      "epoch": 0.13912475506205094,
+      "grad_norm": 5.132678536767327,
+      "learning_rate": 2.875924645658932e-05,
+      "loss": 0.7242,
+      "step": 2130
+    },
+    {
+      "epoch": 0.13977792292619204,
+      "grad_norm": 3.603668060800223,
+      "learning_rate": 2.874683473532131e-05,
+      "loss": 0.7573,
+      "step": 2140
+    },
+    {
+      "epoch": 0.1404310907903331,
+      "grad_norm": 4.072902799746948,
+      "learning_rate": 2.8734363948658892e-05,
+      "loss": 0.715,
+      "step": 2150
+    },
+    {
+      "epoch": 0.1410842586544742,
+      "grad_norm": 22.401733113408323,
+      "learning_rate": 2.8721834150184728e-05,
+      "loss": 0.8006,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1417374265186153,
+      "grad_norm": 6.604851078546388,
+      "learning_rate": 2.8709245393735028e-05,
+      "loss": 0.7337,
+      "step": 2170
+    },
+    {
+      "epoch": 0.14239059438275636,
+      "grad_norm": 52.26699830034364,
+      "learning_rate": 2.869659773339932e-05,
+      "loss": 0.7302,
+      "step": 2180
+    },
+    {
+      "epoch": 0.14304376224689747,
+      "grad_norm": 13.544021605130228,
+      "learning_rate": 2.8683891223520228e-05,
+      "loss": 0.7646,
+      "step": 2190
+    },
+    {
+      "epoch": 0.14369693011103854,
+      "grad_norm": 9.634615764152278,
+      "learning_rate": 2.8671125918693235e-05,
+      "loss": 0.7597,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14435009797517961,
+      "grad_norm": 24.241467420017734,
+      "learning_rate": 2.865830187376643e-05,
+      "loss": 0.725,
+      "step": 2210
+    },
+    {
+      "epoch": 0.14500326583932072,
+      "grad_norm": 5.0945694153152745,
+      "learning_rate": 2.8645419143840317e-05,
+      "loss": 0.7535,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1456564337034618,
+      "grad_norm": 4.106753597412838,
+      "learning_rate": 2.8632477784267512e-05,
+      "loss": 0.6977,
+      "step": 2230
+    },
+    {
+      "epoch": 0.14630960156760286,
+      "grad_norm": 6.34150573419331,
+      "learning_rate": 2.8619477850652566e-05,
+      "loss": 0.7453,
+      "step": 2240
+    },
+    {
+      "epoch": 0.14696276943174397,
+      "grad_norm": 4.133675670803359,
+      "learning_rate": 2.8606419398851704e-05,
+      "loss": 0.7155,
+      "step": 2250
+    },
+    {
+      "epoch": 0.14761593729588504,
+      "grad_norm": 6.462096694964287,
+      "learning_rate": 2.859330248497257e-05,
+      "loss": 0.7986,
+      "step": 2260
+    },
+    {
+      "epoch": 0.14826910516002612,
+      "grad_norm": 6.261836630889005,
+      "learning_rate": 2.8580127165374016e-05,
+      "loss": 0.6918,
+      "step": 2270
+    },
+    {
+      "epoch": 0.14892227302416722,
+      "grad_norm": 2.233205529252193,
+      "learning_rate": 2.8566893496665826e-05,
+      "loss": 0.7545,
+      "step": 2280
+    },
+    {
+      "epoch": 0.1495754408883083,
+      "grad_norm": 6.483858754393392,
+      "learning_rate": 2.8553601535708498e-05,
+      "loss": 0.7325,
+      "step": 2290
+    },
+    {
+      "epoch": 0.15022860875244937,
+      "grad_norm": 5.925565913090568,
+      "learning_rate": 2.8540251339612986e-05,
+      "loss": 0.7474,
+      "step": 2300
+    },
+    {
+      "epoch": 0.15088177661659047,
+      "grad_norm": 8.4240823942178,
+      "learning_rate": 2.852684296574048e-05,
+      "loss": 0.7397,
+      "step": 2310
+    },
+    {
+      "epoch": 0.15153494448073154,
+      "grad_norm": 3.4197951314588866,
+      "learning_rate": 2.851337647170211e-05,
+      "loss": 0.7321,
+      "step": 2320
+    },
+    {
+      "epoch": 0.15218811234487264,
+      "grad_norm": 3.051619840631103,
+      "learning_rate": 2.849985191535875e-05,
+      "loss": 0.6953,
+      "step": 2330
+    },
+    {
+      "epoch": 0.15284128020901372,
+      "grad_norm": 5.255096504929576,
+      "learning_rate": 2.8486269354820743e-05,
+      "loss": 0.7239,
+      "step": 2340
+    },
+    {
+      "epoch": 0.1534944480731548,
+      "grad_norm": 4.604713902237571,
+      "learning_rate": 2.847262884844765e-05,
+      "loss": 0.772,
+      "step": 2350
+    },
+    {
+      "epoch": 0.1541476159372959,
+      "grad_norm": 4.944285498551487,
+      "learning_rate": 2.8458930454848014e-05,
+      "loss": 0.7378,
+      "step": 2360
+    },
+    {
+      "epoch": 0.15480078380143697,
+      "grad_norm": 5.82955432372675,
+      "learning_rate": 2.8445174232879087e-05,
+      "loss": 0.7163,
+      "step": 2370
+    },
+    {
+      "epoch": 0.15545395166557804,
+      "grad_norm": 3.2222937177713744,
+      "learning_rate": 2.8431360241646605e-05,
+      "loss": 0.7484,
+      "step": 2380
+    },
+    {
+      "epoch": 0.15610711952971915,
+      "grad_norm": 9.753891388404778,
+      "learning_rate": 2.8417488540504504e-05,
+      "loss": 0.7307,
+      "step": 2390
+    },
+    {
+      "epoch": 0.15676028739386022,
+      "grad_norm": 3.5077449392766566,
+      "learning_rate": 2.8403559189054692e-05,
+      "loss": 0.7638,
+      "step": 2400
+    },
+    {
+      "epoch": 0.1574134552580013,
+      "grad_norm": 5.776464182538349,
+      "learning_rate": 2.8389572247146772e-05,
+      "loss": 0.7199,
+      "step": 2410
+    },
+    {
+      "epoch": 0.1580666231221424,
+      "grad_norm": 8.250958473708572,
+      "learning_rate": 2.8375527774877795e-05,
+      "loss": 0.7389,
+      "step": 2420
+    },
+    {
+      "epoch": 0.15871979098628347,
+      "grad_norm": 5.5029051650684755,
+      "learning_rate": 2.8361425832592002e-05,
+      "loss": 0.7483,
+      "step": 2430
+    },
+    {
+      "epoch": 0.15937295885042455,
+      "grad_norm": 5.755468339066939,
+      "learning_rate": 2.8347266480880563e-05,
+      "loss": 0.7284,
+      "step": 2440
+    },
+    {
+      "epoch": 0.16002612671456565,
+      "grad_norm": 2.397047503653553,
+      "learning_rate": 2.833304978058131e-05,
+      "loss": 0.7442,
+      "step": 2450
+    },
+    {
+      "epoch": 0.16067929457870672,
+      "grad_norm": 5.380685859554883,
+      "learning_rate": 2.8318775792778497e-05,
+      "loss": 0.7466,
+      "step": 2460
+    },
+    {
+      "epoch": 0.16133246244284782,
+      "grad_norm": 44.37031913032137,
+      "learning_rate": 2.83044445788025e-05,
+      "loss": 0.7185,
+      "step": 2470
+    },
+    {
+      "epoch": 0.1619856303069889,
+      "grad_norm": 66.50805064637852,
+      "learning_rate": 2.82900562002296e-05,
+      "loss": 0.7074,
+      "step": 2480
+    },
+    {
+      "epoch": 0.16263879817112997,
+      "grad_norm": 8.037240610638753,
+      "learning_rate": 2.827561071888168e-05,
+      "loss": 0.7179,
+      "step": 2490
+    },
+    {
+      "epoch": 0.16329196603527107,
+      "grad_norm": 5.618766212652152,
+      "learning_rate": 2.8261108196825972e-05,
+      "loss": 0.7523,
+      "step": 2500
+    },
+    {
+      "epoch": 0.16394513389941215,
+      "grad_norm": 4.727081825163079,
+      "learning_rate": 2.8246548696374808e-05,
+      "loss": 0.7568,
+      "step": 2510
+    },
+    {
+      "epoch": 0.16459830176355322,
+      "grad_norm": 6.317400248396289,
+      "learning_rate": 2.8231932280085312e-05,
+      "loss": 0.7294,
+      "step": 2520
+    },
+    {
+      "epoch": 0.16525146962769433,
+      "grad_norm": 4.105422755599874,
+      "learning_rate": 2.8217259010759185e-05,
+      "loss": 0.7123,
+      "step": 2530
+    },
+    {
+      "epoch": 0.1659046374918354,
+      "grad_norm": 3.912817233451184,
+      "learning_rate": 2.820252895144238e-05,
+      "loss": 0.696,
+      "step": 2540
+    },
+    {
+      "epoch": 0.16655780535597647,
+      "grad_norm": 2.792808858923082,
+      "learning_rate": 2.8187742165424867e-05,
+      "loss": 0.6958,
+      "step": 2550
+    },
+    {
+      "epoch": 0.16721097322011758,
+      "grad_norm": 3.3697500015033515,
+      "learning_rate": 2.8172898716240358e-05,
+      "loss": 0.7514,
+      "step": 2560
+    },
+    {
+      "epoch": 0.16786414108425865,
+      "grad_norm": 12.842632607073732,
+      "learning_rate": 2.8157998667666014e-05,
+      "loss": 0.7384,
+      "step": 2570
+    },
+    {
+      "epoch": 0.16851730894839975,
+      "grad_norm": 3.272066033879194,
+      "learning_rate": 2.8143042083722196e-05,
+      "loss": 0.7468,
+      "step": 2580
+    },
+    {
+      "epoch": 0.16917047681254083,
+      "grad_norm": 7.731601163741944,
+      "learning_rate": 2.8128029028672165e-05,
+      "loss": 0.7475,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1698236446766819,
+      "grad_norm": 3.297949304055716,
+      "learning_rate": 2.8112959567021837e-05,
+      "loss": 0.7589,
+      "step": 2600
+    },
+    {
+      "epoch": 0.170476812540823,
+      "grad_norm": 14.339494194357066,
+      "learning_rate": 2.809783376351947e-05,
+      "loss": 0.7546,
+      "step": 2610
+    },
+    {
+      "epoch": 0.17112998040496408,
+      "grad_norm": 4.46409481839347,
+      "learning_rate": 2.808265168315541e-05,
+      "loss": 0.7425,
+      "step": 2620
+    },
+    {
+      "epoch": 0.17178314826910515,
+      "grad_norm": 9.466307069996041,
+      "learning_rate": 2.80674133911618e-05,
+      "loss": 0.7441,
+      "step": 2630
+    },
+    {
+      "epoch": 0.17243631613324625,
+      "grad_norm": 7.071614823083529,
+      "learning_rate": 2.805211895301233e-05,
+      "loss": 0.72,
+      "step": 2640
+    },
+    {
+      "epoch": 0.17308948399738733,
+      "grad_norm": 4.5445298901654745,
+      "learning_rate": 2.803676843442189e-05,
+      "loss": 0.7564,
+      "step": 2650
+    },
+    {
+      "epoch": 0.1737426518615284,
+      "grad_norm": 4.607374060860952,
+      "learning_rate": 2.8021361901346356e-05,
+      "loss": 0.7509,
+      "step": 2660
+    },
+    {
+      "epoch": 0.1743958197256695,
+      "grad_norm": 3.916261462631034,
+      "learning_rate": 2.8005899419982276e-05,
+      "loss": 0.7363,
+      "step": 2670
+    },
+    {
+      "epoch": 0.17504898758981058,
+      "grad_norm": 4.285270418160644,
+      "learning_rate": 2.7990381056766583e-05,
+      "loss": 0.6819,
+      "step": 2680
+    },
+    {
+      "epoch": 0.17570215545395165,
+      "grad_norm": 30.066405996248537,
+      "learning_rate": 2.7974806878376312e-05,
+      "loss": 0.7257,
+      "step": 2690
+    },
+    {
+      "epoch": 0.17635532331809275,
+      "grad_norm": 6.215186550197811,
+      "learning_rate": 2.7959176951728326e-05,
+      "loss": 0.7205,
+      "step": 2700
+    },
+    {
+      "epoch": 0.17700849118223383,
+      "grad_norm": 6.606929967094239,
+      "learning_rate": 2.7943491343979012e-05,
+      "loss": 0.6971,
+      "step": 2710
+    },
+    {
+      "epoch": 0.17766165904637493,
+      "grad_norm": 4.870287647251215,
+      "learning_rate": 2.7927750122524004e-05,
+      "loss": 0.738,
+      "step": 2720
+    },
+    {
+      "epoch": 0.178314826910516,
+      "grad_norm": 3.840109737338219,
+      "learning_rate": 2.7911953354997882e-05,
+      "loss": 0.7033,
+      "step": 2730
+    },
+    {
+      "epoch": 0.17896799477465708,
+      "grad_norm": 8.663692208470097,
+      "learning_rate": 2.78961011092739e-05,
+      "loss": 0.7536,
+      "step": 2740
+    },
+    {
+      "epoch": 0.17962116263879818,
+      "grad_norm": 8.706739021144726,
+      "learning_rate": 2.7880193453463664e-05,
+      "loss": 0.7418,
+      "step": 2750
+    },
+    {
+      "epoch": 0.18027433050293926,
+      "grad_norm": 4.286484907893715,
+      "learning_rate": 2.786423045591688e-05,
+      "loss": 0.7254,
+      "step": 2760
+    },
+    {
+      "epoch": 0.18092749836708033,
+      "grad_norm": 4.090802478364384,
+      "learning_rate": 2.7848212185221025e-05,
+      "loss": 0.7362,
+      "step": 2770
+    },
+    {
+      "epoch": 0.18158066623122143,
+      "grad_norm": 11.328580271023869,
+      "learning_rate": 2.783213871020106e-05,
+      "loss": 0.7241,
+      "step": 2780
+    },
+    {
+      "epoch": 0.1822338340953625,
+      "grad_norm": 8.48461958671395,
+      "learning_rate": 2.7816010099919157e-05,
+      "loss": 0.6719,
+      "step": 2790
+    },
+    {
+      "epoch": 0.18288700195950358,
+      "grad_norm": 3.2686306933480678,
+      "learning_rate": 2.7799826423674376e-05,
+      "loss": 0.7089,
+      "step": 2800
+    },
+    {
+      "epoch": 0.18354016982364468,
+      "grad_norm": 65.07123500441651,
+      "learning_rate": 2.7783587751002373e-05,
+      "loss": 0.7057,
+      "step": 2810
+    },
+    {
+      "epoch": 0.18419333768778576,
+      "grad_norm": 10.1627349338379,
+      "learning_rate": 2.776729415167511e-05,
+      "loss": 0.7236,
+      "step": 2820
+    },
+    {
+      "epoch": 0.18484650555192683,
+      "grad_norm": 2.5258850888099715,
+      "learning_rate": 2.7750945695700545e-05,
+      "loss": 0.7182,
+      "step": 2830
+    },
+    {
+      "epoch": 0.18549967341606793,
+      "grad_norm": 2.9763625671042457,
+      "learning_rate": 2.773454245332234e-05,
+      "loss": 0.7246,
+      "step": 2840
+    },
+    {
+      "epoch": 0.186152841280209,
+      "grad_norm": 4.674364122198806,
+      "learning_rate": 2.771808449501956e-05,
+      "loss": 0.7512,
+      "step": 2850
+    },
+    {
+      "epoch": 0.1868060091443501,
+      "grad_norm": 8.972997636490977,
+      "learning_rate": 2.770157189150635e-05,
+      "loss": 0.7218,
+      "step": 2860
+    },
+    {
+      "epoch": 0.18745917700849118,
+      "grad_norm": 6.751308738143618,
+      "learning_rate": 2.7685004713731667e-05,
+      "loss": 0.7219,
+      "step": 2870
+    },
+    {
+      "epoch": 0.18811234487263226,
+      "grad_norm": 9.421091936001982,
+      "learning_rate": 2.766838303287894e-05,
+      "loss": 0.6971,
+      "step": 2880
+    },
+    {
+      "epoch": 0.18876551273677336,
+      "grad_norm": 25.71006507749567,
+      "learning_rate": 2.7651706920365778e-05,
+      "loss": 0.7211,
+      "step": 2890
+    },
+    {
+      "epoch": 0.18941868060091444,
+      "grad_norm": 7.097688766178269,
+      "learning_rate": 2.7634976447843673e-05,
+      "loss": 0.7289,
+      "step": 2900
+    },
+    {
+      "epoch": 0.1900718484650555,
+      "grad_norm": 3.012421324402975,
+      "learning_rate": 2.761819168719768e-05,
+      "loss": 0.7334,
+      "step": 2910
+    },
+    {
+      "epoch": 0.1907250163291966,
+      "grad_norm": 6.763816256941492,
+      "learning_rate": 2.760135271054611e-05,
+      "loss": 0.7342,
+      "step": 2920
+    },
+    {
+      "epoch": 0.19137818419333769,
+      "grad_norm": 12.339869717446561,
+      "learning_rate": 2.7584459590240213e-05,
+      "loss": 0.7673,
+      "step": 2930
+    },
+    {
+      "epoch": 0.19203135205747876,
+      "grad_norm": 9.943986419226533,
+      "learning_rate": 2.75675123988639e-05,
+      "loss": 0.7316,
+      "step": 2940
+    },
+    {
+      "epoch": 0.19268451992161986,
+      "grad_norm": 3.9625775254424487,
+      "learning_rate": 2.7550511209233377e-05,
+      "loss": 0.7656,
+      "step": 2950
+    },
+    {
+      "epoch": 0.19333768778576094,
+      "grad_norm": 6.646184770465973,
+      "learning_rate": 2.753345609439689e-05,
+      "loss": 0.7096,
+      "step": 2960
+    },
+    {
+      "epoch": 0.19399085564990204,
+      "grad_norm": 3.363167542629984,
+      "learning_rate": 2.751634712763435e-05,
+      "loss": 0.773,
+      "step": 2970
+    },
+    {
+      "epoch": 0.1946440235140431,
+      "grad_norm": 9.64459690768457,
+      "learning_rate": 2.749918438245709e-05,
+      "loss": 0.7521,
+      "step": 2980
+    },
+    {
+      "epoch": 0.1952971913781842,
+      "grad_norm": 7.66917331463007,
+      "learning_rate": 2.7481967932607478e-05,
+      "loss": 0.7682,
+      "step": 2990
+    },
+    {
+      "epoch": 0.1959503592423253,
+      "grad_norm": 7.4386800903665975,
+      "learning_rate": 2.7464697852058648e-05,
+      "loss": 0.6853,
+      "step": 3000
+    },
+    {
+      "epoch": 0.19660352710646636,
+      "grad_norm": 5.948601553568303,
+      "learning_rate": 2.7447374215014157e-05,
+      "loss": 0.7079,
+      "step": 3010
+    },
+    {
+      "epoch": 0.19725669497060744,
+      "grad_norm": 4.017629160392304,
+      "learning_rate": 2.742999709590769e-05,
+      "loss": 0.6856,
+      "step": 3020
+    },
+    {
+      "epoch": 0.19790986283474854,
+      "grad_norm": 6.6087233490439905,
+      "learning_rate": 2.741256656940272e-05,
+      "loss": 0.7229,
+      "step": 3030
+    },
+    {
+      "epoch": 0.19856303069888961,
+      "grad_norm": 5.738147627571181,
+      "learning_rate": 2.7395082710392183e-05,
+      "loss": 0.721,
+      "step": 3040
+    },
+    {
+      "epoch": 0.1992161985630307,
+      "grad_norm": 4.753298367179536,
+      "learning_rate": 2.7377545593998178e-05,
+      "loss": 0.7019,
+      "step": 3050
+    },
+    {
+      "epoch": 0.1998693664271718,
+      "grad_norm": 6.611797204555651,
+      "learning_rate": 2.7359955295571624e-05,
+      "loss": 0.7115,
+      "step": 3060
+    },
+    {
+      "epoch": 0.20052253429131286,
+      "grad_norm": 4.432909888406863,
+      "learning_rate": 2.7342311890691957e-05,
+      "loss": 0.7427,
+      "step": 3070
+    },
+    {
+      "epoch": 0.20117570215545394,
+      "grad_norm": 12.633636844045604,
+      "learning_rate": 2.7324615455166778e-05,
+      "loss": 0.7014,
+      "step": 3080
+    },
+    {
+      "epoch": 0.20182887001959504,
+      "grad_norm": 3.514345470279303,
+      "learning_rate": 2.7306866065031562e-05,
+      "loss": 0.7306,
+      "step": 3090
+    },
+    {
+      "epoch": 0.20248203788373612,
+      "grad_norm": 7.274767695647385,
+      "learning_rate": 2.728906379654929e-05,
+      "loss": 0.7801,
+      "step": 3100
+    },
+    {
+      "epoch": 0.20313520574787722,
+      "grad_norm": 3.0139129997105454,
+      "learning_rate": 2.727120872621015e-05,
+      "loss": 0.7143,
+      "step": 3110
+    },
+    {
+      "epoch": 0.2037883736120183,
+      "grad_norm": 5.802657459099795,
+      "learning_rate": 2.7253300930731212e-05,
+      "loss": 0.7374,
+      "step": 3120
+    },
+    {
+      "epoch": 0.20444154147615937,
+      "grad_norm": 3.3669875366596598,
+      "learning_rate": 2.7235340487056074e-05,
+      "loss": 0.7172,
+      "step": 3130
+    },
+    {
+      "epoch": 0.20509470934030047,
+      "grad_norm": 5.503009319660073,
+      "learning_rate": 2.7217327472354555e-05,
+      "loss": 0.7321,
+      "step": 3140
+    },
+    {
+      "epoch": 0.20574787720444154,
+      "grad_norm": 5.726618455558819,
+      "learning_rate": 2.7199261964022345e-05,
+      "loss": 0.7416,
+      "step": 3150
+    },
+    {
+      "epoch": 0.20640104506858262,
+      "grad_norm": 14.751272372954604,
+      "learning_rate": 2.7181144039680688e-05,
+      "loss": 0.732,
+      "step": 3160
+    },
+    {
+      "epoch": 0.20705421293272372,
+      "grad_norm": 6.756690771244652,
+      "learning_rate": 2.7162973777176033e-05,
+      "loss": 0.7102,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2077073807968648,
+      "grad_norm": 5.508517519150598,
+      "learning_rate": 2.7144751254579727e-05,
+      "loss": 0.7163,
+      "step": 3180
+    },
+    {
+      "epoch": 0.20836054866100587,
+      "grad_norm": 3.295298267114904,
+      "learning_rate": 2.7126476550187635e-05,
+      "loss": 0.7882,
+      "step": 3190
+    },
+    {
+      "epoch": 0.20901371652514697,
+      "grad_norm": 4.285106842952411,
+      "learning_rate": 2.7108149742519842e-05,
+      "loss": 0.7027,
+      "step": 3200
+    },
+    {
+      "epoch": 0.20966688438928804,
+      "grad_norm": 4.589013804618004,
+      "learning_rate": 2.7089770910320312e-05,
+      "loss": 0.7351,
+      "step": 3210
+    },
+    {
+      "epoch": 0.21032005225342912,
+      "grad_norm": 7.427253910226353,
+      "learning_rate": 2.7071340132556518e-05,
+      "loss": 0.7341,
+      "step": 3220
+    },
+    {
+      "epoch": 0.21097322011757022,
+      "grad_norm": 7.242467453040407,
+      "learning_rate": 2.7052857488419146e-05,
+      "loss": 0.7321,
+      "step": 3230
+    },
+    {
+      "epoch": 0.2116263879817113,
+      "grad_norm": 3.3718873257981823,
+      "learning_rate": 2.703432305732172e-05,
+      "loss": 0.7518,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2122795558458524,
+      "grad_norm": 4.628943334503347,
+      "learning_rate": 2.701573691890029e-05,
+      "loss": 0.7681,
+      "step": 3250
+    },
+    {
+      "epoch": 0.21293272370999347,
+      "grad_norm": 3.692360083654589,
+      "learning_rate": 2.6997099153013053e-05,
+      "loss": 0.7438,
+      "step": 3260
+    },
+    {
+      "epoch": 0.21358589157413455,
+      "grad_norm": 6.026452562923238,
+      "learning_rate": 2.6978409839740045e-05,
+      "loss": 0.6755,
+      "step": 3270
+    },
+    {
+      "epoch": 0.21423905943827565,
+      "grad_norm": 8.388146401979114,
+      "learning_rate": 2.6959669059382787e-05,
+      "loss": 0.7152,
+      "step": 3280
+    },
+    {
+      "epoch": 0.21489222730241672,
+      "grad_norm": 5.489947061474543,
+      "learning_rate": 2.6940876892463924e-05,
+      "loss": 0.7331,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2155453951665578,
+      "grad_norm": 5.238150530271462,
+      "learning_rate": 2.6922033419726903e-05,
+      "loss": 0.7216,
+      "step": 3300
+    },
+    {
+      "epoch": 0.2161985630306989,
+      "grad_norm": 5.76694380401119,
+      "learning_rate": 2.690313872213561e-05,
+      "loss": 0.7041,
+      "step": 3310
+    },
+    {
+      "epoch": 0.21685173089483997,
+      "grad_norm": 4.334325872361498,
+      "learning_rate": 2.6884192880874018e-05,
+      "loss": 0.7035,
+      "step": 3320
+    },
+    {
+      "epoch": 0.21750489875898105,
+      "grad_norm": 3.646615269686379,
+      "learning_rate": 2.6865195977345864e-05,
+      "loss": 0.7135,
+      "step": 3330
+    },
+    {
+      "epoch": 0.21815806662312215,
+      "grad_norm": 3.2031026699869516,
+      "learning_rate": 2.6846148093174266e-05,
+      "loss": 0.7478,
+      "step": 3340
+    },
+    {
+      "epoch": 0.21881123448726322,
+      "grad_norm": 3.294578873577567,
+      "learning_rate": 2.6827049310201392e-05,
+      "loss": 0.6966,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2194644023514043,
+      "grad_norm": 13.531582658656262,
+      "learning_rate": 2.6807899710488118e-05,
+      "loss": 0.7714,
+      "step": 3360
+    },
+    {
+      "epoch": 0.2201175702155454,
+      "grad_norm": 3.1667911772760817,
+      "learning_rate": 2.6788699376313635e-05,
+      "loss": 0.6883,
+      "step": 3370
+    },
+    {
+      "epoch": 0.22077073807968647,
+      "grad_norm": 12.709667365090638,
+      "learning_rate": 2.6769448390175156e-05,
+      "loss": 0.7083,
+      "step": 3380
+    },
+    {
+      "epoch": 0.22142390594382758,
+      "grad_norm": 4.90247272458579,
+      "learning_rate": 2.67501468347875e-05,
+      "loss": 0.7254,
+      "step": 3390
+    },
+    {
+      "epoch": 0.22207707380796865,
+      "grad_norm": 6.586314850084481,
+      "learning_rate": 2.673079479308277e-05,
+      "loss": 0.7053,
+      "step": 3400
+    },
+    {
+      "epoch": 0.22273024167210972,
+      "grad_norm": 8.67302524395751,
+      "learning_rate": 2.671139234821001e-05,
+      "loss": 0.7461,
+      "step": 3410
+    },
+    {
+      "epoch": 0.22338340953625083,
+      "grad_norm": 4.199679298872582,
+      "learning_rate": 2.669193958353481e-05,
+      "loss": 0.7108,
+      "step": 3420
+    },
+    {
+      "epoch": 0.2240365774003919,
+      "grad_norm": 3.7113020225982543,
+      "learning_rate": 2.6672436582638962e-05,
+      "loss": 0.7379,
+      "step": 3430
+    },
+    {
+      "epoch": 0.22468974526453298,
+      "grad_norm": 3.1660823450229953,
+      "learning_rate": 2.6652883429320127e-05,
+      "loss": 0.7458,
+      "step": 3440
+    },
+    {
+      "epoch": 0.22534291312867408,
+      "grad_norm": 4.557885122260074,
+      "learning_rate": 2.6633280207591434e-05,
+      "loss": 0.7384,
+      "step": 3450
+    },
+    {
+      "epoch": 0.22599608099281515,
+      "grad_norm": 3.206803010538826,
+      "learning_rate": 2.6613627001681156e-05,
+      "loss": 0.7837,
+      "step": 3460
+    },
+    {
+      "epoch": 0.22664924885695623,
+      "grad_norm": 5.426168451300572,
+      "learning_rate": 2.659392389603232e-05,
+      "loss": 0.7091,
+      "step": 3470
+    },
+    {
+      "epoch": 0.22730241672109733,
+      "grad_norm": 5.983994434496897,
+      "learning_rate": 2.6574170975302347e-05,
+      "loss": 0.7322,
+      "step": 3480
+    },
+    {
+      "epoch": 0.2279555845852384,
+      "grad_norm": 4.1889732388324346,
+      "learning_rate": 2.6554368324362716e-05,
+      "loss": 0.7274,
+      "step": 3490
+    },
+    {
+      "epoch": 0.2286087524493795,
+      "grad_norm": 4.279384234331947,
+      "learning_rate": 2.653451602829856e-05,
+      "loss": 0.7167,
+      "step": 3500
+    },
+    {
+      "epoch": 0.22926192031352058,
+      "grad_norm": 10.493028415319717,
+      "learning_rate": 2.6514614172408342e-05,
+      "loss": 0.7518,
+      "step": 3510
+    },
+    {
+      "epoch": 0.22991508817766165,
+      "grad_norm": 5.127178380468306,
+      "learning_rate": 2.649466284220344e-05,
+      "loss": 0.7087,
+      "step": 3520
+    },
+    {
+      "epoch": 0.23056825604180275,
+      "grad_norm": 4.447652673443472,
+      "learning_rate": 2.6474662123407827e-05,
+      "loss": 0.7088,
+      "step": 3530
+    },
+    {
+      "epoch": 0.23122142390594383,
+      "grad_norm": 4.424791216908705,
+      "learning_rate": 2.6454612101957676e-05,
+      "loss": 0.7059,
+      "step": 3540
+    },
+    {
+      "epoch": 0.2318745917700849,
+      "grad_norm": 5.883257966824098,
+      "learning_rate": 2.6434512864000988e-05,
+      "loss": 0.7456,
+      "step": 3550
+    },
+    {
+      "epoch": 0.232527759634226,
+      "grad_norm": 4.3204215451597845,
+      "learning_rate": 2.6414364495897242e-05,
+      "loss": 0.7413,
+      "step": 3560
+    },
+    {
+      "epoch": 0.23318092749836708,
+      "grad_norm": 6.554390974464567,
+      "learning_rate": 2.6394167084217005e-05,
+      "loss": 0.7181,
+      "step": 3570
+    },
+    {
+      "epoch": 0.23383409536250815,
+      "grad_norm": 6.3921681525273994,
+      "learning_rate": 2.637392071574157e-05,
+      "loss": 0.7177,
+      "step": 3580
+    },
+    {
+      "epoch": 0.23448726322664926,
+      "grad_norm": 4.851980784102187,
+      "learning_rate": 2.635362547746258e-05,
+      "loss": 0.7402,
+      "step": 3590
+    },
+    {
+      "epoch": 0.23514043109079033,
+      "grad_norm": 3.4119093890395322,
+      "learning_rate": 2.6333281456581654e-05,
+      "loss": 0.691,
+      "step": 3600
+    },
+    {
+      "epoch": 0.2357935989549314,
+      "grad_norm": 3.8612085843913864,
+      "learning_rate": 2.631288874051002e-05,
+      "loss": 0.7416,
+      "step": 3610
+    },
+    {
+      "epoch": 0.2364467668190725,
+      "grad_norm": 17.80518970057431,
+      "learning_rate": 2.6292447416868113e-05,
+      "loss": 0.7185,
+      "step": 3620
+    },
+    {
+      "epoch": 0.23709993468321358,
+      "grad_norm": 8.691379915690442,
+      "learning_rate": 2.6271957573485244e-05,
+      "loss": 0.7232,
+      "step": 3630
+    },
+    {
+      "epoch": 0.23775310254735468,
+      "grad_norm": 7.889136326928524,
+      "learning_rate": 2.6251419298399176e-05,
+      "loss": 0.7473,
+      "step": 3640
+    },
+    {
+      "epoch": 0.23840627041149576,
+      "grad_norm": 4.379578292960399,
+      "learning_rate": 2.6230832679855773e-05,
+      "loss": 0.7269,
+      "step": 3650
+    },
+    {
+      "epoch": 0.23905943827563683,
+      "grad_norm": 5.353751120664064,
+      "learning_rate": 2.6210197806308617e-05,
+      "loss": 0.7662,
+      "step": 3660
+    },
+    {
+      "epoch": 0.23971260613977793,
+      "grad_norm": 2.7709561275339194,
+      "learning_rate": 2.6189514766418625e-05,
+      "loss": 0.7094,
+      "step": 3670
+    },
+    {
+      "epoch": 0.240365774003919,
+      "grad_norm": 3.9103422168330204,
+      "learning_rate": 2.6168783649053666e-05,
+      "loss": 0.7028,
+      "step": 3680
+    },
+    {
+      "epoch": 0.24101894186806008,
+      "grad_norm": 9.505874423721908,
+      "learning_rate": 2.6148004543288178e-05,
+      "loss": 0.7255,
+      "step": 3690
+    },
+    {
+      "epoch": 0.24167210973220118,
+      "grad_norm": 7.653416946344283,
+      "learning_rate": 2.6127177538402795e-05,
+      "loss": 0.7207,
+      "step": 3700
+    },
+    {
+      "epoch": 0.24232527759634226,
+      "grad_norm": 6.251847566275377,
+      "learning_rate": 2.6106302723883952e-05,
+      "loss": 0.7724,
+      "step": 3710
+    },
+    {
+      "epoch": 0.24297844546048333,
+      "grad_norm": 7.31365044828085,
+      "learning_rate": 2.60853801894235e-05,
+      "loss": 0.694,
+      "step": 3720
+    },
+    {
+      "epoch": 0.24363161332462444,
+      "grad_norm": 10.798141206505173,
+      "learning_rate": 2.6064410024918352e-05,
+      "loss": 0.7268,
+      "step": 3730
+    },
+    {
+      "epoch": 0.2442847811887655,
+      "grad_norm": 7.3726815927726195,
+      "learning_rate": 2.6043392320470033e-05,
+      "loss": 0.7419,
+      "step": 3740
+    },
+    {
+      "epoch": 0.24493794905290658,
+      "grad_norm": 8.397591708864457,
+      "learning_rate": 2.6022327166384363e-05,
+      "loss": 0.7248,
+      "step": 3750
+    },
+    {
+      "epoch": 0.24559111691704769,
+      "grad_norm": 6.615639461597031,
+      "learning_rate": 2.600121465317102e-05,
+      "loss": 0.737,
+      "step": 3760
+    },
+    {
+      "epoch": 0.24624428478118876,
+      "grad_norm": 12.252576907293445,
+      "learning_rate": 2.5980054871543167e-05,
+      "loss": 0.678,
+      "step": 3770
