diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,43987 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 6279,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0004778687055731438,
+      "grad_norm": 6.597299531195726,
+      "learning_rate": 0.0,
+      "loss": 0.9182,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009557374111462876,
+      "grad_norm": 6.798268123961526,
+      "learning_rate": 6.369426751592358e-08,
+      "loss": 0.9348,
+      "step": 2
+    },
+    {
+      "epoch": 0.0014336061167194312,
+      "grad_norm": 6.547375642053667,
+      "learning_rate": 1.2738853503184715e-07,
+      "loss": 0.9002,
+      "step": 3
+    },
+    {
+      "epoch": 0.0019114748222925752,
+      "grad_norm": 6.714041140675801,
+      "learning_rate": 1.9108280254777072e-07,
+      "loss": 0.9221,
+      "step": 4
+    },
+    {
+      "epoch": 0.002389343527865719,
+      "grad_norm": 6.482638024201559,
+      "learning_rate": 2.547770700636943e-07,
+      "loss": 0.8989,
+      "step": 5
+    },
+    {
+      "epoch": 0.0028672122334388625,
+      "grad_norm": 6.628677060072881,
+      "learning_rate": 3.1847133757961787e-07,
+      "loss": 0.9246,
+      "step": 6
+    },
+    {
+      "epoch": 0.0033450809390120064,
+      "grad_norm": 6.54934804614698,
+      "learning_rate": 3.8216560509554143e-07,
+      "loss": 0.9118,
+      "step": 7
+    },
+    {
+      "epoch": 0.0038229496445851504,
+      "grad_norm": 6.822862171641388,
+      "learning_rate": 4.45859872611465e-07,
+      "loss": 0.9453,
+      "step": 8
+    },
+    {
+      "epoch": 0.004300818350158294,
+      "grad_norm": 6.436429750787215,
+      "learning_rate": 5.095541401273886e-07,
+      "loss": 0.8919,
+      "step": 9
+    },
+    {
+      "epoch": 0.004778687055731438,
+      "grad_norm": 6.20546872502828,
+      "learning_rate": 5.732484076433121e-07,
+      "loss": 0.9081,
+      "step": 10
+    },
+    {
+      "epoch": 0.0052565557613045814,
+      "grad_norm": 6.1990164932210305,
+      "learning_rate": 6.369426751592357e-07,
+      "loss": 0.9152,
+      "step": 11
+    },
+    {
+      "epoch": 0.005734424466877725,
+      "grad_norm": 6.213474019663624,
+      "learning_rate": 7.006369426751592e-07,
+      "loss": 0.9127,
+      "step": 12
+    },
+    {
+      "epoch": 0.006212293172450869,
+      "grad_norm": 4.774163980815126,
+      "learning_rate": 7.643312101910829e-07,
+      "loss": 0.8724,
+      "step": 13
+    },
+    {
+      "epoch": 0.006690161878024013,
+      "grad_norm": 4.635856494922676,
+      "learning_rate": 8.280254777070064e-07,
+      "loss": 0.8794,
+      "step": 14
+    },
+    {
+      "epoch": 0.007168030583597156,
+      "grad_norm": 4.4643552624491925,
+      "learning_rate": 8.9171974522293e-07,
+      "loss": 0.8574,
+      "step": 15
+    },
+    {
+      "epoch": 0.007645899289170301,
+      "grad_norm": 4.746112409024538,
+      "learning_rate": 9.554140127388537e-07,
+      "loss": 0.9037,
+      "step": 16
+    },
+    {
+      "epoch": 0.008123767994743444,
+      "grad_norm": 3.6663226655371237,
+      "learning_rate": 1.0191082802547772e-06,
+      "loss": 0.8572,
+      "step": 17
+    },
+    {
+      "epoch": 0.008601636700316589,
+      "grad_norm": 3.161655587285928,
+      "learning_rate": 1.0828025477707007e-06,
+      "loss": 0.8504,
+      "step": 18
+    },
+    {
+      "epoch": 0.009079505405889731,
+      "grad_norm": 3.023372031668388,
+      "learning_rate": 1.1464968152866242e-06,
+      "loss": 0.8404,
+      "step": 19
+    },
+    {
+      "epoch": 0.009557374111462876,
+      "grad_norm": 2.8123940051382466,
+      "learning_rate": 1.210191082802548e-06,
+      "loss": 0.8233,
+      "step": 20
+    },
+    {
+      "epoch": 0.01003524281703602,
+      "grad_norm": 2.3319495260663037,
+      "learning_rate": 1.2738853503184715e-06,
+      "loss": 0.8007,
+      "step": 21
+    },
+    {
+      "epoch": 0.010513111522609163,
+      "grad_norm": 2.1046299310129926,
+      "learning_rate": 1.337579617834395e-06,
+      "loss": 0.7893,
+      "step": 22
+    },
+    {
+      "epoch": 0.010990980228182307,
+      "grad_norm": 1.970451770951646,
+      "learning_rate": 1.4012738853503185e-06,
+      "loss": 0.8091,
+      "step": 23
+    },
+    {
+      "epoch": 0.01146884893375545,
+      "grad_norm": 2.2852013132038262,
+      "learning_rate": 1.4649681528662422e-06,
+      "loss": 0.761,
+      "step": 24
+    },
+    {
+      "epoch": 0.011946717639328594,
+      "grad_norm": 2.3125203117243527,
+      "learning_rate": 1.5286624203821657e-06,
+      "loss": 0.7576,
+      "step": 25
+    },
+    {
+      "epoch": 0.012424586344901739,
+      "grad_norm": 2.485516736089154,
+      "learning_rate": 1.5923566878980892e-06,
+      "loss": 0.7298,
+      "step": 26
+    },
+    {
+      "epoch": 0.012902455050474881,
+      "grad_norm": 2.328181944771006,
+      "learning_rate": 1.6560509554140127e-06,
+      "loss": 0.7241,
+      "step": 27
+    },
+    {
+      "epoch": 0.013380323756048026,
+      "grad_norm": 2.1332096308422592,
+      "learning_rate": 1.7197452229299363e-06,
+      "loss": 0.7561,
+      "step": 28
+    },
+    {
+      "epoch": 0.01385819246162117,
+      "grad_norm": 1.9124393579077286,
+      "learning_rate": 1.78343949044586e-06,
+      "loss": 0.7421,
+      "step": 29
+    },
+    {
+      "epoch": 0.014336061167194313,
+      "grad_norm": 1.868955513961223,
+      "learning_rate": 1.8471337579617835e-06,
+      "loss": 0.6917,
+      "step": 30
+    },
+    {
+      "epoch": 0.014813929872767457,
+      "grad_norm": 1.5698888685595818,
+      "learning_rate": 1.9108280254777074e-06,
+      "loss": 0.7297,
+      "step": 31
+    },
+    {
+      "epoch": 0.015291798578340602,
+      "grad_norm": 1.2265156599136022,
+      "learning_rate": 1.974522292993631e-06,
+      "loss": 0.698,
+      "step": 32
+    },
+    {
+      "epoch": 0.015769667283913744,
+      "grad_norm": 1.3091022856057957,
+      "learning_rate": 2.0382165605095544e-06,
+      "loss": 0.6542,
+      "step": 33
+    },
+    {
+      "epoch": 0.01624753598948689,
+      "grad_norm": 1.45790978867409,
+      "learning_rate": 2.101910828025478e-06,
+      "loss": 0.6759,
+      "step": 34
+    },
+    {
+      "epoch": 0.016725404695060033,
+      "grad_norm": 1.482461582005205,
+      "learning_rate": 2.1656050955414015e-06,
+      "loss": 0.6654,
+      "step": 35
+    },
+    {
+      "epoch": 0.017203273400633178,
+      "grad_norm": 1.3245355619344652,
+      "learning_rate": 2.229299363057325e-06,
+      "loss": 0.6541,
+      "step": 36
+    },
+    {
+      "epoch": 0.01768114210620632,
+      "grad_norm": 1.1472782080031219,
+      "learning_rate": 2.2929936305732485e-06,
+      "loss": 0.6515,
+      "step": 37
+    },
+    {
+      "epoch": 0.018159010811779463,
+      "grad_norm": 1.0289175415890455,
+      "learning_rate": 2.356687898089172e-06,
+      "loss": 0.6299,
+      "step": 38
+    },
+    {
+      "epoch": 0.018636879517352607,
+      "grad_norm": 0.9574361352811592,
+      "learning_rate": 2.420382165605096e-06,
+      "loss": 0.6258,
+      "step": 39
+    },
+    {
+      "epoch": 0.01911474822292575,
+      "grad_norm": 0.9277275080285016,
+      "learning_rate": 2.4840764331210194e-06,
+      "loss": 0.6337,
+      "step": 40
+    },
+    {
+      "epoch": 0.019592616928498896,
+      "grad_norm": 0.8945726217645557,
+      "learning_rate": 2.547770700636943e-06,
+      "loss": 0.6198,
+      "step": 41
+    },
+    {
+      "epoch": 0.02007048563407204,
+      "grad_norm": 0.8774284132253958,
+      "learning_rate": 2.6114649681528665e-06,
+      "loss": 0.6102,
+      "step": 42
+    },
+    {
+      "epoch": 0.02054835433964518,
+      "grad_norm": 0.9050997026733555,
+      "learning_rate": 2.67515923566879e-06,
+      "loss": 0.6221,
+      "step": 43
+    },
+    {
+      "epoch": 0.021026223045218326,
+      "grad_norm": 0.8632323727229961,
+      "learning_rate": 2.7388535031847135e-06,
+      "loss": 0.622,
+      "step": 44
+    },
+    {
+      "epoch": 0.02150409175079147,
+      "grad_norm": 0.813517779384421,
+      "learning_rate": 2.802547770700637e-06,
+      "loss": 0.6089,
+      "step": 45
+    },
+    {
+      "epoch": 0.021981960456364615,
+      "grad_norm": 0.8186427508339532,
+      "learning_rate": 2.8662420382165605e-06,
+      "loss": 0.6016,
+      "step": 46
+    },
+    {
+      "epoch": 0.02245982916193776,
+      "grad_norm": 0.8174387059211783,
+      "learning_rate": 2.9299363057324844e-06,
+      "loss": 0.5931,
+      "step": 47
+    },
+    {
+      "epoch": 0.0229376978675109,
+      "grad_norm": 1.305617651951218,
+      "learning_rate": 2.993630573248408e-06,
+      "loss": 0.5692,
+      "step": 48
+    },
+    {
+      "epoch": 0.023415566573084044,
+      "grad_norm": 0.8380237368547312,
+      "learning_rate": 3.0573248407643314e-06,
+      "loss": 0.5887,
+      "step": 49
+    },
+    {
+      "epoch": 0.02389343527865719,
+      "grad_norm": 0.8200238530137682,
+      "learning_rate": 3.121019108280255e-06,
+      "loss": 0.5929,
+      "step": 50
+    },
+    {
+      "epoch": 0.024371303984230333,
+      "grad_norm": 0.8403778738693355,
+      "learning_rate": 3.1847133757961785e-06,
+      "loss": 0.6014,
+      "step": 51
+    },
+    {
+      "epoch": 0.024849172689803477,
+      "grad_norm": 0.9908747877734703,
+      "learning_rate": 3.248407643312102e-06,
+      "loss": 0.5735,
+      "step": 52
+    },
+    {
+      "epoch": 0.025327041395376622,
+      "grad_norm": 0.8075484398827105,
+      "learning_rate": 3.3121019108280255e-06,
+      "loss": 0.5593,
+      "step": 53
+    },
+    {
+      "epoch": 0.025804910100949763,
+      "grad_norm": 0.7626574710843891,
+      "learning_rate": 3.375796178343949e-06,
+      "loss": 0.5806,
+      "step": 54
+    },
+    {
+      "epoch": 0.026282778806522907,
+      "grad_norm": 0.836758410598703,
+      "learning_rate": 3.4394904458598725e-06,
+      "loss": 0.5768,
+      "step": 55
+    },
+    {
+      "epoch": 0.02676064751209605,
+      "grad_norm": 0.7496247956447741,
+      "learning_rate": 3.5031847133757964e-06,
+      "loss": 0.5815,
+      "step": 56
+    },
+    {
+      "epoch": 0.027238516217669196,
+      "grad_norm": 0.7950751813645308,
+      "learning_rate": 3.56687898089172e-06,
+      "loss": 0.5893,
+      "step": 57
+    },
+    {
+      "epoch": 0.02771638492324234,
+      "grad_norm": 0.7934523889506495,
+      "learning_rate": 3.6305732484076435e-06,
+      "loss": 0.5675,
+      "step": 58
+    },
+    {
+      "epoch": 0.02819425362881548,
+      "grad_norm": 0.7469734117900833,
+      "learning_rate": 3.694267515923567e-06,
+      "loss": 0.5679,
+      "step": 59
+    },
+    {
+      "epoch": 0.028672122334388626,
+      "grad_norm": 0.7390161251771473,
+      "learning_rate": 3.757961783439491e-06,
+      "loss": 0.5464,
+      "step": 60
+    },
+    {
+      "epoch": 0.02914999103996177,
+      "grad_norm": 0.756427918659596,
+      "learning_rate": 3.821656050955415e-06,
+      "loss": 0.5422,
+      "step": 61
+    },
+    {
+      "epoch": 0.029627859745534914,
+      "grad_norm": 0.7389039375217277,
+      "learning_rate": 3.885350318471338e-06,
+      "loss": 0.5551,
+      "step": 62
+    },
+    {
+      "epoch": 0.03010572845110806,
+      "grad_norm": 0.7367851829841032,
+      "learning_rate": 3.949044585987262e-06,
+      "loss": 0.5449,
+      "step": 63
+    },
+    {
+      "epoch": 0.030583597156681203,
+      "grad_norm": 0.7123685102640923,
+      "learning_rate": 4.012738853503185e-06,
+      "loss": 0.5458,
+      "step": 64
+    },
+    {
+      "epoch": 0.031061465862254344,
+      "grad_norm": 0.7357352408857197,
+      "learning_rate": 4.076433121019109e-06,
+      "loss": 0.5341,
+      "step": 65
+    },
+    {
+      "epoch": 0.03153933456782749,
+      "grad_norm": 0.7214843181688534,
+      "learning_rate": 4.140127388535032e-06,
+      "loss": 0.5459,
+      "step": 66
+    },
+    {
+      "epoch": 0.032017203273400636,
+      "grad_norm": 0.7467241250442314,
+      "learning_rate": 4.203821656050956e-06,
+      "loss": 0.5396,
+      "step": 67
+    },
+    {
+      "epoch": 0.03249507197897378,
+      "grad_norm": 0.7727746910699702,
+      "learning_rate": 4.26751592356688e-06,
+      "loss": 0.5472,
+      "step": 68
+    },
+    {
+      "epoch": 0.03297294068454692,
+      "grad_norm": 0.7251863973183906,
+      "learning_rate": 4.331210191082803e-06,
+      "loss": 0.5597,
+      "step": 69
+    },
+    {
+      "epoch": 0.033450809390120066,
+      "grad_norm": 0.7272882874909277,
+      "learning_rate": 4.394904458598727e-06,
+      "loss": 0.5556,
+      "step": 70
+    },
+    {
+      "epoch": 0.03392867809569321,
+      "grad_norm": 0.8391711398530735,
+      "learning_rate": 4.45859872611465e-06,
+      "loss": 0.5496,
+      "step": 71
+    },
+    {
+      "epoch": 0.034406546801266355,
+      "grad_norm": 0.9360566101078907,
+      "learning_rate": 4.522292993630574e-06,
+      "loss": 0.5432,
+      "step": 72
+    },
+    {
+      "epoch": 0.034884415506839496,
+      "grad_norm": 0.7797140366402956,
+      "learning_rate": 4.585987261146497e-06,
+      "loss": 0.5395,
+      "step": 73
+    },
+    {
+      "epoch": 0.03536228421241264,
+      "grad_norm": 0.707051554509993,
+      "learning_rate": 4.649681528662421e-06,
+      "loss": 0.5311,
+      "step": 74
+    },
+    {
+      "epoch": 0.035840152917985785,
+      "grad_norm": 0.7834053957637365,
+      "learning_rate": 4.713375796178344e-06,
+      "loss": 0.5487,
+      "step": 75
+    },
+    {
+      "epoch": 0.036318021623558926,
+      "grad_norm": 0.7935586931488725,
+      "learning_rate": 4.777070063694268e-06,
+      "loss": 0.5327,
+      "step": 76
+    },
+    {
+      "epoch": 0.036795890329132074,
+      "grad_norm": 0.770263498050658,
+      "learning_rate": 4.840764331210192e-06,
+      "loss": 0.5146,
+      "step": 77
+    },
+    {
+      "epoch": 0.037273759034705214,
+      "grad_norm": 0.9115020793198642,
+      "learning_rate": 4.904458598726115e-06,
+      "loss": 0.5438,
+      "step": 78
+    },
+    {
+      "epoch": 0.037751627740278355,
+      "grad_norm": 0.8472805895573531,
+      "learning_rate": 4.968152866242039e-06,
+      "loss": 0.549,
+      "step": 79
+    },
+    {
+      "epoch": 0.0382294964458515,
+      "grad_norm": 0.7542037051233978,
+      "learning_rate": 5.031847133757962e-06,
+      "loss": 0.5351,
+      "step": 80
+    },
+    {
+      "epoch": 0.038707365151424644,
+      "grad_norm": 0.747781469023545,
+      "learning_rate": 5.095541401273886e-06,
+      "loss": 0.5084,
+      "step": 81
+    },
+    {
+      "epoch": 0.03918523385699779,
+      "grad_norm": 0.7476744681713057,
+      "learning_rate": 5.159235668789809e-06,
+      "loss": 0.5245,
+      "step": 82
+    },
+    {
+      "epoch": 0.03966310256257093,
+      "grad_norm": 0.7768862389201611,
+      "learning_rate": 5.222929936305733e-06,
+      "loss": 0.5271,
+      "step": 83
+    },
+    {
+      "epoch": 0.04014097126814408,
+      "grad_norm": 0.7111007495704235,
+      "learning_rate": 5.286624203821657e-06,
+      "loss": 0.5266,
+      "step": 84
+    },
+    {
+      "epoch": 0.04061883997371722,
+      "grad_norm": 0.732731140034443,
+      "learning_rate": 5.35031847133758e-06,
+      "loss": 0.5263,
+      "step": 85
+    },
+    {
+      "epoch": 0.04109670867929036,
+      "grad_norm": 0.8652091302250842,
+      "learning_rate": 5.414012738853504e-06,
+      "loss": 0.5273,
+      "step": 86
+    },
+    {
+      "epoch": 0.04157457738486351,
+      "grad_norm": 0.737707359165254,
+      "learning_rate": 5.477707006369427e-06,
+      "loss": 0.5061,
+      "step": 87
+    },
+    {
+      "epoch": 0.04205244609043665,
+      "grad_norm": 0.8065815895775212,
+      "learning_rate": 5.541401273885351e-06,
+      "loss": 0.5098,
+      "step": 88
+    },
+    {
+      "epoch": 0.0425303147960098,
+      "grad_norm": 0.7266312168372544,
+      "learning_rate": 5.605095541401274e-06,
+      "loss": 0.4999,
+      "step": 89
+    },
+    {
+      "epoch": 0.04300818350158294,
+      "grad_norm": 0.7293962461070471,
+      "learning_rate": 5.668789808917198e-06,
+      "loss": 0.5213,
+      "step": 90
+    },
+    {
+      "epoch": 0.04348605220715608,
+      "grad_norm": 0.7774805752991096,
+      "learning_rate": 5.732484076433121e-06,
+      "loss": 0.4995,
+      "step": 91
+    },
+    {
+      "epoch": 0.04396392091272923,
+      "grad_norm": 0.7530825142450961,
+      "learning_rate": 5.796178343949045e-06,
+      "loss": 0.5365,
+      "step": 92
+    },
+    {
+      "epoch": 0.04444178961830237,
+      "grad_norm": 0.6739843506427442,
+      "learning_rate": 5.859872611464969e-06,
+      "loss": 0.5146,
+      "step": 93
+    },
+    {
+      "epoch": 0.04491965832387552,
+      "grad_norm": 0.7872225186802001,
+      "learning_rate": 5.923566878980892e-06,
+      "loss": 0.5171,
+      "step": 94
+    },
+    {
+      "epoch": 0.04539752702944866,
+      "grad_norm": 0.7593936458726057,
+      "learning_rate": 5.987261146496816e-06,
+      "loss": 0.5075,
+      "step": 95
+    },
+    {
+      "epoch": 0.0458753957350218,
+      "grad_norm": 0.7402231844428684,
+      "learning_rate": 6.050955414012739e-06,
+      "loss": 0.5219,
+      "step": 96
+    },
+    {
+      "epoch": 0.04635326444059495,
+      "grad_norm": 0.8456741711717691,
+      "learning_rate": 6.114649681528663e-06,
+      "loss": 0.5179,
+      "step": 97
+    },
+    {
+      "epoch": 0.04683113314616809,
+      "grad_norm": 0.7479830972685164,
+      "learning_rate": 6.178343949044586e-06,
+      "loss": 0.516,
+      "step": 98
+    },
+    {
+      "epoch": 0.047309001851741236,
+      "grad_norm": 0.7472523172623668,
+      "learning_rate": 6.24203821656051e-06,
+      "loss": 0.498,
+      "step": 99
+    },
+    {
+      "epoch": 0.04778687055731438,
+      "grad_norm": 0.8971442569968484,
+      "learning_rate": 6.305732484076433e-06,
+      "loss": 0.51,
+      "step": 100
+    },
+    {
+      "epoch": 0.04826473926288752,
+      "grad_norm": 0.7564195148457464,
+      "learning_rate": 6.369426751592357e-06,
+      "loss": 0.5198,
+      "step": 101
+    },
+    {
+      "epoch": 0.048742607968460666,
+      "grad_norm": 1.0620219552470769,
+      "learning_rate": 6.433121019108281e-06,
+      "loss": 0.5,
+      "step": 102
+    },
+    {
+      "epoch": 0.04922047667403381,
+      "grad_norm": 0.7898571941764354,
+      "learning_rate": 6.496815286624204e-06,
+      "loss": 0.5154,
+      "step": 103
+    },
+    {
+      "epoch": 0.049698345379606955,
+      "grad_norm": 0.7334402709211295,
+      "learning_rate": 6.560509554140128e-06,
+      "loss": 0.499,
+      "step": 104
+    },
+    {
+      "epoch": 0.050176214085180096,
+      "grad_norm": 0.7398643150657089,
+      "learning_rate": 6.624203821656051e-06,
+      "loss": 0.506,
+      "step": 105
+    },
+    {
+      "epoch": 0.050654082790753244,
+      "grad_norm": 0.7563474458560936,
+      "learning_rate": 6.687898089171975e-06,
+      "loss": 0.5153,
+      "step": 106
+    },
+    {
+      "epoch": 0.051131951496326385,
+      "grad_norm": 0.7767148986192806,
+      "learning_rate": 6.751592356687898e-06,
+      "loss": 0.4994,
+      "step": 107
+    },
+    {
+      "epoch": 0.051609820201899526,
+      "grad_norm": 0.7746503252727361,
+      "learning_rate": 6.815286624203822e-06,
+      "loss": 0.4994,
+      "step": 108
+    },
+    {
+      "epoch": 0.05208768890747267,
+      "grad_norm": 0.8270879600000823,
+      "learning_rate": 6.878980891719745e-06,
+      "loss": 0.4912,
+      "step": 109
+    },
+    {
+      "epoch": 0.052565557613045814,
+      "grad_norm": 0.7271341530708423,
+      "learning_rate": 6.942675159235669e-06,
+      "loss": 0.5038,
+      "step": 110
+    },
+    {
+      "epoch": 0.05304342631861896,
+      "grad_norm": 0.7909333113850586,
+      "learning_rate": 7.006369426751593e-06,
+      "loss": 0.4965,
+      "step": 111
+    },
+    {
+      "epoch": 0.0535212950241921,
+      "grad_norm": 1.1920708554598451,
+      "learning_rate": 7.070063694267516e-06,
+      "loss": 0.4835,
+      "step": 112
+    },
+    {
+      "epoch": 0.053999163729765244,
+      "grad_norm": 1.1437207443713215,
+      "learning_rate": 7.13375796178344e-06,
+      "loss": 0.4871,
+      "step": 113
+    },
+    {
+      "epoch": 0.05447703243533839,
+      "grad_norm": 0.7995131778249069,
+      "learning_rate": 7.197452229299363e-06,
+      "loss": 0.5125,
+      "step": 114
+    },
+    {
+      "epoch": 0.05495490114091153,
+      "grad_norm": 0.7658107271492185,
+      "learning_rate": 7.261146496815287e-06,
+      "loss": 0.5043,
+      "step": 115
+    },
+    {
+      "epoch": 0.05543276984648468,
+      "grad_norm": 0.922334411856167,
+      "learning_rate": 7.32484076433121e-06,
+      "loss": 0.4989,
+      "step": 116
+    },
+    {
+      "epoch": 0.05591063855205782,
+      "grad_norm": 0.7836377405415712,
+      "learning_rate": 7.388535031847134e-06,
+      "loss": 0.4899,
+      "step": 117
+    },
+    {
+      "epoch": 0.05638850725763096,
+      "grad_norm": 0.8764338235424465,
+      "learning_rate": 7.452229299363057e-06,
+      "loss": 0.5001,
+      "step": 118
+    },
+    {
+      "epoch": 0.05686637596320411,
+      "grad_norm": 0.8139065283378535,
+      "learning_rate": 7.515923566878982e-06,
+      "loss": 0.5079,
+      "step": 119
+    },
+    {
+      "epoch": 0.05734424466877725,
+      "grad_norm": 0.8340053526116425,
+      "learning_rate": 7.579617834394906e-06,
+      "loss": 0.5087,
+      "step": 120
+    },
+    {
+      "epoch": 0.0578221133743504,
+      "grad_norm": 0.9688304321284679,
+      "learning_rate": 7.64331210191083e-06,
+      "loss": 0.4945,
+      "step": 121
+    },
+    {
+      "epoch": 0.05829998207992354,
+      "grad_norm": 0.8501127062676569,
+      "learning_rate": 7.707006369426753e-06,
+      "loss": 0.4736,
+      "step": 122
+    },
+    {
+      "epoch": 0.05877785078549669,
+      "grad_norm": 0.8146384000193851,
+      "learning_rate": 7.770700636942676e-06,
+      "loss": 0.4917,
+      "step": 123
+    },
+    {
+      "epoch": 0.05925571949106983,
+      "grad_norm": 0.8905411656723172,
+      "learning_rate": 7.8343949044586e-06,
+      "loss": 0.4931,
+      "step": 124
+    },
+    {
+      "epoch": 0.05973358819664297,
+      "grad_norm": 0.8610319305755783,
+      "learning_rate": 7.898089171974524e-06,
+      "loss": 0.5019,
+      "step": 125
+    },
+    {
+      "epoch": 0.06021145690221612,
+      "grad_norm": 0.8674477387963704,
+      "learning_rate": 7.961783439490447e-06,
+      "loss": 0.4896,
+      "step": 126
+    },
+    {
+      "epoch": 0.06068932560778926,
+      "grad_norm": 0.8577218357744752,
+      "learning_rate": 8.02547770700637e-06,
+      "loss": 0.4811,
+      "step": 127
+    },
+    {
+      "epoch": 0.06116719431336241,
+      "grad_norm": 0.8448292497264104,
+      "learning_rate": 8.089171974522295e-06,
+      "loss": 0.504,
+      "step": 128
+    },
+    {
+      "epoch": 0.06164506301893555,
+      "grad_norm": 0.937833097972482,
+      "learning_rate": 8.152866242038218e-06,
+      "loss": 0.4962,
+      "step": 129
+    },
+    {
+      "epoch": 0.06212293172450869,
+      "grad_norm": 0.7840157899279157,
+      "learning_rate": 8.21656050955414e-06,
+      "loss": 0.4985,
+      "step": 130
+    },
+    {
+      "epoch": 0.06260080043008183,
+      "grad_norm": 0.8889842008188884,
+      "learning_rate": 8.280254777070064e-06,
+      "loss": 0.4917,
+      "step": 131
+    },
+    {
+      "epoch": 0.06307866913565498,
+      "grad_norm": 0.8118606083274249,
+      "learning_rate": 8.343949044585989e-06,
+      "loss": 0.5121,
+      "step": 132
+    },
+    {
+      "epoch": 0.06355653784122813,
+      "grad_norm": 0.8160111895836885,
+      "learning_rate": 8.407643312101912e-06,
+      "loss": 0.4975,
+      "step": 133
+    },
+    {
+      "epoch": 0.06403440654680127,
+      "grad_norm": 0.8257529923932002,
+      "learning_rate": 8.471337579617835e-06,
+      "loss": 0.4905,
+      "step": 134
+    },
+    {
+      "epoch": 0.06451227525237441,
+      "grad_norm": 0.9316496549034691,
+      "learning_rate": 8.53503184713376e-06,
+      "loss": 0.4922,
+      "step": 135
+    },
+    {
+      "epoch": 0.06499014395794755,
+      "grad_norm": 3.6885784250742777,
+      "learning_rate": 8.598726114649683e-06,
+      "loss": 0.5088,
+      "step": 136
+    },
+    {
+      "epoch": 0.0654680126635207,
+      "grad_norm": 1.0431096223978598,
+      "learning_rate": 8.662420382165606e-06,
+      "loss": 0.4851,
+      "step": 137
+    },
+    {
+      "epoch": 0.06594588136909384,
+      "grad_norm": 0.9301103008954675,
+      "learning_rate": 8.726114649681529e-06,
+      "loss": 0.5035,
+      "step": 138
+    },
+    {
+      "epoch": 0.06642375007466698,
+      "grad_norm": 0.961097839927529,
+      "learning_rate": 8.789808917197454e-06,
+      "loss": 0.5035,
+      "step": 139
+    },
+    {
+      "epoch": 0.06690161878024013,
+      "grad_norm": 0.8559608590796796,
+      "learning_rate": 8.853503184713377e-06,
+      "loss": 0.4786,
+      "step": 140
+    },
+    {
+      "epoch": 0.06737948748581327,
+      "grad_norm": 0.8702110473814613,
+      "learning_rate": 8.9171974522293e-06,
+      "loss": 0.4769,
+      "step": 141
+    },
+    {
+      "epoch": 0.06785735619138641,
+      "grad_norm": 0.89503054543084,
+      "learning_rate": 8.980891719745225e-06,
+      "loss": 0.4936,
+      "step": 142
+    },
+    {
+      "epoch": 0.06833522489695956,
+      "grad_norm": 0.9405145862374199,
+      "learning_rate": 9.044585987261148e-06,
+      "loss": 0.4994,
+      "step": 143
+    },
+    {
+      "epoch": 0.06881309360253271,
+      "grad_norm": 0.8359505048050953,
+      "learning_rate": 9.10828025477707e-06,
+      "loss": 0.4888,
+      "step": 144
+    },
+    {
+      "epoch": 0.06929096230810584,
+      "grad_norm": 0.8335744765307723,
+      "learning_rate": 9.171974522292994e-06,
+      "loss": 0.486,
+      "step": 145
+    },
+    {
+      "epoch": 0.06976883101367899,
+      "grad_norm": 0.9191064552692454,
+      "learning_rate": 9.235668789808919e-06,
+      "loss": 0.4936,
+      "step": 146
+    },
+    {
+      "epoch": 0.07024669971925214,
+      "grad_norm": 0.8613976280696265,
+      "learning_rate": 9.299363057324842e-06,
+      "loss": 0.4892,
+      "step": 147
+    },
+    {
+      "epoch": 0.07072456842482527,
+      "grad_norm": 0.9085851257000545,
+      "learning_rate": 9.363057324840765e-06,
+      "loss": 0.4959,
+      "step": 148
+    },
+    {
+      "epoch": 0.07120243713039842,
+      "grad_norm": 0.8498504776265091,
+      "learning_rate": 9.426751592356688e-06,
+      "loss": 0.4799,
+      "step": 149
+    },
+    {
+      "epoch": 0.07168030583597157,
+      "grad_norm": 0.981381083456667,
+      "learning_rate": 9.490445859872613e-06,
+      "loss": 0.4759,
+      "step": 150
+    },
+    {
+      "epoch": 0.07215817454154472,
+      "grad_norm": 0.956769830775876,
+      "learning_rate": 9.554140127388536e-06,
+      "loss": 0.4898,
+      "step": 151
+    },
+    {
+      "epoch": 0.07263604324711785,
+      "grad_norm": 1.0038255875882836,
+      "learning_rate": 9.617834394904459e-06,
+      "loss": 0.4775,
+      "step": 152
+    },
+    {
+      "epoch": 0.073113911952691,
+      "grad_norm": 0.8713370885352425,
+      "learning_rate": 9.681528662420384e-06,
+      "loss": 0.4969,
+      "step": 153
+    },
+    {
+      "epoch": 0.07359178065826415,
+      "grad_norm": 0.8141213363595519,
+      "learning_rate": 9.745222929936307e-06,
+      "loss": 0.4705,
+      "step": 154
+    },
+    {
+      "epoch": 0.07406964936383728,
+      "grad_norm": 3.751367127935264,
+      "learning_rate": 9.80891719745223e-06,
+      "loss": 0.4864,
+      "step": 155
+    },
+    {
+      "epoch": 0.07454751806941043,
+      "grad_norm": 1.1148603281454004,
+      "learning_rate": 9.872611464968153e-06,
+      "loss": 0.4822,
+      "step": 156
+    },
+    {
+      "epoch": 0.07502538677498358,
+      "grad_norm": 0.8287232413246658,
+      "learning_rate": 9.936305732484078e-06,
+      "loss": 0.4642,
+      "step": 157
+    },
+    {
+      "epoch": 0.07550325548055671,
+      "grad_norm": 0.8915041269681552,
+      "learning_rate": 1e-05,
+      "loss": 0.4768,
+      "step": 158
+    },
+    {
+      "epoch": 0.07598112418612986,
+      "grad_norm": 0.873722784678586,
+      "learning_rate": 1.0063694267515924e-05,
+      "loss": 0.4808,
+      "step": 159
+    },
+    {
+      "epoch": 0.076458992891703,
+      "grad_norm": 0.8676339202348703,
+      "learning_rate": 1.0127388535031849e-05,
+      "loss": 0.4916,
+      "step": 160
+    },
+    {
+      "epoch": 0.07693686159727615,
+      "grad_norm": 0.9305589175210003,
+      "learning_rate": 1.0191082802547772e-05,
+      "loss": 0.4747,
+      "step": 161
+    },
+    {
+      "epoch": 0.07741473030284929,
+      "grad_norm": 0.8519987127043529,
+      "learning_rate": 1.0254777070063695e-05,
+      "loss": 0.4756,
+      "step": 162
+    },
+    {
+      "epoch": 0.07789259900842244,
+      "grad_norm": 0.8823350335352326,
+      "learning_rate": 1.0318471337579618e-05,
+      "loss": 0.4662,
+      "step": 163
+    },
+    {
+      "epoch": 0.07837046771399558,
+      "grad_norm": 0.7500687237850792,
+      "learning_rate": 1.0382165605095543e-05,
+      "loss": 0.485,
+      "step": 164
+    },
+    {
+      "epoch": 0.07884833641956872,
+      "grad_norm": 0.8517746565202567,
+      "learning_rate": 1.0445859872611466e-05,
+      "loss": 0.4863,
+      "step": 165
+    },
+    {
+      "epoch": 0.07932620512514187,
+      "grad_norm": 1.2439834788261996,
+      "learning_rate": 1.0509554140127389e-05,
+      "loss": 0.4818,
+      "step": 166
+    },
+    {
+      "epoch": 0.07980407383071501,
+      "grad_norm": 0.9673894115074996,
+      "learning_rate": 1.0573248407643314e-05,
+      "loss": 0.4916,
+      "step": 167
+    },
+    {
+      "epoch": 0.08028194253628816,
+      "grad_norm": 0.8778733640888571,
+      "learning_rate": 1.0636942675159237e-05,
+      "loss": 0.4786,
+      "step": 168
+    },
+    {
+      "epoch": 0.0807598112418613,
+      "grad_norm": 0.9193416256035223,
+      "learning_rate": 1.070063694267516e-05,
+      "loss": 0.4806,
+      "step": 169
+    },
+    {
+      "epoch": 0.08123767994743444,
+      "grad_norm": 1.0391078840034496,
+      "learning_rate": 1.0764331210191083e-05,
+      "loss": 0.4601,
+      "step": 170
+    },
+    {
+      "epoch": 0.08171554865300759,
+      "grad_norm": 0.8696302636427024,
+      "learning_rate": 1.0828025477707008e-05,
+      "loss": 0.4766,
+      "step": 171
+    },
+    {
+      "epoch": 0.08219341735858073,
+      "grad_norm": 0.8538992156431188,
+      "learning_rate": 1.089171974522293e-05,
+      "loss": 0.4929,
+      "step": 172
+    },
+    {
+      "epoch": 0.08267128606415387,
+      "grad_norm": 0.8784469546200593,
+      "learning_rate": 1.0955414012738854e-05,
+      "loss": 0.4827,
+      "step": 173
+    },
+    {
+      "epoch": 0.08314915476972702,
+      "grad_norm": 0.7996342900390987,
+      "learning_rate": 1.1019108280254777e-05,
+      "loss": 0.4768,
+      "step": 174
+    },
+    {
+      "epoch": 0.08362702347530016,
+      "grad_norm": 0.9283558874133578,
+      "learning_rate": 1.1082802547770702e-05,
+      "loss": 0.4654,
+      "step": 175
+    },
+    {
+      "epoch": 0.0841048921808733,
+      "grad_norm": 1.1381555316283323,
+      "learning_rate": 1.1146496815286625e-05,
+      "loss": 0.4741,
+      "step": 176
+    },
+    {
+      "epoch": 0.08458276088644645,
+      "grad_norm": 0.9965684448163404,
+      "learning_rate": 1.1210191082802548e-05,
+      "loss": 0.4836,
+      "step": 177
+    },
+    {
+      "epoch": 0.0850606295920196,
+      "grad_norm": 0.8998507742131849,
+      "learning_rate": 1.1273885350318473e-05,
+      "loss": 0.4996,
+      "step": 178
+    },
+    {
+      "epoch": 0.08553849829759273,
+      "grad_norm": 1.0467069720974493,
+      "learning_rate": 1.1337579617834396e-05,
+      "loss": 0.4824,
+      "step": 179
+    },
+    {
+      "epoch": 0.08601636700316588,
+      "grad_norm": 0.865608733409604,
+      "learning_rate": 1.1401273885350319e-05,
+      "loss": 0.4701,
+      "step": 180
+    },
+    {
+      "epoch": 0.08649423570873903,
+      "grad_norm": 0.8930941148308225,
+      "learning_rate": 1.1464968152866242e-05,
+      "loss": 0.4887,
+      "step": 181
+    },
+    {
+      "epoch": 0.08697210441431216,
+      "grad_norm": 0.9632697993858005,
+      "learning_rate": 1.1528662420382167e-05,
+      "loss": 0.4592,
+      "step": 182
+    },
+    {
+      "epoch": 0.08744997311988531,
+      "grad_norm": 0.8648798405759216,
+      "learning_rate": 1.159235668789809e-05,
+      "loss": 0.461,
+      "step": 183
+    },
+    {
+      "epoch": 0.08792784182545846,
+      "grad_norm": 1.1724901516943187,
+      "learning_rate": 1.1656050955414013e-05,
+      "loss": 0.4701,
+      "step": 184
+    },
+    {
+      "epoch": 0.08840571053103159,
+      "grad_norm": 1.428382716298115,
+      "learning_rate": 1.1719745222929938e-05,
+      "loss": 0.4774,
+      "step": 185
+    },
+    {
+      "epoch": 0.08888357923660474,
+      "grad_norm": 0.9014718070520222,
+      "learning_rate": 1.178343949044586e-05,
+      "loss": 0.4825,
+      "step": 186
+    },
+    {
+      "epoch": 0.08936144794217789,
+      "grad_norm": 0.9006881000139725,
+      "learning_rate": 1.1847133757961784e-05,
+      "loss": 0.4729,
+      "step": 187
+    },
+    {
+      "epoch": 0.08983931664775104,
+      "grad_norm": 0.8927655043358175,
+      "learning_rate": 1.1910828025477707e-05,
+      "loss": 0.4725,
+      "step": 188
+    },
+    {
+      "epoch": 0.09031718535332417,
+      "grad_norm": 0.8765220142542697,
+      "learning_rate": 1.1974522292993632e-05,
+      "loss": 0.4739,
+      "step": 189
+    },
+    {
+      "epoch": 0.09079505405889732,
+      "grad_norm": 0.8335515837070985,
+      "learning_rate": 1.2038216560509555e-05,
+      "loss": 0.4721,
+      "step": 190
+    },
+    {
+      "epoch": 0.09127292276447047,
+      "grad_norm": 0.8307106452230681,
+      "learning_rate": 1.2101910828025478e-05,
+      "loss": 0.4846,
+      "step": 191
+    },
+    {
+      "epoch": 0.0917507914700436,
+      "grad_norm": 0.8681247868696225,
+      "learning_rate": 1.2165605095541401e-05,
+      "loss": 0.4728,
+      "step": 192
+    },
+    {
+      "epoch": 0.09222866017561675,
+      "grad_norm": 0.8869954220873089,
+      "learning_rate": 1.2229299363057326e-05,
+      "loss": 0.4721,
+      "step": 193
+    },
+    {
+      "epoch": 0.0927065288811899,
+      "grad_norm": 0.8726364899218871,
+      "learning_rate": 1.2292993630573249e-05,
+      "loss": 0.4676,
+      "step": 194
+    },
+    {
+      "epoch": 0.09318439758676304,
+      "grad_norm": 0.9315747844747434,
+      "learning_rate": 1.2356687898089172e-05,
+      "loss": 0.4662,
+      "step": 195
+    },
+    {
+      "epoch": 0.09366226629233618,
+      "grad_norm": 0.8894305822680878,
+      "learning_rate": 1.2420382165605097e-05,
+      "loss": 0.4593,
+      "step": 196
+    },
+    {
+      "epoch": 0.09414013499790932,
+      "grad_norm": 0.8882716488428526,
+      "learning_rate": 1.248407643312102e-05,
+      "loss": 0.4805,
+      "step": 197
+    },
+    {
+      "epoch": 0.09461800370348247,
+      "grad_norm": 0.9004982374479936,
+      "learning_rate": 1.2547770700636943e-05,
+      "loss": 0.481,
+      "step": 198
+    },
+    {
+      "epoch": 0.0950958724090556,
+      "grad_norm": 0.8961471264918875,
+      "learning_rate": 1.2611464968152866e-05,
+      "loss": 0.4805,
+      "step": 199
+    },
+    {
+      "epoch": 0.09557374111462875,
+      "grad_norm": 0.883909235141162,
+      "learning_rate": 1.267515923566879e-05,
+      "loss": 0.4617,
+      "step": 200
+    },
+    {
+      "epoch": 0.0960516098202019,
+      "grad_norm": 0.8490947466546838,
+      "learning_rate": 1.2738853503184714e-05,
+      "loss": 0.4643,
+      "step": 201
+    },
+    {
+      "epoch": 0.09652947852577504,
+      "grad_norm": 1.0723263514380847,
+      "learning_rate": 1.2802547770700637e-05,
+      "loss": 0.4792,
+      "step": 202
+    },
+    {
+      "epoch": 0.09700734723134818,
+      "grad_norm": 0.8029199282504613,
+      "learning_rate": 1.2866242038216562e-05,
+      "loss": 0.4698,
+      "step": 203
+    },
+    {
+      "epoch": 0.09748521593692133,
+      "grad_norm": 0.8834995304331348,
+      "learning_rate": 1.2929936305732485e-05,
+      "loss": 0.4699,
+      "step": 204
+    },
+    {
+      "epoch": 0.09796308464249448,
+      "grad_norm": 0.8189162195122099,
+      "learning_rate": 1.2993630573248408e-05,
+      "loss": 0.4398,
+      "step": 205
+    },
+    {
+      "epoch": 0.09844095334806761,
+      "grad_norm": 0.9229854446430572,
+      "learning_rate": 1.3057324840764331e-05,
+      "loss": 0.454,
+      "step": 206
+    },
+    {
+      "epoch": 0.09891882205364076,
+      "grad_norm": 0.8530294598476205,
+      "learning_rate": 1.3121019108280256e-05,
+      "loss": 0.4644,
+      "step": 207
+    },
+    {
+      "epoch": 0.09939669075921391,
+      "grad_norm": 1.0909086978949185,
+      "learning_rate": 1.3184713375796179e-05,
+      "loss": 0.4736,
+      "step": 208
+    },
+    {
+      "epoch": 0.09987455946478704,
+      "grad_norm": 0.9272994665983898,
+      "learning_rate": 1.3248407643312102e-05,
+      "loss": 0.4515,
+      "step": 209
+    },
+    {
+      "epoch": 0.10035242817036019,
+      "grad_norm": 1.1625605063974396,
+      "learning_rate": 1.3312101910828025e-05,
+      "loss": 0.4477,
+      "step": 210
+    },
+    {
+      "epoch": 0.10083029687593334,
+      "grad_norm": 0.9352271125871605,
+      "learning_rate": 1.337579617834395e-05,
+      "loss": 0.4878,
+      "step": 211
+    },
+    {
+      "epoch": 0.10130816558150649,
+      "grad_norm": 1.2595391048195885,
+      "learning_rate": 1.3439490445859873e-05,
+      "loss": 0.4694,
+      "step": 212
+    },
+    {
+      "epoch": 0.10178603428707962,
+      "grad_norm": 0.8996354926203881,
+      "learning_rate": 1.3503184713375796e-05,
+      "loss": 0.4564,
+      "step": 213
+    },
+    {
+      "epoch": 0.10226390299265277,
+      "grad_norm": 0.9358758715512773,
+      "learning_rate": 1.356687898089172e-05,
+      "loss": 0.4866,
+      "step": 214
+    },
+    {
+      "epoch": 0.10274177169822592,
+      "grad_norm": 0.8988782843133788,
+      "learning_rate": 1.3630573248407644e-05,
+      "loss": 0.4657,
+      "step": 215
+    },
+    {
+      "epoch": 0.10321964040379905,
+      "grad_norm": 0.9416537008791075,
+      "learning_rate": 1.3694267515923567e-05,
+      "loss": 0.4563,
+      "step": 216
+    },
+    {
+      "epoch": 0.1036975091093722,
+      "grad_norm": 0.8472207027263581,
+      "learning_rate": 1.375796178343949e-05,
+      "loss": 0.4606,
+      "step": 217
+    },
+    {
+      "epoch": 0.10417537781494535,
+      "grad_norm": 0.8585181561986107,
+      "learning_rate": 1.3821656050955415e-05,
+      "loss": 0.4825,
+      "step": 218
+    },
+    {
+      "epoch": 0.10465324652051848,
+      "grad_norm": 1.015742170592993,
+      "learning_rate": 1.3885350318471338e-05,
+      "loss": 0.4576,
+      "step": 219
+    },
+    {
+      "epoch": 0.10513111522609163,
+      "grad_norm": 0.8766329989197548,
+      "learning_rate": 1.3949044585987261e-05,
+      "loss": 0.4648,
+      "step": 220
+    },
+    {
+      "epoch": 0.10560898393166478,
+      "grad_norm": 0.8917767684612441,
+      "learning_rate": 1.4012738853503186e-05,
+      "loss": 0.4713,
+      "step": 221
+    },
+    {
+      "epoch": 0.10608685263723792,
+      "grad_norm": 0.8758511266694687,
+      "learning_rate": 1.4076433121019109e-05,
+      "loss": 0.4741,
+      "step": 222
+    },
+    {
+      "epoch": 0.10656472134281106,
+      "grad_norm": 0.8916775950330903,
+      "learning_rate": 1.4140127388535032e-05,
+      "loss": 0.4544,
+      "step": 223
+    },
+    {
+      "epoch": 0.1070425900483842,
+      "grad_norm": 1.4793552859878714,
+      "learning_rate": 1.4203821656050955e-05,
+      "loss": 0.4497,
+      "step": 224
+    },
+    {
+      "epoch": 0.10752045875395735,
+      "grad_norm": 1.1872856996535994,
+      "learning_rate": 1.426751592356688e-05,
+      "loss": 0.4749,
+      "step": 225
+    },
+    {
+      "epoch": 0.10799832745953049,
+      "grad_norm": 0.8700025220590151,
+      "learning_rate": 1.4331210191082803e-05,
+      "loss": 0.4751,
+      "step": 226
+    },
+    {
+      "epoch": 0.10847619616510364,
+      "grad_norm": 0.9905036123666954,
+      "learning_rate": 1.4394904458598726e-05,
+      "loss": 0.4734,
+      "step": 227
+    },
+    {
+      "epoch": 0.10895406487067678,
+      "grad_norm": 0.9544401508765005,
+      "learning_rate": 1.4458598726114649e-05,
+      "loss": 0.4662,
+      "step": 228
+    },
+    {
+      "epoch": 0.10943193357624993,
+      "grad_norm": 0.8769851114803372,
+      "learning_rate": 1.4522292993630574e-05,
+      "loss": 0.4669,
+      "step": 229
+    },
+    {
+      "epoch": 0.10990980228182307,
+      "grad_norm": 0.8733018885948226,
+      "learning_rate": 1.4585987261146497e-05,
+      "loss": 0.4696,
+      "step": 230
+    },
+    {
+      "epoch": 0.11038767098739621,
+      "grad_norm": 0.9441368673354331,
+      "learning_rate": 1.464968152866242e-05,
+      "loss": 0.4546,
+      "step": 231
+    },
+    {
+      "epoch": 0.11086553969296936,
+      "grad_norm": 0.789316515274095,
+      "learning_rate": 1.4713375796178345e-05,
+      "loss": 0.4927,
+      "step": 232
+    },
+    {
+      "epoch": 0.1113434083985425,
+      "grad_norm": 0.8994006543929829,
+      "learning_rate": 1.4777070063694268e-05,
+      "loss": 0.4768,
+      "step": 233
+    },
+    {
+      "epoch": 0.11182127710411564,
+      "grad_norm": 0.8060765297549664,
+      "learning_rate": 1.4840764331210191e-05,
+      "loss": 0.4663,
+      "step": 234
+    },
+    {
+      "epoch": 0.11229914580968879,
+      "grad_norm": 0.8826274187881293,
+      "learning_rate": 1.4904458598726114e-05,
+      "loss": 0.4635,
+      "step": 235
+    },
+    {
+      "epoch": 0.11277701451526193,
+      "grad_norm": 0.9233679523033971,
+      "learning_rate": 1.4968152866242039e-05,
+      "loss": 0.469,
+      "step": 236
+    },
+    {
+      "epoch": 0.11325488322083507,
+      "grad_norm": 0.9109628801781714,
+      "learning_rate": 1.5031847133757964e-05,
+      "loss": 0.455,
+      "step": 237
+    },
+    {
+      "epoch": 0.11373275192640822,
+      "grad_norm": 0.8186652665707146,
+      "learning_rate": 1.5095541401273888e-05,
+      "loss": 0.4581,
+      "step": 238
+    },
+    {
+      "epoch": 0.11421062063198137,
+      "grad_norm": 0.9309662613156999,
+      "learning_rate": 1.5159235668789811e-05,
+      "loss": 0.4412,
+      "step": 239
+    },
+    {
+      "epoch": 0.1146884893375545,
+      "grad_norm": 3.931134407255136,
+      "learning_rate": 1.5222929936305735e-05,
+      "loss": 0.4509,
+      "step": 240
+    },
+    {
+      "epoch": 0.11516635804312765,
+      "grad_norm": 1.1827238868781336,
+      "learning_rate": 1.528662420382166e-05,
+      "loss": 0.4574,
+      "step": 241
+    },
+    {
+      "epoch": 0.1156442267487008,
+      "grad_norm": 0.8016143281537751,
+      "learning_rate": 1.5350318471337582e-05,
+      "loss": 0.4493,
+      "step": 242
+    },
+    {
+      "epoch": 0.11612209545427393,
+      "grad_norm": 0.968479721862784,
+      "learning_rate": 1.5414012738853506e-05,
+      "loss": 0.4698,
+      "step": 243
+    },
+    {
+      "epoch": 0.11659996415984708,
+      "grad_norm": 2.2168571117681246,
+      "learning_rate": 1.547770700636943e-05,
+      "loss": 0.4564,
+      "step": 244
+    },
+    {
+      "epoch": 0.11707783286542023,
+      "grad_norm": 0.9881765074402735,
+      "learning_rate": 1.5541401273885352e-05,
+      "loss": 0.4639,
+      "step": 245
+    },
+    {
+      "epoch": 0.11755570157099338,
+      "grad_norm": 0.941404090505303,
+      "learning_rate": 1.5605095541401275e-05,
+      "loss": 0.4631,
+      "step": 246
+    },
+    {
+      "epoch": 0.11803357027656651,
+      "grad_norm": 0.9698016606767922,
+      "learning_rate": 1.56687898089172e-05,
+      "loss": 0.444,
+      "step": 247
+    },
+    {
+      "epoch": 0.11851143898213966,
+      "grad_norm": 0.915195330989377,
+      "learning_rate": 1.5732484076433124e-05,
+      "loss": 0.4581,
+      "step": 248
+    },
+    {
+      "epoch": 0.1189893076877128,
+      "grad_norm": 0.8642148011456234,
+      "learning_rate": 1.5796178343949047e-05,
+      "loss": 0.4506,
+      "step": 249
+    },
+    {
+      "epoch": 0.11946717639328594,
+      "grad_norm": 0.935450575737815,
+      "learning_rate": 1.585987261146497e-05,
+      "loss": 0.461,
+      "step": 250
+    },
+    {
+      "epoch": 0.11994504509885909,
+      "grad_norm": 0.892113620702839,
+      "learning_rate": 1.5923566878980894e-05,
+      "loss": 0.4761,
+      "step": 251
+    },
+    {
+      "epoch": 0.12042291380443224,
+      "grad_norm": 0.9066524856260676,
+      "learning_rate": 1.5987261146496817e-05,
+      "loss": 0.4625,
+      "step": 252
+    },
+    {
+      "epoch": 0.12090078251000537,
+      "grad_norm": 1.1306712148067306,
+      "learning_rate": 1.605095541401274e-05,
+      "loss": 0.4496,
+      "step": 253
+    },
+    {
+      "epoch": 0.12137865121557852,
+      "grad_norm": 1.0415733798385562,
+      "learning_rate": 1.6114649681528666e-05,
+      "loss": 0.448,
+      "step": 254
+    },
+    {
+      "epoch": 0.12185651992115167,
+      "grad_norm": 0.8142111756751625,
+      "learning_rate": 1.617834394904459e-05,
+      "loss": 0.4602,
+      "step": 255
+    },
+    {
+      "epoch": 0.12233438862672481,
+      "grad_norm": 0.9011444127512608,
+      "learning_rate": 1.6242038216560512e-05,
+      "loss": 0.4743,
+      "step": 256
+    },
+    {
+      "epoch": 0.12281225733229795,
+      "grad_norm": 0.8423920386335364,
+      "learning_rate": 1.6305732484076436e-05,
+      "loss": 0.4584,
+      "step": 257
+    },
+    {
+      "epoch": 0.1232901260378711,
+      "grad_norm": 0.8878027469639385,
+      "learning_rate": 1.636942675159236e-05,
+      "loss": 0.4586,
+      "step": 258
+    },
+    {
+      "epoch": 0.12376799474344424,
+      "grad_norm": 0.9181872747662326,
+      "learning_rate": 1.643312101910828e-05,
+      "loss": 0.4559,
+      "step": 259
+    },
+    {
+      "epoch": 0.12424586344901738,
+      "grad_norm": 0.8298428319861421,
+      "learning_rate": 1.6496815286624205e-05,
+      "loss": 0.4719,
+      "step": 260
+    },
+    {
+      "epoch": 0.12472373215459052,
+      "grad_norm": 0.8689700122227426,
+      "learning_rate": 1.6560509554140128e-05,
+      "loss": 0.4837,
+      "step": 261
+    },
+    {
+      "epoch": 0.12520160086016366,
+      "grad_norm": 0.8387758000094565,
+      "learning_rate": 1.6624203821656054e-05,
+      "loss": 0.4727,
+      "step": 262
+    },
+    {
+      "epoch": 0.1256794695657368,
+      "grad_norm": 0.7571078182093177,
+      "learning_rate": 1.6687898089171977e-05,