+    },
+    {
+      "epoch": 0.24689745264532986,
+      "grad_norm": 6.008549638823585,
+      "learning_rate": 2.5958847912417065e-05,
+      "loss": 0.7436,
+      "step": 3780
+    },
+    {
+      "epoch": 0.24755062050947094,
+      "grad_norm": 9.731321783664672,
+      "learning_rate": 2.5937593866911694e-05,
+      "loss": 0.6884,
+      "step": 3790
+    },
+    {
+      "epoch": 0.248203788373612,
+      "grad_norm": 4.872437219190829,
+      "learning_rate": 2.5916292826348327e-05,
+      "loss": 0.7551,
+      "step": 3800
+    },
+    {
+      "epoch": 0.2488569562377531,
+      "grad_norm": 3.2546621774905304,
+      "learning_rate": 2.5894944882250177e-05,
+      "loss": 0.7235,
+      "step": 3810
+    },
+    {
+      "epoch": 0.2495101241018942,
+      "grad_norm": 9.424086986405833,
+      "learning_rate": 2.5873550126341963e-05,
+      "loss": 0.6996,
+      "step": 3820
+    },
+    {
+      "epoch": 0.25016329196603526,
+      "grad_norm": 5.514870221849538,
+      "learning_rate": 2.585210865054957e-05,
+      "loss": 0.7349,
+      "step": 3830
+    },
+    {
+      "epoch": 0.25081645983017636,
+      "grad_norm": 6.9013441794446315,
+      "learning_rate": 2.5830620546999587e-05,
+      "loss": 0.7647,
+      "step": 3840
+    },
+    {
+      "epoch": 0.25146962769431747,
+      "grad_norm": 6.6845750918764875,
+      "learning_rate": 2.580908590801897e-05,
+      "loss": 0.7193,
+      "step": 3850
+    },
+    {
+      "epoch": 0.2521227955584585,
+      "grad_norm": 10.961278212697293,
+      "learning_rate": 2.5787504826134613e-05,
+      "loss": 0.7236,
+      "step": 3860
+    },
+    {
+      "epoch": 0.2527759634225996,
+      "grad_norm": 12.837907116460197,
+      "learning_rate": 2.5765877394072965e-05,
+      "loss": 0.6781,
+      "step": 3870
+    },
+    {
+      "epoch": 0.2534291312867407,
+      "grad_norm": 10.64515408154912,
+      "learning_rate": 2.5744203704759616e-05,
+      "loss": 0.7158,
+      "step": 3880
+    },
+    {
+      "epoch": 0.25408229915088176,
+      "grad_norm": 5.828039582881938,
+      "learning_rate": 2.572248385131892e-05,
+      "loss": 0.7243,
+      "step": 3890
+    },
+    {
+      "epoch": 0.25473546701502287,
+      "grad_norm": 4.7378551420496064,
+      "learning_rate": 2.5700717927073572e-05,
+      "loss": 0.686,
+      "step": 3900
+    },
+    {
+      "epoch": 0.25538863487916397,
+      "grad_norm": 5.101256286871471,
+      "learning_rate": 2.5678906025544212e-05,
+      "loss": 0.6929,
+      "step": 3910
+    },
+    {
+      "epoch": 0.256041802743305,
+      "grad_norm": 4.945280487776523,
+      "learning_rate": 2.5657048240449055e-05,
+      "loss": 0.7152,
+      "step": 3920
+    },
+    {
+      "epoch": 0.2566949706074461,
+      "grad_norm": 6.325141259828677,
+      "learning_rate": 2.5635144665703425e-05,
+      "loss": 0.7386,
+      "step": 3930
+    },
+    {
+      "epoch": 0.2573481384715872,
+      "grad_norm": 3.328094625558282,
+      "learning_rate": 2.5613195395419422e-05,
+      "loss": 0.7853,
+      "step": 3940
+    },
+    {
+      "epoch": 0.25800130633572826,
+      "grad_norm": 14.77901385371437,
+      "learning_rate": 2.559120052390546e-05,
+      "loss": 0.7757,
+      "step": 3950
+    },
+    {
+      "epoch": 0.25865447419986937,
+      "grad_norm": 2.926012995140229,
+      "learning_rate": 2.55691601456659e-05,
+      "loss": 0.6735,
+      "step": 3960
+    },
+    {
+      "epoch": 0.25930764206401047,
+      "grad_norm": 5.225338186608472,
+      "learning_rate": 2.5547074355400615e-05,
+      "loss": 0.7265,
+      "step": 3970
+    },
+    {
+      "epoch": 0.2599608099281515,
+      "grad_norm": 4.206340254882354,
+      "learning_rate": 2.5524943248004618e-05,
+      "loss": 0.7145,
+      "step": 3980
+    },
+    {
+      "epoch": 0.2606139777922926,
+      "grad_norm": 6.368887270719955,
+      "learning_rate": 2.550276691856762e-05,
+      "loss": 0.7423,
+      "step": 3990
+    },
+    {
+      "epoch": 0.2612671456564337,
+      "grad_norm": 3.0483125865376963,
+      "learning_rate": 2.548054546237364e-05,
+      "loss": 0.7132,
+      "step": 4000
+    },
+    {
+      "epoch": 0.26192031352057477,
+      "grad_norm": 13.296685653396278,
+      "learning_rate": 2.5458278974900587e-05,
+      "loss": 0.7084,
+      "step": 4010
+    },
+    {
+      "epoch": 0.26257348138471587,
+      "grad_norm": 38.457562726457006,
+      "learning_rate": 2.5435967551819856e-05,
+      "loss": 0.7265,
+      "step": 4020
+    },
+    {
+      "epoch": 0.26322664924885697,
+      "grad_norm": 109.85436026280432,
+      "learning_rate": 2.5413611288995915e-05,
+      "loss": 0.729,
+      "step": 4030
+    },
+    {
+      "epoch": 0.263879817112998,
+      "grad_norm": 8.077784990426677,
+      "learning_rate": 2.53912102824859e-05,
+      "loss": 0.7015,
+      "step": 4040
+    },
+    {
+      "epoch": 0.2645329849771391,
+      "grad_norm": 8.861234493320074,
+      "learning_rate": 2.5368764628539184e-05,
+      "loss": 0.7495,
+      "step": 4050
+    },
+    {
+      "epoch": 0.2651861528412802,
+      "grad_norm": 3.554047847668293,
+      "learning_rate": 2.5346274423596973e-05,
+      "loss": 0.7424,
+      "step": 4060
+    },
+    {
+      "epoch": 0.26583932070542127,
+      "grad_norm": 2.927566126492755,
+      "learning_rate": 2.5323739764291912e-05,
+      "loss": 0.7611,
+      "step": 4070
+    },
+    {
+      "epoch": 0.26649248856956237,
+      "grad_norm": 3.4058451280668876,
+      "learning_rate": 2.5301160747447627e-05,
+      "loss": 0.6972,
+      "step": 4080
+    },
+    {
+      "epoch": 0.26714565643370347,
+      "grad_norm": 3.0373695188526497,
+      "learning_rate": 2.5278537470078352e-05,
+      "loss": 0.7033,
+      "step": 4090
+    },
+    {
+      "epoch": 0.2677988242978446,
+      "grad_norm": 5.525119120836987,
+      "learning_rate": 2.525587002938848e-05,
+      "loss": 0.7608,
+      "step": 4100
+    },
+    {
+      "epoch": 0.2684519921619856,
+      "grad_norm": 5.173021264413592,
+      "learning_rate": 2.5233158522772166e-05,
+      "loss": 0.7528,
+      "step": 4110
+    },
+    {
+      "epoch": 0.2691051600261267,
+      "grad_norm": 7.9505349609036235,
+      "learning_rate": 2.5210403047812896e-05,
+      "loss": 0.7787,
+      "step": 4120
+    },
+    {
+      "epoch": 0.2697583278902678,
+      "grad_norm": 8.093183932153524,
+      "learning_rate": 2.518760370228308e-05,
+      "loss": 0.7288,
+      "step": 4130
+    },
+    {
+      "epoch": 0.27041149575440887,
+      "grad_norm": 7.35362845406468,
+      "learning_rate": 2.516476058414362e-05,
+      "loss": 0.7048,
+      "step": 4140
+    },
+    {
+      "epoch": 0.27106466361855,
+      "grad_norm": 6.063117241653435,
+      "learning_rate": 2.5141873791543494e-05,
+      "loss": 0.7099,
+      "step": 4150
+    },
+    {
+      "epoch": 0.2717178314826911,
+      "grad_norm": 6.2465652182913445,
+      "learning_rate": 2.511894342281933e-05,
+      "loss": 0.7636,
+      "step": 4160
+    },
+    {
+      "epoch": 0.2723709993468321,
+      "grad_norm": 2.4005462620859435,
+      "learning_rate": 2.5095969576494998e-05,
+      "loss": 0.7347,
+      "step": 4170
+    },
+    {
+      "epoch": 0.2730241672109732,
+      "grad_norm": 4.287489527920315,
+      "learning_rate": 2.5072952351281166e-05,
+      "loss": 0.7047,
+      "step": 4180
+    },
+    {
+      "epoch": 0.2736773350751143,
+      "grad_norm": 10.924198105492543,
+      "learning_rate": 2.504989184607489e-05,
+      "loss": 0.7135,
+      "step": 4190
+    },
+    {
+      "epoch": 0.27433050293925537,
+      "grad_norm": 3.344600105958022,
+      "learning_rate": 2.502678815995919e-05,
+      "loss": 0.7566,
+      "step": 4200
+    },
+    {
+      "epoch": 0.2749836708033965,
+      "grad_norm": 8.178872258203794,
+      "learning_rate": 2.500364139220261e-05,
+      "loss": 0.7012,
+      "step": 4210
+    },
+    {
+      "epoch": 0.2756368386675376,
+      "grad_norm": 13.792796170330694,
+      "learning_rate": 2.4980451642258807e-05,
+      "loss": 0.7214,
+      "step": 4220
+    },
+    {
+      "epoch": 0.2762900065316786,
+      "grad_norm": 4.686318820906931,
+      "learning_rate": 2.495721900976611e-05,
+      "loss": 0.7561,
+      "step": 4230
+    },
+    {
+      "epoch": 0.2769431743958197,
+      "grad_norm": 52.23368325147696,
+      "learning_rate": 2.4933943594547116e-05,
+      "loss": 0.7079,
+      "step": 4240
+    },
+    {
+      "epoch": 0.2775963422599608,
+      "grad_norm": 3.3254162864810706,
+      "learning_rate": 2.4910625496608227e-05,
+      "loss": 0.7151,
+      "step": 4250
+    },
+    {
+      "epoch": 0.2782495101241019,
+      "grad_norm": 6.428077097460673,
+      "learning_rate": 2.488726481613925e-05,
+      "loss": 0.7109,
+      "step": 4260
+    },
+    {
+      "epoch": 0.278902677988243,
+      "grad_norm": 13.474459188221212,
+      "learning_rate": 2.4863861653512947e-05,
+      "loss": 0.7513,
+      "step": 4270
+    },
+    {
+      "epoch": 0.2795558458523841,
+      "grad_norm": 5.57604753719738,
+      "learning_rate": 2.484041610928461e-05,
+      "loss": 0.7523,
+      "step": 4280
+    },
+    {
+      "epoch": 0.2802090137165251,
+      "grad_norm": 7.417290679168985,
+      "learning_rate": 2.4816928284191642e-05,
+      "loss": 0.7464,
+      "step": 4290
+    },
+    {
+      "epoch": 0.2808621815806662,
+      "grad_norm": 7.4368198628749616,
+      "learning_rate": 2.4793398279153098e-05,
+      "loss": 0.7309,
+      "step": 4300
+    },
+    {
+      "epoch": 0.28151534944480733,
+      "grad_norm": 5.404604799289123,
+      "learning_rate": 2.4769826195269276e-05,
+      "loss": 0.6812,
+      "step": 4310
+    },
+    {
+      "epoch": 0.2821685173089484,
+      "grad_norm": 5.67385871603199,
+      "learning_rate": 2.474621213382126e-05,
+      "loss": 0.7485,
+      "step": 4320
+    },
+    {
+      "epoch": 0.2828216851730895,
+      "grad_norm": 3.327246034991726,
+      "learning_rate": 2.4722556196270516e-05,
+      "loss": 0.7491,
+      "step": 4330
+    },
+    {
+      "epoch": 0.2834748530372306,
+      "grad_norm": 4.298927510493578,
+      "learning_rate": 2.4698858484258413e-05,
+      "loss": 0.7447,
+      "step": 4340
+    },
+    {
+      "epoch": 0.2841280209013716,
+      "grad_norm": 2.830645539219132,
+      "learning_rate": 2.4675119099605832e-05,
+      "loss": 0.6611,
+      "step": 4350
+    },
+    {
+      "epoch": 0.2847811887655127,
+      "grad_norm": 2.892594280107655,
+      "learning_rate": 2.46513381443127e-05,
+      "loss": 0.7574,
+      "step": 4360
+    },
+    {
+      "epoch": 0.28543435662965383,
+      "grad_norm": 2.8096820701671392,
+      "learning_rate": 2.462751572055755e-05,
+      "loss": 0.7352,
+      "step": 4370
+    },
+    {
+      "epoch": 0.28608752449379493,
+      "grad_norm": 6.625013819266248,
+      "learning_rate": 2.46036519306971e-05,
+      "loss": 0.6894,
+      "step": 4380
+    },
+    {
+      "epoch": 0.286740692357936,
+      "grad_norm": 5.553982675208715,
+      "learning_rate": 2.457974687726581e-05,
+      "loss": 0.69,
+      "step": 4390
+    },
+    {
+      "epoch": 0.2873938602220771,
+      "grad_norm": 10.463948438278225,
+      "learning_rate": 2.4555800662975415e-05,
+      "loss": 0.7381,
+      "step": 4400
+    },
+    {
+      "epoch": 0.2880470280862182,
+      "grad_norm": 5.237933324400403,
+      "learning_rate": 2.4531813390714523e-05,
+      "loss": 0.705,
+      "step": 4410
+    },
+    {
+      "epoch": 0.28870019595035923,
+      "grad_norm": 29.3661171326619,
+      "learning_rate": 2.4507785163548145e-05,
+      "loss": 0.6982,
+      "step": 4420
+    },
+    {
+      "epoch": 0.28935336381450033,
+      "grad_norm": 5.569785158681364,
+      "learning_rate": 2.448371608471726e-05,
+      "loss": 0.7602,
+      "step": 4430
+    },
+    {
+      "epoch": 0.29000653167864143,
+      "grad_norm": 2.7943484866446475,
+      "learning_rate": 2.4459606257638375e-05,
+      "loss": 0.7343,
+      "step": 4440
+    },
+    {
+      "epoch": 0.2906596995427825,
+      "grad_norm": 2.9668248409437354,
+      "learning_rate": 2.4435455785903088e-05,
+      "loss": 0.6952,
+      "step": 4450
+    },
+    {
+      "epoch": 0.2913128674069236,
+      "grad_norm": 4.0679875797768545,
+      "learning_rate": 2.441126477327761e-05,
+      "loss": 0.7106,
+      "step": 4460
+    },
+    {
+      "epoch": 0.2919660352710647,
+      "grad_norm": 12.16575396175707,
+      "learning_rate": 2.4387033323702364e-05,
+      "loss": 0.712,
+      "step": 4470
+    },
+    {
+      "epoch": 0.29261920313520573,
+      "grad_norm": 2.8440806437673234,
+      "learning_rate": 2.4362761541291502e-05,
+      "loss": 0.7658,
+      "step": 4480
+    },
+    {
+      "epoch": 0.29327237099934683,
+      "grad_norm": 2.970610346903458,
+      "learning_rate": 2.433844953033249e-05,
+      "loss": 0.7424,
+      "step": 4490
+    },
+    {
+      "epoch": 0.29392553886348793,
+      "grad_norm": 6.591526614830109,
+      "learning_rate": 2.431409739528562e-05,
+      "loss": 0.7057,
+      "step": 4500
+    },
+    {
+      "epoch": 0.294578706727629,
+      "grad_norm": 3.5000170298570765,
+      "learning_rate": 2.42897052407836e-05,
+      "loss": 0.7207,
+      "step": 4510
+    },
+    {
+      "epoch": 0.2952318745917701,
+      "grad_norm": 4.09870087115134,
+      "learning_rate": 2.4265273171631077e-05,
+      "loss": 0.7326,
+      "step": 4520
+    },
+    {
+      "epoch": 0.2958850424559112,
+      "grad_norm": 3.634194650545928,
+      "learning_rate": 2.42408012928042e-05,
+      "loss": 0.7509,
+      "step": 4530
+    },
+    {
+      "epoch": 0.29653821032005223,
+      "grad_norm": 4.634696693084635,
+      "learning_rate": 2.4216289709450176e-05,
+      "loss": 0.7639,
+      "step": 4540
+    },
+    {
+      "epoch": 0.29719137818419333,
+      "grad_norm": 1.674361047291477,
+      "learning_rate": 2.4191738526886794e-05,
+      "loss": 0.7451,
+      "step": 4550
+    },
+    {
+      "epoch": 0.29784454604833444,
+      "grad_norm": 4.679840951482941,
+      "learning_rate": 2.4167147850601998e-05,
+      "loss": 0.6997,
+      "step": 4560
+    },
+    {
+      "epoch": 0.2984977139124755,
+      "grad_norm": 15.092169321509138,
+      "learning_rate": 2.414251778625342e-05,
+      "loss": 0.7105,
+      "step": 4570
+    },
+    {
+      "epoch": 0.2991508817766166,
+      "grad_norm": 4.95638690004173,
+      "learning_rate": 2.411784843966793e-05,
+      "loss": 0.7067,
+      "step": 4580
+    },
+    {
+      "epoch": 0.2998040496407577,
+      "grad_norm": 9.521604766867073,
+      "learning_rate": 2.4093139916841172e-05,
+      "loss": 0.7225,
+      "step": 4590
+    },
+    {
+      "epoch": 0.30045721750489873,
+      "grad_norm": 10.371086393594378,
+      "learning_rate": 2.4068392323937125e-05,
+      "loss": 0.6966,
+      "step": 4600
+    },
+    {
+      "epoch": 0.30111038536903983,
+      "grad_norm": 6.1990655256347535,
+      "learning_rate": 2.4043605767287643e-05,
+      "loss": 0.7165,
+      "step": 4610
+    },
+    {
+      "epoch": 0.30176355323318094,
+      "grad_norm": 7.466217760942189,
+      "learning_rate": 2.4018780353391978e-05,
+      "loss": 0.6912,
+      "step": 4620
+    },
+    {
+      "epoch": 0.30241672109732204,
+      "grad_norm": 5.232860562458902,
+      "learning_rate": 2.3993916188916348e-05,
+      "loss": 0.7015,
+      "step": 4630
+    },
+    {
+      "epoch": 0.3030698889614631,
+      "grad_norm": 7.8628232759082195,
+      "learning_rate": 2.396901338069348e-05,
+      "loss": 0.701,
+      "step": 4640
+    },
+    {
+      "epoch": 0.3037230568256042,
+      "grad_norm": 5.6841565040341315,
+      "learning_rate": 2.394407203572211e-05,
+      "loss": 0.7673,
+      "step": 4650
+    },
+    {
+      "epoch": 0.3043762246897453,
+      "grad_norm": 8.54205331314945,
+      "learning_rate": 2.3919092261166584e-05,
+      "loss": 0.6904,
+      "step": 4660
+    },
+    {
+      "epoch": 0.30502939255388634,
+      "grad_norm": 11.156476357778907,
+      "learning_rate": 2.3894074164356353e-05,
+      "loss": 0.6855,
+      "step": 4670
+    },
+    {
+      "epoch": 0.30568256041802744,
+      "grad_norm": 7.870284591240595,
+      "learning_rate": 2.3869017852785525e-05,
+      "loss": 0.724,
+      "step": 4680
+    },
+    {
+      "epoch": 0.30633572828216854,
+      "grad_norm": 3.3237816492979007,
+      "learning_rate": 2.3843923434112402e-05,
+      "loss": 0.7228,
+      "step": 4690
+    },
+    {
+      "epoch": 0.3069888961463096,
+      "grad_norm": 2.480450775802249,
+      "learning_rate": 2.3818791016159022e-05,
+      "loss": 0.7184,
+      "step": 4700
+    },
+    {
+      "epoch": 0.3076420640104507,
+      "grad_norm": 7.916224283272688,
+      "learning_rate": 2.3793620706910696e-05,
+      "loss": 0.7388,
+      "step": 4710
+    },
+    {
+      "epoch": 0.3082952318745918,
+      "grad_norm": 2.943344146547367,
+      "learning_rate": 2.3768412614515536e-05,
+      "loss": 0.6956,
+      "step": 4720
+    },
+    {
+      "epoch": 0.30894839973873284,
+      "grad_norm": 3.327029912173773,
+      "learning_rate": 2.3743166847283995e-05,
+      "loss": 0.7233,
+      "step": 4730
+    },
+    {
+      "epoch": 0.30960156760287394,
+      "grad_norm": 6.533958348071706,
+      "learning_rate": 2.3717883513688405e-05,
+      "loss": 0.7083,
+      "step": 4740
+    },
+    {
+      "epoch": 0.31025473546701504,
+      "grad_norm": 2.2289245925613885,
+      "learning_rate": 2.3692562722362508e-05,
+      "loss": 0.746,
+      "step": 4750
+    },
+    {
+      "epoch": 0.3109079033311561,
+      "grad_norm": 3.7748651745700323,
+      "learning_rate": 2.3667204582100984e-05,
+      "loss": 0.6574,
+      "step": 4760
+    },
+    {
+      "epoch": 0.3115610711952972,
+      "grad_norm": 3.0117659311519964,
+      "learning_rate": 2.3641809201858996e-05,
+      "loss": 0.732,
+      "step": 4770
+    },
+    {
+      "epoch": 0.3122142390594383,
+      "grad_norm": 6.5912357978805245,
+      "learning_rate": 2.3616376690751703e-05,
+      "loss": 0.7292,
+      "step": 4780
+    },
+    {
+      "epoch": 0.31286740692357934,
+      "grad_norm": 4.055781986644913,
+      "learning_rate": 2.359090715805381e-05,
+      "loss": 0.7227,
+      "step": 4790
+    },
+    {
+      "epoch": 0.31352057478772044,
+      "grad_norm": 5.128394514646037,
+      "learning_rate": 2.3565400713199095e-05,
+      "loss": 0.7592,
+      "step": 4800
+    },
+    {
+      "epoch": 0.31417374265186154,
+      "grad_norm": 8.779130673134816,
+      "learning_rate": 2.3539857465779925e-05,
+      "loss": 0.6676,
+      "step": 4810
+    },
+    {
+      "epoch": 0.3148269105160026,
+      "grad_norm": 5.429036887728827,
+      "learning_rate": 2.3514277525546803e-05,
+      "loss": 0.7009,
+      "step": 4820
+    },
+    {
+      "epoch": 0.3154800783801437,
+      "grad_norm": 7.8542004760770014,
+      "learning_rate": 2.348866100240789e-05,
+      "loss": 0.7233,
+      "step": 4830
+    },
+    {
+      "epoch": 0.3161332462442848,
+      "grad_norm": 6.478615135277363,
+      "learning_rate": 2.3463008006428506e-05,
+      "loss": 0.7015,
+      "step": 4840
+    },
+    {
+      "epoch": 0.31678641410842584,
+      "grad_norm": 3.4721596945324875,
+      "learning_rate": 2.343731864783073e-05,
+      "loss": 0.7295,
+      "step": 4850
+    },
+    {
+      "epoch": 0.31743958197256694,
+      "grad_norm": 5.134712506262902,
+      "learning_rate": 2.3411593036992835e-05,
+      "loss": 0.7464,
+      "step": 4860
+    },
+    {
+      "epoch": 0.31809274983670804,
+      "grad_norm": 5.049954351842052,
+      "learning_rate": 2.3385831284448873e-05,
+      "loss": 0.7473,
+      "step": 4870
+    },
+    {
+      "epoch": 0.3187459177008491,
+      "grad_norm": 7.027799717697589,
+      "learning_rate": 2.336003350088819e-05,
+      "loss": 0.688,
+      "step": 4880
+    },
+    {
+      "epoch": 0.3193990855649902,
+      "grad_norm": 4.380900953812384,
+      "learning_rate": 2.3334199797154936e-05,
+      "loss": 0.6974,
+      "step": 4890
+    },
+    {
+      "epoch": 0.3200522534291313,
+      "grad_norm": 1.802352319241488,
+      "learning_rate": 2.3308330284247605e-05,
+      "loss": 0.6911,
+      "step": 4900
+    },
+    {
+      "epoch": 0.3207054212932724,
+      "grad_norm": 3.6480027783199236,
+      "learning_rate": 2.3282425073318546e-05,
+      "loss": 0.7089,
+      "step": 4910
+    },
+    {
+      "epoch": 0.32135858915741344,
+      "grad_norm": 8.846036258285976,
+      "learning_rate": 2.3256484275673486e-05,
+      "loss": 0.7138,
+      "step": 4920
+    },
+    {
+      "epoch": 0.32201175702155455,
+      "grad_norm": 5.168086297852975,
+      "learning_rate": 2.3230508002771067e-05,
+      "loss": 0.7055,
+      "step": 4930
+    },
+    {
+      "epoch": 0.32266492488569565,
+      "grad_norm": 3.9475777647489396,
+      "learning_rate": 2.320449636622235e-05,
+      "loss": 0.7257,
+      "step": 4940
+    },
+    {
+      "epoch": 0.3233180927498367,
+      "grad_norm": 4.931806927889069,
+      "learning_rate": 2.3178449477790325e-05,
+      "loss": 0.7071,
+      "step": 4950
+    },
+    {
+      "epoch": 0.3239712606139778,
+      "grad_norm": 2.2215574528530437,
+      "learning_rate": 2.3152367449389483e-05,
+      "loss": 0.7037,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3246244284781189,
+      "grad_norm": 8.690398027898421,
+      "learning_rate": 2.312625039308528e-05,
+      "loss": 0.7256,
+      "step": 4970
+    },
+    {
+      "epoch": 0.32527759634225994,
+      "grad_norm": 10.206026981904913,
+      "learning_rate": 2.3100098421093655e-05,
+      "loss": 0.6887,
+      "step": 4980
+    },
+    {
+      "epoch": 0.32593076420640105,
+      "grad_norm": 3.80442191887017,
+      "learning_rate": 2.3073911645780602e-05,
+      "loss": 0.7179,
+      "step": 4990
+    },
+    {
+      "epoch": 0.32658393207054215,
+      "grad_norm": 4.728308085787767,
+      "learning_rate": 2.304769017966163e-05,
+      "loss": 0.7809,
+      "step": 5000
+    },
+    {
+      "epoch": 0.3272370999346832,
+      "grad_norm": 11.391935045550042,
+      "learning_rate": 2.302143413540132e-05,
+      "loss": 0.7669,
+      "step": 5010
+    },
+    {
+      "epoch": 0.3278902677988243,
+      "grad_norm": 11.911310517206102,
+      "learning_rate": 2.2995143625812804e-05,
+      "loss": 0.7085,
+      "step": 5020
+    },
+    {
+      "epoch": 0.3285434356629654,
+      "grad_norm": 7.262195721290905,
+      "learning_rate": 2.296881876385731e-05,
+      "loss": 0.7089,
+      "step": 5030
+    },
+    {
+      "epoch": 0.32919660352710645,
+      "grad_norm": 8.385837204775681,
+      "learning_rate": 2.2942459662643667e-05,
+      "loss": 0.7249,
+      "step": 5040
+    },
+    {
+      "epoch": 0.32984977139124755,
+      "grad_norm": 9.484288524755074,
+      "learning_rate": 2.291606643542782e-05,
+      "loss": 0.7342,
+      "step": 5050
+    },
+    {
+      "epoch": 0.33050293925538865,
+      "grad_norm": 4.6070263067396535,
+      "learning_rate": 2.288963919561233e-05,
+      "loss": 0.7352,
+      "step": 5060
+    },
+    {
+      "epoch": 0.3311561071195297,
+      "grad_norm": 4.701683051508864,
+      "learning_rate": 2.2863178056745913e-05,
+      "loss": 0.737,
+      "step": 5070
+    },
+    {
+      "epoch": 0.3318092749836708,
+      "grad_norm": 3.5407757854087034,
+      "learning_rate": 2.2836683132522927e-05,
+      "loss": 0.7017,
+      "step": 5080
+    },
+    {
+      "epoch": 0.3324624428478119,
+      "grad_norm": 8.776372349033894,
+      "learning_rate": 2.2810154536782903e-05,
+      "loss": 0.6874,
+      "step": 5090
+    },
+    {
+      "epoch": 0.33311561071195295,
+      "grad_norm": 14.30802960940852,
+      "learning_rate": 2.2783592383510038e-05,
+      "loss": 0.7418,
+      "step": 5100
+    },
+    {
+      "epoch": 0.33376877857609405,
+      "grad_norm": 15.268248599076554,
+      "learning_rate": 2.275699678683272e-05,
+      "loss": 0.7243,
+      "step": 5110
+    },
+    {
+      "epoch": 0.33442194644023515,
+      "grad_norm": 3.8522854449552493,
+      "learning_rate": 2.2730367861023023e-05,
+      "loss": 0.6963,
+      "step": 5120
+    },
+    {
+      "epoch": 0.3350751143043762,
+      "grad_norm": 15.097323735018765,
+      "learning_rate": 2.2703705720496235e-05,
+      "loss": 0.7404,
+      "step": 5130
+    },
+    {
+      "epoch": 0.3357282821685173,
+      "grad_norm": 3.289494356993917,
+      "learning_rate": 2.2677010479810362e-05,
+      "loss": 0.7096,
+      "step": 5140
+    },
+    {
+      "epoch": 0.3363814500326584,
+      "grad_norm": 3.3373795905520107,
+      "learning_rate": 2.2650282253665605e-05,
+      "loss": 0.7068,
+      "step": 5150
+    },
+    {
+      "epoch": 0.3370346178967995,
+      "grad_norm": 22.639667962001727,
+      "learning_rate": 2.2623521156903914e-05,
+      "loss": 0.6999,
+      "step": 5160
+    },
+    {
+      "epoch": 0.33768778576094055,
+      "grad_norm": 8.066217556055694,
+      "learning_rate": 2.2596727304508474e-05,
+      "loss": 0.7185,
+      "step": 5170
+    },
+    {
+      "epoch": 0.33834095362508165,
+      "grad_norm": 3.5470860624323395,
+      "learning_rate": 2.256990081160319e-05,
+      "loss": 0.6573,
+      "step": 5180
+    },
+    {
+      "epoch": 0.33899412148922276,
+      "grad_norm": 2.833517423894922,
+      "learning_rate": 2.2543041793452228e-05,
+      "loss": 0.6982,
+      "step": 5190
+    },
+    {
+      "epoch": 0.3396472893533638,
+      "grad_norm": 5.640485006270031,
+      "learning_rate": 2.2516150365459507e-05,
+      "loss": 0.7177,
+      "step": 5200
+    },
+    {
+      "epoch": 0.3403004572175049,
+      "grad_norm": 3.839458519984891,
+      "learning_rate": 2.2489226643168183e-05,
+      "loss": 0.7174,
+      "step": 5210
+    },
+    {
+      "epoch": 0.340953625081646,
+      "grad_norm": 6.786572033585219,
+      "learning_rate": 2.246227074226018e-05,
+      "loss": 0.7393,
+      "step": 5220
+    },
+    {
+      "epoch": 0.34160679294578705,
+      "grad_norm": 8.830342296217026,
+      "learning_rate": 2.243528277855568e-05,
+      "loss": 0.6876,
+      "step": 5230
+    },
+    {
+      "epoch": 0.34225996080992815,
+      "grad_norm": 12.293778646461922,
+      "learning_rate": 2.2408262868012635e-05,
+      "loss": 0.7381,
+      "step": 5240
+    },
+    {
+      "epoch": 0.34291312867406926,
+      "grad_norm": 6.77122748377005,
+      "learning_rate": 2.2381211126726255e-05,
+      "loss": 0.7216,
+      "step": 5250
+    },
+    {
+      "epoch": 0.3435662965382103,
+      "grad_norm": 3.839666787432626,
+      "learning_rate": 2.2354127670928513e-05,
+      "loss": 0.6704,
+      "step": 5260
+    },
+    {
+      "epoch": 0.3442194644023514,
+      "grad_norm": 3.5517692240588605,
+      "learning_rate": 2.2327012616987646e-05,
+      "loss": 0.6805,
+      "step": 5270
+    },
+    {
+      "epoch": 0.3448726322664925,
+      "grad_norm": 5.315718450066826,
+      "learning_rate": 2.2299866081407676e-05,
+      "loss": 0.729,
+      "step": 5280
+    },
+    {
+      "epoch": 0.34552580013063355,
+      "grad_norm": 6.848787865977037,
+      "learning_rate": 2.227268818082787e-05,
+      "loss": 0.6739,
+      "step": 5290
+    },
+    {
+      "epoch": 0.34617896799477466,
+      "grad_norm": 3.6917748968535564,
+      "learning_rate": 2.2245479032022272e-05,
+      "loss": 0.7269,
+      "step": 5300
+    },
+    {
+      "epoch": 0.34683213585891576,
+      "grad_norm": 9.160335319411757,
+      "learning_rate": 2.2218238751899174e-05,
+      "loss": 0.6839,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3474853037230568,
+      "grad_norm": 5.424577814365955,
+      "learning_rate": 2.2190967457500646e-05,
+      "loss": 0.6939,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3481384715871979,
+      "grad_norm": 3.6648874879145015,
+      "learning_rate": 2.2163665266002007e-05,
+      "loss": 0.7073,
+      "step": 5330
+    },
+    {
+      "epoch": 0.348791639451339,
+      "grad_norm": 9.13369002747884,
+      "learning_rate": 2.213633229471133e-05,
+      "loss": 0.7327,
+      "step": 5340
+    },
+    {
+      "epoch": 0.34944480731548005,
+      "grad_norm": 2.881754363096986,
+      "learning_rate": 2.210896866106894e-05,
+      "loss": 0.7052,
+      "step": 5350
+    },
+    {
+      "epoch": 0.35009797517962116,
+      "grad_norm": 5.529284801624932,
+      "learning_rate": 2.2081574482646903e-05,
+      "loss": 0.7253,
+      "step": 5360
+    },
+    {
+      "epoch": 0.35075114304376226,
+      "grad_norm": 2.9786955632123235,
+      "learning_rate": 2.205414987714854e-05,
+      "loss": 0.6812,
+      "step": 5370
+    },
+    {
+      "epoch": 0.3514043109079033,
+      "grad_norm": 3.993821378974443,
+      "learning_rate": 2.202669496240788e-05,
+      "loss": 0.6951,
+      "step": 5380
+    },
+    {
+      "epoch": 0.3520574787720444,
+      "grad_norm": 10.693892502406564,
+      "learning_rate": 2.1999209856389215e-05,
+      "loss": 0.6938,
+      "step": 5390
+    },
+    {
+      "epoch": 0.3527106466361855,
+      "grad_norm": 10.783647265180912,
+      "learning_rate": 2.1971694677186523e-05,
+      "loss": 0.7188,
+      "step": 5400
+    },
+    {
+      "epoch": 0.35336381450032656,
+      "grad_norm": 5.596700220664436,
+      "learning_rate": 2.194414954302302e-05,
+      "loss": 0.7373,
+      "step": 5410
+    },
+    {
+      "epoch": 0.35401698236446766,
+      "grad_norm": 10.407517181316326,
+      "learning_rate": 2.191657457225062e-05,
+      "loss": 0.7397,
+      "step": 5420
+    },
+    {
+      "epoch": 0.35467015022860876,
+      "grad_norm": 5.081907583105732,
+      "learning_rate": 2.1888969883349436e-05,
+      "loss": 0.6908,
+      "step": 5430
+    },
+    {
+      "epoch": 0.35532331809274986,
+      "grad_norm": 30.697198457905934,
+      "learning_rate": 2.1861335594927264e-05,
+      "loss": 0.7265,
+      "step": 5440
+    },
+    {
+      "epoch": 0.3559764859568909,
+      "grad_norm": 3.973571491869874,
+      "learning_rate": 2.1833671825719092e-05,
+      "loss": 0.7273,
+      "step": 5450
+    },
+    {
+      "epoch": 0.356629653821032,
+      "grad_norm": 3.6562494783319606,
+      "learning_rate": 2.1805978694586564e-05,
+      "loss": 0.742,
+      "step": 5460
+    },
+    {
+      "epoch": 0.3572828216851731,
+      "grad_norm": 17.645169020048908,
+      "learning_rate": 2.1778256320517485e-05,
+      "loss": 0.7066,
+      "step": 5470
+    },
+    {
+      "epoch": 0.35793598954931416,
+      "grad_norm": 4.218794080715886,
+      "learning_rate": 2.1750504822625316e-05,
+      "loss": 0.6779,
+      "step": 5480
+    },
+    {
+      "epoch": 0.35858915741345526,
+      "grad_norm": 6.070136540242526,
+      "learning_rate": 2.172272432014864e-05,
+      "loss": 0.6884,
+      "step": 5490
+    },
+    {
+      "epoch": 0.35924232527759636,
+      "grad_norm": 4.296131638468927,
+      "learning_rate": 2.169491493245066e-05,
+      "loss": 0.7549,
+      "step": 5500
+    },
+    {
+      "epoch": 0.3598954931417374,
+      "grad_norm": 4.45693239441751,
+      "learning_rate": 2.1667076779018708e-05,
+      "loss": 0.7081,
+      "step": 5510
+    },
+    {
+      "epoch": 0.3605486610058785,
+      "grad_norm": 6.126532226736765,
+      "learning_rate": 2.16392099794637e-05,
+      "loss": 0.7233,
+      "step": 5520
+    },
+    {
+      "epoch": 0.3612018288700196,
+      "grad_norm": 10.690768628798947,
+      "learning_rate": 2.1611314653519633e-05,
+      "loss": 0.7494,
+      "step": 5530
+    },
+    {
+      "epoch": 0.36185499673416066,
+      "grad_norm": 6.005580532144763,
+      "learning_rate": 2.1583390921043074e-05,
+      "loss": 0.7013,
+      "step": 5540
+    },
+    {
+      "epoch": 0.36250816459830176,
+      "grad_norm": 4.400593474851796,
+      "learning_rate": 2.1555438902012644e-05,
+      "loss": 0.6549,
+      "step": 5550
+    },
+    {
+      "epoch": 0.36316133246244287,
+      "grad_norm": 2.4175705589561884,
+      "learning_rate": 2.152745871652851e-05,
+      "loss": 0.7459,
+      "step": 5560
+    },
+    {
+      "epoch": 0.3638145003265839,
+      "grad_norm": 8.380975214402772,
+      "learning_rate": 2.1499450484811836e-05,
+      "loss": 0.7475,
+      "step": 5570
+    },
+    {
+      "epoch": 0.364467668190725,
+      "grad_norm": 7.535104105735855,
+      "learning_rate": 2.1471414327204325e-05,
+      "loss": 0.7463,
+      "step": 5580
+    },
+    {
+      "epoch": 0.3651208360548661,
+      "grad_norm": 10.183550297183858,
+      "learning_rate": 2.1443350364167635e-05,
+      "loss": 0.7225,
+      "step": 5590
+    },
+    {
+      "epoch": 0.36577400391900716,
+      "grad_norm": 6.961617279255035,
+      "learning_rate": 2.1415258716282912e-05,
+      "loss": 0.7151,
+      "step": 5600
+    },
+    {
+      "epoch": 0.36642717178314826,
+      "grad_norm": 1.940684132435257,
+      "learning_rate": 2.1387139504250254e-05,
+      "loss": 0.7195,
+      "step": 5610
+    },
+    {
+      "epoch": 0.36708033964728937,
+      "grad_norm": 3.744684393791611,
+      "learning_rate": 2.135899284888819e-05,
+      "loss": 0.6486,
+      "step": 5620
+    },
+    {
+      "epoch": 0.3677335075114304,
+      "grad_norm": 6.069719872334578,
+      "learning_rate": 2.1330818871133164e-05,
+      "loss": 0.7101,
+      "step": 5630
+    },
+    {
+      "epoch": 0.3683866753755715,
+      "grad_norm": 12.369927493771815,
+      "learning_rate": 2.130261769203901e-05,
+      "loss": 0.7299,
+      "step": 5640
+    },
+    {
+      "epoch": 0.3690398432397126,
+      "grad_norm": 5.184241316936024,
+      "learning_rate": 2.1274389432776442e-05,
+      "loss": 0.6976,
+      "step": 5650
+    },
+    {
+      "epoch": 0.36969301110385366,
+      "grad_norm": 5.144146765126566,
+      "learning_rate": 2.124613421463253e-05,
+      "loss": 0.7232,
+      "step": 5660
+    },
+    {
+      "epoch": 0.37034617896799477,
+      "grad_norm": 4.746113040084255,
+      "learning_rate": 2.121785215901018e-05,
+      "loss": 0.7095,
+      "step": 5670
+    },
+    {
+      "epoch": 0.37099934683213587,
+      "grad_norm": 5.32268400735139,
+      "learning_rate": 2.118954338742759e-05,
+      "loss": 0.7106,
+      "step": 5680
+    },
+    {
+      "epoch": 0.37165251469627697,
+      "grad_norm": 7.237241370763546,
+      "learning_rate": 2.1161208021517766e-05,
+      "loss": 0.7085,
+      "step": 5690
+    },
+    {
+      "epoch": 0.372305682560418,
+      "grad_norm": 4.077001658098951,
+      "learning_rate": 2.1132846183027978e-05,
+      "loss": 0.7134,
+      "step": 5700
+    },
+    {
+      "epoch": 0.3729588504245591,
+      "grad_norm": 5.488216261292269,
+      "learning_rate": 2.1104457993819237e-05,
+      "loss": 0.7269,
+      "step": 5710
+    },
+    {
+      "epoch": 0.3736120182887002,
+      "grad_norm": 17.11430526139107,
+      "learning_rate": 2.1076043575865768e-05,
+      "loss": 0.6564,
+      "step": 5720
+    },
+    {
+      "epoch": 0.37426518615284127,
+      "grad_norm": 10.97778991135696,
+      "learning_rate": 2.10476030512545e-05,
+      "loss": 0.7019,
+      "step": 5730
+    },
+    {
+      "epoch": 0.37491835401698237,
+      "grad_norm": 8.044336485211486,
+      "learning_rate": 2.1019136542184534e-05,
+      "loss": 0.6942,
+      "step": 5740
+    },
+    {
+      "epoch": 0.37557152188112347,
+      "grad_norm": 18.021968575001992,
+      "learning_rate": 2.099064417096662e-05,
+      "loss": 0.7102,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3762246897452645,
+      "grad_norm": 7.597040915393705,
+      "learning_rate": 2.0962126060022603e-05,
+      "loss": 0.726,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3768778576094056,
+      "grad_norm": 4.866355469739272,
+      "learning_rate": 2.0933582331884967e-05,
+      "loss": 0.6813,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3775310254735467,
+      "grad_norm": 9.713816871869412,
+      "learning_rate": 2.0905013109196217e-05,
+      "loss": 0.6868,
+      "step": 5780
+    },
+    {
+      "epoch": 0.37818419333768777,
+      "grad_norm": 10.125773082465265,
+      "learning_rate": 2.0876418514708442e-05,
+      "loss": 0.7506,
+      "step": 5790
+    },
+    {
+      "epoch": 0.37883736120182887,
+      "grad_norm": 6.982779164809807,
+      "learning_rate": 2.0847798671282706e-05,
+      "loss": 0.6706,
+      "step": 5800
+    },
+    {
+      "epoch": 0.37949052906597,
+      "grad_norm": 3.4051762559620142,
+      "learning_rate": 2.081915370188859e-05,
+      "loss": 0.7132,
+      "step": 5810
+    },
+    {
+      "epoch": 0.380143696930111,
+      "grad_norm": 3.4718152498585506,
+      "learning_rate": 2.0790483729603624e-05,
+      "loss": 0.6742,
+      "step": 5820
+    },
+    {
+      "epoch": 0.3807968647942521,
+      "grad_norm": 22.035510447982027,
+      "learning_rate": 2.0761788877612746e-05,
+      "loss": 0.7255,
+      "step": 5830
+    },
+    {
+      "epoch": 0.3814500326583932,
+      "grad_norm": 2.0183516401641395,
+      "learning_rate": 2.0733069269207828e-05,
+      "loss": 0.7506,
+      "step": 5840
+    },
+    {
+      "epoch": 0.38210320052253427,
+      "grad_norm": 40.64552816683451,
+      "learning_rate": 2.0704325027787085e-05,
+      "loss": 0.7109,
+      "step": 5850
+    },
+    {
+      "epoch": 0.38275636838667537,
+      "grad_norm": 11.592439068405213,
+      "learning_rate": 2.0675556276854588e-05,
+      "loss": 0.6948,
+      "step": 5860
+    },
+    {
+      "epoch": 0.3834095362508165,
+      "grad_norm": 8.94221169646604,
+      "learning_rate": 2.0646763140019702e-05,
+      "loss": 0.7036,
+      "step": 5870
+    },
+    {
+      "epoch": 0.3840627041149575,
+      "grad_norm": 3.4153432996898463,
+      "learning_rate": 2.0617945740996583e-05,
+      "loss": 0.7016,
+      "step": 5880
+    },
+    {
+      "epoch": 0.3847158719790986,
+      "grad_norm": 4.83458645240433,
+      "learning_rate": 2.0589104203603624e-05,
+      "loss": 0.6892,
+      "step": 5890
+    },
+    {
+      "epoch": 0.3853690398432397,
+      "grad_norm": 2.992026860080857,
+      "learning_rate": 2.056023865176294e-05,
+      "loss": 0.7367,
+      "step": 5900
+    },
+    {
+      "epoch": 0.38602220770738077,
+      "grad_norm": 7.321261903148404,
+      "learning_rate": 2.0531349209499822e-05,
+      "loss": 0.7121,
+      "step": 5910
+    },
+    {
+      "epoch": 0.3866753755715219,
+      "grad_norm": 4.135261026651754,
+      "learning_rate": 2.0502436000942206e-05,
+      "loss": 0.7015,
+      "step": 5920
+    },
+    {
+      "epoch": 0.387328543435663,
+      "grad_norm": 5.739710682346368,
+      "learning_rate": 2.047349915032016e-05,
+      "loss": 0.6811,
+      "step": 5930
+    },
+    {
+      "epoch": 0.3879817112998041,
+      "grad_norm": 7.817442528367328,
+      "learning_rate": 2.0444538781965324e-05,
+      "loss": 0.7079,
+      "step": 5940
+    },
+    {
+      "epoch": 0.3886348791639451,
+      "grad_norm": 1.7125273681458524,
+      "learning_rate": 2.041555502031037e-05,
+      "loss": 0.6939,
+      "step": 5950
+    },
+    {
+      "epoch": 0.3892880470280862,
+      "grad_norm": 7.063769471170878,
+      "learning_rate": 2.0386547989888514e-05,
+      "loss": 0.7409,
+      "step": 5960
+    },
+    {
+      "epoch": 0.38994121489222733,
+      "grad_norm": 6.649069768334491,
+      "learning_rate": 2.0357517815332918e-05,
+      "loss": 0.7039,
+      "step": 5970
+    },
+    {
+      "epoch": 0.3905943827563684,
+      "grad_norm": 4.9949670256327225,
+      "learning_rate": 2.0328464621376216e-05,
+      "loss": 0.708,
+      "step": 5980
+    },
+    {
+      "epoch": 0.3912475506205095,
+      "grad_norm": 10.223302544529933,
+      "learning_rate": 2.0299388532849922e-05,
+      "loss": 0.7625,
+      "step": 5990
+    },
+    {
+      "epoch": 0.3919007184846506,
+      "grad_norm": 3.4792253561647155,
+      "learning_rate": 2.027028967468394e-05,
+      "loss": 0.7313,
+      "step": 6000
+    },
+    {
+      "epoch": 0.3925538863487916,
+      "grad_norm": 6.173284673862917,
+      "learning_rate": 2.0241168171906002e-05,
+      "loss": 0.6961,
+      "step": 6010
+    },
+    {
+      "epoch": 0.3932070542129327,
+      "grad_norm": 9.069327472622437,
+      "learning_rate": 2.0212024149641124e-05,
+      "loss": 0.6839,
+      "step": 6020
+    },
+    {
+      "epoch": 0.39386022207707383,
+      "grad_norm": 4.354312267637299,
+      "learning_rate": 2.0182857733111094e-05,
+      "loss": 0.7299,
+      "step": 6030
+    },
+    {
+      "epoch": 0.3945133899412149,
+      "grad_norm": 10.938547313138859,
+      "learning_rate": 2.015366904763392e-05,
+      "loss": 0.737,
+      "step": 6040
+    },
+    {
+      "epoch": 0.395166557805356,
+      "grad_norm": 3.368581390407729,
+      "learning_rate": 2.012445821862329e-05,
+      "loss": 0.6936,
+      "step": 6050
+    },
+    {
+      "epoch": 0.3958197256694971,
+      "grad_norm": 22.651820687722875,
+      "learning_rate": 2.0095225371588023e-05,
+      "loss": 0.6701,
+      "step": 6060
+    },
+    {
+      "epoch": 0.3964728935336381,
+      "grad_norm": 7.087662693961875,
+      "learning_rate": 2.006597063213156e-05,
+      "loss": 0.7057,
+      "step": 6070
+    },
+    {
+      "epoch": 0.39712606139777923,
+      "grad_norm": 3.6259318502326736,
+      "learning_rate": 2.0036694125951395e-05,
+      "loss": 0.747,
+      "step": 6080
+    },
+    {
+      "epoch": 0.39777922926192033,
+      "grad_norm": 2.4208395436437615,
+      "learning_rate": 2.0007395978838556e-05,
+      "loss": 0.7042,
+      "step": 6090
+    },
+    {
+      "epoch": 0.3984323971260614,
+      "grad_norm": 3.7718790757505682,
+      "learning_rate": 1.9978076316677035e-05,
+      "loss": 0.7039,
+      "step": 6100
+    },
+    {
+      "epoch": 0.3990855649902025,
+      "grad_norm": 4.715238565697955,
+      "learning_rate": 1.9948735265443297e-05,
+      "loss": 0.7006,
+      "step": 6110
+    },
+    {
+      "epoch": 0.3997387328543436,
+      "grad_norm": 3.656750114993287,
+      "learning_rate": 1.9919372951205675e-05,
+      "loss": 0.6848,
+      "step": 6120
+    },
+    {
+      "epoch": 0.40039190071848463,
+      "grad_norm": 6.2809029098948335,
+      "learning_rate": 1.9889989500123896e-05,
+      "loss": 0.7516,
+      "step": 6130
+    },
+    {
+      "epoch": 0.40104506858262573,
+      "grad_norm": 4.45791400589284,
+      "learning_rate": 1.9860585038448472e-05,
+      "loss": 0.7293,
+      "step": 6140
+    },
+    {
+      "epoch": 0.40169823644676683,
+      "grad_norm": 6.671689461980347,
+      "learning_rate": 1.9831159692520208e-05,
+      "loss": 0.7139,
+      "step": 6150
+    },
+    {
+      "epoch": 0.4023514043109079,
+      "grad_norm": 7.413217349665978,
+      "learning_rate": 1.9801713588769643e-05,
+      "loss": 0.6818,
+      "step": 6160
+    },
+    {
+      "epoch": 0.403004572175049,
+      "grad_norm": 3.0801533633605467,
+      "learning_rate": 1.9772246853716497e-05,
+      "loss": 0.6924,
+      "step": 6170
+    },
+    {
+      "epoch": 0.4036577400391901,
+      "grad_norm": 8.663708499869044,
+      "learning_rate": 1.9742759613969136e-05,
+      "loss": 0.7205,
+      "step": 6180
+    },
+    {
+      "epoch": 0.40431090790333113,
+      "grad_norm": 6.16162556292061,
+      "learning_rate": 1.9713251996224037e-05,
+      "loss": 0.7135,
+      "step": 6190
+    },
+    {
+      "epoch": 0.40496407576747223,
+      "grad_norm": 4.790487901849723,
+      "learning_rate": 1.9683724127265228e-05,
+      "loss": 0.7015,
+      "step": 6200
+    },
+    {
+      "epoch": 0.40561724363161333,
+      "grad_norm": 3.6618688723110346,
+      "learning_rate": 1.965417613396375e-05,
+      "loss": 0.6915,
+      "step": 6210
+    },
+    {
+      "epoch": 0.40627041149575444,
+      "grad_norm": 16.05473312037314,
+      "learning_rate": 1.96246081432771e-05,
+      "loss": 0.7518,
+      "step": 6220
+    },
+    {
+      "epoch": 0.4069235793598955,
+      "grad_norm": 4.88766672771986,
+      "learning_rate": 1.959502028224872e-05,
+      "loss": 0.7345,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4075767472240366,
+      "grad_norm": 12.426776499556185,
+      "learning_rate": 1.9565412678007414e-05,
+      "loss": 0.7051,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4082299150881777,
+      "grad_norm": 17.259950283504946,
+      "learning_rate": 1.9535785457766816e-05,
+      "loss": 0.6707,
+      "step": 6250
+    },
+    {
+      "epoch": 0.40888308295231873,
+      "grad_norm": 3.9776106316833317,
+      "learning_rate": 1.950613874882484e-05,
+      "loss": 0.6904,
+      "step": 6260
+    },
+    {
+      "epoch": 0.40953625081645983,
+      "grad_norm": 4.209309505641302,
+      "learning_rate": 1.947647267856314e-05,
+      "loss": 0.7336,
+      "step": 6270
+    },
+    {
+      "epoch": 0.41018941868060094,
+      "grad_norm": 4.510058288197555,
+      "learning_rate": 1.9446787374446574e-05,
+      "loss": 0.7202,
+      "step": 6280
+    },
+    {
+      "epoch": 0.410842586544742,
+      "grad_norm": 6.194243484377249,
+      "learning_rate": 1.9417082964022605e-05,
+      "loss": 0.6833,
+      "step": 6290
+    },
+    {
+      "epoch": 0.4114957544088831,
+      "grad_norm": 3.311152634015392,
+      "learning_rate": 1.938735957492083e-05,
+      "loss": 0.6833,
+      "step": 6300
+    },
+    {
+      "epoch": 0.4121489222730242,
+      "grad_norm": 57.47319772756037,
+      "learning_rate": 1.935761733485236e-05,
+      "loss": 0.7215,
+      "step": 6310
+    },
+    {
+      "epoch": 0.41280209013716523,
+      "grad_norm": 6.884796376261111,
+      "learning_rate": 1.9327856371609327e-05,
+      "loss": 0.7029,
+      "step": 6320
+    },
+    {
+      "epoch": 0.41345525800130634,
+      "grad_norm": 4.169035556055065,
+      "learning_rate": 1.9298076813064282e-05,
+      "loss": 0.7448,
+      "step": 6330
+    },
+    {
+      "epoch": 0.41410842586544744,
+      "grad_norm": 5.96368320968989,
+      "learning_rate": 1.9268278787169696e-05,
+      "loss": 0.7163,
+      "step": 6340
+    },
+    {
+      "epoch": 0.4147615937295885,
+      "grad_norm": 4.222858152405998,
+      "learning_rate": 1.923846242195738e-05,
+      "loss": 0.7119,
+      "step": 6350
+    },
+    {
+      "epoch": 0.4154147615937296,
+      "grad_norm": 11.408875045751458,
+      "learning_rate": 1.9208627845537946e-05,
+      "loss": 0.7135,
+      "step": 6360
+    },
+    {
+      "epoch": 0.4160679294578707,
+      "grad_norm": 6.302076439551584,
+      "learning_rate": 1.9178775186100245e-05,
+      "loss": 0.6941,
+      "step": 6370
+    },
+    {
+      "epoch": 0.41672109732201174,
+      "grad_norm": 7.347215719366371,
+      "learning_rate": 1.914890457191083e-05,
+      "loss": 0.7243,
+      "step": 6380
+    },
+    {
+      "epoch": 0.41737426518615284,
+      "grad_norm": 4.592029067528999,
+      "learning_rate": 1.91190161313134e-05,
+      "loss": 0.6766,
+      "step": 6390
+    },
+    {
+      "epoch": 0.41802743305029394,
+      "grad_norm": 5.2659124188411335,
+      "learning_rate": 1.9089109992728253e-05,
+      "loss": 0.6972,
+      "step": 6400
+    },
+    {
+      "epoch": 0.418680600914435,
+      "grad_norm": 3.8701089324208575,
+      "learning_rate": 1.9059186284651714e-05,
+      "loss": 0.6983,
+      "step": 6410
+    },
+    {
+      "epoch": 0.4193337687785761,
+      "grad_norm": 9.765641406236218,
+      "learning_rate": 1.902924513565561e-05,
+      "loss": 0.713,
+      "step": 6420
+    },
+    {
+      "epoch": 0.4199869366427172,
+      "grad_norm": 47.84550535643343,
+      "learning_rate": 1.8999286674386712e-05,
+      "loss": 0.7405,
+      "step": 6430
+    },
+    {
+      "epoch": 0.42064010450685824,
+      "grad_norm": 4.640806868871516,
+      "learning_rate": 1.8969311029566158e-05,
+      "loss": 0.7386,
+      "step": 6440
+    },
+    {
+      "epoch": 0.42129327237099934,
+      "grad_norm": 4.9456925811064645,
+      "learning_rate": 1.8939318329988924e-05,
+      "loss": 0.6878,
+      "step": 6450
+    },
+    {
+      "epoch": 0.42194644023514044,
+      "grad_norm": 3.7569417916267196,
+      "learning_rate": 1.890930870452327e-05,
+      "loss": 0.719,
+      "step": 6460
+    },
+    {
+      "epoch": 0.42259960809928154,
+      "grad_norm": 4.046929672048653,
+      "learning_rate": 1.8879282282110183e-05,
+      "loss": 0.6635,
+      "step": 6470
+    },
+    {
+      "epoch": 0.4232527759634226,
+      "grad_norm": 3.241109572317114,
+      "learning_rate": 1.8849239191762807e-05,
+      "loss": 0.7271,
+      "step": 6480
+    },
+    {
+      "epoch": 0.4239059438275637,
+      "grad_norm": 3.5379414582873,
+      "learning_rate": 1.881917956256591e-05,
+      "loss": 0.7063,
+      "step": 6490
+    },
+    {
+      "epoch": 0.4245591116917048,
+      "grad_norm": 5.3498102814022355,
+      "learning_rate": 1.878910352367533e-05,
+      "loss": 0.666,
+      "step": 6500
+    },
+    {
+      "epoch": 0.42521227955584584,
+      "grad_norm": 3.062766083037805,
+      "learning_rate": 1.8759011204317403e-05,
+      "loss": 0.6783,
+      "step": 6510
+    },
+    {
+      "epoch": 0.42586544741998694,
+      "grad_norm": 11.9116282093037,
+      "learning_rate": 1.872890273378841e-05,
+      "loss": 0.6585,
+      "step": 6520
+    },
+    {
+      "epoch": 0.42651861528412804,
+      "grad_norm": 18.417492786135245,
+      "learning_rate": 1.8698778241454048e-05,
+      "loss": 0.6759,
+      "step": 6530
+    },
+    {
+      "epoch": 0.4271717831482691,
+      "grad_norm": 8.22022213823648,
+      "learning_rate": 1.8668637856748826e-05,
+      "loss": 0.7064,
+      "step": 6540
+    },
+    {
+      "epoch": 0.4278249510124102,
+      "grad_norm": 9.406437361869447,
+      "learning_rate": 1.8638481709175566e-05,
+      "loss": 0.7451,
+      "step": 6550
+    },
+    {
+      "epoch": 0.4284781188765513,
+      "grad_norm": 24.665543539235813,
+      "learning_rate": 1.8608309928304797e-05,
+      "loss": 0.7097,
+      "step": 6560
+    },
+    {
+      "epoch": 0.42913128674069234,
+      "grad_norm": 6.918551299318292,
+      "learning_rate": 1.857812264377423e-05,
+      "loss": 0.7246,
+      "step": 6570
+    },
+    {
+      "epoch": 0.42978445460483344,
+      "grad_norm": 4.661685101269544,
+      "learning_rate": 1.8547919985288183e-05,
+      "loss": 0.7204,
+      "step": 6580
+    },
+    {
+      "epoch": 0.43043762246897455,
+      "grad_norm": 6.459041072664668,
+      "learning_rate": 1.851770208261704e-05,
+      "loss": 0.6513,
+      "step": 6590
+    },
+    {
+      "epoch": 0.4310907903331156,
+      "grad_norm": 15.875963076716992,
+      "learning_rate": 1.8487469065596668e-05,
+      "loss": 0.7152,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4317439581972567,
+      "grad_norm": 5.891032642456025,
+      "learning_rate": 1.845722106412789e-05,
+      "loss": 0.6918,
+      "step": 6610
+    },
+    {
+      "epoch": 0.4323971260613978,
+      "grad_norm": 5.232261663805651,
+      "learning_rate": 1.842695820817591e-05,
+      "loss": 0.6852,
+      "step": 6620
+    },
+    {
+      "epoch": 0.43305029392553884,
+      "grad_norm": 12.336035217709867,
+      "learning_rate": 1.8396680627769753e-05,
+      "loss": 0.7119,
+      "step": 6630
+    },
+    {
+      "epoch": 0.43370346178967994,
+      "grad_norm": 11.538495193823614,
+      "learning_rate": 1.8366388453001702e-05,
+      "loss": 0.7276,
+      "step": 6640
+    },
+    {
+      "epoch": 0.43435662965382105,
+      "grad_norm": 2.5758268343611985,
+      "learning_rate": 1.833608181402676e-05,
+      "loss": 0.6689,
+      "step": 6650
+    },
+    {
+      "epoch": 0.4350097975179621,
+      "grad_norm": 4.745025640163145,
+      "learning_rate": 1.830576084106208e-05,
+      "loss": 0.7138,
+      "step": 6660
+    },
+    {
+      "epoch": 0.4356629653821032,
+      "grad_norm": 5.142215007475423,
+      "learning_rate": 1.8275425664386385e-05,
+      "loss": 0.6803,
+      "step": 6670
+    },
+    {
+      "epoch": 0.4363161332462443,
+      "grad_norm": 5.252139790751175,
+      "learning_rate": 1.8245076414339438e-05,
+      "loss": 0.6771,
+      "step": 6680
+    },
+    {
+      "epoch": 0.43696930111038534,
+      "grad_norm": 5.820963942150845,
+      "learning_rate": 1.821471322132148e-05,
+      "loss": 0.7338,
+      "step": 6690
+    },
+    {
+      "epoch": 0.43762246897452645,
+      "grad_norm": 13.496280864296235,
+      "learning_rate": 1.8184336215792644e-05,
+      "loss": 0.6952,
+      "step": 6700
+    },
+    {
+      "epoch": 0.43827563683866755,
+      "grad_norm": 9.016756353678497,
+      "learning_rate": 1.8153945528272415e-05,
+      "loss": 0.7139,
+      "step": 6710
+    },
+    {
+      "epoch": 0.4389288047028086,
+      "grad_norm": 4.822974955108365,
+      "learning_rate": 1.8123541289339068e-05,
+      "loss": 0.7025,
+      "step": 6720
+    },
+    {
+      "epoch": 0.4395819725669497,
+      "grad_norm": 4.348630290498651,
+      "learning_rate": 1.8093123629629105e-05,
+      "loss": 0.7191,
+      "step": 6730
+    },
+    {
+      "epoch": 0.4402351404310908,
+      "grad_norm": 4.015291548959369,
+      "learning_rate": 1.8062692679836684e-05,
+      "loss": 0.707,
+      "step": 6740
+    },
+    {
+      "epoch": 0.4408883082952319,
+      "grad_norm": 5.595847306715769,
+      "learning_rate": 1.803224857071307e-05,
+      "loss": 0.708,
+      "step": 6750
+    },
+    {
+      "epoch": 0.44154147615937295,
+      "grad_norm": 6.389126553007975,
+      "learning_rate": 1.8001791433066082e-05,
+      "loss": 0.6708,
+      "step": 6760
+    },
+    {
+      "epoch": 0.44219464402351405,
+      "grad_norm": 8.602448545134093,
+      "learning_rate": 1.7971321397759495e-05,
+      "loss": 0.7064,
+      "step": 6770
+    },
+    {
+      "epoch": 0.44284781188765515,
+      "grad_norm": 9.192310968011101,
+      "learning_rate": 1.7940838595712522e-05,
+      "loss": 0.6585,
+      "step": 6780
+    },
+    {
+      "epoch": 0.4435009797517962,
+      "grad_norm": 6.114239953281097,
+      "learning_rate": 1.791034315789921e-05,
+      "loss": 0.7215,
+      "step": 6790
+    },
+    {
+      "epoch": 0.4441541476159373,
+      "grad_norm": 5.817934510545314,
+      "learning_rate": 1.7879835215347915e-05,
+      "loss": 0.7057,
+      "step": 6800
+    },
+    {
+      "epoch": 0.4448073154800784,
+      "grad_norm": 5.461675111129483,
+      "learning_rate": 1.784931489914072e-05,
+      "loss": 0.6925,
+      "step": 6810
+    },
+    {
+      "epoch": 0.44546048334421945,
+      "grad_norm": 13.13059369734387,
+      "learning_rate": 1.781878234041286e-05,
+      "loss": 0.6884,
+      "step": 6820
+    },
+    {
+      "epoch": 0.44611365120836055,
+      "grad_norm": 7.136465651727822,
+      "learning_rate": 1.7788237670352176e-05,
+      "loss": 0.737,
+      "step": 6830
+    },
+    {
+      "epoch": 0.44676681907250165,
+      "grad_norm": 13.671072521091867,
+      "learning_rate": 1.7757681020198557e-05,
+      "loss": 0.7404,
+      "step": 6840
+    },
+    {
+      "epoch": 0.4474199869366427,
+      "grad_norm": 5.397882237012338,
+      "learning_rate": 1.7727112521243362e-05,
+      "loss": 0.7256,
+      "step": 6850
+    },
+    {
+      "epoch": 0.4480731548007838,
+      "grad_norm": 6.12908954723118,
+      "learning_rate": 1.769653230482886e-05,
+      "loss": 0.6802,
+      "step": 6860
+    },
+    {
+      "epoch": 0.4487263226649249,
+      "grad_norm": 3.6504708286902066,
+      "learning_rate": 1.7665940502347654e-05,
+      "loss": 0.7045,
+      "step": 6870
+    },
+    {
+      "epoch": 0.44937949052906595,
+      "grad_norm": 58.966599142676735,
+      "learning_rate": 1.763533724524215e-05,
+      "loss": 0.741,
+      "step": 6880
+    },
+    {
+      "epoch": 0.45003265839320705,
+      "grad_norm": 3.9881784274514915,
+      "learning_rate": 1.760472266500396e-05,
+      "loss": 0.7162,
+      "step": 6890
+    },
+    {
+      "epoch": 0.45068582625734815,
+      "grad_norm": 4.002812827071041,
+      "learning_rate": 1.7574096893173336e-05,
+      "loss": 0.724,
+      "step": 6900
+    },
+    {
+      "epoch": 0.4513389941214892,
+      "grad_norm": 4.939553714933158,
+      "learning_rate": 1.7543460061338636e-05,
+      "loss": 0.6901,
+      "step": 6910
+    },
+    {
+      "epoch": 0.4519921619856303,
+      "grad_norm": 5.030454395011258,
+      "learning_rate": 1.7512812301135726e-05,
+      "loss": 0.7254,
+      "step": 6920
+    },
+    {
+      "epoch": 0.4526453298497714,
+      "grad_norm": 22.27855846953453,
+      "learning_rate": 1.748215374424744e-05,
+      "loss": 0.7051,
+      "step": 6930
+    },
+    {
+      "epoch": 0.45329849771391245,
+      "grad_norm": 7.470444746348905,
+      "learning_rate": 1.7451484522402983e-05,
+      "loss": 0.7057,
+      "step": 6940
+    },
+    {
+      "epoch": 0.45395166557805355,
+      "grad_norm": 11.682170774889052,
+      "learning_rate": 1.7420804767377398e-05,
+      "loss": 0.7149,
+      "step": 6950
+    },
+    {
+      "epoch": 0.45460483344219466,
+      "grad_norm": 5.335499780442737,
+      "learning_rate": 1.739011461099098e-05,
+      "loss": 0.7296,
+      "step": 6960
+    },
+    {
+      "epoch": 0.4552580013063357,
+      "grad_norm": 4.741493237122223,
+      "learning_rate": 1.7359414185108727e-05,
+      "loss": 0.6507,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4559111691704768,
+      "grad_norm": 3.041945945425016,
+      "learning_rate": 1.7328703621639737e-05,
+      "loss": 0.7245,
+      "step": 6980
+    },
+    {
+      "epoch": 0.4565643370346179,
+      "grad_norm": 4.6202574462500365,
+      "learning_rate": 1.7297983052536683e-05,
+      "loss": 0.6651,
+      "step": 6990
+    },
+    {
+      "epoch": 0.457217504898759,
+      "grad_norm": 11.336858163854028,
+      "learning_rate": 1.7267252609795236e-05,
+      "loss": 0.7212,
+      "step": 7000
+    },
+    {
+      "epoch": 0.45787067276290006,
+      "grad_norm": 6.600581819725192,
+      "learning_rate": 1.723651242545347e-05,
+      "loss": 0.7195,
+      "step": 7010
+    },
+    {
+      "epoch": 0.45852384062704116,
+      "grad_norm": 7.634299938630818,
+      "learning_rate": 1.7205762631591323e-05,
+      "loss": 0.6876,
+      "step": 7020
+    },
+    {
+      "epoch": 0.45917700849118226,
+      "grad_norm": 5.545937214410351,
+      "learning_rate": 1.7175003360330027e-05,
+      "loss": 0.7441,
+      "step": 7030
+    },
+    {
+      "epoch": 0.4598301763553233,
+      "grad_norm": 2.453391382612166,
+      "learning_rate": 1.7144234743831538e-05,
+      "loss": 0.7199,
+      "step": 7040
+    },
+    {
+      "epoch": 0.4604833442194644,
+      "grad_norm": 5.482371080990207,
+      "learning_rate": 1.7113456914297956e-05,
+      "loss": 0.7049,
+      "step": 7050
+    },
+    {
+      "epoch": 0.4611365120836055,
+      "grad_norm": 5.677867067700385,
+      "learning_rate": 1.7082670003970968e-05,
+      "loss": 0.7171,
+      "step": 7060
+    },
+    {
+      "epoch": 0.46178967994774656,
+      "grad_norm": 6.625917766987666,
+      "learning_rate": 1.7051874145131276e-05,
+      "loss": 0.7625,
+      "step": 7070
+    },
+    {
+      "epoch": 0.46244284781188766,