+      "loss": 0.47,
+      "step": 263
+    },
+    {
+      "epoch": 0.12615733827130995,
+      "grad_norm": 0.9152714319886641,
+      "learning_rate": 1.67515923566879e-05,
+      "loss": 0.4498,
+      "step": 264
+    },
+    {
+      "epoch": 0.1266352069768831,
+      "grad_norm": 0.7676184859264714,
+      "learning_rate": 1.6815286624203824e-05,
+      "loss": 0.452,
+      "step": 265
+    },
+    {
+      "epoch": 0.12711307568245625,
+      "grad_norm": 0.893887271149616,
+      "learning_rate": 1.6878980891719747e-05,
+      "loss": 0.4571,
+      "step": 266
+    },
+    {
+      "epoch": 0.1275909443880294,
+      "grad_norm": 0.7664207431197599,
+      "learning_rate": 1.694267515923567e-05,
+      "loss": 0.4605,
+      "step": 267
+    },
+    {
+      "epoch": 0.12806881309360255,
+      "grad_norm": 0.8462357846724339,
+      "learning_rate": 1.7006369426751593e-05,
+      "loss": 0.4627,
+      "step": 268
+    },
+    {
+      "epoch": 0.12854668179917567,
+      "grad_norm": 0.7917183376178123,
+      "learning_rate": 1.707006369426752e-05,
+      "loss": 0.4587,
+      "step": 269
+    },
+    {
+      "epoch": 0.12902455050474881,
+      "grad_norm": 0.8751109922497367,
+      "learning_rate": 1.7133757961783442e-05,
+      "loss": 0.4593,
+      "step": 270
+    },
+    {
+      "epoch": 0.12950241921032196,
+      "grad_norm": 0.8869857964850081,
+      "learning_rate": 1.7197452229299365e-05,
+      "loss": 0.4455,
+      "step": 271
+    },
+    {
+      "epoch": 0.1299802879158951,
+      "grad_norm": 0.8652417097770274,
+      "learning_rate": 1.726114649681529e-05,
+      "loss": 0.4532,
+      "step": 272
+    },
+    {
+      "epoch": 0.13045815662146826,
+      "grad_norm": 0.8319161291537053,
+      "learning_rate": 1.732484076433121e-05,
+      "loss": 0.4453,
+      "step": 273
+    },
+    {
+      "epoch": 0.1309360253270414,
+      "grad_norm": 0.816357574302165,
+      "learning_rate": 1.7388535031847135e-05,
+      "loss": 0.446,
+      "step": 274
+    },
+    {
+      "epoch": 0.13141389403261453,
+      "grad_norm": 0.8885905160698827,
+      "learning_rate": 1.7452229299363058e-05,
+      "loss": 0.4695,
+      "step": 275
+    },
+    {
+      "epoch": 0.13189176273818767,
+      "grad_norm": 1.0861739881800032,
+      "learning_rate": 1.7515923566878984e-05,
+      "loss": 0.4567,
+      "step": 276
+    },
+    {
+      "epoch": 0.13236963144376082,
+      "grad_norm": 0.7809548635639034,
+      "learning_rate": 1.7579617834394907e-05,
+      "loss": 0.459,
+      "step": 277
+    },
+    {
+      "epoch": 0.13284750014933397,
+      "grad_norm": 0.7635231358009456,
+      "learning_rate": 1.764331210191083e-05,
+      "loss": 0.4589,
+      "step": 278
+    },
+    {
+      "epoch": 0.13332536885490712,
+      "grad_norm": 0.7560665612792802,
+      "learning_rate": 1.7707006369426754e-05,
+      "loss": 0.4528,
+      "step": 279
+    },
+    {
+      "epoch": 0.13380323756048026,
+      "grad_norm": 0.8086543758881466,
+      "learning_rate": 1.7770700636942677e-05,
+      "loss": 0.472,
+      "step": 280
+    },
+    {
+      "epoch": 0.1342811062660534,
+      "grad_norm": 0.7550004521071806,
+      "learning_rate": 1.78343949044586e-05,
+      "loss": 0.4551,
+      "step": 281
+    },
+    {
+      "epoch": 0.13475897497162653,
+      "grad_norm": 0.7788235365516403,
+      "learning_rate": 1.7898089171974523e-05,
+      "loss": 0.4759,
+      "step": 282
+    },
+    {
+      "epoch": 0.13523684367719968,
+      "grad_norm": 0.7296889721103695,
+      "learning_rate": 1.796178343949045e-05,
+      "loss": 0.4504,
+      "step": 283
+    },
+    {
+      "epoch": 0.13571471238277283,
+      "grad_norm": 0.8658688096621077,
+      "learning_rate": 1.8025477707006372e-05,
+      "loss": 0.4645,
+      "step": 284
+    },
+    {
+      "epoch": 0.13619258108834598,
+      "grad_norm": 0.7398631780173166,
+      "learning_rate": 1.8089171974522295e-05,
+      "loss": 0.4476,
+      "step": 285
+    },
+    {
+      "epoch": 0.13667044979391912,
+      "grad_norm": 0.8865227853224859,
+      "learning_rate": 1.815286624203822e-05,
+      "loss": 0.465,
+      "step": 286
+    },
+    {
+      "epoch": 0.13714831849949227,
+      "grad_norm": 0.8311750620796008,
+      "learning_rate": 1.821656050955414e-05,
+      "loss": 0.4652,
+      "step": 287
+    },
+    {
+      "epoch": 0.13762618720506542,
+      "grad_norm": 0.8878939015819043,
+      "learning_rate": 1.8280254777070065e-05,
+      "loss": 0.4714,
+      "step": 288
+    },
+    {
+      "epoch": 0.13810405591063854,
+      "grad_norm": 0.8154300177125134,
+      "learning_rate": 1.8343949044585988e-05,
+      "loss": 0.4458,
+      "step": 289
+    },
+    {
+      "epoch": 0.1385819246162117,
+      "grad_norm": 0.8769838209673224,
+      "learning_rate": 1.8407643312101914e-05,
+      "loss": 0.4425,
+      "step": 290
+    },
+    {
+      "epoch": 0.13905979332178484,
+      "grad_norm": 0.8218246714550715,
+      "learning_rate": 1.8471337579617837e-05,
+      "loss": 0.4555,
+      "step": 291
+    },
+    {
+      "epoch": 0.13953766202735798,
+      "grad_norm": 0.9038739220509564,
+      "learning_rate": 1.853503184713376e-05,
+      "loss": 0.4542,
+      "step": 292
+    },
+    {
+      "epoch": 0.14001553073293113,
+      "grad_norm": 0.9290101054130957,
+      "learning_rate": 1.8598726114649684e-05,
+      "loss": 0.4573,
+      "step": 293
+    },
+    {
+      "epoch": 0.14049339943850428,
+      "grad_norm": 0.8231788424091275,
+      "learning_rate": 1.8662420382165607e-05,
+      "loss": 0.454,
+      "step": 294
+    },
+    {
+      "epoch": 0.14097126814407743,
+      "grad_norm": 0.8068528909020721,
+      "learning_rate": 1.872611464968153e-05,
+      "loss": 0.4703,
+      "step": 295
+    },
+    {
+      "epoch": 0.14144913684965055,
+      "grad_norm": 0.8841324441539453,
+      "learning_rate": 1.8789808917197453e-05,
+      "loss": 0.4576,
+      "step": 296
+    },
+    {
+      "epoch": 0.1419270055552237,
+      "grad_norm": 0.8103277577643994,
+      "learning_rate": 1.8853503184713376e-05,
+      "loss": 0.4624,
+      "step": 297
+    },
+    {
+      "epoch": 0.14240487426079684,
+      "grad_norm": 0.7407111914182409,
+      "learning_rate": 1.8917197452229302e-05,
+      "loss": 0.4539,
+      "step": 298
+    },
+    {
+      "epoch": 0.14288274296637,
+      "grad_norm": 0.8494222480318973,
+      "learning_rate": 1.8980891719745225e-05,
+      "loss": 0.4633,
+      "step": 299
+    },
+    {
+      "epoch": 0.14336061167194314,
+      "grad_norm": 0.760718892931761,
+      "learning_rate": 1.904458598726115e-05,
+      "loss": 0.4491,
+      "step": 300
+    },
+    {
+      "epoch": 0.1438384803775163,
+      "grad_norm": 0.883575953116712,
+      "learning_rate": 1.910828025477707e-05,
+      "loss": 0.4579,
+      "step": 301
+    },
+    {
+      "epoch": 0.14431634908308943,
+      "grad_norm": 0.8706151177176045,
+      "learning_rate": 1.9171974522292995e-05,
+      "loss": 0.4747,
+      "step": 302
+    },
+    {
+      "epoch": 0.14479421778866255,
+      "grad_norm": 0.8387944329782351,
+      "learning_rate": 1.9235668789808918e-05,
+      "loss": 0.4484,
+      "step": 303
+    },
+    {
+      "epoch": 0.1452720864942357,
+      "grad_norm": 1.1446786646678901,
+      "learning_rate": 1.929936305732484e-05,
+      "loss": 0.4694,
+      "step": 304
+    },
+    {
+      "epoch": 0.14574995519980885,
+      "grad_norm": 0.895347496956798,
+      "learning_rate": 1.9363057324840767e-05,
+      "loss": 0.4488,
+      "step": 305
+    },
+    {
+      "epoch": 0.146227823905382,
+      "grad_norm": 1.5869264897706326,
+      "learning_rate": 1.942675159235669e-05,
+      "loss": 0.4478,
+      "step": 306
+    },
+    {
+      "epoch": 0.14670569261095515,
+      "grad_norm": 0.9876414045670644,
+      "learning_rate": 1.9490445859872614e-05,
+      "loss": 0.4593,
+      "step": 307
+    },
+    {
+      "epoch": 0.1471835613165283,
+      "grad_norm": 0.8428176446808334,
+      "learning_rate": 1.9554140127388537e-05,
+      "loss": 0.4603,
+      "step": 308
+    },
+    {
+      "epoch": 0.14766143002210141,
+      "grad_norm": 0.9091407347525033,
+      "learning_rate": 1.961783439490446e-05,
+      "loss": 0.4635,
+      "step": 309
+    },
+    {
+      "epoch": 0.14813929872767456,
+      "grad_norm": 0.8541069012050329,
+      "learning_rate": 1.9681528662420383e-05,
+      "loss": 0.4556,
+      "step": 310
+    },
+    {
+      "epoch": 0.1486171674332477,
+      "grad_norm": 0.8866008034684022,
+      "learning_rate": 1.9745222929936306e-05,
+      "loss": 0.4595,
+      "step": 311
+    },
+    {
+      "epoch": 0.14909503613882086,
+      "grad_norm": 1.0041418570948004,
+      "learning_rate": 1.9808917197452232e-05,
+      "loss": 0.4575,
+      "step": 312
+    },
+    {
+      "epoch": 0.149572904844394,
+      "grad_norm": 0.8608989524709532,
+      "learning_rate": 1.9872611464968155e-05,
+      "loss": 0.4592,
+      "step": 313
+    },
+    {
+      "epoch": 0.15005077354996715,
+      "grad_norm": 0.9193782324674579,
+      "learning_rate": 1.993630573248408e-05,
+      "loss": 0.4702,
+      "step": 314
+    },
+    {
+      "epoch": 0.1505286422555403,
+      "grad_norm": 0.8953098442701494,
+      "learning_rate": 2e-05,
+      "loss": 0.4372,
+      "step": 315
+    },
+    {
+      "epoch": 0.15100651096111342,
+      "grad_norm": 0.9315345476094137,
+      "learning_rate": 1.99999986130882e-05,
+      "loss": 0.4561,
+      "step": 316
+    },
+    {
+      "epoch": 0.15148437966668657,
+      "grad_norm": 0.8193614925815302,
+      "learning_rate": 1.999999445235318e-05,
+      "loss": 0.4419,
+      "step": 317
+    },
+    {
+      "epoch": 0.15196224837225972,
+      "grad_norm": 0.8580436146437501,
+      "learning_rate": 1.9999987517796095e-05,
+      "loss": 0.4517,
+      "step": 318
+    },
+    {
+      "epoch": 0.15244011707783287,
+      "grad_norm": 0.823969742063536,
+      "learning_rate": 1.9999977809418872e-05,
+      "loss": 0.465,
+      "step": 319
+    },
+    {
+      "epoch": 0.152917985783406,
+      "grad_norm": 0.878097134652308,
+      "learning_rate": 1.99999653272242e-05,
+      "loss": 0.4515,
+      "step": 320
+    },
+    {
+      "epoch": 0.15339585448897916,
+      "grad_norm": 0.8475324958361699,
+      "learning_rate": 1.999995007121554e-05,
+      "loss": 0.4491,
+      "step": 321
+    },
+    {
+      "epoch": 0.1538737231945523,
+      "grad_norm": 0.8481861989745667,
+      "learning_rate": 1.9999932041397128e-05,
+      "loss": 0.4623,
+      "step": 322
+    },
+    {
+      "epoch": 0.15435159190012543,
+      "grad_norm": 0.7962447154311731,
+      "learning_rate": 1.9999911237773964e-05,
+      "loss": 0.4525,
+      "step": 323
+    },
+    {
+      "epoch": 0.15482946060569858,
+      "grad_norm": 0.7407282750796474,
+      "learning_rate": 1.999988766035182e-05,
+      "loss": 0.4465,
+      "step": 324
+    },
+    {
+      "epoch": 0.15530732931127172,
+      "grad_norm": 0.8298386043292453,
+      "learning_rate": 1.9999861309137232e-05,
+      "loss": 0.4542,
+      "step": 325
+    },
+    {
+      "epoch": 0.15578519801684487,
+      "grad_norm": 0.7485540031873075,
+      "learning_rate": 1.999983218413751e-05,
+      "loss": 0.4413,
+      "step": 326
+    },
+    {
+      "epoch": 0.15626306672241802,
+      "grad_norm": 1.5089153257896015,
+      "learning_rate": 1.9999800285360736e-05,
+      "loss": 0.4631,
+      "step": 327
+    },
+    {
+      "epoch": 0.15674093542799117,
+      "grad_norm": 0.8158254057514103,
+      "learning_rate": 1.999976561281576e-05,
+      "loss": 0.4548,
+      "step": 328
+    },
+    {
+      "epoch": 0.15721880413356432,
+      "grad_norm": 0.8034746528062102,
+      "learning_rate": 1.9999728166512187e-05,
+      "loss": 0.4399,
+      "step": 329
+    },
+    {
+      "epoch": 0.15769667283913744,
+      "grad_norm": 0.7519006657356639,
+      "learning_rate": 1.999968794646042e-05,
+      "loss": 0.4503,
+      "step": 330
+    },
+    {
+      "epoch": 0.15817454154471058,
+      "grad_norm": 0.8336919761973133,
+      "learning_rate": 1.9999644952671604e-05,
+      "loss": 0.4574,
+      "step": 331
+    },
+    {
+      "epoch": 0.15865241025028373,
+      "grad_norm": 0.7895923256720561,
+      "learning_rate": 1.9999599185157673e-05,
+      "loss": 0.445,
+      "step": 332
+    },
+    {
+      "epoch": 0.15913027895585688,
+      "grad_norm": 0.8990492990399171,
+      "learning_rate": 1.9999550643931313e-05,
+      "loss": 0.4542,
+      "step": 333
+    },
+    {
+      "epoch": 0.15960814766143003,
+      "grad_norm": 0.7540142922122045,
+      "learning_rate": 1.9999499329005995e-05,
+      "loss": 0.4701,
+      "step": 334
+    },
+    {
+      "epoch": 0.16008601636700318,
+      "grad_norm": 0.8744480175459001,
+      "learning_rate": 1.9999445240395953e-05,
+      "loss": 0.4587,
+      "step": 335
+    },
+    {
+      "epoch": 0.16056388507257632,
+      "grad_norm": 0.7760610520617973,
+      "learning_rate": 1.9999388378116186e-05,
+      "loss": 0.4431,
+      "step": 336
+    },
+    {
+      "epoch": 0.16104175377814944,
+      "grad_norm": 0.9367580126034849,
+      "learning_rate": 1.9999328742182472e-05,
+      "loss": 0.4471,
+      "step": 337
+    },
+    {
+      "epoch": 0.1615196224837226,
+      "grad_norm": 0.8695840825168545,
+      "learning_rate": 1.999926633261135e-05,
+      "loss": 0.4369,
+      "step": 338
+    },
+    {
+      "epoch": 0.16199749118929574,
+      "grad_norm": 0.8085802780374626,
+      "learning_rate": 1.9999201149420128e-05,
+      "loss": 0.4412,
+      "step": 339
+    },
+    {
+      "epoch": 0.1624753598948689,
+      "grad_norm": 0.9521212177197098,
+      "learning_rate": 1.9999133192626893e-05,
+      "loss": 0.434,
+      "step": 340
+    },
+    {
+      "epoch": 0.16295322860044203,
+      "grad_norm": 0.821086659719905,
+      "learning_rate": 1.999906246225049e-05,
+      "loss": 0.4601,
+      "step": 341
+    },
+    {
+      "epoch": 0.16343109730601518,
+      "grad_norm": 0.8520130760092065,
+      "learning_rate": 1.9998988958310542e-05,
+      "loss": 0.4604,
+      "step": 342
+    },
+    {
+      "epoch": 0.1639089660115883,
+      "grad_norm": 0.90728532792054,
+      "learning_rate": 1.9998912680827436e-05,
+      "loss": 0.44,
+      "step": 343
+    },
+    {
+      "epoch": 0.16438683471716145,
+      "grad_norm": 0.8144812444652371,
+      "learning_rate": 1.9998833629822328e-05,
+      "loss": 0.4398,
+      "step": 344
+    },
+    {
+      "epoch": 0.1648647034227346,
+      "grad_norm": 0.9394455637761423,
+      "learning_rate": 1.9998751805317152e-05,
+      "loss": 0.4472,
+      "step": 345
+    },
+    {
+      "epoch": 0.16534257212830775,
+      "grad_norm": 0.7942127753027973,
+      "learning_rate": 1.9998667207334596e-05,
+      "loss": 0.4518,
+      "step": 346
+    },
+    {
+      "epoch": 0.1658204408338809,
+      "grad_norm": 0.8729986142728988,
+      "learning_rate": 1.9998579835898135e-05,
+      "loss": 0.4671,
+      "step": 347
+    },
+    {
+      "epoch": 0.16629830953945404,
+      "grad_norm": 0.8038525219516176,
+      "learning_rate": 1.9998489691031994e-05,
+      "loss": 0.4475,
+      "step": 348
+    },
+    {
+      "epoch": 0.1667761782450272,
+      "grad_norm": 0.7793008423025405,
+      "learning_rate": 1.9998396772761186e-05,
+      "loss": 0.4397,
+      "step": 349
+    },
+    {
+      "epoch": 0.1672540469506003,
+      "grad_norm": 0.8347706044801682,
+      "learning_rate": 1.999830108111148e-05,
+      "loss": 0.4471,
+      "step": 350
+    },
+    {
+      "epoch": 0.16773191565617346,
+      "grad_norm": 0.775692619220453,
+      "learning_rate": 1.999820261610942e-05,
+      "loss": 0.4539,
+      "step": 351
+    },
+    {
+      "epoch": 0.1682097843617466,
+      "grad_norm": 0.823420887104832,
+      "learning_rate": 1.9998101377782322e-05,
+      "loss": 0.4544,
+      "step": 352
+    },
+    {
+      "epoch": 0.16868765306731975,
+      "grad_norm": 0.81678869462245,
+      "learning_rate": 1.9997997366158264e-05,
+      "loss": 0.4365,
+      "step": 353
+    },
+    {
+      "epoch": 0.1691655217728929,
+      "grad_norm": 0.7743470553025751,
+      "learning_rate": 1.99978905812661e-05,
+      "loss": 0.4554,
+      "step": 354
+    },
+    {
+      "epoch": 0.16964339047846605,
+      "grad_norm": 0.7641953458649339,
+      "learning_rate": 1.999778102313545e-05,
+      "loss": 0.4578,
+      "step": 355
+    },
+    {
+      "epoch": 0.1701212591840392,
+      "grad_norm": 1.0296472768958496,
+      "learning_rate": 1.99976686917967e-05,
+      "loss": 0.4389,
+      "step": 356
+    },
+    {
+      "epoch": 0.17059912788961232,
+      "grad_norm": 0.7672656783937326,
+      "learning_rate": 1.9997553587281012e-05,
+      "loss": 0.4253,
+      "step": 357
+    },
+    {
+      "epoch": 0.17107699659518547,
+      "grad_norm": 0.7273787482171606,
+      "learning_rate": 1.999743570962031e-05,
+      "loss": 0.4458,
+      "step": 358
+    },
+    {
+      "epoch": 0.1715548653007586,
+      "grad_norm": 0.8352123781017939,
+      "learning_rate": 1.9997315058847296e-05,
+      "loss": 0.4449,
+      "step": 359
+    },
+    {
+      "epoch": 0.17203273400633176,
+      "grad_norm": 0.8661207476943021,
+      "learning_rate": 1.999719163499543e-05,
+      "loss": 0.4416,
+      "step": 360
+    },
+    {
+      "epoch": 0.1725106027119049,
+      "grad_norm": 0.814549911316121,
+      "learning_rate": 1.999706543809896e-05,
+      "loss": 0.4512,
+      "step": 361
+    },
+    {
+      "epoch": 0.17298847141747806,
+      "grad_norm": 0.783161823300539,
+      "learning_rate": 1.9996936468192874e-05,
+      "loss": 0.4426,
+      "step": 362
+    },
+    {
+      "epoch": 0.1734663401230512,
+      "grad_norm": 0.8535445094255457,
+      "learning_rate": 1.9996804725312963e-05,
+      "loss": 0.4492,
+      "step": 363
+    },
+    {
+      "epoch": 0.17394420882862432,
+      "grad_norm": 0.8668749659680126,
+      "learning_rate": 1.9996670209495757e-05,
+      "loss": 0.4574,
+      "step": 364
+    },
+    {
+      "epoch": 0.17442207753419747,
+      "grad_norm": 0.9384280918032981,
+      "learning_rate": 1.999653292077857e-05,
+      "loss": 0.4308,
+      "step": 365
+    },
+    {
+      "epoch": 0.17489994623977062,
+      "grad_norm": 0.8148125307985487,
+      "learning_rate": 1.999639285919949e-05,
+      "loss": 0.4501,
+      "step": 366
+    },
+    {
+      "epoch": 0.17537781494534377,
+      "grad_norm": 0.8327358415612538,
+      "learning_rate": 1.9996250024797364e-05,
+      "loss": 0.4318,
+      "step": 367
+    },
+    {
+      "epoch": 0.17585568365091692,
+      "grad_norm": 0.8013196691254347,
+      "learning_rate": 1.9996104417611815e-05,
+      "loss": 0.4501,
+      "step": 368
+    },
+    {
+      "epoch": 0.17633355235649006,
+      "grad_norm": 0.7247950871190025,
+      "learning_rate": 1.9995956037683225e-05,
+      "loss": 0.4371,
+      "step": 369
+    },
+    {
+      "epoch": 0.17681142106206318,
+      "grad_norm": 0.8177374937697399,
+      "learning_rate": 1.999580488505276e-05,
+      "loss": 0.4561,
+      "step": 370
+    },
+    {
+      "epoch": 0.17728928976763633,
+      "grad_norm": 0.7929393091666411,
+      "learning_rate": 1.999565095976234e-05,
+      "loss": 0.454,
+      "step": 371
+    },
+    {
+      "epoch": 0.17776715847320948,
+      "grad_norm": 0.7843978974842676,
+      "learning_rate": 1.999549426185466e-05,
+      "loss": 0.4514,
+      "step": 372
+    },
+    {
+      "epoch": 0.17824502717878263,
+      "grad_norm": 0.7290779188482008,
+      "learning_rate": 1.9995334791373194e-05,
+      "loss": 0.456,
+      "step": 373
+    },
+    {
+      "epoch": 0.17872289588435578,
+      "grad_norm": 0.7580582343922024,
+      "learning_rate": 1.9995172548362172e-05,
+      "loss": 0.4519,
+      "step": 374
+    },
+    {
+      "epoch": 0.17920076458992892,
+      "grad_norm": 0.8860454919596451,
+      "learning_rate": 1.9995007532866594e-05,
+      "loss": 0.4356,
+      "step": 375
+    },
+    {
+      "epoch": 0.17967863329550207,
+      "grad_norm": 0.9228385090291951,
+      "learning_rate": 1.9994839744932237e-05,
+      "loss": 0.4411,
+      "step": 376
+    },
+    {
+      "epoch": 0.1801565020010752,
+      "grad_norm": 0.7361256362322965,
+      "learning_rate": 1.9994669184605642e-05,
+      "loss": 0.4644,
+      "step": 377
+    },
+    {
+      "epoch": 0.18063437070664834,
+      "grad_norm": 0.7285467924733723,
+      "learning_rate": 1.9994495851934116e-05,
+      "loss": 0.4504,
+      "step": 378
+    },
+    {
+      "epoch": 0.1811122394122215,
+      "grad_norm": 0.6852384080647346,
+      "learning_rate": 1.9994319746965743e-05,
+      "loss": 0.4367,
+      "step": 379
+    },
+    {
+      "epoch": 0.18159010811779464,
+      "grad_norm": 0.7016915322810489,
+      "learning_rate": 1.9994140869749366e-05,
+      "loss": 0.4423,
+      "step": 380
+    },
+    {
+      "epoch": 0.18206797682336778,
+      "grad_norm": 0.7419915375299211,
+      "learning_rate": 1.9993959220334608e-05,
+      "loss": 0.4319,
+      "step": 381
+    },
+    {
+      "epoch": 0.18254584552894093,
+      "grad_norm": 0.691366860088611,
+      "learning_rate": 1.999377479877185e-05,
+      "loss": 0.4396,
+      "step": 382
+    },
+    {
+      "epoch": 0.18302371423451408,
+      "grad_norm": 0.8101619011356389,
+      "learning_rate": 1.9993587605112252e-05,
+      "loss": 0.437,
+      "step": 383
+    },
+    {
+      "epoch": 0.1835015829400872,
+      "grad_norm": 0.7680909891808453,
+      "learning_rate": 1.9993397639407736e-05,
+      "loss": 0.4415,
+      "step": 384
+    },
+    {
+      "epoch": 0.18397945164566035,
+      "grad_norm": 0.7391098105965442,
+      "learning_rate": 1.9993204901710995e-05,
+      "loss": 0.4519,
+      "step": 385
+    },
+    {
+      "epoch": 0.1844573203512335,
+      "grad_norm": 0.7022079848845771,
+      "learning_rate": 1.999300939207549e-05,
+      "loss": 0.4378,
+      "step": 386
+    },
+    {
+      "epoch": 0.18493518905680664,
+      "grad_norm": 0.8413020737595335,
+      "learning_rate": 1.999281111055545e-05,
+      "loss": 0.4259,
+      "step": 387
+    },
+    {
+      "epoch": 0.1854130577623798,
+      "grad_norm": 0.675966589527395,
+      "learning_rate": 1.9992610057205888e-05,
+      "loss": 0.4401,
+      "step": 388
+    },
+    {
+      "epoch": 0.18589092646795294,
+      "grad_norm": 0.7936827487639535,
+      "learning_rate": 1.9992406232082557e-05,
+      "loss": 0.4364,
+      "step": 389
+    },
+    {
+      "epoch": 0.1863687951735261,
+      "grad_norm": 0.6543232226417673,
+      "learning_rate": 1.9992199635241997e-05,
+      "loss": 0.4505,
+      "step": 390
+    },
+    {
+      "epoch": 0.1868466638790992,
+      "grad_norm": 0.8239352462676031,
+      "learning_rate": 1.9991990266741524e-05,
+      "loss": 0.4405,
+      "step": 391
+    },
+    {
+      "epoch": 0.18732453258467235,
+      "grad_norm": 0.7573116961718065,
+      "learning_rate": 1.9991778126639202e-05,
+      "loss": 0.4219,
+      "step": 392
+    },
+    {
+      "epoch": 0.1878024012902455,
+      "grad_norm": 0.7835516359179223,
+      "learning_rate": 1.9991563214993885e-05,
+      "loss": 0.4374,
+      "step": 393
+    },
+    {
+      "epoch": 0.18828026999581865,
+      "grad_norm": 0.7604499069077229,
+      "learning_rate": 1.9991345531865173e-05,
+      "loss": 0.4589,
+      "step": 394
+    },
+    {
+      "epoch": 0.1887581387013918,
+      "grad_norm": 0.76870199204744,
+      "learning_rate": 1.999112507731346e-05,
+      "loss": 0.4508,
+      "step": 395
+    },
+    {
+      "epoch": 0.18923600740696495,
+      "grad_norm": 0.6786887636756039,
+      "learning_rate": 1.999090185139989e-05,
+      "loss": 0.4342,
+      "step": 396
+    },
+    {
+      "epoch": 0.1897138761125381,
+      "grad_norm": 0.6999332563120491,
+      "learning_rate": 1.9990675854186384e-05,
+      "loss": 0.4407,
+      "step": 397
+    },
+    {
+      "epoch": 0.1901917448181112,
+      "grad_norm": 0.8334575931899786,
+      "learning_rate": 1.9990447085735624e-05,
+      "loss": 0.4429,
+      "step": 398
+    },
+    {
+      "epoch": 0.19066961352368436,
+      "grad_norm": 0.7685745229409217,
+      "learning_rate": 1.9990215546111074e-05,
+      "loss": 0.4398,
+      "step": 399
+    },
+    {
+      "epoch": 0.1911474822292575,
+      "grad_norm": 0.7005717268603726,
+      "learning_rate": 1.9989981235376956e-05,
+      "loss": 0.4486,
+      "step": 400
+    },
+    {
+      "epoch": 0.19162535093483066,
+      "grad_norm": 0.6603349007597211,
+      "learning_rate": 1.9989744153598264e-05,
+      "loss": 0.4581,
+      "step": 401
+    },
+    {
+      "epoch": 0.1921032196404038,
+      "grad_norm": 0.7880156496415623,
+      "learning_rate": 1.998950430084076e-05,
+      "loss": 0.4381,
+      "step": 402
+    },
+    {
+      "epoch": 0.19258108834597695,
+      "grad_norm": 0.7736124963496666,
+      "learning_rate": 1.998926167717097e-05,
+      "loss": 0.4202,
+      "step": 403
+    },
+    {
+      "epoch": 0.19305895705155007,
+      "grad_norm": 0.7008884273159984,
+      "learning_rate": 1.9989016282656207e-05,
+      "loss": 0.4539,
+      "step": 404
+    },
+    {
+      "epoch": 0.19353682575712322,
+      "grad_norm": 0.7575001466030076,
+      "learning_rate": 1.9988768117364526e-05,
+      "loss": 0.4367,
+      "step": 405
+    },
+    {
+      "epoch": 0.19401469446269637,
+      "grad_norm": 0.7446911712881855,
+      "learning_rate": 1.9988517181364767e-05,
+      "loss": 0.4587,
+      "step": 406
+    },
+    {
+      "epoch": 0.19449256316826952,
+      "grad_norm": 0.7076619424650085,
+      "learning_rate": 1.9988263474726536e-05,
+      "loss": 0.4322,
+      "step": 407
+    },
+    {
+      "epoch": 0.19497043187384266,
+      "grad_norm": 0.7386357724353199,
+      "learning_rate": 1.9988006997520208e-05,
+      "loss": 0.446,
+      "step": 408
+    },
+    {
+      "epoch": 0.1954483005794158,
+      "grad_norm": 0.7180682823947508,
+      "learning_rate": 1.9987747749816923e-05,
+      "loss": 0.4367,
+      "step": 409
+    },
+    {
+      "epoch": 0.19592616928498896,
+      "grad_norm": 0.8350524472024324,
+      "learning_rate": 1.9987485731688595e-05,
+      "loss": 0.4691,
+      "step": 410
+    },
+    {
+      "epoch": 0.19640403799056208,
+      "grad_norm": 0.7262068323613972,
+      "learning_rate": 1.9987220943207903e-05,
+      "loss": 0.4412,
+      "step": 411
+    },
+    {
+      "epoch": 0.19688190669613523,
+      "grad_norm": 0.7033003276566966,
+      "learning_rate": 1.998695338444829e-05,
+      "loss": 0.4516,
+      "step": 412
+    },
+    {
+      "epoch": 0.19735977540170838,
+      "grad_norm": 0.740747812864168,
+      "learning_rate": 1.9986683055483975e-05,
+      "loss": 0.4444,
+      "step": 413
+    },
+    {
+      "epoch": 0.19783764410728152,
+      "grad_norm": 0.8106322648081262,
+      "learning_rate": 1.9986409956389946e-05,
+      "loss": 0.421,
+      "step": 414
+    },
+    {
+      "epoch": 0.19831551281285467,
+      "grad_norm": 1.343586386680477,
+      "learning_rate": 1.998613408724195e-05,
+      "loss": 0.4443,
+      "step": 415
+    },
+    {
+      "epoch": 0.19879338151842782,
+      "grad_norm": 0.8243575543930032,
+      "learning_rate": 1.9985855448116507e-05,
+      "loss": 0.4256,
+      "step": 416
+    },
+    {
+      "epoch": 0.19927125022400097,
+      "grad_norm": 0.7616433208806905,
+      "learning_rate": 1.9985574039090912e-05,
+      "loss": 0.4333,
+      "step": 417
+    },
+    {
+      "epoch": 0.1997491189295741,
+      "grad_norm": 0.8197988566665843,
+      "learning_rate": 1.9985289860243222e-05,
+      "loss": 0.4292,
+      "step": 418
+    },
+    {
+      "epoch": 0.20022698763514724,
+      "grad_norm": 0.6780749889189807,
+      "learning_rate": 1.9985002911652262e-05,
+      "loss": 0.4454,
+      "step": 419
+    },
+    {
+      "epoch": 0.20070485634072038,
+      "grad_norm": 0.8323686570996995,
+      "learning_rate": 1.998471319339763e-05,
+      "loss": 0.4357,
+      "step": 420
+    },
+    {
+      "epoch": 0.20118272504629353,
+      "grad_norm": 0.7320142206171304,
+      "learning_rate": 1.998442070555968e-05,
+      "loss": 0.4145,
+      "step": 421
+    },
+    {
+      "epoch": 0.20166059375186668,
+      "grad_norm": 0.782381732360279,
+      "learning_rate": 1.9984125448219555e-05,
+      "loss": 0.4424,
+      "step": 422
+    },
+    {
+      "epoch": 0.20213846245743983,
+      "grad_norm": 0.748058238639162,
+      "learning_rate": 1.998382742145914e-05,
+      "loss": 0.4426,
+      "step": 423
+    },
+    {
+      "epoch": 0.20261633116301297,
+      "grad_norm": 0.7799942137521412,
+      "learning_rate": 1.9983526625361115e-05,
+      "loss": 0.4427,
+      "step": 424
+    },
+    {
+      "epoch": 0.2030941998685861,
+      "grad_norm": 0.703037954101251,
+      "learning_rate": 1.9983223060008908e-05,
+      "loss": 0.4298,
+      "step": 425
+    },
+    {
+      "epoch": 0.20357206857415924,
+      "grad_norm": 0.7344901897075886,
+      "learning_rate": 1.998291672548673e-05,
+      "loss": 0.4369,
+      "step": 426
+    },
+    {
+      "epoch": 0.2040499372797324,
+      "grad_norm": 0.743895441020455,
+      "learning_rate": 1.9982607621879545e-05,
+      "loss": 0.4235,
+      "step": 427
+    },
+    {
+      "epoch": 0.20452780598530554,
+      "grad_norm": 0.6646898981475607,
+      "learning_rate": 1.9982295749273093e-05,
+      "loss": 0.4378,
+      "step": 428
+    },
+    {
+      "epoch": 0.2050056746908787,
+      "grad_norm": 0.7093024591727115,
+      "learning_rate": 1.998198110775389e-05,
+      "loss": 0.428,
+      "step": 429
+    },
+    {
+      "epoch": 0.20548354339645183,
+      "grad_norm": 0.7212954419357257,
+      "learning_rate": 1.9981663697409203e-05,
+      "loss": 0.4356,
+      "step": 430
+    },
+    {
+      "epoch": 0.20596141210202498,
+      "grad_norm": 0.723396603925273,
+      "learning_rate": 1.998134351832708e-05,
+      "loss": 0.4309,
+      "step": 431
+    },
+    {
+      "epoch": 0.2064392808075981,
+      "grad_norm": 0.7369743637876824,
+      "learning_rate": 1.9981020570596334e-05,
+      "loss": 0.446,
+      "step": 432
+    },
+    {
+      "epoch": 0.20691714951317125,
+      "grad_norm": 0.7097781588794618,
+      "learning_rate": 1.9980694854306545e-05,
+      "loss": 0.4302,
+      "step": 433
+    },
+    {
+      "epoch": 0.2073950182187444,
+      "grad_norm": 0.6844965007718748,
+      "learning_rate": 1.998036636954806e-05,
+      "loss": 0.4293,
+      "step": 434
+    },
+    {
+      "epoch": 0.20787288692431755,
+      "grad_norm": 0.7056310983568724,
+      "learning_rate": 1.998003511641199e-05,
+      "loss": 0.4541,
+      "step": 435
+    },
+    {
+      "epoch": 0.2083507556298907,
+      "grad_norm": 0.6540869976167216,
+      "learning_rate": 1.9979701094990226e-05,
+      "loss": 0.4187,
+      "step": 436
+    },
+    {
+      "epoch": 0.20882862433546384,
+      "grad_norm": 0.7375068029416245,
+      "learning_rate": 1.997936430537542e-05,
+      "loss": 0.4377,
+      "step": 437
+    },
+    {
+      "epoch": 0.20930649304103696,
+      "grad_norm": 0.6814790625160986,
+      "learning_rate": 1.9979024747660985e-05,
+      "loss": 0.4445,
+      "step": 438
+    },
+    {
+      "epoch": 0.2097843617466101,
+      "grad_norm": 0.7103604757794988,
+      "learning_rate": 1.9978682421941114e-05,
+      "loss": 0.4313,
+      "step": 439
+    },
+    {
+      "epoch": 0.21026223045218326,
+      "grad_norm": 0.7165788798454129,
+      "learning_rate": 1.997833732831076e-05,
+      "loss": 0.4419,
+      "step": 440
+    },
+    {
+      "epoch": 0.2107400991577564,
+      "grad_norm": 0.7673501597057077,
+      "learning_rate": 1.9977989466865645e-05,
+      "loss": 0.435,
+      "step": 441
+    },
+    {
+      "epoch": 0.21121796786332955,
+      "grad_norm": 0.6931708197147421,
+      "learning_rate": 1.9977638837702263e-05,
+      "loss": 0.4317,
+      "step": 442
+    },
+    {
+      "epoch": 0.2116958365689027,
+      "grad_norm": 0.6687440502755292,
+      "learning_rate": 1.997728544091787e-05,
+      "loss": 0.4575,
+      "step": 443
+    },
+    {
+      "epoch": 0.21217370527447585,
+      "grad_norm": 0.7234453514272886,
+      "learning_rate": 1.997692927661049e-05,
+      "loss": 0.4186,
+      "step": 444
+    },
+    {
+      "epoch": 0.21265157398004897,
+      "grad_norm": 0.8989542148130306,
+      "learning_rate": 1.9976570344878916e-05,
+      "loss": 0.4277,
+      "step": 445
+    },
+    {
+      "epoch": 0.21312944268562212,
+      "grad_norm": 0.6798823076977999,
+      "learning_rate": 1.9976208645822716e-05,
+      "loss": 0.4203,
+      "step": 446
+    },
+    {
+      "epoch": 0.21360731139119526,
+      "grad_norm": 0.7297558856068362,
+      "learning_rate": 1.9975844179542213e-05,
+      "loss": 0.4361,
+      "step": 447
+    },
+    {
+      "epoch": 0.2140851800967684,
+      "grad_norm": 0.7211293256955031,
+      "learning_rate": 1.9975476946138506e-05,
+      "loss": 0.4296,
+      "step": 448
+    },
+    {
+      "epoch": 0.21456304880234156,
+      "grad_norm": 0.7882118710161115,
+      "learning_rate": 1.997510694571346e-05,
+      "loss": 0.4462,
+      "step": 449
+    },
+    {
+      "epoch": 0.2150409175079147,
+      "grad_norm": 0.7179853617013857,
+      "learning_rate": 1.9974734178369702e-05,
+      "loss": 0.4296,
+      "step": 450
+    },
+    {
+      "epoch": 0.21551878621348786,
+      "grad_norm": 0.7095782872691374,
+      "learning_rate": 1.9974358644210635e-05,
+      "loss": 0.4411,
+      "step": 451
+    },
+    {
+      "epoch": 0.21599665491906098,
+      "grad_norm": 0.6772035963394542,
+      "learning_rate": 1.997398034334043e-05,
+      "loss": 0.4317,
+      "step": 452
+    },
+    {
+      "epoch": 0.21647452362463412,
+      "grad_norm": 0.6743627379910052,
+      "learning_rate": 1.997359927586401e-05,
+      "loss": 0.4401,
+      "step": 453
+    },
+    {
+      "epoch": 0.21695239233020727,
+      "grad_norm": 0.7837750007770807,
+      "learning_rate": 1.9973215441887085e-05,
+      "loss": 0.4448,
+      "step": 454
+    },
+    {
+      "epoch": 0.21743026103578042,
+      "grad_norm": 0.7047243533444423,
+      "learning_rate": 1.997282884151612e-05,
+      "loss": 0.4473,
+      "step": 455
+    },
+    {
+      "epoch": 0.21790812974135357,
+      "grad_norm": 1.4379264367718632,
+      "learning_rate": 1.9972439474858348e-05,
+      "loss": 0.454,
+      "step": 456
+    },
+    {
+      "epoch": 0.21838599844692672,
+      "grad_norm": 0.8191228771916826,
+      "learning_rate": 1.9972047342021784e-05,
+      "loss": 0.4322,
+      "step": 457
+    },
+    {
+      "epoch": 0.21886386715249986,
+      "grad_norm": 0.7133604179165115,
+      "learning_rate": 1.9971652443115186e-05,
+      "loss": 0.4153,
+      "step": 458
+    },
+    {
+      "epoch": 0.21934173585807298,
+      "grad_norm": 0.6701729667469203,
+      "learning_rate": 1.99712547782481e-05,
+      "loss": 0.4286,
+      "step": 459
+    },
+    {
+      "epoch": 0.21981960456364613,
+      "grad_norm": 0.7794564597547953,
+      "learning_rate": 1.9970854347530828e-05,
+      "loss": 0.4385,
+      "step": 460
+    },
+    {
+      "epoch": 0.22029747326921928,
+      "grad_norm": 0.8024059316138707,
+      "learning_rate": 1.9970451151074442e-05,
+      "loss": 0.4374,
+      "step": 461
+    },
+    {
+      "epoch": 0.22077534197479243,
+      "grad_norm": 0.6631388176963441,
+      "learning_rate": 1.997004518899078e-05,
+      "loss": 0.4368,
+      "step": 462
+    },
+    {
+      "epoch": 0.22125321068036558,
+      "grad_norm": 0.7042713011158211,
+      "learning_rate": 1.9969636461392454e-05,
+      "loss": 0.4398,
+      "step": 463
+    },
+    {
+      "epoch": 0.22173107938593872,
+      "grad_norm": 0.7010671119076932,
+      "learning_rate": 1.9969224968392837e-05,
+      "loss": 0.4541,
+      "step": 464
+    },
+    {
+      "epoch": 0.22220894809151187,
+      "grad_norm": 0.7354792559659475,
+      "learning_rate": 1.9968810710106065e-05,
+      "loss": 0.4397,
+      "step": 465
+    },
+    {
+      "epoch": 0.222686816797085,
+      "grad_norm": 0.7095812173991619,
+      "learning_rate": 1.9968393686647046e-05,
+      "loss": 0.4487,
+      "step": 466
+    },
+    {
+      "epoch": 0.22316468550265814,
+      "grad_norm": 0.6677575026918666,
+      "learning_rate": 1.9967973898131462e-05,
+      "loss": 0.4468,
+      "step": 467
+    },
+    {
+      "epoch": 0.2236425542082313,
+      "grad_norm": 0.7939463667123086,
+      "learning_rate": 1.9967551344675752e-05,
+      "loss": 0.4281,
+      "step": 468
+    },
+    {
+      "epoch": 0.22412042291380443,
+      "grad_norm": 0.7618631702940217,
+      "learning_rate": 1.996712602639712e-05,
+      "loss": 0.4283,
+      "step": 469
+    },
+    {
+      "epoch": 0.22459829161937758,
+      "grad_norm": 0.7333732414541805,
+      "learning_rate": 1.9966697943413548e-05,
+      "loss": 0.4305,
+      "step": 470
+    },
+    {
+      "epoch": 0.22507616032495073,
+      "grad_norm": 0.7779594408406364,
+      "learning_rate": 1.9966267095843776e-05,
+      "loss": 0.4052,
+      "step": 471
+    },
+    {
+      "epoch": 0.22555402903052385,
+      "grad_norm": 0.6757227788709603,
+      "learning_rate": 1.996583348380731e-05,
+      "loss": 0.4196,
+      "step": 472
+    },
+    {
+      "epoch": 0.226031897736097,
+      "grad_norm": 0.7580164038914531,
+      "learning_rate": 1.9965397107424434e-05,
+      "loss": 0.404,
+      "step": 473
+    },
+    {
+      "epoch": 0.22650976644167015,
+      "grad_norm": 0.7558024725157709,
+      "learning_rate": 1.9964957966816184e-05,
+      "loss": 0.4426,
+      "step": 474
+    },
+    {
+      "epoch": 0.2269876351472433,
+      "grad_norm": 0.780681584126318,
+      "learning_rate": 1.9964516062104377e-05,
+      "loss": 0.4283,
+      "step": 475
+    },
+    {
+      "epoch": 0.22746550385281644,
+      "grad_norm": 0.9899203517841794,
+      "learning_rate": 1.996407139341158e-05,
+      "loss": 0.4241,
+      "step": 476
+    },
+    {
+      "epoch": 0.2279433725583896,
+      "grad_norm": 0.6521695714753446,
+      "learning_rate": 1.9963623960861144e-05,
+      "loss": 0.4198,
+      "step": 477
+    },
+    {
+      "epoch": 0.22842124126396274,
+      "grad_norm": 0.725943514393016,
+      "learning_rate": 1.9963173764577178e-05,
+      "loss": 0.4447,
+      "step": 478
+    },
+    {
+      "epoch": 0.22889910996953586,
+      "grad_norm": 0.6805293528269828,
+      "learning_rate": 1.9962720804684555e-05,
+      "loss": 0.4276,
+      "step": 479
+    },
+    {
+      "epoch": 0.229376978675109,
+      "grad_norm": 0.7471831254807846,
+      "learning_rate": 1.996226508130892e-05,
+      "loss": 0.4209,
+      "step": 480
+    },
+    {
+      "epoch": 0.22985484738068215,
+      "grad_norm": 0.6443809058619084,
+      "learning_rate": 1.9961806594576684e-05,
+      "loss": 0.4351,
+      "step": 481
+    },
+    {
+      "epoch": 0.2303327160862553,
+      "grad_norm": 1.3415000224918492,
+      "learning_rate": 1.996134534461502e-05,
+      "loss": 0.459,
+      "step": 482
+    },
+    {
+      "epoch": 0.23081058479182845,
+      "grad_norm": 0.7134420590449075,
+      "learning_rate": 1.996088133155188e-05,
+      "loss": 0.4387,
+      "step": 483
+    },
+    {
+      "epoch": 0.2312884534974016,
+      "grad_norm": 0.7813274776781214,
+      "learning_rate": 1.9960414555515958e-05,
+      "loss": 0.4126,
+      "step": 484
+    },
+    {
+      "epoch": 0.23176632220297474,
+      "grad_norm": 0.6898643605397596,
+      "learning_rate": 1.995994501663674e-05,
+      "loss": 0.4246,
+      "step": 485
+    },
+    {
+      "epoch": 0.23224419090854787,
+      "grad_norm": 0.9121587741249505,
+      "learning_rate": 1.9959472715044463e-05,
+      "loss": 0.4244,
+      "step": 486
+    },
+    {
+      "epoch": 0.232722059614121,
+      "grad_norm": 0.7147251675369962,
+      "learning_rate": 1.9958997650870137e-05,
+      "loss": 0.4224,
+      "step": 487
+    },
+    {
+      "epoch": 0.23319992831969416,
+      "grad_norm": 0.7899586618920265,
+      "learning_rate": 1.9958519824245536e-05,
+      "loss": 0.4112,
+      "step": 488
+    },
+    {
+      "epoch": 0.2336777970252673,
+      "grad_norm": 0.7904708831006497,
+      "learning_rate": 1.99580392353032e-05,
+      "loss": 0.4389,
+      "step": 489
+    },
+    {
+      "epoch": 0.23415566573084046,
+      "grad_norm": 0.6741342579240754,
+      "learning_rate": 1.995755588417644e-05,
+      "loss": 0.4243,
+      "step": 490
+    },
+    {
+      "epoch": 0.2346335344364136,
+      "grad_norm": 0.738629652421803,
+      "learning_rate": 1.9957069770999324e-05,
+      "loss": 0.4437,
+      "step": 491
+    },
+    {
+      "epoch": 0.23511140314198675,
+      "grad_norm": 0.6924217432219697,
+      "learning_rate": 1.9956580895906694e-05,
+      "loss": 0.43,
+      "step": 492
+    },
+    {
+      "epoch": 0.23558927184755987,
+      "grad_norm": 0.7369180221087179,
+      "learning_rate": 1.9956089259034154e-05,
+      "loss": 0.4339,
+      "step": 493
+    },
+    {
+      "epoch": 0.23606714055313302,
+      "grad_norm": 0.805559465873891,
+      "learning_rate": 1.995559486051808e-05,
+      "loss": 0.4361,
+      "step": 494
+    },
+    {
+      "epoch": 0.23654500925870617,
+      "grad_norm": 0.7823248018398075,
+      "learning_rate": 1.99550977004956e-05,
+      "loss": 0.4359,
+      "step": 495
+    },
+    {
+      "epoch": 0.23702287796427932,
+      "grad_norm": 0.7554148112351713,
+      "learning_rate": 1.9954597779104624e-05,
+      "loss": 0.4242,
+      "step": 496
+    },
+    {
+      "epoch": 0.23750074666985246,
+      "grad_norm": 0.7027609576658066,
+      "learning_rate": 1.995409509648382e-05,
+      "loss": 0.4309,
+      "step": 497
+    },
+    {
+      "epoch": 0.2379786153754256,
+      "grad_norm": 0.705129151371866,
+      "learning_rate": 1.9953589652772627e-05,
+      "loss": 0.4464,
+      "step": 498
+    },
+    {
+      "epoch": 0.23845648408099873,
+      "grad_norm": 0.7214815482253628,
+      "learning_rate": 1.995308144811124e-05,
+      "loss": 0.4427,
+      "step": 499
+    },
+    {
+      "epoch": 0.23893435278657188,
+      "grad_norm": 0.7183066127583475,
+      "learning_rate": 1.9952570482640628e-05,
+      "loss": 0.4301,
+      "step": 500
+    },
+    {
+      "epoch": 0.23941222149214503,
+      "grad_norm": 0.797831280147801,
+      "learning_rate": 1.9952056756502525e-05,
+      "loss": 0.4367,
+      "step": 501
+    },
+    {
+      "epoch": 0.23989009019771818,
+      "grad_norm": 0.7379315395749698,
+      "learning_rate": 1.9951540269839428e-05,
+      "loss": 0.4399,
+      "step": 502
+    },
+    {
+      "epoch": 0.24036795890329132,
+      "grad_norm": 0.7945392228374598,
+      "learning_rate": 1.9951021022794602e-05,
+      "loss": 0.4234,
+      "step": 503
+    },
+    {
+      "epoch": 0.24084582760886447,
+      "grad_norm": 0.7264140524348444,
+      "learning_rate": 1.995049901551208e-05,
+      "loss": 0.4264,
+      "step": 504
+    },
+    {
+      "epoch": 0.24132369631443762,
+      "grad_norm": 0.668632865034176,
+      "learning_rate": 1.9949974248136655e-05,
+      "loss": 0.4288,
+      "step": 505
+    },
+    {
+      "epoch": 0.24180156502001074,
+      "grad_norm": 0.8309640004970509,
+      "learning_rate": 1.9949446720813886e-05,
+      "loss": 0.4438,
+      "step": 506
+    },
+    {
+      "epoch": 0.2422794337255839,
+      "grad_norm": 0.6904435675318091,
+      "learning_rate": 1.9948916433690103e-05,
+      "loss": 0.4439,
+      "step": 507
+    },
+    {
+      "epoch": 0.24275730243115703,
+      "grad_norm": 0.7354434100825048,
+      "learning_rate": 1.99483833869124e-05,
+      "loss": 0.4283,
+      "step": 508
+    },
+    {
+      "epoch": 0.24323517113673018,
+      "grad_norm": 0.6765398629661438,
+      "learning_rate": 1.9947847580628625e-05,
+      "loss": 0.4387,
+      "step": 509
+    },
+    {
+      "epoch": 0.24371303984230333,
+      "grad_norm": 0.7239562081327908,
+      "learning_rate": 1.9947309014987414e-05,
+      "loss": 0.4375,
+      "step": 510
+    },
+    {
+      "epoch": 0.24419090854787648,
+      "grad_norm": 0.6978300565774074,
+      "learning_rate": 1.9946767690138146e-05,
+      "loss": 0.4356,
+      "step": 511
+    },
+    {
+      "epoch": 0.24466877725344963,
+      "grad_norm": 0.7117480219425797,
+      "learning_rate": 1.994622360623098e-05,
+      "loss": 0.4403,
+      "step": 512
+    },
+    {
+      "epoch": 0.24514664595902275,
+      "grad_norm": 0.7477545746890273,
+      "learning_rate": 1.994567676341683e-05,
+      "loss": 0.4245,
+      "step": 513
+    },
+    {
+      "epoch": 0.2456245146645959,
+      "grad_norm": 0.7374013791361971,
+      "learning_rate": 1.9945127161847393e-05,
+      "loss": 0.4223,
+      "step": 514
+    },
+    {
+      "epoch": 0.24610238337016904,
+      "grad_norm": 0.6642173278949521,
+      "learning_rate": 1.9944574801675106e-05,
+      "loss": 0.4341,
+      "step": 515
+    },
+    {
+      "epoch": 0.2465802520757422,
+      "grad_norm": 0.7259774358658097,
+      "learning_rate": 1.994401968305319e-05,
+      "loss": 0.4252,
+      "step": 516
+    },
+    {
+      "epoch": 0.24705812078131534,
+      "grad_norm": 0.7180794651515113,
+      "learning_rate": 1.994346180613562e-05,
+      "loss": 0.4348,
+      "step": 517
+    },
+    {
+      "epoch": 0.24753598948688849,
+      "grad_norm": 0.6955289749178415,
+      "learning_rate": 1.9942901171077146e-05,
+      "loss": 0.4375,
+      "step": 518
+    },
+    {
+      "epoch": 0.24801385819246163,
+      "grad_norm": 0.7245432909819279,
+      "learning_rate": 1.994233777803328e-05,
+      "loss": 0.4399,
+      "step": 519
+    },
+    {
+      "epoch": 0.24849172689803475,
+      "grad_norm": 0.6067174282133342,
+      "learning_rate": 1.9941771627160287e-05,
+      "loss": 0.4217,
+      "step": 520
+    },
+    {
+      "epoch": 0.2489695956036079,
+      "grad_norm": 0.703988857071276,
+      "learning_rate": 1.994120271861522e-05,
+      "loss": 0.4412,
+      "step": 521
+    },
+    {
+      "epoch": 0.24944746430918105,
+      "grad_norm": 0.8991621677494505,
+      "learning_rate": 1.9940631052555882e-05,
+      "loss": 0.3995,
+      "step": 522
+    },
+    {
+      "epoch": 0.2499253330147542,
+      "grad_norm": 0.7976604424556334,
+      "learning_rate": 1.9940056629140835e-05,
+      "loss": 0.4105,
+      "step": 523
+    },
+    {
+      "epoch": 0.2504032017203273,
+      "grad_norm": 0.6019680965372577,
+      "learning_rate": 1.9939479448529418e-05,
+      "loss": 0.4192,
+      "step": 524
+    },
+    {
+      "epoch": 0.25088107042590047,
+      "grad_norm": 0.704506878051367,
+      "learning_rate": 1.9938899510881732e-05,
+      "loss": 0.4196,
+      "step": 525
+    },
+    {
+      "epoch": 0.2513589391314736,
+      "grad_norm": 0.6554375963262496,
+      "learning_rate": 1.9938316816358644e-05,
+      "loss": 0.4343,