+      "grad_norm": 3.9787150554458712,
+      "learning_rate": 1.7021069470098048e-05,
+      "loss": 0.6928,
+      "step": 7080
+    },
+    {
+      "epoch": 0.46309601567602876,
+      "grad_norm": 6.607647609209365,
+      "learning_rate": 1.6990256111228306e-05,
+      "loss": 0.7183,
+      "step": 7090
+    },
+    {
+      "epoch": 0.4637491835401698,
+      "grad_norm": 26.05901856805948,
+      "learning_rate": 1.69594342009164e-05,
+      "loss": 0.7199,
+      "step": 7100
+    },
+    {
+      "epoch": 0.4644023514043109,
+      "grad_norm": 2.609610541214694,
+      "learning_rate": 1.6928603871593417e-05,
+      "loss": 0.6709,
+      "step": 7110
+    },
+    {
+      "epoch": 0.465055519268452,
+      "grad_norm": 3.437628310149699,
+      "learning_rate": 1.6897765255726626e-05,
+      "loss": 0.6821,
+      "step": 7120
+    },
+    {
+      "epoch": 0.46570868713259306,
+      "grad_norm": 3.694418672043116,
+      "learning_rate": 1.6866918485818883e-05,
+      "loss": 0.7009,
+      "step": 7130
+    },
+    {
+      "epoch": 0.46636185499673416,
+      "grad_norm": 5.064209072648378,
+      "learning_rate": 1.6836063694408095e-05,
+      "loss": 0.7264,
+      "step": 7140
+    },
+    {
+      "epoch": 0.46701502286087526,
+      "grad_norm": 4.097679523587385,
+      "learning_rate": 1.680520101406663e-05,
+      "loss": 0.7179,
+      "step": 7150
+    },
+    {
+      "epoch": 0.4676681907250163,
+      "grad_norm": 6.233472103811337,
+      "learning_rate": 1.6774330577400752e-05,
+      "loss": 0.6983,
+      "step": 7160
+    },
+    {
+      "epoch": 0.4683213585891574,
+      "grad_norm": 5.792189357233483,
+      "learning_rate": 1.6743452517050048e-05,
+      "loss": 0.7063,
+      "step": 7170
+    },
+    {
+      "epoch": 0.4689745264532985,
+      "grad_norm": 7.292501637926463,
+      "learning_rate": 1.6712566965686864e-05,
+      "loss": 0.7,
+      "step": 7180
+    },
+    {
+      "epoch": 0.46962769431743956,
+      "grad_norm": 10.071292424078878,
+      "learning_rate": 1.6681674056015738e-05,
+      "loss": 0.7445,
+      "step": 7190
+    },
+    {
+      "epoch": 0.47028086218158066,
+      "grad_norm": 3.1886962441070827,
+      "learning_rate": 1.6650773920772813e-05,
+      "loss": 0.726,
+      "step": 7200
+    },
+    {
+      "epoch": 0.47093403004572176,
+      "grad_norm": 4.3591848899379215,
+      "learning_rate": 1.661986669272528e-05,
+      "loss": 0.7052,
+      "step": 7210
+    },
+    {
+      "epoch": 0.4715871979098628,
+      "grad_norm": 6.972903805566907,
+      "learning_rate": 1.658895250467081e-05,
+      "loss": 0.7293,
+      "step": 7220
+    },
+    {
+      "epoch": 0.4722403657740039,
+      "grad_norm": 4.495721902724383,
+      "learning_rate": 1.6558031489436987e-05,
+      "loss": 0.6583,
+      "step": 7230
+    },
+    {
+      "epoch": 0.472893533638145,
+      "grad_norm": 3.427324388987157,
+      "learning_rate": 1.652710377988071e-05,
+      "loss": 0.6369,
+      "step": 7240
+    },
+    {
+      "epoch": 0.47354670150228606,
+      "grad_norm": 2.2380193054416533,
+      "learning_rate": 1.6496169508887645e-05,
+      "loss": 0.6559,
+      "step": 7250
+    },
+    {
+      "epoch": 0.47419986936642716,
+      "grad_norm": 5.534199246287345,
+      "learning_rate": 1.6465228809371666e-05,
+      "loss": 0.7328,
+      "step": 7260
+    },
+    {
+      "epoch": 0.47485303723056826,
+      "grad_norm": 5.194906990459634,
+      "learning_rate": 1.6434281814274257e-05,
+      "loss": 0.7506,
+      "step": 7270
+    },
+    {
+      "epoch": 0.47550620509470937,
+      "grad_norm": 25.079647296086396,
+      "learning_rate": 1.6403328656563948e-05,
+      "loss": 0.6865,
+      "step": 7280
+    },
+    {
+      "epoch": 0.4761593729588504,
+      "grad_norm": 29.382249562541475,
+      "learning_rate": 1.6372369469235756e-05,
+      "loss": 0.705,
+      "step": 7290
+    },
+    {
+      "epoch": 0.4768125408229915,
+      "grad_norm": 6.934988306059051,
+      "learning_rate": 1.6341404385310592e-05,
+      "loss": 0.7332,
+      "step": 7300
+    },
+    {
+      "epoch": 0.4774657086871326,
+      "grad_norm": 5.0690173359655555,
+      "learning_rate": 1.631043353783473e-05,
+      "loss": 0.7018,
+      "step": 7310
+    },
+    {
+      "epoch": 0.47811887655127366,
+      "grad_norm": 6.086800862922415,
+      "learning_rate": 1.6279457059879173e-05,
+      "loss": 0.7089,
+      "step": 7320
+    },
+    {
+      "epoch": 0.47877204441541477,
+      "grad_norm": 3.6627542400974775,
+      "learning_rate": 1.6248475084539137e-05,
+      "loss": 0.7287,
+      "step": 7330
+    },
+    {
+      "epoch": 0.47942521227955587,
+      "grad_norm": 9.57059828973796,
+      "learning_rate": 1.6217487744933466e-05,
+      "loss": 0.7164,
+      "step": 7340
+    },
+    {
+      "epoch": 0.4800783801436969,
+      "grad_norm": 5.553447257648302,
+      "learning_rate": 1.618649517420403e-05,
+      "loss": 0.7384,
+      "step": 7350
+    },
+    {
+      "epoch": 0.480731548007838,
+      "grad_norm": 17.982629659825115,
+      "learning_rate": 1.615549750551519e-05,
+      "loss": 0.7176,
+      "step": 7360
+    },
+    {
+      "epoch": 0.4813847158719791,
+      "grad_norm": 5.054362597628958,
+      "learning_rate": 1.6124494872053204e-05,
+      "loss": 0.6956,
+      "step": 7370
+    },
+    {
+      "epoch": 0.48203788373612017,
+      "grad_norm": 5.3668664117133345,
+      "learning_rate": 1.609348740702567e-05,
+      "loss": 0.6896,
+      "step": 7380
+    },
+    {
+      "epoch": 0.48269105160026127,
+      "grad_norm": 6.833581067068372,
+      "learning_rate": 1.6062475243660942e-05,
+      "loss": 0.6546,
+      "step": 7390
+    },
+    {
+      "epoch": 0.48334421946440237,
+      "grad_norm": 8.426135738805158,
+      "learning_rate": 1.6031458515207552e-05,
+      "loss": 0.7207,
+      "step": 7400
+    },
+    {
+      "epoch": 0.4839973873285434,
+      "grad_norm": 6.445906991995997,
+      "learning_rate": 1.6000437354933664e-05,
+      "loss": 0.6955,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4846505551926845,
+      "grad_norm": 9.185210273507879,
+      "learning_rate": 1.5969411896126465e-05,
+      "loss": 0.7361,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4853037230568256,
+      "grad_norm": 9.628498580999864,
+      "learning_rate": 1.5938382272091635e-05,
+      "loss": 0.6671,
+      "step": 7430
+    },
+    {
+      "epoch": 0.48595689092096667,
+      "grad_norm": 18.780973765647442,
+      "learning_rate": 1.5907348616152722e-05,
+      "loss": 0.6964,
+      "step": 7440
+    },
+    {
+      "epoch": 0.48661005878510777,
+      "grad_norm": 3.4109938598953313,
+      "learning_rate": 1.5876311061650625e-05,
+      "loss": 0.7258,
+      "step": 7450
+    },
+    {
+      "epoch": 0.48726322664924887,
+      "grad_norm": 5.148159268491969,
+      "learning_rate": 1.584526974194297e-05,
+      "loss": 0.7413,
+      "step": 7460
+    },
+    {
+      "epoch": 0.4879163945133899,
+      "grad_norm": 5.825994107139637,
+      "learning_rate": 1.581422479040358e-05,
+      "loss": 0.6995,
+      "step": 7470
+    },
+    {
+      "epoch": 0.488569562377531,
+      "grad_norm": 8.15661433199093,
+      "learning_rate": 1.5783176340421877e-05,
+      "loss": 0.737,
+      "step": 7480
+    },
+    {
+      "epoch": 0.4892227302416721,
+      "grad_norm": 2.488291310719702,
+      "learning_rate": 1.57521245254023e-05,
+      "loss": 0.6931,
+      "step": 7490
+    },
+    {
+      "epoch": 0.48987589810581317,
+      "grad_norm": 4.250185794135119,
+      "learning_rate": 1.572106947876377e-05,
+      "loss": 0.6854,
+      "step": 7500
+    },
+    {
+      "epoch": 0.49052906596995427,
+      "grad_norm": 3.8635990640808355,
+      "learning_rate": 1.5690011333939074e-05,
+      "loss": 0.7205,
+      "step": 7510
+    },
+    {
+      "epoch": 0.49118223383409537,
+      "grad_norm": 9.085150254077263,
+      "learning_rate": 1.565895022437432e-05,
+      "loss": 0.7455,
+      "step": 7520
+    },
+    {
+      "epoch": 0.4918354016982365,
+      "grad_norm": 4.516782829977407,
+      "learning_rate": 1.562788628352836e-05,
+      "loss": 0.6965,
+      "step": 7530
+    },
+    {
+      "epoch": 0.4924885695623775,
+      "grad_norm": 4.2329680297711265,
+      "learning_rate": 1.5596819644872195e-05,
+      "loss": 0.7156,
+      "step": 7540
+    },
+    {
+      "epoch": 0.4931417374265186,
+      "grad_norm": 13.379779282646771,
+      "learning_rate": 1.556575044188843e-05,
+      "loss": 0.688,
+      "step": 7550
+    },
+    {
+      "epoch": 0.4937949052906597,
+      "grad_norm": 10.879140832861319,
+      "learning_rate": 1.553467880807069e-05,
+      "loss": 0.6622,
+      "step": 7560
+    },
+    {
+      "epoch": 0.49444807315480077,
+      "grad_norm": 4.213562665524711,
+      "learning_rate": 1.5503604876923035e-05,
+      "loss": 0.7061,
+      "step": 7570
+    },
+    {
+      "epoch": 0.4951012410189419,
+      "grad_norm": 3.7923437638813424,
+      "learning_rate": 1.5472528781959402e-05,
+      "loss": 0.7046,
+      "step": 7580
+    },
+    {
+      "epoch": 0.495754408883083,
+      "grad_norm": 4.996725536064961,
+      "learning_rate": 1.5441450656703012e-05,
+      "loss": 0.6894,
+      "step": 7590
+    },
+    {
+      "epoch": 0.496407576747224,
+      "grad_norm": 13.749534460336342,
+      "learning_rate": 1.5410370634685835e-05,
+      "loss": 0.7193,
+      "step": 7600
+    },
+    {
+      "epoch": 0.4970607446113651,
+      "grad_norm": 14.370136532914689,
+      "learning_rate": 1.5379288849447964e-05,
+      "loss": 0.7396,
+      "step": 7610
+    },
+    {
+      "epoch": 0.4977139124755062,
+      "grad_norm": 2.9206834291164174,
+      "learning_rate": 1.5348205434537098e-05,
+      "loss": 0.6699,
+      "step": 7620
+    },
+    {
+      "epoch": 0.4983670803396473,
+      "grad_norm": 22.590699893533344,
+      "learning_rate": 1.5317120523507904e-05,
+      "loss": 0.706,
+      "step": 7630
+    },
+    {
+      "epoch": 0.4990202482037884,
+      "grad_norm": 39.07888094030937,
+      "learning_rate": 1.5286034249921495e-05,
+      "loss": 0.7308,
+      "step": 7640
+    },
+    {
+      "epoch": 0.4996734160679295,
+      "grad_norm": 6.180342416505246,
+      "learning_rate": 1.5254946747344843e-05,
+      "loss": 0.7015,
+      "step": 7650
+    },
+    {
+      "epoch": 0.5003265839320705,
+      "grad_norm": 3.506114659272752,
+      "learning_rate": 1.52238581493502e-05,
+      "loss": 0.6883,
+      "step": 7660
+    },
+    {
+      "epoch": 0.5009797517962117,
+      "grad_norm": 3.506978434952958,
+      "learning_rate": 1.5192768589514508e-05,
+      "loss": 0.7279,
+      "step": 7670
+    },
+    {
+      "epoch": 0.5016329196603527,
+      "grad_norm": 6.328901379843207,
+      "learning_rate": 1.5161678201418857e-05,
+      "loss": 0.7585,
+      "step": 7680
+    },
+    {
+      "epoch": 0.5022860875244938,
+      "grad_norm": 3.59766393737303,
+      "learning_rate": 1.5130587118647891e-05,
+      "loss": 0.6758,
+      "step": 7690
+    },
+    {
+      "epoch": 0.5029392553886349,
+      "grad_norm": 3.098792850125791,
+      "learning_rate": 1.5099495474789243e-05,
+      "loss": 0.7318,
+      "step": 7700
+    },
+    {
+      "epoch": 0.503592423252776,
+      "grad_norm": 4.313613747783113,
+      "learning_rate": 1.5068403403432948e-05,
+      "loss": 0.6997,
+      "step": 7710
+    },
+    {
+      "epoch": 0.504245591116917,
+      "grad_norm": 4.960621068279318,
+      "learning_rate": 1.5037311038170888e-05,
+      "loss": 0.6933,
+      "step": 7720
+    },
+    {
+      "epoch": 0.5048987589810582,
+      "grad_norm": 4.50244603965065,
+      "learning_rate": 1.5006218512596204e-05,
+      "loss": 0.7122,
+      "step": 7730
+    },
+    {
+      "epoch": 0.5055519268451992,
+      "grad_norm": 5.402529456090525,
+      "learning_rate": 1.4975125960302718e-05,
+      "loss": 0.7356,
+      "step": 7740
+    },
+    {
+      "epoch": 0.5062050947093403,
+      "grad_norm": 5.251731768589659,
+      "learning_rate": 1.4944033514884378e-05,
+      "loss": 0.7091,
+      "step": 7750
+    },
+    {
+      "epoch": 0.5068582625734814,
+      "grad_norm": 4.527582618132294,
+      "learning_rate": 1.4912941309934673e-05,
+      "loss": 0.7349,
+      "step": 7760
+    },
+    {
+      "epoch": 0.5075114304376225,
+      "grad_norm": 5.063886664473719,
+      "learning_rate": 1.4881849479046042e-05,
+      "loss": 0.7231,
+      "step": 7770
+    },
+    {
+      "epoch": 0.5081645983017635,
+      "grad_norm": 7.340580162798091,
+      "learning_rate": 1.485075815580934e-05,
+      "loss": 0.6708,
+      "step": 7780
+    },
+    {
+      "epoch": 0.5088177661659047,
+      "grad_norm": 8.279851370689158,
+      "learning_rate": 1.481966747381323e-05,
+      "loss": 0.6817,
+      "step": 7790
+    },
+    {
+      "epoch": 0.5094709340300457,
+      "grad_norm": 4.007759673902586,
+      "learning_rate": 1.4788577566643612e-05,
+      "loss": 0.7361,
+      "step": 7800
+    },
+    {
+      "epoch": 0.5101241018941868,
+      "grad_norm": 3.607676977732233,
+      "learning_rate": 1.4757488567883066e-05,
+      "loss": 0.6834,
+      "step": 7810
+    },
+    {
+      "epoch": 0.5107772697583279,
+      "grad_norm": 6.825186579463042,
+      "learning_rate": 1.472640061111027e-05,
+      "loss": 0.6966,
+      "step": 7820
+    },
+    {
+      "epoch": 0.511430437622469,
+      "grad_norm": 4.199100198181181,
+      "learning_rate": 1.4695313829899421e-05,
+      "loss": 0.7256,
+      "step": 7830
+    },
+    {
+      "epoch": 0.51208360548661,
+      "grad_norm": 3.8617562426661523,
+      "learning_rate": 1.4664228357819667e-05,
+      "loss": 0.7319,
+      "step": 7840
+    },
+    {
+      "epoch": 0.5127367733507512,
+      "grad_norm": 59.94700214694851,
+      "learning_rate": 1.4633144328434534e-05,
+      "loss": 0.7075,
+      "step": 7850
+    },
+    {
+      "epoch": 0.5133899412148922,
+      "grad_norm": 4.66758709867641,
+      "learning_rate": 1.4602061875301339e-05,
+      "loss": 0.6796,
+      "step": 7860
+    },
+    {
+      "epoch": 0.5140431090790333,
+      "grad_norm": 9.719456098118954,
+      "learning_rate": 1.4570981131970636e-05,
+      "loss": 0.7226,
+      "step": 7870
+    },
+    {
+      "epoch": 0.5146962769431744,
+      "grad_norm": 5.463175870167361,
+      "learning_rate": 1.4539902231985631e-05,
+      "loss": 0.6783,
+      "step": 7880
+    },
+    {
+      "epoch": 0.5153494448073155,
+      "grad_norm": 5.382574555075144,
+      "learning_rate": 1.4508825308881605e-05,
+      "loss": 0.6674,
+      "step": 7890
+    },
+    {
+      "epoch": 0.5160026126714565,
+      "grad_norm": 4.868251531153773,
+      "learning_rate": 1.4477750496185348e-05,
+      "loss": 0.7024,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5166557805355977,
+      "grad_norm": 3.7294213547214325,
+      "learning_rate": 1.4446677927414587e-05,
+      "loss": 0.7643,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5173089483997387,
+      "grad_norm": 3.3941905879029948,
+      "learning_rate": 1.44156077360774e-05,
+      "loss": 0.6935,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5179621162638798,
+      "grad_norm": 11.955624228102595,
+      "learning_rate": 1.4384540055671652e-05,
+      "loss": 0.6563,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5186152841280209,
+      "grad_norm": 4.223751917576418,
+      "learning_rate": 1.4353475019684431e-05,
+      "loss": 0.709,
+      "step": 7940
+    },
+    {
+      "epoch": 0.519268451992162,
+      "grad_norm": 3.743270303247067,
+      "learning_rate": 1.4322412761591441e-05,
+      "loss": 0.6992,
+      "step": 7950
+    },
+    {
+      "epoch": 0.519921619856303,
+      "grad_norm": 5.432489476270528,
+      "learning_rate": 1.4291353414856466e-05,
+      "loss": 0.6729,
+      "step": 7960
+    },
+    {
+      "epoch": 0.5205747877204442,
+      "grad_norm": 4.5886135319405374,
+      "learning_rate": 1.4260297112930774e-05,
+      "loss": 0.654,
+      "step": 7970
+    },
+    {
+      "epoch": 0.5212279555845852,
+      "grad_norm": 3.5771128459787525,
+      "learning_rate": 1.4229243989252554e-05,
+      "loss": 0.6779,
+      "step": 7980
+    },
+    {
+      "epoch": 0.5218811234487263,
+      "grad_norm": 7.088599397946011,
+      "learning_rate": 1.4198194177246343e-05,
+      "loss": 0.7128,
+      "step": 7990
+    },
+    {
+      "epoch": 0.5225342913128674,
+      "grad_norm": 10.356939935810402,
+      "learning_rate": 1.4167147810322438e-05,
+      "loss": 0.718,
+      "step": 8000
+    },
+    {
+      "epoch": 0.5231874591770085,
+      "grad_norm": 7.230293589180331,
+      "learning_rate": 1.4136105021876346e-05,
+      "loss": 0.6918,
+      "step": 8010
+    },
+    {
+      "epoch": 0.5238406270411495,
+      "grad_norm": 3.9909166676063683,
+      "learning_rate": 1.4105065945288196e-05,
+      "loss": 0.7007,
+      "step": 8020
+    },
+    {
+      "epoch": 0.5244937949052907,
+      "grad_norm": 5.794003497753727,
+      "learning_rate": 1.4074030713922151e-05,
+      "loss": 0.6887,
+      "step": 8030
+    },
+    {
+      "epoch": 0.5251469627694317,
+      "grad_norm": 2.9783249490709203,
+      "learning_rate": 1.4042999461125876e-05,
+      "loss": 0.7596,
+      "step": 8040
+    },
+    {
+      "epoch": 0.5258001306335728,
+      "grad_norm": 26.692386939122034,
+      "learning_rate": 1.4011972320229934e-05,
+      "loss": 0.6974,
+      "step": 8050
+    },
+    {
+      "epoch": 0.5264532984977139,
+      "grad_norm": 2.6819712524637076,
+      "learning_rate": 1.398094942454721e-05,
+      "loss": 0.6884,
+      "step": 8060
+    },
+    {
+      "epoch": 0.527106466361855,
+      "grad_norm": 4.990492364294924,
+      "learning_rate": 1.3949930907372363e-05,
+      "loss": 0.6969,
+      "step": 8070
+    },
+    {
+      "epoch": 0.527759634225996,
+      "grad_norm": 5.327689323623585,
+      "learning_rate": 1.3918916901981234e-05,
+      "loss": 0.7193,
+      "step": 8080
+    },
+    {
+      "epoch": 0.5284128020901372,
+      "grad_norm": 2.7492206509522057,
+      "learning_rate": 1.3887907541630272e-05,
+      "loss": 0.7205,
+      "step": 8090
+    },
+    {
+      "epoch": 0.5290659699542782,
+      "grad_norm": 2.8357777616212565,
+      "learning_rate": 1.3856902959555987e-05,
+      "loss": 0.7318,
+      "step": 8100
+    },
+    {
+      "epoch": 0.5297191378184193,
+      "grad_norm": 3.9368304939861085,
+      "learning_rate": 1.3825903288974329e-05,
+      "loss": 0.6837,
+      "step": 8110
+    },
+    {
+      "epoch": 0.5303723056825604,
+      "grad_norm": 3.977284424009092,
+      "learning_rate": 1.3794908663080165e-05,
+      "loss": 0.6738,
+      "step": 8120
+    },
+    {
+      "epoch": 0.5310254735467015,
+      "grad_norm": 8.306483794883848,
+      "learning_rate": 1.376391921504669e-05,
+      "loss": 0.6866,
+      "step": 8130
+    },
+    {
+      "epoch": 0.5316786414108425,
+      "grad_norm": 14.970355431587016,
+      "learning_rate": 1.3732935078024839e-05,
+      "loss": 0.688,
+      "step": 8140
+    },
+    {
+      "epoch": 0.5323318092749837,
+      "grad_norm": 3.481786339363183,
+      "learning_rate": 1.3701956385142732e-05,
+      "loss": 0.6571,
+      "step": 8150
+    },
+    {
+      "epoch": 0.5329849771391247,
+      "grad_norm": 4.091950698182643,
+      "learning_rate": 1.3670983269505098e-05,
+      "loss": 0.6786,
+      "step": 8160
+    },
+    {
+      "epoch": 0.5336381450032658,
+      "grad_norm": 4.133504020078944,
+      "learning_rate": 1.3640015864192709e-05,
+      "loss": 0.7041,
+      "step": 8170
+    },
+    {
+      "epoch": 0.5342913128674069,
+      "grad_norm": 4.756759050876008,
+      "learning_rate": 1.3609054302261787e-05,
+      "loss": 0.718,
+      "step": 8180
+    },
+    {
+      "epoch": 0.534944480731548,
+      "grad_norm": 4.278232151247267,
+      "learning_rate": 1.3578098716743457e-05,
+      "loss": 0.7106,
+      "step": 8190
+    },
+    {
+      "epoch": 0.5355976485956891,
+      "grad_norm": 2.0249684555045726,
+      "learning_rate": 1.3547149240643165e-05,
+      "loss": 0.6737,
+      "step": 8200
+    },
+    {
+      "epoch": 0.5362508164598302,
+      "grad_norm": 5.223443576125692,
+      "learning_rate": 1.3516206006940108e-05,
+      "loss": 0.6646,
+      "step": 8210
+    },
+    {
+      "epoch": 0.5369039843239712,
+      "grad_norm": 4.667809596201585,
+      "learning_rate": 1.3485269148586655e-05,
+      "loss": 0.7258,
+      "step": 8220
+    },
+    {
+      "epoch": 0.5375571521881124,
+      "grad_norm": 7.584805216491663,
+      "learning_rate": 1.3454338798507793e-05,
+      "loss": 0.7059,
+      "step": 8230
+    },
+    {
+      "epoch": 0.5382103200522534,
+      "grad_norm": 3.369443911847391,
+      "learning_rate": 1.3423415089600531e-05,
+      "loss": 0.718,
+      "step": 8240
+    },
+    {
+      "epoch": 0.5388634879163945,
+      "grad_norm": 2.1903022256230056,
+      "learning_rate": 1.3392498154733359e-05,
+      "loss": 0.6787,
+      "step": 8250
+    },
+    {
+      "epoch": 0.5395166557805356,
+      "grad_norm": 16.6815682487673,
+      "learning_rate": 1.3361588126745646e-05,
+      "loss": 0.7269,
+      "step": 8260
+    },
+    {
+      "epoch": 0.5401698236446767,
+      "grad_norm": 3.702406157567106,
+      "learning_rate": 1.3330685138447095e-05,
+      "loss": 0.7176,
+      "step": 8270
+    },
+    {
+      "epoch": 0.5408229915088177,
+      "grad_norm": 3.4301661546239473,
+      "learning_rate": 1.3299789322617156e-05,
+      "loss": 0.6611,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5414761593729589,
+      "grad_norm": 4.73145186322431,
+      "learning_rate": 1.3268900812004468e-05,
+      "loss": 0.7399,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5421293272371,
+      "grad_norm": 6.065859463546294,
+      "learning_rate": 1.3238019739326275e-05,
+      "loss": 0.7051,
+      "step": 8300
+    },
+    {
+      "epoch": 0.542782495101241,
+      "grad_norm": 3.5219919830092383,
+      "learning_rate": 1.3207146237267866e-05,
+      "loss": 0.7082,
+      "step": 8310
+    },
+    {
+      "epoch": 0.5434356629653821,
+      "grad_norm": 5.153036144779057,
+      "learning_rate": 1.3176280438482007e-05,
+      "loss": 0.696,
+      "step": 8320
+    },
+    {
+      "epoch": 0.5440888308295232,
+      "grad_norm": 18.004351619708324,
+      "learning_rate": 1.3145422475588357e-05,
+      "loss": 0.7294,
+      "step": 8330
+    },
+    {
+      "epoch": 0.5447419986936642,
+      "grad_norm": 3.241110896411055,
+      "learning_rate": 1.3114572481172905e-05,
+      "loss": 0.6854,
+      "step": 8340
+    },
+    {
+      "epoch": 0.5453951665578054,
+      "grad_norm": 29.532297848473245,
+      "learning_rate": 1.3083730587787416e-05,
+      "loss": 0.7101,
+      "step": 8350
+    },
+    {
+      "epoch": 0.5460483344219464,
+      "grad_norm": 7.991969846743042,
+      "learning_rate": 1.305289692794883e-05,
+      "loss": 0.6968,
+      "step": 8360
+    },
+    {
+      "epoch": 0.5467015022860875,
+      "grad_norm": 3.3552311379053354,
+      "learning_rate": 1.3022071634138723e-05,
+      "loss": 0.6995,
+      "step": 8370
+    },
+    {
+      "epoch": 0.5473546701502287,
+      "grad_norm": 3.2979728714825955,
+      "learning_rate": 1.2991254838802722e-05,
+      "loss": 0.7522,
+      "step": 8380
+    },
+    {
+      "epoch": 0.5480078380143697,
+      "grad_norm": 14.98352824852725,
+      "learning_rate": 1.2960446674349939e-05,
+      "loss": 0.7072,
+      "step": 8390
+    },
+    {
+      "epoch": 0.5486610058785107,
+      "grad_norm": 8.870482678617899,
+      "learning_rate": 1.2929647273152407e-05,
+      "loss": 0.7086,
+      "step": 8400
+    },
+    {
+      "epoch": 0.5493141737426519,
+      "grad_norm": 13.259488145434055,
+      "learning_rate": 1.2898856767544486e-05,
+      "loss": 0.6932,
+      "step": 8410
+    },
+    {
+      "epoch": 0.549967341606793,
+      "grad_norm": 6.786398062995833,
+      "learning_rate": 1.286807528982234e-05,
+      "loss": 0.6994,
+      "step": 8420
+    },
+    {
+      "epoch": 0.550620509470934,
+      "grad_norm": 21.918977139408003,
+      "learning_rate": 1.2837302972243331e-05,
+      "loss": 0.6888,
+      "step": 8430
+    },
+    {
+      "epoch": 0.5512736773350752,
+      "grad_norm": 5.569225574398577,
+      "learning_rate": 1.2806539947025465e-05,
+      "loss": 0.7531,
+      "step": 8440
+    },
+    {
+      "epoch": 0.5519268451992162,
+      "grad_norm": 7.651003912583063,
+      "learning_rate": 1.277578634634682e-05,
+      "loss": 0.6703,
+      "step": 8450
+    },
+    {
+      "epoch": 0.5525800130633572,
+      "grad_norm": 3.8123580405899546,
+      "learning_rate": 1.274504230234498e-05,
+      "loss": 0.7247,
+      "step": 8460
+    },
+    {
+      "epoch": 0.5532331809274984,
+      "grad_norm": 4.623355057456077,
+      "learning_rate": 1.2714307947116473e-05,
+      "loss": 0.7145,
+      "step": 8470
+    },
+    {
+      "epoch": 0.5538863487916394,
+      "grad_norm": 2.574327944116888,
+      "learning_rate": 1.2683583412716202e-05,
+      "loss": 0.6898,
+      "step": 8480
+    },
+    {
+      "epoch": 0.5545395166557805,
+      "grad_norm": 3.2663389456748404,
+      "learning_rate": 1.2652868831156846e-05,
+      "loss": 0.7128,
+      "step": 8490
+    },
+    {
+      "epoch": 0.5551926845199217,
+      "grad_norm": 4.498572229008819,
+      "learning_rate": 1.262216433440835e-05,
+      "loss": 0.6927,
+      "step": 8500
+    },
+    {
+      "epoch": 0.5558458523840627,
+      "grad_norm": 3.2539273521224508,
+      "learning_rate": 1.2591470054397316e-05,
+      "loss": 0.7144,
+      "step": 8510
+    },
+    {
+      "epoch": 0.5564990202482037,
+      "grad_norm": 6.184850260248766,
+      "learning_rate": 1.256078612300645e-05,
+      "loss": 0.7184,
+      "step": 8520
+    },
+    {
+      "epoch": 0.5571521881123449,
+      "grad_norm": 4.851817445306759,
+      "learning_rate": 1.253011267207399e-05,
+      "loss": 0.7139,
+      "step": 8530
+    },
+    {
+      "epoch": 0.557805355976486,
+      "grad_norm": 4.124875615874441,
+      "learning_rate": 1.2499449833393147e-05,
+      "loss": 0.7228,
+      "step": 8540
+    },
+    {
+      "epoch": 0.558458523840627,
+      "grad_norm": 16.611404251919918,
+      "learning_rate": 1.2468797738711543e-05,
+      "loss": 0.7002,
+      "step": 8550
+    },
+    {
+      "epoch": 0.5591116917047682,
+      "grad_norm": 5.330949767268735,
+      "learning_rate": 1.2438156519730613e-05,
+      "loss": 0.6904,
+      "step": 8560
+    },
+    {
+      "epoch": 0.5597648595689092,
+      "grad_norm": 3.6213033835624406,
+      "learning_rate": 1.240752630810508e-05,
+      "loss": 0.7664,
+      "step": 8570
+    },
+    {
+      "epoch": 0.5604180274330502,
+      "grad_norm": 5.470156417925566,
+      "learning_rate": 1.2376907235442377e-05,
+      "loss": 0.6961,
+      "step": 8580
+    },
+    {
+      "epoch": 0.5610711952971914,
+      "grad_norm": 4.024325315426769,
+      "learning_rate": 1.2346299433302067e-05,
+      "loss": 0.7538,
+      "step": 8590
+    },
+    {
+      "epoch": 0.5617243631613325,
+      "grad_norm": 5.837701451748462,
+      "learning_rate": 1.2315703033195285e-05,
+      "loss": 0.7085,
+      "step": 8600
+    },
+    {
+      "epoch": 0.5623775310254735,
+      "grad_norm": 4.8934203133055,
+      "learning_rate": 1.228511816658419e-05,
+      "loss": 0.6848,
+      "step": 8610
+    },
+    {
+      "epoch": 0.5630306988896147,
+      "grad_norm": 3.7706313667515503,
+      "learning_rate": 1.2254544964881364e-05,
+      "loss": 0.7374,
+      "step": 8620
+    },
+    {
+      "epoch": 0.5636838667537557,
+      "grad_norm": 4.437415968408255,
+      "learning_rate": 1.2223983559449292e-05,
+      "loss": 0.6693,
+      "step": 8630
+    },
+    {
+      "epoch": 0.5643370346178967,
+      "grad_norm": 10.73476128251022,
+      "learning_rate": 1.2193434081599758e-05,
+      "loss": 0.7481,
+      "step": 8640
+    },
+    {
+      "epoch": 0.5649902024820379,
+      "grad_norm": 4.852001422186861,
+      "learning_rate": 1.2162896662593297e-05,
+      "loss": 0.6515,
+      "step": 8650
+    },
+    {
+      "epoch": 0.565643370346179,
+      "grad_norm": 5.824228247078944,
+      "learning_rate": 1.2132371433638643e-05,
+      "loss": 0.6588,
+      "step": 8660
+    },
+    {
+      "epoch": 0.56629653821032,
+      "grad_norm": 5.839446905935213,
+      "learning_rate": 1.2101858525892147e-05,
+      "loss": 0.7222,
+      "step": 8670
+    },
+    {
+      "epoch": 0.5669497060744612,
+      "grad_norm": 3.12927883944411,
+      "learning_rate": 1.207135807045722e-05,
+      "loss": 0.676,
+      "step": 8680
+    },
+    {
+      "epoch": 0.5676028739386022,
+      "grad_norm": 4.3550286177561,
+      "learning_rate": 1.204087019838377e-05,
+      "loss": 0.7436,
+      "step": 8690
+    },
+    {
+      "epoch": 0.5682560418027433,
+      "grad_norm": 4.882541008077315,
+      "learning_rate": 1.2010395040667642e-05,
+      "loss": 0.664,
+      "step": 8700
+    },
+    {
+      "epoch": 0.5689092096668844,
+      "grad_norm": 5.572176970355917,
+      "learning_rate": 1.1979932728250045e-05,
+      "loss": 0.7028,
+      "step": 8710
+    },
+    {
+      "epoch": 0.5695623775310255,
+      "grad_norm": 9.180272587737047,
+      "learning_rate": 1.1949483392016997e-05,
+      "loss": 0.7212,
+      "step": 8720
+    },
+    {
+      "epoch": 0.5702155453951666,
+      "grad_norm": 7.546332108773005,
+      "learning_rate": 1.1919047162798773e-05,