+      "step": 526
+    },
+    {
+      "epoch": 0.25183680783704676,
+      "grad_norm": 0.992457986922531,
+      "learning_rate": 1.9937731365121777e-05,
+      "loss": 0.4328,
+      "step": 527
+    },
+    {
+      "epoch": 0.2523146765426199,
+      "grad_norm": 0.6658083513417471,
+      "learning_rate": 1.9937143157333528e-05,
+      "loss": 0.4298,
+      "step": 528
+    },
+    {
+      "epoch": 0.25279254524819306,
+      "grad_norm": 0.7222172239047646,
+      "learning_rate": 1.9936552193157055e-05,
+      "loss": 0.428,
+      "step": 529
+    },
+    {
+      "epoch": 0.2532704139537662,
+      "grad_norm": 0.6390063477285883,
+      "learning_rate": 1.9935958472756283e-05,
+      "loss": 0.4348,
+      "step": 530
+    },
+    {
+      "epoch": 0.25374828265933935,
+      "grad_norm": 0.749202476640269,
+      "learning_rate": 1.9935361996295896e-05,
+      "loss": 0.4424,
+      "step": 531
+    },
+    {
+      "epoch": 0.2542261513649125,
+      "grad_norm": 0.7192147324773318,
+      "learning_rate": 1.993476276394135e-05,
+      "loss": 0.447,
+      "step": 532
+    },
+    {
+      "epoch": 0.25470402007048565,
+      "grad_norm": 0.7498159675525952,
+      "learning_rate": 1.9934160775858856e-05,
+      "loss": 0.416,
+      "step": 533
+    },
+    {
+      "epoch": 0.2551818887760588,
+      "grad_norm": 0.7187313720282974,
+      "learning_rate": 1.9933556032215402e-05,
+      "loss": 0.4033,
+      "step": 534
+    },
+    {
+      "epoch": 0.25565975748163194,
+      "grad_norm": 0.7075911635545219,
+      "learning_rate": 1.993294853317873e-05,
+      "loss": 0.4233,
+      "step": 535
+    },
+    {
+      "epoch": 0.2561376261872051,
+      "grad_norm": 0.7116488897227271,
+      "learning_rate": 1.9932338278917348e-05,
+      "loss": 0.4469,
+      "step": 536
+    },
+    {
+      "epoch": 0.2566154948927782,
+      "grad_norm": 0.6870926308030396,
+      "learning_rate": 1.993172526960053e-05,
+      "loss": 0.4429,
+      "step": 537
+    },
+    {
+      "epoch": 0.25709336359835133,
+      "grad_norm": 0.7109693256013045,
+      "learning_rate": 1.9931109505398316e-05,
+      "loss": 0.4263,
+      "step": 538
+    },
+    {
+      "epoch": 0.2575712323039245,
+      "grad_norm": 0.6545214091784688,
+      "learning_rate": 1.9930490986481507e-05,
+      "loss": 0.4346,
+      "step": 539
+    },
+    {
+      "epoch": 0.25804910100949763,
+      "grad_norm": 0.8547757226945168,
+      "learning_rate": 1.9929869713021668e-05,
+      "loss": 0.4203,
+      "step": 540
+    },
+    {
+      "epoch": 0.2585269697150708,
+      "grad_norm": 0.6813604542557925,
+      "learning_rate": 1.9929245685191133e-05,
+      "loss": 0.4267,
+      "step": 541
+    },
+    {
+      "epoch": 0.2590048384206439,
+      "grad_norm": 0.6641936751299076,
+      "learning_rate": 1.9928618903162992e-05,
+      "loss": 0.4339,
+      "step": 542
+    },
+    {
+      "epoch": 0.25948270712621707,
+      "grad_norm": 0.9793051618462918,
+      "learning_rate": 1.9927989367111102e-05,
+      "loss": 0.4389,
+      "step": 543
+    },
+    {
+      "epoch": 0.2599605758317902,
+      "grad_norm": 0.5970607719650806,
+      "learning_rate": 1.9927357077210093e-05,
+      "loss": 0.4285,
+      "step": 544
+    },
+    {
+      "epoch": 0.26043844453736337,
+      "grad_norm": 0.6677945714629177,
+      "learning_rate": 1.9926722033635343e-05,
+      "loss": 0.4185,
+      "step": 545
+    },
+    {
+      "epoch": 0.2609163132429365,
+      "grad_norm": 0.7032868326166338,
+      "learning_rate": 1.992608423656301e-05,
+      "loss": 0.4248,
+      "step": 546
+    },
+    {
+      "epoch": 0.26139418194850966,
+      "grad_norm": 0.675972367560421,
+      "learning_rate": 1.9925443686169998e-05,
+      "loss": 0.4358,
+      "step": 547
+    },
+    {
+      "epoch": 0.2618720506540828,
+      "grad_norm": 0.6648018975438753,
+      "learning_rate": 1.992480038263399e-05,
+      "loss": 0.4243,
+      "step": 548
+    },
+    {
+      "epoch": 0.26234991935965596,
+      "grad_norm": 0.7300070842008606,
+      "learning_rate": 1.992415432613343e-05,
+      "loss": 0.426,
+      "step": 549
+    },
+    {
+      "epoch": 0.26282778806522905,
+      "grad_norm": 0.6966820393669727,
+      "learning_rate": 1.9923505516847514e-05,
+      "loss": 0.4462,
+      "step": 550
+    },
+    {
+      "epoch": 0.2633056567708022,
+      "grad_norm": 0.6799521833338832,
+      "learning_rate": 1.9922853954956217e-05,
+      "loss": 0.4165,
+      "step": 551
+    },
+    {
+      "epoch": 0.26378352547637535,
+      "grad_norm": 0.6842858880612487,
+      "learning_rate": 1.9922199640640272e-05,
+      "loss": 0.4215,
+      "step": 552
+    },
+    {
+      "epoch": 0.2642613941819485,
+      "grad_norm": 0.684498742912379,
+      "learning_rate": 1.9921542574081168e-05,
+      "loss": 0.4376,
+      "step": 553
+    },
+    {
+      "epoch": 0.26473926288752164,
+      "grad_norm": 0.7472555717042567,
+      "learning_rate": 1.992088275546117e-05,
+      "loss": 0.4281,
+      "step": 554
+    },
+    {
+      "epoch": 0.2652171315930948,
+      "grad_norm": 0.6526006080700162,
+      "learning_rate": 1.9920220184963296e-05,
+      "loss": 0.4203,
+      "step": 555
+    },
+    {
+      "epoch": 0.26569500029866794,
+      "grad_norm": 0.6864252975558336,
+      "learning_rate": 1.991955486277133e-05,
+      "loss": 0.4253,
+      "step": 556
+    },
+    {
+      "epoch": 0.2661728690042411,
+      "grad_norm": 0.7010911608893313,
+      "learning_rate": 1.9918886789069824e-05,
+      "loss": 0.4268,
+      "step": 557
+    },
+    {
+      "epoch": 0.26665073770981423,
+      "grad_norm": 0.6432530703558913,
+      "learning_rate": 1.991821596404409e-05,
+      "loss": 0.4095,
+      "step": 558
+    },
+    {
+      "epoch": 0.2671286064153874,
+      "grad_norm": 0.6714812788316195,
+      "learning_rate": 1.99175423878802e-05,
+      "loss": 0.439,
+      "step": 559
+    },
+    {
+      "epoch": 0.26760647512096053,
+      "grad_norm": 0.7313943296188351,
+      "learning_rate": 1.9916866060764994e-05,
+      "loss": 0.4207,
+      "step": 560
+    },
+    {
+      "epoch": 0.2680843438265337,
+      "grad_norm": 0.6753023749397835,
+      "learning_rate": 1.9916186982886074e-05,
+      "loss": 0.4232,
+      "step": 561
+    },
+    {
+      "epoch": 0.2685622125321068,
+      "grad_norm": 0.6812854395562149,
+      "learning_rate": 1.9915505154431806e-05,
+      "loss": 0.4312,
+      "step": 562
+    },
+    {
+      "epoch": 0.26904008123768,
+      "grad_norm": 0.7326468942961418,
+      "learning_rate": 1.9914820575591315e-05,
+      "loss": 0.4106,
+      "step": 563
+    },
+    {
+      "epoch": 0.26951794994325307,
+      "grad_norm": 0.710879982047124,
+      "learning_rate": 1.9914133246554486e-05,
+      "loss": 0.4312,
+      "step": 564
+    },
+    {
+      "epoch": 0.2699958186488262,
+      "grad_norm": 0.6098533019354313,
+      "learning_rate": 1.991344316751198e-05,
+      "loss": 0.4269,
+      "step": 565
+    },
+    {
+      "epoch": 0.27047368735439936,
+      "grad_norm": 0.6527319247618629,
+      "learning_rate": 1.9912750338655207e-05,
+      "loss": 0.4253,
+      "step": 566
+    },
+    {
+      "epoch": 0.2709515560599725,
+      "grad_norm": 0.6554170109236619,
+      "learning_rate": 1.9912054760176352e-05,
+      "loss": 0.4175,
+      "step": 567
+    },
+    {
+      "epoch": 0.27142942476554566,
+      "grad_norm": 0.6487793205519652,
+      "learning_rate": 1.991135643226835e-05,
+      "loss": 0.4306,
+      "step": 568
+    },
+    {
+      "epoch": 0.2719072934711188,
+      "grad_norm": 0.6830594174977985,
+      "learning_rate": 1.9910655355124905e-05,
+      "loss": 0.4062,
+      "step": 569
+    },
+    {
+      "epoch": 0.27238516217669195,
+      "grad_norm": 0.7070510514831456,
+      "learning_rate": 1.9909951528940485e-05,
+      "loss": 0.4244,
+      "step": 570
+    },
+    {
+      "epoch": 0.2728630308822651,
+      "grad_norm": 0.6919752747886392,
+      "learning_rate": 1.9909244953910324e-05,
+      "loss": 0.4374,
+      "step": 571
+    },
+    {
+      "epoch": 0.27334089958783825,
+      "grad_norm": 0.627083497227152,
+      "learning_rate": 1.990853563023041e-05,
+      "loss": 0.4293,
+      "step": 572
+    },
+    {
+      "epoch": 0.2738187682934114,
+      "grad_norm": 0.6811946382742817,
+      "learning_rate": 1.990782355809749e-05,
+      "loss": 0.4314,
+      "step": 573
+    },
+    {
+      "epoch": 0.27429663699898454,
+      "grad_norm": 0.8011136341064639,
+      "learning_rate": 1.9907108737709088e-05,
+      "loss": 0.4313,
+      "step": 574
+    },
+    {
+      "epoch": 0.2747745057045577,
+      "grad_norm": 0.7022098540651683,
+      "learning_rate": 1.990639116926348e-05,
+      "loss": 0.4102,
+      "step": 575
+    },
+    {
+      "epoch": 0.27525237441013084,
+      "grad_norm": 0.7139794448818395,
+      "learning_rate": 1.9905670852959707e-05,
+      "loss": 0.4131,
+      "step": 576
+    },
+    {
+      "epoch": 0.275730243115704,
+      "grad_norm": 0.6566170557292315,
+      "learning_rate": 1.9904947788997572e-05,
+      "loss": 0.4492,
+      "step": 577
+    },
+    {
+      "epoch": 0.2762081118212771,
+      "grad_norm": 0.6960200627247798,
+      "learning_rate": 1.9904221977577644e-05,
+      "loss": 0.4236,
+      "step": 578
+    },
+    {
+      "epoch": 0.27668598052685023,
+      "grad_norm": 0.7038180733792689,
+      "learning_rate": 1.9903493418901246e-05,
+      "loss": 0.4156,
+      "step": 579
+    },
+    {
+      "epoch": 0.2771638492324234,
+      "grad_norm": 0.6995395704949278,
+      "learning_rate": 1.9902762113170467e-05,
+      "loss": 0.427,
+      "step": 580
+    },
+    {
+      "epoch": 0.2776417179379965,
+      "grad_norm": 0.7141523375442532,
+      "learning_rate": 1.990202806058816e-05,
+      "loss": 0.4075,
+      "step": 581
+    },
+    {
+      "epoch": 0.27811958664356967,
+      "grad_norm": 0.6927565201255096,
+      "learning_rate": 1.990129126135794e-05,
+      "loss": 0.4238,
+      "step": 582
+    },
+    {
+      "epoch": 0.2785974553491428,
+      "grad_norm": 0.6417220798020897,
+      "learning_rate": 1.9900551715684175e-05,
+      "loss": 0.4144,
+      "step": 583
+    },
+    {
+      "epoch": 0.27907532405471597,
+      "grad_norm": 0.6392690865600442,
+      "learning_rate": 1.989980942377201e-05,
+      "loss": 0.415,
+      "step": 584
+    },
+    {
+      "epoch": 0.2795531927602891,
+      "grad_norm": 0.639620744803394,
+      "learning_rate": 1.989906438582734e-05,
+      "loss": 0.4251,
+      "step": 585
+    },
+    {
+      "epoch": 0.28003106146586226,
+      "grad_norm": 0.6707471902399823,
+      "learning_rate": 1.9898316602056825e-05,
+      "loss": 0.4225,
+      "step": 586
+    },
+    {
+      "epoch": 0.2805089301714354,
+      "grad_norm": 0.6292145878798392,
+      "learning_rate": 1.989756607266789e-05,
+      "loss": 0.4281,
+      "step": 587
+    },
+    {
+      "epoch": 0.28098679887700856,
+      "grad_norm": 0.6650049148813777,
+      "learning_rate": 1.9896812797868714e-05,
+      "loss": 0.4273,
+      "step": 588
+    },
+    {
+      "epoch": 0.2814646675825817,
+      "grad_norm": 0.6989760949516968,
+      "learning_rate": 1.9896056777868245e-05,
+      "loss": 0.4133,
+      "step": 589
+    },
+    {
+      "epoch": 0.28194253628815485,
+      "grad_norm": 0.6273420717769904,
+      "learning_rate": 1.9895298012876192e-05,
+      "loss": 0.4117,
+      "step": 590
+    },
+    {
+      "epoch": 0.28242040499372795,
+      "grad_norm": 0.6513446444776289,
+      "learning_rate": 1.9894536503103018e-05,
+      "loss": 0.4243,
+      "step": 591
+    },
+    {
+      "epoch": 0.2828982736993011,
+      "grad_norm": 0.7257067616669561,
+      "learning_rate": 1.9893772248759956e-05,
+      "loss": 0.4377,
+      "step": 592
+    },
+    {
+      "epoch": 0.28337614240487424,
+      "grad_norm": 0.6449292794210849,
+      "learning_rate": 1.9893005250058994e-05,
+      "loss": 0.4262,
+      "step": 593
+    },
+    {
+      "epoch": 0.2838540111104474,
+      "grad_norm": 0.7101669933858333,
+      "learning_rate": 1.9892235507212885e-05,
+      "loss": 0.4166,
+      "step": 594
+    },
+    {
+      "epoch": 0.28433187981602054,
+      "grad_norm": 0.9532443107189144,
+      "learning_rate": 1.9891463020435144e-05,
+      "loss": 0.412,
+      "step": 595
+    },
+    {
+      "epoch": 0.2848097485215937,
+      "grad_norm": 0.6763594227901946,
+      "learning_rate": 1.9890687789940044e-05,
+      "loss": 0.4305,
+      "step": 596
+    },
+    {
+      "epoch": 0.28528761722716683,
+      "grad_norm": 0.666645047214093,
+      "learning_rate": 1.9889909815942615e-05,
+      "loss": 0.4262,
+      "step": 597
+    },
+    {
+      "epoch": 0.28576548593274,
+      "grad_norm": 0.7187479088271995,
+      "learning_rate": 1.9889129098658662e-05,
+      "loss": 0.4127,
+      "step": 598
+    },
+    {
+      "epoch": 0.28624335463831313,
+      "grad_norm": 0.7191161821405772,
+      "learning_rate": 1.9888345638304737e-05,
+      "loss": 0.4345,
+      "step": 599
+    },
+    {
+      "epoch": 0.2867212233438863,
+      "grad_norm": 0.706994011714395,
+      "learning_rate": 1.9887559435098162e-05,
+      "loss": 0.41,
+      "step": 600
+    },
+    {
+      "epoch": 0.2871990920494594,
+      "grad_norm": 0.7497992817297052,
+      "learning_rate": 1.988677048925701e-05,
+      "loss": 0.4285,
+      "step": 601
+    },
+    {
+      "epoch": 0.2876769607550326,
+      "grad_norm": 0.652784327493937,
+      "learning_rate": 1.9885978801000124e-05,
+      "loss": 0.4256,
+      "step": 602
+    },
+    {
+      "epoch": 0.2881548294606057,
+      "grad_norm": 0.776621607360754,
+      "learning_rate": 1.988518437054711e-05,
+      "loss": 0.4322,
+      "step": 603
+    },
+    {
+      "epoch": 0.28863269816617887,
+      "grad_norm": 0.5839322380726024,
+      "learning_rate": 1.9884387198118316e-05,
+      "loss": 0.4167,
+      "step": 604
+    },
+    {
+      "epoch": 0.28911056687175196,
+      "grad_norm": 0.7907030900607616,
+      "learning_rate": 1.9883587283934875e-05,
+      "loss": 0.406,
+      "step": 605
+    },
+    {
+      "epoch": 0.2895884355773251,
+      "grad_norm": 1.0014227126496111,
+      "learning_rate": 1.988278462821866e-05,
+      "loss": 0.4151,
+      "step": 606
+    },
+    {
+      "epoch": 0.29006630428289826,
+      "grad_norm": 0.66994522088662,
+      "learning_rate": 1.9881979231192323e-05,
+      "loss": 0.4392,
+      "step": 607
+    },
+    {
+      "epoch": 0.2905441729884714,
+      "grad_norm": 0.5824214225034886,
+      "learning_rate": 1.9881171093079264e-05,
+      "loss": 0.4267,
+      "step": 608
+    },
+    {
+      "epoch": 0.29102204169404455,
+      "grad_norm": 0.6495953436821396,
+      "learning_rate": 1.988036021410364e-05,
+      "loss": 0.4334,
+      "step": 609
+    },
+    {
+      "epoch": 0.2914999103996177,
+      "grad_norm": 0.6878938109775274,
+      "learning_rate": 1.9879546594490383e-05,
+      "loss": 0.41,
+      "step": 610
+    },
+    {
+      "epoch": 0.29197777910519085,
+      "grad_norm": 0.7296069094771699,
+      "learning_rate": 1.987873023446517e-05,
+      "loss": 0.4134,
+      "step": 611
+    },
+    {
+      "epoch": 0.292455647810764,
+      "grad_norm": 0.7455591671402041,
+      "learning_rate": 1.987791113425445e-05,
+      "loss": 0.424,
+      "step": 612
+    },
+    {
+      "epoch": 0.29293351651633714,
+      "grad_norm": 0.6695483498248275,
+      "learning_rate": 1.9877089294085424e-05,
+      "loss": 0.425,
+      "step": 613
+    },
+    {
+      "epoch": 0.2934113852219103,
+      "grad_norm": 0.6796642781672113,
+      "learning_rate": 1.9876264714186054e-05,
+      "loss": 0.3983,
+      "step": 614
+    },
+    {
+      "epoch": 0.29388925392748344,
+      "grad_norm": 0.6095965308498806,
+      "learning_rate": 1.987543739478507e-05,
+      "loss": 0.4225,
+      "step": 615
+    },
+    {
+      "epoch": 0.2943671226330566,
+      "grad_norm": 0.6305677380212332,
+      "learning_rate": 1.987460733611195e-05,
+      "loss": 0.4356,
+      "step": 616
+    },
+    {
+      "epoch": 0.29484499133862974,
+      "grad_norm": 0.7050343310604908,
+      "learning_rate": 1.9873774538396945e-05,
+      "loss": 0.4283,
+      "step": 617
+    },
+    {
+      "epoch": 0.29532286004420283,
+      "grad_norm": 0.6530927062141915,
+      "learning_rate": 1.987293900187105e-05,
+      "loss": 0.4316,
+      "step": 618
+    },
+    {
+      "epoch": 0.295800728749776,
+      "grad_norm": 0.6928306836676819,
+      "learning_rate": 1.9872100726766028e-05,
+      "loss": 0.4066,
+      "step": 619
+    },
+    {
+      "epoch": 0.2962785974553491,
+      "grad_norm": 0.7451946223590281,
+      "learning_rate": 1.987125971331441e-05,
+      "loss": 0.4227,
+      "step": 620
+    },
+    {
+      "epoch": 0.29675646616092227,
+      "grad_norm": 0.7569155679553317,
+      "learning_rate": 1.9870415961749472e-05,
+      "loss": 0.4171,
+      "step": 621
+    },
+    {
+      "epoch": 0.2972343348664954,
+      "grad_norm": 0.7729456988114033,
+      "learning_rate": 1.986956947230526e-05,
+      "loss": 0.4374,
+      "step": 622
+    },
+    {
+      "epoch": 0.29771220357206857,
+      "grad_norm": 0.7222068350990688,
+      "learning_rate": 1.986872024521657e-05,
+      "loss": 0.4055,
+      "step": 623
+    },
+    {
+      "epoch": 0.2981900722776417,
+      "grad_norm": 0.777255093182333,
+      "learning_rate": 1.9867868280718966e-05,
+      "loss": 0.4184,
+      "step": 624
+    },
+    {
+      "epoch": 0.29866794098321486,
+      "grad_norm": 0.707058786281402,
+      "learning_rate": 1.9867013579048765e-05,
+      "loss": 0.4278,
+      "step": 625
+    },
+    {
+      "epoch": 0.299145809688788,
+      "grad_norm": 0.6978123671395932,
+      "learning_rate": 1.986615614044305e-05,
+      "loss": 0.3976,
+      "step": 626
+    },
+    {
+      "epoch": 0.29962367839436116,
+      "grad_norm": 0.6497844128498012,
+      "learning_rate": 1.9865295965139654e-05,
+      "loss": 0.4233,
+      "step": 627
+    },
+    {
+      "epoch": 0.3001015470999343,
+      "grad_norm": 0.7279868951567827,
+      "learning_rate": 1.9864433053377183e-05,
+      "loss": 0.421,
+      "step": 628
+    },
+    {
+      "epoch": 0.30057941580550745,
+      "grad_norm": 0.6228075334334027,
+      "learning_rate": 1.9863567405394987e-05,
+      "loss": 0.4182,
+      "step": 629
+    },
+    {
+      "epoch": 0.3010572845110806,
+      "grad_norm": 1.2081410007291293,
+      "learning_rate": 1.9862699021433186e-05,
+      "loss": 0.43,
+      "step": 630
+    },
+    {
+      "epoch": 0.30153515321665375,
+      "grad_norm": 0.6584806729516838,
+      "learning_rate": 1.9861827901732645e-05,
+      "loss": 0.4243,
+      "step": 631
+    },
+    {
+      "epoch": 0.30201302192222684,
+      "grad_norm": 0.7145104561793435,
+      "learning_rate": 1.986095404653501e-05,
+      "loss": 0.4149,
+      "step": 632
+    },
+    {
+      "epoch": 0.3024908906278,
+      "grad_norm": 0.6293456214888333,
+      "learning_rate": 1.986007745608266e-05,
+      "loss": 0.4059,
+      "step": 633
+    },
+    {
+      "epoch": 0.30296875933337314,
+      "grad_norm": 0.6105778855320209,
+      "learning_rate": 1.985919813061876e-05,
+      "loss": 0.4075,
+      "step": 634
+    },
+    {
+      "epoch": 0.3034466280389463,
+      "grad_norm": 0.5850039212779303,
+      "learning_rate": 1.9858316070387208e-05,
+      "loss": 0.4265,
+      "step": 635
+    },
+    {
+      "epoch": 0.30392449674451943,
+      "grad_norm": 0.6552242491777841,
+      "learning_rate": 1.985743127563268e-05,
+      "loss": 0.4211,
+      "step": 636
+    },
+    {
+      "epoch": 0.3044023654500926,
+      "grad_norm": 0.6302333591257354,
+      "learning_rate": 1.9856543746600593e-05,
+      "loss": 0.4223,
+      "step": 637
+    },
+    {
+      "epoch": 0.30488023415566573,
+      "grad_norm": 0.6537304777031439,
+      "learning_rate": 1.985565348353714e-05,
+      "loss": 0.4278,
+      "step": 638
+    },
+    {
+      "epoch": 0.3053581028612389,
+      "grad_norm": 0.6730389531773375,
+      "learning_rate": 1.9854760486689257e-05,
+      "loss": 0.4222,
+      "step": 639
+    },
+    {
+      "epoch": 0.305835971566812,
+      "grad_norm": 0.64240760182782,
+      "learning_rate": 1.9853864756304654e-05,
+      "loss": 0.4157,
+      "step": 640
+    },
+    {
+      "epoch": 0.3063138402723852,
+      "grad_norm": 0.6417456096649262,
+      "learning_rate": 1.9852966292631785e-05,
+      "loss": 0.3962,
+      "step": 641
+    },
+    {
+      "epoch": 0.3067917089779583,
+      "grad_norm": 0.6690827386602772,
+      "learning_rate": 1.985206509591987e-05,
+      "loss": 0.4155,
+      "step": 642
+    },
+    {
+      "epoch": 0.30726957768353147,
+      "grad_norm": 0.6670159655708577,
+      "learning_rate": 1.9851161166418888e-05,
+      "loss": 0.4209,
+      "step": 643
+    },
+    {
+      "epoch": 0.3077474463891046,
+      "grad_norm": 0.6837870559936386,
+      "learning_rate": 1.9850254504379568e-05,
+      "loss": 0.4239,
+      "step": 644
+    },
+    {
+      "epoch": 0.3082253150946777,
+      "grad_norm": 0.635771654798046,
+      "learning_rate": 1.9849345110053405e-05,
+      "loss": 0.4115,
+      "step": 645
+    },
+    {
+      "epoch": 0.30870318380025086,
+      "grad_norm": 0.6317438803320911,
+      "learning_rate": 1.9848432983692642e-05,
+      "loss": 0.4167,
+      "step": 646
+    },
+    {
+      "epoch": 0.309181052505824,
+      "grad_norm": 0.7355553259854325,
+      "learning_rate": 1.98475181255503e-05,
+      "loss": 0.4191,
+      "step": 647
+    },
+    {
+      "epoch": 0.30965892121139715,
+      "grad_norm": 0.6759418052818299,
+      "learning_rate": 1.9846600535880135e-05,
+      "loss": 0.4287,
+      "step": 648
+    },
+    {
+      "epoch": 0.3101367899169703,
+      "grad_norm": 0.8036676894573562,
+      "learning_rate": 1.9845680214936668e-05,
+      "loss": 0.4001,
+      "step": 649
+    },
+    {
+      "epoch": 0.31061465862254345,
+      "grad_norm": 0.6256985874959373,
+      "learning_rate": 1.984475716297519e-05,
+      "loss": 0.4126,
+      "step": 650
+    },
+    {
+      "epoch": 0.3110925273281166,
+      "grad_norm": 0.6830281771469799,
+      "learning_rate": 1.984383138025173e-05,
+      "loss": 0.4133,
+      "step": 651
+    },
+    {
+      "epoch": 0.31157039603368974,
+      "grad_norm": 0.7108602267936108,
+      "learning_rate": 1.984290286702309e-05,
+      "loss": 0.4056,
+      "step": 652
+    },
+    {
+      "epoch": 0.3120482647392629,
+      "grad_norm": 0.6539475916242298,
+      "learning_rate": 1.984197162354682e-05,
+      "loss": 0.4332,
+      "step": 653
+    },
+    {
+      "epoch": 0.31252613344483604,
+      "grad_norm": 0.7003890411182644,
+      "learning_rate": 1.984103765008123e-05,
+      "loss": 0.4152,
+      "step": 654
+    },
+    {
+      "epoch": 0.3130040021504092,
+      "grad_norm": 0.8734062183276717,
+      "learning_rate": 1.984010094688539e-05,
+      "loss": 0.4308,
+      "step": 655
+    },
+    {
+      "epoch": 0.31348187085598234,
+      "grad_norm": 0.6483836791788339,
+      "learning_rate": 1.9839161514219125e-05,
+      "loss": 0.4335,
+      "step": 656
+    },
+    {
+      "epoch": 0.3139597395615555,
+      "grad_norm": 0.7224253461656356,
+      "learning_rate": 1.983821935234301e-05,
+      "loss": 0.4079,
+      "step": 657
+    },
+    {
+      "epoch": 0.31443760826712863,
+      "grad_norm": 0.6023125747576976,
+      "learning_rate": 1.9837274461518393e-05,
+      "loss": 0.4208,
+      "step": 658
+    },
+    {
+      "epoch": 0.3149154769727017,
+      "grad_norm": 0.6840565183544368,
+      "learning_rate": 1.9836326842007368e-05,
+      "loss": 0.4179,
+      "step": 659
+    },
+    {
+      "epoch": 0.3153933456782749,
+      "grad_norm": 0.6070851699714096,
+      "learning_rate": 1.9835376494072788e-05,
+      "loss": 0.421,
+      "step": 660
+    },
+    {
+      "epoch": 0.315871214383848,
+      "grad_norm": 0.7883725816487038,
+      "learning_rate": 1.9834423417978258e-05,
+      "loss": 0.4161,
+      "step": 661
+    },
+    {
+      "epoch": 0.31634908308942117,
+      "grad_norm": 0.7058787543409192,
+      "learning_rate": 1.983346761398815e-05,
+      "loss": 0.4077,
+      "step": 662
+    },
+    {
+      "epoch": 0.3168269517949943,
+      "grad_norm": 0.6615940212203109,
+      "learning_rate": 1.983250908236759e-05,
+      "loss": 0.4149,
+      "step": 663
+    },
+    {
+      "epoch": 0.31730482050056746,
+      "grad_norm": 0.7006736509156779,
+      "learning_rate": 1.9831547823382446e-05,
+      "loss": 0.4088,
+      "step": 664
+    },
+    {
+      "epoch": 0.3177826892061406,
+      "grad_norm": 0.6646570566525254,
+      "learning_rate": 1.9830583837299363e-05,
+      "loss": 0.4131,
+      "step": 665
+    },
+    {
+      "epoch": 0.31826055791171376,
+      "grad_norm": 0.6530528056282681,
+      "learning_rate": 1.9829617124385732e-05,
+      "loss": 0.396,
+      "step": 666
+    },
+    {
+      "epoch": 0.3187384266172869,
+      "grad_norm": 0.9328303194521024,
+      "learning_rate": 1.9828647684909703e-05,
+      "loss": 0.3951,
+      "step": 667
+    },
+    {
+      "epoch": 0.31921629532286006,
+      "grad_norm": 0.6959860169532506,
+      "learning_rate": 1.9827675519140183e-05,
+      "loss": 0.427,
+      "step": 668
+    },
+    {
+      "epoch": 0.3196941640284332,
+      "grad_norm": 0.6996213590788233,
+      "learning_rate": 1.9826700627346825e-05,
+      "loss": 0.4142,
+      "step": 669
+    },
+    {
+      "epoch": 0.32017203273400635,
+      "grad_norm": 0.6591077715112674,
+      "learning_rate": 1.9825723009800058e-05,
+      "loss": 0.4163,
+      "step": 670
+    },
+    {
+      "epoch": 0.3206499014395795,
+      "grad_norm": 0.6696332483194858,
+      "learning_rate": 1.982474266677105e-05,
+      "loss": 0.407,
+      "step": 671
+    },
+    {
+      "epoch": 0.32112777014515265,
+      "grad_norm": 0.792144854122436,
+      "learning_rate": 1.9823759598531732e-05,
+      "loss": 0.4204,
+      "step": 672
+    },
+    {
+      "epoch": 0.32160563885072574,
+      "grad_norm": 0.7610226047172992,
+      "learning_rate": 1.9822773805354788e-05,
+      "loss": 0.412,
+      "step": 673
+    },
+    {
+      "epoch": 0.3220835075562989,
+      "grad_norm": 0.7518454562539055,
+      "learning_rate": 1.982178528751366e-05,
+      "loss": 0.4263,
+      "step": 674
+    },
+    {
+      "epoch": 0.32256137626187203,
+      "grad_norm": 0.6802817311894305,
+      "learning_rate": 1.9820794045282553e-05,
+      "loss": 0.4107,
+      "step": 675
+    },
+    {
+      "epoch": 0.3230392449674452,
+      "grad_norm": 0.7023162206342487,
+      "learning_rate": 1.9819800078936406e-05,
+      "loss": 0.4114,
+      "step": 676
+    },
+    {
+      "epoch": 0.32351711367301833,
+      "grad_norm": 0.6685874705533328,
+      "learning_rate": 1.981880338875094e-05,
+      "loss": 0.4185,
+      "step": 677
+    },
+    {
+      "epoch": 0.3239949823785915,
+      "grad_norm": 0.6244222422392075,
+      "learning_rate": 1.9817803975002614e-05,
+      "loss": 0.4241,
+      "step": 678
+    },
+    {
+      "epoch": 0.3244728510841646,
+      "grad_norm": 0.6522844807838616,
+      "learning_rate": 1.9816801837968647e-05,
+      "loss": 0.4089,
+      "step": 679
+    },
+    {
+      "epoch": 0.3249507197897378,
+      "grad_norm": 0.7102809988759002,
+      "learning_rate": 1.9815796977927015e-05,
+      "loss": 0.4094,
+      "step": 680
+    },
+    {
+      "epoch": 0.3254285884953109,
+      "grad_norm": 0.9736649325340913,
+      "learning_rate": 1.981478939515645e-05,
+      "loss": 0.3995,
+      "step": 681
+    },
+    {
+      "epoch": 0.32590645720088407,
+      "grad_norm": 0.6342355412436739,
+      "learning_rate": 1.981377908993644e-05,
+      "loss": 0.4086,
+      "step": 682
+    },
+    {
+      "epoch": 0.3263843259064572,
+      "grad_norm": 0.6557900154548946,
+      "learning_rate": 1.9812766062547218e-05,
+      "loss": 0.4167,
+      "step": 683
+    },
+    {
+      "epoch": 0.32686219461203037,
+      "grad_norm": 0.7089582460412668,
+      "learning_rate": 1.9811750313269785e-05,
+      "loss": 0.4184,
+      "step": 684
+    },
+    {
+      "epoch": 0.3273400633176035,
+      "grad_norm": 0.643622934429116,
+      "learning_rate": 1.9810731842385892e-05,
+      "loss": 0.4198,
+      "step": 685
+    },
+    {
+      "epoch": 0.3278179320231766,
+      "grad_norm": 0.6544401734455819,
+      "learning_rate": 1.9809710650178043e-05,
+      "loss": 0.4297,
+      "step": 686
+    },
+    {
+      "epoch": 0.32829580072874975,
+      "grad_norm": 0.6350037232509057,
+      "learning_rate": 1.9808686736929507e-05,
+      "loss": 0.4139,
+      "step": 687
+    },
+    {
+      "epoch": 0.3287736694343229,
+      "grad_norm": 0.6453998273855003,
+      "learning_rate": 1.9807660102924285e-05,
+      "loss": 0.4104,
+      "step": 688
+    },
+    {
+      "epoch": 0.32925153813989605,
+      "grad_norm": 0.6118435089195414,
+      "learning_rate": 1.980663074844716e-05,
+      "loss": 0.4439,
+      "step": 689
+    },
+    {
+      "epoch": 0.3297294068454692,
+      "grad_norm": 0.7137050965412601,
+      "learning_rate": 1.9805598673783644e-05,
+      "loss": 0.4051,
+      "step": 690
+    },
+    {
+      "epoch": 0.33020727555104235,
+      "grad_norm": 0.660701532438893,
+      "learning_rate": 1.980456387922003e-05,
+      "loss": 0.4247,
+      "step": 691
+    },
+    {
+      "epoch": 0.3306851442566155,
+      "grad_norm": 0.6219578421483081,
+      "learning_rate": 1.9803526365043342e-05,
+      "loss": 0.4311,
+      "step": 692
+    },
+    {
+      "epoch": 0.33116301296218864,
+      "grad_norm": 0.7398002604532392,
+      "learning_rate": 1.980248613154137e-05,
+      "loss": 0.4162,
+      "step": 693
+    },
+    {
+      "epoch": 0.3316408816677618,
+      "grad_norm": 0.594653754515913,
+      "learning_rate": 1.9801443179002664e-05,
+      "loss": 0.4291,
+      "step": 694
+    },
+    {
+      "epoch": 0.33211875037333494,
+      "grad_norm": 0.7039036720424287,
+      "learning_rate": 1.980039750771651e-05,
+      "loss": 0.4154,
+      "step": 695
+    },
+    {
+      "epoch": 0.3325966190789081,
+      "grad_norm": 0.6360396889801934,
+      "learning_rate": 1.9799349117972966e-05,
+      "loss": 0.4097,
+      "step": 696
+    },
+    {
+      "epoch": 0.33307448778448123,
+      "grad_norm": 0.650673130557327,
+      "learning_rate": 1.9798298010062834e-05,
+      "loss": 0.408,
+      "step": 697
+    },
+    {
+      "epoch": 0.3335523564900544,
+      "grad_norm": 0.6449814851092306,
+      "learning_rate": 1.979724418427767e-05,
+      "loss": 0.4237,
+      "step": 698
+    },
+    {
+      "epoch": 0.33403022519562753,
+      "grad_norm": 0.6952694171904437,
+      "learning_rate": 1.9796187640909793e-05,
+      "loss": 0.414,
+      "step": 699
+    },
+    {
+      "epoch": 0.3345080939012006,
+      "grad_norm": 0.6255490632329522,
+      "learning_rate": 1.9795128380252263e-05,
+      "loss": 0.4128,
+      "step": 700
+    },
+    {
+      "epoch": 0.33498596260677377,
+      "grad_norm": 0.7726068650204944,
+      "learning_rate": 1.9794066402598905e-05,
+      "loss": 0.3941,
+      "step": 701
+    },
+    {
+      "epoch": 0.3354638313123469,
+      "grad_norm": 0.7260875044062942,
+      "learning_rate": 1.9793001708244293e-05,
+      "loss": 0.4278,
+      "step": 702
+    },
+    {
+      "epoch": 0.33594170001792006,
+      "grad_norm": 0.6812812508438351,
+      "learning_rate": 1.9791934297483754e-05,
+      "loss": 0.4121,
+      "step": 703
+    },
+    {
+      "epoch": 0.3364195687234932,
+      "grad_norm": 0.6382204766381031,
+      "learning_rate": 1.9790864170613363e-05,
+      "loss": 0.4128,
+      "step": 704
+    },
+    {
+      "epoch": 0.33689743742906636,
+      "grad_norm": 0.7007245066237098,
+      "learning_rate": 1.978979132792996e-05,
+      "loss": 0.4226,
+      "step": 705
+    },
+    {
+      "epoch": 0.3373753061346395,
+      "grad_norm": 0.6498578233016445,
+      "learning_rate": 1.978871576973113e-05,
+      "loss": 0.4003,
+      "step": 706
+    },
+    {
+      "epoch": 0.33785317484021266,
+      "grad_norm": 0.8549420085188715,
+      "learning_rate": 1.9787637496315223e-05,
+      "loss": 0.4117,
+      "step": 707
+    },
+    {
+      "epoch": 0.3383310435457858,
+      "grad_norm": 0.7414633332196651,
+      "learning_rate": 1.978655650798132e-05,
+      "loss": 0.3992,
+      "step": 708
+    },
+    {
+      "epoch": 0.33880891225135895,
+      "grad_norm": 0.6427548458668187,
+      "learning_rate": 1.9785472805029274e-05,
+      "loss": 0.4097,
+      "step": 709
+    },
+    {
+      "epoch": 0.3392867809569321,
+      "grad_norm": 0.6277498179900414,
+      "learning_rate": 1.9784386387759684e-05,
+      "loss": 0.419,
+      "step": 710
+    },
+    {
+      "epoch": 0.33976464966250525,
+      "grad_norm": 0.7227930165728,
+      "learning_rate": 1.9783297256473905e-05,
+      "loss": 0.3995,
+      "step": 711
+    },
+    {
+      "epoch": 0.3402425183680784,
+      "grad_norm": 0.6041073332058529,
+      "learning_rate": 1.9782205411474042e-05,
+      "loss": 0.4028,
+      "step": 712
+    },
+    {
+      "epoch": 0.3407203870736515,
+      "grad_norm": 0.697676563953654,
+      "learning_rate": 1.978111085306295e-05,
+      "loss": 0.4131,
+      "step": 713
+    },
+    {
+      "epoch": 0.34119825577922464,
+      "grad_norm": 0.6626931153008405,
+      "learning_rate": 1.9780013581544245e-05,
+      "loss": 0.4364,
+      "step": 714
+    },
+    {
+      "epoch": 0.3416761244847978,
+      "grad_norm": 0.6841487019463591,
+      "learning_rate": 1.977891359722229e-05,
+      "loss": 0.423,
+      "step": 715
+    },
+    {
+      "epoch": 0.34215399319037093,
+      "grad_norm": 0.7564619275323727,
+      "learning_rate": 1.9777810900402203e-05,
+      "loss": 0.4161,
+      "step": 716
+    },
+    {
+      "epoch": 0.3426318618959441,
+      "grad_norm": 1.3658709278718753,
+      "learning_rate": 1.9776705491389844e-05,
+      "loss": 0.4109,
+      "step": 717
+    },
+    {
+      "epoch": 0.3431097306015172,
+      "grad_norm": 0.7907397360768815,
+      "learning_rate": 1.9775597370491838e-05,
+      "loss": 0.4051,
+      "step": 718
+    },
+    {
+      "epoch": 0.3435875993070904,
+      "grad_norm": 0.6514561615066782,
+      "learning_rate": 1.977448653801557e-05,
+      "loss": 0.4034,
+      "step": 719
+    },
+    {
+      "epoch": 0.3440654680126635,
+      "grad_norm": 0.6824032819742455,
+      "learning_rate": 1.9773372994269147e-05,
+      "loss": 0.4276,
+      "step": 720
+    },
+    {
+      "epoch": 0.34454333671823667,
+      "grad_norm": 0.6883595228191173,
+      "learning_rate": 1.9772256739561454e-05,
+      "loss": 0.4267,
+      "step": 721
+    },
+    {
+      "epoch": 0.3450212054238098,
+      "grad_norm": 0.7266627480718902,
+      "learning_rate": 1.9771137774202126e-05,
+      "loss": 0.4171,
+      "step": 722
+    },
+    {
+      "epoch": 0.34549907412938297,
+      "grad_norm": 0.6857743582824786,
+      "learning_rate": 1.9770016098501535e-05,
+      "loss": 0.4133,
+      "step": 723
+    },
+    {
+      "epoch": 0.3459769428349561,
+      "grad_norm": 0.8685391109702237,
+      "learning_rate": 1.9768891712770824e-05,
+      "loss": 0.394,
+      "step": 724
+    },
+    {
+      "epoch": 0.34645481154052926,
+      "grad_norm": 0.6842643433044535,
+      "learning_rate": 1.976776461732187e-05,
+      "loss": 0.4152,
+      "step": 725
+    },
+    {
+      "epoch": 0.3469326802461024,
+      "grad_norm": 0.7671231877298523,
+      "learning_rate": 1.976663481246731e-05,
+      "loss": 0.4238,
+      "step": 726
+    },
+    {
+      "epoch": 0.3474105489516755,
+      "grad_norm": 0.6577325703350719,
+      "learning_rate": 1.9765502298520534e-05,
+      "loss": 0.412,
+      "step": 727
+    },
+    {
+      "epoch": 0.34788841765724865,
+      "grad_norm": 0.6791922311294679,
+      "learning_rate": 1.976436707579568e-05,
+      "loss": 0.4071,
+      "step": 728
+    },
+    {
+      "epoch": 0.3483662863628218,
+      "grad_norm": 0.7164375380905718,
+      "learning_rate": 1.9763229144607643e-05,
+      "loss": 0.4213,
+      "step": 729
+    },
+    {
+      "epoch": 0.34884415506839495,
+      "grad_norm": 0.6448079596578629,
+      "learning_rate": 1.976208850527206e-05,
+      "loss": 0.4264,
+      "step": 730
+    },
+    {
+      "epoch": 0.3493220237739681,
+      "grad_norm": 0.6810860347020636,
+      "learning_rate": 1.9760945158105326e-05,
+      "loss": 0.4401,
+      "step": 731
+    },
+    {
+      "epoch": 0.34979989247954124,
+      "grad_norm": 0.6508629347760229,
+      "learning_rate": 1.975979910342458e-05,
+      "loss": 0.404,
+      "step": 732
+    },
+    {
+      "epoch": 0.3502777611851144,
+      "grad_norm": 0.667860785115438,
+      "learning_rate": 1.975865034154773e-05,
+      "loss": 0.4189,
+      "step": 733
+    },
+    {
+      "epoch": 0.35075562989068754,
+      "grad_norm": 0.5960633232464225,
+      "learning_rate": 1.975749887279341e-05,
+      "loss": 0.4063,
+      "step": 734
+    },
+    {
+      "epoch": 0.3512334985962607,
+      "grad_norm": 0.6676835117003757,
+      "learning_rate": 1.9756344697481027e-05,
+      "loss": 0.4168,
+      "step": 735
+    },
+    {
+      "epoch": 0.35171136730183383,
+      "grad_norm": 0.6027086678736178,
+      "learning_rate": 1.975518781593072e-05,
+      "loss": 0.3967,
+      "step": 736
+    },
+    {
+      "epoch": 0.352189236007407,
+      "grad_norm": 0.7011001055447835,
+      "learning_rate": 1.975402822846339e-05,
+      "loss": 0.409,
+      "step": 737
+    },
+    {
+      "epoch": 0.35266710471298013,
+      "grad_norm": 0.6173498609221076,
+      "learning_rate": 1.9752865935400692e-05,
+      "loss": 0.4069,
+      "step": 738
+    },
+    {
+      "epoch": 0.3531449734185533,
+      "grad_norm": 0.7500007957464343,
+      "learning_rate": 1.975170093706502e-05,
+      "loss": 0.4039,
+      "step": 739
+    },
+    {
+      "epoch": 0.35362284212412637,
+      "grad_norm": 0.586011297625854,
+      "learning_rate": 1.975053323377952e-05,
+      "loss": 0.4071,
+      "step": 740
+    },
+    {
+      "epoch": 0.3541007108296995,
+      "grad_norm": 0.749881324377058,
+      "learning_rate": 1.9749362825868105e-05,
+      "loss": 0.4236,
+      "step": 741
+    },
+    {
+      "epoch": 0.35457857953527266,
+      "grad_norm": 0.6097150903798377,
+      "learning_rate": 1.9748189713655414e-05,
+      "loss": 0.4343,
+      "step": 742
+    },
+    {
+      "epoch": 0.3550564482408458,
+      "grad_norm": 0.5888296839949839,
+      "learning_rate": 1.9747013897466852e-05,
+      "loss": 0.413,
+      "step": 743
+    },
+    {
+      "epoch": 0.35553431694641896,
+      "grad_norm": 0.6218972588587672,
+      "learning_rate": 1.974583537762857e-05,
+      "loss": 0.4002,
+      "step": 744
+    },
+    {
+      "epoch": 0.3560121856519921,
+      "grad_norm": 0.6244729217199203,
+      "learning_rate": 1.9744654154467468e-05,
+      "loss": 0.4249,
+      "step": 745
+    },
+    {
+      "epoch": 0.35649005435756526,
+      "grad_norm": 0.6019630528796382,
+      "learning_rate": 1.9743470228311195e-05,
+      "loss": 0.4108,
+      "step": 746
+    },
+    {
+      "epoch": 0.3569679230631384,
+      "grad_norm": 0.6053335199202433,
+      "learning_rate": 1.9742283599488154e-05,
+      "loss": 0.4164,
+      "step": 747
+    },
+    {
+      "epoch": 0.35744579176871155,
+      "grad_norm": 0.5390394035508936,
+      "learning_rate": 1.974109426832749e-05,
+      "loss": 0.4269,
+      "step": 748
+    },
+    {
+      "epoch": 0.3579236604742847,
+      "grad_norm": 0.6864448137729777,
+      "learning_rate": 1.973990223515911e-05,
+      "loss": 0.4151,
+      "step": 749
+    },
+    {
+      "epoch": 0.35840152917985785,
+      "grad_norm": 0.6117555520134601,
+      "learning_rate": 1.9738707500313655e-05,
+      "loss": 0.4127,
+      "step": 750
+    },
+    {
+      "epoch": 0.358879397885431,
+      "grad_norm": 0.6231362351908816,
+      "learning_rate": 1.973751006412253e-05,
+      "loss": 0.4107,
+      "step": 751
+    },
+    {
+      "epoch": 0.35935726659100414,
+      "grad_norm": 0.624929534918908,
+      "learning_rate": 1.973630992691788e-05,
+      "loss": 0.4006,
+      "step": 752
+    },
+    {
+      "epoch": 0.3598351352965773,
+      "grad_norm": 0.6397155829395167,
+      "learning_rate": 1.97351070890326e-05,
+      "loss": 0.4036,
+      "step": 753
+    },
+    {
+      "epoch": 0.3603130040021504,
+      "grad_norm": 0.6568169413470653,
+      "learning_rate": 1.9733901550800342e-05,
+      "loss": 0.3832,
+      "step": 754
+    },
+    {
+      "epoch": 0.36079087270772353,
+      "grad_norm": 0.6058849776764632,
+      "learning_rate": 1.9732693312555492e-05,
+      "loss": 0.4093,
+      "step": 755
+    },
+    {
+      "epoch": 0.3612687414132967,
+      "grad_norm": 0.6417564281631485,
+      "learning_rate": 1.9731482374633203e-05,
+      "loss": 0.4261,
+      "step": 756
+    },
+    {
+      "epoch": 0.3617466101188698,
+      "grad_norm": 0.6139727827460876,
+      "learning_rate": 1.973026873736936e-05,
+      "loss": 0.4099,
+      "step": 757
+    },
+    {
+      "epoch": 0.362224478824443,
+      "grad_norm": 0.6510466515537121,
+      "learning_rate": 1.972905240110061e-05,
+      "loss": 0.4054,
+      "step": 758
+    },
+    {
+      "epoch": 0.3627023475300161,
+      "grad_norm": 0.6367387723523807,
+      "learning_rate": 1.9727833366164342e-05,
+      "loss": 0.4288,
+      "step": 759
+    },
+    {
+      "epoch": 0.36318021623558927,
+      "grad_norm": 0.7047833872183114,
+      "learning_rate": 1.9726611632898693e-05,
+      "loss": 0.4184,
+      "step": 760
+    },
+    {
+      "epoch": 0.3636580849411624,
+      "grad_norm": 0.6625607174095874,
+      "learning_rate": 1.9725387201642553e-05,
+      "loss": 0.3984,
+      "step": 761
+    },
+    {
+      "epoch": 0.36413595364673557,
+      "grad_norm": 0.7543210649256787,
+      "learning_rate": 1.9724160072735553e-05,
+      "loss": 0.4157,
+      "step": 762
+    },
+    {
+      "epoch": 0.3646138223523087,
+      "grad_norm": 0.8409356525795596,
+      "learning_rate": 1.9722930246518083e-05,
+      "loss": 0.3997,
+      "step": 763
+    },
+    {
+      "epoch": 0.36509169105788186,
+      "grad_norm": 0.6419317088471457,
+      "learning_rate": 1.9721697723331273e-05,
+      "loss": 0.4059,
+      "step": 764
+    },
+    {
+      "epoch": 0.365569559763455,
+      "grad_norm": 0.739982349078809,
+      "learning_rate": 1.9720462503517e-05,
+      "loss": 0.4118,
+      "step": 765
+    },
+    {
+      "epoch": 0.36604742846902816,
+      "grad_norm": 0.6898565679952424,
+      "learning_rate": 1.9719224587417896e-05,
+      "loss": 0.4039,
+      "step": 766
+    },
+    {
+      "epoch": 0.3665252971746013,
+      "grad_norm": 0.7612996038765697,
+      "learning_rate": 1.971798397537733e-05,
+      "loss": 0.4153,
+      "step": 767
+    },
+    {
+      "epoch": 0.3670031658801744,
+      "grad_norm": 0.6660225921682122,
+      "learning_rate": 1.971674066773944e-05,
+      "loss": 0.4148,
+      "step": 768
+    },
+    {
+      "epoch": 0.36748103458574755,
+      "grad_norm": 0.7220884614015122,
+      "learning_rate": 1.9715494664849088e-05,
+      "loss": 0.4055,
+      "step": 769
+    },
+    {
+      "epoch": 0.3679589032913207,
+      "grad_norm": 0.677877244256832,
+      "learning_rate": 1.971424596705189e-05,
+      "loss": 0.4119,
+      "step": 770
+    },
+    {
+      "epoch": 0.36843677199689384,
+      "grad_norm": 0.7526211652684072,
+      "learning_rate": 1.971299457469422e-05,
+      "loss": 0.4177,
+      "step": 771
+    },
+    {
+      "epoch": 0.368914640702467,
+      "grad_norm": 0.6114869523814213,
+      "learning_rate": 1.971174048812319e-05,
+      "loss": 0.4101,
+      "step": 772
+    },
+    {
+      "epoch": 0.36939250940804014,
+      "grad_norm": 0.6899914780670549,
+      "learning_rate": 1.971048370768666e-05,
+      "loss": 0.4351,
+      "step": 773
+    },
+    {
+      "epoch": 0.3698703781136133,
+      "grad_norm": 0.6192089565424979,
+      "learning_rate": 1.9709224233733236e-05,
+      "loss": 0.4042,
+      "step": 774
+    },
+    {
+      "epoch": 0.37034824681918643,
+      "grad_norm": 0.7355290743652367,
+      "learning_rate": 1.9707962066612278e-05,
+      "loss": 0.4027,
+      "step": 775
+    },
+    {
+      "epoch": 0.3708261155247596,
+      "grad_norm": 0.7403645775826277,
+      "learning_rate": 1.9706697206673892e-05,
+      "loss": 0.4023,
+      "step": 776
+    },
+    {
+      "epoch": 0.37130398423033273,
+      "grad_norm": 0.6999584425787945,
+      "learning_rate": 1.9705429654268925e-05,
+      "loss": 0.4097,
+      "step": 777
+    },
+    {
+      "epoch": 0.3717818529359059,
+      "grad_norm": 0.6998503016090031,
+      "learning_rate": 1.9704159409748967e-05,
+      "loss": 0.4081,
+      "step": 778
+    },
+    {
+      "epoch": 0.372259721641479,
+      "grad_norm": 0.6443494078717972,
+      "learning_rate": 1.9702886473466368e-05,
+      "loss": 0.4077,
+      "step": 779
+    },
+    {
+      "epoch": 0.3727375903470522,
+      "grad_norm": 0.6852044402539894,
+      "learning_rate": 1.970161084577422e-05,
+      "loss": 0.4193,
+      "step": 780
+    },
+    {
+      "epoch": 0.37321545905262526,
+      "grad_norm": 0.6236384893553105,
+      "learning_rate": 1.970033252702636e-05,
+      "loss": 0.4031,
+      "step": 781
+    },
+    {
+      "epoch": 0.3736933277581984,
+      "grad_norm": 0.6469866739392119,
+      "learning_rate": 1.969905151757736e-05,
+      "loss": 0.4099,
+      "step": 782
+    },
+    {
+      "epoch": 0.37417119646377156,
+      "grad_norm": 0.6504567142773998,
+      "learning_rate": 1.9697767817782565e-05,
+      "loss": 0.4075,
+      "step": 783
+    },
+    {
+      "epoch": 0.3746490651693447,
+      "grad_norm": 0.7743161467373921,
+      "learning_rate": 1.969648142799804e-05,
+      "loss": 0.4241,
+      "step": 784
+    },
+    {
+      "epoch": 0.37512693387491786,
+      "grad_norm": 0.61791736904947,
+      "learning_rate": 1.9695192348580606e-05,
+      "loss": 0.4073,
+      "step": 785
+    },
+    {
+      "epoch": 0.375604802580491,
+      "grad_norm": 0.6551482938668043,
+      "learning_rate": 1.969390057988784e-05,
+      "loss": 0.4157,
+      "step": 786
+    },
+    {
+      "epoch": 0.37608267128606415,
+      "grad_norm": 0.6241776375892095,
+      "learning_rate": 1.9692606122278047e-05,
+      "loss": 0.3991,
+      "step": 787
+    },
+    {
+      "epoch": 0.3765605399916373,
+      "grad_norm": 0.8697779384559257,
+      "learning_rate": 1.9691308976110293e-05,
+      "loss": 0.3943,
+      "step": 788
+    },
+    {
+      "epoch": 0.37703840869721045,
+      "grad_norm": 0.6646525801024487,
+      "learning_rate": 1.969000914174438e-05,
+      "loss": 0.4011,
+      "step": 789
+    },
+    {
+      "epoch": 0.3775162774027836,
+      "grad_norm": 0.6842360997446302,
+      "learning_rate": 1.9688706619540863e-05,
+      "loss": 0.4229,
+      "step": 790
+    },
+    {
+      "epoch": 0.37799414610835674,
+      "grad_norm": 0.6269010992340172,
+      "learning_rate": 1.9687401409861032e-05,
+      "loss": 0.4165,
+      "step": 791
+    },
+    {
+      "epoch": 0.3784720148139299,
+      "grad_norm": 0.6948929157285494,
+      "learning_rate": 1.9686093513066933e-05,
+      "loss": 0.4189,
+      "step": 792
+    },
+    {
+      "epoch": 0.37894988351950304,
+      "grad_norm": 0.6938250597089114,
+      "learning_rate": 1.9684782929521355e-05,
+      "loss": 0.4382,
+      "step": 793
+    },
+    {
+      "epoch": 0.3794277522250762,
+      "grad_norm": 0.710536040353905,
+      "learning_rate": 1.9683469659587826e-05,
+      "loss": 0.4018,
+      "step": 794
+    },
+    {
+      "epoch": 0.3799056209306493,
+      "grad_norm": 0.7765930152534725,
+      "learning_rate": 1.968215370363063e-05,
+      "loss": 0.4025,
+      "step": 795
+    },
+    {
+      "epoch": 0.3803834896362224,
+      "grad_norm": 0.673786846379775,
+      "learning_rate": 1.9680835062014784e-05,
+      "loss": 0.4072,
+      "step": 796
+    },
+    {
+      "epoch": 0.3808613583417956,
+      "grad_norm": 0.6624205844915216,
+      "learning_rate": 1.967951373510606e-05,
+      "loss": 0.4029,
+      "step": 797
+    },
+    {
+      "epoch": 0.3813392270473687,
+      "grad_norm": 0.6702031183000383,
+      "learning_rate": 1.967818972327097e-05,
+      "loss": 0.3933,
+      "step": 798
+    },
+    {
+      "epoch": 0.38181709575294187,
+      "grad_norm": 0.6643830669329299,
+      "learning_rate": 1.967686302687677e-05,
+      "loss": 0.4078,
+      "step": 799
+    },
+    {
+      "epoch": 0.382294964458515,
+      "grad_norm": 0.6820114775894023,
+      "learning_rate": 1.9675533646291463e-05,
+      "loss": 0.4019,
+      "step": 800
+    },
+    {
+      "epoch": 0.38277283316408817,
+      "grad_norm": 0.6654337293955677,
+      "learning_rate": 1.9674201581883796e-05,
+      "loss": 0.4143,
+      "step": 801
+    },
+    {
+      "epoch": 0.3832507018696613,
+      "grad_norm": 0.6277257073491326,
+      "learning_rate": 1.9672866834023263e-05,
+      "loss": 0.3954,
+      "step": 802
+    },
+    {
+      "epoch": 0.38372857057523446,
+      "grad_norm": 1.181246701089062,
+      "learning_rate": 1.9671529403080095e-05,
+      "loss": 0.4115,
+      "step": 803
+    },
+    {
+      "epoch": 0.3842064392808076,
+      "grad_norm": 0.6999373807721258,
+      "learning_rate": 1.9670189289425273e-05,
+      "loss": 0.4085,
+      "step": 804
+    },
+    {
+      "epoch": 0.38468430798638076,
+      "grad_norm": 0.6351836651328577,
+      "learning_rate": 1.9668846493430522e-05,
+      "loss": 0.4024,
+      "step": 805
+    },
+    {
+      "epoch": 0.3851621766919539,
+      "grad_norm": 0.6200352215300432,
+      "learning_rate": 1.966750101546831e-05,
+      "loss": 0.4217,
+      "step": 806
+    },
+    {
+      "epoch": 0.38564004539752705,
+      "grad_norm": 0.6312607230160627,
+      "learning_rate": 1.9666152855911845e-05,
+      "loss": 0.4148,
+      "step": 807
+    },
+    {
+      "epoch": 0.38611791410310015,
+      "grad_norm": 0.6247739745978539,
+      "learning_rate": 1.966480201513509e-05,
+      "loss": 0.4129,
+      "step": 808
+    },
+    {
+      "epoch": 0.3865957828086733,
+      "grad_norm": 23.521606298262835,
+      "learning_rate": 1.966344849351274e-05,
+      "loss": 0.3985,
+      "step": 809
+    },
+    {
+      "epoch": 0.38707365151424644,
+      "grad_norm": 0.7307789504862988,
+      "learning_rate": 1.9662092291420233e-05,
+      "loss": 0.4355,
+      "step": 810
+    },
+    {
+      "epoch": 0.3875515202198196,
+      "grad_norm": 0.6338895925736729,
+      "learning_rate": 1.9660733409233763e-05,
+      "loss": 0.4114,
+      "step": 811
+    },
+    {
+      "epoch": 0.38802938892539274,
+      "grad_norm": 1.022414001164045,
+      "learning_rate": 1.965937184733026e-05,
+      "loss": 0.4177,
+      "step": 812
+    },
+    {
+      "epoch": 0.3885072576309659,
+      "grad_norm": 0.6257154243428295,
+      "learning_rate": 1.965800760608739e-05,
+      "loss": 0.4238,
+      "step": 813
+    },
+    {
+      "epoch": 0.38898512633653903,
+      "grad_norm": 0.6578526200723673,
+      "learning_rate": 1.965664068588358e-05,
+      "loss": 0.4027,
+      "step": 814
+    },
+    {
+      "epoch": 0.3894629950421122,
+      "grad_norm": 0.6394160956438139,
+      "learning_rate": 1.965527108709798e-05,
+      "loss": 0.4047,
+      "step": 815
+    },
+    {
+      "epoch": 0.38994086374768533,
+      "grad_norm": 0.6469488408202481,
+      "learning_rate": 1.96538988101105e-05,
+      "loss": 0.4035,
+      "step": 816
+    },
+    {
+      "epoch": 0.3904187324532585,
+      "grad_norm": 0.6625645754583935,
+      "learning_rate": 1.9652523855301783e-05,
+      "loss": 0.4144,
+      "step": 817
+    },
+    {
+      "epoch": 0.3908966011588316,
+      "grad_norm": 0.6841818131947363,
+      "learning_rate": 1.9651146223053213e-05,
+      "loss": 0.389,
+      "step": 818
+    },
+    {
+      "epoch": 0.3913744698644048,
+      "grad_norm": 0.6438625312526988,
+      "learning_rate": 1.9649765913746923e-05,
+      "loss": 0.4111,
+      "step": 819
+    },
+    {
+      "epoch": 0.3918523385699779,
+      "grad_norm": 1.01601163370014,
+      "learning_rate": 1.964838292776579e-05,
+      "loss": 0.4031,
+      "step": 820
+    },
+    {
+      "epoch": 0.39233020727555107,
+      "grad_norm": 0.677050070767038,
+      "learning_rate": 1.9646997265493426e-05,
+      "loss": 0.4149,
+      "step": 821
+    },
+    {
+      "epoch": 0.39280807598112416,
+      "grad_norm": 0.6738615859070755,
+      "learning_rate": 1.9645608927314194e-05,
+      "loss": 0.4123,
+      "step": 822
+    },
+    {
+      "epoch": 0.3932859446866973,
+      "grad_norm": 0.6651159143963696,
+      "learning_rate": 1.9644217913613187e-05,
+      "loss": 0.4057,
+      "step": 823
+    },
+    {
+      "epoch": 0.39376381339227046,
+      "grad_norm": 0.7691000084987812,
+      "learning_rate": 1.9642824224776252e-05,
+      "loss": 0.4046,
+      "step": 824
+    },
+    {
+      "epoch": 0.3942416820978436,
+      "grad_norm": 0.7211243084062602,
+      "learning_rate": 1.9641427861189973e-05,
+      "loss": 0.4089,
+      "step": 825
+    },
+    {
+      "epoch": 0.39471955080341675,
+      "grad_norm": 0.63694129075116,
+      "learning_rate": 1.964002882324168e-05,
+      "loss": 0.3879,
+      "step": 826
+    },
+    {
+      "epoch": 0.3951974195089899,
+      "grad_norm": 0.5886615227252643,
+      "learning_rate": 1.9638627111319437e-05,
+      "loss": 0.4098,
+      "step": 827
+    },
+    {
+      "epoch": 0.39567528821456305,
+      "grad_norm": 0.6720674176320514,
+      "learning_rate": 1.963722272581206e-05,
+      "loss": 0.4014,
+      "step": 828
+    },
+    {
+      "epoch": 0.3961531569201362,
+      "grad_norm": 0.5791735119683044,
+      "learning_rate": 1.963581566710909e-05,
+      "loss": 0.4008,
+      "step": 829
+    },
+    {
+      "epoch": 0.39663102562570934,
+      "grad_norm": 0.6464775042145513,
+      "learning_rate": 1.963440593560083e-05,
+      "loss": 0.4164,
+      "step": 830
+    },
+    {
+      "epoch": 0.3971088943312825,
+      "grad_norm": 0.6144881514988998,
+      "learning_rate": 1.963299353167831e-05,
+      "loss": 0.411,
+      "step": 831
+    },
+    {
+      "epoch": 0.39758676303685564,
+      "grad_norm": 0.670072169690497,
+      "learning_rate": 1.9631578455733307e-05,
+      "loss": 0.4089,
+      "step": 832
+    },
+    {
+      "epoch": 0.3980646317424288,
+      "grad_norm": 0.6989267984529208,
+      "learning_rate": 1.9630160708158343e-05,
+      "loss": 0.406,
+      "step": 833
+    },
+    {
+      "epoch": 0.39854250044800194,
+      "grad_norm": 1.073909712530086,
+      "learning_rate": 1.9628740289346668e-05,
+      "loss": 0.395,
+      "step": 834
+    },
+    {
+      "epoch": 0.3990203691535751,
+      "grad_norm": 0.9121845119973022,
+      "learning_rate": 1.9627317199692287e-05,
+      "loss": 0.4055,
+      "step": 835
+    },
+    {
+      "epoch": 0.3994982378591482,
+      "grad_norm": 0.6116187864851871,
+      "learning_rate": 1.9625891439589933e-05,
+      "loss": 0.3942,
+      "step": 836
+    },
+    {
+      "epoch": 0.3999761065647213,
+      "grad_norm": 0.6622950577654838,
+      "learning_rate": 1.9624463009435097e-05,
+      "loss": 0.4037,
+      "step": 837
+    },
+    {
+      "epoch": 0.40045397527029447,
+      "grad_norm": 0.7953106318152663,
+      "learning_rate": 1.9623031909623993e-05,
+      "loss": 0.4165,
+      "step": 838
+    },
+    {
+      "epoch": 0.4009318439758676,
+      "grad_norm": 0.6653838350357102,
+      "learning_rate": 1.962159814055358e-05,
+      "loss": 0.4084,
+      "step": 839
+    },
+    {
+      "epoch": 0.40140971268144077,
+      "grad_norm": 0.6890324725741804,
+      "learning_rate": 1.962016170262157e-05,
+      "loss": 0.4236,
+      "step": 840
+    },
+    {
+      "epoch": 0.4018875813870139,
+      "grad_norm": 0.6550797312828021,
+      "learning_rate": 1.96187225962264e-05,
+      "loss": 0.4096,
+      "step": 841
+    },
+    {
+      "epoch": 0.40236545009258706,
+      "grad_norm": 0.7180548191723333,
+      "learning_rate": 1.9617280821767253e-05,
+      "loss": 0.3889,
+      "step": 842
+    },
+    {
+      "epoch": 0.4028433187981602,
+      "grad_norm": 0.6990218911792896,
+      "learning_rate": 1.9615836379644054e-05,
+      "loss": 0.3876,
+      "step": 843
+    },
+    {
+      "epoch": 0.40332118750373336,
+      "grad_norm": 0.8178032329800812,
+      "learning_rate": 1.961438927025746e-05,
+      "loss": 0.3978,
+      "step": 844
+    },
+    {
+      "epoch": 0.4037990562093065,
+      "grad_norm": 0.5792010743662441,
+      "learning_rate": 1.961293949400888e-05,
+      "loss": 0.4065,
+      "step": 845
+    },
+    {
+      "epoch": 0.40427692491487965,
+      "grad_norm": 0.6641398130174404,
+      "learning_rate": 1.9611487051300454e-05,
+      "loss": 0.3914,
+      "step": 846
+    },
+    {
+      "epoch": 0.4047547936204528,
+      "grad_norm": 0.6195936771890139,
+      "learning_rate": 1.961003194253506e-05,
+      "loss": 0.402,
+      "step": 847
+    },
+    {
+      "epoch": 0.40523266232602595,
+      "grad_norm": 0.6768052082314968,
+      "learning_rate": 1.9608574168116324e-05,
+      "loss": 0.4199,
+      "step": 848
+    },
+    {
+      "epoch": 0.40571053103159904,
+      "grad_norm": 0.6082693604746662,
+      "learning_rate": 1.960711372844861e-05,
+      "loss": 0.4074,
+      "step": 849
+    },
+    {
+      "epoch": 0.4061883997371722,
+      "grad_norm": 0.6268219384593703,
+      "learning_rate": 1.960565062393701e-05,
+      "loss": 0.4157,
+      "step": 850
+    },
+    {
+      "epoch": 0.40666626844274534,
+      "grad_norm": 0.6158932362492403,
+      "learning_rate": 1.960418485498737e-05,
+      "loss": 0.3942,
+      "step": 851
+    },
+    {
+      "epoch": 0.4071441371483185,
+      "grad_norm": 0.7217988468496267,
+      "learning_rate": 1.9602716422006266e-05,
+      "loss": 0.4123,
+      "step": 852
+    },
+    {
+      "epoch": 0.40762200585389163,
+      "grad_norm": 0.6549013187083295,
+      "learning_rate": 1.9601245325401016e-05,
+      "loss": 0.3881,
+      "step": 853
+    },
+    {
+      "epoch": 0.4080998745594648,
+      "grad_norm": 0.6163590041751829,
+      "learning_rate": 1.9599771565579673e-05,
+      "loss": 0.4224,
+      "step": 854
+    },
+    {
+      "epoch": 0.40857774326503793,
+      "grad_norm": 0.8712876541825809,
+      "learning_rate": 1.9598295142951035e-05,
+      "loss": 0.4042,
+      "step": 855
+    },
+    {
+      "epoch": 0.4090556119706111,
+      "grad_norm": 0.7278398539493856,
+      "learning_rate": 1.959681605792464e-05,
+      "loss": 0.4164,
+      "step": 856
+    },
+    {
+      "epoch": 0.4095334806761842,
+      "grad_norm": 0.600386439964583,
+      "learning_rate": 1.9595334310910753e-05,
+      "loss": 0.4135,
+      "step": 857
+    },
+    {
+      "epoch": 0.4100113493817574,
+      "grad_norm": 0.7081480085926267,
+      "learning_rate": 1.9593849902320386e-05,
+      "loss": 0.4027,
+      "step": 858
+    },
+    {
+      "epoch": 0.4104892180873305,
+      "grad_norm": 0.6188620046952817,
+      "learning_rate": 1.9592362832565287e-05,
+      "loss": 0.4016,
+      "step": 859
+    },
+    {
+      "epoch": 0.41096708679290367,
+      "grad_norm": 0.682019844909097,
+      "learning_rate": 1.9590873102057948e-05,
+      "loss": 0.4071,
+      "step": 860
+    },
+    {
+      "epoch": 0.4114449554984768,
+      "grad_norm": 0.6603131181430578,
+      "learning_rate": 1.9589380711211588e-05,
+      "loss": 0.3946,
+      "step": 861
+    },
+    {
+      "epoch": 0.41192282420404996,
+      "grad_norm": 0.6577106617903441,
+      "learning_rate": 1.9587885660440176e-05,
+      "loss": 0.4133,
+      "step": 862
+    },
+    {
+      "epoch": 0.41240069290962306,
+      "grad_norm": 0.6628545122914132,
+      "learning_rate": 1.9586387950158406e-05,
+      "loss": 0.4131,
+      "step": 863
+    },
+    {
+      "epoch": 0.4128785616151962,
+      "grad_norm": 0.6190751337142328,
+      "learning_rate": 1.958488758078172e-05,
+      "loss": 0.4191,
+      "step": 864
+    },
+    {
+      "epoch": 0.41335643032076935,
+      "grad_norm": 0.7460256010176625,
+      "learning_rate": 1.9583384552726294e-05,
+      "loss": 0.3936,
+      "step": 865
+    },
+    {
+      "epoch": 0.4138342990263425,
+      "grad_norm": 0.5886477110046239,
+      "learning_rate": 1.9581878866409042e-05,
+      "loss": 0.4069,
+      "step": 866
+    },
+    {
+      "epoch": 0.41431216773191565,
+      "grad_norm": 0.6886448680334144,
+      "learning_rate": 1.9580370522247614e-05,
+      "loss": 0.397,
+      "step": 867
+    },
+    {
+      "epoch": 0.4147900364374888,
+      "grad_norm": 0.6232904922457612,
+      "learning_rate": 1.9578859520660396e-05,
+      "loss": 0.4029,
+      "step": 868
+    },
+    {
+      "epoch": 0.41526790514306194,
+      "grad_norm": 0.5941334358676716,
+      "learning_rate": 1.9577345862066518e-05,
+      "loss": 0.3878,
+      "step": 869
+    },
+    {
+      "epoch": 0.4157457738486351,
+      "grad_norm": 1.0139675059105102,
+      "learning_rate": 1.957582954688584e-05,
+      "loss": 0.3952,
+      "step": 870
+    },
+    {
+      "epoch": 0.41622364255420824,
+      "grad_norm": 3.476711388368822,
+      "learning_rate": 1.9574310575538956e-05,
+      "loss": 0.4103,
+      "step": 871
+    },
+    {
+      "epoch": 0.4167015112597814,
+      "grad_norm": 0.7611765590691459,
+      "learning_rate": 1.9572788948447206e-05,
+      "loss": 0.4179,
+      "step": 872
+    },
+    {
+      "epoch": 0.41717937996535454,
+      "grad_norm": 0.600202395984294,
+      "learning_rate": 1.9571264666032667e-05,
+      "loss": 0.4074,
+      "step": 873
+    },
+    {
+      "epoch": 0.4176572486709277,
+      "grad_norm": 0.651033950877546,
+      "learning_rate": 1.9569737728718143e-05,
+      "loss": 0.4007,
+      "step": 874
+    },
+    {
+      "epoch": 0.41813511737650083,
+      "grad_norm": 0.6518303558888998,
+      "learning_rate": 1.9568208136927177e-05,
+      "loss": 0.4078,
+      "step": 875
+    },
+    {
+      "epoch": 0.4186129860820739,
+      "grad_norm": 0.6107727192896184,
+      "learning_rate": 1.956667589108406e-05,
+      "loss": 0.4205,
+      "step": 876
+    },
+    {
+      "epoch": 0.41909085478764707,
+      "grad_norm": 0.5977988275186865,
+      "learning_rate": 1.95651409916138e-05,
+      "loss": 0.3981,
+      "step": 877
+    },
+    {
+      "epoch": 0.4195687234932202,
+      "grad_norm": 0.6053736216356521,
+      "learning_rate": 1.9563603438942155e-05,
+      "loss": 0.41,
+      "step": 878
+    },
+    {
+      "epoch": 0.42004659219879337,
+      "grad_norm": 0.7201616837003876,
+      "learning_rate": 1.9562063233495615e-05,
+      "loss": 0.4086,
+      "step": 879
+    },
+    {
+      "epoch": 0.4205244609043665,
+      "grad_norm": 0.6109033092609298,
+      "learning_rate": 1.9560520375701408e-05,
+      "loss": 0.4061,
+      "step": 880
+    },
+    {
+      "epoch": 0.42100232960993966,
+      "grad_norm": 0.6186670618629603,
+      "learning_rate": 1.9558974865987494e-05,
+      "loss": 0.4207,
+      "step": 881
+    },
+    {
+      "epoch": 0.4214801983155128,
+      "grad_norm": 0.6147640108712225,
+      "learning_rate": 1.9557426704782564e-05,
+      "loss": 0.4166,
+      "step": 882
+    },
+    {
+      "epoch": 0.42195806702108596,
+      "grad_norm": 0.6202320300282839,
+      "learning_rate": 1.9555875892516064e-05,
+      "loss": 0.4055,
+      "step": 883
+    },
+    {
+      "epoch": 0.4224359357266591,
+      "grad_norm": 0.6438580742369829,
+      "learning_rate": 1.955432242961815e-05,
+      "loss": 0.4077,
+      "step": 884
+    },
+    {
+      "epoch": 0.42291380443223225,
+      "grad_norm": 0.6294856559072706,
+      "learning_rate": 1.9552766316519726e-05,
+      "loss": 0.4149,
+      "step": 885
+    },
+    {
+      "epoch": 0.4233916731378054,
+      "grad_norm": 0.6161283088521784,
+      "learning_rate": 1.955120755365244e-05,
+      "loss": 0.411,
+      "step": 886
+    },
+    {
+      "epoch": 0.42386954184337855,
+      "grad_norm": 0.6014577535152139,
+      "learning_rate": 1.9549646141448657e-05,
+      "loss": 0.3914,
+      "step": 887
+    },
+    {
+      "epoch": 0.4243474105489517,
+      "grad_norm": 0.6848333020026872,
+      "learning_rate": 1.9548082080341486e-05,
+      "loss": 0.4001,
+      "step": 888
+    },
+    {
+      "epoch": 0.42482527925452485,
+      "grad_norm": 0.6720870979122917,
+      "learning_rate": 1.954651537076477e-05,
+      "loss": 0.4238,
+      "step": 889
+    },
+    {
+      "epoch": 0.42530314796009794,
+      "grad_norm": 0.5818304388573133,
+      "learning_rate": 1.9544946013153093e-05,
+      "loss": 0.4042,
+      "step": 890
+    },
+    {
+      "epoch": 0.4257810166656711,
+      "grad_norm": 0.7744562017315486,
+      "learning_rate": 1.9543374007941756e-05,
+      "loss": 0.4238,
+      "step": 891
+    },
+    {
+      "epoch": 0.42625888537124423,
+      "grad_norm": 0.7762559725025642,
+      "learning_rate": 1.9541799355566813e-05,
+      "loss": 0.4111,
+      "step": 892
+    },
+    {
+      "epoch": 0.4267367540768174,
+      "grad_norm": 0.6599610950235643,
+      "learning_rate": 1.9540222056465046e-05,
+      "loss": 0.411,
+      "step": 893
+    },
+    {
+      "epoch": 0.42721462278239053,
+      "grad_norm": 0.7413734398856531,
+      "learning_rate": 1.9538642111073966e-05,
+      "loss": 0.417,
+      "step": 894
+    },
+    {
+      "epoch": 0.4276924914879637,
+      "grad_norm": 0.6502451354585309,
+      "learning_rate": 1.9537059519831822e-05,
+      "loss": 0.416,
+      "step": 895
+    },
+    {
+      "epoch": 0.4281703601935368,
+      "grad_norm": 0.8797654195920014,
+      "learning_rate": 1.9535474283177597e-05,
+      "loss": 0.4156,
+      "step": 896
+    },
+    {
+      "epoch": 0.42864822889911,
+      "grad_norm": 0.5595634974263763,
+      "learning_rate": 1.953388640155101e-05,
+      "loss": 0.4082,
+      "step": 897
+    },
+    {
+      "epoch": 0.4291260976046831,
+      "grad_norm": 0.6530429716085568,
+      "learning_rate": 1.953229587539251e-05,
+      "loss": 0.4025,
+      "step": 898
+    },
+    {
+      "epoch": 0.42960396631025627,
+      "grad_norm": 0.6657647909949589,
+      "learning_rate": 1.953070270514328e-05,
+      "loss": 0.4143,
+      "step": 899
+    },
+    {
+      "epoch": 0.4300818350158294,
+      "grad_norm": 0.7463671068027646,
+      "learning_rate": 1.9529106891245244e-05,
+      "loss": 0.3887,
+      "step": 900
+    },
+    {
+      "epoch": 0.43055970372140256,
+      "grad_norm": 0.6059170836222347,
+      "learning_rate": 1.952750843414104e-05,
+      "loss": 0.4028,
+      "step": 901
+    },
+    {
+      "epoch": 0.4310375724269757,
+      "grad_norm": 0.6074415817250135,
+      "learning_rate": 1.9525907334274063e-05,
+      "loss": 0.3867,
+      "step": 902
+    },
+    {
+      "epoch": 0.4315154411325488,
+      "grad_norm": 0.6460281093432895,
+      "learning_rate": 1.9524303592088424e-05,
+      "loss": 0.403,
+      "step": 903
+    },
+    {
+      "epoch": 0.43199330983812195,
+      "grad_norm": 0.6192867647004487,
+      "learning_rate": 1.9522697208028975e-05,
+      "loss": 0.4004,
+      "step": 904
+    },
+    {
+      "epoch": 0.4324711785436951,
+      "grad_norm": 0.7247569688470655,
+      "learning_rate": 1.9521088182541298e-05,
+      "loss": 0.4167,
+      "step": 905
+    },
+    {
+      "epoch": 0.43294904724926825,
+      "grad_norm": 0.7590859276414997,
+      "learning_rate": 1.9519476516071706e-05,
+      "loss": 0.4002,
+      "step": 906
+    },
+    {
+      "epoch": 0.4334269159548414,
+      "grad_norm": 0.5841622232228564,
+      "learning_rate": 1.951786220906725e-05,
+      "loss": 0.4182,
+      "step": 907
+    },
+    {
+      "epoch": 0.43390478466041454,
+      "grad_norm": 0.6692111403784917,
+      "learning_rate": 1.951624526197571e-05,
+      "loss": 0.401,
+      "step": 908
+    },
+    {
+      "epoch": 0.4343826533659877,
+      "grad_norm": 0.6097916498450394,
+      "learning_rate": 1.95146256752456e-05,
+      "loss": 0.4001,
+      "step": 909
+    },
+    {
+      "epoch": 0.43486052207156084,
+      "grad_norm": 0.7357451300421103,
+      "learning_rate": 1.951300344932616e-05,
+      "loss": 0.4279,
+      "step": 910
+    },
+    {
+      "epoch": 0.435338390777134,
+      "grad_norm": 1.4472772781586367,
+      "learning_rate": 1.9511378584667372e-05,
+      "loss": 0.409,
+      "step": 911
+    },
+    {
+      "epoch": 0.43581625948270714,
+      "grad_norm": 0.6412845231251528,
+      "learning_rate": 1.950975108171994e-05,
+      "loss": 0.4232,
+      "step": 912
+    },
+    {
+      "epoch": 0.4362941281882803,
+      "grad_norm": 0.6815096104930907,
+      "learning_rate": 1.950812094093531e-05,
+      "loss": 0.3989,
+      "step": 913
+    },
+    {
+      "epoch": 0.43677199689385343,
+      "grad_norm": 0.6287441394449275,
+      "learning_rate": 1.950648816276565e-05,
+      "loss": 0.4141,
+      "step": 914
+    },
+    {
+      "epoch": 0.4372498655994266,
+      "grad_norm": 0.6475760769332336,
+      "learning_rate": 1.9504852747663862e-05,
+      "loss": 0.4116,
+      "step": 915
+    },
+    {
+      "epoch": 0.4377277343049997,
+      "grad_norm": 0.5697529534279385,
+      "learning_rate": 1.9503214696083587e-05,
+      "loss": 0.3941,
+      "step": 916
+    },
+    {
+      "epoch": 0.4382056030105728,
+      "grad_norm": 0.6214698860437475,
+      "learning_rate": 1.9501574008479188e-05,
+      "loss": 0.4199,
+      "step": 917
+    },
+    {
+      "epoch": 0.43868347171614597,
+      "grad_norm": 0.5978885664378845,
+      "learning_rate": 1.9499930685305767e-05,
+      "loss": 0.3966,
+      "step": 918
+    },
+    {
+      "epoch": 0.4391613404217191,
+      "grad_norm": 0.6051257671018444,
+      "learning_rate": 1.949828472701915e-05,
+      "loss": 0.4008,
+      "step": 919
+    },
+    {
+      "epoch": 0.43963920912729226,
+      "grad_norm": 0.6920182979659302,
+      "learning_rate": 1.9496636134075894e-05,
+      "loss": 0.3959,
+      "step": 920
+    },
+    {
+      "epoch": 0.4401170778328654,
+      "grad_norm": 0.650235126033297,
+      "learning_rate": 1.9494984906933293e-05,
+      "loss": 0.3951,
+      "step": 921
+    },
+    {
+      "epoch": 0.44059494653843856,
+      "grad_norm": 0.683430884067445,
+      "learning_rate": 1.9493331046049366e-05,
+      "loss": 0.398,
+      "step": 922
+    },
+    {
+      "epoch": 0.4410728152440117,
+      "grad_norm": 0.6452456126774382,
+      "learning_rate": 1.9491674551882867e-05,
+      "loss": 0.389,
+      "step": 923
+    },
+    {
+      "epoch": 0.44155068394958485,
+      "grad_norm": 0.695330236916925,
+      "learning_rate": 1.9490015424893277e-05,
+      "loss": 0.4016,
+      "step": 924
+    },
+    {
+      "epoch": 0.442028552655158,
+      "grad_norm": 0.6316596603137281,
+      "learning_rate": 1.9488353665540813e-05,
+      "loss": 0.3978,
+      "step": 925
+    },
+    {
+      "epoch": 0.44250642136073115,
+      "grad_norm": 0.8094826470406035,
+      "learning_rate": 1.9486689274286413e-05,
+      "loss": 0.3837,
+      "step": 926
+    },
+    {
+      "epoch": 0.4429842900663043,
+      "grad_norm": 0.8597357950448963,
+      "learning_rate": 1.9485022251591744e-05,
+      "loss": 0.4175,
+      "step": 927
+    },
+    {
+      "epoch": 0.44346215877187745,
+      "grad_norm": 0.6850435114555503,
+      "learning_rate": 1.9483352597919222e-05,
+      "loss": 0.416,
+      "step": 928
+    },
+    {
+      "epoch": 0.4439400274774506,
+      "grad_norm": 0.6105257300932792,
+      "learning_rate": 1.9481680313731973e-05,
+      "loss": 0.4087,
+      "step": 929
+    },
+    {
+      "epoch": 0.44441789618302374,
+      "grad_norm": 0.7783073049281799,
+      "learning_rate": 1.9480005399493857e-05,
+      "loss": 0.4084,
+      "step": 930
+    },
+    {
+      "epoch": 0.44489576488859683,
+      "grad_norm": 0.8160008351900309,
+      "learning_rate": 1.9478327855669468e-05,
+      "loss": 0.4102,
+      "step": 931
+    },
+    {
+      "epoch": 0.44537363359417,
+      "grad_norm": 0.6251532938868944,
+      "learning_rate": 1.9476647682724125e-05,
+      "loss": 0.4028,
+      "step": 932
+    },
+    {
+      "epoch": 0.44585150229974313,
+      "grad_norm": 0.7279316882881499,
+      "learning_rate": 1.9474964881123883e-05,
+      "loss": 0.4085,
+      "step": 933
+    },
+    {
+      "epoch": 0.4463293710053163,
+      "grad_norm": 0.6648782797147017,
+      "learning_rate": 1.9473279451335517e-05,
+      "loss": 0.412,
+      "step": 934
+    },
+    {
+      "epoch": 0.4468072397108894,
+      "grad_norm": 0.7245218627960338,
+      "learning_rate": 1.9471591393826536e-05,
+      "loss": 0.3952,
+      "step": 935
+    },
+    {
+      "epoch": 0.4472851084164626,
+      "grad_norm": 0.7468126769642631,
+      "learning_rate": 1.9469900709065176e-05,
+      "loss": 0.4059,
+      "step": 936
+    },
+    {
+      "epoch": 0.4477629771220357,
+      "grad_norm": 0.7157504712908894,
+      "learning_rate": 1.9468207397520413e-05,
+      "loss": 0.3994,
+      "step": 937
+    },
+    {
+      "epoch": 0.44824084582760887,
+      "grad_norm": 0.6310072705677632,
+      "learning_rate": 1.946651145966193e-05,
+      "loss": 0.406,
+      "step": 938
+    },
+    {
+      "epoch": 0.448718714533182,
+      "grad_norm": 0.6784353372400441,
+      "learning_rate": 1.9464812895960152e-05,
+      "loss": 0.4009,
+      "step": 939
+    },
+    {
+      "epoch": 0.44919658323875516,
+      "grad_norm": 0.6479984452905685,
+      "learning_rate": 1.9463111706886234e-05,
+      "loss": 0.4168,
+      "step": 940
+    },
+    {
+      "epoch": 0.4496744519443283,
+      "grad_norm": 0.6287167548848303,
+      "learning_rate": 1.9461407892912055e-05,
+      "loss": 0.4014,
+      "step": 941
+    },
+    {
+      "epoch": 0.45015232064990146,
+      "grad_norm": 0.7024078893645669,
+      "learning_rate": 1.9459701454510228e-05,
+      "loss": 0.4097,
+      "step": 942
+    },
+    {
+      "epoch": 0.4506301893554746,
+      "grad_norm": 0.6977676751624544,
+      "learning_rate": 1.945799239215408e-05,
+      "loss": 0.4026,
+      "step": 943
+    },
+    {
+      "epoch": 0.4511080580610477,
+      "grad_norm": 0.64820079330831,
+      "learning_rate": 1.945628070631768e-05,
+      "loss": 0.3986,
+      "step": 944
+    },
+    {
+      "epoch": 0.45158592676662085,
+      "grad_norm": 0.7286318274900389,
+      "learning_rate": 1.9454566397475813e-05,
+      "loss": 0.3888,
+      "step": 945
+    },
+    {
+      "epoch": 0.452063795472194,
+      "grad_norm": 1.4877125746903945,
+      "learning_rate": 1.9452849466104008e-05,
+      "loss": 0.4151,
+      "step": 946
+    },
+    {
+      "epoch": 0.45254166417776714,
+      "grad_norm": 0.6879684254386736,
+      "learning_rate": 1.9451129912678506e-05,
+      "loss": 0.3934,
+      "step": 947
+    },
+    {
+      "epoch": 0.4530195328833403,
+      "grad_norm": 0.6050674990436092,
+      "learning_rate": 1.9449407737676277e-05,
+      "loss": 0.4019,
+      "step": 948
+    },
+    {
+      "epoch": 0.45349740158891344,
+      "grad_norm": 0.6424207582798632,
+      "learning_rate": 1.9447682941575032e-05,
+      "loss": 0.4142,
+      "step": 949
+    },
+    {
+      "epoch": 0.4539752702944866,
+      "grad_norm": 0.6584464274725602,
+      "learning_rate": 1.944595552485319e-05,
+      "loss": 0.4112,
+      "step": 950
+    },
+    {
+      "epoch": 0.45445313900005974,
+      "grad_norm": 1.8201606803835755,
+      "learning_rate": 1.9444225487989912e-05,
+      "loss": 0.3986,
+      "step": 951
+    },
+    {
+      "epoch": 0.4549310077056329,
+      "grad_norm": 0.6649041164921152,
+      "learning_rate": 1.9442492831465075e-05,
+      "loss": 0.4124,
+      "step": 952
+    },
+    {
+      "epoch": 0.45540887641120603,
+      "grad_norm": 0.6103343936848789,
+      "learning_rate": 1.944075755575929e-05,
+      "loss": 0.402,
+      "step": 953
+    },
+    {
+      "epoch": 0.4558867451167792,
+      "grad_norm": 1.1183419241012187,
+      "learning_rate": 1.943901966135389e-05,
+      "loss": 0.4032,
+      "step": 954
+    },
+    {
+      "epoch": 0.4563646138223523,
+      "grad_norm": 0.6278256875792309,
+      "learning_rate": 1.943727914873094e-05,
+      "loss": 0.397,
+      "step": 955
+    },
+    {
+      "epoch": 0.4568424825279255,
+      "grad_norm": 0.5725976955845966,
+      "learning_rate": 1.943553601837322e-05,
+      "loss": 0.4117,
+      "step": 956
+    },
+    {
+      "epoch": 0.4573203512334986,
+      "grad_norm": 0.7341816058812604,
+      "learning_rate": 1.943379027076425e-05,
+      "loss": 0.389,
+      "step": 957
+    },
+    {
+      "epoch": 0.4577982199390717,
+      "grad_norm": 0.6161118906924941,
+      "learning_rate": 1.943204190638827e-05,
+      "loss": 0.402,
+      "step": 958
+    },
+    {
+      "epoch": 0.45827608864464486,
+      "grad_norm": 0.5931536750814665,
+      "learning_rate": 1.9430290925730245e-05,
+      "loss": 0.4064,
+      "step": 959
+    },
+    {
+      "epoch": 0.458753957350218,
+      "grad_norm": 0.6397321208831805,
+      "learning_rate": 1.9428537329275862e-05,
+      "loss": 0.4029,
+      "step": 960
+    },
+    {
+      "epoch": 0.45923182605579116,
+      "grad_norm": 0.5770485075045017,
+      "learning_rate": 1.942678111751154e-05,
+      "loss": 0.4154,
+      "step": 961
+    },
+    {
+      "epoch": 0.4597096947613643,
+      "grad_norm": 0.6983859402707672,
+      "learning_rate": 1.942502229092442e-05,
+      "loss": 0.3865,
+      "step": 962
+    },
+    {
+      "epoch": 0.46018756346693745,
+      "grad_norm": 0.566499169480957,
+      "learning_rate": 1.9423260850002375e-05,
+      "loss": 0.391,
+      "step": 963
+    },
+    {
+      "epoch": 0.4606654321725106,
+      "grad_norm": 0.6674856926358066,
+      "learning_rate": 1.9421496795233995e-05,
+      "loss": 0.4177,
+      "step": 964
+    },
+    {
+      "epoch": 0.46114330087808375,
+      "grad_norm": 0.6077827552367149,
+      "learning_rate": 1.941973012710859e-05,
+      "loss": 0.4064,
+      "step": 965
+    },
+    {
+      "epoch": 0.4616211695836569,
+      "grad_norm": 0.6393677236982209,
+      "learning_rate": 1.9417960846116214e-05,
+      "loss": 0.4079,
+      "step": 966
+    },
+    {
+      "epoch": 0.46209903828923005,
+      "grad_norm": 0.6681026939421527,
+      "learning_rate": 1.941618895274763e-05,
+      "loss": 0.3953,
+      "step": 967
+    },
+    {
+      "epoch": 0.4625769069948032,
+      "grad_norm": 0.6460174077865963,
+      "learning_rate": 1.9414414447494326e-05,
+      "loss": 0.4012,
+      "step": 968
+    },
+    {
+      "epoch": 0.46305477570037634,
+      "grad_norm": 0.6582432277442615,
+      "learning_rate": 1.9412637330848524e-05,
+      "loss": 0.4101,
+      "step": 969
+    },
+    {
+      "epoch": 0.4635326444059495,
+      "grad_norm": 0.6318462060701575,
+      "learning_rate": 1.941085760330316e-05,
+      "loss": 0.4059,
+      "step": 970
+    },
+    {
+      "epoch": 0.4640105131115226,
+      "grad_norm": 0.7203162939351494,
+      "learning_rate": 1.9409075265351904e-05,
+      "loss": 0.4137,
+      "step": 971
+    },
+    {
+      "epoch": 0.46448838181709573,
+      "grad_norm": 0.6894708541476223,
+      "learning_rate": 1.940729031748914e-05,
+      "loss": 0.409,
+      "step": 972
+    },
+    {
+      "epoch": 0.4649662505226689,
+      "grad_norm": 0.6082868118061796,
+      "learning_rate": 1.9405502760209988e-05,
+      "loss": 0.3811,
+      "step": 973
+    },
+    {
+      "epoch": 0.465444119228242,
+      "grad_norm": 0.575259488569313,
+      "learning_rate": 1.9403712594010275e-05,
+      "loss": 0.3959,
+      "step": 974
+    },
+    {
+      "epoch": 0.4659219879338152,
+      "grad_norm": 0.6187113385107887,
+      "learning_rate": 1.940191981938657e-05,
+      "loss": 0.3927,
+      "step": 975
+    },
+    {
+      "epoch": 0.4663998566393883,
+      "grad_norm": 0.5980885678751732,
+      "learning_rate": 1.9400124436836155e-05,
+      "loss": 0.403,
+      "step": 976
+    },
+    {
+      "epoch": 0.46687772534496147,
+      "grad_norm": 0.59607435141776,
+      "learning_rate": 1.9398326446857034e-05,
+      "loss": 0.4037,
+      "step": 977
+    },
+    {
+      "epoch": 0.4673555940505346,
+      "grad_norm": 0.6121952890654038,
+      "learning_rate": 1.939652584994794e-05,
+      "loss": 0.3928,
+      "step": 978
+    },
+    {
+      "epoch": 0.46783346275610777,
+      "grad_norm": 0.6519511555849895,
+      "learning_rate": 1.9394722646608332e-05,
+      "loss": 0.3984,
+      "step": 979
+    },
+    {
+      "epoch": 0.4683113314616809,
+      "grad_norm": 0.5807245005292517,
+      "learning_rate": 1.9392916837338376e-05,
+      "loss": 0.4097,
+      "step": 980
+    },
+    {
+      "epoch": 0.46878920016725406,
+      "grad_norm": 0.5679167625106275,
+      "learning_rate": 1.939110842263898e-05,
+      "loss": 0.3963,
+      "step": 981
+    },
+    {
+      "epoch": 0.4692670688728272,
+      "grad_norm": 0.6227411181855099,
+      "learning_rate": 1.9389297403011767e-05,
+      "loss": 0.4104,
+      "step": 982
+    },
+    {
+      "epoch": 0.46974493757840036,
+      "grad_norm": 0.5888976376216906,
+      "learning_rate": 1.9387483778959075e-05,
+      "loss": 0.3936,
+      "step": 983
+    },
+    {
+      "epoch": 0.4702228062839735,
+      "grad_norm": 1.09297261340797,
+      "learning_rate": 1.9385667550983974e-05,
+      "loss": 0.4219,
+      "step": 984
+    },
+    {
+      "epoch": 0.4707006749895466,
+      "grad_norm": 0.5838151733918109,
+      "learning_rate": 1.9383848719590257e-05,
+      "loss": 0.4059,
+      "step": 985
+    },
+    {
+      "epoch": 0.47117854369511974,
+      "grad_norm": 0.5684738684896777,
+      "learning_rate": 1.9382027285282437e-05,
+      "loss": 0.386,
+      "step": 986
+    },
+    {
+      "epoch": 0.4716564124006929,
+      "grad_norm": 0.5797945596171794,
+      "learning_rate": 1.9380203248565738e-05,
+      "loss": 0.406,
+      "step": 987
+    },
+    {
+      "epoch": 0.47213428110626604,
+      "grad_norm": 0.5735206647369371,
+      "learning_rate": 1.9378376609946126e-05,
+      "loss": 0.4109,
+      "step": 988
+    },
+    {
+      "epoch": 0.4726121498118392,
+      "grad_norm": 0.5591211466804421,
+      "learning_rate": 1.937654736993027e-05,
+      "loss": 0.4157,
+      "step": 989
+    },
+    {
+      "epoch": 0.47309001851741234,
+      "grad_norm": 0.5955292703363935,
+      "learning_rate": 1.9374715529025575e-05,
+      "loss": 0.3973,
+      "step": 990
+    },
+    {
+      "epoch": 0.4735678872229855,
+      "grad_norm": 0.967343371666392,
+      "learning_rate": 1.9372881087740162e-05,
+      "loss": 0.3997,
+      "step": 991
+    },
+    {
+      "epoch": 0.47404575592855863,
+      "grad_norm": 0.5801781650436966,
+      "learning_rate": 1.9371044046582867e-05,
+      "loss": 0.4082,
+      "step": 992
+    },
+    {
+      "epoch": 0.4745236246341318,
+      "grad_norm": 0.6145894286769649,
+      "learning_rate": 1.9369204406063257e-05,
+      "loss": 0.4004,
+      "step": 993
+    },
+    {
+      "epoch": 0.4750014933397049,
+      "grad_norm": 0.6330844181843237,
+      "learning_rate": 1.936736216669161e-05,
+      "loss": 0.3982,
+      "step": 994
+    },
+    {
+      "epoch": 0.4754793620452781,
+      "grad_norm": 0.5704791125251828,
+      "learning_rate": 1.9365517328978943e-05,
+      "loss": 0.4129,
+      "step": 995
+    },
+    {
+      "epoch": 0.4759572307508512,
+      "grad_norm": 0.5887093260184034,
+      "learning_rate": 1.936366989343697e-05,
+      "loss": 0.4071,
+      "step": 996
+    },
+    {
+      "epoch": 0.47643509945642437,
+      "grad_norm": 0.6293352936706972,
+      "learning_rate": 1.9361819860578143e-05,
+      "loss": 0.4105,
+      "step": 997
+    },
+    {
+      "epoch": 0.47691296816199746,
+      "grad_norm": 0.5368324375592416,
+      "learning_rate": 1.9359967230915622e-05,
+      "loss": 0.3939,
+      "step": 998
+    },
+    {
+      "epoch": 0.4773908368675706,
+      "grad_norm": 0.6637884556032649,
+      "learning_rate": 1.9358112004963304e-05,
+      "loss": 0.4003,
+      "step": 999
+    },
+    {
+      "epoch": 0.47786870557314376,
+      "grad_norm": 0.6269569640161438,
+      "learning_rate": 1.9356254183235785e-05,
+      "loss": 0.3977,
+      "step": 1000
+    },
+    {
+      "epoch": 0.4783465742787169,
+      "grad_norm": 0.5907094927745844,
+      "learning_rate": 1.93543937662484e-05,
+      "loss": 0.3888,
+      "step": 1001
+    },
+    {
+      "epoch": 0.47882444298429006,
+      "grad_norm": 0.6356082307658898,
+      "learning_rate": 1.935253075451719e-05,
+      "loss": 0.4009,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4793023116898632,
+      "grad_norm": 0.598063943812148,
+      "learning_rate": 1.935066514855893e-05,
+      "loss": 0.3996,
+      "step": 1003
+    },
+    {
+      "epoch": 0.47978018039543635,
+      "grad_norm": 0.6878231195525895,
+      "learning_rate": 1.9348796948891094e-05,
+      "loss": 0.3917,
+      "step": 1004
+    },
+    {
+      "epoch": 0.4802580491010095,
+      "grad_norm": 0.5939172442059919,
+      "learning_rate": 1.93469261560319e-05,
+      "loss": 0.3907,
+      "step": 1005
+    },
+    {
+      "epoch": 0.48073591780658265,
+      "grad_norm": 0.7585300644554811,
+      "learning_rate": 1.9345052770500262e-05,
+      "loss": 0.3896,
+      "step": 1006
+    },
+    {
+      "epoch": 0.4812137865121558,
+      "grad_norm": 0.5717895671680949,
+      "learning_rate": 1.934317679281583e-05,
+      "loss": 0.4091,
+      "step": 1007
+    },
+    {
+      "epoch": 0.48169165521772894,
+      "grad_norm": 0.6208962871977249,
+      "learning_rate": 1.9341298223498973e-05,
+      "loss": 0.3867,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4821695239233021,
+      "grad_norm": 0.66712603907002,
+      "learning_rate": 1.933941706307076e-05,
+      "loss": 0.3796,
+      "step": 1009
+    },
+    {
+      "epoch": 0.48264739262887524,
+      "grad_norm": 0.586042042111701,
+      "learning_rate": 1.9337533312053002e-05,
+      "loss": 0.4176,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4831252613344484,
+      "grad_norm": 0.6269719997423575,
+      "learning_rate": 1.9335646970968214e-05,
+      "loss": 0.3991,
+      "step": 1011
+    },
+    {
+      "epoch": 0.4836031300400215,
+      "grad_norm": 0.55742694480261,
+      "learning_rate": 1.933375804033963e-05,
+      "loss": 0.3899,
+      "step": 1012
+    },
+    {
+      "epoch": 0.4840809987455946,
+      "grad_norm": 0.6669718179555711,
+      "learning_rate": 1.9331866520691214e-05,
+      "loss": 0.397,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4845588674511678,
+      "grad_norm": 0.6257739428321304,
+      "learning_rate": 1.9329972412547637e-05,
+      "loss": 0.3954,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4850367361567409,
+      "grad_norm": 0.6256701257508585,
+      "learning_rate": 1.9328075716434287e-05,
+      "loss": 0.3933,
+      "step": 1015
+    },
+    {
+      "epoch": 0.48551460486231407,
+      "grad_norm": 0.6720722566805102,
+      "learning_rate": 1.932617643287728e-05,
+      "loss": 0.3918,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4859924735678872,
+      "grad_norm": 0.6950543974048602,
+      "learning_rate": 1.9324274562403446e-05,
+      "loss": 0.4021,
+      "step": 1017
+    },
+    {
+      "epoch": 0.48647034227346037,
+      "grad_norm": 0.6220688444979767,
+      "learning_rate": 1.9322370105540317e-05,
+      "loss": 0.4057,
+      "step": 1018
+    },
+    {
+      "epoch": 0.4869482109790335,
+      "grad_norm": 0.6165966442005212,
+      "learning_rate": 1.932046306281617e-05,
+      "loss": 0.3905,
+      "step": 1019
+    },
+    {
+      "epoch": 0.48742607968460666,
+      "grad_norm": 1.0211125046724039,
+      "learning_rate": 1.931855343475998e-05,
+      "loss": 0.4144,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4879039483901798,
+      "grad_norm": 0.5733822946288798,
+      "learning_rate": 1.931664122190144e-05,
+      "loss": 0.3917,
+      "step": 1021
+    },
+    {
+      "epoch": 0.48838181709575296,
+      "grad_norm": 0.6130027608133938,
+      "learning_rate": 1.931472642477097e-05,
+      "loss": 0.413,
+      "step": 1022
+    },
+    {
+      "epoch": 0.4888596858013261,
+      "grad_norm": 0.6073794838231226,
+      "learning_rate": 1.93128090438997e-05,
+      "loss": 0.3926,
+      "step": 1023
+    },
+    {
+      "epoch": 0.48933755450689925,
+      "grad_norm": 0.6699051309493238,
+      "learning_rate": 1.9310889079819474e-05,
+      "loss": 0.3945,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4898154232124724,
+      "grad_norm": 0.6983896244728203,
+      "learning_rate": 1.930896653306286e-05,
+      "loss": 0.4017,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4902932919180455,
+      "grad_norm": 0.5963071167388689,
+      "learning_rate": 1.9307041404163135e-05,
+      "loss": 0.4059,
+      "step": 1026
+    },
+    {
+      "epoch": 0.49077116062361864,
+      "grad_norm": 0.6671617092464391,
+      "learning_rate": 1.93051136936543e-05,
+      "loss": 0.3939,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4912490293291918,
+      "grad_norm": 0.6222549383683285,
+      "learning_rate": 1.9303183402071065e-05,
+      "loss": 0.4081,
+      "step": 1028
+    },
+    {
+      "epoch": 0.49172689803476494,
+      "grad_norm": 0.6266179396807892,
+      "learning_rate": 1.930125052994886e-05,
+      "loss": 0.4013,
+      "step": 1029
+    },
+    {
+      "epoch": 0.4922047667403381,
+      "grad_norm": 0.5868185951865348,
+      "learning_rate": 1.929931507782383e-05,
+      "loss": 0.3842,
+      "step": 1030
+    },
+    {
+      "epoch": 0.49268263544591123,
+      "grad_norm": 0.5854933919657626,
+      "learning_rate": 1.9297377046232833e-05,
+      "loss": 0.3961,
+      "step": 1031
+    },
+    {
+      "epoch": 0.4931605041514844,
+      "grad_norm": 0.5729556005202394,
+      "learning_rate": 1.9295436435713447e-05,
+      "loss": 0.3947,
+      "step": 1032
+    },
+    {
+      "epoch": 0.49363837285705753,
+      "grad_norm": 0.6238098390423348,
+      "learning_rate": 1.9293493246803962e-05,
+      "loss": 0.3816,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4941162415626307,
+      "grad_norm": 0.5629873685546706,
+      "learning_rate": 1.9291547480043385e-05,
+      "loss": 0.4014,
+      "step": 1034
+    },
+    {
+      "epoch": 0.4945941102682038,
+      "grad_norm": 0.5648249030788314,
+      "learning_rate": 1.9289599135971437e-05,
+      "loss": 0.4241,
+      "step": 1035
+    },
+    {
+      "epoch": 0.49507197897377697,
+      "grad_norm": 0.6247491915063801,
+      "learning_rate": 1.9287648215128553e-05,
+      "loss": 0.4027,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4955498476793501,
+      "grad_norm": 0.5182781354785916,
+      "learning_rate": 1.928569471805589e-05,
+      "loss": 0.3816,
+      "step": 1037
+    },
+    {
+      "epoch": 0.49602771638492327,
+      "grad_norm": 0.5842774907201587,
+      "learning_rate": 1.9283738645295304e-05,
+      "loss": 0.405,
+      "step": 1038
+    },
+    {
+      "epoch": 0.49650558509049636,
+      "grad_norm": 0.7006381584927244,
+      "learning_rate": 1.928177999738938e-05,
+      "loss": 0.3895,
+      "step": 1039
+    },
+    {
+      "epoch": 0.4969834537960695,
+      "grad_norm": 0.6100647649213121,
+      "learning_rate": 1.9279818774881418e-05,
+      "loss": 0.388,
+      "step": 1040
+    },
+    {
+      "epoch": 0.49746132250164266,
+      "grad_norm": 0.5485271123253042,
+      "learning_rate": 1.9277854978315415e-05,
+      "loss": 0.3972,
+      "step": 1041
+    },
+    {
+      "epoch": 0.4979391912072158,
+      "grad_norm": 0.5710883012225493,
+      "learning_rate": 1.9275888608236104e-05,
+      "loss": 0.4059,
+      "step": 1042
+    },
+    {
+      "epoch": 0.49841705991278895,
+      "grad_norm": 0.5630532049128263,
+      "learning_rate": 1.9273919665188913e-05,
+      "loss": 0.3959,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4988949286183621,
+      "grad_norm": 0.5595415721679393,
+      "learning_rate": 1.9271948149719998e-05,
+      "loss": 0.4058,
+      "step": 1044
+    },
+    {
+      "epoch": 0.49937279732393525,
+      "grad_norm": 0.5890600892979772,
+      "learning_rate": 1.9269974062376224e-05,
+      "loss": 0.3893,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4998506660295084,
+      "grad_norm": 0.5984090472815659,
+      "learning_rate": 1.926799740370516e-05,
+      "loss": 0.3982,
+      "step": 1046
+    },
+    {
+      "epoch": 0.5003285347350815,
+      "grad_norm": 0.6502431881696009,
+      "learning_rate": 1.92660181742551e-05,
+      "loss": 0.3988,
+      "step": 1047
+    },
+    {
+      "epoch": 0.5008064034406546,
+      "grad_norm": 0.5794688514925903,
+      "learning_rate": 1.926403637457505e-05,
+      "loss": 0.3978,
+      "step": 1048
+    },
+    {
+      "epoch": 0.5012842721462278,
+      "grad_norm": 0.5486583842322683,
+      "learning_rate": 1.926205200521473e-05,
+      "loss": 0.405,
+      "step": 1049
+    },
+    {
+      "epoch": 0.5017621408518009,
+      "grad_norm": 0.5964484082934959,
+      "learning_rate": 1.926006506672456e-05,
+      "loss": 0.4075,
+      "step": 1050
+    },
+    {
+      "epoch": 0.5022400095573741,
+      "grad_norm": 0.6015432115221289,
+      "learning_rate": 1.925807555965568e-05,
+      "loss": 0.4011,
+      "step": 1051
+    },
+    {
+      "epoch": 0.5027178782629472,
+      "grad_norm": 0.594525307065842,
+      "learning_rate": 1.9256083484559953e-05,
+      "loss": 0.3979,
+      "step": 1052
+    },
+    {
+      "epoch": 0.5031957469685204,
+      "grad_norm": 0.6660777792568852,
+      "learning_rate": 1.9254088841989943e-05,