+      "loss": 0.6982,
+      "step": 8730
+    },
+    {
+      "epoch": 0.5708687132593077,
+      "grad_norm": 10.472934899896995,
+      "learning_rate": 1.1888624171369315e-05,
+      "loss": 0.6695,
+      "step": 8740
+    },
+    {
+      "epoch": 0.5715218811234487,
+      "grad_norm": 6.384248494945576,
+      "learning_rate": 1.1858214548445698e-05,
+      "loss": 0.6871,
+      "step": 8750
+    },
+    {
+      "epoch": 0.5721750489875899,
+      "grad_norm": 2.402182869835395,
+      "learning_rate": 1.1827818424687554e-05,
+      "loss": 0.7061,
+      "step": 8760
+    },
+    {
+      "epoch": 0.5728282168517309,
+      "grad_norm": 6.648042445206821,
+      "learning_rate": 1.1797435930696518e-05,
+      "loss": 0.7003,
+      "step": 8770
+    },
+    {
+      "epoch": 0.573481384715872,
+      "grad_norm": 9.54863736482344,
+      "learning_rate": 1.1767067197015658e-05,
+      "loss": 0.6984,
+      "step": 8780
+    },
+    {
+      "epoch": 0.5741345525800131,
+      "grad_norm": 19.88307419400473,
+      "learning_rate": 1.1736712354128914e-05,
+      "loss": 0.6891,
+      "step": 8790
+    },
+    {
+      "epoch": 0.5747877204441542,
+      "grad_norm": 10.644397576568926,
+      "learning_rate": 1.1706371532460546e-05,
+      "loss": 0.679,
+      "step": 8800
+    },
+    {
+      "epoch": 0.5754408883082952,
+      "grad_norm": 2.3336898554043213,
+      "learning_rate": 1.1676044862374584e-05,
+      "loss": 0.7158,
+      "step": 8810
+    },
+    {
+      "epoch": 0.5760940561724364,
+      "grad_norm": 6.048589734364383,
+      "learning_rate": 1.1645732474174225e-05,
+      "loss": 0.6977,
+      "step": 8820
+    },
+    {
+      "epoch": 0.5767472240365774,
+      "grad_norm": 4.169729302491312,
+      "learning_rate": 1.1615434498101325e-05,
+      "loss": 0.6741,
+      "step": 8830
+    },
+    {
+      "epoch": 0.5774003919007185,
+      "grad_norm": 14.92365469180896,
+      "learning_rate": 1.1585151064335811e-05,
+      "loss": 0.7175,
+      "step": 8840
+    },
+    {
+      "epoch": 0.5780535597648596,
+      "grad_norm": 10.701959120465434,
+      "learning_rate": 1.1554882302995118e-05,
+      "loss": 0.6797,
+      "step": 8850
+    },
+    {
+      "epoch": 0.5787067276290007,
+      "grad_norm": 8.957942605184153,
+      "learning_rate": 1.1524628344133653e-05,
+      "loss": 0.7207,
+      "step": 8860
+    },
+    {
+      "epoch": 0.5793598954931417,
+      "grad_norm": 4.787159179029991,
+      "learning_rate": 1.1494389317742204e-05,
+      "loss": 0.7303,
+      "step": 8870
+    },
+    {
+      "epoch": 0.5800130633572829,
+      "grad_norm": 4.314912867368067,
+      "learning_rate": 1.1464165353747412e-05,
+      "loss": 0.7074,
+      "step": 8880
+    },
+    {
+      "epoch": 0.5806662312214239,
+      "grad_norm": 11.290676584978511,
+      "learning_rate": 1.14339565820112e-05,
+      "loss": 0.6827,
+      "step": 8890
+    },
+    {
+      "epoch": 0.581319399085565,
+      "grad_norm": 16.338731976431227,
+      "learning_rate": 1.1403763132330214e-05,
+      "loss": 0.7293,
+      "step": 8900
+    },
+    {
+      "epoch": 0.5819725669497061,
+      "grad_norm": 4.281523981168196,
+      "learning_rate": 1.1373585134435257e-05,
+      "loss": 0.6684,
+      "step": 8910
+    },
+    {
+      "epoch": 0.5826257348138472,
+      "grad_norm": 2.8456531905150486,
+      "learning_rate": 1.1343422717990753e-05,
+      "loss": 0.6874,
+      "step": 8920
+    },
+    {
+      "epoch": 0.5832789026779882,
+      "grad_norm": 52.77777843453033,
+      "learning_rate": 1.1313276012594174e-05,
+      "loss": 0.721,
+      "step": 8930
+    },
+    {
+      "epoch": 0.5839320705421294,
+      "grad_norm": 7.929333909142418,
+      "learning_rate": 1.1283145147775493e-05,
+      "loss": 0.6996,
+      "step": 8940
+    },
+    {
+      "epoch": 0.5845852384062704,
+      "grad_norm": 2.8746395714437747,
+      "learning_rate": 1.12530302529966e-05,
+      "loss": 0.725,
+      "step": 8950
+    },
+    {
+      "epoch": 0.5852384062704115,
+      "grad_norm": 4.914356701994025,
+      "learning_rate": 1.1222931457650792e-05,
+      "loss": 0.7305,
+      "step": 8960
+    },
+    {
+      "epoch": 0.5858915741345526,
+      "grad_norm": 7.837217247347996,
+      "learning_rate": 1.1192848891062181e-05,
+      "loss": 0.6508,
+      "step": 8970
+    },
+    {
+      "epoch": 0.5865447419986937,
+      "grad_norm": 3.4712253021031962,
+      "learning_rate": 1.1162782682485152e-05,
+      "loss": 0.6367,
+      "step": 8980
+    },
+    {
+      "epoch": 0.5871979098628347,
+      "grad_norm": 2.877607904262043,
+      "learning_rate": 1.1132732961103808e-05,
+      "loss": 0.6545,
+      "step": 8990
+    },
+    {
+      "epoch": 0.5878510777269759,
+      "grad_norm": 8.107637605161019,
+      "learning_rate": 1.11026998560314e-05,
+      "loss": 0.7188,
+      "step": 9000
+    },
+    {
+      "epoch": 0.5885042455911169,
+      "grad_norm": 4.6981324726393945,
+      "learning_rate": 1.1072683496309804e-05,
+      "loss": 0.7503,
+      "step": 9010
+    },
+    {
+      "epoch": 0.589157413455258,
+      "grad_norm": 3.394614970378694,
+      "learning_rate": 1.1042684010908929e-05,
+      "loss": 0.7021,
+      "step": 9020
+    },
+    {
+      "epoch": 0.5898105813193991,
+      "grad_norm": 8.911408439658569,
+      "learning_rate": 1.1012701528726187e-05,
+      "loss": 0.7031,
+      "step": 9030
+    },
+    {
+      "epoch": 0.5904637491835402,
+      "grad_norm": 4.013093499233945,
+      "learning_rate": 1.0982736178585939e-05,
+      "loss": 0.6802,
+      "step": 9040
+    },
+    {
+      "epoch": 0.5911169170476812,
+      "grad_norm": 7.257045873716964,
+      "learning_rate": 1.0952788089238924e-05,
+      "loss": 0.7254,
+      "step": 9050
+    },
+    {
+      "epoch": 0.5917700849118224,
+      "grad_norm": 14.982688068983514,
+      "learning_rate": 1.0922857389361734e-05,
+      "loss": 0.6788,
+      "step": 9060
+    },
+    {
+      "epoch": 0.5924232527759634,
+      "grad_norm": 91.51916103165227,
+      "learning_rate": 1.0892944207556227e-05,
+      "loss": 0.6519,
+      "step": 9070
+    },
+    {
+      "epoch": 0.5930764206401045,
+      "grad_norm": 6.205600657688591,
+      "learning_rate": 1.0863048672349008e-05,
+      "loss": 0.7368,
+      "step": 9080
+    },
+    {
+      "epoch": 0.5937295885042456,
+      "grad_norm": 6.509641392641457,
+      "learning_rate": 1.0833170912190846e-05,
+      "loss": 0.6625,
+      "step": 9090
+    },
+    {
+      "epoch": 0.5943827563683867,
+      "grad_norm": 2.6445889487797176,
+      "learning_rate": 1.0803311055456139e-05,
+      "loss": 0.7142,
+      "step": 9100
+    },
+    {
+      "epoch": 0.5950359242325277,
+      "grad_norm": 5.711217703774923,
+      "learning_rate": 1.0773469230442372e-05,
+      "loss": 0.6801,
+      "step": 9110
+    },
+    {
+      "epoch": 0.5956890920966689,
+      "grad_norm": 7.889073950278876,
+      "learning_rate": 1.074364556536954e-05,
+      "loss": 0.6735,
+      "step": 9120
+    },
+    {
+      "epoch": 0.5963422599608099,
+      "grad_norm": 3.353698192285854,
+      "learning_rate": 1.071384018837962e-05,
+      "loss": 0.657,
+      "step": 9130
+    },
+    {
+      "epoch": 0.596995427824951,
+      "grad_norm": 2.6333231897601395,
+      "learning_rate": 1.0684053227536007e-05,
+      "loss": 0.7279,
+      "step": 9140
+    },
+    {
+      "epoch": 0.5976485956890921,
+      "grad_norm": 5.586013814764972,
+      "learning_rate": 1.0654284810822972e-05,
+      "loss": 0.71,
+      "step": 9150
+    },
+    {
+      "epoch": 0.5983017635532332,
+      "grad_norm": 7.257685041180998,
+      "learning_rate": 1.0624535066145103e-05,
+      "loss": 0.6757,
+      "step": 9160
+    },
+    {
+      "epoch": 0.5989549314173742,
+      "grad_norm": 6.444891084708104,
+      "learning_rate": 1.0594804121326773e-05,
+      "loss": 0.7004,
+      "step": 9170
+    },
+    {
+      "epoch": 0.5996080992815154,
+      "grad_norm": 6.434512732255839,
+      "learning_rate": 1.0565092104111555e-05,
+      "loss": 0.6643,
+      "step": 9180
+    },
+    {
+      "epoch": 0.6002612671456564,
+      "grad_norm": 4.260366922972642,
+      "learning_rate": 1.0535399142161722e-05,
+      "loss": 0.7153,
+      "step": 9190
+    },
+    {
+      "epoch": 0.6009144350097975,
+      "grad_norm": 17.9026416385046,
+      "learning_rate": 1.050572536305765e-05,
+      "loss": 0.6703,
+      "step": 9200
+    },
+    {
+      "epoch": 0.6015676028739386,
+      "grad_norm": 6.1533366634295135,
+      "learning_rate": 1.0476070894297319e-05,
+      "loss": 0.6978,
+      "step": 9210
+    },
+    {
+      "epoch": 0.6022207707380797,
+      "grad_norm": 6.951694298470321,
+      "learning_rate": 1.0446435863295713e-05,
+      "loss": 0.6933,
+      "step": 9220
+    },
+    {
+      "epoch": 0.6028739386022207,
+      "grad_norm": 4.479107145165328,
+      "learning_rate": 1.041682039738432e-05,
+      "loss": 0.6497,
+      "step": 9230
+    },
+    {
+      "epoch": 0.6035271064663619,
+      "grad_norm": 7.84728121965074,
+      "learning_rate": 1.0387224623810553e-05,
+      "loss": 0.7165,
+      "step": 9240
+    },
+    {
+      "epoch": 0.6041802743305029,
+      "grad_norm": 3.5520424192576256,
+      "learning_rate": 1.0357648669737207e-05,
+      "loss": 0.7187,
+      "step": 9250
+    },
+    {
+      "epoch": 0.6048334421946441,
+      "grad_norm": 11.145100811330805,
+      "learning_rate": 1.0328092662241934e-05,
+      "loss": 0.7626,
+      "step": 9260
+    },
+    {
+      "epoch": 0.6054866100587851,
+      "grad_norm": 4.628007890145608,
+      "learning_rate": 1.0298556728316677e-05,
+      "loss": 0.7583,
+      "step": 9270
+    },
+    {
+      "epoch": 0.6061397779229262,
+      "grad_norm": 2.4723596379906847,
+      "learning_rate": 1.0269040994867126e-05,
+      "loss": 0.7218,
+      "step": 9280
+    },
+    {
+      "epoch": 0.6067929457870673,
+      "grad_norm": 3.549410413755828,
+      "learning_rate": 1.023954558871218e-05,
+      "loss": 0.7301,
+      "step": 9290
+    },
+    {
+      "epoch": 0.6074461136512084,
+      "grad_norm": 3.8402240987792906,
+      "learning_rate": 1.0210070636583397e-05,
+      "loss": 0.7194,
+      "step": 9300
+    },
+    {
+      "epoch": 0.6080992815153494,
+      "grad_norm": 9.342316890776898,
+      "learning_rate": 1.0180616265124454e-05,
+      "loss": 0.6945,
+      "step": 9310
+    },
+    {
+      "epoch": 0.6087524493794906,
+      "grad_norm": 8.938765856323773,
+      "learning_rate": 1.0151182600890605e-05,
+      "loss": 0.6951,
+      "step": 9320
+    },
+    {
+      "epoch": 0.6094056172436316,
+      "grad_norm": 6.274849282692598,
+      "learning_rate": 1.012176977034811e-05,
+      "loss": 0.6848,
+      "step": 9330
+    },
+    {
+      "epoch": 0.6100587851077727,
+      "grad_norm": 4.769425874462992,
+      "learning_rate": 1.0092377899873738e-05,
+      "loss": 0.6752,
+      "step": 9340
+    },
+    {
+      "epoch": 0.6107119529719138,
+      "grad_norm": 3.329658405780108,
+      "learning_rate": 1.006300711575419e-05,
+      "loss": 0.6937,
+      "step": 9350
+    },
+    {
+      "epoch": 0.6113651208360549,
+      "grad_norm": 5.471425475561112,
+      "learning_rate": 1.0033657544185567e-05,
+      "loss": 0.6984,
+      "step": 9360
+    },
+    {
+      "epoch": 0.6120182887001959,
+      "grad_norm": 4.918442658610056,
+      "learning_rate": 1.0004329311272832e-05,
+      "loss": 0.6955,
+      "step": 9370
+    },
+    {
+      "epoch": 0.6126714565643371,
+      "grad_norm": 4.707819525216195,
+      "learning_rate": 9.97502254302925e-06,
+      "loss": 0.6808,
+      "step": 9380
+    },
+    {
+      "epoch": 0.6133246244284781,
+      "grad_norm": 5.203240104066543,
+      "learning_rate": 9.945737365375876e-06,
+      "loss": 0.6939,
+      "step": 9390
+    },
+    {
+      "epoch": 0.6139777922926192,
+      "grad_norm": 9.333326975502391,
+      "learning_rate": 9.916473904140984e-06,
+      "loss": 0.6521,
+      "step": 9400
+    },
+    {
+      "epoch": 0.6146309601567603,
+      "grad_norm": 3.09437801267708,
+      "learning_rate": 9.887232285059548e-06,
+      "loss": 0.6954,
+      "step": 9410
+    },
+    {
+      "epoch": 0.6152841280209014,
+      "grad_norm": 32.01391680411232,
+      "learning_rate": 9.85801263377269e-06,
+      "loss": 0.7049,
+      "step": 9420
+    },
+    {
+      "epoch": 0.6159372958850424,
+      "grad_norm": 5.260083641866695,
+      "learning_rate": 9.828815075827148e-06,
+      "loss": 0.6741,
+      "step": 9430
+    },
+    {
+      "epoch": 0.6165904637491836,
+      "grad_norm": 4.0416270513002255,
+      "learning_rate": 9.799639736674729e-06,
+      "loss": 0.7333,
+      "step": 9440
+    },
+    {
+      "epoch": 0.6172436316133246,
+      "grad_norm": 5.155925948624099,
+      "learning_rate": 9.770486741671777e-06,
+      "loss": 0.7013,
+      "step": 9450
+    },
+    {
+      "epoch": 0.6178967994774657,
+      "grad_norm": 3.91699790906619,
+      "learning_rate": 9.74135621607863e-06,
+      "loss": 0.7214,
+      "step": 9460
+    },
+    {
+      "epoch": 0.6185499673416068,
+      "grad_norm": 6.112641926933419,
+      "learning_rate": 9.712248285059079e-06,
+      "loss": 0.6964,
+      "step": 9470
+    },
+    {
+      "epoch": 0.6192031352057479,
+      "grad_norm": 3.154827892850623,
+      "learning_rate": 9.683163073679831e-06,
+      "loss": 0.6868,
+      "step": 9480
+    },
+    {
+      "epoch": 0.6198563030698889,
+      "grad_norm": 4.0062924006415805,
+      "learning_rate": 9.65410070690999e-06,
+      "loss": 0.671,
+      "step": 9490
+    },
+    {
+      "epoch": 0.6205094709340301,
+      "grad_norm": 3.703242561631074,
+      "learning_rate": 9.625061309620487e-06,
+      "loss": 0.6945,
+      "step": 9500
+    },
+    {
+      "epoch": 0.6211626387981711,
+      "grad_norm": 6.392048497671309,
+      "learning_rate": 9.59604500658357e-06,
+      "loss": 0.6774,
+      "step": 9510
+    },
+    {
+      "epoch": 0.6218158066623122,
+      "grad_norm": 5.178076400022287,
+      "learning_rate": 9.56705192247226e-06,
+      "loss": 0.7251,
+      "step": 9520
+    },
+    {
+      "epoch": 0.6224689745264533,
+      "grad_norm": 2.5070876740750947,
+      "learning_rate": 9.53808218185981e-06,
+      "loss": 0.6909,
+      "step": 9530
+    },
+    {
+      "epoch": 0.6231221423905944,
+      "grad_norm": 4.151677190296909,
+      "learning_rate": 9.509135909219178e-06,
+      "loss": 0.6889,
+      "step": 9540
+    },
+    {
+      "epoch": 0.6237753102547354,
+      "grad_norm": 3.542263040633155,
+      "learning_rate": 9.48021322892249e-06,
+      "loss": 0.6894,
+      "step": 9550
+    },
+    {
+      "epoch": 0.6244284781188766,
+      "grad_norm": 6.583784192577517,
+      "learning_rate": 9.451314265240489e-06,
+      "loss": 0.6659,
+      "step": 9560
+    },
+    {
+      "epoch": 0.6250816459830176,
+      "grad_norm": 5.24457887775372,
+      "learning_rate": 9.422439142342035e-06,
+      "loss": 0.6829,
+      "step": 9570
+    },
+    {
+      "epoch": 0.6257348138471587,
+      "grad_norm": 7.216483152433425,
+      "learning_rate": 9.393587984293546e-06,
+      "loss": 0.7424,
+      "step": 9580
+    },
+    {
+      "epoch": 0.6263879817112998,
+      "grad_norm": 5.40830328432894,
+      "learning_rate": 9.36476091505846e-06,
+      "loss": 0.7183,
+      "step": 9590
+    },
+    {
+      "epoch": 0.6270411495754409,
+      "grad_norm": 2.612648013361852,
+      "learning_rate": 9.335958058496734e-06,
+      "loss": 0.6786,
+      "step": 9600
+    },
+    {
+      "epoch": 0.6276943174395819,
+      "grad_norm": 3.9662598735254826,
+      "learning_rate": 9.307179538364274e-06,
+      "loss": 0.7128,
+      "step": 9610
+    },
+    {
+      "epoch": 0.6283474853037231,
+      "grad_norm": 5.322804409368759,
+      "learning_rate": 9.278425478312437e-06,
+      "loss": 0.6438,
+      "step": 9620
+    },
+    {
+      "epoch": 0.6290006531678641,
+      "grad_norm": 4.24650699461021,
+      "learning_rate": 9.249696001887462e-06,
+      "loss": 0.7191,
+      "step": 9630
+    },
+    {
+      "epoch": 0.6296538210320052,
+      "grad_norm": 4.311181129354523,
+      "learning_rate": 9.220991232529977e-06,
+      "loss": 0.6694,
+      "step": 9640
+    },
+    {
+      "epoch": 0.6303069888961463,
+      "grad_norm": 2.438796652186847,
+      "learning_rate": 9.192311293574452e-06,
+      "loss": 0.6893,
+      "step": 9650
+    },
+    {
+      "epoch": 0.6309601567602874,
+      "grad_norm": 2.3407674504161,
+      "learning_rate": 9.163656308248666e-06,
+      "loss": 0.6814,
+      "step": 9660
+    },
+    {
+      "epoch": 0.6316133246244284,
+      "grad_norm": 5.193439989443432,
+      "learning_rate": 9.135026399673175e-06,
+      "loss": 0.7126,
+      "step": 9670
+    },
+    {
+      "epoch": 0.6322664924885696,
+      "grad_norm": 11.270517585736275,
+      "learning_rate": 9.106421690860796e-06,
+      "loss": 0.7057,
+      "step": 9680
+    },
+    {
+      "epoch": 0.6329196603527106,
+      "grad_norm": 5.1149136915038165,
+      "learning_rate": 9.077842304716069e-06,
+      "loss": 0.7197,
+      "step": 9690
+    },
+    {
+      "epoch": 0.6335728282168517,
+      "grad_norm": 4.952294505678321,
+      "learning_rate": 9.049288364034742e-06,
+      "loss": 0.6765,
+      "step": 9700
+    },
+    {
+      "epoch": 0.6342259960809928,
+      "grad_norm": 7.154019337112617,
+      "learning_rate": 9.020759991503207e-06,
+      "loss": 0.726,
+      "step": 9710
+    },
+    {
+      "epoch": 0.6348791639451339,
+      "grad_norm": 16.238808871780552,
+      "learning_rate": 8.99225730969802e-06,
+      "loss": 0.7069,
+      "step": 9720
+    },
+    {
+      "epoch": 0.6355323318092749,
+      "grad_norm": 2.991481926898025,
+      "learning_rate": 8.963780441085347e-06,
+      "loss": 0.6926,
+      "step": 9730
+    },
+    {
+      "epoch": 0.6361854996734161,
+      "grad_norm": 3.3243848532788682,
+      "learning_rate": 8.935329508020446e-06,
+      "loss": 0.7167,
+      "step": 9740
+    },
+    {
+      "epoch": 0.6368386675375571,
+      "grad_norm": 4.070596461723911,
+      "learning_rate": 8.906904632747137e-06,
+      "loss": 0.7356,
+      "step": 9750
+    },
+    {
+      "epoch": 0.6374918354016982,
+      "grad_norm": 3.7465849267000575,
+      "learning_rate": 8.878505937397272e-06,
+      "loss": 0.6961,
+      "step": 9760
+    },
+    {
+      "epoch": 0.6381450032658393,
+      "grad_norm": 3.790695499226875,
+      "learning_rate": 8.850133543990228e-06,
+      "loss": 0.6806,
+      "step": 9770
+    },
+    {
+      "epoch": 0.6387981711299804,
+      "grad_norm": 7.764921037477569,
+      "learning_rate": 8.821787574432371e-06,
+      "loss": 0.6691,
+      "step": 9780
+    },
+    {
+      "epoch": 0.6394513389941215,
+      "grad_norm": 5.566876179450992,
+      "learning_rate": 8.793468150516517e-06,
+      "loss": 0.7063,
+      "step": 9790
+    },
+    {
+      "epoch": 0.6401045068582626,
+      "grad_norm": 3.4053216071478345,
+      "learning_rate": 8.765175393921441e-06,
+      "loss": 0.6836,
+      "step": 9800
+    },
+    {
+      "epoch": 0.6407576747224036,
+      "grad_norm": 5.502089363662013,
+      "learning_rate": 8.736909426211335e-06,
+      "loss": 0.6388,
+      "step": 9810
+    },
+    {
+      "epoch": 0.6414108425865448,
+      "grad_norm": 2.7638694234276517,
+      "learning_rate": 8.708670368835286e-06,
+      "loss": 0.6598,
+      "step": 9820
+    },
+    {
+      "epoch": 0.6420640104506858,
+      "grad_norm": 12.233396527562816,
+      "learning_rate": 8.680458343126753e-06,
+      "loss": 0.7278,
+      "step": 9830
+    },
+    {
+      "epoch": 0.6427171783148269,
+      "grad_norm": 4.917255280169664,
+      "learning_rate": 8.65227347030306e-06,
+      "loss": 0.6496,
+      "step": 9840
+    },
+    {
+      "epoch": 0.643370346178968,
+      "grad_norm": 4.7377241016215725,
+      "learning_rate": 8.624115871464852e-06,
+      "loss": 0.6733,
+      "step": 9850
+    },
+    {
+      "epoch": 0.6440235140431091,
+      "grad_norm": 2.7426327074990207,
+      "learning_rate": 8.595985667595596e-06,
+      "loss": 0.7166,
+      "step": 9860
+    },
+    {
+      "epoch": 0.6446766819072501,
+      "grad_norm": 73.36784262898324,
+      "learning_rate": 8.56788297956104e-06,
+      "loss": 0.7123,
+      "step": 9870
+    },
+    {
+      "epoch": 0.6453298497713913,
+      "grad_norm": 3.134712080929007,
+      "learning_rate": 8.539807928108728e-06,
+      "loss": 0.6646,
+      "step": 9880
+    },
+    {
+      "epoch": 0.6459830176355323,
+      "grad_norm": 8.323293003427095,
+      "learning_rate": 8.511760633867436e-06,
+      "loss": 0.649,
+      "step": 9890
+    },
+    {
+      "epoch": 0.6466361854996734,
+      "grad_norm": 7.07615085043799,
+      "learning_rate": 8.483741217346696e-06,
+      "loss": 0.6633,
+      "step": 9900
+    },
+    {
+      "epoch": 0.6472893533638145,
+      "grad_norm": 7.961109043478078,
+      "learning_rate": 8.455749798936245e-06,
+      "loss": 0.7161,
+      "step": 9910
+    },
+    {
+      "epoch": 0.6479425212279556,
+      "grad_norm": 2.4133702739326104,
+      "learning_rate": 8.42778649890552e-06,
+      "loss": 0.665,
+      "step": 9920
+    },
+    {
+      "epoch": 0.6485956890920966,
+      "grad_norm": 7.42532223602706,
+      "learning_rate": 8.399851437403172e-06,
+      "loss": 0.6916,
+      "step": 9930
+    },
+    {
+      "epoch": 0.6492488569562378,
+      "grad_norm": 3.5245869658336026,
+      "learning_rate": 8.371944734456469e-06,
+      "loss": 0.7338,
+      "step": 9940
+    },
+    {
+      "epoch": 0.6499020248203788,
+      "grad_norm": 5.328557956142334,
+      "learning_rate": 8.344066509970884e-06,
+      "loss": 0.6911,
+      "step": 9950
+    },
+    {
+      "epoch": 0.6505551926845199,
+      "grad_norm": 7.668967882961343,
+      "learning_rate": 8.316216883729493e-06,
+      "loss": 0.6889,
+      "step": 9960
+    },
+    {
+      "epoch": 0.651208360548661,
+      "grad_norm": 8.063885237104227,
+      "learning_rate": 8.288395975392515e-06,
+      "loss": 0.7142,
+      "step": 9970
+    },
+    {
+      "epoch": 0.6518615284128021,
+      "grad_norm": 10.741056048367856,
+      "learning_rate": 8.260603904496769e-06,
+      "loss": 0.7316,
+      "step": 9980
+    },
+    {
+      "epoch": 0.6525146962769431,
+      "grad_norm": 4.3002787477499185,
+      "learning_rate": 8.232840790455173e-06,
+      "loss": 0.6964,
+      "step": 9990
+    },
+    {
+      "epoch": 0.6531678641410843,
+      "grad_norm": 4.695114258502196,
+      "learning_rate": 8.205106752556227e-06,
+      "loss": 0.7238,
+      "step": 10000
+    },
+    {
+      "epoch": 0.6538210320052253,
+      "grad_norm": 11.537353885010582,
+      "learning_rate": 8.177401909963496e-06,
+      "loss": 0.7204,
+      "step": 10010
+    },
+    {
+      "epoch": 0.6544741998693664,
+      "grad_norm": 13.72025313430244,
+      "learning_rate": 8.149726381715108e-06,
+      "loss": 0.7004,
+      "step": 10020
+    },
+    {
+      "epoch": 0.6551273677335075,
+      "grad_norm": 4.275313668148756,
+      "learning_rate": 8.122080286723233e-06,
+      "loss": 0.6829,
+      "step": 10030
+    },
+    {
+      "epoch": 0.6557805355976486,
+      "grad_norm": 2.261053906695268,
+      "learning_rate": 8.094463743773587e-06,
+      "loss": 0.7081,
+      "step": 10040
+    },
+    {
+      "epoch": 0.6564337034617896,
+      "grad_norm": 15.40058816554506,
+      "learning_rate": 8.066876871524893e-06,
+      "loss": 0.7099,
+      "step": 10050
+    },
+    {
+      "epoch": 0.6570868713259308,
+      "grad_norm": 9.434056746531544,
+      "learning_rate": 8.039319788508413e-06,
+      "loss": 0.6964,
+      "step": 10060
+    },
+    {
+      "epoch": 0.6577400391900718,
+      "grad_norm": 4.916230564927696,
+      "learning_rate": 8.011792613127389e-06,
+      "loss": 0.6974,
+      "step": 10070
+    },
+    {
+      "epoch": 0.6583932070542129,
+      "grad_norm": 3.415752629777518,
+      "learning_rate": 7.984295463656591e-06,
+      "loss": 0.6811,
+      "step": 10080
+    },
+    {
+      "epoch": 0.659046374918354,
+      "grad_norm": 3.159574664982398,
+      "learning_rate": 7.956828458241738e-06,
+      "loss": 0.7241,
+      "step": 10090
+    },
+    {
+      "epoch": 0.6596995427824951,
+      "grad_norm": 7.408267490822411,
+      "learning_rate": 7.929391714899066e-06,
+      "loss": 0.6528,
+      "step": 10100
+    },
+    {
+      "epoch": 0.6603527106466361,
+      "grad_norm": 8.520105705500969,
+      "learning_rate": 7.901985351514772e-06,
+      "loss": 0.7304,
+      "step": 10110
+    },
+    {
+      "epoch": 0.6610058785107773,
+      "grad_norm": 4.930976711189574,
+      "learning_rate": 7.874609485844513e-06,
+      "loss": 0.6879,
+      "step": 10120
+    },
+    {
+      "epoch": 0.6616590463749183,
+      "grad_norm": 4.4618731317544285,
+      "learning_rate": 7.847264235512924e-06,
+      "loss": 0.6824,
+      "step": 10130
+    },
+    {
+      "epoch": 0.6623122142390594,
+      "grad_norm": 5.219787677232259,
+      "learning_rate": 7.819949718013077e-06,
+      "loss": 0.7147,
+      "step": 10140
+    },
+    {
+      "epoch": 0.6629653821032006,
+      "grad_norm": 3.99345983845127,
+      "learning_rate": 7.792666050706023e-06,
+      "loss": 0.6897,
+      "step": 10150
+    },
+    {
+      "epoch": 0.6636185499673416,
+      "grad_norm": 5.542787577146197,
+      "learning_rate": 7.765413350820236e-06,
+      "loss": 0.6884,
+      "step": 10160
+    },
+    {
+      "epoch": 0.6642717178314826,
+      "grad_norm": 30.100215102532907,
+      "learning_rate": 7.73819173545114e-06,
+      "loss": 0.67,
+      "step": 10170
+    },
+    {
+      "epoch": 0.6649248856956238,
+      "grad_norm": 4.793597664333635,
+      "learning_rate": 7.711001321560596e-06,
+      "loss": 0.685,
+      "step": 10180
+    },
+    {
+      "epoch": 0.6655780535597648,
+      "grad_norm": 8.678629566067418,
+      "learning_rate": 7.683842225976423e-06,
+      "loss": 0.6828,
+      "step": 10190
+    },
+    {
+      "epoch": 0.6662312214239059,
+      "grad_norm": 3.6349803731984025,
+      "learning_rate": 7.656714565391852e-06,
+      "loss": 0.6821,
+      "step": 10200
+    },
+    {
+      "epoch": 0.666884389288047,
+      "grad_norm": 8.673927297972229,
+      "learning_rate": 7.629618456365055e-06,
+      "loss": 0.7103,
+      "step": 10210
+    },
+    {
+      "epoch": 0.6675375571521881,
+      "grad_norm": 8.057835375428068,
+      "learning_rate": 7.60255401531865e-06,
+      "loss": 0.6935,
+      "step": 10220
+    },
+    {
+      "epoch": 0.6681907250163291,
+      "grad_norm": 9.950279991005985,
+      "learning_rate": 7.5755213585391775e-06,
+      "loss": 0.6699,
+      "step": 10230
+    },
+    {
+      "epoch": 0.6688438928804703,
+      "grad_norm": 6.578518164839038,
+      "learning_rate": 7.548520602176613e-06,
+      "loss": 0.6962,
+      "step": 10240
+    },
+    {
+      "epoch": 0.6694970607446113,
+      "grad_norm": 3.6595824615281742,
+      "learning_rate": 7.521551862243861e-06,
+      "loss": 0.6775,
+      "step": 10250
+    },
+    {
+      "epoch": 0.6701502286087524,
+      "grad_norm": 6.542277248906721,
+      "learning_rate": 7.4946152546162815e-06,
+      "loss": 0.7179,
+      "step": 10260
+    },
+    {
+      "epoch": 0.6708033964728936,
+      "grad_norm": 3.1715830400767593,
+      "learning_rate": 7.46771089503115e-06,
+      "loss": 0.6577,
+      "step": 10270
+    },
+    {
+      "epoch": 0.6714565643370346,
+      "grad_norm": 9.166993800019167,
+      "learning_rate": 7.4408388990872086e-06,
+      "loss": 0.7217,
+      "step": 10280
+    },
+    {
+      "epoch": 0.6721097322011756,
+      "grad_norm": 3.1457370128580266,
+      "learning_rate": 7.41399938224412e-06,
+      "loss": 0.7238,
+      "step": 10290
+    },
+    {
+      "epoch": 0.6727629000653168,
+      "grad_norm": 5.087128249324293,
+      "learning_rate": 7.387192459822002e-06,
+      "loss": 0.6834,
+      "step": 10300
+    },
+    {
+      "epoch": 0.6734160679294579,
+      "grad_norm": 3.3234894038686136,
+      "learning_rate": 7.360418247000945e-06,
+      "loss": 0.6807,
+      "step": 10310
+    },
+    {
+      "epoch": 0.674069235793599,
+      "grad_norm": 3.0582560502629152,
+      "learning_rate": 7.333676858820461e-06,
+      "loss": 0.721,
+      "step": 10320
+    },
+    {
+      "epoch": 0.67472240365774,
+      "grad_norm": 2.7130806226776465,
+      "learning_rate": 7.3069684101790594e-06,
+      "loss": 0.6809,
+      "step": 10330
+    },
+    {
+      "epoch": 0.6753755715218811,
+      "grad_norm": 3.7810845457086617,
+      "learning_rate": 7.2802930158336974e-06,
+      "loss": 0.7271,
+      "step": 10340
+    },
+    {
+      "epoch": 0.6760287393860223,
+      "grad_norm": 5.206439454660348,
+      "learning_rate": 7.253650790399333e-06,
+      "loss": 0.6402,
+      "step": 10350
+    },
+    {
+      "epoch": 0.6766819072501633,
+      "grad_norm": 10.397540139012824,
+      "learning_rate": 7.2270418483483785e-06,
+      "loss": 0.6979,
+      "step": 10360
+    },
+    {
+      "epoch": 0.6773350751143044,
+      "grad_norm": 14.072917690695897,
+      "learning_rate": 7.2004663040102666e-06,
+      "loss": 0.7332,
+      "step": 10370
+    },
+    {
+      "epoch": 0.6779882429784455,
+      "grad_norm": 20.837793142434133,
+      "learning_rate": 7.173924271570917e-06,
+      "loss": 0.7167,
+      "step": 10380
+    },
+    {
+      "epoch": 0.6786414108425866,
+      "grad_norm": 2.95535033006669,