+      "loss": 0.3944,
+      "step": 1053
+    },
+    {
+      "epoch": 0.5036736156740935,
+      "grad_norm": 0.589208348377751,
+      "learning_rate": 1.9252091632498926e-05,
+      "loss": 0.3863,
+      "step": 1054
+    },
+    {
+      "epoch": 0.5041514843796667,
+      "grad_norm": 0.635977217231809,
+      "learning_rate": 1.9250091856640895e-05,
+      "loss": 0.4113,
+      "step": 1055
+    },
+    {
+      "epoch": 0.5046293530852398,
+      "grad_norm": 0.6828960314162973,
+      "learning_rate": 1.9248089514970553e-05,
+      "loss": 0.3817,
+      "step": 1056
+    },
+    {
+      "epoch": 0.505107221790813,
+      "grad_norm": 0.5707103289605037,
+      "learning_rate": 1.924608460804331e-05,
+      "loss": 0.3892,
+      "step": 1057
+    },
+    {
+      "epoch": 0.5055850904963861,
+      "grad_norm": 0.5594636089450963,
+      "learning_rate": 1.9244077136415298e-05,
+      "loss": 0.3878,
+      "step": 1058
+    },
+    {
+      "epoch": 0.5060629592019593,
+      "grad_norm": 0.566732821838709,
+      "learning_rate": 1.924206710064335e-05,
+      "loss": 0.4012,
+      "step": 1059
+    },
+    {
+      "epoch": 0.5065408279075324,
+      "grad_norm": 0.5934091175624526,
+      "learning_rate": 1.9240054501285015e-05,
+      "loss": 0.3814,
+      "step": 1060
+    },
+    {
+      "epoch": 0.5070186966131055,
+      "grad_norm": 0.5563172693307783,
+      "learning_rate": 1.9238039338898554e-05,
+      "loss": 0.4011,
+      "step": 1061
+    },
+    {
+      "epoch": 0.5074965653186787,
+      "grad_norm": 0.5998159022961965,
+      "learning_rate": 1.9236021614042936e-05,
+      "loss": 0.3854,
+      "step": 1062
+    },
+    {
+      "epoch": 0.5079744340242518,
+      "grad_norm": 0.5778247983953256,
+      "learning_rate": 1.9234001327277842e-05,
+      "loss": 0.3855,
+      "step": 1063
+    },
+    {
+      "epoch": 0.508452302729825,
+      "grad_norm": 0.5761650392879268,
+      "learning_rate": 1.9231978479163666e-05,
+      "loss": 0.4016,
+      "step": 1064
+    },
+    {
+      "epoch": 0.5089301714353981,
+      "grad_norm": 0.5599768828396271,
+      "learning_rate": 1.922995307026151e-05,
+      "loss": 0.3971,
+      "step": 1065
+    },
+    {
+      "epoch": 0.5094080401409713,
+      "grad_norm": 0.8720261103107927,
+      "learning_rate": 1.9227925101133184e-05,
+      "loss": 0.4097,
+      "step": 1066
+    },
+    {
+      "epoch": 0.5098859088465444,
+      "grad_norm": 0.5924340309412149,
+      "learning_rate": 1.922589457234121e-05,
+      "loss": 0.3871,
+      "step": 1067
+    },
+    {
+      "epoch": 0.5103637775521176,
+      "grad_norm": 0.6004985435021857,
+      "learning_rate": 1.9223861484448833e-05,
+      "loss": 0.3951,
+      "step": 1068
+    },
+    {
+      "epoch": 0.5108416462576907,
+      "grad_norm": 0.6146759569604939,
+      "learning_rate": 1.922182583801998e-05,
+      "loss": 0.3922,
+      "step": 1069
+    },
+    {
+      "epoch": 0.5113195149632639,
+      "grad_norm": 0.5821995968641629,
+      "learning_rate": 1.921978763361931e-05,
+      "loss": 0.4043,
+      "step": 1070
+    },
+    {
+      "epoch": 0.511797383668837,
+      "grad_norm": 0.6337816029347101,
+      "learning_rate": 1.9217746871812184e-05,
+      "loss": 0.4012,
+      "step": 1071
+    },
+    {
+      "epoch": 0.5122752523744102,
+      "grad_norm": 0.5637983730800908,
+      "learning_rate": 1.9215703553164676e-05,
+      "loss": 0.4098,
+      "step": 1072
+    },
+    {
+      "epoch": 0.5127531210799833,
+      "grad_norm": 0.6159837626313385,
+      "learning_rate": 1.921365767824356e-05,
+      "loss": 0.402,
+      "step": 1073
+    },
+    {
+      "epoch": 0.5132309897855564,
+      "grad_norm": 0.6617255676511058,
+      "learning_rate": 1.9211609247616335e-05,
+      "loss": 0.4129,
+      "step": 1074
+    },
+    {
+      "epoch": 0.5137088584911296,
+      "grad_norm": 0.5934568209062026,
+      "learning_rate": 1.9209558261851194e-05,
+      "loss": 0.3833,
+      "step": 1075
+    },
+    {
+      "epoch": 0.5141867271967027,
+      "grad_norm": 0.6221409035389256,
+      "learning_rate": 1.9207504721517047e-05,
+      "loss": 0.3989,
+      "step": 1076
+    },
+    {
+      "epoch": 0.5146645959022759,
+      "grad_norm": 0.5989406482919697,
+      "learning_rate": 1.92054486271835e-05,
+      "loss": 0.3982,
+      "step": 1077
+    },
+    {
+      "epoch": 0.515142464607849,
+      "grad_norm": 0.6476320490359937,
+      "learning_rate": 1.9203389979420893e-05,
+      "loss": 0.3723,
+      "step": 1078
+    },
+    {
+      "epoch": 0.5156203333134222,
+      "grad_norm": 0.5853567195202979,
+      "learning_rate": 1.9201328778800247e-05,
+      "loss": 0.4077,
+      "step": 1079
+    },
+    {
+      "epoch": 0.5160982020189953,
+      "grad_norm": 0.5900737762201892,
+      "learning_rate": 1.919926502589331e-05,
+      "loss": 0.3866,
+      "step": 1080
+    },
+    {
+      "epoch": 0.5165760707245685,
+      "grad_norm": 0.6051998816584814,
+      "learning_rate": 1.9197198721272527e-05,
+      "loss": 0.4011,
+      "step": 1081
+    },
+    {
+      "epoch": 0.5170539394301416,
+      "grad_norm": 0.5901233973103801,
+      "learning_rate": 1.919512986551105e-05,
+      "loss": 0.3873,
+      "step": 1082
+    },
+    {
+      "epoch": 0.5175318081357148,
+      "grad_norm": 0.5723558837242273,
+      "learning_rate": 1.9193058459182747e-05,
+      "loss": 0.4008,
+      "step": 1083
+    },
+    {
+      "epoch": 0.5180096768412878,
+      "grad_norm": 0.586853172304618,
+      "learning_rate": 1.9190984502862196e-05,
+      "loss": 0.372,
+      "step": 1084
+    },
+    {
+      "epoch": 0.518487545546861,
+      "grad_norm": 0.5970057412271904,
+      "learning_rate": 1.9188907997124666e-05,
+      "loss": 0.4002,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5189654142524341,
+      "grad_norm": 0.5890213702570649,
+      "learning_rate": 1.9186828942546148e-05,
+      "loss": 0.3988,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5194432829580072,
+      "grad_norm": 0.5345329489947781,
+      "learning_rate": 1.9184747339703334e-05,
+      "loss": 0.3914,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5199211516635804,
+      "grad_norm": 0.6791735135926099,
+      "learning_rate": 1.9182663189173625e-05,
+      "loss": 0.3988,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5203990203691535,
+      "grad_norm": 0.5596329309086331,
+      "learning_rate": 1.9180576491535125e-05,
+      "loss": 0.3962,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5208768890747267,
+      "grad_norm": 0.5628392832054393,
+      "learning_rate": 1.9178487247366652e-05,
+      "loss": 0.3747,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5213547577802998,
+      "grad_norm": 0.569936759080529,
+      "learning_rate": 1.9176395457247722e-05,
+      "loss": 0.4074,
+      "step": 1091
+    },
+    {
+      "epoch": 0.521832626485873,
+      "grad_norm": 0.5528738761628226,
+      "learning_rate": 1.917430112175856e-05,
+      "loss": 0.3992,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5223104951914461,
+      "grad_norm": 0.6109126673803167,
+      "learning_rate": 1.9172204241480098e-05,
+      "loss": 0.4046,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5227883638970193,
+      "grad_norm": 0.5513494264708154,
+      "learning_rate": 1.9170104816993973e-05,
+      "loss": 0.388,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5232662326025924,
+      "grad_norm": 0.5523842737345893,
+      "learning_rate": 1.916800284888253e-05,
+      "loss": 0.3811,
+      "step": 1095
+    },
+    {
+      "epoch": 0.5237441013081656,
+      "grad_norm": 0.5443431872996419,
+      "learning_rate": 1.9165898337728818e-05,
+      "loss": 0.4142,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5242219700137387,
+      "grad_norm": 0.5903680509886946,
+      "learning_rate": 1.9163791284116592e-05,
+      "loss": 0.3954,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5246998387193119,
+      "grad_norm": 0.7018103685542808,
+      "learning_rate": 1.916168168863031e-05,
+      "loss": 0.4088,
+      "step": 1098
+    },
+    {
+      "epoch": 0.525177707424885,
+      "grad_norm": 0.5915389193430727,
+      "learning_rate": 1.915956955185514e-05,
+      "loss": 0.4055,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5256555761304581,
+      "grad_norm": 0.6388748207474322,
+      "learning_rate": 1.915745487437694e-05,
+      "loss": 0.4055,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5261334448360313,
+      "grad_norm": 0.5374359808793951,
+      "learning_rate": 1.91553376567823e-05,
+      "loss": 0.4095,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5266113135416044,
+      "grad_norm": 0.6253661283286657,
+      "learning_rate": 1.915321789965849e-05,
+      "loss": 0.3993,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5270891822471776,
+      "grad_norm": 0.6007927062737137,
+      "learning_rate": 1.915109560359349e-05,
+      "loss": 0.3789,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5275670509527507,
+      "grad_norm": 0.5773114426599819,
+      "learning_rate": 1.9148970769176e-05,
+      "loss": 0.4031,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5280449196583239,
+      "grad_norm": 1.5644927863876732,
+      "learning_rate": 1.9146843396995396e-05,
+      "loss": 0.3807,
+      "step": 1105
+    },
+    {
+      "epoch": 0.528522788363897,
+      "grad_norm": 0.7068271199032458,
+      "learning_rate": 1.9144713487641786e-05,
+      "loss": 0.404,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5290006570694702,
+      "grad_norm": 0.5842159117368475,
+      "learning_rate": 1.9142581041705964e-05,
+      "loss": 0.3733,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5294785257750433,
+      "grad_norm": 0.629512416750542,
+      "learning_rate": 1.914044605977943e-05,
+      "loss": 0.4029,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5299563944806165,
+      "grad_norm": 0.5591127726288794,
+      "learning_rate": 1.91383085424544e-05,
+      "loss": 0.3975,
+      "step": 1109
+    },
+    {
+      "epoch": 0.5304342631861896,
+      "grad_norm": 0.5508233170263688,
+      "learning_rate": 1.9136168490323772e-05,
+      "loss": 0.3975,
+      "step": 1110
+    },
+    {
+      "epoch": 0.5309121318917628,
+      "grad_norm": 0.6177805307841441,
+      "learning_rate": 1.9134025903981163e-05,
+      "loss": 0.3835,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5313900005973359,
+      "grad_norm": 0.5833143529773639,
+      "learning_rate": 1.9131880784020893e-05,
+      "loss": 0.4139,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5318678693029091,
+      "grad_norm": 0.5967595237608566,
+      "learning_rate": 1.9129733131037977e-05,
+      "loss": 0.4106,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5323457380084822,
+      "grad_norm": 0.6172221950353266,
+      "learning_rate": 1.9127582945628135e-05,
+      "loss": 0.3902,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5328236067140553,
+      "grad_norm": 0.5682174098963437,
+      "learning_rate": 1.9125430228387794e-05,
+      "loss": 0.3874,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5333014754196285,
+      "grad_norm": 0.6193518812671628,
+      "learning_rate": 1.9123274979914076e-05,
+      "loss": 0.4007,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5337793441252016,
+      "grad_norm": 0.604016269024644,
+      "learning_rate": 1.9121117200804812e-05,
+      "loss": 0.3932,
+      "step": 1117
+    },
+    {
+      "epoch": 0.5342572128307748,
+      "grad_norm": 0.6209761683524013,
+      "learning_rate": 1.9118956891658526e-05,
+      "loss": 0.3821,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5347350815363479,
+      "grad_norm": 0.6931704696111478,
+      "learning_rate": 1.9116794053074458e-05,
+      "loss": 0.4071,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5352129502419211,
+      "grad_norm": 0.632080163475699,
+      "learning_rate": 1.9114628685652535e-05,
+      "loss": 0.3839,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5356908189474942,
+      "grad_norm": 0.5748984989211147,
+      "learning_rate": 1.9112460789993394e-05,
+      "loss": 0.3929,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5361686876530674,
+      "grad_norm": 0.5635604927408805,
+      "learning_rate": 1.9110290366698373e-05,
+      "loss": 0.4027,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5366465563586404,
+      "grad_norm": 0.7558151194826072,
+      "learning_rate": 1.9108117416369502e-05,
+      "loss": 0.3986,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5371244250642137,
+      "grad_norm": 0.5980206275007937,
+      "learning_rate": 1.910594193960953e-05,
+      "loss": 0.4006,
+      "step": 1124
+    },
+    {
+      "epoch": 0.5376022937697867,
+      "grad_norm": 0.6780188301974273,
+      "learning_rate": 1.9103763937021887e-05,
+      "loss": 0.3826,
+      "step": 1125
+    },
+    {
+      "epoch": 0.53808016247536,
+      "grad_norm": 0.5790597345181424,
+      "learning_rate": 1.9101583409210714e-05,
+      "loss": 0.3835,
+      "step": 1126
+    },
+    {
+      "epoch": 0.538558031180933,
+      "grad_norm": 0.6487135859049837,
+      "learning_rate": 1.9099400356780857e-05,
+      "loss": 0.3995,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5390358998865061,
+      "grad_norm": 0.5780178648506041,
+      "learning_rate": 1.909721478033785e-05,
+      "loss": 0.3925,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5395137685920793,
+      "grad_norm": 0.642292787021177,
+      "learning_rate": 1.909502668048793e-05,
+      "loss": 0.4016,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5399916372976524,
+      "grad_norm": 0.5939162000913498,
+      "learning_rate": 1.909283605783805e-05,
+      "loss": 0.4057,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5404695060032256,
+      "grad_norm": 0.6390737833949265,
+      "learning_rate": 1.9090642912995836e-05,
+      "loss": 0.3997,
+      "step": 1131
+    },
+    {
+      "epoch": 0.5409473747087987,
+      "grad_norm": 0.5629045570791384,
+      "learning_rate": 1.9088447246569638e-05,
+      "loss": 0.3948,
+      "step": 1132
+    },
+    {
+      "epoch": 0.5414252434143719,
+      "grad_norm": 0.7058159449700776,
+      "learning_rate": 1.908624905916849e-05,
+      "loss": 0.4006,
+      "step": 1133
+    },
+    {
+      "epoch": 0.541903112119945,
+      "grad_norm": 0.5821061957075071,
+      "learning_rate": 1.9084048351402135e-05,
+      "loss": 0.4029,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5423809808255182,
+      "grad_norm": 0.5745676634359547,
+      "learning_rate": 1.9081845123881002e-05,
+      "loss": 0.4067,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5428588495310913,
+      "grad_norm": 0.5799046046456168,
+      "learning_rate": 1.9079639377216236e-05,
+      "loss": 0.3895,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5433367182366645,
+      "grad_norm": 0.5423826236018965,
+      "learning_rate": 1.9077431112019666e-05,
+      "loss": 0.392,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5438145869422376,
+      "grad_norm": 0.5614547514964748,
+      "learning_rate": 1.9075220328903833e-05,
+      "loss": 0.419,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5442924556478108,
+      "grad_norm": 0.7294948459972235,
+      "learning_rate": 1.907300702848196e-05,
+      "loss": 0.3725,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5447703243533839,
+      "grad_norm": 0.7211253855583865,
+      "learning_rate": 1.9070791211367984e-05,
+      "loss": 0.3958,
+      "step": 1140
+    },
+    {
+      "epoch": 0.545248193058957,
+      "grad_norm": 0.7264564481802851,
+      "learning_rate": 1.9068572878176535e-05,
+      "loss": 0.3969,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5457260617645302,
+      "grad_norm": 0.5947660314640929,
+      "learning_rate": 1.9066352029522933e-05,
+      "loss": 0.4076,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5462039304701033,
+      "grad_norm": 0.6646952552501058,
+      "learning_rate": 1.906412866602321e-05,
+      "loss": 0.4096,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5466817991756765,
+      "grad_norm": 0.6440292968735994,
+      "learning_rate": 1.906190278829408e-05,
+      "loss": 0.3947,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5471596678812496,
+      "grad_norm": 0.6531777155361128,
+      "learning_rate": 1.9059674396952963e-05,
+      "loss": 0.3921,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5476375365868228,
+      "grad_norm": 0.6263069312765026,
+      "learning_rate": 1.9057443492617983e-05,
+      "loss": 0.3908,
+      "step": 1146
+    },
+    {
+      "epoch": 0.5481154052923959,
+      "grad_norm": 0.6009448091108315,
+      "learning_rate": 1.9055210075907946e-05,
+      "loss": 0.3872,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5485932739979691,
+      "grad_norm": 0.7151497490060705,
+      "learning_rate": 1.9052974147442365e-05,
+      "loss": 0.4047,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5490711427035422,
+      "grad_norm": 0.6210354172889246,
+      "learning_rate": 1.905073570784145e-05,
+      "loss": 0.3875,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5495490114091154,
+      "grad_norm": 0.6071292868798474,
+      "learning_rate": 1.90484947577261e-05,
+      "loss": 0.3786,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5500268801146885,
+      "grad_norm": 0.5393751375916709,
+      "learning_rate": 1.9046251297717915e-05,
+      "loss": 0.3815,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5505047488202617,
+      "grad_norm": 0.549917579146435,
+      "learning_rate": 1.9044005328439197e-05,
+      "loss": 0.3955,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5509826175258348,
+      "grad_norm": 0.7395520846508221,
+      "learning_rate": 1.9041756850512932e-05,
+      "loss": 0.4056,
+      "step": 1153
+    },
+    {
+      "epoch": 0.551460486231408,
+      "grad_norm": 0.5862288750441804,
+      "learning_rate": 1.9039505864562812e-05,
+      "loss": 0.3972,
+      "step": 1154
+    },
+    {
+      "epoch": 0.5519383549369811,
+      "grad_norm": 0.5831709652721829,
+      "learning_rate": 1.903725237121322e-05,
+      "loss": 0.3919,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5524162236425542,
+      "grad_norm": 0.5916849238602208,
+      "learning_rate": 1.9034996371089233e-05,
+      "loss": 0.3856,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5528940923481274,
+      "grad_norm": 0.6411973294376119,
+      "learning_rate": 1.9032737864816627e-05,
+      "loss": 0.4067,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5533719610537005,
+      "grad_norm": 0.5832416797946984,
+      "learning_rate": 1.9030476853021875e-05,
+      "loss": 0.389,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5538498297592737,
+      "grad_norm": 0.622158136439798,
+      "learning_rate": 1.9028213336332135e-05,
+      "loss": 0.3893,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5543276984648468,
+      "grad_norm": 0.5762766222613097,
+      "learning_rate": 1.902594731537527e-05,
+      "loss": 0.394,
+      "step": 1160
+    },
+    {
+      "epoch": 0.55480556717042,
+      "grad_norm": 0.550291838034785,
+      "learning_rate": 1.9023678790779838e-05,
+      "loss": 0.3888,
+      "step": 1161
+    },
+    {
+      "epoch": 0.555283435875993,
+      "grad_norm": 0.5662654245608815,
+      "learning_rate": 1.9021407763175083e-05,
+      "loss": 0.3926,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5557613045815663,
+      "grad_norm": 0.6334933581251789,
+      "learning_rate": 1.901913423319095e-05,
+      "loss": 0.4011,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5562391732871393,
+      "grad_norm": 0.5575895045734839,
+      "learning_rate": 1.9016858201458075e-05,
+      "loss": 0.3972,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5567170419927125,
+      "grad_norm": 0.5966619103161227,
+      "learning_rate": 1.901457966860779e-05,
+      "loss": 0.3824,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5571949106982856,
+      "grad_norm": 0.5870555991442512,
+      "learning_rate": 1.9012298635272117e-05,
+      "loss": 0.4003,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5576727794038588,
+      "grad_norm": 0.569548018108478,
+      "learning_rate": 1.9010015102083778e-05,
+      "loss": 0.3992,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5581506481094319,
+      "grad_norm": 0.5537323419722724,
+      "learning_rate": 1.9007729069676185e-05,
+      "loss": 0.4126,
+      "step": 1168
+    },
+    {
+      "epoch": 0.558628516815005,
+      "grad_norm": 0.6214624827198821,
+      "learning_rate": 1.9005440538683443e-05,
+      "loss": 0.4034,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5591063855205782,
+      "grad_norm": 0.5794512137484054,
+      "learning_rate": 1.9003149509740347e-05,
+      "loss": 0.4042,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5595842542261513,
+      "grad_norm": 0.6471077379542854,
+      "learning_rate": 1.9000855983482386e-05,
+      "loss": 0.3955,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5600621229317245,
+      "grad_norm": 0.7713823924796754,
+      "learning_rate": 1.8998559960545753e-05,
+      "loss": 0.3894,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5605399916372976,
+      "grad_norm": 0.669563924210895,
+      "learning_rate": 1.8996261441567318e-05,
+      "loss": 0.4061,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5610178603428708,
+      "grad_norm": 0.6358401107367022,
+      "learning_rate": 1.8993960427184647e-05,
+      "loss": 0.4049,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5614957290484439,
+      "grad_norm": 0.632834423434781,
+      "learning_rate": 1.899165691803601e-05,
+      "loss": 0.3974,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5619735977540171,
+      "grad_norm": 0.7647633906590625,
+      "learning_rate": 1.8989350914760348e-05,
+      "loss": 0.3967,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5624514664595902,
+      "grad_norm": 0.6692881107694713,
+      "learning_rate": 1.8987042417997313e-05,
+      "loss": 0.4027,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5629293351651634,
+      "grad_norm": 0.831620122043209,
+      "learning_rate": 1.898473142838724e-05,
+      "loss": 0.4012,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5634072038707365,
+      "grad_norm": 1.1686193693749636,
+      "learning_rate": 1.898241794657116e-05,
+      "loss": 0.405,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5638850725763097,
+      "grad_norm": 0.6351278759311869,
+      "learning_rate": 1.8980101973190787e-05,
+      "loss": 0.3754,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5643629412818828,
+      "grad_norm": 0.5873130279029877,
+      "learning_rate": 1.8977783508888535e-05,
+      "loss": 0.3965,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5648408099874559,
+      "grad_norm": 0.5989794788052755,
+      "learning_rate": 1.89754625543075e-05,
+      "loss": 0.399,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5653186786930291,
+      "grad_norm": 0.5698375975862137,
+      "learning_rate": 1.8973139110091477e-05,
+      "loss": 0.4015,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5657965473986022,
+      "grad_norm": 0.6737645692540567,
+      "learning_rate": 1.897081317688495e-05,
+      "loss": 0.4223,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5662744161041754,
+      "grad_norm": 0.593327748620439,
+      "learning_rate": 1.896848475533309e-05,
+      "loss": 0.4066,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5667522848097485,
+      "grad_norm": 0.5883337985827269,
+      "learning_rate": 1.896615384608176e-05,
+      "loss": 0.3736,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5672301535153217,
+      "grad_norm": 0.9321510339697601,
+      "learning_rate": 1.896382044977751e-05,
+      "loss": 0.4119,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5677080222208948,
+      "grad_norm": 0.5793456890053796,
+      "learning_rate": 1.896148456706759e-05,
+      "loss": 0.3928,
+      "step": 1188
+    },
+    {
+      "epoch": 0.568185890926468,
+      "grad_norm": 0.8624241078357229,
+      "learning_rate": 1.8959146198599928e-05,
+      "loss": 0.4138,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5686637596320411,
+      "grad_norm": 0.6540030367738001,
+      "learning_rate": 1.8956805345023145e-05,
+      "loss": 0.3981,
+      "step": 1190
+    },
+    {
+      "epoch": 0.5691416283376143,
+      "grad_norm": 0.6327764655932094,
+      "learning_rate": 1.8954462006986557e-05,
+      "loss": 0.4082,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5696194970431874,
+      "grad_norm": 0.6018338777252428,
+      "learning_rate": 1.8952116185140164e-05,
+      "loss": 0.3884,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5700973657487606,
+      "grad_norm": 0.5645211426190209,
+      "learning_rate": 1.8949767880134652e-05,
+      "loss": 0.4008,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5705752344543337,
+      "grad_norm": 0.62032674267378,
+      "learning_rate": 1.89474170926214e-05,
+      "loss": 0.387,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5710531031599068,
+      "grad_norm": 0.5331504995268993,
+      "learning_rate": 1.894506382325248e-05,
+      "loss": 0.3944,
+      "step": 1195
+    },
+    {
+      "epoch": 0.57153097186548,
+      "grad_norm": 0.5571879177321089,
+      "learning_rate": 1.8942708072680637e-05,
+      "loss": 0.3873,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5720088405710531,
+      "grad_norm": 0.5434082219860525,
+      "learning_rate": 1.8940349841559325e-05,
+      "loss": 0.3819,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5724867092766263,
+      "grad_norm": 0.5556021619843906,
+      "learning_rate": 1.8937989130542672e-05,
+      "loss": 0.3819,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5729645779821994,
+      "grad_norm": 0.6319621645757422,
+      "learning_rate": 1.8935625940285502e-05,
+      "loss": 0.3921,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5734424466877726,
+      "grad_norm": 0.6282911906906646,
+      "learning_rate": 1.8933260271443313e-05,
+      "loss": 0.4087,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5739203153933456,
+      "grad_norm": 0.5581191438877338,
+      "learning_rate": 1.8930892124672303e-05,
+      "loss": 0.3932,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5743981840989189,
+      "grad_norm": 0.5463800521790595,
+      "learning_rate": 1.892852150062936e-05,
+      "loss": 0.394,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5748760528044919,
+      "grad_norm": 0.5931390296947407,
+      "learning_rate": 1.8926148399972047e-05,
+      "loss": 0.3773,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5753539215100651,
+      "grad_norm": 0.5452845931438769,
+      "learning_rate": 1.8923772823358624e-05,
+      "loss": 0.3901,
+      "step": 1204
+    },
+    {
+      "epoch": 0.5758317902156382,
+      "grad_norm": 0.5840615558705706,
+      "learning_rate": 1.8921394771448032e-05,
+      "loss": 0.4034,
+      "step": 1205
+    },
+    {
+      "epoch": 0.5763096589212114,
+      "grad_norm": 0.6291436076221341,
+      "learning_rate": 1.89190142448999e-05,
+      "loss": 0.3905,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5767875276267845,
+      "grad_norm": 0.5327483631263784,
+      "learning_rate": 1.8916631244374548e-05,
+      "loss": 0.3999,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5772653963323577,
+      "grad_norm": 0.6494525583362151,
+      "learning_rate": 1.891424577053297e-05,
+      "loss": 0.3893,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5777432650379308,
+      "grad_norm": 0.5510447847761317,
+      "learning_rate": 1.8911857824036863e-05,
+      "loss": 0.3835,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5782211337435039,
+      "grad_norm": 0.6227648730898245,
+      "learning_rate": 1.89094674055486e-05,
+      "loss": 0.3999,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5786990024490771,
+      "grad_norm": 0.5541558527671858,
+      "learning_rate": 1.8907074515731236e-05,
+      "loss": 0.3994,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5791768711546502,
+      "grad_norm": 0.5676146792664295,
+      "learning_rate": 1.8904679155248523e-05,
+      "loss": 0.3874,
+      "step": 1212
+    },
+    {
+      "epoch": 0.5796547398602234,
+      "grad_norm": 0.7499627866819074,
+      "learning_rate": 1.890228132476488e-05,
+      "loss": 0.3683,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5801326085657965,
+      "grad_norm": 0.5526639952242319,
+      "learning_rate": 1.889988102494544e-05,
+      "loss": 0.3977,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5806104772713697,
+      "grad_norm": 0.6492832611355143,
+      "learning_rate": 1.889747825645599e-05,
+      "loss": 0.3861,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5810883459769428,
+      "grad_norm": 0.5411532191238587,
+      "learning_rate": 1.8895073019963022e-05,
+      "loss": 0.3934,
+      "step": 1216
+    },
+    {
+      "epoch": 0.581566214682516,
+      "grad_norm": 0.6698750283375203,
+      "learning_rate": 1.8892665316133706e-05,
+      "loss": 0.4089,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5820440833880891,
+      "grad_norm": 0.604672289766661,
+      "learning_rate": 1.8890255145635895e-05,
+      "loss": 0.3985,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5825219520936623,
+      "grad_norm": 0.5809198566541879,
+      "learning_rate": 1.888784250913813e-05,
+      "loss": 0.4029,
+      "step": 1219
+    },
+    {
+      "epoch": 0.5829998207992354,
+      "grad_norm": 0.5575695641227716,
+      "learning_rate": 1.8885427407309627e-05,
+      "loss": 0.3918,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5834776895048086,
+      "grad_norm": 0.6012108926363768,
+      "learning_rate": 1.88830098408203e-05,
+      "loss": 0.3864,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5839555582103817,
+      "grad_norm": 0.5621575791142275,
+      "learning_rate": 1.8880589810340734e-05,
+      "loss": 0.38,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5844334269159548,
+      "grad_norm": 0.5831322537081107,
+      "learning_rate": 1.8878167316542207e-05,
+      "loss": 0.3894,
+      "step": 1223
+    },
+    {
+      "epoch": 0.584911295621528,
+      "grad_norm": 0.5871042303485483,
+      "learning_rate": 1.8875742360096675e-05,
+      "loss": 0.3896,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5853891643271011,
+      "grad_norm": 0.5564242439821103,
+      "learning_rate": 1.887331494167678e-05,
+      "loss": 0.3933,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5858670330326743,
+      "grad_norm": 0.5819982187143664,
+      "learning_rate": 1.887088506195584e-05,
+      "loss": 0.4114,
+      "step": 1226
+    },
+    {
+      "epoch": 0.5863449017382474,
+      "grad_norm": 0.5090058539216136,
+      "learning_rate": 1.8868452721607865e-05,
+      "loss": 0.3962,
+      "step": 1227
+    },
+    {
+      "epoch": 0.5868227704438206,
+      "grad_norm": 0.5875412543113175,
+      "learning_rate": 1.8866017921307544e-05,
+      "loss": 0.4056,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5873006391493937,
+      "grad_norm": 0.6166139308626585,
+      "learning_rate": 1.886358066173024e-05,
+      "loss": 0.3882,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5877785078549669,
+      "grad_norm": 0.5888127004124792,
+      "learning_rate": 1.8861140943552014e-05,
+      "loss": 0.3912,
+      "step": 1230
+    },
+    {
+      "epoch": 0.58825637656054,
+      "grad_norm": 0.6162981466724767,
+      "learning_rate": 1.8858698767449598e-05,
+      "loss": 0.3937,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5887342452661132,
+      "grad_norm": 0.6715967270841622,
+      "learning_rate": 1.8856254134100408e-05,
+      "loss": 0.401,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5892121139716863,
+      "grad_norm": 0.5366020380685378,
+      "learning_rate": 1.8853807044182544e-05,
+      "loss": 0.3875,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5896899826772595,
+      "grad_norm": 0.591930559738908,
+      "learning_rate": 1.8851357498374785e-05,
+      "loss": 0.4082,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5901678513828326,
+      "grad_norm": 0.5911447561891597,
+      "learning_rate": 1.884890549735659e-05,
+      "loss": 0.3904,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5906457200884057,
+      "grad_norm": 0.5226862065805489,
+      "learning_rate": 1.8846451041808102e-05,
+      "loss": 0.3937,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5911235887939789,
+      "grad_norm": 0.6596621571985957,
+      "learning_rate": 1.8843994132410143e-05,
+      "loss": 0.403,
+      "step": 1237
+    },
+    {
+      "epoch": 0.591601457499552,
+      "grad_norm": 0.5701923237556955,
+      "learning_rate": 1.884153476984422e-05,
+      "loss": 0.4008,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5920793262051252,
+      "grad_norm": 0.5930242333237962,
+      "learning_rate": 1.883907295479251e-05,
+      "loss": 0.3795,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5925571949106982,
+      "grad_norm": 0.7314294978915366,
+      "learning_rate": 1.8836608687937883e-05,
+      "loss": 0.3928,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5930350636162715,
+      "grad_norm": 0.5401549968229234,
+      "learning_rate": 1.883414196996388e-05,
+      "loss": 0.3828,
+      "step": 1241
+    },
+    {
+      "epoch": 0.5935129323218445,
+      "grad_norm": 0.5649220265715015,
+      "learning_rate": 1.8831672801554726e-05,
+      "loss": 0.4082,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5939908010274177,
+      "grad_norm": 0.5284937507489004,
+      "learning_rate": 1.882920118339533e-05,
+      "loss": 0.3894,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5944686697329908,
+      "grad_norm": 0.5645488013407426,
+      "learning_rate": 1.8826727116171264e-05,
+      "loss": 0.3931,
+      "step": 1244
+    },
+    {
+      "epoch": 0.594946538438564,
+      "grad_norm": 0.5376862100448556,
+      "learning_rate": 1.8824250600568798e-05,
+      "loss": 0.3859,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5954244071441371,
+      "grad_norm": 0.601826503242195,
+      "learning_rate": 1.8821771637274874e-05,
+      "loss": 0.3896,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5959022758497103,
+      "grad_norm": 0.547959448752086,
+      "learning_rate": 1.881929022697711e-05,
+      "loss": 0.3716,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5963801445552834,
+      "grad_norm": 0.5519251178713137,
+      "learning_rate": 1.881680637036381e-05,
+      "loss": 0.4019,
+      "step": 1248
+    },
+    {
+      "epoch": 0.5968580132608566,
+      "grad_norm": 0.5714630327545864,
+      "learning_rate": 1.8814320068123945e-05,
+      "loss": 0.3937,
+      "step": 1249
+    },
+    {
+      "epoch": 0.5973358819664297,
+      "grad_norm": 0.5470336630560245,
+      "learning_rate": 1.8811831320947177e-05,
+      "loss": 0.4043,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5978137506720028,
+      "grad_norm": 0.5569332027496797,
+      "learning_rate": 1.8809340129523835e-05,
+      "loss": 0.3718,
+      "step": 1251
+    },
+    {
+      "epoch": 0.598291619377576,
+      "grad_norm": 0.5598645339474388,
+      "learning_rate": 1.880684649454494e-05,
+      "loss": 0.4028,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5987694880831491,
+      "grad_norm": 0.5364480060048303,
+      "learning_rate": 1.8804350416702174e-05,
+      "loss": 0.3782,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5992473567887223,
+      "grad_norm": 0.5873593424397728,
+      "learning_rate": 1.880185189668791e-05,
+      "loss": 0.392,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5997252254942954,
+      "grad_norm": 0.5235640424479179,
+      "learning_rate": 1.879935093519519e-05,
+      "loss": 0.3805,
+      "step": 1255
+    },
+    {
+      "epoch": 0.6002030941998686,
+      "grad_norm": 0.620909525268386,
+      "learning_rate": 1.8796847532917743e-05,
+      "loss": 0.4068,
+      "step": 1256
+    },
+    {
+      "epoch": 0.6006809629054417,
+      "grad_norm": 0.5489235045804418,
+      "learning_rate": 1.879434169054996e-05,
+      "loss": 0.3848,
+      "step": 1257
+    },
+    {
+      "epoch": 0.6011588316110149,
+      "grad_norm": 0.5347491918019722,
+      "learning_rate": 1.8791833408786922e-05,
+      "loss": 0.3819,
+      "step": 1258
+    },
+    {
+      "epoch": 0.601636700316588,
+      "grad_norm": 0.5915201608837969,
+      "learning_rate": 1.878932268832438e-05,
+      "loss": 0.3951,
+      "step": 1259
+    },
+    {
+      "epoch": 0.6021145690221612,
+      "grad_norm": 0.5356326976446867,
+      "learning_rate": 1.878680952985877e-05,
+      "loss": 0.3905,
+      "step": 1260
+    },
+    {
+      "epoch": 0.6025924377277343,
+      "grad_norm": 0.6181767591406878,
+      "learning_rate": 1.878429393408719e-05,
+      "loss": 0.4024,
+      "step": 1261
+    },
+    {
+      "epoch": 0.6030703064333075,
+      "grad_norm": 0.5632469969713791,
+      "learning_rate": 1.8781775901707425e-05,
+      "loss": 0.3937,
+      "step": 1262
+    },
+    {
+      "epoch": 0.6035481751388806,
+      "grad_norm": 0.5877175455362745,
+      "learning_rate": 1.8779255433417935e-05,
+      "loss": 0.3943,
+      "step": 1263
+    },
+    {
+      "epoch": 0.6040260438444537,
+      "grad_norm": 0.5503903372236624,
+      "learning_rate": 1.8776732529917846e-05,
+      "loss": 0.3984,
+      "step": 1264
+    },
+    {
+      "epoch": 0.6045039125500269,
+      "grad_norm": 0.5649581602924505,
+      "learning_rate": 1.8774207191906976e-05,
+      "loss": 0.4017,
+      "step": 1265
+    },
+    {
+      "epoch": 0.6049817812556,
+      "grad_norm": 0.563853155134363,
+      "learning_rate": 1.8771679420085805e-05,
+      "loss": 0.3891,
+      "step": 1266
+    },
+    {
+      "epoch": 0.6054596499611732,
+      "grad_norm": 0.5651839005098936,
+      "learning_rate": 1.8769149215155497e-05,
+      "loss": 0.3913,
+      "step": 1267
+    },
+    {
+      "epoch": 0.6059375186667463,
+      "grad_norm": 0.5963210739445259,
+      "learning_rate": 1.8766616577817875e-05,
+      "loss": 0.4133,
+      "step": 1268
+    },
+    {
+      "epoch": 0.6064153873723195,
+      "grad_norm": 0.55539068455758,
+      "learning_rate": 1.876408150877546e-05,
+      "loss": 0.3797,
+      "step": 1269
+    },
+    {
+      "epoch": 0.6068932560778926,
+      "grad_norm": 0.5856042654690761,
+      "learning_rate": 1.8761544008731426e-05,
+      "loss": 0.4058,
+      "step": 1270
+    },
+    {
+      "epoch": 0.6073711247834658,
+      "grad_norm": 0.542197554269779,
+      "learning_rate": 1.8759004078389635e-05,
+      "loss": 0.39,
+      "step": 1271
+    },
+    {
+      "epoch": 0.6078489934890389,
+      "grad_norm": 0.5353927253355668,
+      "learning_rate": 1.8756461718454622e-05,
+      "loss": 0.4147,
+      "step": 1272
+    },
+    {
+      "epoch": 0.6083268621946121,
+      "grad_norm": 0.7529582857438246,
+      "learning_rate": 1.8753916929631586e-05,
+      "loss": 0.4046,
+      "step": 1273
+    },
+    {
+      "epoch": 0.6088047309001852,
+      "grad_norm": 0.5958352414182106,
+      "learning_rate": 1.8751369712626413e-05,
+      "loss": 0.385,
+      "step": 1274
+    },
+    {
+      "epoch": 0.6092825996057584,
+      "grad_norm": 0.5841015694045317,
+      "learning_rate": 1.874882006814565e-05,
+      "loss": 0.3854,
+      "step": 1275
+    },
+    {
+      "epoch": 0.6097604683113315,
+      "grad_norm": 0.6682019912149343,
+      "learning_rate": 1.874626799689653e-05,
+      "loss": 0.3775,
+      "step": 1276
+    },
+    {
+      "epoch": 0.6102383370169046,
+      "grad_norm": 0.5543315924200671,
+      "learning_rate": 1.874371349958695e-05,
+      "loss": 0.3767,
+      "step": 1277
+    },
+    {
+      "epoch": 0.6107162057224778,
+      "grad_norm": 0.5842970394229539,
+      "learning_rate": 1.874115657692548e-05,
+      "loss": 0.3992,
+      "step": 1278
+    },
+    {
+      "epoch": 0.6111940744280508,
+      "grad_norm": 0.5711349862291916,
+      "learning_rate": 1.8738597229621368e-05,
+      "loss": 0.3913,
+      "step": 1279
+    },
+    {
+      "epoch": 0.611671943133624,
+      "grad_norm": 0.6023071222272888,
+      "learning_rate": 1.8736035458384528e-05,
+      "loss": 0.3896,
+      "step": 1280
+    },
+    {
+      "epoch": 0.6121498118391971,
+      "grad_norm": 0.5777611740504066,
+      "learning_rate": 1.8733471263925553e-05,
+      "loss": 0.4059,
+      "step": 1281
+    },
+    {
+      "epoch": 0.6126276805447703,
+      "grad_norm": 0.5729042128264256,
+      "learning_rate": 1.8730904646955706e-05,
+      "loss": 0.4117,
+      "step": 1282
+    },
+    {
+      "epoch": 0.6131055492503434,
+      "grad_norm": 0.5307580555984255,
+      "learning_rate": 1.8728335608186923e-05,
+      "loss": 0.3967,
+      "step": 1283
+    },
+    {
+      "epoch": 0.6135834179559166,
+      "grad_norm": 0.5972005721716636,
+      "learning_rate": 1.8725764148331804e-05,
+      "loss": 0.3851,
+      "step": 1284
+    },
+    {
+      "epoch": 0.6140612866614897,
+      "grad_norm": 0.5486242431937549,
+      "learning_rate": 1.8723190268103634e-05,
+      "loss": 0.3921,
+      "step": 1285
+    },
+    {
+      "epoch": 0.6145391553670629,
+      "grad_norm": 0.5129417046829837,
+      "learning_rate": 1.8720613968216356e-05,
+      "loss": 0.3799,
+      "step": 1286
+    },
+    {
+      "epoch": 0.615017024072636,
+      "grad_norm": 0.5389718288002542,
+      "learning_rate": 1.871803524938459e-05,
+      "loss": 0.3784,
+      "step": 1287
+    },
+    {
+      "epoch": 0.6154948927782092,
+      "grad_norm": 0.5413932174652784,
+      "learning_rate": 1.871545411232363e-05,
+      "loss": 0.3883,
+      "step": 1288
+    },
+    {
+      "epoch": 0.6159727614837823,
+      "grad_norm": 0.5576540640582818,
+      "learning_rate": 1.871287055774944e-05,
+      "loss": 0.3725,
+      "step": 1289
+    },
+    {
+      "epoch": 0.6164506301893554,
+      "grad_norm": 0.5438843907191524,
+      "learning_rate": 1.8710284586378645e-05,
+      "loss": 0.3853,
+      "step": 1290
+    },
+    {
+      "epoch": 0.6169284988949286,
+      "grad_norm": 0.5921910838665108,
+      "learning_rate": 1.870769619892856e-05,
+      "loss": 0.3881,
+      "step": 1291
+    },
+    {
+      "epoch": 0.6174063676005017,
+      "grad_norm": 0.5281942703357103,
+      "learning_rate": 1.8705105396117145e-05,
+      "loss": 0.4026,
+      "step": 1292
+    },
+    {
+      "epoch": 0.6178842363060749,
+      "grad_norm": 0.6001108596932736,
+      "learning_rate": 1.870251217866305e-05,
+      "loss": 0.3875,
+      "step": 1293
+    },
+    {
+      "epoch": 0.618362105011648,
+      "grad_norm": 0.610945572306042,
+      "learning_rate": 1.8699916547285583e-05,
+      "loss": 0.3812,
+      "step": 1294
+    },
+    {
+      "epoch": 0.6188399737172212,
+      "grad_norm": 0.5695158823092838,
+      "learning_rate": 1.8697318502704734e-05,
+      "loss": 0.371,
+      "step": 1295
+    },
+    {
+      "epoch": 0.6193178424227943,
+      "grad_norm": 0.569180451522841,
+      "learning_rate": 1.869471804564115e-05,
+      "loss": 0.4033,
+      "step": 1296
+    },
+    {
+      "epoch": 0.6197957111283675,
+      "grad_norm": 0.5328007824386596,
+      "learning_rate": 1.869211517681615e-05,
+      "loss": 0.4016,
+      "step": 1297
+    },
+    {
+      "epoch": 0.6202735798339406,
+      "grad_norm": 0.5787431750554224,
+      "learning_rate": 1.868950989695173e-05,
+      "loss": 0.3875,
+      "step": 1298
+    },
+    {
+      "epoch": 0.6207514485395138,
+      "grad_norm": 0.7099261830662204,
+      "learning_rate": 1.8686902206770542e-05,
+      "loss": 0.3925,
+      "step": 1299
+    },
+    {
+      "epoch": 0.6212293172450869,
+      "grad_norm": 0.5775144630618253,
+      "learning_rate": 1.8684292106995916e-05,
+      "loss": 0.3876,
+      "step": 1300
+    },
+    {
+      "epoch": 0.6217071859506601,
+      "grad_norm": 0.5960304043808599,
+      "learning_rate": 1.868167959835185e-05,
+      "loss": 0.3966,
+      "step": 1301
+    },
+    {
+      "epoch": 0.6221850546562332,
+      "grad_norm": 0.6193706361120256,
+      "learning_rate": 1.8679064681563005e-05,
+      "loss": 0.4009,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6226629233618064,
+      "grad_norm": 0.6051679338180642,
+      "learning_rate": 1.867644735735471e-05,
+      "loss": 0.385,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6231407920673795,
+      "grad_norm": 0.6745787304908575,
+      "learning_rate": 1.8673827626452972e-05,
+      "loss": 0.4009,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6236186607729526,
+      "grad_norm": 0.591651344835618,
+      "learning_rate": 1.8671205489584453e-05,
+      "loss": 0.3935,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6240965294785258,
+      "grad_norm": 0.5499144339680516,
+      "learning_rate": 1.8668580947476487e-05,
+      "loss": 0.3803,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6245743981840989,
+      "grad_norm": 0.5742789490169716,
+      "learning_rate": 1.8665954000857077e-05,
+      "loss": 0.3976,
+      "step": 1307
+    },
+    {
+      "epoch": 0.6250522668896721,
+      "grad_norm": 0.5342157539830956,
+      "learning_rate": 1.8663324650454896e-05,
+      "loss": 0.3966,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6255301355952452,
+      "grad_norm": 0.557937880620906,
+      "learning_rate": 1.8660692896999272e-05,
+      "loss": 0.3891,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6260080043008184,
+      "grad_norm": 0.6049761505152792,
+      "learning_rate": 1.865805874122021e-05,
+      "loss": 0.3903,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6264858730063915,
+      "grad_norm": 0.5638630151330256,
+      "learning_rate": 1.865542218384838e-05,
+      "loss": 0.3987,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6269637417119647,
+      "grad_norm": 0.5558339859479475,
+      "learning_rate": 1.865278322561512e-05,
+      "loss": 0.3838,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6274416104175378,
+      "grad_norm": 0.7599896980096237,
+      "learning_rate": 1.8650141867252418e-05,
+      "loss": 0.3783,
+      "step": 1313
+    },
+    {
+      "epoch": 0.627919479123111,
+      "grad_norm": 0.5377592981486465,
+      "learning_rate": 1.8647498109492952e-05,
+      "loss": 0.3748,
+      "step": 1314
+    },
+    {
+      "epoch": 0.6283973478286841,
+      "grad_norm": 0.5747378065579319,
+      "learning_rate": 1.8644851953070045e-05,
+      "loss": 0.3797,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6288752165342573,
+      "grad_norm": 0.5611211100705439,
+      "learning_rate": 1.8642203398717704e-05,
+      "loss": 0.374,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6293530852398304,