+      "learning_rate": 7.147415865072263e-06,
+      "loss": 0.691,
+      "step": 10390
+    },
+    {
+      "epoch": 0.6792945787067276,
+      "grad_norm": 5.20187262048636,
+      "learning_rate": 7.120941198411757e-06,
+      "loss": 0.7248,
+      "step": 10400
+    },
+    {
+      "epoch": 0.6799477465708688,
+      "grad_norm": 14.561008884563188,
+      "learning_rate": 7.094500385341882e-06,
+      "loss": 0.7156,
+      "step": 10410
+    },
+    {
+      "epoch": 0.6806009144350098,
+      "grad_norm": 6.659356337135833,
+      "learning_rate": 7.068093539469674e-06,
+      "loss": 0.68,
+      "step": 10420
+    },
+    {
+      "epoch": 0.6812540822991509,
+      "grad_norm": 7.950217324411396,
+      "learning_rate": 7.0417207742562106e-06,
+      "loss": 0.6544,
+      "step": 10430
+    },
+    {
+      "epoch": 0.681907250163292,
+      "grad_norm": 2.5071824848115853,
+      "learning_rate": 7.015382203016151e-06,
+      "loss": 0.7305,
+      "step": 10440
+    },
+    {
+      "epoch": 0.6825604180274331,
+      "grad_norm": 6.4374528901681884,
+      "learning_rate": 6.989077938917218e-06,
+      "loss": 0.7117,
+      "step": 10450
+    },
+    {
+      "epoch": 0.6832135858915741,
+      "grad_norm": 3.7611806134016508,
+      "learning_rate": 6.96280809497975e-06,
+      "loss": 0.6836,
+      "step": 10460
+    },
+    {
+      "epoch": 0.6838667537557153,
+      "grad_norm": 5.700978499632452,
+      "learning_rate": 6.93657278407616e-06,
+      "loss": 0.6625,
+      "step": 10470
+    },
+    {
+      "epoch": 0.6845199216198563,
+      "grad_norm": 2.581514477053975,
+      "learning_rate": 6.910372118930523e-06,
+      "loss": 0.6672,
+      "step": 10480
+    },
+    {
+      "epoch": 0.6851730894839974,
+      "grad_norm": 12.617504349736718,
+      "learning_rate": 6.8842062121180274e-06,
+      "loss": 0.6884,
+      "step": 10490
+    },
+    {
+      "epoch": 0.6858262573481385,
+      "grad_norm": 2.5495646497807947,
+      "learning_rate": 6.858075176064523e-06,
+      "loss": 0.7237,
+      "step": 10500
+    },
+    {
+      "epoch": 0.6864794252122796,
+      "grad_norm": 9.075860218651238,
+      "learning_rate": 6.831979123046042e-06,
+      "loss": 0.6929,
+      "step": 10510
+    },
+    {
+      "epoch": 0.6871325930764206,
+      "grad_norm": 11.79800801678243,
+      "learning_rate": 6.805918165188288e-06,
+      "loss": 0.6983,
+      "step": 10520
+    },
+    {
+      "epoch": 0.6877857609405618,
+      "grad_norm": 16.602144291276336,
+      "learning_rate": 6.779892414466196e-06,
+      "loss": 0.7003,
+      "step": 10530
+    },
+    {
+      "epoch": 0.6884389288047028,
+      "grad_norm": 16.30785786139135,
+      "learning_rate": 6.75390198270341e-06,
+      "loss": 0.7112,
+      "step": 10540
+    },
+    {
+      "epoch": 0.6890920966688439,
+      "grad_norm": 2.458437276248381,
+      "learning_rate": 6.727946981571826e-06,
+      "loss": 0.7271,
+      "step": 10550
+    },
+    {
+      "epoch": 0.689745264532985,
+      "grad_norm": 5.955957574861067,
+      "learning_rate": 6.702027522591101e-06,
+      "loss": 0.6631,
+      "step": 10560
+    },
+    {
+      "epoch": 0.6903984323971261,
+      "grad_norm": 2.8278119667432966,
+      "learning_rate": 6.676143717128197e-06,
+      "loss": 0.6936,
+      "step": 10570
+    },
+    {
+      "epoch": 0.6910516002612671,
+      "grad_norm": 4.3483105327397835,
+      "learning_rate": 6.65029567639687e-06,
+      "loss": 0.6828,
+      "step": 10580
+    },
+    {
+      "epoch": 0.6917047681254083,
+      "grad_norm": 3.947818013403805,
+      "learning_rate": 6.624483511457204e-06,
+      "loss": 0.712,
+      "step": 10590
+    },
+    {
+      "epoch": 0.6923579359895493,
+      "grad_norm": 3.374901805438211,
+      "learning_rate": 6.598707333215154e-06,
+      "loss": 0.7021,
+      "step": 10600
+    },
+    {
+      "epoch": 0.6930111038536904,
+      "grad_norm": 6.057687009033569,
+      "learning_rate": 6.5729672524220365e-06,
+      "loss": 0.6851,
+      "step": 10610
+    },
+    {
+      "epoch": 0.6936642717178315,
+      "grad_norm": 3.843916602517143,
+      "learning_rate": 6.5472633796740885e-06,
+      "loss": 0.6767,
+      "step": 10620
+    },
+    {
+      "epoch": 0.6943174395819726,
+      "grad_norm": 4.808172683191776,
+      "learning_rate": 6.521595825411942e-06,
+      "loss": 0.7512,
+      "step": 10630
+    },
+    {
+      "epoch": 0.6949706074461136,
+      "grad_norm": 2.3894514806323017,
+      "learning_rate": 6.495964699920215e-06,
+      "loss": 0.674,
+      "step": 10640
+    },
+    {
+      "epoch": 0.6956237753102548,
+      "grad_norm": 4.533425486319881,
+      "learning_rate": 6.4703701133269795e-06,
+      "loss": 0.7744,
+      "step": 10650
+    },
+    {
+      "epoch": 0.6962769431743958,
+      "grad_norm": 4.777684895896499,
+      "learning_rate": 6.444812175603333e-06,
+      "loss": 0.6787,
+      "step": 10660
+    },
+    {
+      "epoch": 0.6969301110385369,
+      "grad_norm": 6.419177600320241,
+      "learning_rate": 6.419290996562885e-06,
+      "loss": 0.6809,
+      "step": 10670
+    },
+    {
+      "epoch": 0.697583278902678,
+      "grad_norm": 7.25828815129346,
+      "learning_rate": 6.393806685861316e-06,
+      "loss": 0.7134,
+      "step": 10680
+    },
+    {
+      "epoch": 0.6982364467668191,
+      "grad_norm": 4.27661929331523,
+      "learning_rate": 6.368359352995906e-06,
+      "loss": 0.7359,
+      "step": 10690
+    },
+    {
+      "epoch": 0.6988896146309601,
+      "grad_norm": 3.8979699842570548,
+      "learning_rate": 6.342949107305026e-06,
+      "loss": 0.6823,
+      "step": 10700
+    },
+    {
+      "epoch": 0.6995427824951013,
+      "grad_norm": 3.649185659413299,
+      "learning_rate": 6.317576057967728e-06,
+      "loss": 0.6787,
+      "step": 10710
+    },
+    {
+      "epoch": 0.7001959503592423,
+      "grad_norm": 10.476570510576728,
+      "learning_rate": 6.292240314003217e-06,
+      "loss": 0.6764,
+      "step": 10720
+    },
+    {
+      "epoch": 0.7008491182233834,
+      "grad_norm": 20.378338312121556,
+      "learning_rate": 6.266941984270434e-06,
+      "loss": 0.6936,
+      "step": 10730
+    },
+    {
+      "epoch": 0.7015022860875245,
+      "grad_norm": 4.145143282154411,
+      "learning_rate": 6.241681177467542e-06,
+      "loss": 0.648,
+      "step": 10740
+    },
+    {
+      "epoch": 0.7021554539516656,
+      "grad_norm": 4.968717562971461,
+      "learning_rate": 6.216458002131502e-06,
+      "loss": 0.6815,
+      "step": 10750
+    },
+    {
+      "epoch": 0.7028086218158066,
+      "grad_norm": 3.49150362081721,
+      "learning_rate": 6.1912725666375695e-06,
+      "loss": 0.6583,
+      "step": 10760
+    },
+    {
+      "epoch": 0.7034617896799478,
+      "grad_norm": 3.0513662248766322,
+      "learning_rate": 6.166124979198849e-06,
+      "loss": 0.6995,
+      "step": 10770
+    },
+    {
+      "epoch": 0.7041149575440888,
+      "grad_norm": 5.397307127156423,
+      "learning_rate": 6.141015347865828e-06,
+      "loss": 0.6558,
+      "step": 10780
+    },
+    {
+      "epoch": 0.7047681254082299,
+      "grad_norm": 3.1775879172611146,
+      "learning_rate": 6.1159437805259e-06,
+      "loss": 0.7086,
+      "step": 10790
+    },
+    {
+      "epoch": 0.705421293272371,
+      "grad_norm": 6.070511077028698,
+      "learning_rate": 6.090910384902932e-06,
+      "loss": 0.7152,
+      "step": 10800
+    },
+    {
+      "epoch": 0.7060744611365121,
+      "grad_norm": 11.451698703609093,
+      "learning_rate": 6.065915268556756e-06,
+      "loss": 0.6721,
+      "step": 10810
+    },
+    {
+      "epoch": 0.7067276290006531,
+      "grad_norm": 4.040931608599787,
+      "learning_rate": 6.040958538882752e-06,
+      "loss": 0.6974,
+      "step": 10820
+    },
+    {
+      "epoch": 0.7073807968647943,
+      "grad_norm": 8.50160291928914,
+      "learning_rate": 6.016040303111346e-06,
+      "loss": 0.6284,
+      "step": 10830
+    },
+    {
+      "epoch": 0.7080339647289353,
+      "grad_norm": 9.50722108755325,
+      "learning_rate": 5.991160668307587e-06,
+      "loss": 0.6882,
+      "step": 10840
+    },
+    {
+      "epoch": 0.7086871325930765,
+      "grad_norm": 7.7646464109268205,
+      "learning_rate": 5.966319741370658e-06,
+      "loss": 0.7234,
+      "step": 10850
+    },
+    {
+      "epoch": 0.7093403004572175,
+      "grad_norm": 5.587117786690515,
+      "learning_rate": 5.941517629033432e-06,
+      "loss": 0.6888,
+      "step": 10860
+    },
+    {
+      "epoch": 0.7099934683213586,
+      "grad_norm": 4.801651654593505,
+      "learning_rate": 5.916754437862004e-06,
+      "loss": 0.7223,
+      "step": 10870
+    },
+    {
+      "epoch": 0.7106466361854997,
+      "grad_norm": 10.77526477037024,
+      "learning_rate": 5.89203027425524e-06,
+      "loss": 0.667,
+      "step": 10880
+    },
+    {
+      "epoch": 0.7112998040496408,
+      "grad_norm": 4.811645630139854,
+      "learning_rate": 5.867345244444328e-06,
+      "loss": 0.6292,
+      "step": 10890
+    },
+    {
+      "epoch": 0.7119529719137818,
+      "grad_norm": 3.9493363511020103,
+      "learning_rate": 5.8426994544922955e-06,
+      "loss": 0.7109,
+      "step": 10900
+    },
+    {
+      "epoch": 0.712606139777923,
+      "grad_norm": 4.308805569433444,
+      "learning_rate": 5.818093010293586e-06,
+      "loss": 0.722,
+      "step": 10910
+    },
+    {
+      "epoch": 0.713259307642064,
+      "grad_norm": 3.8841534453939537,
+      "learning_rate": 5.793526017573577e-06,
+      "loss": 0.7473,
+      "step": 10920
+    },
+    {
+      "epoch": 0.7139124755062051,
+      "grad_norm": 2.9837089714102847,
+      "learning_rate": 5.768998581888138e-06,
+      "loss": 0.6982,
+      "step": 10930
+    },
+    {
+      "epoch": 0.7145656433703462,
+      "grad_norm": 2.9982269133645367,
+      "learning_rate": 5.7445108086231715e-06,
+      "loss": 0.6836,
+      "step": 10940
+    },
+    {
+      "epoch": 0.7152188112344873,
+      "grad_norm": 4.547014476826066,
+      "learning_rate": 5.720062802994181e-06,
+      "loss": 0.6605,
+      "step": 10950
+    },
+    {
+      "epoch": 0.7158719790986283,
+      "grad_norm": 2.1769079718774704,
+      "learning_rate": 5.6956546700457885e-06,
+      "loss": 0.6883,
+      "step": 10960
+    },
+    {
+      "epoch": 0.7165251469627695,
+      "grad_norm": 5.03574147163945,
+      "learning_rate": 5.671286514651289e-06,
+      "loss": 0.7189,
+      "step": 10970
+    },
+    {
+      "epoch": 0.7171783148269105,
+      "grad_norm": 4.862398145535597,
+      "learning_rate": 5.646958441512234e-06,
+      "loss": 0.7095,
+      "step": 10980
+    },
+    {
+      "epoch": 0.7178314826910516,
+      "grad_norm": 15.308086895953474,
+      "learning_rate": 5.622670555157924e-06,
+      "loss": 0.6784,
+      "step": 10990
+    },
+    {
+      "epoch": 0.7184846505551927,
+      "grad_norm": 9.436744508124669,
+      "learning_rate": 5.5984229599450275e-06,
+      "loss": 0.709,
+      "step": 11000
+    },
+    {
+      "epoch": 0.7191378184193338,
+      "grad_norm": 2.532456946200012,
+      "learning_rate": 5.57421576005705e-06,
+      "loss": 0.7034,
+      "step": 11010
+    },
+    {
+      "epoch": 0.7197909862834748,
+      "grad_norm": 5.041112107392264,
+      "learning_rate": 5.550049059503976e-06,
+      "loss": 0.711,
+      "step": 11020
+    },
+    {
+      "epoch": 0.720444154147616,
+      "grad_norm": 5.007525312309261,
+      "learning_rate": 5.525922962121746e-06,
+      "loss": 0.6772,
+      "step": 11030
+    },
+    {
+      "epoch": 0.721097322011757,
+      "grad_norm": 3.8826844100241056,
+      "learning_rate": 5.50183757157187e-06,
+      "loss": 0.6671,
+      "step": 11040
+    },
+    {
+      "epoch": 0.7217504898758981,
+      "grad_norm": 3.8447345000076814,
+      "learning_rate": 5.477792991340932e-06,
+      "loss": 0.7535,
+      "step": 11050
+    },
+    {
+      "epoch": 0.7224036577400392,
+      "grad_norm": 6.033444336490015,
+      "learning_rate": 5.453789324740175e-06,
+      "loss": 0.7139,
+      "step": 11060
+    },
+    {
+      "epoch": 0.7230568256041803,
+      "grad_norm": 4.553578681833169,
+      "learning_rate": 5.4298266749050616e-06,
+      "loss": 0.6643,
+      "step": 11070
+    },
+    {
+      "epoch": 0.7237099934683213,
+      "grad_norm": 7.790695891534031,
+      "learning_rate": 5.405905144794807e-06,
+      "loss": 0.7111,
+      "step": 11080
+    },
+    {
+      "epoch": 0.7243631613324625,
+      "grad_norm": 22.8654485053664,
+      "learning_rate": 5.38202483719195e-06,
+      "loss": 0.6625,
+      "step": 11090
+    },
+    {
+      "epoch": 0.7250163291966035,
+      "grad_norm": 7.7791755086408925,
+      "learning_rate": 5.3581858547019095e-06,
+      "loss": 0.7008,
+      "step": 11100
+    },
+    {
+      "epoch": 0.7256694970607446,
+      "grad_norm": 3.286051116880111,
+      "learning_rate": 5.334388299752559e-06,
+      "loss": 0.6656,
+      "step": 11110
+    },
+    {
+      "epoch": 0.7263226649248857,
+      "grad_norm": 3.587098841551492,
+      "learning_rate": 5.310632274593751e-06,
+      "loss": 0.6798,
+      "step": 11120
+    },
+    {
+      "epoch": 0.7269758327890268,
+      "grad_norm": 7.054947046720567,
+      "learning_rate": 5.286917881296918e-06,
+      "loss": 0.7412,
+      "step": 11130
+    },
+    {
+      "epoch": 0.7276290006531678,
+      "grad_norm": 14.569440636615642,
+      "learning_rate": 5.263245221754604e-06,
+      "loss": 0.6485,
+      "step": 11140
+    },
+    {
+      "epoch": 0.728282168517309,
+      "grad_norm": 4.516233832158538,
+      "learning_rate": 5.239614397680038e-06,
+      "loss": 0.6619,
+      "step": 11150
+    },
+    {
+      "epoch": 0.72893533638145,
+      "grad_norm": 4.493184756832641,
+      "learning_rate": 5.216025510606698e-06,
+      "loss": 0.6696,
+      "step": 11160
+    },
+    {
+      "epoch": 0.7295885042455911,
+      "grad_norm": 5.640137712136578,
+      "learning_rate": 5.192478661887869e-06,
+      "loss": 0.681,
+      "step": 11170
+    },
+    {
+      "epoch": 0.7302416721097322,
+      "grad_norm": 6.6556384741462,
+      "learning_rate": 5.168973952696225e-06,
+      "loss": 0.7231,
+      "step": 11180
+    },
+    {
+      "epoch": 0.7308948399738733,
+      "grad_norm": 7.515936640823098,
+      "learning_rate": 5.1455114840233636e-06,
+      "loss": 0.673,
+      "step": 11190
+    },
+    {
+      "epoch": 0.7315480078380143,
+      "grad_norm": 3.5093368242859024,
+      "learning_rate": 5.122091356679405e-06,
+      "loss": 0.7244,
+      "step": 11200
+    },
+    {
+      "epoch": 0.7322011757021555,
+      "grad_norm": 15.219204185021983,
+      "learning_rate": 5.098713671292531e-06,
+      "loss": 0.6844,
+      "step": 11210
+    },
+    {
+      "epoch": 0.7328543435662965,
+      "grad_norm": 4.243075170782062,
+      "learning_rate": 5.075378528308577e-06,
+      "loss": 0.6838,
+      "step": 11220
+    },
+    {
+      "epoch": 0.7335075114304376,
+      "grad_norm": 5.773093578537657,
+      "learning_rate": 5.052086027990578e-06,
+      "loss": 0.6706,
+      "step": 11230
+    },
+    {
+      "epoch": 0.7341606792945787,
+      "grad_norm": 2.574491124656257,
+      "learning_rate": 5.028836270418352e-06,
+      "loss": 0.7104,
+      "step": 11240
+    },
+    {
+      "epoch": 0.7348138471587198,
+      "grad_norm": 3.6829042406272094,
+      "learning_rate": 5.005629355488066e-06,
+      "loss": 0.6982,
+      "step": 11250
+    },
+    {
+      "epoch": 0.7354670150228608,
+      "grad_norm": 8.464757675064218,
+      "learning_rate": 4.9824653829118015e-06,
+      "loss": 0.7169,
+      "step": 11260
+    },
+    {
+      "epoch": 0.736120182887002,
+      "grad_norm": 8.172113535446657,
+      "learning_rate": 4.959344452217148e-06,
+      "loss": 0.7048,
+      "step": 11270
+    },
+    {
+      "epoch": 0.736773350751143,
+      "grad_norm": 5.29669855324662,
+      "learning_rate": 4.936266662746737e-06,
+      "loss": 0.7039,
+      "step": 11280
+    },
+    {
+      "epoch": 0.7374265186152841,
+      "grad_norm": 3.060578035099832,
+      "learning_rate": 4.91323211365786e-06,
+      "loss": 0.6704,
+      "step": 11290
+    },
+    {
+      "epoch": 0.7380796864794252,
+      "grad_norm": 4.77220405088633,
+      "learning_rate": 4.890240903922002e-06,
+      "loss": 0.6563,
+      "step": 11300
+    },
+    {
+      "epoch": 0.7387328543435663,
+      "grad_norm": 5.885766735187112,
+      "learning_rate": 4.867293132324439e-06,
+      "loss": 0.6889,
+      "step": 11310
+    },
+    {
+      "epoch": 0.7393860222077073,
+      "grad_norm": 4.012360786090751,
+      "learning_rate": 4.8443888974638035e-06,
+      "loss": 0.6678,
+      "step": 11320
+    },
+    {
+      "epoch": 0.7400391900718485,
+      "grad_norm": 7.5356316067983915,
+      "learning_rate": 4.821528297751682e-06,
+      "loss": 0.7078,
+      "step": 11330
+    },
+    {
+      "epoch": 0.7406923579359895,
+      "grad_norm": 6.360271118911102,
+      "learning_rate": 4.798711431412161e-06,
+      "loss": 0.7228,
+      "step": 11340
+    },
+    {
+      "epoch": 0.7413455258001306,
+      "grad_norm": 7.053731961346386,
+      "learning_rate": 4.775938396481417e-06,
+      "loss": 0.6698,
+      "step": 11350
+    },
+    {
+      "epoch": 0.7419986936642717,
+      "grad_norm": 3.6864863150034894,
+      "learning_rate": 4.753209290807314e-06,
+      "loss": 0.6992,
+      "step": 11360
+    },
+    {
+      "epoch": 0.7426518615284128,
+      "grad_norm": 5.584661164553851,
+      "learning_rate": 4.730524212048951e-06,
+      "loss": 0.7142,
+      "step": 11370
+    },
+    {
+      "epoch": 0.7433050293925539,
+      "grad_norm": 3.6761825942391004,
+      "learning_rate": 4.7078832576762796e-06,
+      "loss": 0.697,
+      "step": 11380
+    },
+    {
+      "epoch": 0.743958197256695,
+      "grad_norm": 6.215721121189116,
+      "learning_rate": 4.685286524969629e-06,
+      "loss": 0.7307,
+      "step": 11390
+    },
+    {
+      "epoch": 0.744611365120836,
+      "grad_norm": 2.1010462470722593,
+      "learning_rate": 4.66273411101936e-06,
+      "loss": 0.721,
+      "step": 11400
+    },
+    {
+      "epoch": 0.7452645329849772,
+      "grad_norm": 2.9120407659276193,
+      "learning_rate": 4.640226112725385e-06,
+      "loss": 0.7089,
+      "step": 11410
+    },
+    {
+      "epoch": 0.7459177008491182,
+      "grad_norm": 3.369323477827884,
+      "learning_rate": 4.6177626267967995e-06,
+      "loss": 0.6807,
+      "step": 11420
+    },
+    {
+      "epoch": 0.7465708687132593,
+      "grad_norm": 6.767124353473997,
+      "learning_rate": 4.595343749751426e-06,
+      "loss": 0.6976,
+      "step": 11430
+    },
+    {
+      "epoch": 0.7472240365774004,
+      "grad_norm": 8.934916516358554,
+      "learning_rate": 4.5729695779154226e-06,
+      "loss": 0.6557,
+      "step": 11440
+    },
+    {
+      "epoch": 0.7478772044415415,
+      "grad_norm": 5.360081573264151,
+      "learning_rate": 4.550640207422877e-06,
+      "loss": 0.6952,
+      "step": 11450
+    },
+    {
+      "epoch": 0.7485303723056825,
+      "grad_norm": 9.06782995876813,
+      "learning_rate": 4.528355734215366e-06,
+      "loss": 0.7008,
+      "step": 11460
+    },
+    {
+      "epoch": 0.7491835401698237,
+      "grad_norm": 4.1971231598815235,
+      "learning_rate": 4.506116254041564e-06,
+      "loss": 0.6627,
+      "step": 11470
+    },
+    {
+      "epoch": 0.7498367080339647,
+      "grad_norm": 17.670004667271773,
+      "learning_rate": 4.483921862456819e-06,
+      "loss": 0.6703,
+      "step": 11480
+    },
+    {
+      "epoch": 0.7504898758981058,
+      "grad_norm": 7.1153667981089255,
+      "learning_rate": 4.4617726548227675e-06,
+      "loss": 0.7008,
+      "step": 11490
+    },
+    {
+      "epoch": 0.7511430437622469,
+      "grad_norm": 6.198148561728875,
+      "learning_rate": 4.439668726306884e-06,
+      "loss": 0.6873,
+      "step": 11500
+    },
+    {
+      "epoch": 0.751796211626388,
+      "grad_norm": 4.730018751820983,
+      "learning_rate": 4.417610171882114e-06,
+      "loss": 0.6713,
+      "step": 11510
+    },
+    {
+      "epoch": 0.752449379490529,
+      "grad_norm": 11.858758681696122,
+      "learning_rate": 4.395597086326432e-06,
+      "loss": 0.7241,
+      "step": 11520
+    },
+    {
+      "epoch": 0.7531025473546702,
+      "grad_norm": 2.5604194009971657,
+      "learning_rate": 4.373629564222452e-06,
+      "loss": 0.7211,
+      "step": 11530
+    },
+    {
+      "epoch": 0.7537557152188112,
+      "grad_norm": 6.127721376563953,
+      "learning_rate": 4.351707699957022e-06,
+      "loss": 0.6998,
+      "step": 11540
+    },
+    {
+      "epoch": 0.7544088830829523,
+      "grad_norm": 4.5795583726944695,
+      "learning_rate": 4.329831587720802e-06,
+      "loss": 0.7306,
+      "step": 11550
+    },
+    {
+      "epoch": 0.7550620509470934,
+      "grad_norm": 5.700473953084509,
+      "learning_rate": 4.308001321507894e-06,
+      "loss": 0.7182,
+      "step": 11560
+    },
+    {
+      "epoch": 0.7557152188112345,
+      "grad_norm": 8.045066259856924,
+      "learning_rate": 4.2862169951153876e-06,
+      "loss": 0.6862,
+      "step": 11570
+    },
+    {
+      "epoch": 0.7563683866753755,
+      "grad_norm": 30.07752840317419,
+      "learning_rate": 4.264478702143012e-06,
+      "loss": 0.7133,
+      "step": 11580
+    },
+    {
+      "epoch": 0.7570215545395167,
+      "grad_norm": 9.266849327574121,
+      "learning_rate": 4.242786535992684e-06,
+      "loss": 0.7105,
+      "step": 11590
+    },
+    {
+      "epoch": 0.7576747224036577,
+      "grad_norm": 6.445159799327492,
+      "learning_rate": 4.221140589868147e-06,
+      "loss": 0.6981,
+      "step": 11600
+    },
+    {
+      "epoch": 0.7583278902677988,
+      "grad_norm": 6.142205187903812,
+      "learning_rate": 4.199540956774541e-06,
+      "loss": 0.6832,
+      "step": 11610
+    },
+    {
+      "epoch": 0.75898105813194,
+      "grad_norm": 4.052417391594583,
+      "learning_rate": 4.177987729518021e-06,
+      "loss": 0.6577,
+      "step": 11620
+    },
+    {
+      "epoch": 0.759634225996081,
+      "grad_norm": 6.323134425360704,
+      "learning_rate": 4.156481000705346e-06,
+      "loss": 0.6844,
+      "step": 11630
+    },
+    {
+      "epoch": 0.760287393860222,
+      "grad_norm": 3.8227515050705354,
+      "learning_rate": 4.135020862743491e-06,
+      "loss": 0.6884,
+      "step": 11640
+    },
+    {
+      "epoch": 0.7609405617243632,
+      "grad_norm": 6.410463017777649,
+      "learning_rate": 4.113607407839253e-06,
+      "loss": 0.6684,
+      "step": 11650
+    },
+    {
+      "epoch": 0.7615937295885042,
+      "grad_norm": 4.813746699567568,
+      "learning_rate": 4.0922407279988335e-06,
+      "loss": 0.6729,
+      "step": 11660
+    },
+    {
+      "epoch": 0.7622468974526453,
+      "grad_norm": 3.2257376071290094,
+      "learning_rate": 4.070920915027476e-06,
+      "loss": 0.6794,
+      "step": 11670
+    },
+    {
+      "epoch": 0.7629000653167864,
+      "grad_norm": 2.888089279221371,
+      "learning_rate": 4.049648060529033e-06,
+      "loss": 0.6889,
+      "step": 11680
+    },
+    {
+      "epoch": 0.7635532331809275,
+      "grad_norm": 5.197354252520824,
+      "learning_rate": 4.028422255905616e-06,
+      "loss": 0.7189,
+      "step": 11690
+    },
+    {
+      "epoch": 0.7642064010450685,
+      "grad_norm": 4.325704069656428,
+      "learning_rate": 4.007243592357151e-06,
+      "loss": 0.6961,
+      "step": 11700
+    },
+    {
+      "epoch": 0.7648595689092097,
+      "grad_norm": 2.574492698992469,
+      "learning_rate": 3.98611216088104e-06,
+      "loss": 0.7196,
+      "step": 11710
+    },
+    {
+      "epoch": 0.7655127367733507,
+      "grad_norm": 15.270702380581223,
+      "learning_rate": 3.965028052271734e-06,
+      "loss": 0.6795,
+      "step": 11720
+    },
+    {
+      "epoch": 0.7661659046374918,
+      "grad_norm": 4.859452519917749,
+      "learning_rate": 3.94399135712035e-06,
+      "loss": 0.6708,
+      "step": 11730
+    },
+    {
+      "epoch": 0.766819072501633,
+      "grad_norm": 6.767605181222972,
+      "learning_rate": 3.923002165814301e-06,
+      "loss": 0.6824,
+      "step": 11740
+    },
+    {
+      "epoch": 0.767472240365774,
+      "grad_norm": 26.047999750760614,
+      "learning_rate": 3.902060568536873e-06,
+      "loss": 0.6562,
+      "step": 11750
+    },
+    {
+      "epoch": 0.768125408229915,
+      "grad_norm": 5.14331132911694,
+      "learning_rate": 3.881166655266879e-06,
+      "loss": 0.7051,
+      "step": 11760
+    },
+    {
+      "epoch": 0.7687785760940562,
+      "grad_norm": 3.0823853728057458,
+      "learning_rate": 3.860320515778224e-06,
+      "loss": 0.6721,
+      "step": 11770
+    },
+    {
+      "epoch": 0.7694317439581972,
+      "grad_norm": 4.808685772880641,
+      "learning_rate": 3.8395222396395685e-06,
+      "loss": 0.6915,
+      "step": 11780
+    },
+    {
+      "epoch": 0.7700849118223383,
+      "grad_norm": 5.018712314639996,
+      "learning_rate": 3.818771916213906e-06,
+      "loss": 0.6796,
+      "step": 11790
+    },
+    {
+      "epoch": 0.7707380796864794,
+      "grad_norm": 3.4740197644766804,
+      "learning_rate": 3.798069634658208e-06,
+      "loss": 0.6998,
+      "step": 11800
+    },
+    {
+      "epoch": 0.7713912475506205,
+      "grad_norm": 12.853249556821103,
+      "learning_rate": 3.7774154839230135e-06,
+      "loss": 0.6859,
+      "step": 11810
+    },
+    {
+      "epoch": 0.7720444154147615,
+      "grad_norm": 5.520376088143,
+      "learning_rate": 3.756809552752059e-06,
+      "loss": 0.7069,
+      "step": 11820
+    },
+    {
+      "epoch": 0.7726975832789027,
+      "grad_norm": 7.54059220839262,
+      "learning_rate": 3.736251929681914e-06,
+      "loss": 0.6804,
+      "step": 11830
+    },
+    {
+      "epoch": 0.7733507511430437,
+      "grad_norm": 5.109724027045723,
+      "learning_rate": 3.7157427030415714e-06,
+      "loss": 0.6751,
+      "step": 11840
+    },
+    {
+      "epoch": 0.7740039190071848,
+      "grad_norm": 7.887641085709507,
+      "learning_rate": 3.6952819609520826e-06,
+      "loss": 0.716,
+      "step": 11850
+    },
+    {
+      "epoch": 0.774657086871326,
+      "grad_norm": 4.475583913204593,
+      "learning_rate": 3.674869791326179e-06,
+      "loss": 0.7377,
+      "step": 11860
+    },
+    {
+      "epoch": 0.775310254735467,
+      "grad_norm": 10.331787568014837,
+      "learning_rate": 3.654506281867898e-06,
+      "loss": 0.7288,
+      "step": 11870
+    },
+    {
+      "epoch": 0.7759634225996082,
+      "grad_norm": 7.437261240997636,
+      "learning_rate": 3.634191520072191e-06,
+      "loss": 0.7145,
+      "step": 11880
+    },
+    {
+      "epoch": 0.7766165904637492,
+      "grad_norm": 3.740876734779163,
+      "learning_rate": 3.6139255932245707e-06,
+      "loss": 0.6993,
+      "step": 11890
+    },
+    {
+      "epoch": 0.7772697583278902,
+      "grad_norm": 7.71460714629273,
+      "learning_rate": 3.593708588400714e-06,
+      "loss": 0.7172,
+      "step": 11900
+    },
+    {
+      "epoch": 0.7779229261920314,
+      "grad_norm": 5.226329667132641,
+      "learning_rate": 3.5735405924660914e-06,
+      "loss": 0.6739,
+      "step": 11910
+    },
+    {
+      "epoch": 0.7785760940561725,
+      "grad_norm": 6.898912769850468,
+      "learning_rate": 3.5534216920756185e-06,
+      "loss": 0.6955,
+      "step": 11920
+    },
+    {
+      "epoch": 0.7792292619203135,
+      "grad_norm": 3.142973519932529,
+      "learning_rate": 3.533351973673238e-06,
+      "loss": 0.7032,
+      "step": 11930
+    },
+    {
+      "epoch": 0.7798824297844547,
+      "grad_norm": 3.9251708509836387,
+      "learning_rate": 3.5133315234915984e-06,
+      "loss": 0.6849,
+      "step": 11940
+    },
+    {
+      "epoch": 0.7805355976485957,
+      "grad_norm": 9.332540569242788,
+      "learning_rate": 3.493360427551643e-06,
+      "loss": 0.7404,
+      "step": 11950
+    },
+    {
+      "epoch": 0.7811887655127367,
+      "grad_norm": 7.326016966754033,
+      "learning_rate": 3.4734387716622724e-06,
+      "loss": 0.6932,
+      "step": 11960
+    },
+    {
+      "epoch": 0.7818419333768779,
+      "grad_norm": 8.101112816744926,
+      "learning_rate": 3.453566641419942e-06,
+      "loss": 0.6652,
+      "step": 11970
+    },
+    {
+      "epoch": 0.782495101241019,
+      "grad_norm": 43.286231944771565,
+      "learning_rate": 3.4337441222083316e-06,
+      "loss": 0.6901,
+      "step": 11980
+    },
+    {
+      "epoch": 0.78314826910516,
+      "grad_norm": 4.646261266777329,
+      "learning_rate": 3.4139712991979487e-06,
+      "loss": 0.6991,
+      "step": 11990
+    },
+    {
+      "epoch": 0.7838014369693012,
+      "grad_norm": 4.87134473654209,
+      "learning_rate": 3.3942482573457716e-06,
+      "loss": 0.7127,
+      "step": 12000
+    },
+    {
+      "epoch": 0.7844546048334422,
+      "grad_norm": 2.988937963991205,
+      "learning_rate": 3.374575081394891e-06,
+      "loss": 0.7015,
+      "step": 12010
+    },
+    {
+      "epoch": 0.7851077726975833,
+      "grad_norm": 4.634966859174349,
+      "learning_rate": 3.354951855874136e-06,
+      "loss": 0.6939,
+      "step": 12020
+    },
+    {
+      "epoch": 0.7857609405617244,
+      "grad_norm": 4.259983454210986,
+      "learning_rate": 3.33537866509773e-06,
+      "loss": 0.6993,
+      "step": 12030
+    },
+    {
+      "epoch": 0.7864141084258655,
+      "grad_norm": 3.421594338251067,
+      "learning_rate": 3.3158555931648915e-06,