+      "grad_norm": 0.5398727147760198,
+      "learning_rate": 1.8639552447170586e-05,
+      "loss": 0.4111,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6298309539454034,
+      "grad_norm": 0.6478100942138757,
+      "learning_rate": 1.8636899099164016e-05,
+      "loss": 0.373,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6303088226509767,
+      "grad_norm": 0.5239375476824424,
+      "learning_rate": 1.863424335543399e-05,
+      "loss": 0.3934,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6307866913565497,
+      "grad_norm": 0.5594162158750057,
+      "learning_rate": 1.863158521671716e-05,
+      "loss": 0.3914,
+      "step": 1320
+    },
+    {
+      "epoch": 0.631264560062123,
+      "grad_norm": 0.5605106637954282,
+      "learning_rate": 1.862892468375085e-05,
+      "loss": 0.3907,
+      "step": 1321
+    },
+    {
+      "epoch": 0.631742428767696,
+      "grad_norm": 0.5566958695945204,
+      "learning_rate": 1.8626261757273047e-05,
+      "loss": 0.3853,
+      "step": 1322
+    },
+    {
+      "epoch": 0.6322202974732692,
+      "grad_norm": 0.5930367871951245,
+      "learning_rate": 1.8623596438022395e-05,
+      "loss": 0.3697,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6326981661788423,
+      "grad_norm": 0.5744435257090056,
+      "learning_rate": 1.862092872673821e-05,
+      "loss": 0.3937,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6331760348844155,
+      "grad_norm": 0.6885689441398278,
+      "learning_rate": 1.8618258624160465e-05,
+      "loss": 0.395,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6336539035899886,
+      "grad_norm": 0.5363058119426725,
+      "learning_rate": 1.86155861310298e-05,
+      "loss": 0.3835,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6341317722955618,
+      "grad_norm": 0.5648056274555979,
+      "learning_rate": 1.8612911248087523e-05,
+      "loss": 0.401,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6346096410011349,
+      "grad_norm": 0.5517054491297217,
+      "learning_rate": 1.8610233976075595e-05,
+      "loss": 0.4094,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6350875097067081,
+      "grad_norm": 0.610689280585854,
+      "learning_rate": 1.860755431573664e-05,
+      "loss": 0.4007,
+      "step": 1329
+    },
+    {
+      "epoch": 0.6355653784122812,
+      "grad_norm": 0.5856790883141125,
+      "learning_rate": 1.8604872267813954e-05,
+      "loss": 0.3868,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6360432471178543,
+      "grad_norm": 0.5170823321710041,
+      "learning_rate": 1.8602187833051487e-05,
+      "loss": 0.3772,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6365211158234275,
+      "grad_norm": 0.6525237313804392,
+      "learning_rate": 1.859950101219386e-05,
+      "loss": 0.3696,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6369989845290006,
+      "grad_norm": 0.5357547476404161,
+      "learning_rate": 1.859681180598634e-05,
+      "loss": 0.3866,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6374768532345738,
+      "grad_norm": 0.5669350371510699,
+      "learning_rate": 1.859412021517487e-05,
+      "loss": 0.3887,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6379547219401469,
+      "grad_norm": 0.5936322359971947,
+      "learning_rate": 1.859142624050605e-05,
+      "loss": 0.4002,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6384325906457201,
+      "grad_norm": 0.572891703767859,
+      "learning_rate": 1.8588729882727142e-05,
+      "loss": 0.3722,
+      "step": 1336
+    },
+    {
+      "epoch": 0.6389104593512932,
+      "grad_norm": 0.6958298176911252,
+      "learning_rate": 1.8586031142586073e-05,
+      "loss": 0.3872,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6393883280568664,
+      "grad_norm": 0.6299358564635291,
+      "learning_rate": 1.858333002083141e-05,
+      "loss": 0.4038,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6398661967624395,
+      "grad_norm": 0.5567224363270409,
+      "learning_rate": 1.8580626518212413e-05,
+      "loss": 0.3945,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6403440654680127,
+      "grad_norm": 0.5356857164607767,
+      "learning_rate": 1.8577920635478976e-05,
+      "loss": 0.4022,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6408219341735858,
+      "grad_norm": 0.5602058275360913,
+      "learning_rate": 1.8575212373381672e-05,
+      "loss": 0.3972,
+      "step": 1341
+    },
+    {
+      "epoch": 0.641299802879159,
+      "grad_norm": 0.5391006769979308,
+      "learning_rate": 1.8572501732671714e-05,
+      "loss": 0.4078,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6417776715847321,
+      "grad_norm": 0.5274111377005911,
+      "learning_rate": 1.8569788714100993e-05,
+      "loss": 0.3887,
+      "step": 1343
+    },
+    {
+      "epoch": 0.6422555402903053,
+      "grad_norm": 0.5803845464955352,
+      "learning_rate": 1.8567073318422053e-05,
+      "loss": 0.402,
+      "step": 1344
+    },
+    {
+      "epoch": 0.6427334089958784,
+      "grad_norm": 0.5602953934235403,
+      "learning_rate": 1.8564355546388094e-05,
+      "loss": 0.4069,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6432112777014515,
+      "grad_norm": 0.5353517895711432,
+      "learning_rate": 1.856163539875298e-05,
+      "loss": 0.3795,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6436891464070247,
+      "grad_norm": 0.5562727303504088,
+      "learning_rate": 1.855891287627123e-05,
+      "loss": 0.388,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6441670151125978,
+      "grad_norm": 0.6540276248275614,
+      "learning_rate": 1.8556187979698024e-05,
+      "loss": 0.3935,
+      "step": 1348
+    },
+    {
+      "epoch": 0.644644883818171,
+      "grad_norm": 0.5318329859411215,
+      "learning_rate": 1.85534607097892e-05,
+      "loss": 0.391,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6451227525237441,
+      "grad_norm": 0.5984520705371762,
+      "learning_rate": 1.855073106730126e-05,
+      "loss": 0.3743,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6456006212293173,
+      "grad_norm": 0.8656630009596894,
+      "learning_rate": 1.8547999052991353e-05,
+      "loss": 0.3831,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6460784899348904,
+      "grad_norm": 0.5900719253503339,
+      "learning_rate": 1.854526466761729e-05,
+      "loss": 0.391,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6465563586404636,
+      "grad_norm": 0.5874685110312172,
+      "learning_rate": 1.8542527911937546e-05,
+      "loss": 0.3891,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6470342273460367,
+      "grad_norm": 0.5208727410164187,
+      "learning_rate": 1.8539788786711247e-05,
+      "loss": 0.388,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6475120960516099,
+      "grad_norm": 0.5379067771394836,
+      "learning_rate": 1.8537047292698175e-05,
+      "loss": 0.3847,
+      "step": 1355
+    },
+    {
+      "epoch": 0.647989964757183,
+      "grad_norm": 0.5686849032925404,
+      "learning_rate": 1.853430343065878e-05,
+      "loss": 0.3902,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6484678334627562,
+      "grad_norm": 0.5425524249952016,
+      "learning_rate": 1.853155720135415e-05,
+      "loss": 0.3964,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6489457021683293,
+      "grad_norm": 0.5446815678020227,
+      "learning_rate": 1.8528808605546053e-05,
+      "loss": 0.4028,
+      "step": 1358
+    },
+    {
+      "epoch": 0.6494235708739023,
+      "grad_norm": 0.5502869577791677,
+      "learning_rate": 1.852605764399689e-05,
+      "loss": 0.4149,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6499014395794755,
+      "grad_norm": 0.5356342796004244,
+      "learning_rate": 1.852330431746973e-05,
+      "loss": 0.3927,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6503793082850486,
+      "grad_norm": 0.5620414346740084,
+      "learning_rate": 1.852054862672831e-05,
+      "loss": 0.3847,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6508571769906218,
+      "grad_norm": 0.5494824917990645,
+      "learning_rate": 1.8517790572536996e-05,
+      "loss": 0.3893,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6513350456961949,
+      "grad_norm": 0.6052620118835617,
+      "learning_rate": 1.851503015566083e-05,
+      "loss": 0.3853,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6518129144017681,
+      "grad_norm": 0.5723298054745771,
+      "learning_rate": 1.85122673768655e-05,
+      "loss": 0.3837,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6522907831073412,
+      "grad_norm": 1.0042668593415573,
+      "learning_rate": 1.8509502236917353e-05,
+      "loss": 0.3901,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6527686518129144,
+      "grad_norm": 0.7999380988229664,
+      "learning_rate": 1.8506734736583388e-05,
+      "loss": 0.3922,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6532465205184875,
+      "grad_norm": 0.583257327373365,
+      "learning_rate": 1.850396487663127e-05,
+      "loss": 0.3916,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6537243892240607,
+      "grad_norm": 0.563446303201131,
+      "learning_rate": 1.85011926578293e-05,
+      "loss": 0.3973,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6542022579296338,
+      "grad_norm": 0.5276619379868954,
+      "learning_rate": 1.8498418080946444e-05,
+      "loss": 0.3765,
+      "step": 1369
+    },
+    {
+      "epoch": 0.654680126635207,
+      "grad_norm": 0.5838719171402456,
+      "learning_rate": 1.8495641146752322e-05,
+      "loss": 0.4094,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6551579953407801,
+      "grad_norm": 0.5690336348547724,
+      "learning_rate": 1.8492861856017206e-05,
+      "loss": 0.3861,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6556358640463532,
+      "grad_norm": 0.6359126253694918,
+      "learning_rate": 1.8490080209512024e-05,
+      "loss": 0.4018,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6561137327519264,
+      "grad_norm": 0.6504835153018964,
+      "learning_rate": 1.848729620800835e-05,
+      "loss": 0.4012,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6565916014574995,
+      "grad_norm": 0.5305539981752121,
+      "learning_rate": 1.8484509852278426e-05,
+      "loss": 0.387,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6570694701630727,
+      "grad_norm": 0.5413728281865292,
+      "learning_rate": 1.848172114309513e-05,
+      "loss": 0.3996,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6575473388686458,
+      "grad_norm": 0.5770591202114015,
+      "learning_rate": 1.847893008123201e-05,
+      "loss": 0.3918,
+      "step": 1376
+    },
+    {
+      "epoch": 0.658025207574219,
+      "grad_norm": 0.5418743549514499,
+      "learning_rate": 1.8476136667463246e-05,
+      "loss": 0.3675,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6585030762797921,
+      "grad_norm": 0.5384300809696159,
+      "learning_rate": 1.8473340902563686e-05,
+      "loss": 0.3804,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6589809449853653,
+      "grad_norm": 0.582974175378902,
+      "learning_rate": 1.847054278730883e-05,
+      "loss": 0.3924,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6594588136909384,
+      "grad_norm": 0.521915823350355,
+      "learning_rate": 1.8467742322474822e-05,
+      "loss": 0.3678,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6599366823965116,
+      "grad_norm": 0.552505801390087,
+      "learning_rate": 1.846493950883846e-05,
+      "loss": 0.3673,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6604145511020847,
+      "grad_norm": 0.5738503571000322,
+      "learning_rate": 1.84621343471772e-05,
+      "loss": 0.3784,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6608924198076579,
+      "grad_norm": 0.543933586148958,
+      "learning_rate": 1.8459326838269137e-05,
+      "loss": 0.3883,
+      "step": 1383
+    },
+    {
+      "epoch": 0.661370288513231,
+      "grad_norm": 0.5624988585079359,
+      "learning_rate": 1.8456516982893036e-05,
+      "loss": 0.3841,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6618481572188041,
+      "grad_norm": 0.8760123821285536,
+      "learning_rate": 1.845370478182829e-05,
+      "loss": 0.3789,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6623260259243773,
+      "grad_norm": 0.5255149377123545,
+      "learning_rate": 1.8450890235854958e-05,
+      "loss": 0.3995,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6628038946299504,
+      "grad_norm": 0.8238825404916186,
+      "learning_rate": 1.8448073345753746e-05,
+      "loss": 0.381,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6632817633355236,
+      "grad_norm": 0.5368697433670593,
+      "learning_rate": 1.8445254112306013e-05,
+      "loss": 0.3901,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6637596320410967,
+      "grad_norm": 0.5844500955136835,
+      "learning_rate": 1.8442432536293756e-05,
+      "loss": 0.3841,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6642375007466699,
+      "grad_norm": 0.5317108098879734,
+      "learning_rate": 1.8439608618499637e-05,
+      "loss": 0.3846,
+      "step": 1390
+    },
+    {
+      "epoch": 0.664715369452243,
+      "grad_norm": 0.5486204761949643,
+      "learning_rate": 1.843678235970696e-05,
+      "loss": 0.4078,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6651932381578162,
+      "grad_norm": 0.5621459280922341,
+      "learning_rate": 1.8433953760699678e-05,
+      "loss": 0.3711,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6656711068633893,
+      "grad_norm": 0.626320660146386,
+      "learning_rate": 1.8431122822262398e-05,
+      "loss": 0.3721,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6661489755689625,
+      "grad_norm": 0.5634692726095906,
+      "learning_rate": 1.8428289545180367e-05,
+      "loss": 0.3997,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6666268442745356,
+      "grad_norm": 0.5806101182458545,
+      "learning_rate": 1.842545393023949e-05,
+      "loss": 0.3834,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6671047129801088,
+      "grad_norm": 0.5508052241942929,
+      "learning_rate": 1.8422615978226313e-05,
+      "loss": 0.398,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6675825816856819,
+      "grad_norm": 0.5741650592949158,
+      "learning_rate": 1.8419775689928035e-05,
+      "loss": 0.3806,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6680604503912551,
+      "grad_norm": 0.5163994653334596,
+      "learning_rate": 1.8416933066132507e-05,
+      "loss": 0.3815,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6685383190968281,
+      "grad_norm": 0.6173088911528165,
+      "learning_rate": 1.8414088107628215e-05,
+      "loss": 0.3758,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6690161878024012,
+      "grad_norm": 0.6040108460649616,
+      "learning_rate": 1.841124081520431e-05,
+      "loss": 0.3933,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6694940565079744,
+      "grad_norm": 0.6180489249655141,
+      "learning_rate": 1.840839118965057e-05,
+      "loss": 0.3931,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6699719252135475,
+      "grad_norm": 0.5294472563078757,
+      "learning_rate": 1.8405539231757435e-05,
+      "loss": 0.3943,
+      "step": 1402
+    },
+    {
+      "epoch": 0.6704497939191207,
+      "grad_norm": 0.6292773643061201,
+      "learning_rate": 1.840268494231599e-05,
+      "loss": 0.3938,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6709276626246938,
+      "grad_norm": 0.713996443805077,
+      "learning_rate": 1.839982832211796e-05,
+      "loss": 0.3781,
+      "step": 1404
+    },
+    {
+      "epoch": 0.671405531330267,
+      "grad_norm": 0.5799035626176681,
+      "learning_rate": 1.8396969371955724e-05,
+      "loss": 0.4023,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6718834000358401,
+      "grad_norm": 0.5700183080878829,
+      "learning_rate": 1.8394108092622307e-05,
+      "loss": 0.3844,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6723612687414133,
+      "grad_norm": 0.6189918262496643,
+      "learning_rate": 1.839124448491137e-05,
+      "loss": 0.3948,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6728391374469864,
+      "grad_norm": 0.5222965203774697,
+      "learning_rate": 1.8388378549617238e-05,
+      "loss": 0.4008,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6733170061525596,
+      "grad_norm": 0.7906792457229658,
+      "learning_rate": 1.838551028753486e-05,
+      "loss": 0.3897,
+      "step": 1409
+    },
+    {
+      "epoch": 0.6737948748581327,
+      "grad_norm": 0.6020606169233097,
+      "learning_rate": 1.838263969945985e-05,
+      "loss": 0.3839,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6742727435637059,
+      "grad_norm": 0.5357579566712205,
+      "learning_rate": 1.8379766786188457e-05,
+      "loss": 0.3836,
+      "step": 1411
+    },
+    {
+      "epoch": 0.674750612269279,
+      "grad_norm": 0.5633942493212672,
+      "learning_rate": 1.8376891548517566e-05,
+      "loss": 0.3847,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6752284809748521,
+      "grad_norm": 0.5798470271651702,
+      "learning_rate": 1.837401398724473e-05,
+      "loss": 0.3813,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6757063496804253,
+      "grad_norm": 0.5636278137204155,
+      "learning_rate": 1.837113410316813e-05,
+      "loss": 0.3861,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6761842183859984,
+      "grad_norm": 0.6008374318361374,
+      "learning_rate": 1.836825189708659e-05,
+      "loss": 0.3771,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6766620870915716,
+      "grad_norm": 0.5521729571309322,
+      "learning_rate": 1.836536736979959e-05,
+      "loss": 0.3854,
+      "step": 1416
+    },
+    {
+      "epoch": 0.6771399557971447,
+      "grad_norm": 0.557361883029169,
+      "learning_rate": 1.8362480522107244e-05,
+      "loss": 0.3679,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6776178245027179,
+      "grad_norm": 0.5435824654885472,
+      "learning_rate": 1.8359591354810313e-05,
+      "loss": 0.391,
+      "step": 1418
+    },
+    {
+      "epoch": 0.678095693208291,
+      "grad_norm": 0.5417613234703497,
+      "learning_rate": 1.8356699868710198e-05,
+      "loss": 0.3881,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6785735619138642,
+      "grad_norm": 0.5308236388209336,
+      "learning_rate": 1.8353806064608953e-05,
+      "loss": 0.4036,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6790514306194373,
+      "grad_norm": 0.5884130068215864,
+      "learning_rate": 1.8350909943309262e-05,
+      "loss": 0.3732,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6795292993250105,
+      "grad_norm": 0.5409490290787906,
+      "learning_rate": 1.8348011505614462e-05,
+      "loss": 0.3896,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6800071680305836,
+      "grad_norm": 0.5970144011295364,
+      "learning_rate": 1.8345110752328527e-05,
+      "loss": 0.3834,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6804850367361568,
+      "grad_norm": 0.5261236913058273,
+      "learning_rate": 1.834220768425607e-05,
+      "loss": 0.389,
+      "step": 1424
+    },
+    {
+      "epoch": 0.6809629054417299,
+      "grad_norm": 0.5800849754387479,
+      "learning_rate": 1.833930230220236e-05,
+      "loss": 0.3769,
+      "step": 1425
+    },
+    {
+      "epoch": 0.681440774147303,
+      "grad_norm": 0.5513763570891704,
+      "learning_rate": 1.8336394606973293e-05,
+      "loss": 0.3755,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6819186428528762,
+      "grad_norm": 0.5923393622326396,
+      "learning_rate": 1.8333484599375414e-05,
+      "loss": 0.3781,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6823965115584493,
+      "grad_norm": 0.6542084688710019,
+      "learning_rate": 1.8330572280215904e-05,
+      "loss": 0.3802,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6828743802640225,
+      "grad_norm": 0.5657355120408663,
+      "learning_rate": 1.8327657650302596e-05,
+      "loss": 0.3916,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6833522489695956,
+      "grad_norm": 0.6676995536636485,
+      "learning_rate": 1.8324740710443955e-05,
+      "loss": 0.3841,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6838301176751688,
+      "grad_norm": 0.5639564465708701,
+      "learning_rate": 1.8321821461449084e-05,
+      "loss": 0.3787,
+      "step": 1431
+    },
+    {
+      "epoch": 0.6843079863807419,
+      "grad_norm": 0.5518394477924401,
+      "learning_rate": 1.831889990412773e-05,
+      "loss": 0.389,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6847858550863151,
+      "grad_norm": 0.5756892172775409,
+      "learning_rate": 1.831597603929029e-05,
+      "loss": 0.391,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6852637237918882,
+      "grad_norm": 0.538534263526804,
+      "learning_rate": 1.8313049867747788e-05,
+      "loss": 0.3897,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6857415924974614,
+      "grad_norm": 0.5680170895960552,
+      "learning_rate": 1.831012139031189e-05,
+      "loss": 0.3967,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6862194612030345,
+      "grad_norm": 0.586463487068828,
+      "learning_rate": 1.830719060779491e-05,
+      "loss": 0.3813,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6866973299086077,
+      "grad_norm": 0.6179525461545747,
+      "learning_rate": 1.830425752100979e-05,
+      "loss": 0.3918,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6871751986141807,
+      "grad_norm": 0.6674306038089053,
+      "learning_rate": 1.8301322130770117e-05,
+      "loss": 0.3828,
+      "step": 1438
+    },
+    {
+      "epoch": 0.687653067319754,
+      "grad_norm": 2.332261711848657,
+      "learning_rate": 1.829838443789012e-05,
+      "loss": 0.3838,
+      "step": 1439
+    },
+    {
+      "epoch": 0.688130936025327,
+      "grad_norm": 0.6981750566847236,
+      "learning_rate": 1.829544444318466e-05,
+      "loss": 0.3946,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6886088047309001,
+      "grad_norm": 0.5303694950239685,
+      "learning_rate": 1.829250214746924e-05,
+      "loss": 0.3809,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6890866734364733,
+      "grad_norm": 0.7179900691839765,
+      "learning_rate": 1.8289557551560002e-05,
+      "loss": 0.3862,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6895645421420464,
+      "grad_norm": 0.5345675046456023,
+      "learning_rate": 1.8286610656273724e-05,
+      "loss": 0.3658,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6900424108476196,
+      "grad_norm": 0.6704734216159564,
+      "learning_rate": 1.828366146242782e-05,
+      "loss": 0.4057,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6905202795531927,
+      "grad_norm": 0.6051417776741872,
+      "learning_rate": 1.8280709970840352e-05,
+      "loss": 0.4059,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6909981482587659,
+      "grad_norm": 0.5450521205173916,
+      "learning_rate": 1.8277756182330008e-05,
+      "loss": 0.3914,
+      "step": 1446
+    },
+    {
+      "epoch": 0.691476016964339,
+      "grad_norm": 0.6382596888643569,
+      "learning_rate": 1.8274800097716113e-05,
+      "loss": 0.3682,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6919538856699122,
+      "grad_norm": 0.5469859901894466,
+      "learning_rate": 1.827184171781864e-05,
+      "loss": 0.396,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6924317543754853,
+      "grad_norm": 0.6169779108480756,
+      "learning_rate": 1.8268881043458183e-05,
+      "loss": 0.3852,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6929096230810585,
+      "grad_norm": 0.5515949799511892,
+      "learning_rate": 1.8265918075455985e-05,
+      "loss": 0.3879,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6933874917866316,
+      "grad_norm": 0.5852784051902299,
+      "learning_rate": 1.8262952814633927e-05,
+      "loss": 0.3777,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6938653604922048,
+      "grad_norm": 0.5747324569861376,
+      "learning_rate": 1.8259985261814506e-05,
+      "loss": 0.3708,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6943432291977779,
+      "grad_norm": 0.5584613448774921,
+      "learning_rate": 1.825701541782088e-05,
+      "loss": 0.3939,
+      "step": 1453
+    },
+    {
+      "epoch": 0.694821097903351,
+      "grad_norm": 0.6166551656589301,
+      "learning_rate": 1.825404328347683e-05,
+      "loss": 0.3956,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6952989666089242,
+      "grad_norm": 0.5834938296403458,
+      "learning_rate": 1.8251068859606777e-05,
+      "loss": 0.3797,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6957768353144973,
+      "grad_norm": 0.6153031826664653,
+      "learning_rate": 1.8248092147035762e-05,
+      "loss": 0.3873,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6962547040200705,
+      "grad_norm": 0.5549036883726521,
+      "learning_rate": 1.8245113146589478e-05,
+      "loss": 0.3811,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6967325727256436,
+      "grad_norm": 0.6100500222117385,
+      "learning_rate": 1.8242131859094253e-05,
+      "loss": 0.3739,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6972104414312168,
+      "grad_norm": 0.5644164022778724,
+      "learning_rate": 1.823914828537704e-05,
+      "loss": 0.3805,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6976883101367899,
+      "grad_norm": 0.5720320474251424,
+      "learning_rate": 1.823616242626542e-05,
+      "loss": 0.3716,
+      "step": 1460
+    },
+    {
+      "epoch": 0.6981661788423631,
+      "grad_norm": 0.5353119220152026,
+      "learning_rate": 1.8233174282587636e-05,
+      "loss": 0.3736,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6986440475479362,
+      "grad_norm": 0.6632595208658023,
+      "learning_rate": 1.823018385517253e-05,
+      "loss": 0.3728,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6991219162535094,
+      "grad_norm": 0.5332034499910223,
+      "learning_rate": 1.8227191144849606e-05,
+      "loss": 0.3905,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6995997849590825,
+      "grad_norm": 0.6114377061257181,
+      "learning_rate": 1.822419615244898e-05,
+      "loss": 0.3729,
+      "step": 1464
+    },
+    {
+      "epoch": 0.7000776536646557,
+      "grad_norm": 0.5341668385517253,
+      "learning_rate": 1.8221198878801415e-05,
+      "loss": 0.3839,
+      "step": 1465
+    },
+    {
+      "epoch": 0.7005555223702288,
+      "grad_norm": 0.6080766324288399,
+      "learning_rate": 1.82181993247383e-05,
+      "loss": 0.3795,
+      "step": 1466
+    },
+    {
+      "epoch": 0.7010333910758019,
+      "grad_norm": 0.5588790456997228,
+      "learning_rate": 1.8215197491091657e-05,
+      "loss": 0.4014,
+      "step": 1467
+    },
+    {
+      "epoch": 0.7015112597813751,
+      "grad_norm": 0.5512763293405722,
+      "learning_rate": 1.8212193378694145e-05,
+      "loss": 0.3927,
+      "step": 1468
+    },
+    {
+      "epoch": 0.7019891284869482,
+      "grad_norm": 0.5615590012756244,
+      "learning_rate": 1.8209186988379053e-05,
+      "loss": 0.3922,
+      "step": 1469
+    },
+    {
+      "epoch": 0.7024669971925214,
+      "grad_norm": 0.5651868825673871,
+      "learning_rate": 1.8206178320980295e-05,
+      "loss": 0.3872,
+      "step": 1470
+    },
+    {
+      "epoch": 0.7029448658980945,
+      "grad_norm": 0.5672597491790865,
+      "learning_rate": 1.8203167377332428e-05,
+      "loss": 0.3755,
+      "step": 1471
+    },
+    {
+      "epoch": 0.7034227346036677,
+      "grad_norm": 0.6384822322202582,
+      "learning_rate": 1.820015415827063e-05,
+      "loss": 0.388,
+      "step": 1472
+    },
+    {
+      "epoch": 0.7039006033092408,
+      "grad_norm": 0.6101936905626903,
+      "learning_rate": 1.8197138664630714e-05,
+      "loss": 0.4005,
+      "step": 1473
+    },
+    {
+      "epoch": 0.704378472014814,
+      "grad_norm": 0.5780890528276408,
+      "learning_rate": 1.819412089724913e-05,
+      "loss": 0.3919,
+      "step": 1474
+    },
+    {
+      "epoch": 0.704856340720387,
+      "grad_norm": 0.5709336635510021,
+      "learning_rate": 1.819110085696295e-05,
+      "loss": 0.3851,
+      "step": 1475
+    },
+    {
+      "epoch": 0.7053342094259603,
+      "grad_norm": 0.5653007211064114,
+      "learning_rate": 1.8188078544609885e-05,
+      "loss": 0.3736,
+      "step": 1476
+    },
+    {
+      "epoch": 0.7058120781315333,
+      "grad_norm": 0.5489304827020616,
+      "learning_rate": 1.8185053961028262e-05,
+      "loss": 0.3939,
+      "step": 1477
+    },
+    {
+      "epoch": 0.7062899468371066,
+      "grad_norm": 0.57579168376753,
+      "learning_rate": 1.8182027107057054e-05,
+      "loss": 0.3797,
+      "step": 1478
+    },
+    {
+      "epoch": 0.7067678155426796,
+      "grad_norm": 0.5694198738113794,
+      "learning_rate": 1.8178997983535852e-05,
+      "loss": 0.3695,
+      "step": 1479
+    },
+    {
+      "epoch": 0.7072456842482527,
+      "grad_norm": 0.6271433046445659,
+      "learning_rate": 1.817596659130489e-05,
+      "loss": 0.3941,
+      "step": 1480
+    },
+    {
+      "epoch": 0.7077235529538259,
+      "grad_norm": 0.5356698848547248,
+      "learning_rate": 1.8172932931205018e-05,
+      "loss": 0.3814,
+      "step": 1481
+    },
+    {
+      "epoch": 0.708201421659399,
+      "grad_norm": 0.5350117309625678,
+      "learning_rate": 1.8169897004077714e-05,
+      "loss": 0.3684,
+      "step": 1482
+    },
+    {
+      "epoch": 0.7086792903649722,
+      "grad_norm": 0.563828736209021,
+      "learning_rate": 1.8166858810765093e-05,
+      "loss": 0.3759,
+      "step": 1483
+    },
+    {
+      "epoch": 0.7091571590705453,
+      "grad_norm": 0.5015045069229582,
+      "learning_rate": 1.8163818352109905e-05,
+      "loss": 0.3837,
+      "step": 1484
+    },
+    {
+      "epoch": 0.7096350277761185,
+      "grad_norm": 0.5494681307134424,
+      "learning_rate": 1.816077562895551e-05,
+      "loss": 0.3835,
+      "step": 1485
+    },
+    {
+      "epoch": 0.7101128964816916,
+      "grad_norm": 0.8231790789588842,
+      "learning_rate": 1.8157730642145912e-05,
+      "loss": 0.3928,
+      "step": 1486
+    },
+    {
+      "epoch": 0.7105907651872648,
+      "grad_norm": 0.5469496135283491,
+      "learning_rate": 1.815468339252573e-05,
+      "loss": 0.3996,
+      "step": 1487
+    },
+    {
+      "epoch": 0.7110686338928379,
+      "grad_norm": 0.5496003239400815,
+      "learning_rate": 1.8151633880940226e-05,
+      "loss": 0.3915,
+      "step": 1488
+    },
+    {
+      "epoch": 0.7115465025984111,
+      "grad_norm": 0.5228411620381861,
+      "learning_rate": 1.8148582108235274e-05,
+      "loss": 0.3725,
+      "step": 1489
+    },
+    {
+      "epoch": 0.7120243713039842,
+      "grad_norm": 0.5704507706283268,
+      "learning_rate": 1.814552807525738e-05,
+      "loss": 0.3919,
+      "step": 1490
+    },
+    {
+      "epoch": 0.7125022400095574,
+      "grad_norm": 0.648766640982509,
+      "learning_rate": 1.8142471782853686e-05,
+      "loss": 0.3908,
+      "step": 1491
+    },
+    {
+      "epoch": 0.7129801087151305,
+      "grad_norm": 0.8667271303998992,
+      "learning_rate": 1.813941323187195e-05,
+      "loss": 0.386,
+      "step": 1492
+    },
+    {
+      "epoch": 0.7134579774207037,
+      "grad_norm": 0.5901056863467703,
+      "learning_rate": 1.813635242316056e-05,
+      "loss": 0.3683,
+      "step": 1493
+    },
+    {
+      "epoch": 0.7139358461262768,
+      "grad_norm": 0.5222770355952134,
+      "learning_rate": 1.813328935756853e-05,
+      "loss": 0.396,
+      "step": 1494
+    },
+    {
+      "epoch": 0.7144137148318499,
+      "grad_norm": 0.6020991844265385,
+      "learning_rate": 1.81302240359455e-05,
+      "loss": 0.3642,
+      "step": 1495
+    },
+    {
+      "epoch": 0.7148915835374231,
+      "grad_norm": 0.573707434054128,
+      "learning_rate": 1.812715645914174e-05,
+      "loss": 0.3872,
+      "step": 1496
+    },
+    {
+      "epoch": 0.7153694522429962,
+      "grad_norm": 0.5132272553395804,
+      "learning_rate": 1.8124086628008137e-05,
+      "loss": 0.3992,
+      "step": 1497
+    },
+    {
+      "epoch": 0.7158473209485694,
+      "grad_norm": 0.5814015590506686,
+      "learning_rate": 1.812101454339621e-05,
+      "loss": 0.3813,
+      "step": 1498
+    },
+    {
+      "epoch": 0.7163251896541425,
+      "grad_norm": 0.5029998615126221,
+      "learning_rate": 1.81179402061581e-05,
+      "loss": 0.3932,
+      "step": 1499
+    },
+    {
+      "epoch": 0.7168030583597157,
+      "grad_norm": 0.5827864769681101,
+      "learning_rate": 1.8114863617146576e-05,
+      "loss": 0.3809,
+      "step": 1500
+    },
+    {
+      "epoch": 0.7172809270652888,
+      "grad_norm": 0.6147481989073579,
+      "learning_rate": 1.8111784777215028e-05,
+      "loss": 0.3914,
+      "step": 1501
+    },
+    {
+      "epoch": 0.717758795770862,
+      "grad_norm": 0.6022165391681865,
+      "learning_rate": 1.8108703687217474e-05,
+      "loss": 0.3926,
+      "step": 1502
+    },
+    {
+      "epoch": 0.7182366644764351,
+      "grad_norm": 0.5749879738062318,
+      "learning_rate": 1.8105620348008548e-05,
+      "loss": 0.3698,
+      "step": 1503
+    },
+    {
+      "epoch": 0.7187145331820083,
+      "grad_norm": 0.5144494515929651,
+      "learning_rate": 1.810253476044352e-05,
+      "loss": 0.3903,
+      "step": 1504
+    },
+    {
+      "epoch": 0.7191924018875814,
+      "grad_norm": 0.6680656863681617,
+      "learning_rate": 1.8099446925378278e-05,
+      "loss": 0.4031,
+      "step": 1505
+    },
+    {
+      "epoch": 0.7196702705931546,
+      "grad_norm": 0.5363554834856091,
+      "learning_rate": 1.809635684366933e-05,
+      "loss": 0.3897,
+      "step": 1506
+    },
+    {
+      "epoch": 0.7201481392987277,
+      "grad_norm": 0.57060574097619,
+      "learning_rate": 1.809326451617381e-05,
+      "loss": 0.3847,
+      "step": 1507
+    },
+    {
+      "epoch": 0.7206260080043008,
+      "grad_norm": 0.5082209282023774,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.386,
+      "step": 1508
+    },
+    {
+      "epoch": 0.721103876709874,
+      "grad_norm": 0.531721764246233,
+      "learning_rate": 1.8087073127254708e-05,
+      "loss": 0.379,
+      "step": 1509
+    },
+    {
+      "epoch": 0.7215817454154471,
+      "grad_norm": 0.5672223837229015,
+      "learning_rate": 1.8083974067548506e-05,
+      "loss": 0.3743,
+      "step": 1510
+    },
+    {
+      "epoch": 0.7220596141210203,
+      "grad_norm": 0.557363288254791,
+      "learning_rate": 1.80808727654905e-05,
+      "loss": 0.3922,
+      "step": 1511
+    },
+    {
+      "epoch": 0.7225374828265934,
+      "grad_norm": 0.5524986533091046,
+      "learning_rate": 1.807776922194093e-05,
+      "loss": 0.3828,
+      "step": 1512
+    },
+    {
+      "epoch": 0.7230153515321666,
+      "grad_norm": 0.5624030376925013,
+      "learning_rate": 1.807466343776067e-05,
+      "loss": 0.3879,
+      "step": 1513
+    },
+    {
+      "epoch": 0.7234932202377397,
+      "grad_norm": 0.5587652959424021,
+      "learning_rate": 1.8071555413811202e-05,
+      "loss": 0.3977,
+      "step": 1514
+    },
+    {
+      "epoch": 0.7239710889433129,
+      "grad_norm": 0.5243505053339486,
+      "learning_rate": 1.806844515095465e-05,
+      "loss": 0.3756,
+      "step": 1515
+    },
+    {
+      "epoch": 0.724448957648886,
+      "grad_norm": 0.5136588134139919,
+      "learning_rate": 1.8065332650053733e-05,
+      "loss": 0.386,
+      "step": 1516
+    },
+    {
+      "epoch": 0.7249268263544592,
+      "grad_norm": 0.765740907023463,
+      "learning_rate": 1.806221791197181e-05,
+      "loss": 0.3731,
+      "step": 1517
+    },
+    {
+      "epoch": 0.7254046950600322,
+      "grad_norm": 0.5765742196934072,
+      "learning_rate": 1.8059100937572853e-05,
+      "loss": 0.3859,
+      "step": 1518
+    },
+    {
+      "epoch": 0.7258825637656054,
+      "grad_norm": 0.508218622133711,
+      "learning_rate": 1.8055981727721454e-05,
+      "loss": 0.3887,
+      "step": 1519
+    },
+    {
+      "epoch": 0.7263604324711785,
+      "grad_norm": 0.5169436575801677,
+      "learning_rate": 1.8052860283282832e-05,
+      "loss": 0.3969,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7268383011767516,
+      "grad_norm": 0.5627369158971857,
+      "learning_rate": 1.8049736605122814e-05,
+      "loss": 0.3957,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7273161698823248,
+      "grad_norm": 0.5642116374092477,
+      "learning_rate": 1.804661069410786e-05,
+      "loss": 0.3877,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7277940385878979,
+      "grad_norm": 0.5632828254864813,
+      "learning_rate": 1.8043482551105038e-05,
+      "loss": 0.3846,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7282719072934711,
+      "grad_norm": 0.6720264991530787,
+      "learning_rate": 1.8040352176982042e-05,
+      "loss": 0.3742,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7287497759990442,
+      "grad_norm": 0.5295580482800217,
+      "learning_rate": 1.8037219572607177e-05,
+      "loss": 0.3812,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7292276447046174,
+      "grad_norm": 0.5343714521991437,
+      "learning_rate": 1.803408473884938e-05,
+      "loss": 0.3767,
+      "step": 1526
+    },
+    {
+      "epoch": 0.7297055134101905,
+      "grad_norm": 0.5713754536699026,
+      "learning_rate": 1.8030947676578198e-05,
+      "loss": 0.3829,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7301833821157637,
+      "grad_norm": 0.5439449953744703,
+      "learning_rate": 1.802780838666379e-05,
+      "loss": 0.3883,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7306612508213368,
+      "grad_norm": 1.4893821900130726,
+      "learning_rate": 1.8024666869976946e-05,
+      "loss": 0.3874,
+      "step": 1529
+    },
+    {
+      "epoch": 0.73113911952691,
+      "grad_norm": 0.5642566066926745,
+      "learning_rate": 1.8021523127389066e-05,
+      "loss": 0.3707,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7316169882324831,
+      "grad_norm": 0.5131365637356557,
+      "learning_rate": 1.8018377159772163e-05,
+      "loss": 0.3842,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7320948569380563,
+      "grad_norm": 0.6078967074642584,
+      "learning_rate": 1.801522896799888e-05,
+      "loss": 0.3802,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7325727256436294,
+      "grad_norm": 0.5556930415295182,
+      "learning_rate": 1.801207855294247e-05,
+      "loss": 0.3974,
+      "step": 1533
+    },
+    {
+      "epoch": 0.7330505943492026,
+      "grad_norm": 0.5433320644721947,
+      "learning_rate": 1.8008925915476795e-05,
+      "loss": 0.3817,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7335284630547757,
+      "grad_norm": 0.5725907910354026,
+      "learning_rate": 1.800577105647635e-05,
+      "loss": 0.3862,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7340063317603488,
+      "grad_norm": 0.5275822555651795,
+      "learning_rate": 1.800261397681623e-05,
+      "loss": 0.3662,
+      "step": 1536
+    },
+    {
+      "epoch": 0.734484200465922,
+      "grad_norm": 0.5958868088113626,
+      "learning_rate": 1.799945467737216e-05,
+      "loss": 0.3861,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7349620691714951,
+      "grad_norm": 0.528135241277979,
+      "learning_rate": 1.7996293159020468e-05,
+      "loss": 0.3633,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7354399378770683,
+      "grad_norm": 0.5629034073505758,
+      "learning_rate": 1.799312942263811e-05,
+      "loss": 0.3752,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7359178065826414,
+      "grad_norm": 0.5369379864183004,
+      "learning_rate": 1.7989963469102643e-05,
+      "loss": 0.3788,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7363956752882146,
+      "grad_norm": 0.5181107364251343,
+      "learning_rate": 1.798679529929225e-05,
+      "loss": 0.3908,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7368735439937877,
+      "grad_norm": 0.5889689370863749,
+      "learning_rate": 1.7983624914085726e-05,
+      "loss": 0.38,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7373514126993609,
+      "grad_norm": 0.5087601460758195,
+      "learning_rate": 1.7980452314362482e-05,
+      "loss": 0.3816,
+      "step": 1543
+    },
+    {
+      "epoch": 0.737829281404934,
+      "grad_norm": 0.5940221012894452,
+      "learning_rate": 1.7977277501002538e-05,
+      "loss": 0.3637,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7383071501105072,
+      "grad_norm": 0.5633247453068111,
+      "learning_rate": 1.797410047488653e-05,
+      "loss": 0.3969,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7387850188160803,
+      "grad_norm": 0.5712343756297124,
+      "learning_rate": 1.797092123689571e-05,
+      "loss": 0.3811,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7392628875216535,
+      "grad_norm": 0.7181347071469072,
+      "learning_rate": 1.7967739787911946e-05,
+      "loss": 0.3998,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7397407562272266,
+      "grad_norm": 0.5637357459690837,
+      "learning_rate": 1.7964556128817713e-05,
+      "loss": 0.3927,
+      "step": 1548
+    },
+    {
+      "epoch": 0.7402186249327997,
+      "grad_norm": 0.5365699147490028,
+      "learning_rate": 1.79613702604961e-05,
+      "loss": 0.3888,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7406964936383729,
+      "grad_norm": 0.5518595874098879,
+      "learning_rate": 1.7958182183830816e-05,
+      "loss": 0.3996,
+      "step": 1550
+    },
+    {
+      "epoch": 0.741174362343946,
+      "grad_norm": 0.5208148323047387,
+      "learning_rate": 1.795499189970617e-05,
+      "loss": 0.382,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7416522310495192,
+      "grad_norm": 0.5007085516775489,
+      "learning_rate": 1.79517994090071e-05,
+      "loss": 0.4056,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7421300997550923,
+      "grad_norm": 0.509836881110819,
+      "learning_rate": 1.794860471261914e-05,
+      "loss": 0.3774,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7426079684606655,
+      "grad_norm": 0.5656747401049432,
+      "learning_rate": 1.794540781142844e-05,
+      "loss": 0.3795,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7430858371662385,
+      "grad_norm": 0.5513113429693962,
+      "learning_rate": 1.794220870632177e-05,
+      "loss": 0.373,
+      "step": 1555
+    },
+    {
+      "epoch": 0.7435637058718118,
+      "grad_norm": 0.5180245500331098,
+      "learning_rate": 1.7939007398186507e-05,
+      "loss": 0.3891,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7440415745773848,
+      "grad_norm": 0.5856101098872712,
+      "learning_rate": 1.793580388791063e-05,
+      "loss": 0.3906,
+      "step": 1557
+    },
+    {
+      "epoch": 0.744519443282958,
+      "grad_norm": 0.5597746606677304,
+      "learning_rate": 1.7932598176382735e-05,
+      "loss": 0.3768,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7449973119885311,
+      "grad_norm": 0.5634426580173892,
+      "learning_rate": 1.792939026449204e-05,
+      "loss": 0.3708,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7454751806941043,
+      "grad_norm": 0.5467128365442397,
+      "learning_rate": 1.7926180153128358e-05,
+      "loss": 0.3797,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7459530493996774,
+      "grad_norm": 0.5203223948570378,
+      "learning_rate": 1.7922967843182113e-05,
+      "loss": 0.3872,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7464309181052505,
+      "grad_norm": 0.5963679934080937,
+      "learning_rate": 1.7919753335544352e-05,
+      "loss": 0.3822,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7469087868108237,
+      "grad_norm": 0.5959650909771645,
+      "learning_rate": 1.7916536631106714e-05,
+      "loss": 0.3906,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7473866555163968,
+      "grad_norm": 0.5685204218306926,
+      "learning_rate": 1.7913317730761463e-05,
+      "loss": 0.3845,
+      "step": 1564
+    },
+    {
+      "epoch": 0.74786452422197,
+      "grad_norm": 0.6029950556254274,
+      "learning_rate": 1.791009663540146e-05,
+      "loss": 0.3794,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7483423929275431,
+      "grad_norm": 0.5300367366403645,
+      "learning_rate": 1.790687334592018e-05,
+      "loss": 0.3836,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7488202616331163,
+      "grad_norm": 0.5994658946073593,
+      "learning_rate": 1.790364786321171e-05,
+      "loss": 0.3881,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7492981303386894,
+      "grad_norm": 0.5964406475861134,
+      "learning_rate": 1.7900420188170745e-05,
+      "loss": 0.3905,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7497759990442626,
+      "grad_norm": 1.7504865632734914,
+      "learning_rate": 1.7897190321692578e-05,
+      "loss": 0.3996,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7502538677498357,
+      "grad_norm": 0.5830761379235893,
+      "learning_rate": 1.789395826467312e-05,
+      "loss": 0.3859,
+      "step": 1570
+    },
+    {
+      "epoch": 0.7507317364554089,
+      "grad_norm": 0.5242952519103954,
+      "learning_rate": 1.7890724018008883e-05,
+      "loss": 0.3939,
+      "step": 1571
+    },
+    {
+      "epoch": 0.751209605160982,
+      "grad_norm": 0.543050764985698,
+      "learning_rate": 1.7887487582596995e-05,
+      "loss": 0.394,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7516874738665552,
+      "grad_norm": 0.5259949143161994,
+      "learning_rate": 1.7884248959335186e-05,
+      "loss": 0.3858,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7521653425721283,
+      "grad_norm": 0.5710757710219158,
+      "learning_rate": 1.7881008149121793e-05,
+      "loss": 0.3782,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7526432112777014,
+      "grad_norm": 0.5437004375278933,
+      "learning_rate": 1.7877765152855757e-05,
+      "loss": 0.3744,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7531210799832746,
+      "grad_norm": 0.5158132068664555,
+      "learning_rate": 1.7874519971436627e-05,
+      "loss": 0.378,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7535989486888477,
+      "grad_norm": 0.5706843234145509,
+      "learning_rate": 1.787127260576456e-05,
+      "loss": 0.3843,
+      "step": 1577
+    },
+    {
+      "epoch": 0.7540768173944209,
+      "grad_norm": 0.5829854434809196,
+      "learning_rate": 1.7868023056740323e-05,
+      "loss": 0.383,
+      "step": 1578
+    },
+    {
+      "epoch": 0.754554686099994,
+      "grad_norm": 0.5491598281489602,
+      "learning_rate": 1.7864771325265276e-05,
+      "loss": 0.3843,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7550325548055672,
+      "grad_norm": 0.5283360065954008,
+      "learning_rate": 1.78615174122414e-05,