+      "loss": 0.691,
+      "step": 12040
+    },
+    {
+      "epoch": 0.7870672762900065,
+      "grad_norm": 4.756566377761871,
+      "learning_rate": 3.296382723959521e-06,
+      "loss": 0.6625,
+      "step": 12050
+    },
+    {
+      "epoch": 0.7877204441541477,
+      "grad_norm": 6.204055063009476,
+      "learning_rate": 3.2769601411497917e-06,
+      "loss": 0.6749,
+      "step": 12060
+    },
+    {
+      "epoch": 0.7883736120182887,
+      "grad_norm": 4.677416114890872,
+      "learning_rate": 3.2575879281878387e-06,
+      "loss": 0.7238,
+      "step": 12070
+    },
+    {
+      "epoch": 0.7890267798824298,
+      "grad_norm": 7.883616018032296,
+      "learning_rate": 3.238266168309341e-06,
+      "loss": 0.6868,
+      "step": 12080
+    },
+    {
+      "epoch": 0.7896799477465709,
+      "grad_norm": 4.685982827389975,
+      "learning_rate": 3.218994944533235e-06,
+      "loss": 0.7452,
+      "step": 12090
+    },
+    {
+      "epoch": 0.790333115610712,
+      "grad_norm": 7.214355980990887,
+      "learning_rate": 3.199774339661299e-06,
+      "loss": 0.6869,
+      "step": 12100
+    },
+    {
+      "epoch": 0.790986283474853,
+      "grad_norm": 4.079648255912239,
+      "learning_rate": 3.1806044362778184e-06,
+      "loss": 0.6876,
+      "step": 12110
+    },
+    {
+      "epoch": 0.7916394513389942,
+      "grad_norm": 3.734472169769081,
+      "learning_rate": 3.161485316749248e-06,
+      "loss": 0.6945,
+      "step": 12120
+    },
+    {
+      "epoch": 0.7922926192031352,
+      "grad_norm": 2.452217663175586,
+      "learning_rate": 3.142417063223822e-06,
+      "loss": 0.7043,
+      "step": 12130
+    },
+    {
+      "epoch": 0.7929457870672763,
+      "grad_norm": 3.530193609704299,
+      "learning_rate": 3.1233997576312453e-06,
+      "loss": 0.6959,
+      "step": 12140
+    },
+    {
+      "epoch": 0.7935989549314174,
+      "grad_norm": 2.5668358755276413,
+      "learning_rate": 3.1044334816822856e-06,
+      "loss": 0.7136,
+      "step": 12150
+    },
+    {
+      "epoch": 0.7942521227955585,
+      "grad_norm": 4.931477990820825,
+      "learning_rate": 3.085518316868482e-06,
+      "loss": 0.6985,
+      "step": 12160
+    },
+    {
+      "epoch": 0.7949052906596995,
+      "grad_norm": 8.704483227583205,
+      "learning_rate": 3.06665434446175e-06,
+      "loss": 0.6761,
+      "step": 12170
+    },
+    {
+      "epoch": 0.7955584585238407,
+      "grad_norm": 20.461859228952505,
+      "learning_rate": 3.04784164551406e-06,
+      "loss": 0.7015,
+      "step": 12180
+    },
+    {
+      "epoch": 0.7962116263879817,
+      "grad_norm": 3.6639577366669425,
+      "learning_rate": 3.0290803008570716e-06,
+      "loss": 0.7254,
+      "step": 12190
+    },
+    {
+      "epoch": 0.7968647942521228,
+      "grad_norm": 8.385447348465478,
+      "learning_rate": 3.010370391101788e-06,
+      "loss": 0.6877,
+      "step": 12200
+    },
+    {
+      "epoch": 0.7975179621162639,
+      "grad_norm": 92.3924258343859,
+      "learning_rate": 2.9917119966382296e-06,
+      "loss": 0.681,
+      "step": 12210
+    },
+    {
+      "epoch": 0.798171129980405,
+      "grad_norm": 6.476743615060319,
+      "learning_rate": 2.9731051976350605e-06,
+      "loss": 0.7358,
+      "step": 12220
+    },
+    {
+      "epoch": 0.798824297844546,
+      "grad_norm": 6.862513630059609,
+      "learning_rate": 2.954550074039258e-06,
+      "loss": 0.7084,
+      "step": 12230
+    },
+    {
+      "epoch": 0.7994774657086872,
+      "grad_norm": 5.881076997488694,
+      "learning_rate": 2.93604670557577e-06,
+      "loss": 0.7179,
+      "step": 12240
+    },
+    {
+      "epoch": 0.8001306335728282,
+      "grad_norm": 2.7363017242268954,
+      "learning_rate": 2.917595171747178e-06,
+      "loss": 0.6615,
+      "step": 12250
+    },
+    {
+      "epoch": 0.8007838014369693,
+      "grad_norm": 13.487462756624321,
+      "learning_rate": 2.8991955518333353e-06,
+      "loss": 0.6833,
+      "step": 12260
+    },
+    {
+      "epoch": 0.8014369693011104,
+      "grad_norm": 4.4447295706736645,
+      "learning_rate": 2.8808479248910484e-06,
+      "loss": 0.7112,
+      "step": 12270
+    },
+    {
+      "epoch": 0.8020901371652515,
+      "grad_norm": 6.063889806264584,
+      "learning_rate": 2.862552369753725e-06,
+      "loss": 0.6724,
+      "step": 12280
+    },
+    {
+      "epoch": 0.8027433050293925,
+      "grad_norm": 6.6405287791742165,
+      "learning_rate": 2.8443089650310313e-06,
+      "loss": 0.7205,
+      "step": 12290
+    },
+    {
+      "epoch": 0.8033964728935337,
+      "grad_norm": 3.9419553466853516,
+      "learning_rate": 2.8261177891085803e-06,
+      "loss": 0.6971,
+      "step": 12300
+    },
+    {
+      "epoch": 0.8040496407576747,
+      "grad_norm": 4.354713052784848,
+      "learning_rate": 2.807978920147547e-06,
+      "loss": 0.6968,
+      "step": 12310
+    },
+    {
+      "epoch": 0.8047028086218158,
+      "grad_norm": 13.069194629609031,
+      "learning_rate": 2.789892436084393e-06,
+      "loss": 0.6771,
+      "step": 12320
+    },
+    {
+      "epoch": 0.8053559764859569,
+      "grad_norm": 8.064868194770664,
+      "learning_rate": 2.7718584146304727e-06,
+      "loss": 0.6947,
+      "step": 12330
+    },
+    {
+      "epoch": 0.806009144350098,
+      "grad_norm": 2.3428135336163436,
+      "learning_rate": 2.7538769332717486e-06,
+      "loss": 0.6939,
+      "step": 12340
+    },
+    {
+      "epoch": 0.806662312214239,
+      "grad_norm": 7.1717066526912445,
+      "learning_rate": 2.73594806926842e-06,
+      "loss": 0.6727,
+      "step": 12350
+    },
+    {
+      "epoch": 0.8073154800783802,
+      "grad_norm": 7.3523299073095165,
+      "learning_rate": 2.7180718996546223e-06,
+      "loss": 0.6875,
+      "step": 12360
+    },
+    {
+      "epoch": 0.8079686479425212,
+      "grad_norm": 3.342750043282877,
+      "learning_rate": 2.700248501238068e-06,
+      "loss": 0.6709,
+      "step": 12370
+    },
+    {
+      "epoch": 0.8086218158066623,
+      "grad_norm": 2.8621407562447256,
+      "learning_rate": 2.6824779505997387e-06,
+      "loss": 0.6844,
+      "step": 12380
+    },
+    {
+      "epoch": 0.8092749836708034,
+      "grad_norm": 5.293840737674243,
+      "learning_rate": 2.6647603240935416e-06,
+      "loss": 0.7063,
+      "step": 12390
+    },
+    {
+      "epoch": 0.8099281515349445,
+      "grad_norm": 2.758612843514663,
+      "learning_rate": 2.6470956978459894e-06,
+      "loss": 0.7293,
+      "step": 12400
+    },
+    {
+      "epoch": 0.8105813193990856,
+      "grad_norm": 3.4769473430781326,
+      "learning_rate": 2.6294841477558746e-06,
+      "loss": 0.6904,
+      "step": 12410
+    },
+    {
+      "epoch": 0.8112344872632267,
+      "grad_norm": 3.5772224187425494,
+      "learning_rate": 2.6119257494939338e-06,
+      "loss": 0.7102,
+      "step": 12420
+    },
+    {
+      "epoch": 0.8118876551273677,
+      "grad_norm": 49.35632558244181,
+      "learning_rate": 2.594420578502537e-06,
+      "loss": 0.6637,
+      "step": 12430
+    },
+    {
+      "epoch": 0.8125408229915089,
+      "grad_norm": 2.4670140404148997,
+      "learning_rate": 2.576968709995342e-06,
+      "loss": 0.6894,
+      "step": 12440
+    },
+    {
+      "epoch": 0.8131939908556499,
+      "grad_norm": 3.1293163245147513,
+      "learning_rate": 2.5595702189570034e-06,
+      "loss": 0.6928,
+      "step": 12450
+    },
+    {
+      "epoch": 0.813847158719791,
+      "grad_norm": 4.722752774421984,
+      "learning_rate": 2.542225180142807e-06,
+      "loss": 0.7045,
+      "step": 12460
+    },
+    {
+      "epoch": 0.8145003265839321,
+      "grad_norm": 4.282456499492312,
+      "learning_rate": 2.524933668078393e-06,
+      "loss": 0.6874,
+      "step": 12470
+    },
+    {
+      "epoch": 0.8151534944480732,
+      "grad_norm": 7.445379875334576,
+      "learning_rate": 2.507695757059406e-06,
+      "loss": 0.693,
+      "step": 12480
+    },
+    {
+      "epoch": 0.8158066623122142,
+      "grad_norm": 9.103161742545142,
+      "learning_rate": 2.490511521151187e-06,
+      "loss": 0.6902,
+      "step": 12490
+    },
+    {
+      "epoch": 0.8164598301763554,
+      "grad_norm": 23.66529921042094,
+      "learning_rate": 2.473381034188457e-06,
+      "loss": 0.6828,
+      "step": 12500
+    },
+    {
+      "epoch": 0.8171129980404964,
+      "grad_norm": 7.930931677716769,
+      "learning_rate": 2.45630436977499e-06,
+      "loss": 0.6911,
+      "step": 12510
+    },
+    {
+      "epoch": 0.8177661659046375,
+      "grad_norm": 6.202448810456874,
+      "learning_rate": 2.439281601283313e-06,
+      "loss": 0.7151,
+      "step": 12520
+    },
+    {
+      "epoch": 0.8184193337687786,
+      "grad_norm": 4.083723529675006,
+      "learning_rate": 2.4223128018543698e-06,
+      "loss": 0.6664,
+      "step": 12530
+    },
+    {
+      "epoch": 0.8190725016329197,
+      "grad_norm": 7.755204698823513,
+      "learning_rate": 2.4053980443972262e-06,
+      "loss": 0.7058,
+      "step": 12540
+    },
+    {
+      "epoch": 0.8197256694970607,
+      "grad_norm": 2.526437776845295,
+      "learning_rate": 2.388537401588738e-06,
+      "loss": 0.7069,
+      "step": 12550
+    },
+    {
+      "epoch": 0.8203788373612019,
+      "grad_norm": 7.229351499395132,
+      "learning_rate": 2.371730945873264e-06,
+      "loss": 0.6185,
+      "step": 12560
+    },
+    {
+      "epoch": 0.8210320052253429,
+      "grad_norm": 2.0170822198816527,
+      "learning_rate": 2.3549787494623277e-06,
+      "loss": 0.6377,
+      "step": 12570
+    },
+    {
+      "epoch": 0.821685173089484,
+      "grad_norm": 6.849124278873261,
+      "learning_rate": 2.3382808843343225e-06,
+      "loss": 0.6372,
+      "step": 12580
+    },
+    {
+      "epoch": 0.8223383409536251,
+      "grad_norm": 3.949714123967933,
+      "learning_rate": 2.321637422234203e-06,
+      "loss": 0.6714,
+      "step": 12590
+    },
+    {
+      "epoch": 0.8229915088177662,
+      "grad_norm": 5.466741225656826,
+      "learning_rate": 2.305048434673168e-06,
+      "loss": 0.6814,
+      "step": 12600
+    },
+    {
+      "epoch": 0.8236446766819072,
+      "grad_norm": 9.13517549226144,
+      "learning_rate": 2.2885139929283605e-06,
+      "loss": 0.6825,
+      "step": 12610
+    },
+    {
+      "epoch": 0.8242978445460484,
+      "grad_norm": 3.5636581412270987,
+      "learning_rate": 2.2720341680425514e-06,
+      "loss": 0.6953,
+      "step": 12620
+    },
+    {
+      "epoch": 0.8249510124101894,
+      "grad_norm": 5.3332834638807185,
+      "learning_rate": 2.255609030823859e-06,
+      "loss": 0.706,
+      "step": 12630
+    },
+    {
+      "epoch": 0.8256041802743305,
+      "grad_norm": 6.947948938053646,
+      "learning_rate": 2.239238651845409e-06,
+      "loss": 0.7044,
+      "step": 12640
+    },
+    {
+      "epoch": 0.8262573481384716,
+      "grad_norm": 10.389458566460489,
+      "learning_rate": 2.2229231014450648e-06,
+      "loss": 0.6701,
+      "step": 12650
+    },
+    {
+      "epoch": 0.8269105160026127,
+      "grad_norm": 4.448630973164709,
+      "learning_rate": 2.2066624497251005e-06,
+      "loss": 0.7235,
+      "step": 12660
+    },
+    {
+      "epoch": 0.8275636838667537,
+      "grad_norm": 7.655730311603453,
+      "learning_rate": 2.1904567665519086e-06,
+      "loss": 0.7614,
+      "step": 12670
+    },
+    {
+      "epoch": 0.8282168517308949,
+      "grad_norm": 10.563719188142853,
+      "learning_rate": 2.1743061215557148e-06,
+      "loss": 0.732,
+      "step": 12680
+    },
+    {
+      "epoch": 0.8288700195950359,
+      "grad_norm": 11.322455477804937,
+      "learning_rate": 2.1582105841302425e-06,
+      "loss": 0.6881,
+      "step": 12690
+    },
+    {
+      "epoch": 0.829523187459177,
+      "grad_norm": 10.512372674522478,
+      "learning_rate": 2.1421702234324587e-06,
+      "loss": 0.6815,
+      "step": 12700
+    },
+    {
+      "epoch": 0.8301763553233181,
+      "grad_norm": 4.34459453029744,
+      "learning_rate": 2.1261851083822383e-06,
+      "loss": 0.6837,
+      "step": 12710
+    },
+    {
+      "epoch": 0.8308295231874592,
+      "grad_norm": 15.14965475463287,
+      "learning_rate": 2.110255307662101e-06,
+      "loss": 0.6867,
+      "step": 12720
+    },
+    {
+      "epoch": 0.8314826910516002,
+      "grad_norm": 6.047047851615411,
+      "learning_rate": 2.094380889716881e-06,
+      "loss": 0.7015,
+      "step": 12730
+    },
+    {
+      "epoch": 0.8321358589157414,
+      "grad_norm": 3.228008989758837,
+      "learning_rate": 2.078561922753471e-06,
+      "loss": 0.6774,
+      "step": 12740
+    },
+    {
+      "epoch": 0.8327890267798824,
+      "grad_norm": 5.460843195940031,
+      "learning_rate": 2.062798474740496e-06,
+      "loss": 0.6863,
+      "step": 12750
+    },
+    {
+      "epoch": 0.8334421946440235,
+      "grad_norm": 29.725573858395695,
+      "learning_rate": 2.047090613408043e-06,
+      "loss": 0.6753,
+      "step": 12760
+    },
+    {
+      "epoch": 0.8340953625081646,
+      "grad_norm": 4.378649687928872,
+      "learning_rate": 2.0314384062473564e-06,
+      "loss": 0.7087,
+      "step": 12770
+    },
+    {
+      "epoch": 0.8347485303723057,
+      "grad_norm": 6.7044943809033395,
+      "learning_rate": 2.0158419205105545e-06,
+      "loss": 0.6614,
+      "step": 12780
+    },
+    {
+      "epoch": 0.8354016982364467,
+      "grad_norm": 16.866259310380592,
+      "learning_rate": 2.0003012232103496e-06,
+      "loss": 0.6765,
+      "step": 12790
+    },
+    {
+      "epoch": 0.8360548661005879,
+      "grad_norm": 4.040076241104312,
+      "learning_rate": 1.9848163811197375e-06,
+      "loss": 0.6687,
+      "step": 12800
+    },
+    {
+      "epoch": 0.8367080339647289,
+      "grad_norm": 3.751237410791482,
+      "learning_rate": 1.9693874607717334e-06,
+      "loss": 0.674,
+      "step": 12810
+    },
+    {
+      "epoch": 0.83736120182887,
+      "grad_norm": 6.810935024428175,
+      "learning_rate": 1.9540145284590656e-06,
+      "loss": 0.7296,
+      "step": 12820
+    },
+    {
+      "epoch": 0.8380143696930111,
+      "grad_norm": 22.30397259690559,
+      "learning_rate": 1.9386976502339195e-06,
+      "loss": 0.6802,
+      "step": 12830
+    },
+    {
+      "epoch": 0.8386675375571522,
+      "grad_norm": 9.640858454172971,
+      "learning_rate": 1.923436891907608e-06,
+      "loss": 0.697,
+      "step": 12840
+    },
+    {
+      "epoch": 0.8393207054212932,
+      "grad_norm": 5.1550539421271075,
+      "learning_rate": 1.9082323190503403e-06,
+      "loss": 0.6743,
+      "step": 12850
+    },
+    {
+      "epoch": 0.8399738732854344,
+      "grad_norm": 3.6143646760410846,
+      "learning_rate": 1.8930839969909075e-06,
+      "loss": 0.7415,
+      "step": 12860
+    },
+    {
+      "epoch": 0.8406270411495754,
+      "grad_norm": 5.464168525956508,
+      "learning_rate": 1.877991990816405e-06,
+      "loss": 0.6898,
+      "step": 12870
+    },
+    {
+      "epoch": 0.8412802090137165,
+      "grad_norm": 3.456997732226285,
+      "learning_rate": 1.8629563653719705e-06,
+      "loss": 0.6824,
+      "step": 12880
+    },
+    {
+      "epoch": 0.8419333768778576,
+      "grad_norm": 3.9004033393633533,
+      "learning_rate": 1.8479771852604805e-06,
+      "loss": 0.7254,
+      "step": 12890
+    },
+    {
+      "epoch": 0.8425865447419987,
+      "grad_norm": 14.7940208316794,
+      "learning_rate": 1.8330545148422966e-06,
+      "loss": 0.6697,
+      "step": 12900
+    },
+    {
+      "epoch": 0.8432397126061397,
+      "grad_norm": 5.865402823921409,
+      "learning_rate": 1.8181884182349707e-06,
+      "loss": 0.6707,
+      "step": 12910
+    },
+    {
+      "epoch": 0.8438928804702809,
+      "grad_norm": 5.3692230407006765,
+      "learning_rate": 1.8033789593129763e-06,
+      "loss": 0.7178,
+      "step": 12920
+    },
+    {
+      "epoch": 0.8445460483344219,
+      "grad_norm": 2.7141742127684214,
+      "learning_rate": 1.788626201707434e-06,
+      "loss": 0.6836,
+      "step": 12930
+    },
+    {
+      "epoch": 0.8451992161985631,
+      "grad_norm": 13.91576329266843,
+      "learning_rate": 1.773930208805849e-06,
+      "loss": 0.6617,
+      "step": 12940
+    },
+    {
+      "epoch": 0.8458523840627041,
+      "grad_norm": 6.069386768818718,
+      "learning_rate": 1.7592910437518134e-06,
+      "loss": 0.6836,
+      "step": 12950
+    },
+    {
+      "epoch": 0.8465055519268452,
+      "grad_norm": 6.304680844366065,
+      "learning_rate": 1.7447087694447577e-06,
+      "loss": 0.693,
+      "step": 12960
+    },
+    {
+      "epoch": 0.8471587197909863,
+      "grad_norm": 3.3827802202829838,
+      "learning_rate": 1.7301834485396733e-06,
+      "loss": 0.7326,
+      "step": 12970
+    },
+    {
+      "epoch": 0.8478118876551274,
+      "grad_norm": 3.4211838951874456,
+      "learning_rate": 1.7157151434468371e-06,
+      "loss": 0.7205,
+      "step": 12980
+    },
+    {
+      "epoch": 0.8484650555192684,
+      "grad_norm": 4.341322186081593,
+      "learning_rate": 1.7013039163315602e-06,
+      "loss": 0.6569,
+      "step": 12990
+    },
+    {
+      "epoch": 0.8491182233834096,
+      "grad_norm": 4.980096876435401,
+      "learning_rate": 1.6869498291138886e-06,
+      "loss": 0.6956,
+      "step": 13000
+    },
+    {
+      "epoch": 0.8497713912475506,
+      "grad_norm": 4.701793190597957,
+      "learning_rate": 1.6726529434683808e-06,
+      "loss": 0.6784,
+      "step": 13010
+    },
+    {
+      "epoch": 0.8504245591116917,
+      "grad_norm": 2.3121382971719107,
+      "learning_rate": 1.6584133208238023e-06,
+      "loss": 0.6594,
+      "step": 13020
+    },
+    {
+      "epoch": 0.8510777269758328,
+      "grad_norm": 4.607070398447369,
+      "learning_rate": 1.6442310223628936e-06,
+      "loss": 0.6819,
+      "step": 13030
+    },
+    {
+      "epoch": 0.8517308948399739,
+      "grad_norm": 5.658518973644752,
+      "learning_rate": 1.6301061090220825e-06,
+      "loss": 0.6855,
+      "step": 13040
+    },
+    {
+      "epoch": 0.8523840627041149,
+      "grad_norm": 3.675512970915112,
+      "learning_rate": 1.6160386414912354e-06,
+      "loss": 0.6815,
+      "step": 13050
+    },
+    {
+      "epoch": 0.8530372305682561,
+      "grad_norm": 3.722639598188754,
+      "learning_rate": 1.6020286802134027e-06,
+      "loss": 0.6526,
+      "step": 13060
+    },
+    {
+      "epoch": 0.8536903984323971,
+      "grad_norm": 6.002349711311793,
+      "learning_rate": 1.5880762853845294e-06,
+      "loss": 0.6802,
+      "step": 13070
+    },
+    {
+      "epoch": 0.8543435662965382,
+      "grad_norm": 20.86184076601306,
+      "learning_rate": 1.5741815169532398e-06,
+      "loss": 0.6974,
+      "step": 13080
+    },
+    {
+      "epoch": 0.8549967341606793,
+      "grad_norm": 3.7061664951431528,
+      "learning_rate": 1.560344434620543e-06,
+      "loss": 0.6771,
+      "step": 13090
+    },
+    {
+      "epoch": 0.8556499020248204,
+      "grad_norm": 3.2473742808850674,
+      "learning_rate": 1.5465650978396035e-06,
+      "loss": 0.6931,
+      "step": 13100
+    },
+    {
+      "epoch": 0.8563030698889614,
+      "grad_norm": 4.710037980625067,
+      "learning_rate": 1.5328435658154565e-06,
+      "loss": 0.694,
+      "step": 13110
+    },
+    {
+      "epoch": 0.8569562377531026,
+      "grad_norm": 6.6459503004091,
+      "learning_rate": 1.5191798975047889e-06,
+      "loss": 0.6468,
+      "step": 13120
+    },
+    {
+      "epoch": 0.8576094056172436,
+      "grad_norm": 24.670829765303676,
+      "learning_rate": 1.5055741516156519e-06,
+      "loss": 0.7096,
+      "step": 13130
+    },
+    {
+      "epoch": 0.8582625734813847,
+      "grad_norm": 4.199668625974335,
+      "learning_rate": 1.4920263866072314e-06,
+      "loss": 0.718,
+      "step": 13140
+    },
+    {
+      "epoch": 0.8589157413455258,
+      "grad_norm": 31.907476179303377,
+      "learning_rate": 1.4785366606895879e-06,
+      "loss": 0.6497,
+      "step": 13150
+    },
+    {
+      "epoch": 0.8595689092096669,
+      "grad_norm": 7.202527863416497,
+      "learning_rate": 1.4651050318234055e-06,
+      "loss": 0.7042,
+      "step": 13160
+    },
+    {
+      "epoch": 0.8602220770738079,
+      "grad_norm": 40.873267834092985,
+      "learning_rate": 1.451731557719752e-06,
+      "loss": 0.7122,
+      "step": 13170
+    },
+    {
+      "epoch": 0.8608752449379491,
+      "grad_norm": 5.618784882943198,
+      "learning_rate": 1.4384162958398166e-06,
+      "loss": 0.703,
+      "step": 13180
+    },
+    {
+      "epoch": 0.8615284128020901,
+      "grad_norm": 4.327833475699335,
+      "learning_rate": 1.4251593033946803e-06,
+      "loss": 0.6976,
+      "step": 13190
+    },
+    {
+      "epoch": 0.8621815806662312,
+      "grad_norm": 3.571729116055556,
+      "learning_rate": 1.4119606373450455e-06,
+      "loss": 0.6785,
+      "step": 13200
+    },
+    {
+      "epoch": 0.8628347485303723,
+      "grad_norm": 3.4649354047382745,
+      "learning_rate": 1.3988203544010292e-06,
+      "loss": 0.6709,
+      "step": 13210
+    },
+    {
+      "epoch": 0.8634879163945134,
+      "grad_norm": 7.301633145272006,
+      "learning_rate": 1.3857385110218668e-06,
+      "loss": 0.6924,
+      "step": 13220
+    },
+    {
+      "epoch": 0.8641410842586544,
+      "grad_norm": 3.4298172146330854,
+      "learning_rate": 1.3727151634157249e-06,
+      "loss": 0.7395,
+      "step": 13230
+    },
+    {
+      "epoch": 0.8647942521227956,
+      "grad_norm": 107.76265151943878,
+      "learning_rate": 1.3597503675394225e-06,
+      "loss": 0.6282,
+      "step": 13240
+    },
+    {
+      "epoch": 0.8654474199869366,
+      "grad_norm": 3.904019382647567,
+      "learning_rate": 1.3468441790981983e-06,
+      "loss": 0.7327,
+      "step": 13250
+    },
+    {
+      "epoch": 0.8661005878510777,
+      "grad_norm": 5.708366264006957,
+      "learning_rate": 1.3339966535454861e-06,
+      "loss": 0.6733,
+      "step": 13260
+    },
+    {
+      "epoch": 0.8667537557152188,
+      "grad_norm": 4.535509514665049,
+      "learning_rate": 1.321207846082656e-06,
+      "loss": 0.7116,
+      "step": 13270
+    },
+    {
+      "epoch": 0.8674069235793599,
+      "grad_norm": 11.624418510742403,
+      "learning_rate": 1.3084778116587948e-06,
+      "loss": 0.6779,
+      "step": 13280
+    },
+    {
+      "epoch": 0.8680600914435009,
+      "grad_norm": 4.095446751856266,
+      "learning_rate": 1.2958066049704564e-06,
+      "loss": 0.6732,
+      "step": 13290
+    },
+    {
+      "epoch": 0.8687132593076421,
+      "grad_norm": 3.6972034608317315,
+      "learning_rate": 1.2831942804614306e-06,
+      "loss": 0.7488,
+      "step": 13300
+    },
+    {
+      "epoch": 0.8693664271717831,
+      "grad_norm": 3.3586477512686637,
+      "learning_rate": 1.2706408923225138e-06,
+      "loss": 0.7328,
+      "step": 13310
+    },
+    {
+      "epoch": 0.8700195950359242,
+      "grad_norm": 4.6623427704591744,
+      "learning_rate": 1.2581464944912774e-06,
+      "loss": 0.7201,
+      "step": 13320
+    },
+    {
+      "epoch": 0.8706727629000653,
+      "grad_norm": 15.700944344857493,
+      "learning_rate": 1.245711140651825e-06,
+      "loss": 0.6937,
+      "step": 13330
+    },
+    {
+      "epoch": 0.8713259307642064,
+      "grad_norm": 4.451147450157866,
+      "learning_rate": 1.2333348842345687e-06,
+      "loss": 0.6852,
+      "step": 13340
+    },
+    {
+      "epoch": 0.8719790986283474,
+      "grad_norm": 2.738144548734267,
+      "learning_rate": 1.2210177784160064e-06,
+      "loss": 0.7138,
+      "step": 13350
+    },
+    {
+      "epoch": 0.8726322664924886,
+      "grad_norm": 2.8632800842532586,
+      "learning_rate": 1.2087598761184765e-06,
+      "loss": 0.6942,
+      "step": 13360
+    },
+    {
+      "epoch": 0.8732854343566296,
+      "grad_norm": 7.694040880817023,
+      "learning_rate": 1.1965612300099555e-06,
+      "loss": 0.7027,
+      "step": 13370
+    },
+    {
+      "epoch": 0.8739386022207707,
+      "grad_norm": 14.48342526973317,
+      "learning_rate": 1.1844218925037953e-06,
+      "loss": 0.6937,
+      "step": 13380
+    },
+    {
+      "epoch": 0.8745917700849118,
+      "grad_norm": 2.2731035649293525,
+      "learning_rate": 1.1723419157585386e-06,
+      "loss": 0.7297,
+      "step": 13390
+    },
+    {
+      "epoch": 0.8752449379490529,
+      "grad_norm": 5.450732911551555,
+      "learning_rate": 1.16032135167766e-06,
+      "loss": 0.6678,
+      "step": 13400
+    },
+    {
+      "epoch": 0.8758981058131939,
+      "grad_norm": 2.4707591898591983,
+      "learning_rate": 1.148360251909374e-06,
+      "loss": 0.7047,
+      "step": 13410
+    },
+    {
+      "epoch": 0.8765512736773351,
+      "grad_norm": 4.69178149194048,
+      "learning_rate": 1.1364586678463868e-06,
+      "loss": 0.6913,
+      "step": 13420
+    },
+    {
+      "epoch": 0.8772044415414761,
+      "grad_norm": 5.475472291406911,
+      "learning_rate": 1.1246166506256834e-06,
+      "loss": 0.7222,
+      "step": 13430
+    },
+    {
+      "epoch": 0.8778576094056172,
+      "grad_norm": 6.649352605756714,
+      "learning_rate": 1.1128342511283278e-06,
+      "loss": 0.6908,
+      "step": 13440
+    },
+    {
+      "epoch": 0.8785107772697583,
+      "grad_norm": 4.445498497797621,
+      "learning_rate": 1.1011115199792032e-06,
+      "loss": 0.6752,
+      "step": 13450
+    },
+    {
+      "epoch": 0.8791639451338994,
+      "grad_norm": 10.860958812672406,
+      "learning_rate": 1.0894485075468385e-06,
+      "loss": 0.6967,
+      "step": 13460
+    },
+    {
+      "epoch": 0.8798171129980406,
+      "grad_norm": 19.130271434463495,
+      "learning_rate": 1.0778452639431585e-06,
+      "loss": 0.6903,
+      "step": 13470
+    },
+    {
+      "epoch": 0.8804702808621816,
+      "grad_norm": 6.029789723135012,
+      "learning_rate": 1.0663018390232947e-06,
+      "loss": 0.6964,
+      "step": 13480
+    },
+    {
+      "epoch": 0.8811234487263226,
+      "grad_norm": 4.7575663485688935,
+      "learning_rate": 1.0548182823853463e-06,
+      "loss": 0.7022,
+      "step": 13490
+    },
+    {
+      "epoch": 0.8817766165904638,
+      "grad_norm": 5.457678175572029,
+      "learning_rate": 1.0433946433701896e-06,
+      "loss": 0.7079,
+      "step": 13500
+    },
+    {
+      "epoch": 0.8824297844546048,
+      "grad_norm": 7.560645262211808,
+      "learning_rate": 1.0320309710612469e-06,
+      "loss": 0.6716,
+      "step": 13510
+    },
+    {
+      "epoch": 0.8830829523187459,
+      "grad_norm": 12.568101839096094,
+      "learning_rate": 1.0207273142842899e-06,
+      "loss": 0.6721,
+      "step": 13520
+    },
+    {
+      "epoch": 0.883736120182887,
+      "grad_norm": 5.211823395284773,
+      "learning_rate": 1.00948372160722e-06,
+      "loss": 0.6953,
+      "step": 13530
+    },
+    {
+      "epoch": 0.8843892880470281,
+      "grad_norm": 6.709679162531643,
+      "learning_rate": 9.983002413398635e-07,
+      "loss": 0.694,
+      "step": 13540
+    },
+    {
+      "epoch": 0.8850424559111691,
+      "grad_norm": 9.576647297562653,
+      "learning_rate": 9.871769215337744e-07,
+      "loss": 0.6808,
+      "step": 13550
+    },
+    {
+      "epoch": 0.8856956237753103,
+      "grad_norm": 3.263554550975047,
+      "learning_rate": 9.76113809982006e-07,
+      "loss": 0.6822,
+      "step": 13560
+    },
+    {
+      "epoch": 0.8863487916394513,
+      "grad_norm": 2.721918474062616,
+      "learning_rate": 9.651109542189246e-07,
+      "loss": 0.6771,
+      "step": 13570
+    },
+    {
+      "epoch": 0.8870019595035924,
+      "grad_norm": 6.588269310751218,
+      "learning_rate": 9.541684015199937e-07,
+      "loss": 0.6984,
+      "step": 13580
+    },
+    {
+      "epoch": 0.8876551273677336,
+      "grad_norm": 4.5199986169188415,
+      "learning_rate": 9.432861989015806e-07,
+      "loss": 0.6777,
+      "step": 13590
+    },
+    {
+      "epoch": 0.8883082952318746,
+      "grad_norm": 4.463527161594297,
+      "learning_rate": 9.324643931207438e-07,
+      "loss": 0.6895,
+      "step": 13600
+    },
+    {
+      "epoch": 0.8889614630960156,
+      "grad_norm": 2.5564012291530616,
+      "learning_rate": 9.217030306750424e-07,
+      "loss": 0.6993,
+      "step": 13610
+    },
+    {
+      "epoch": 0.8896146309601568,
+      "grad_norm": 3.939372389518821,
+      "learning_rate": 9.110021578023265e-07,
+      "loss": 0.7517,
+      "step": 13620
+    },
+    {
+      "epoch": 0.8902677988242979,
+      "grad_norm": 3.3916239167939524,
+      "learning_rate": 9.003618204805458e-07,
+      "loss": 0.6961,
+      "step": 13630
+    },
+    {
+      "epoch": 0.8909209666884389,
+      "grad_norm": 3.676860784580647,
+      "learning_rate": 8.897820644275517e-07,
+      "loss": 0.6922,
+      "step": 13640
+    },
+    {
+      "epoch": 0.89157413455258,
+      "grad_norm": 84.63930247489932,
+      "learning_rate": 8.792629351008935e-07,
+      "loss": 0.7671,
+      "step": 13650
+    },
+    {
+      "epoch": 0.8922273024167211,
+      "grad_norm": 3.7924065060240566,
+      "learning_rate": 8.688044776976373e-07,
+      "loss": 0.7007,
+      "step": 13660
+    },
+    {
+      "epoch": 0.8928804702808621,
+      "grad_norm": 5.923617546949608,
+      "learning_rate": 8.584067371541543e-07,
+      "loss": 0.669,
+      "step": 13670
+    },
+    {
+      "epoch": 0.8935336381450033,
+      "grad_norm": 6.261930142057152,
+      "learning_rate": 8.480697581459379e-07,
+      "loss": 0.6811,
+      "step": 13680
+    },
+    {
+      "epoch": 0.8941868060091444,