+      "loss": 0.3867,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7555104235111403,
+      "grad_norm": 0.5361665431481477,
+      "learning_rate": 1.7858261318571265e-05,
+      "loss": 0.3836,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7559882922167135,
+      "grad_norm": 0.5773171737336307,
+      "learning_rate": 1.785500304515806e-05,
+      "loss": 0.4076,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7564661609222866,
+      "grad_norm": 0.5639322568045663,
+      "learning_rate": 1.785174259290557e-05,
+      "loss": 0.3856,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7569440296278598,
+      "grad_norm": 0.5659655900325912,
+      "learning_rate": 1.7848479962718183e-05,
+      "loss": 0.3686,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7574218983334329,
+      "grad_norm": 0.6962795442570291,
+      "learning_rate": 1.78452151555009e-05,
+      "loss": 0.3881,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7578997670390061,
+      "grad_norm": 0.5417577285230938,
+      "learning_rate": 1.7841948172159322e-05,
+      "loss": 0.3794,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7583776357445792,
+      "grad_norm": 0.5211348547780497,
+      "learning_rate": 1.783867901359965e-05,
+      "loss": 0.388,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7588555044501524,
+      "grad_norm": 0.5104365801361086,
+      "learning_rate": 1.7835407680728695e-05,
+      "loss": 0.3777,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7593333731557255,
+      "grad_norm": 0.540783001344937,
+      "learning_rate": 1.783213417445386e-05,
+      "loss": 0.3805,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7598112418612986,
+      "grad_norm": 0.5258232259033346,
+      "learning_rate": 1.7828858495683162e-05,
+      "loss": 0.3767,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7602891105668718,
+      "grad_norm": 0.5222845479618173,
+      "learning_rate": 1.7825580645325217e-05,
+      "loss": 0.3866,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7607669792724449,
+      "grad_norm": 0.5399407703301767,
+      "learning_rate": 1.7822300624289242e-05,
+      "loss": 0.3754,
+      "step": 1592
+    },
+    {
+      "epoch": 0.7612448479780181,
+      "grad_norm": 0.4996656893078933,
+      "learning_rate": 1.7819018433485055e-05,
+      "loss": 0.3792,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7617227166835912,
+      "grad_norm": 0.5274635062067855,
+      "learning_rate": 1.7815734073823078e-05,
+      "loss": 0.3822,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7622005853891644,
+      "grad_norm": 0.6480497541249455,
+      "learning_rate": 1.781244754621434e-05,
+      "loss": 0.3848,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7626784540947374,
+      "grad_norm": 0.5526359451420565,
+      "learning_rate": 1.7809158851570463e-05,
+      "loss": 0.3725,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7631563228003106,
+      "grad_norm": 0.49088307821668725,
+      "learning_rate": 1.780586799080367e-05,
+      "loss": 0.3823,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7636341915058837,
+      "grad_norm": 0.5511921737669051,
+      "learning_rate": 1.780257496482679e-05,
+      "loss": 0.3808,
+      "step": 1598
+    },
+    {
+      "epoch": 0.764112060211457,
+      "grad_norm": 0.6775694510669721,
+      "learning_rate": 1.7799279774553246e-05,
+      "loss": 0.3802,
+      "step": 1599
+    },
+    {
+      "epoch": 0.76458992891703,
+      "grad_norm": 0.5310444266054842,
+      "learning_rate": 1.779598242089707e-05,
+      "loss": 0.3809,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7650677976226032,
+      "grad_norm": 0.4885991078950156,
+      "learning_rate": 1.7792682904772894e-05,
+      "loss": 0.3891,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7655456663281763,
+      "grad_norm": 0.7799415079754219,
+      "learning_rate": 1.7789381227095938e-05,
+      "loss": 0.4004,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7660235350337494,
+      "grad_norm": 0.48613695514881117,
+      "learning_rate": 1.7786077388782034e-05,
+      "loss": 0.3995,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7665014037393226,
+      "grad_norm": 0.5366975828343058,
+      "learning_rate": 1.77827713907476e-05,
+      "loss": 0.3757,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7669792724448957,
+      "grad_norm": 0.5182970925432647,
+      "learning_rate": 1.7779463233909677e-05,
+      "loss": 0.3675,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7674571411504689,
+      "grad_norm": 0.566425081373052,
+      "learning_rate": 1.7776152919185873e-05,
+      "loss": 0.3983,
+      "step": 1606
+    },
+    {
+      "epoch": 0.767935009856042,
+      "grad_norm": 0.5637827498293969,
+      "learning_rate": 1.777284044749442e-05,
+      "loss": 0.3732,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7684128785616152,
+      "grad_norm": 0.6832393610992347,
+      "learning_rate": 1.7769525819754138e-05,
+      "loss": 0.3705,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7688907472671883,
+      "grad_norm": 0.6312177052802135,
+      "learning_rate": 1.7766209036884447e-05,
+      "loss": 0.3672,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7693686159727615,
+      "grad_norm": 0.5350057787652999,
+      "learning_rate": 1.7762890099805362e-05,
+      "loss": 0.3753,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7698464846783346,
+      "grad_norm": 0.5139252783072269,
+      "learning_rate": 1.77595690094375e-05,
+      "loss": 0.3825,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7703243533839078,
+      "grad_norm": 0.849473799017312,
+      "learning_rate": 1.7756245766702068e-05,
+      "loss": 0.3682,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7708022220894809,
+      "grad_norm": 0.5132435684021898,
+      "learning_rate": 1.775292037252088e-05,
+      "loss": 0.3896,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7712800907950541,
+      "grad_norm": 0.5945126305429924,
+      "learning_rate": 1.774959282781634e-05,
+      "loss": 0.3948,
+      "step": 1614
+    },
+    {
+      "epoch": 0.7717579595006272,
+      "grad_norm": 0.5332717557684908,
+      "learning_rate": 1.774626313351145e-05,
+      "loss": 0.3671,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7722358282062003,
+      "grad_norm": 0.5596768760181904,
+      "learning_rate": 1.7742931290529804e-05,
+      "loss": 0.3519,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7727136969117735,
+      "grad_norm": 0.801668548538205,
+      "learning_rate": 1.7739597299795606e-05,
+      "loss": 0.3759,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7731915656173466,
+      "grad_norm": 0.5443956847817307,
+      "learning_rate": 1.773626116223364e-05,
+      "loss": 0.3701,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7736694343229198,
+      "grad_norm": 0.5223788847180523,
+      "learning_rate": 1.7732922878769298e-05,
+      "loss": 0.3815,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7741473030284929,
+      "grad_norm": 0.5153447344508111,
+      "learning_rate": 1.7729582450328547e-05,
+      "loss": 0.388,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7746251717340661,
+      "grad_norm": 0.5399342464141584,
+      "learning_rate": 1.7726239877837977e-05,
+      "loss": 0.3771,
+      "step": 1621
+    },
+    {
+      "epoch": 0.7751030404396392,
+      "grad_norm": 0.5827130864299022,
+      "learning_rate": 1.7722895162224752e-05,
+      "loss": 0.3812,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7755809091452124,
+      "grad_norm": 0.5264498651194451,
+      "learning_rate": 1.7719548304416638e-05,
+      "loss": 0.37,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7760587778507855,
+      "grad_norm": 0.5525289832285949,
+      "learning_rate": 1.7716199305341993e-05,
+      "loss": 0.3641,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7765366465563587,
+      "grad_norm": 0.6073361920730956,
+      "learning_rate": 1.771284816592978e-05,
+      "loss": 0.3751,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7770145152619318,
+      "grad_norm": 0.5261064606183812,
+      "learning_rate": 1.770949488710953e-05,
+      "loss": 0.3831,
+      "step": 1626
+    },
+    {
+      "epoch": 0.777492383967505,
+      "grad_norm": 0.5346800888345036,
+      "learning_rate": 1.7706139469811395e-05,
+      "loss": 0.4004,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7779702526730781,
+      "grad_norm": 0.585135599645527,
+      "learning_rate": 1.77027819149661e-05,
+      "loss": 0.3757,
+      "step": 1628
+    },
+    {
+      "epoch": 0.7784481213786513,
+      "grad_norm": 0.5412022951804295,
+      "learning_rate": 1.7699422223504983e-05,
+      "loss": 0.3903,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7789259900842244,
+      "grad_norm": 0.5087555055370976,
+      "learning_rate": 1.7696060396359956e-05,
+      "loss": 0.3858,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7794038587897975,
+      "grad_norm": 0.5392498385275896,
+      "learning_rate": 1.7692696434463527e-05,
+      "loss": 0.3908,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7798817274953707,
+      "grad_norm": 0.5248886136108247,
+      "learning_rate": 1.768933033874881e-05,
+      "loss": 0.3818,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7803595962009438,
+      "grad_norm": 0.6344637610053451,
+      "learning_rate": 1.768596211014949e-05,
+      "loss": 0.3705,
+      "step": 1633
+    },
+    {
+      "epoch": 0.780837464906517,
+      "grad_norm": 0.5079797518477398,
+      "learning_rate": 1.7682591749599856e-05,
+      "loss": 0.3876,
+      "step": 1634
+    },
+    {
+      "epoch": 0.78131533361209,
+      "grad_norm": 0.7193783569627261,
+      "learning_rate": 1.7679219258034798e-05,
+      "loss": 0.4143,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7817932023176632,
+      "grad_norm": 0.5441633206776926,
+      "learning_rate": 1.767584463638977e-05,
+      "loss": 0.3819,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7822710710232363,
+      "grad_norm": 0.5501870910998597,
+      "learning_rate": 1.767246788560084e-05,
+      "loss": 0.3675,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7827489397288095,
+      "grad_norm": 1.1914603909692592,
+      "learning_rate": 1.766908900660466e-05,
+      "loss": 0.3664,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7832268084343826,
+      "grad_norm": 0.5737236436900817,
+      "learning_rate": 1.7665708000338472e-05,
+      "loss": 0.3912,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7837046771399558,
+      "grad_norm": 0.5111903213622228,
+      "learning_rate": 1.7662324867740102e-05,
+      "loss": 0.3729,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7841825458455289,
+      "grad_norm": 0.5448282748052444,
+      "learning_rate": 1.7658939609747978e-05,
+      "loss": 0.3828,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7846604145511021,
+      "grad_norm": 0.5536887303747314,
+      "learning_rate": 1.7655552227301105e-05,
+      "loss": 0.3734,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7851382832566752,
+      "grad_norm": 0.5055836319468296,
+      "learning_rate": 1.7652162721339085e-05,
+      "loss": 0.3666,
+      "step": 1643
+    },
+    {
+      "epoch": 0.7856161519622483,
+      "grad_norm": 0.588666775287373,
+      "learning_rate": 1.764877109280211e-05,
+      "loss": 0.3917,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7860940206678215,
+      "grad_norm": 0.5912101719269106,
+      "learning_rate": 1.7645377342630956e-05,
+      "loss": 0.3778,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7865718893733946,
+      "grad_norm": 0.5355115123869113,
+      "learning_rate": 1.764198147176699e-05,
+      "loss": 0.3786,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7870497580789678,
+      "grad_norm": 0.5273687503114382,
+      "learning_rate": 1.7638583481152164e-05,
+      "loss": 0.375,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7875276267845409,
+      "grad_norm": 0.6354283594908688,
+      "learning_rate": 1.7635183371729022e-05,
+      "loss": 0.393,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7880054954901141,
+      "grad_norm": 0.5268426064765231,
+      "learning_rate": 1.7631781144440697e-05,
+      "loss": 0.3834,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7884833641956872,
+      "grad_norm": 0.5851405613750261,
+      "learning_rate": 1.76283768002309e-05,
+      "loss": 0.369,
+      "step": 1650
+    },
+    {
+      "epoch": 0.7889612329012604,
+      "grad_norm": 0.5514500060660226,
+      "learning_rate": 1.7624970340043948e-05,
+      "loss": 0.4001,
+      "step": 1651
+    },
+    {
+      "epoch": 0.7894391016068335,
+      "grad_norm": 0.5303902826647586,
+      "learning_rate": 1.7621561764824724e-05,
+      "loss": 0.383,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7899169703124067,
+      "grad_norm": 0.5956286874077665,
+      "learning_rate": 1.7618151075518706e-05,
+      "loss": 0.3741,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7903948390179798,
+      "grad_norm": 0.6017706875819347,
+      "learning_rate": 1.7614738273071963e-05,
+      "loss": 0.3778,
+      "step": 1654
+    },
+    {
+      "epoch": 0.790872707723553,
+      "grad_norm": 0.5218421728887426,
+      "learning_rate": 1.7611323358431145e-05,
+      "loss": 0.3719,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7913505764291261,
+      "grad_norm": 0.5695441720555854,
+      "learning_rate": 1.7607906332543486e-05,
+      "loss": 0.394,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7918284451346992,
+      "grad_norm": 0.5713620867531857,
+      "learning_rate": 1.7604487196356818e-05,
+      "loss": 0.3932,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7923063138402724,
+      "grad_norm": 0.5142635919293859,
+      "learning_rate": 1.7601065950819536e-05,
+      "loss": 0.393,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7927841825458455,
+      "grad_norm": 0.5563424936264186,
+      "learning_rate": 1.7597642596880642e-05,
+      "loss": 0.365,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7932620512514187,
+      "grad_norm": 0.5538505182272977,
+      "learning_rate": 1.759421713548971e-05,
+      "loss": 0.3783,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7937399199569918,
+      "grad_norm": 0.5142957210677673,
+      "learning_rate": 1.7590789567596908e-05,
+      "loss": 0.3953,
+      "step": 1661
+    },
+    {
+      "epoch": 0.794217788662565,
+      "grad_norm": 0.5688406081526746,
+      "learning_rate": 1.758735989415298e-05,
+      "loss": 0.3934,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7946956573681381,
+      "grad_norm": 0.5009105683579143,
+      "learning_rate": 1.758392811610925e-05,
+      "loss": 0.3844,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7951735260737113,
+      "grad_norm": 0.5603598984766094,
+      "learning_rate": 1.7580494234417645e-05,
+      "loss": 0.3952,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7956513947792844,
+      "grad_norm": 0.5385081611958358,
+      "learning_rate": 1.757705825003065e-05,
+      "loss": 0.3791,
+      "step": 1665
+    },
+    {
+      "epoch": 0.7961292634848576,
+      "grad_norm": 0.550224901020255,
+      "learning_rate": 1.7573620163901362e-05,
+      "loss": 0.3723,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7966071321904307,
+      "grad_norm": 0.6130708493238943,
+      "learning_rate": 1.7570179976983433e-05,
+      "loss": 0.3807,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7970850008960039,
+      "grad_norm": 0.5930943169087218,
+      "learning_rate": 1.7566737690231113e-05,
+      "loss": 0.3436,
+      "step": 1668
+    },
+    {
+      "epoch": 0.797562869601577,
+      "grad_norm": 0.5666432561035308,
+      "learning_rate": 1.756329330459923e-05,
+      "loss": 0.3721,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7980407383071502,
+      "grad_norm": 0.5599329725398824,
+      "learning_rate": 1.7559846821043205e-05,
+      "loss": 0.3896,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7985186070127233,
+      "grad_norm": 0.520091169063164,
+      "learning_rate": 1.7556398240519025e-05,
+      "loss": 0.3766,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7989964757182964,
+      "grad_norm": 0.5987277410784928,
+      "learning_rate": 1.755294756398326e-05,
+      "loss": 0.3887,
+      "step": 1672
+    },
+    {
+      "epoch": 0.7994743444238696,
+      "grad_norm": 0.4945539006204766,
+      "learning_rate": 1.7549494792393077e-05,
+      "loss": 0.3861,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7999522131294426,
+      "grad_norm": 0.536193701750873,
+      "learning_rate": 1.7546039926706206e-05,
+      "loss": 0.3772,
+      "step": 1674
+    },
+    {
+      "epoch": 0.8004300818350158,
+      "grad_norm": 0.5264371388469762,
+      "learning_rate": 1.754258296788097e-05,
+      "loss": 0.3936,
+      "step": 1675
+    },
+    {
+      "epoch": 0.8009079505405889,
+      "grad_norm": 0.49365045426457704,
+      "learning_rate": 1.753912391687627e-05,
+      "loss": 0.3854,
+      "step": 1676
+    },
+    {
+      "epoch": 0.8013858192461621,
+      "grad_norm": 0.572847244678029,
+      "learning_rate": 1.753566277465158e-05,
+      "loss": 0.3892,
+      "step": 1677
+    },
+    {
+      "epoch": 0.8018636879517352,
+      "grad_norm": 0.5167500232282058,
+      "learning_rate": 1.7532199542166967e-05,
+      "loss": 0.3847,
+      "step": 1678
+    },
+    {
+      "epoch": 0.8023415566573084,
+      "grad_norm": 0.5616325554937263,
+      "learning_rate": 1.7528734220383065e-05,
+      "loss": 0.3672,
+      "step": 1679
+    },
+    {
+      "epoch": 0.8028194253628815,
+      "grad_norm": 0.5320597666482352,
+      "learning_rate": 1.7525266810261096e-05,
+      "loss": 0.387,
+      "step": 1680
+    },
+    {
+      "epoch": 0.8032972940684547,
+      "grad_norm": 0.5555593972083235,
+      "learning_rate": 1.7521797312762854e-05,
+      "loss": 0.3973,
+      "step": 1681
+    },
+    {
+      "epoch": 0.8037751627740278,
+      "grad_norm": 0.5880958847721388,
+      "learning_rate": 1.7518325728850722e-05,
+      "loss": 0.3683,
+      "step": 1682
+    },
+    {
+      "epoch": 0.804253031479601,
+      "grad_norm": 0.5565593460383242,
+      "learning_rate": 1.7514852059487652e-05,
+      "loss": 0.3751,
+      "step": 1683
+    },
+    {
+      "epoch": 0.8047309001851741,
+      "grad_norm": 0.5268326213678646,
+      "learning_rate": 1.7511376305637183e-05,
+      "loss": 0.3874,
+      "step": 1684
+    },
+    {
+      "epoch": 0.8052087688907472,
+      "grad_norm": 0.5685440803479197,
+      "learning_rate": 1.7507898468263422e-05,
+      "loss": 0.367,
+      "step": 1685
+    },
+    {
+      "epoch": 0.8056866375963204,
+      "grad_norm": 0.5404765320360659,
+      "learning_rate": 1.7504418548331065e-05,
+      "loss": 0.3781,
+      "step": 1686
+    },
+    {
+      "epoch": 0.8061645063018935,
+      "grad_norm": 0.5458851473805669,
+      "learning_rate": 1.750093654680538e-05,
+      "loss": 0.3834,
+      "step": 1687
+    },
+    {
+      "epoch": 0.8066423750074667,
+      "grad_norm": 0.53472908598489,
+      "learning_rate": 1.7497452464652207e-05,
+      "loss": 0.3931,
+      "step": 1688
+    },
+    {
+      "epoch": 0.8071202437130398,
+      "grad_norm": 0.5339306448364256,
+      "learning_rate": 1.7493966302837978e-05,
+      "loss": 0.3765,
+      "step": 1689
+    },
+    {
+      "epoch": 0.807598112418613,
+      "grad_norm": 0.5452629105131837,
+      "learning_rate": 1.7490478062329686e-05,
+      "loss": 0.3854,
+      "step": 1690
+    },
+    {
+      "epoch": 0.8080759811241861,
+      "grad_norm": 0.5007685109159807,
+      "learning_rate": 1.7486987744094905e-05,
+      "loss": 0.3822,
+      "step": 1691
+    },
+    {
+      "epoch": 0.8085538498297593,
+      "grad_norm": 0.5114501233435106,
+      "learning_rate": 1.74834953491018e-05,
+      "loss": 0.3751,
+      "step": 1692
+    },
+    {
+      "epoch": 0.8090317185353324,
+      "grad_norm": 0.5141980507523353,
+      "learning_rate": 1.7480000878319084e-05,
+      "loss": 0.3684,
+      "step": 1693
+    },
+    {
+      "epoch": 0.8095095872409056,
+      "grad_norm": 0.5034630274486521,
+      "learning_rate": 1.7476504332716072e-05,
+      "loss": 0.3758,
+      "step": 1694
+    },
+    {
+      "epoch": 0.8099874559464787,
+      "grad_norm": 0.5387147348927491,
+      "learning_rate": 1.7473005713262644e-05,
+      "loss": 0.37,
+      "step": 1695
+    },
+    {
+      "epoch": 0.8104653246520519,
+      "grad_norm": 0.5442093019130165,
+      "learning_rate": 1.7469505020929252e-05,
+      "loss": 0.3686,
+      "step": 1696
+    },
+    {
+      "epoch": 0.810943193357625,
+      "grad_norm": 0.4947860144971616,
+      "learning_rate": 1.7466002256686925e-05,
+      "loss": 0.3813,
+      "step": 1697
+    },
+    {
+      "epoch": 0.8114210620631981,
+      "grad_norm": 0.5537428519276251,
+      "learning_rate": 1.746249742150727e-05,
+      "loss": 0.3845,
+      "step": 1698
+    },
+    {
+      "epoch": 0.8118989307687713,
+      "grad_norm": 0.5074492556160554,
+      "learning_rate": 1.7458990516362468e-05,
+      "loss": 0.3706,
+      "step": 1699
+    },
+    {
+      "epoch": 0.8123767994743444,
+      "grad_norm": 0.5464251538547661,
+      "learning_rate": 1.7455481542225272e-05,
+      "loss": 0.3771,
+      "step": 1700
+    },
+    {
+      "epoch": 0.8128546681799176,
+      "grad_norm": 0.5386465641293379,
+      "learning_rate": 1.7451970500069007e-05,
+      "loss": 0.3802,
+      "step": 1701
+    },
+    {
+      "epoch": 0.8133325368854907,
+      "grad_norm": 0.6455115743390833,
+      "learning_rate": 1.7448457390867575e-05,
+      "loss": 0.3586,
+      "step": 1702
+    },
+    {
+      "epoch": 0.8138104055910639,
+      "grad_norm": 0.5305807234406511,
+      "learning_rate": 1.744494221559545e-05,
+      "loss": 0.3918,
+      "step": 1703
+    },
+    {
+      "epoch": 0.814288274296637,
+      "grad_norm": 0.5536892586879414,
+      "learning_rate": 1.7441424975227685e-05,
+      "loss": 0.3853,
+      "step": 1704
+    },
+    {
+      "epoch": 0.8147661430022102,
+      "grad_norm": 0.5262189394069378,
+      "learning_rate": 1.7437905670739893e-05,
+      "loss": 0.3738,
+      "step": 1705
+    },
+    {
+      "epoch": 0.8152440117077833,
+      "grad_norm": 0.5550027476074038,
+      "learning_rate": 1.7434384303108273e-05,
+      "loss": 0.3778,
+      "step": 1706
+    },
+    {
+      "epoch": 0.8157218804133565,
+      "grad_norm": 0.5346890732228916,
+      "learning_rate": 1.7430860873309586e-05,
+      "loss": 0.3702,
+      "step": 1707
+    },
+    {
+      "epoch": 0.8161997491189296,
+      "grad_norm": 0.5437600270856643,
+      "learning_rate": 1.7427335382321173e-05,
+      "loss": 0.3684,
+      "step": 1708
+    },
+    {
+      "epoch": 0.8166776178245028,
+      "grad_norm": 0.6463467252286066,
+      "learning_rate": 1.742380783112094e-05,
+      "loss": 0.383,
+      "step": 1709
+    },
+    {
+      "epoch": 0.8171554865300759,
+      "grad_norm": 0.5461358713423635,
+      "learning_rate": 1.7420278220687366e-05,
+      "loss": 0.3672,
+      "step": 1710
+    },
+    {
+      "epoch": 0.817633355235649,
+      "grad_norm": 0.5179296288711173,
+      "learning_rate": 1.7416746551999504e-05,
+      "loss": 0.3631,
+      "step": 1711
+    },
+    {
+      "epoch": 0.8181112239412222,
+      "grad_norm": 0.62856546922575,
+      "learning_rate": 1.741321282603698e-05,
+      "loss": 0.3587,
+      "step": 1712
+    },
+    {
+      "epoch": 0.8185890926467952,
+      "grad_norm": 0.5022602921463213,
+      "learning_rate": 1.7409677043779986e-05,
+      "loss": 0.3827,
+      "step": 1713
+    },
+    {
+      "epoch": 0.8190669613523685,
+      "grad_norm": 0.5287967693920427,
+      "learning_rate": 1.7406139206209283e-05,
+      "loss": 0.3813,
+      "step": 1714
+    },
+    {
+      "epoch": 0.8195448300579415,
+      "grad_norm": 0.5110900484564262,
+      "learning_rate": 1.7402599314306207e-05,
+      "loss": 0.3874,
+      "step": 1715
+    },
+    {
+      "epoch": 0.8200226987635147,
+      "grad_norm": 0.5781567849028013,
+      "learning_rate": 1.739905736905266e-05,
+      "loss": 0.3567,
+      "step": 1716
+    },
+    {
+      "epoch": 0.8205005674690878,
+      "grad_norm": 0.49833085488616846,
+      "learning_rate": 1.739551337143112e-05,
+      "loss": 0.3574,
+      "step": 1717
+    },
+    {
+      "epoch": 0.820978436174661,
+      "grad_norm": 0.5263629142124272,
+      "learning_rate": 1.739196732242462e-05,
+      "loss": 0.3899,
+      "step": 1718
+    },
+    {
+      "epoch": 0.8214563048802341,
+      "grad_norm": 0.5022935161884177,
+      "learning_rate": 1.7388419223016778e-05,
+      "loss": 0.3576,
+      "step": 1719
+    },
+    {
+      "epoch": 0.8219341735858073,
+      "grad_norm": 0.5254848193160996,
+      "learning_rate": 1.7384869074191777e-05,
+      "loss": 0.3779,
+      "step": 1720
+    },
+    {
+      "epoch": 0.8224120422913804,
+      "grad_norm": 0.5213123941213047,
+      "learning_rate": 1.738131687693436e-05,
+      "loss": 0.3586,
+      "step": 1721
+    },
+    {
+      "epoch": 0.8228899109969536,
+      "grad_norm": 0.5156498268841546,
+      "learning_rate": 1.737776263222984e-05,
+      "loss": 0.3724,
+      "step": 1722
+    },
+    {
+      "epoch": 0.8233677797025267,
+      "grad_norm": 0.5704476991648906,
+      "learning_rate": 1.737420634106411e-05,
+      "loss": 0.3705,
+      "step": 1723
+    },
+    {
+      "epoch": 0.8238456484080999,
+      "grad_norm": 0.5546790809955634,
+      "learning_rate": 1.7370648004423623e-05,
+      "loss": 0.3704,
+      "step": 1724
+    },
+    {
+      "epoch": 0.824323517113673,
+      "grad_norm": 0.618179020385375,
+      "learning_rate": 1.7367087623295394e-05,
+      "loss": 0.3723,
+      "step": 1725
+    },
+    {
+      "epoch": 0.8248013858192461,
+      "grad_norm": 0.5663350618454257,
+      "learning_rate": 1.7363525198667013e-05,
+      "loss": 0.3964,
+      "step": 1726
+    },
+    {
+      "epoch": 0.8252792545248193,
+      "grad_norm": 0.5530576617160258,
+      "learning_rate": 1.735996073152663e-05,
+      "loss": 0.3638,
+      "step": 1727
+    },
+    {
+      "epoch": 0.8257571232303924,
+      "grad_norm": 0.5518889071407768,
+      "learning_rate": 1.7356394222862966e-05,
+      "loss": 0.3777,
+      "step": 1728
+    },
+    {
+      "epoch": 0.8262349919359656,
+      "grad_norm": 0.5219424160449163,
+      "learning_rate": 1.7352825673665313e-05,
+      "loss": 0.3773,
+      "step": 1729
+    },
+    {
+      "epoch": 0.8267128606415387,
+      "grad_norm": 0.5416629798429636,
+      "learning_rate": 1.7349255084923517e-05,
+      "loss": 0.3728,
+      "step": 1730
+    },
+    {
+      "epoch": 0.8271907293471119,
+      "grad_norm": 0.542222397838335,
+      "learning_rate": 1.7345682457627998e-05,
+      "loss": 0.3846,
+      "step": 1731
+    },
+    {
+      "epoch": 0.827668598052685,
+      "grad_norm": 0.5564300713817316,
+      "learning_rate": 1.7342107792769747e-05,
+      "loss": 0.3854,
+      "step": 1732
+    },
+    {
+      "epoch": 0.8281464667582582,
+      "grad_norm": 0.5047965719419824,
+      "learning_rate": 1.7338531091340304e-05,
+      "loss": 0.3684,
+      "step": 1733
+    },
+    {
+      "epoch": 0.8286243354638313,
+      "grad_norm": 0.5525238950280134,
+      "learning_rate": 1.7334952354331783e-05,
+      "loss": 0.3604,
+      "step": 1734
+    },
+    {
+      "epoch": 0.8291022041694045,
+      "grad_norm": 0.5386457884764131,
+      "learning_rate": 1.7331371582736864e-05,
+      "loss": 0.382,
+      "step": 1735
+    },
+    {
+      "epoch": 0.8295800728749776,
+      "grad_norm": 0.5721485678679854,
+      "learning_rate": 1.7327788777548796e-05,
+      "loss": 0.3591,
+      "step": 1736
+    },
+    {
+      "epoch": 0.8300579415805508,
+      "grad_norm": 0.9233246568993144,
+      "learning_rate": 1.732420393976138e-05,
+      "loss": 0.3704,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8305358102861239,
+      "grad_norm": 0.5345664368757062,
+      "learning_rate": 1.7320617070368985e-05,
+      "loss": 0.3895,
+      "step": 1738
+    },
+    {
+      "epoch": 0.831013678991697,
+      "grad_norm": 0.5011860668006918,
+      "learning_rate": 1.731702817036655e-05,
+      "loss": 0.3743,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8314915476972702,
+      "grad_norm": 0.49769794527863426,
+      "learning_rate": 1.731343724074957e-05,
+      "loss": 0.3729,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8319694164028433,
+      "grad_norm": 0.4910631974353203,
+      "learning_rate": 1.730984428251411e-05,
+      "loss": 0.3775,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8324472851084165,
+      "grad_norm": 0.4981131547731618,
+      "learning_rate": 1.7306249296656784e-05,
+      "loss": 0.3819,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8329251538139896,
+      "grad_norm": 0.5176839501122841,
+      "learning_rate": 1.7302652284174785e-05,
+      "loss": 0.3719,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8334030225195628,
+      "grad_norm": 0.5152962449518306,
+      "learning_rate": 1.729905324606586e-05,
+      "loss": 0.3804,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8338808912251359,
+      "grad_norm": 0.5558715961134091,
+      "learning_rate": 1.7295452183328317e-05,
+      "loss": 0.3828,
+      "step": 1745
+    },
+    {
+      "epoch": 0.8343587599307091,
+      "grad_norm": 0.5422705807257295,
+      "learning_rate": 1.7291849096961027e-05,
+      "loss": 0.3728,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8348366286362822,
+      "grad_norm": 0.5132286453110108,
+      "learning_rate": 1.7288243987963423e-05,
+      "loss": 0.3915,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8353144973418554,
+      "grad_norm": 0.5896569071524325,
+      "learning_rate": 1.7284636857335503e-05,
+      "loss": 0.3721,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8357923660474285,
+      "grad_norm": 0.5315131515573899,
+      "learning_rate": 1.7281027706077815e-05,
+      "loss": 0.3916,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8362702347530017,
+      "grad_norm": 0.5270182046627034,
+      "learning_rate": 1.7277416535191478e-05,
+      "loss": 0.3867,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8367481034585748,
+      "grad_norm": 0.5016720673015266,
+      "learning_rate": 1.7273803345678163e-05,
+      "loss": 0.3793,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8372259721641478,
+      "grad_norm": 0.538715167203939,
+      "learning_rate": 1.7270188138540106e-05,
+      "loss": 0.3835,
+      "step": 1752
+    },
+    {
+      "epoch": 0.837703840869721,
+      "grad_norm": 0.5152936944221022,
+      "learning_rate": 1.726657091478011e-05,
+      "loss": 0.3848,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8381817095752941,
+      "grad_norm": 0.5144600147503787,
+      "learning_rate": 1.7262951675401517e-05,
+      "loss": 0.375,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8386595782808673,
+      "grad_norm": 0.5789439256338922,
+      "learning_rate": 1.7259330421408247e-05,
+      "loss": 0.3686,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8391374469864404,
+      "grad_norm": 0.5385957955102172,
+      "learning_rate": 1.7255707153804772e-05,
+      "loss": 0.3666,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8396153156920136,
+      "grad_norm": 0.6176394509383132,
+      "learning_rate": 1.725208187359612e-05,
+      "loss": 0.3837,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8400931843975867,
+      "grad_norm": 0.5271166709229913,
+      "learning_rate": 1.724845458178788e-05,
+      "loss": 0.3616,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8405710531031599,
+      "grad_norm": 0.5785356635347916,
+      "learning_rate": 1.72448252793862e-05,
+      "loss": 0.3735,
+      "step": 1759
+    },
+    {
+      "epoch": 0.841048921808733,
+      "grad_norm": 0.5396169495403562,
+      "learning_rate": 1.7241193967397784e-05,
+      "loss": 0.3863,
+      "step": 1760
+    },
+    {
+      "epoch": 0.8415267905143062,
+      "grad_norm": 0.5968305743269035,
+      "learning_rate": 1.7237560646829893e-05,
+      "loss": 0.3844,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8420046592198793,
+      "grad_norm": 0.5739568058873602,
+      "learning_rate": 1.723392531869035e-05,
+      "loss": 0.3623,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8424825279254525,
+      "grad_norm": 0.5245097908307391,
+      "learning_rate": 1.7230287983987524e-05,
+      "loss": 0.3704,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8429603966310256,
+      "grad_norm": 0.5908273151839591,
+      "learning_rate": 1.722664864373035e-05,
+      "loss": 0.3819,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8434382653365988,
+      "grad_norm": 0.5155637558679567,
+      "learning_rate": 1.7223007298928322e-05,
+      "loss": 0.3592,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8439161340421719,
+      "grad_norm": 0.5175194051496753,
+      "learning_rate": 1.7219363950591482e-05,
+      "loss": 0.3819,
+      "step": 1766
+    },
+    {
+      "epoch": 0.844394002747745,
+      "grad_norm": 0.5743671321894646,
+      "learning_rate": 1.7215718599730427e-05,
+      "loss": 0.3766,
+      "step": 1767
+    },
+    {
+      "epoch": 0.8448718714533182,
+      "grad_norm": 0.501602630302012,
+      "learning_rate": 1.7212071247356316e-05,
+      "loss": 0.3879,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8453497401588913,
+      "grad_norm": 0.49054416633074915,
+      "learning_rate": 1.720842189448086e-05,
+      "loss": 0.385,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8458276088644645,
+      "grad_norm": 0.5155620766796909,
+      "learning_rate": 1.7204770542116326e-05,
+      "loss": 0.3691,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8463054775700376,
+      "grad_norm": 0.7473764351264286,
+      "learning_rate": 1.720111719127553e-05,
+      "loss": 0.3691,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8467833462756108,
+      "grad_norm": 0.5182869320161151,
+      "learning_rate": 1.7197461842971854e-05,
+      "loss": 0.3742,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8472612149811839,
+      "grad_norm": 0.539699951292783,
+      "learning_rate": 1.7193804498219222e-05,
+      "loss": 0.3665,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8477390836867571,
+      "grad_norm": 0.49195363335336606,
+      "learning_rate": 1.719014515803212e-05,
+      "loss": 0.3781,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8482169523923302,
+      "grad_norm": 0.5107276989908707,
+      "learning_rate": 1.7186483823425582e-05,
+      "loss": 0.3844,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8486948210979034,
+      "grad_norm": 0.5454810223277834,
+      "learning_rate": 1.7182820495415197e-05,
+      "loss": 0.3615,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8491726898034765,
+      "grad_norm": 0.5294996389762363,
+      "learning_rate": 1.7179155175017115e-05,
+      "loss": 0.3734,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8496505585090497,
+      "grad_norm": 0.5499916031625075,
+      "learning_rate": 1.717548786324802e-05,
+      "loss": 0.3849,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8501284272146228,
+      "grad_norm": 0.4973577029191277,
+      "learning_rate": 1.7171818561125168e-05,
+      "loss": 0.3731,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8506062959201959,
+      "grad_norm": 0.5851851900260322,
+      "learning_rate": 1.7168147269666357e-05,
+      "loss": 0.3799,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8510841646257691,
+      "grad_norm": 0.4822034404377725,
+      "learning_rate": 1.7164473989889937e-05,
+      "loss": 0.3693,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8515620333313422,
+      "grad_norm": 1.2985086241604744,
+      "learning_rate": 1.7160798722814808e-05,
+      "loss": 0.3677,
+      "step": 1782
+    },
+    {
+      "epoch": 0.8520399020369154,
+      "grad_norm": 0.5822593257906487,
+      "learning_rate": 1.7157121469460428e-05,
+      "loss": 0.3575,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8525177707424885,
+      "grad_norm": 0.6086802336624687,
+      "learning_rate": 1.7153442230846808e-05,
+      "loss": 0.3706,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8529956394480617,
+      "grad_norm": 0.5605480795542676,
+      "learning_rate": 1.714976100799449e-05,
+      "loss": 0.3898,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8534735081536348,
+      "grad_norm": 0.4972956581461137,
+      "learning_rate": 1.7146077801924593e-05,
+      "loss": 0.3694,
+      "step": 1786
+    },
+    {
+      "epoch": 0.853951376859208,
+      "grad_norm": 0.5805310199986851,
+      "learning_rate": 1.7142392613658764e-05,
+      "loss": 0.3718,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8544292455647811,
+      "grad_norm": 0.5287792910272413,
+      "learning_rate": 1.7138705444219215e-05,
+      "loss": 0.3714,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8549071142703543,
+      "grad_norm": 0.5101894702630763,
+      "learning_rate": 1.7135016294628703e-05,
+      "loss": 0.3621,
+      "step": 1789
+    },
+    {
+      "epoch": 0.8553849829759274,
+      "grad_norm": 0.6079137123999032,
+      "learning_rate": 1.713132516591053e-05,
+      "loss": 0.3794,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8558628516815006,
+      "grad_norm": 0.5669666874135264,
+      "learning_rate": 1.7127632059088547e-05,
+      "loss": 0.3777,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8563407203870737,
+      "grad_norm": 0.545562892761849,
+      "learning_rate": 1.7123936975187164e-05,
+      "loss": 0.3797,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8568185890926467,
+      "grad_norm": 0.6098893498728746,
+      "learning_rate": 1.7120239915231326e-05,
+      "loss": 0.3689,
+      "step": 1793
+    },
+    {
+      "epoch": 0.85729645779822,
+      "grad_norm": 0.5161732997271268,
+      "learning_rate": 1.7116540880246536e-05,
+      "loss": 0.3894,
+      "step": 1794
+    },
+    {
+      "epoch": 0.857774326503793,
+      "grad_norm": 0.5403314394420677,
+      "learning_rate": 1.7112839871258838e-05,
+      "loss": 0.3683,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8582521952093662,
+      "grad_norm": 0.5784670620134145,
+      "learning_rate": 1.710913688929483e-05,
+      "loss": 0.3815,
+      "step": 1796
+    },
+    {
+      "epoch": 0.8587300639149393,
+      "grad_norm": 0.5025205431269636,
+      "learning_rate": 1.710543193538165e-05,
+      "loss": 0.3879,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8592079326205125,
+      "grad_norm": 0.5892397369469627,
+      "learning_rate": 1.7101725010546988e-05,
+      "loss": 0.3956,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8596858013260856,
+      "grad_norm": 0.47462809973773024,
+      "learning_rate": 1.7098016115819082e-05,
+      "loss": 0.3783,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8601636700316588,
+      "grad_norm": 0.5288631373582081,
+      "learning_rate": 1.7094305252226713e-05,
+      "loss": 0.3958,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8606415387372319,
+      "grad_norm": 0.530585661371997,
+      "learning_rate": 1.7090592420799206e-05,
+      "loss": 0.3794,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8611194074428051,
+      "grad_norm": 0.5288225495006263,
+      "learning_rate": 1.708687762256644e-05,
+      "loss": 0.3562,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8615972761483782,
+      "grad_norm": 0.5052081248332658,
+      "learning_rate": 1.708316085855883e-05,
+      "loss": 0.3788,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8620751448539514,
+      "grad_norm": 0.5037294594905739,
+      "learning_rate": 1.7079442129807345e-05,
+      "loss": 0.3726,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8625530135595245,
+      "grad_norm": 0.49391286749648405,
+      "learning_rate": 1.7075721437343488e-05,
+      "loss": 0.3716,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8630308822650976,
+      "grad_norm": 0.4971338729755265,
+      "learning_rate": 1.707199878219932e-05,
+      "loss": 0.3703,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8635087509706708,
+      "grad_norm": 0.5233353158987447,
+      "learning_rate": 1.7068274165407438e-05,
+      "loss": 0.3863,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8639866196762439,
+      "grad_norm": 0.5533231164405442,
+      "learning_rate": 1.706454758800099e-05,
+      "loss": 0.3771,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8644644883818171,
+      "grad_norm": 0.5368308793227745,
+      "learning_rate": 1.706081905101365e-05,
+      "loss": 0.3606,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8649423570873902,
+      "grad_norm": 0.5642371259255226,
+      "learning_rate": 1.705708855547966e-05,
+      "loss": 0.3536,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8654202257929634,
+      "grad_norm": 0.5545194035776633,
+      "learning_rate": 1.7053356102433786e-05,
+      "loss": 0.3738,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8658980944985365,
+      "grad_norm": 0.66942068698917,
+      "learning_rate": 1.704962169291135e-05,
+      "loss": 0.3773,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8663759632041097,
+      "grad_norm": 0.5154302230933852,
+      "learning_rate": 1.704588532794821e-05,
+      "loss": 0.3703,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8668538319096828,
+      "grad_norm": 0.5317590606844941,
+      "learning_rate": 1.7042147008580768e-05,
+      "loss": 0.3655,
+      "step": 1814
+    },
+    {
+      "epoch": 0.867331700615256,
+      "grad_norm": 0.5285596576650338,
+      "learning_rate": 1.7038406735845967e-05,
+      "loss": 0.3862,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8678095693208291,
+      "grad_norm": 0.556947670099024,
+      "learning_rate": 1.7034664510781294e-05,
+      "loss": 0.3494,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8682874380264023,
+      "grad_norm": 0.5007734114771821,
+      "learning_rate": 1.7030920334424774e-05,
+      "loss": 0.3784,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8687653067319754,
+      "grad_norm": 0.5306943675714777,
+      "learning_rate": 1.7027174207814977e-05,
+      "loss": 0.3735,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8692431754375486,
+      "grad_norm": 0.49311406820837095,
+      "learning_rate": 1.7023426131991008e-05,
+      "loss": 0.3745,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8697210441431217,
+      "grad_norm": 2.167028523901367,
+      "learning_rate": 1.7019676107992523e-05,
+      "loss": 0.384,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8701989128486948,
+      "grad_norm": 0.5774159028549396,
+      "learning_rate": 1.701592413685971e-05,
+      "loss": 0.3688,
+      "step": 1821
+    },
+    {
+      "epoch": 0.870676781554268,
+      "grad_norm": 0.5107936854972478,
+      "learning_rate": 1.7012170219633306e-05,
+      "loss": 0.3823,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8711546502598411,
+      "grad_norm": 0.5153038414963254,
+      "learning_rate": 1.700841435735457e-05,
+      "loss": 0.3906,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8716325189654143,
+      "grad_norm": 0.6927821263237308,
+      "learning_rate": 1.7004656551065317e-05,
+      "loss": 0.3549,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8721103876709874,
+      "grad_norm": 0.5401362110735732,
+      "learning_rate": 1.70008968018079e-05,
+      "loss": 0.3759,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8725882563765606,
+      "grad_norm": 0.5538300760634917,
+      "learning_rate": 1.6997135110625203e-05,
+      "loss": 0.3658,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8730661250821337,
+      "grad_norm": 0.6412840566926354,
+      "learning_rate": 1.6993371478560652e-05,
+      "loss": 0.3801,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8735439937877069,
+      "grad_norm": 0.5385013396738852,
+      "learning_rate": 1.6989605906658217e-05,
+      "loss": 0.3696,
+      "step": 1828
+    },
+    {
+      "epoch": 0.87402186249328,
+      "grad_norm": 0.5786447420911893,
+      "learning_rate": 1.6985838395962397e-05,
+      "loss": 0.3773,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8744997311988532,
+      "grad_norm": 0.528475998356483,
+      "learning_rate": 1.6982068947518235e-05,
+      "loss": 0.3615,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8749775999044263,
+      "grad_norm": 0.9317925627901069,
+      "learning_rate": 1.6978297562371304e-05,
+      "loss": 0.3869,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8754554686099995,
+      "grad_norm": 0.491169996492111,
+      "learning_rate": 1.6974524241567726e-05,
+      "loss": 0.3694,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8759333373155725,
+      "grad_norm": 0.5289533891513116,
+      "learning_rate": 1.6970748986154153e-05,
+      "loss": 0.3659,
+      "step": 1833
+    },
+    {
+      "epoch": 0.8764112060211456,
+      "grad_norm": 0.5541497878429626,
+      "learning_rate": 1.6966971797177777e-05,
+      "loss": 0.3617,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8768890747267188,
+      "grad_norm": 0.5832036352429374,
+      "learning_rate": 1.6963192675686312e-05,
+      "loss": 0.3667,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8773669434322919,
+      "grad_norm": 0.5650619500446709,
+      "learning_rate": 1.6959411622728034e-05,
+      "loss": 0.3808,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8778448121378651,
+      "grad_norm": 0.5699075770705608,
+      "learning_rate": 1.695562863935173e-05,
+      "loss": 0.3656,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8783226808434382,
+      "grad_norm": 0.551562318574458,
+      "learning_rate": 1.695184372660674e-05,
+      "loss": 0.3712,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8788005495490114,
+      "grad_norm": 0.5240567810273391,
+      "learning_rate": 1.6948056885542925e-05,
+      "loss": 0.3543,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8792784182545845,
+      "grad_norm": 0.5020036514932894,
+      "learning_rate": 1.694426811721069e-05,
+      "loss": 0.3679,
+      "step": 1840
+    },
+    {
+      "epoch": 0.8797562869601577,
+      "grad_norm": 0.556752437838373,
+      "learning_rate": 1.6940477422660976e-05,
+      "loss": 0.3704,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8802341556657308,
+      "grad_norm": 0.5191567869165117,
+      "learning_rate": 1.6936684802945255e-05,
+      "loss": 0.3773,
+      "step": 1842
+    },
+    {
+      "epoch": 0.880712024371304,
+      "grad_norm": 0.5533663083233841,
+      "learning_rate": 1.693289025911553e-05,
+      "loss": 0.3779,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8811898930768771,
+      "grad_norm": 0.5503898647969782,