+      "grad_norm": 4.144322540176693,
+      "learning_rate": 8.377935850874136e-07,
+      "loss": 0.6428,
+      "step": 13690
+    },
+    {
+      "epoch": 0.8948399738732854,
+      "grad_norm": 5.285507743069797,
+      "learning_rate": 8.275782621317424e-07,
+      "loss": 0.6897,
+      "step": 13700
+    },
+    {
+      "epoch": 0.8954931417374266,
+      "grad_norm": 5.880334907388088,
+      "learning_rate": 8.174238331706346e-07,
+      "loss": 0.7171,
+      "step": 13710
+    },
+    {
+      "epoch": 0.8961463096015676,
+      "grad_norm": 4.729785268155556,
+      "learning_rate": 8.073303418341582e-07,
+      "loss": 0.7273,
+      "step": 13720
+    },
+    {
+      "epoch": 0.8967994774657086,
+      "grad_norm": 4.18389236139603,
+      "learning_rate": 7.972978314905572e-07,
+      "loss": 0.6712,
+      "step": 13730
+    },
+    {
+      "epoch": 0.8974526453298498,
+      "grad_norm": 36.68213932833151,
+      "learning_rate": 7.873263452460533e-07,
+      "loss": 0.7055,
+      "step": 13740
+    },
+    {
+      "epoch": 0.8981058131939909,
+      "grad_norm": 6.203815873346086,
+      "learning_rate": 7.774159259446834e-07,
+      "loss": 0.7088,
+      "step": 13750
+    },
+    {
+      "epoch": 0.8987589810581319,
+      "grad_norm": 6.0828838792008675,
+      "learning_rate": 7.675666161680822e-07,
+      "loss": 0.6817,
+      "step": 13760
+    },
+    {
+      "epoch": 0.8994121489222731,
+      "grad_norm": 15.42583812742743,
+      "learning_rate": 7.577784582353314e-07,
+      "loss": 0.6922,
+      "step": 13770
+    },
+    {
+      "epoch": 0.9000653167864141,
+      "grad_norm": 5.40224506897626,
+      "learning_rate": 7.480514942027595e-07,
+      "loss": 0.6646,
+      "step": 13780
+    },
+    {
+      "epoch": 0.9007184846505552,
+      "grad_norm": 11.010252596168959,
+      "learning_rate": 7.383857658637699e-07,
+      "loss": 0.6413,
+      "step": 13790
+    },
+    {
+      "epoch": 0.9013716525146963,
+      "grad_norm": 10.280050004770693,
+      "learning_rate": 7.287813147486522e-07,
+      "loss": 0.6916,
+      "step": 13800
+    },
+    {
+      "epoch": 0.9020248203788374,
+      "grad_norm": 27.838447999335717,
+      "learning_rate": 7.192381821244076e-07,
+      "loss": 0.6672,
+      "step": 13810
+    },
+    {
+      "epoch": 0.9026779882429784,
+      "grad_norm": 3.5603793340741716,
+      "learning_rate": 7.097564089945819e-07,
+      "loss": 0.6974,
+      "step": 13820
+    },
+    {
+      "epoch": 0.9033311561071196,
+      "grad_norm": 1.8945309353857664,
+      "learning_rate": 7.003360360990713e-07,
+      "loss": 0.695,
+      "step": 13830
+    },
+    {
+      "epoch": 0.9039843239712606,
+      "grad_norm": 3.734134937694426,
+      "learning_rate": 6.909771039139618e-07,
+      "loss": 0.6432,
+      "step": 13840
+    },
+    {
+      "epoch": 0.9046374918354017,
+      "grad_norm": 8.454806289303644,
+      "learning_rate": 6.816796526513469e-07,
+      "loss": 0.6882,
+      "step": 13850
+    },
+    {
+      "epoch": 0.9052906596995428,
+      "grad_norm": 6.13037154100083,
+      "learning_rate": 6.724437222591601e-07,
+      "loss": 0.7562,
+      "step": 13860
+    },
+    {
+      "epoch": 0.9059438275636839,
+      "grad_norm": 37.35666127481901,
+      "learning_rate": 6.632693524209993e-07,
+      "loss": 0.693,
+      "step": 13870
+    },
+    {
+      "epoch": 0.9065969954278249,
+      "grad_norm": 2.5934276150658575,
+      "learning_rate": 6.541565825559608e-07,
+      "loss": 0.6659,
+      "step": 13880
+    },
+    {
+      "epoch": 0.9072501632919661,
+      "grad_norm": 4.4479991658024725,
+      "learning_rate": 6.451054518184613e-07,
+      "loss": 0.7033,
+      "step": 13890
+    },
+    {
+      "epoch": 0.9079033311561071,
+      "grad_norm": 3.9779535944376616,
+      "learning_rate": 6.361159990980836e-07,
+      "loss": 0.7081,
+      "step": 13900
+    },
+    {
+      "epoch": 0.9085564990202482,
+      "grad_norm": 10.652074806950573,
+      "learning_rate": 6.271882630193931e-07,
+      "loss": 0.7306,
+      "step": 13910
+    },
+    {
+      "epoch": 0.9092096668843893,
+      "grad_norm": 8.959014393548467,
+      "learning_rate": 6.183222819417822e-07,
+      "loss": 0.7099,
+      "step": 13920
+    },
+    {
+      "epoch": 0.9098628347485304,
+      "grad_norm": 7.508628205476429,
+      "learning_rate": 6.09518093959312e-07,
+      "loss": 0.6944,
+      "step": 13930
+    },
+    {
+      "epoch": 0.9105160026126714,
+      "grad_norm": 3.6530980274780824,
+      "learning_rate": 6.007757369005278e-07,
+      "loss": 0.6975,
+      "step": 13940
+    },
+    {
+      "epoch": 0.9111691704768126,
+      "grad_norm": 16.31255518878356,
+      "learning_rate": 5.920952483283159e-07,
+      "loss": 0.6187,
+      "step": 13950
+    },
+    {
+      "epoch": 0.9118223383409536,
+      "grad_norm": 3.3490188300490993,
+      "learning_rate": 5.834766655397334e-07,
+      "loss": 0.6923,
+      "step": 13960
+    },
+    {
+      "epoch": 0.9124755062050947,
+      "grad_norm": 5.306609893923255,
+      "learning_rate": 5.749200255658516e-07,
+      "loss": 0.6937,
+      "step": 13970
+    },
+    {
+      "epoch": 0.9131286740692358,
+      "grad_norm": 7.073860955629541,
+      "learning_rate": 5.664253651715917e-07,
+      "loss": 0.6991,
+      "step": 13980
+    },
+    {
+      "epoch": 0.9137818419333769,
+      "grad_norm": 8.503460348036263,
+      "learning_rate": 5.579927208555713e-07,
+      "loss": 0.7047,
+      "step": 13990
+    },
+    {
+      "epoch": 0.914435009797518,
+      "grad_norm": 8.502466629263433,
+      "learning_rate": 5.496221288499459e-07,
+      "loss": 0.6506,
+      "step": 14000
+    },
+    {
+      "epoch": 0.9150881776616591,
+      "grad_norm": 3.533571813440642,
+      "learning_rate": 5.413136251202544e-07,
+      "loss": 0.7007,
+      "step": 14010
+    },
+    {
+      "epoch": 0.9157413455258001,
+      "grad_norm": 4.645811939047689,
+      "learning_rate": 5.330672453652657e-07,
+      "loss": 0.7376,
+      "step": 14020
+    },
+    {
+      "epoch": 0.9163945133899413,
+      "grad_norm": 9.018121912413273,
+      "learning_rate": 5.248830250168174e-07,
+      "loss": 0.7021,
+      "step": 14030
+    },
+    {
+      "epoch": 0.9170476812540823,
+      "grad_norm": 2.049483638954049,
+      "learning_rate": 5.167609992396788e-07,
+      "loss": 0.691,
+      "step": 14040
+    },
+    {
+      "epoch": 0.9177008491182234,
+      "grad_norm": 3.587822578628003,
+      "learning_rate": 5.087012029313832e-07,
+      "loss": 0.7162,
+      "step": 14050
+    },
+    {
+      "epoch": 0.9183540169823645,
+      "grad_norm": 5.337187010546555,
+      "learning_rate": 5.007036707220874e-07,
+      "loss": 0.6885,
+      "step": 14060
+    },
+    {
+      "epoch": 0.9190071848465056,
+      "grad_norm": 4.327237365263435,
+      "learning_rate": 4.927684369744195e-07,
+      "loss": 0.7006,
+      "step": 14070
+    },
+    {
+      "epoch": 0.9196603527106466,
+      "grad_norm": 4.396522769226511,
+      "learning_rate": 4.848955357833396e-07,
+      "loss": 0.6899,
+      "step": 14080
+    },
+    {
+      "epoch": 0.9203135205747878,
+      "grad_norm": 5.83333667573379,
+      "learning_rate": 4.770850009759769e-07,
+      "loss": 0.7049,
+      "step": 14090
+    },
+    {
+      "epoch": 0.9209666884389288,
+      "grad_norm": 3.0477622280944137,
+      "learning_rate": 4.693368661114988e-07,
+      "loss": 0.7459,
+      "step": 14100
+    },
+    {
+      "epoch": 0.9216198563030699,
+      "grad_norm": 8.532140217261535,
+      "learning_rate": 4.6165116448096346e-07,
+      "loss": 0.6397,
+      "step": 14110
+    },
+    {
+      "epoch": 0.922273024167211,
+      "grad_norm": 10.03387663966036,
+      "learning_rate": 4.5402792910717026e-07,
+      "loss": 0.6632,
+      "step": 14120
+    },
+    {
+      "epoch": 0.9229261920313521,
+      "grad_norm": 7.781927975452916,
+      "learning_rate": 4.4646719274452685e-07,
+      "loss": 0.6627,
+      "step": 14130
+    },
+    {
+      "epoch": 0.9235793598954931,
+      "grad_norm": 4.884147473757335,
+      "learning_rate": 4.3896898787889885e-07,
+      "loss": 0.643,
+      "step": 14140
+    },
+    {
+      "epoch": 0.9242325277596343,
+      "grad_norm": 2.4835514646988126,
+      "learning_rate": 4.315333467274851e-07,
+      "loss": 0.7385,
+      "step": 14150
+    },
+    {
+      "epoch": 0.9248856956237753,
+      "grad_norm": 3.3527394609036048,
+      "learning_rate": 4.2416030123865634e-07,
+      "loss": 0.6989,
+      "step": 14160
+    },
+    {
+      "epoch": 0.9255388634879164,
+      "grad_norm": 6.4679488271034815,
+      "learning_rate": 4.1684988309184656e-07,
+      "loss": 0.6573,
+      "step": 14170
+    },
+    {
+      "epoch": 0.9261920313520575,
+      "grad_norm": 13.210624488086639,
+      "learning_rate": 4.0960212369739016e-07,
+      "loss": 0.6825,
+      "step": 14180
+    },
+    {
+      "epoch": 0.9268451992161986,
+      "grad_norm": 3.6521328951919405,
+      "learning_rate": 4.024170541964017e-07,
+      "loss": 0.6929,
+      "step": 14190
+    },
+    {
+      "epoch": 0.9274983670803396,
+      "grad_norm": 7.8804629314783465,
+      "learning_rate": 3.9529470546064315e-07,
+      "loss": 0.705,
+      "step": 14200
+    },
+    {
+      "epoch": 0.9281515349444808,
+      "grad_norm": 6.128700539787899,
+      "learning_rate": 3.8823510809238184e-07,
+      "loss": 0.6672,
+      "step": 14210
+    },
+    {
+      "epoch": 0.9288047028086218,
+      "grad_norm": 15.312114352605443,
+      "learning_rate": 3.8123829242426577e-07,
+      "loss": 0.6954,
+      "step": 14220
+    },
+    {
+      "epoch": 0.9294578706727629,
+      "grad_norm": 9.102448069176269,
+      "learning_rate": 3.743042885191922e-07,
+      "loss": 0.6824,
+      "step": 14230
+    },
+    {
+      "epoch": 0.930111038536904,
+      "grad_norm": 2.7931428014768844,
+      "learning_rate": 3.6743312617017745e-07,
+      "loss": 0.6554,
+      "step": 14240
+    },
+    {
+      "epoch": 0.9307642064010451,
+      "grad_norm": 2.903822346384219,
+      "learning_rate": 3.6062483490023056e-07,
+      "loss": 0.7012,
+      "step": 14250
+    },
+    {
+      "epoch": 0.9314173742651861,
+      "grad_norm": 11.335648094600034,
+      "learning_rate": 3.538794439622234e-07,
+      "loss": 0.6825,
+      "step": 14260
+    },
+    {
+      "epoch": 0.9320705421293273,
+      "grad_norm": 4.648415752969088,
+      "learning_rate": 3.471969823387705e-07,
+      "loss": 0.6567,
+      "step": 14270
+    },
+    {
+      "epoch": 0.9327237099934683,
+      "grad_norm": 5.642285892343845,
+      "learning_rate": 3.4057747874209457e-07,
+      "loss": 0.6932,
+      "step": 14280
+    },
+    {
+      "epoch": 0.9333768778576094,
+      "grad_norm": 2.1561865313490833,
+      "learning_rate": 3.340209616139145e-07,
+      "loss": 0.6706,
+      "step": 14290
+    },
+    {
+      "epoch": 0.9340300457217505,
+      "grad_norm": 4.556269114261926,
+      "learning_rate": 3.2752745912531743e-07,
+      "loss": 0.6885,
+      "step": 14300
+    },
+    {
+      "epoch": 0.9346832135858916,
+      "grad_norm": 3.377248791911224,
+      "learning_rate": 3.2109699917663713e-07,
+      "loss": 0.6943,
+      "step": 14310
+    },
+    {
+      "epoch": 0.9353363814500326,
+      "grad_norm": 4.49289672560348,
+      "learning_rate": 3.1472960939733566e-07,
+      "loss": 0.6961,
+      "step": 14320
+    },
+    {
+      "epoch": 0.9359895493141738,
+      "grad_norm": 5.042183316243519,
+      "learning_rate": 3.0842531714588673e-07,
+      "loss": 0.6841,
+      "step": 14330
+    },
+    {
+      "epoch": 0.9366427171783148,
+      "grad_norm": 8.043654543674085,
+      "learning_rate": 3.0218414950964944e-07,
+      "loss": 0.6923,
+      "step": 14340
+    },
+    {
+      "epoch": 0.9372958850424559,
+      "grad_norm": 4.451546800933181,
+      "learning_rate": 2.9600613330476814e-07,
+      "loss": 0.6797,
+      "step": 14350
+    },
+    {
+      "epoch": 0.937949052906597,
+      "grad_norm": 3.1482886191633104,
+      "learning_rate": 2.8989129507603904e-07,
+      "loss": 0.6939,
+      "step": 14360
+    },
+    {
+      "epoch": 0.9386022207707381,
+      "grad_norm": 3.7098862614638266,
+      "learning_rate": 2.8383966109680747e-07,
+      "loss": 0.6863,
+      "step": 14370
+    },
+    {
+      "epoch": 0.9392553886348791,
+      "grad_norm": 5.3282925308156,
+      "learning_rate": 2.778512573688491e-07,
+      "loss": 0.6897,
+      "step": 14380
+    },
+    {
+      "epoch": 0.9399085564990203,
+      "grad_norm": 4.219253623006551,
+      "learning_rate": 2.719261096222669e-07,
+      "loss": 0.7354,
+      "step": 14390
+    },
+    {
+      "epoch": 0.9405617243631613,
+      "grad_norm": 3.929316539205132,
+      "learning_rate": 2.660642433153698e-07,
+      "loss": 0.6779,
+      "step": 14400
+    },
+    {
+      "epoch": 0.9412148922273024,
+      "grad_norm": 5.935845074156229,
+      "learning_rate": 2.602656836345707e-07,
+      "loss": 0.7155,
+      "step": 14410
+    },
+    {
+      "epoch": 0.9418680600914435,
+      "grad_norm": 17.773980074135935,
+      "learning_rate": 2.545304554942751e-07,
+      "loss": 0.7328,
+      "step": 14420
+    },
+    {
+      "epoch": 0.9425212279555846,
+      "grad_norm": 7.140119338547583,
+      "learning_rate": 2.4885858353677295e-07,
+      "loss": 0.6906,
+      "step": 14430
+    },
+    {
+      "epoch": 0.9431743958197256,
+      "grad_norm": 3.561871657255324,
+      "learning_rate": 2.4325009213214177e-07,
+      "loss": 0.696,
+      "step": 14440
+    },
+    {
+      "epoch": 0.9438275636838668,
+      "grad_norm": 9.022172530381402,
+      "learning_rate": 2.3770500537812211e-07,
+      "loss": 0.6945,
+      "step": 14450
+    },
+    {
+      "epoch": 0.9444807315480078,
+      "grad_norm": 2.9858760553941734,
+      "learning_rate": 2.32223347100039e-07,
+      "loss": 0.7145,
+      "step": 14460
+    },
+    {
+      "epoch": 0.9451338994121489,
+      "grad_norm": 10.192499162843738,
+      "learning_rate": 2.2680514085068049e-07,
+      "loss": 0.6805,
+      "step": 14470
+    },
+    {
+      "epoch": 0.94578706727629,
+      "grad_norm": 2.867441607864355,
+      "learning_rate": 2.214504099102044e-07,
+      "loss": 0.6791,
+      "step": 14480
+    },
+    {
+      "epoch": 0.9464402351404311,
+      "grad_norm": 3.496606271353713,
+      "learning_rate": 2.161591772860383e-07,
+      "loss": 0.6719,
+      "step": 14490
+    },
+    {
+      "epoch": 0.9470934030045721,
+      "grad_norm": 9.951317352040162,
+      "learning_rate": 2.109314657127781e-07,
+      "loss": 0.7256,
+      "step": 14500
+    },
+    {
+      "epoch": 0.9477465708687133,
+      "grad_norm": 3.795601588076029,
+      "learning_rate": 2.0576729765209468e-07,
+      "loss": 0.6505,
+      "step": 14510
+    },
+    {
+      "epoch": 0.9483997387328543,
+      "grad_norm": 3.0206069510686486,
+      "learning_rate": 2.0066669529262726e-07,
+      "loss": 0.7397,
+      "step": 14520
+    },
+    {
+      "epoch": 0.9490529065969955,
+      "grad_norm": 4.714697506081651,
+      "learning_rate": 1.9562968054990693e-07,
+      "loss": 0.6803,
+      "step": 14530
+    },
+    {
+      "epoch": 0.9497060744611365,
+      "grad_norm": 3.9588049089040362,
+      "learning_rate": 1.9065627506623663e-07,
+      "loss": 0.6761,
+      "step": 14540
+    },
+    {
+      "epoch": 0.9503592423252776,
+      "grad_norm": 2.6319925633183385,
+      "learning_rate": 1.8574650021062622e-07,
+      "loss": 0.6709,
+      "step": 14550
+    },
+    {
+      "epoch": 0.9510124101894187,
+      "grad_norm": 7.501251879162482,
+      "learning_rate": 1.8090037707867602e-07,
+      "loss": 0.697,
+      "step": 14560
+    },
+    {
+      "epoch": 0.9516655780535598,
+      "grad_norm": 11.39937787951777,
+      "learning_rate": 1.7611792649250168e-07,
+      "loss": 0.7192,
+      "step": 14570
+    },
+    {
+      "epoch": 0.9523187459177008,
+      "grad_norm": 26.26072372101657,
+      "learning_rate": 1.7139916900064111e-07,
+      "loss": 0.6729,
+      "step": 14580
+    },
+    {
+      "epoch": 0.952971913781842,
+      "grad_norm": 22.837025290547228,
+      "learning_rate": 1.6674412487796109e-07,
+      "loss": 0.6705,
+      "step": 14590
+    },
+    {
+      "epoch": 0.953625081645983,
+      "grad_norm": 17.314522635400518,
+      "learning_rate": 1.6215281412557737e-07,
+      "loss": 0.7002,
+      "step": 14600
+    },
+    {
+      "epoch": 0.9542782495101241,
+      "grad_norm": 5.138758914896164,
+      "learning_rate": 1.5762525647076308e-07,
+      "loss": 0.6723,
+      "step": 14610
+    },
+    {
+      "epoch": 0.9549314173742652,
+      "grad_norm": 11.749498072515893,
+      "learning_rate": 1.5316147136687053e-07,
+      "loss": 0.7342,
+      "step": 14620
+    },
+    {
+      "epoch": 0.9555845852384063,
+      "grad_norm": 9.208858239128203,
+      "learning_rate": 1.4876147799323613e-07,
+      "loss": 0.6792,
+      "step": 14630
+    },
+    {
+      "epoch": 0.9562377531025473,
+      "grad_norm": 2.8501151814525647,
+      "learning_rate": 1.4442529525511395e-07,
+      "loss": 0.6651,
+      "step": 14640
+    },
+    {
+      "epoch": 0.9568909209666885,
+      "grad_norm": 3.7411306397999913,
+      "learning_rate": 1.4015294178357895e-07,
+      "loss": 0.6937,
+      "step": 14650
+    },
+    {
+      "epoch": 0.9575440888308295,
+      "grad_norm": 17.34969017785557,
+      "learning_rate": 1.3594443593545724e-07,
+      "loss": 0.6594,
+      "step": 14660
+    },
+    {
+      "epoch": 0.9581972566949706,
+      "grad_norm": 5.181828903418344,
+      "learning_rate": 1.3179979579324265e-07,
+      "loss": 0.6961,
+      "step": 14670
+    },
+    {
+      "epoch": 0.9588504245591117,
+      "grad_norm": 3.4066746998174007,
+      "learning_rate": 1.2771903916502014e-07,
+      "loss": 0.6623,
+      "step": 14680
+    },
+    {
+      "epoch": 0.9595035924232528,
+      "grad_norm": 5.360315357154784,
+      "learning_rate": 1.23702183584391e-07,
+      "loss": 0.6627,
+      "step": 14690
+    },
+    {
+      "epoch": 0.9601567602873938,
+      "grad_norm": 3.624361179231387,
+      "learning_rate": 1.1974924631039108e-07,
+      "loss": 0.6754,
+      "step": 14700
+    },
+    {
+      "epoch": 0.960809928151535,
+      "grad_norm": 6.004317002153827,
+      "learning_rate": 1.1586024432742759e-07,
+      "loss": 0.6712,
+      "step": 14710
+    },
+    {
+      "epoch": 0.961463096015676,
+      "grad_norm": 3.1624037173846915,
+      "learning_rate": 1.1203519434519582e-07,
+      "loss": 0.6449,
+      "step": 14720
+    },
+    {
+      "epoch": 0.9621162638798171,
+      "grad_norm": 12.210848778068598,
+      "learning_rate": 1.082741127986142e-07,
+      "loss": 0.6341,
+      "step": 14730
+    },
+    {
+      "epoch": 0.9627694317439582,
+      "grad_norm": 4.19471051605192,
+      "learning_rate": 1.0457701584774936e-07,
+      "loss": 0.6853,
+      "step": 14740
+    },
+    {
+      "epoch": 0.9634225996080993,
+      "grad_norm": 7.093214107218928,
+      "learning_rate": 1.0094391937774617e-07,
+      "loss": 0.7483,
+      "step": 14750
+    },
+    {
+      "epoch": 0.9640757674722403,
+      "grad_norm": 45.34717692511988,
+      "learning_rate": 9.737483899876443e-08,
+      "loss": 0.682,
+      "step": 14760
+    },
+    {
+      "epoch": 0.9647289353363815,
+      "grad_norm": 3.034575064258203,
+      "learning_rate": 9.386979004590734e-08,
+      "loss": 0.7301,
+      "step": 14770
+    },
+    {
+      "epoch": 0.9653821032005225,
+      "grad_norm": 7.736430440263213,
+      "learning_rate": 9.042878757915984e-08,
+      "loss": 0.6526,
+      "step": 14780
+    },
+    {
+      "epoch": 0.9660352710646636,
+      "grad_norm": 6.196227728838487,
+      "learning_rate": 8.705184638331698e-08,
+      "loss": 0.6768,
+      "step": 14790
+    },
+    {
+      "epoch": 0.9666884389288047,
+      "grad_norm": 3.6317626748795067,
+      "learning_rate": 8.373898096793065e-08,
+      "loss": 0.6733,
+      "step": 14800
+    },
+    {
+      "epoch": 0.9673416067929458,
+      "grad_norm": 21.752350077330007,
+      "learning_rate": 8.049020556723464e-08,
+      "loss": 0.6687,
+      "step": 14810
+    },
+    {
+      "epoch": 0.9679947746570868,
+      "grad_norm": 2.414092721101793,
+      "learning_rate": 7.730553414009466e-08,
+      "loss": 0.6846,
+      "step": 14820
+    },
+    {
+      "epoch": 0.968647942521228,
+      "grad_norm": 6.9056862993753825,
+      "learning_rate": 7.418498036994182e-08,
+      "loss": 0.7105,
+      "step": 14830
+    },
+    {
+      "epoch": 0.969301110385369,
+      "grad_norm": 20.5521136085634,
+      "learning_rate": 7.112855766471749e-08,
+      "loss": 0.6715,
+      "step": 14840
+    },
+    {
+      "epoch": 0.9699542782495101,
+      "grad_norm": 6.202034573357447,
+      "learning_rate": 6.813627915681186e-08,
+      "loss": 0.6898,
+      "step": 14850
+    },
+    {
+      "epoch": 0.9706074461136512,
+      "grad_norm": 5.266374212104467,
+      "learning_rate": 6.520815770301058e-08,
+      "loss": 0.7049,
+      "step": 14860
+    },
+    {
+      "epoch": 0.9712606139777923,
+      "grad_norm": 8.003394360461222,
+      "learning_rate": 6.234420588443978e-08,
+      "loss": 0.6853,
+      "step": 14870
+    },
+    {
+      "epoch": 0.9719137818419333,
+      "grad_norm": 21.539384830513455,
+      "learning_rate": 5.954443600650783e-08,
+      "loss": 0.6738,
+      "step": 14880
+    },
+    {
+      "epoch": 0.9725669497060745,
+      "grad_norm": 13.756086077616871,
+      "learning_rate": 5.680886009886199e-08,
+      "loss": 0.7059,
+      "step": 14890
+    },
+    {
+      "epoch": 0.9732201175702155,
+      "grad_norm": 4.776557963789963,
+      "learning_rate": 5.413748991532019e-08,
+      "loss": 0.6703,
+      "step": 14900
+    },
+    {
+      "epoch": 0.9738732854343566,
+      "grad_norm": 5.150915542256727,
+      "learning_rate": 5.153033693384101e-08,
+      "loss": 0.6762,
+      "step": 14910
+    },
+    {
+      "epoch": 0.9745264532984977,
+      "grad_norm": 4.569623220258107,
+      "learning_rate": 4.898741235645543e-08,
+      "loss": 0.711,
+      "step": 14920
+    },
+    {
+      "epoch": 0.9751796211626388,
+      "grad_norm": 2.9152966869512986,
+      "learning_rate": 4.650872710923349e-08,
+      "loss": 0.6874,
+      "step": 14930
+    },
+    {
+      "epoch": 0.9758327890267798,
+      "grad_norm": 4.391462470174728,
+      "learning_rate": 4.4094291842227684e-08,
+      "loss": 0.663,
+      "step": 14940
+    },
+    {
+      "epoch": 0.976485956890921,
+      "grad_norm": 4.087919553170458,
+      "learning_rate": 4.174411692943136e-08,
+      "loss": 0.677,
+      "step": 14950
+    },
+    {
+      "epoch": 0.977139124755062,
+      "grad_norm": 2.1554312879370157,
+      "learning_rate": 3.945821246873205e-08,
+      "loss": 0.6425,
+      "step": 14960
+    },
+    {
+      "epoch": 0.9777922926192031,
+      "grad_norm": 9.7690536949704,
+      "learning_rate": 3.723658828187149e-08,
+      "loss": 0.7033,
+      "step": 14970
+    },
+    {
+      "epoch": 0.9784454604833442,
+      "grad_norm": 6.956938857033174,
+      "learning_rate": 3.50792539144007e-08,
+      "loss": 0.643,
+      "step": 14980
+    },
+    {
+      "epoch": 0.9790986283474853,
+      "grad_norm": 18.59785275114106,
+      "learning_rate": 3.298621863564e-08,
+      "loss": 0.6852,
+      "step": 14990
+    },
+    {
+      "epoch": 0.9797517962116263,
+      "grad_norm": 6.774665694768924,
+      "learning_rate": 3.095749143863735e-08,
+      "loss": 0.6704,
+      "step": 15000
+    },
+    {
+      "epoch": 0.9804049640757675,
+      "grad_norm": 6.242109276731404,
+      "learning_rate": 2.8993081040130098e-08,
+      "loss": 0.7305,
+      "step": 15010
+    },
+    {
+      "epoch": 0.9810581319399085,
+      "grad_norm": 6.9854383235254,
+      "learning_rate": 2.7092995880513283e-08,
+      "loss": 0.7011,
+      "step": 15020
+    },
+    {
+      "epoch": 0.9817112998040496,
+      "grad_norm": 2.284336666928396,
+      "learning_rate": 2.525724412379471e-08,
+      "loss": 0.7013,
+      "step": 15030
+    },
+    {
+      "epoch": 0.9823644676681907,
+      "grad_norm": 25.699388263709825,
+      "learning_rate": 2.3485833657563293e-08,
+      "loss": 0.6755,
+      "step": 15040
+    },
+    {
+      "epoch": 0.9830176355323318,
+      "grad_norm": 6.009184166094514,
+      "learning_rate": 2.1778772092959086e-08,
+      "loss": 0.7308,
+      "step": 15050
+    },
+    {
+      "epoch": 0.983670803396473,
+      "grad_norm": 11.053594248061634,
+      "learning_rate": 2.013606676463331e-08,
+      "loss": 0.7425,
+      "step": 15060
+    },
+    {
+      "epoch": 0.984323971260614,
+      "grad_norm": 3.607102772922988,
+      "learning_rate": 1.8557724730725035e-08,
+      "loss": 0.6991,
+      "step": 15070
+    },
+    {
+      "epoch": 0.984977139124755,
+      "grad_norm": 3.205612333365301,
+      "learning_rate": 1.7043752772822886e-08,
+      "loss": 0.6464,
+      "step": 15080
+    },
+    {
+      "epoch": 0.9856303069888962,
+      "grad_norm": 5.556643714300547,
+      "learning_rate": 1.5594157395940056e-08,
+      "loss": 0.7331,
+      "step": 15090
+    },
+    {
+      "epoch": 0.9862834748530372,
+      "grad_norm": 3.5464258119596814,
+      "learning_rate": 1.4208944828486003e-08,
+      "loss": 0.7062,
+      "step": 15100
+    },
+    {
+      "epoch": 0.9869366427171783,
+      "grad_norm": 3.722206624948899,
+      "learning_rate": 1.2888121022243126e-08,
+      "loss": 0.6512,
+      "step": 15110
+    },
+    {
+      "epoch": 0.9875898105813194,
+      "grad_norm": 2.2409677410944533,
+      "learning_rate": 1.16316916523318e-08,
+      "loss": 0.7349,
+      "step": 15120
+    },
+    {
+      "epoch": 0.9882429784454605,
+      "grad_norm": 3.878672305578926,
+      "learning_rate": 1.043966211719538e-08,
+      "loss": 0.7203,
+      "step": 15130
+    },
+    {
+      "epoch": 0.9888961463096015,
+      "grad_norm": 5.454242056298144,
+      "learning_rate": 9.312037538571905e-09,
+      "loss": 0.717,
+      "step": 15140
+    },
+    {
+      "epoch": 0.9895493141737427,
+      "grad_norm": 5.726823288208327,
+      "learning_rate": 8.24882276147576e-09,
+      "loss": 0.7076,
+      "step": 15150
+    },
+    {
+      "epoch": 0.9902024820378837,
+      "grad_norm": 14.755082628670916,
+      "learning_rate": 7.250022354171048e-09,
+      "loss": 0.6764,
+      "step": 15160
+    },
+    {
+      "epoch": 0.9908556499020248,
+      "grad_norm": 3.6312013396320113,
+      "learning_rate": 6.315640608158257e-09,
+      "loss": 0.669,
+      "step": 15170
+    },
+    {
+      "epoch": 0.991508817766166,
+      "grad_norm": 4.409239275794741,
+      "learning_rate": 5.445681538154279e-09,
+      "loss": 0.6649,
+      "step": 15180
+    },
+    {
+      "epoch": 0.992161985630307,
+      "grad_norm": 5.127595755856114,
+      "learning_rate": 4.640148882069095e-09,
+      "loss": 0.6749,
+      "step": 15190
+    },
+    {
+      "epoch": 0.992815153494448,
+      "grad_norm": 4.635445835436082,
+      "learning_rate": 3.899046101000781e-09,
+      "loss": 0.7167,
+      "step": 15200
+    },
+    {
+      "epoch": 0.9934683213585892,
+      "grad_norm": 7.473855033630282,
+      "learning_rate": 3.2223763792121884e-09,
+      "loss": 0.7007,
+      "step": 15210
+    },
+    {
+      "epoch": 0.9941214892227302,
+      "grad_norm": 2.7252881588946205,
+      "learning_rate": 2.610142624115963e-09,
+      "loss": 0.7108,
+      "step": 15220
+    },
+    {
+      "epoch": 0.9947746570868713,
+      "grad_norm": 13.861173872125017,
+      "learning_rate": 2.0623474662712085e-09,
+      "loss": 0.6722,
+      "step": 15230
+    },
+    {
+      "epoch": 0.9954278249510125,
+      "grad_norm": 6.165396310361432,
+      "learning_rate": 1.5789932593635037e-09,
+      "loss": 0.6937,
+      "step": 15240
+    },
+    {
+      "epoch": 0.9960809928151535,
+      "grad_norm": 2.3860816417389623,
+      "learning_rate": 1.1600820801982437e-09,
+      "loss": 0.6757,
+      "step": 15250
+    },
+    {
+      "epoch": 0.9967341606792945,
+      "grad_norm": 6.100593050284992,
+      "learning_rate": 8.056157286923104e-10,
+      "loss": 0.6631,
+      "step": 15260
+    },
+    {
+      "epoch": 0.9973873285434357,
+      "grad_norm": 21.92836751794244,
+      "learning_rate": 5.155957278657475e-10,
+      "loss": 0.7129,
+      "step": 15270
+    },
+    {
+      "epoch": 0.9980404964075767,
+      "grad_norm": 3.9263976263245834,
+      "learning_rate": 2.900233238334327e-10,
+      "loss": 0.6716,
+      "step": 15280
+    },
+    {
+      "epoch": 0.9986936642717178,
+      "grad_norm": 8.924314362505408,
+      "learning_rate": 1.2889948580174783e-10,
+      "loss": 0.7055,
+      "step": 15290
+    },
+    {
+      "epoch": 0.999346832135859,
+      "grad_norm": 15.640463539532494,
+      "learning_rate": 3.222490606524797e-11,
+      "loss": 0.7206,
+      "step": 15300
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 3.167310766334676,
+      "learning_rate": 0.0,
+      "loss": 0.6538,
+      "step": 15310
+    },
+    {
+      "epoch": 1.0,
+      "step": 15310,
+      "total_flos": 4.185096573664887e+19,
+      "train_loss": 0.7165804752721108,
+      "train_runtime": 78866.3439,
+      "train_samples_per_second": 12.424,
+      "train_steps_per_second": 0.194
+    }
+  ],
+  "logging_steps": 10,
+  "max_steps": 15310,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.185096573664887e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}