+      "learning_rate": 1.692909379222434e-05,
+      "loss": 0.3791,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8816677617824503,
+      "grad_norm": 0.5551067074917195,
+      "learning_rate": 1.6925295403324758e-05,
+      "loss": 0.3821,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8821456304880234,
+      "grad_norm": 0.5622953147656727,
+      "learning_rate": 1.6921495093470394e-05,
+      "loss": 0.3873,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8826234991935965,
+      "grad_norm": 0.5365186748712514,
+      "learning_rate": 1.6917692863715384e-05,
+      "loss": 0.3717,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8831013678991697,
+      "grad_norm": 0.5478602830128808,
+      "learning_rate": 1.69138887151144e-05,
+      "loss": 0.3723,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8835792366047428,
+      "grad_norm": 0.5427879963525947,
+      "learning_rate": 1.6910082648722643e-05,
+      "loss": 0.3716,
+      "step": 1849
+    },
+    {
+      "epoch": 0.884057105310316,
+      "grad_norm": 0.5101893372822378,
+      "learning_rate": 1.6906274665595854e-05,
+      "loss": 0.369,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8845349740158891,
+      "grad_norm": 0.5434862693905707,
+      "learning_rate": 1.6902464766790295e-05,
+      "loss": 0.3731,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8850128427214623,
+      "grad_norm": 0.5250634322593507,
+      "learning_rate": 1.6898652953362765e-05,
+      "loss": 0.3736,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8854907114270354,
+      "grad_norm": 0.5732211434436395,
+      "learning_rate": 1.68948392263706e-05,
+      "loss": 0.3756,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8859685801326086,
+      "grad_norm": 0.5470798537140039,
+      "learning_rate": 1.6891023586871654e-05,
+      "loss": 0.3834,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8864464488381817,
+      "grad_norm": 0.5126373876854253,
+      "learning_rate": 1.688720603592432e-05,
+      "loss": 0.3726,
+      "step": 1855
+    },
+    {
+      "epoch": 0.8869243175437549,
+      "grad_norm": 0.574477636771307,
+      "learning_rate": 1.6883386574587524e-05,
+      "loss": 0.3726,
+      "step": 1856
+    },
+    {
+      "epoch": 0.887402186249328,
+      "grad_norm": 0.5760651136056916,
+      "learning_rate": 1.687956520392071e-05,
+      "loss": 0.3623,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8878800549549012,
+      "grad_norm": 0.8682852157187648,
+      "learning_rate": 1.6875741924983865e-05,
+      "loss": 0.3741,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8883579236604743,
+      "grad_norm": 0.623287286436862,
+      "learning_rate": 1.687191673883749e-05,
+      "loss": 0.359,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8888357923660475,
+      "grad_norm": 0.5331585373043249,
+      "learning_rate": 1.6868089646542632e-05,
+      "loss": 0.3788,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8893136610716206,
+      "grad_norm": 0.5578200837319638,
+      "learning_rate": 1.686426064916086e-05,
+      "loss": 0.3794,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8897915297771937,
+      "grad_norm": 0.5422159312602259,
+      "learning_rate": 1.6860429747754267e-05,
+      "loss": 0.3772,
+      "step": 1862
+    },
+    {
+      "epoch": 0.8902693984827669,
+      "grad_norm": 0.5228366276138506,
+      "learning_rate": 1.685659694338548e-05,
+      "loss": 0.3806,
+      "step": 1863
+    },
+    {
+      "epoch": 0.89074726718834,
+      "grad_norm": 0.5358911036348268,
+      "learning_rate": 1.6852762237117643e-05,
+      "loss": 0.3834,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8912251358939132,
+      "grad_norm": 0.528212445670647,
+      "learning_rate": 1.6848925630014445e-05,
+      "loss": 0.3882,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8917030045994863,
+      "grad_norm": 0.5542155155507105,
+      "learning_rate": 1.684508712314009e-05,
+      "loss": 0.3753,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8921808733050595,
+      "grad_norm": 0.504104107379108,
+      "learning_rate": 1.6841246717559316e-05,
+      "loss": 0.3875,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8926587420106326,
+      "grad_norm": 0.5434960586896977,
+      "learning_rate": 1.6837404414337374e-05,
+      "loss": 0.3558,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8931366107162058,
+      "grad_norm": 0.5567958899249101,
+      "learning_rate": 1.683356021454006e-05,
+      "loss": 0.3672,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8936144794217789,
+      "grad_norm": 0.5022263534465929,
+      "learning_rate": 1.6829714119233688e-05,
+      "loss": 0.3537,
+      "step": 1870
+    },
+    {
+      "epoch": 0.894092348127352,
+      "grad_norm": 0.549284213737301,
+      "learning_rate": 1.6825866129485088e-05,
+      "loss": 0.3574,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8945702168329251,
+      "grad_norm": 0.4705829377264607,
+      "learning_rate": 1.6822016246361633e-05,
+      "loss": 0.3851,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8950480855384984,
+      "grad_norm": 0.5062114611657378,
+      "learning_rate": 1.681816447093121e-05,
+      "loss": 0.3609,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8955259542440714,
+      "grad_norm": 0.5826140193721359,
+      "learning_rate": 1.6814310804262225e-05,
+      "loss": 0.3726,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8960038229496445,
+      "grad_norm": 0.5520132821306097,
+      "learning_rate": 1.6810455247423634e-05,
+      "loss": 0.3782,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8964816916552177,
+      "grad_norm": 0.5170660835598903,
+      "learning_rate": 1.680659780148489e-05,
+      "loss": 0.357,
+      "step": 1876
+    },
+    {
+      "epoch": 0.8969595603607908,
+      "grad_norm": 0.583946284953352,
+      "learning_rate": 1.680273846751598e-05,
+      "loss": 0.3776,
+      "step": 1877
+    },
+    {
+      "epoch": 0.897437429066364,
+      "grad_norm": 0.5723420342490975,
+      "learning_rate": 1.6798877246587418e-05,
+      "loss": 0.3895,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8979152977719371,
+      "grad_norm": 0.5082366243810379,
+      "learning_rate": 1.679501413977024e-05,
+      "loss": 0.3808,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8983931664775103,
+      "grad_norm": 0.507443343885867,
+      "learning_rate": 1.6791149148136003e-05,
+      "loss": 0.3536,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8988710351830834,
+      "grad_norm": 0.5703634285838638,
+      "learning_rate": 1.6787282272756784e-05,
+      "loss": 0.3682,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8993489038886566,
+      "grad_norm": 0.5158030787718338,
+      "learning_rate": 1.6783413514705186e-05,
+      "loss": 0.3813,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8998267725942297,
+      "grad_norm": 0.5015158481243369,
+      "learning_rate": 1.677954287505434e-05,
+      "loss": 0.3775,
+      "step": 1883
+    },
+    {
+      "epoch": 0.9003046412998029,
+      "grad_norm": 0.5098825854109337,
+      "learning_rate": 1.6775670354877888e-05,
+      "loss": 0.3526,
+      "step": 1884
+    },
+    {
+      "epoch": 0.900782510005376,
+      "grad_norm": 0.5242066060684257,
+      "learning_rate": 1.677179595525e-05,
+      "loss": 0.4053,
+      "step": 1885
+    },
+    {
+      "epoch": 0.9012603787109492,
+      "grad_norm": 0.5351980918146104,
+      "learning_rate": 1.6767919677245367e-05,
+      "loss": 0.3714,
+      "step": 1886
+    },
+    {
+      "epoch": 0.9017382474165223,
+      "grad_norm": 0.5328925764874355,
+      "learning_rate": 1.6764041521939194e-05,
+      "loss": 0.3675,
+      "step": 1887
+    },
+    {
+      "epoch": 0.9022161161220954,
+      "grad_norm": 0.5494270094887949,
+      "learning_rate": 1.6760161490407227e-05,
+      "loss": 0.388,
+      "step": 1888
+    },
+    {
+      "epoch": 0.9026939848276686,
+      "grad_norm": 0.5031329675709398,
+      "learning_rate": 1.675627958372571e-05,
+      "loss": 0.3811,
+      "step": 1889
+    },
+    {
+      "epoch": 0.9031718535332417,
+      "grad_norm": 0.5262746250301709,
+      "learning_rate": 1.675239580297141e-05,
+      "loss": 0.3935,
+      "step": 1890
+    },
+    {
+      "epoch": 0.9036497222388149,
+      "grad_norm": 0.5069957733853212,
+      "learning_rate": 1.6748510149221623e-05,
+      "loss": 0.3777,
+      "step": 1891
+    },
+    {
+      "epoch": 0.904127590944388,
+      "grad_norm": 0.5428615459596646,
+      "learning_rate": 1.6744622623554166e-05,
+      "loss": 0.3802,
+      "step": 1892
+    },
+    {
+      "epoch": 0.9046054596499612,
+      "grad_norm": 0.5266303210825704,
+      "learning_rate": 1.6740733227047365e-05,
+      "loss": 0.3776,
+      "step": 1893
+    },
+    {
+      "epoch": 0.9050833283555343,
+      "grad_norm": 0.5348979320077515,
+      "learning_rate": 1.673684196078007e-05,
+      "loss": 0.3848,
+      "step": 1894
+    },
+    {
+      "epoch": 0.9055611970611075,
+      "grad_norm": 0.5314119413401441,
+      "learning_rate": 1.6732948825831657e-05,
+      "loss": 0.364,
+      "step": 1895
+    },
+    {
+      "epoch": 0.9060390657666806,
+      "grad_norm": 0.5516947593383675,
+      "learning_rate": 1.6729053823282e-05,
+      "loss": 0.3553,
+      "step": 1896
+    },
+    {
+      "epoch": 0.9065169344722538,
+      "grad_norm": 0.5090831476794752,
+      "learning_rate": 1.6725156954211516e-05,
+      "loss": 0.3713,
+      "step": 1897
+    },
+    {
+      "epoch": 0.9069948031778269,
+      "grad_norm": 0.6262571396693599,
+      "learning_rate": 1.672125821970112e-05,
+      "loss": 0.3689,
+      "step": 1898
+    },
+    {
+      "epoch": 0.9074726718834001,
+      "grad_norm": 0.5193279263946532,
+      "learning_rate": 1.6717357620832256e-05,
+      "loss": 0.3783,
+      "step": 1899
+    },
+    {
+      "epoch": 0.9079505405889732,
+      "grad_norm": 0.503064317241693,
+      "learning_rate": 1.671345515868688e-05,
+      "loss": 0.3793,
+      "step": 1900
+    },
+    {
+      "epoch": 0.9084284092945463,
+      "grad_norm": 0.5130059203403388,
+      "learning_rate": 1.6709550834347463e-05,
+      "loss": 0.3715,
+      "step": 1901
+    },
+    {
+      "epoch": 0.9089062780001195,
+      "grad_norm": 0.5369725030475303,
+      "learning_rate": 1.6705644648897004e-05,
+      "loss": 0.3592,
+      "step": 1902
+    },
+    {
+      "epoch": 0.9093841467056926,
+      "grad_norm": 0.5150663544165971,
+      "learning_rate": 1.6701736603419002e-05,
+      "loss": 0.3637,
+      "step": 1903
+    },
+    {
+      "epoch": 0.9098620154112658,
+      "grad_norm": 0.5399311902064715,
+      "learning_rate": 1.6697826698997483e-05,
+      "loss": 0.3594,
+      "step": 1904
+    },
+    {
+      "epoch": 0.9103398841168389,
+      "grad_norm": 0.5518934019533673,
+      "learning_rate": 1.6693914936716983e-05,
+      "loss": 0.3682,
+      "step": 1905
+    },
+    {
+      "epoch": 0.9108177528224121,
+      "grad_norm": 0.5209140835926345,
+      "learning_rate": 1.6690001317662563e-05,
+      "loss": 0.382,
+      "step": 1906
+    },
+    {
+      "epoch": 0.9112956215279852,
+      "grad_norm": 0.5660288276754546,
+      "learning_rate": 1.6686085842919784e-05,
+      "loss": 0.3801,
+      "step": 1907
+    },
+    {
+      "epoch": 0.9117734902335584,
+      "grad_norm": 0.47006676537993314,
+      "learning_rate": 1.668216851357473e-05,
+      "loss": 0.3724,
+      "step": 1908
+    },
+    {
+      "epoch": 0.9122513589391315,
+      "grad_norm": 0.5577644040191538,
+      "learning_rate": 1.667824933071401e-05,
+      "loss": 0.3622,
+      "step": 1909
+    },
+    {
+      "epoch": 0.9127292276447047,
+      "grad_norm": 0.5340230250631273,
+      "learning_rate": 1.6674328295424723e-05,
+      "loss": 0.369,
+      "step": 1910
+    },
+    {
+      "epoch": 0.9132070963502777,
+      "grad_norm": 0.5079231886115417,
+      "learning_rate": 1.6670405408794498e-05,
+      "loss": 0.3715,
+      "step": 1911
+    },
+    {
+      "epoch": 0.913684965055851,
+      "grad_norm": 0.5325193692349207,
+      "learning_rate": 1.666648067191148e-05,
+      "loss": 0.3789,
+      "step": 1912
+    },
+    {
+      "epoch": 0.914162833761424,
+      "grad_norm": 0.5323003539826137,
+      "learning_rate": 1.666255408586432e-05,
+      "loss": 0.3818,
+      "step": 1913
+    },
+    {
+      "epoch": 0.9146407024669972,
+      "grad_norm": 0.4869815217201792,
+      "learning_rate": 1.6658625651742178e-05,
+      "loss": 0.3733,
+      "step": 1914
+    },
+    {
+      "epoch": 0.9151185711725703,
+      "grad_norm": 0.5576240488695627,
+      "learning_rate": 1.6654695370634738e-05,
+      "loss": 0.3589,
+      "step": 1915
+    },
+    {
+      "epoch": 0.9155964398781434,
+      "grad_norm": 0.49247551368135417,
+      "learning_rate": 1.6650763243632187e-05,
+      "loss": 0.3585,
+      "step": 1916
+    },
+    {
+      "epoch": 0.9160743085837166,
+      "grad_norm": 0.5148567366006771,
+      "learning_rate": 1.664682927182523e-05,
+      "loss": 0.3838,
+      "step": 1917
+    },
+    {
+      "epoch": 0.9165521772892897,
+      "grad_norm": 0.49717329992717463,
+      "learning_rate": 1.6642893456305086e-05,
+      "loss": 0.3915,
+      "step": 1918
+    },
+    {
+      "epoch": 0.9170300459948629,
+      "grad_norm": 0.5628249235481696,
+      "learning_rate": 1.663895579816347e-05,
+      "loss": 0.3689,
+      "step": 1919
+    },
+    {
+      "epoch": 0.917507914700436,
+      "grad_norm": 0.5136903278787316,
+      "learning_rate": 1.6635016298492628e-05,
+      "loss": 0.3838,
+      "step": 1920
+    },
+    {
+      "epoch": 0.9179857834060092,
+      "grad_norm": 0.5334625808445861,
+      "learning_rate": 1.6631074958385304e-05,
+      "loss": 0.374,
+      "step": 1921
+    },
+    {
+      "epoch": 0.9184636521115823,
+      "grad_norm": 0.5386195590160799,
+      "learning_rate": 1.6627131778934755e-05,
+      "loss": 0.3751,
+      "step": 1922
+    },
+    {
+      "epoch": 0.9189415208171555,
+      "grad_norm": 0.5379052116006212,
+      "learning_rate": 1.662318676123475e-05,
+      "loss": 0.3809,
+      "step": 1923
+    },
+    {
+      "epoch": 0.9194193895227286,
+      "grad_norm": 0.5100602264626746,
+      "learning_rate": 1.6619239906379574e-05,
+      "loss": 0.3701,
+      "step": 1924
+    },
+    {
+      "epoch": 0.9198972582283018,
+      "grad_norm": 0.5239150729526496,
+      "learning_rate": 1.6615291215464005e-05,
+      "loss": 0.3877,
+      "step": 1925
+    },
+    {
+      "epoch": 0.9203751269338749,
+      "grad_norm": 0.6375609494234671,
+      "learning_rate": 1.6611340689583343e-05,
+      "loss": 0.3844,
+      "step": 1926
+    },
+    {
+      "epoch": 0.9208529956394481,
+      "grad_norm": 0.5123037134022694,
+      "learning_rate": 1.66073883298334e-05,
+      "loss": 0.3651,
+      "step": 1927
+    },
+    {
+      "epoch": 0.9213308643450212,
+      "grad_norm": 0.5110576525109256,
+      "learning_rate": 1.6603434137310482e-05,
+      "loss": 0.3707,
+      "step": 1928
+    },
+    {
+      "epoch": 0.9218087330505943,
+      "grad_norm": 0.5021261687202652,
+      "learning_rate": 1.6599478113111424e-05,
+      "loss": 0.3682,
+      "step": 1929
+    },
+    {
+      "epoch": 0.9222866017561675,
+      "grad_norm": 0.5371433965125264,
+      "learning_rate": 1.6595520258333545e-05,
+      "loss": 0.3624,
+      "step": 1930
+    },
+    {
+      "epoch": 0.9227644704617406,
+      "grad_norm": 0.5366454399722713,
+      "learning_rate": 1.659156057407469e-05,
+      "loss": 0.3809,
+      "step": 1931
+    },
+    {
+      "epoch": 0.9232423391673138,
+      "grad_norm": 0.9136617720639402,
+      "learning_rate": 1.6587599061433207e-05,
+      "loss": 0.3738,
+      "step": 1932
+    },
+    {
+      "epoch": 0.9237202078728869,
+      "grad_norm": 0.8488713857773934,
+      "learning_rate": 1.6583635721507944e-05,
+      "loss": 0.3637,
+      "step": 1933
+    },
+    {
+      "epoch": 0.9241980765784601,
+      "grad_norm": 0.49495459546373427,
+      "learning_rate": 1.6579670555398268e-05,
+      "loss": 0.367,
+      "step": 1934
+    },
+    {
+      "epoch": 0.9246759452840332,
+      "grad_norm": 0.5422350097681802,
+      "learning_rate": 1.657570356420404e-05,
+      "loss": 0.3747,
+      "step": 1935
+    },
+    {
+      "epoch": 0.9251538139896064,
+      "grad_norm": 0.5485293026148379,
+      "learning_rate": 1.657173474902564e-05,
+      "loss": 0.3767,
+      "step": 1936
+    },
+    {
+      "epoch": 0.9256316826951795,
+      "grad_norm": 0.7216669697310142,
+      "learning_rate": 1.6567764110963948e-05,
+      "loss": 0.3761,
+      "step": 1937
+    },
+    {
+      "epoch": 0.9261095514007527,
+      "grad_norm": 0.4879951676041884,
+      "learning_rate": 1.6563791651120336e-05,
+      "loss": 0.3637,
+      "step": 1938
+    },
+    {
+      "epoch": 0.9265874201063258,
+      "grad_norm": 0.49428473376773735,
+      "learning_rate": 1.6559817370596708e-05,
+      "loss": 0.3843,
+      "step": 1939
+    },
+    {
+      "epoch": 0.927065288811899,
+      "grad_norm": 0.48309373277907747,
+      "learning_rate": 1.6555841270495456e-05,
+      "loss": 0.3765,
+      "step": 1940
+    },
+    {
+      "epoch": 0.9275431575174721,
+      "grad_norm": 0.5013632030999634,
+      "learning_rate": 1.6551863351919478e-05,
+      "loss": 0.3694,
+      "step": 1941
+    },
+    {
+      "epoch": 0.9280210262230452,
+      "grad_norm": 0.5110638525672886,
+      "learning_rate": 1.6547883615972176e-05,
+      "loss": 0.363,
+      "step": 1942
+    },
+    {
+      "epoch": 0.9284988949286184,
+      "grad_norm": 0.5185504025173149,
+      "learning_rate": 1.6543902063757462e-05,
+      "loss": 0.3801,
+      "step": 1943
+    },
+    {
+      "epoch": 0.9289767636341915,
+      "grad_norm": 0.5049387315294666,
+      "learning_rate": 1.653991869637975e-05,
+      "loss": 0.369,
+      "step": 1944
+    },
+    {
+      "epoch": 0.9294546323397647,
+      "grad_norm": 0.4912665471357074,
+      "learning_rate": 1.6535933514943955e-05,
+      "loss": 0.3702,
+      "step": 1945
+    },
+    {
+      "epoch": 0.9299325010453378,
+      "grad_norm": 0.5118031125606555,
+      "learning_rate": 1.653194652055549e-05,
+      "loss": 0.3732,
+      "step": 1946
+    },
+    {
+      "epoch": 0.930410369750911,
+      "grad_norm": 0.6431555235235055,
+      "learning_rate": 1.6527957714320283e-05,
+      "loss": 0.3725,
+      "step": 1947
+    },
+    {
+      "epoch": 0.930888238456484,
+      "grad_norm": 0.5097523807572212,
+      "learning_rate": 1.6523967097344763e-05,
+      "loss": 0.3712,
+      "step": 1948
+    },
+    {
+      "epoch": 0.9313661071620573,
+      "grad_norm": 0.5046812127135673,
+      "learning_rate": 1.6519974670735846e-05,
+      "loss": 0.3737,
+      "step": 1949
+    },
+    {
+      "epoch": 0.9318439758676303,
+      "grad_norm": 0.5194776247965222,
+      "learning_rate": 1.6515980435600965e-05,
+      "loss": 0.3742,
+      "step": 1950
+    },
+    {
+      "epoch": 0.9323218445732036,
+      "grad_norm": 0.5404940696555603,
+      "learning_rate": 1.6511984393048055e-05,
+      "loss": 0.381,
+      "step": 1951
+    },
+    {
+      "epoch": 0.9327997132787766,
+      "grad_norm": 0.5553063994108789,
+      "learning_rate": 1.6507986544185543e-05,
+      "loss": 0.3649,
+      "step": 1952
+    },
+    {
+      "epoch": 0.9332775819843498,
+      "grad_norm": 0.5499668048968661,
+      "learning_rate": 1.650398689012236e-05,
+      "loss": 0.3793,
+      "step": 1953
+    },
+    {
+      "epoch": 0.9337554506899229,
+      "grad_norm": 0.563687647978333,
+      "learning_rate": 1.649998543196794e-05,
+      "loss": 0.3808,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9342333193954961,
+      "grad_norm": 0.545135564437546,
+      "learning_rate": 1.6495982170832224e-05,
+      "loss": 0.3706,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9347111881010692,
+      "grad_norm": 0.5303943311116299,
+      "learning_rate": 1.6491977107825642e-05,
+      "loss": 0.3685,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9351890568066423,
+      "grad_norm": 0.5619908532760816,
+      "learning_rate": 1.648797024405912e-05,
+      "loss": 0.3883,
+      "step": 1957
+    },
+    {
+      "epoch": 0.9356669255122155,
+      "grad_norm": 0.5038859263688038,
+      "learning_rate": 1.64839615806441e-05,
+      "loss": 0.3766,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9361447942177886,
+      "grad_norm": 0.521422602845323,
+      "learning_rate": 1.6479951118692515e-05,
+      "loss": 0.3895,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9366226629233618,
+      "grad_norm": 0.5282657221332856,
+      "learning_rate": 1.6475938859316795e-05,
+      "loss": 0.3605,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9371005316289349,
+      "grad_norm": 0.6351039349256459,
+      "learning_rate": 1.6471924803629867e-05,
+      "loss": 0.3651,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9375784003345081,
+      "grad_norm": 0.5764981472703029,
+      "learning_rate": 1.6467908952745163e-05,
+      "loss": 0.38,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9380562690400812,
+      "grad_norm": 8.536515562815161,
+      "learning_rate": 1.6463891307776606e-05,
+      "loss": 0.3956,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9385341377456544,
+      "grad_norm": 0.6278724935607395,
+      "learning_rate": 1.645987186983862e-05,
+      "loss": 0.3739,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9390120064512275,
+      "grad_norm": 0.48387133354028933,
+      "learning_rate": 1.6455850640046134e-05,
+      "loss": 0.3888,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9394898751568007,
+      "grad_norm": 0.531275176529521,
+      "learning_rate": 1.6451827619514552e-05,
+      "loss": 0.3649,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9399677438623738,
+      "grad_norm": 0.569449658310809,
+      "learning_rate": 1.6447802809359802e-05,
+      "loss": 0.3699,
+      "step": 1967
+    },
+    {
+      "epoch": 0.940445612567947,
+      "grad_norm": 0.5120339336372176,
+      "learning_rate": 1.6443776210698288e-05,
+      "loss": 0.3846,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9409234812735201,
+      "grad_norm": 0.5460326477017806,
+      "learning_rate": 1.643974782464692e-05,
+      "loss": 0.3465,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9414013499790932,
+      "grad_norm": 0.518882665215977,
+      "learning_rate": 1.6435717652323097e-05,
+      "loss": 0.3839,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9418792186846664,
+      "grad_norm": 0.49852258072808264,
+      "learning_rate": 1.6431685694844725e-05,
+      "loss": 0.3816,
+      "step": 1971
+    },
+    {
+      "epoch": 0.9423570873902395,
+      "grad_norm": 0.5815555518919842,
+      "learning_rate": 1.6427651953330196e-05,
+      "loss": 0.381,
+      "step": 1972
+    },
+    {
+      "epoch": 0.9428349560958127,
+      "grad_norm": 0.48319880339598503,
+      "learning_rate": 1.6423616428898392e-05,
+      "loss": 0.3701,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9433128248013858,
+      "grad_norm": 0.5167840266697357,
+      "learning_rate": 1.6419579122668704e-05,
+      "loss": 0.3556,
+      "step": 1974
+    },
+    {
+      "epoch": 0.943790693506959,
+      "grad_norm": 0.5618010439515246,
+      "learning_rate": 1.6415540035761008e-05,
+      "loss": 0.3744,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9442685622125321,
+      "grad_norm": 0.5051139249357881,
+      "learning_rate": 1.641149916929567e-05,
+      "loss": 0.3876,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9447464309181053,
+      "grad_norm": 0.5611312662809763,
+      "learning_rate": 1.6407456524393562e-05,
+      "loss": 0.3686,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9452242996236784,
+      "grad_norm": 0.5732842999649458,
+      "learning_rate": 1.640341210217604e-05,
+      "loss": 0.3865,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9457021683292516,
+      "grad_norm": 0.550740123702534,
+      "learning_rate": 1.6399365903764956e-05,
+      "loss": 0.3743,
+      "step": 1979
+    },
+    {
+      "epoch": 0.9461800370348247,
+      "grad_norm": 0.5379350100432237,
+      "learning_rate": 1.639531793028265e-05,
+      "loss": 0.3808,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9466579057403979,
+      "grad_norm": 0.5315866561065092,
+      "learning_rate": 1.6391268182851963e-05,
+      "loss": 0.3734,
+      "step": 1981
+    },
+    {
+      "epoch": 0.947135774445971,
+      "grad_norm": 0.5224017338900245,
+      "learning_rate": 1.638721666259622e-05,
+      "loss": 0.3859,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9476136431515441,
+      "grad_norm": 0.5967691599781494,
+      "learning_rate": 1.638316337063925e-05,
+      "loss": 0.3581,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9480915118571173,
+      "grad_norm": 0.589119592504554,
+      "learning_rate": 1.6379108308105354e-05,
+      "loss": 0.3835,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9485693805626904,
+      "grad_norm": 0.51383408409613,
+      "learning_rate": 1.637505147611934e-05,
+      "loss": 0.3586,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9490472492682636,
+      "grad_norm": 0.4994797959754618,
+      "learning_rate": 1.63709928758065e-05,
+      "loss": 0.377,
+      "step": 1986
+    },
+    {
+      "epoch": 0.9495251179738367,
+      "grad_norm": 0.5048066876195055,
+      "learning_rate": 1.6366932508292618e-05,
+      "loss": 0.3731,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9500029866794099,
+      "grad_norm": 0.524811235817121,
+      "learning_rate": 1.6362870374703967e-05,
+      "loss": 0.3793,
+      "step": 1988
+    },
+    {
+      "epoch": 0.950480855384983,
+      "grad_norm": 0.5735930473974293,
+      "learning_rate": 1.6358806476167316e-05,
+      "loss": 0.3656,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9509587240905562,
+      "grad_norm": 0.4832184523655007,
+      "learning_rate": 1.6354740813809917e-05,
+      "loss": 0.373,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9514365927961292,
+      "grad_norm": 0.5128768000065408,
+      "learning_rate": 1.635067338875951e-05,
+      "loss": 0.3821,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9519144615017024,
+      "grad_norm": 0.5022274034535443,
+      "learning_rate": 1.6346604202144326e-05,
+      "loss": 0.3901,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9523923302072755,
+      "grad_norm": 0.5489970781237143,
+      "learning_rate": 1.634253325509309e-05,
+      "loss": 0.3858,
+      "step": 1993
+    },
+    {
+      "epoch": 0.9528701989128487,
+      "grad_norm": 0.4975450932340785,
+      "learning_rate": 1.6338460548735015e-05,
+      "loss": 0.3686,
+      "step": 1994
+    },
+    {
+      "epoch": 0.9533480676184218,
+      "grad_norm": 0.48127408958137835,
+      "learning_rate": 1.6334386084199787e-05,
+      "loss": 0.3745,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9538259363239949,
+      "grad_norm": 0.5704221304645912,
+      "learning_rate": 1.6330309862617598e-05,
+      "loss": 0.3613,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9543038050295681,
+      "grad_norm": 0.5148160729331526,
+      "learning_rate": 1.6326231885119117e-05,
+      "loss": 0.3827,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9547816737351412,
+      "grad_norm": 0.6284728203646394,
+      "learning_rate": 1.63221521528355e-05,
+      "loss": 0.3824,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9552595424407144,
+      "grad_norm": 0.49271041299162915,
+      "learning_rate": 1.63180706668984e-05,
+      "loss": 0.3691,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9557374111462875,
+      "grad_norm": 0.5390149320147591,
+      "learning_rate": 1.631398742843995e-05,
+      "loss": 0.3652,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9562152798518607,
+      "grad_norm": 0.5025205155225639,
+      "learning_rate": 1.6309902438592762e-05,
+      "loss": 0.3694,
+      "step": 2001
+    },
+    {
+      "epoch": 0.9566931485574338,
+      "grad_norm": 0.5222794351472518,
+      "learning_rate": 1.6305815698489938e-05,
+      "loss": 0.3681,
+      "step": 2002
+    },
+    {
+      "epoch": 0.957171017263007,
+      "grad_norm": 0.5145118353547241,
+      "learning_rate": 1.6301727209265077e-05,
+      "loss": 0.3719,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9576488859685801,
+      "grad_norm": 0.4908469670116591,
+      "learning_rate": 1.629763697205225e-05,
+      "loss": 0.3705,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9581267546741533,
+      "grad_norm": 0.5012664663409069,
+      "learning_rate": 1.629354498798601e-05,
+      "loss": 0.3719,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9586046233797264,
+      "grad_norm": 0.518588891339538,
+      "learning_rate": 1.628945125820141e-05,
+      "loss": 0.3709,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9590824920852996,
+      "grad_norm": 0.5117699235235205,
+      "learning_rate": 1.628535578383397e-05,
+      "loss": 0.3737,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9595603607908727,
+      "grad_norm": 0.5356226048939458,
+      "learning_rate": 1.6281258566019712e-05,
+      "loss": 0.3764,
+      "step": 2008
+    },
+    {
+      "epoch": 0.9600382294964459,
+      "grad_norm": 0.5083134780109303,
+      "learning_rate": 1.6277159605895124e-05,
+      "loss": 0.3723,
+      "step": 2009
+    },
+    {
+      "epoch": 0.960516098202019,
+      "grad_norm": 0.5023579232084872,
+      "learning_rate": 1.627305890459719e-05,
+      "loss": 0.3704,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9609939669075921,
+      "grad_norm": 0.46631779465683365,
+      "learning_rate": 1.6268956463263372e-05,
+      "loss": 0.3526,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9614718356131653,
+      "grad_norm": 0.547788682791615,
+      "learning_rate": 1.6264852283031614e-05,
+      "loss": 0.3779,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9619497043187384,
+      "grad_norm": 0.5104177809427011,
+      "learning_rate": 1.6260746365040342e-05,
+      "loss": 0.3729,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9624275730243116,
+      "grad_norm": 0.5421693818087101,
+      "learning_rate": 1.6256638710428468e-05,
+      "loss": 0.3698,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9629054417298847,
+      "grad_norm": 0.5878685955211844,
+      "learning_rate": 1.625252932033538e-05,
+      "loss": 0.3702,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9633833104354579,
+      "grad_norm": 0.5392634273182241,
+      "learning_rate": 1.6248418195900944e-05,
+      "loss": 0.3714,
+      "step": 2016
+    },
+    {
+      "epoch": 0.963861179141031,
+      "grad_norm": 0.5041453371523026,
+      "learning_rate": 1.6244305338265528e-05,
+      "loss": 0.3757,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9643390478466042,
+      "grad_norm": 0.5611684018900728,
+      "learning_rate": 1.6240190748569958e-05,
+      "loss": 0.3615,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9648169165521773,
+      "grad_norm": 0.4958027277870733,
+      "learning_rate": 1.6236074427955547e-05,
+      "loss": 0.3643,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9652947852577505,
+      "grad_norm": 0.5065207954211904,
+      "learning_rate": 1.6231956377564095e-05,
+      "loss": 0.372,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9657726539633236,
+      "grad_norm": 0.48383262645888736,
+      "learning_rate": 1.6227836598537874e-05,
+      "loss": 0.3924,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9662505226688968,
+      "grad_norm": 0.5200139461232003,
+      "learning_rate": 1.6223715092019636e-05,
+      "loss": 0.3671,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9667283913744699,
+      "grad_norm": 0.4953988372654815,
+      "learning_rate": 1.6219591859152618e-05,
+      "loss": 0.3812,
+      "step": 2023
+    },
+    {
+      "epoch": 0.967206260080043,
+      "grad_norm": 0.5102837388791924,
+      "learning_rate": 1.621546690108053e-05,
+      "loss": 0.3652,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9676841287856162,
+      "grad_norm": 0.5629208320653784,
+      "learning_rate": 1.621134021894756e-05,
+      "loss": 0.3611,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9681619974911893,
+      "grad_norm": 0.4952840329409754,
+      "learning_rate": 1.6207211813898377e-05,
+      "loss": 0.3972,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9686398661967625,
+      "grad_norm": 0.5037099311393208,
+      "learning_rate": 1.6203081687078136e-05,
+      "loss": 0.373,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9691177349023355,
+      "grad_norm": 0.5838992353089708,
+      "learning_rate": 1.6198949839632453e-05,
+      "loss": 0.3793,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9695956036079088,
+      "grad_norm": 0.49580068754399503,
+      "learning_rate": 1.619481627270743e-05,
+      "loss": 0.3613,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9700734723134818,
+      "grad_norm": 0.532208761854008,
+      "learning_rate": 1.619068098744965e-05,
+      "loss": 0.3866,
+      "step": 2030
+    },
+    {
+      "epoch": 0.970551341019055,
+      "grad_norm": 0.5224695569046617,
+      "learning_rate": 1.6186543985006164e-05,
+      "loss": 0.3699,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9710292097246281,
+      "grad_norm": 0.5227062263015988,
+      "learning_rate": 1.6182405266524507e-05,
+      "loss": 0.3912,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9715070784302013,
+      "grad_norm": 0.52945280171711,
+      "learning_rate": 1.6178264833152688e-05,
+      "loss": 0.3403,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9719849471357744,
+      "grad_norm": 0.5629792708553427,
+      "learning_rate": 1.6174122686039182e-05,
+      "loss": 0.3784,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9724628158413476,
+      "grad_norm": 0.5163999238740223,
+      "learning_rate": 1.6169978826332955e-05,
+      "loss": 0.3659,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9729406845469207,
+      "grad_norm": 0.5133995557530827,
+      "learning_rate": 1.6165833255183438e-05,
+      "loss": 0.3694,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9734185532524938,
+      "grad_norm": 0.5617728810922827,
+      "learning_rate": 1.616168597374054e-05,
+      "loss": 0.363,
+      "step": 2037
+    },
+    {
+      "epoch": 0.973896421958067,
+      "grad_norm": 0.5237674641752671,
+      "learning_rate": 1.615753698315464e-05,
+      "loss": 0.3803,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9743742906636401,
+      "grad_norm": 0.5228216583365605,
+      "learning_rate": 1.61533862845766e-05,
+      "loss": 0.3681,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9748521593692133,
+      "grad_norm": 0.4805229482972204,
+      "learning_rate": 1.6149233879157747e-05,
+      "loss": 0.3658,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9753300280747864,
+      "grad_norm": 0.5534895159460543,
+      "learning_rate": 1.614507976804989e-05,
+      "loss": 0.3784,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9758078967803596,
+      "grad_norm": 0.5486724592299036,
+      "learning_rate": 1.6140923952405302e-05,
+      "loss": 0.3786,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9762857654859327,
+      "grad_norm": 0.8080884579911872,
+      "learning_rate": 1.6136766433376728e-05,
+      "loss": 0.3723,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9767636341915059,
+      "grad_norm": 0.5359191028098362,
+      "learning_rate": 1.6132607212117404e-05,
+      "loss": 0.3526,
+      "step": 2044
+    },
+    {
+      "epoch": 0.977241502897079,
+      "grad_norm": 0.5053369417965282,
+      "learning_rate": 1.6128446289781012e-05,
+      "loss": 0.3772,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9777193716026522,
+      "grad_norm": 0.5211994750741029,
+      "learning_rate": 1.6124283667521727e-05,
+      "loss": 0.3808,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9781972403082253,
+      "grad_norm": 0.5676346296630453,
+      "learning_rate": 1.612011934649418e-05,
+      "loss": 0.3715,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9786751090137985,
+      "grad_norm": 0.5248006849417738,
+      "learning_rate": 1.611595332785348e-05,
+      "loss": 0.3553,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9791529777193716,
+      "grad_norm": 0.508623732364023,
+      "learning_rate": 1.6111785612755214e-05,
+      "loss": 0.3754,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9796308464249448,
+      "grad_norm": 0.5266899133269225,
+      "learning_rate": 1.610761620235543e-05,
+      "loss": 0.3859,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9801087151305179,
+      "grad_norm": 0.4751459644310755,
+      "learning_rate": 1.610344509781065e-05,
+      "loss": 0.387,
+      "step": 2051
+    },
+    {
+      "epoch": 0.980586583836091,
+      "grad_norm": 0.5316863726383205,
+      "learning_rate": 1.609927230027786e-05,
+      "loss": 0.3959,
+      "step": 2052
+    },
+    {
+      "epoch": 0.9810644525416642,
+      "grad_norm": 0.5453039434436993,
+      "learning_rate": 1.609509781091452e-05,
+      "loss": 0.3697,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9815423212472373,
+      "grad_norm": 0.4825716401466539,
+      "learning_rate": 1.6090921630878568e-05,
+      "loss": 0.3835,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9820201899528105,
+      "grad_norm": 0.5178510139332422,
+      "learning_rate": 1.60867437613284e-05,
+      "loss": 0.3717,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9824980586583836,
+      "grad_norm": 0.524887791800861,
+      "learning_rate": 1.6082564203422876e-05,
+      "loss": 0.3653,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9829759273639568,
+      "grad_norm": 0.5213297178508581,
+      "learning_rate": 1.6078382958321336e-05,
+      "loss": 0.3725,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9834537960695299,
+      "grad_norm": 0.5282681790101221,
+      "learning_rate": 1.6074200027183584e-05,
+      "loss": 0.3695,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9839316647751031,
+      "grad_norm": 0.489145805606624,
+      "learning_rate": 1.6070015411169896e-05,
+      "loss": 0.3623,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9844095334806762,
+      "grad_norm": 0.5331793559039849,
+      "learning_rate": 1.6065829111441e-05,
+      "loss": 0.3556,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9848874021862494,
+      "grad_norm": 0.5443653643393922,
+      "learning_rate": 1.6061641129158112e-05,
+      "loss": 0.3988,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9853652708918225,
+      "grad_norm": 0.8351330747470432,
+      "learning_rate": 1.60574514654829e-05,
+      "loss": 0.3804,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9858431395973957,
+      "grad_norm": 0.5030615466689651,
+      "learning_rate": 1.6053260121577503e-05,
+      "loss": 0.3618,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9863210083029688,
+      "grad_norm": 0.5471550674629175,
+      "learning_rate": 1.6049067098604523e-05,
+      "loss": 0.3557,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9867988770085419,
+      "grad_norm": 0.48069192998466753,
+      "learning_rate": 1.6044872397727037e-05,
+      "loss": 0.3612,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9872767457141151,
+      "grad_norm": 0.5253037063797116,
+      "learning_rate": 1.6040676020108577e-05,
+      "loss": 0.3655,
+      "step": 2066
+    },
+    {
+      "epoch": 0.9877546144196881,
+      "grad_norm": 0.5275280865811858,
+      "learning_rate": 1.6036477966913143e-05,
+      "loss": 0.3852,
+      "step": 2067
+    },
+    {
+      "epoch": 0.9882324831252614,
+      "grad_norm": 0.47485994365355333,
+      "learning_rate": 1.6032278239305204e-05,
+      "loss": 0.3638,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9887103518308344,
+      "grad_norm": 0.5362179099114655,
+      "learning_rate": 1.6028076838449692e-05,
+      "loss": 0.3725,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9891882205364076,
+      "grad_norm": 0.5119867606693408,
+      "learning_rate": 1.6023873765511993e-05,
+      "loss": 0.3757,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9896660892419807,
+      "grad_norm": 0.5570349801753108,
+      "learning_rate": 1.6019669021657972e-05,
+      "loss": 0.362,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9901439579475539,
+      "grad_norm": 0.5023908521667267,
+      "learning_rate": 1.601546260805395e-05,
+      "loss": 0.3664,
+      "step": 2072
+    },
+    {
+      "epoch": 0.990621826653127,
+      "grad_norm": 0.4919070804327344,
+      "learning_rate": 1.6011254525866715e-05,
+      "loss": 0.3746,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9910996953587002,
+      "grad_norm": 0.5274265284295131,
+      "learning_rate": 1.600704477626351e-05,
+      "loss": 0.3554,
+      "step": 2074
+    },
+    {
+      "epoch": 0.9915775640642733,
+      "grad_norm": 0.4973674393943186,
+      "learning_rate": 1.6002833360412044e-05,
+      "loss": 0.3734,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9920554327698465,
+      "grad_norm": 0.6994137247566886,
+      "learning_rate": 1.599862027948049e-05,
+      "loss": 0.3785,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9925333014754196,
+      "grad_norm": 0.5005262403669962,
+      "learning_rate": 1.5994405534637487e-05,
+      "loss": 0.3508,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9930111701809927,
+      "grad_norm": 0.48923707105340497,
+      "learning_rate": 1.5990189127052128e-05,
+      "loss": 0.3767,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9934890388865659,
+      "grad_norm": 0.5287580848621545,
+      "learning_rate": 1.5985971057893973e-05,
+      "loss": 0.3754,
+      "step": 2079
+    },
+    {
+      "epoch": 0.993966907592139,
+      "grad_norm": 0.4820940737829678,
+      "learning_rate": 1.5981751328333036e-05,
+      "loss": 0.3631,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9944447762977122,
+      "grad_norm": 0.47169419878350755,
+      "learning_rate": 1.5977529939539794e-05,
+      "loss": 0.3737,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9949226450032853,
+      "grad_norm": 0.4819615935924389,
+      "learning_rate": 1.597330689268519e-05,
+      "loss": 0.3776,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9954005137088585,
+      "grad_norm": 0.4915767151527447,
+      "learning_rate": 1.5969082188940623e-05,
+      "loss": 0.3855,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9958783824144316,
+      "grad_norm": 0.5283871739026111,
+      "learning_rate": 1.5964855829477946e-05,
+      "loss": 0.3643,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9963562511200048,
+      "grad_norm": 0.5657962366230493,
+      "learning_rate": 1.5960627815469486e-05,
+      "loss": 0.3671,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9968341198255779,
+      "grad_norm": 0.47929648182468015,
+      "learning_rate": 1.5956398148088007e-05,
+      "loss": 0.3733,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9973119885311511,
+      "grad_norm": 0.5282895558309355,
+      "learning_rate": 1.5952166828506754e-05,
+      "loss": 0.3546,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9977898572367242,
+      "grad_norm": 0.49273301808521663,
+      "learning_rate": 1.5947933857899418e-05,
+      "loss": 0.3794,
+      "step": 2088
+    },
+    {
+      "epoch": 0.9982677259422974,
+      "grad_norm": 0.5307309337337829,
+      "learning_rate": 1.594369923744015e-05,
+      "loss": 0.367,
+      "step": 2089
+    },
+    {
+      "epoch": 0.9987455946478705,
+      "grad_norm": 0.5584015875131911,
+      "learning_rate": 1.5939462968303554e-05,
+      "loss": 0.3699,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9992234633534436,
+      "grad_norm": 0.5866978710305609,
+      "learning_rate": 1.5935225051664708e-05,
+      "loss": 0.3759,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9997013320590168,
+      "grad_norm": 0.48873807272784037,
+      "learning_rate": 1.593098548869912e-05,
+      "loss": 0.3608,
+      "step": 2092
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5723334542119868,
+      "learning_rate": 1.5926744280582786e-05,
+      "loss": 0.3643,
+      "step": 2093
+    },
+    {
+      "epoch": 1.000477868705573,
+      "grad_norm": 0.6300673986894956,
+      "learning_rate": 1.5922501428492126e-05,
+      "loss": 0.333,
+      "step": 2094
+    },
+    {
+      "epoch": 1.0009557374111462,
+      "grad_norm": 0.5750968585044649,
+      "learning_rate": 1.5918256933604047e-05,
+      "loss": 0.3407,
+      "step": 2095
+    },
+    {
+      "epoch": 1.0014336061167195,
+      "grad_norm": 0.5649190535406365,
+      "learning_rate": 1.591401079709589e-05,
+      "loss": 0.3313,
+      "step": 2096
+    },
+    {
+      "epoch": 1.0019114748222926,
+      "grad_norm": 0.6470098562888061,
+      "learning_rate": 1.590976302014546e-05,
+      "loss": 0.3193,
+      "step": 2097
+    },
+    {
+      "epoch": 1.0023893435278657,
+      "grad_norm": 0.6124457488598071,
+      "learning_rate": 1.5905513603931013e-05,
+      "loss": 0.3453,
+      "step": 2098
+    },
+    {
+      "epoch": 1.0028672122334388,
+      "grad_norm": 0.5569653564403705,
+      "learning_rate": 1.5901262549631266e-05,
+      "loss": 0.3308,
+      "step": 2099
+    },
+    {
+      "epoch": 1.003345080939012,
+      "grad_norm": 0.5630136553754124,
+      "learning_rate": 1.589700985842538e-05,
+      "loss": 0.3401,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0038229496445852,
+      "grad_norm": 0.5491390499825274,
+      "learning_rate": 1.5892755531492986e-05,
+      "loss": 0.3368,
+      "step": 2101
+    },
+    {
+      "epoch": 1.0043008183501583,
+      "grad_norm": 0.5802336855773184,
+      "learning_rate": 1.5888499570014152e-05,
+      "loss": 0.3379,
+      "step": 2102
+    },
+    {
+      "epoch": 1.0047786870557314,
+      "grad_norm": 0.5027412997145168,
+      "learning_rate": 1.5884241975169406e-05,
+      "loss": 0.3239,
+      "step": 2103
+    },
+    {
+      "epoch": 1.0052565557613047,
+      "grad_norm": 0.5219023282328462,
+      "learning_rate": 1.5879982748139738e-05,
+      "loss": 0.3316,
+      "step": 2104
+    },
+    {
+      "epoch": 1.0057344244668778,
+      "grad_norm": 0.6198051403881502,
+      "learning_rate": 1.5875721890106574e-05,
+      "loss": 0.3355,
+      "step": 2105
+    },
+    {
+      "epoch": 1.0062122931724509,
+      "grad_norm": 0.5823835508633309,
+      "learning_rate": 1.58714594022518e-05,
+      "loss": 0.3421,
+      "step": 2106
+    },
+    {
+      "epoch": 1.006690161878024,
+      "grad_norm": 0.6314654286023943,
+      "learning_rate": 1.586719528575776e-05,
+      "loss": 0.3492,
+      "step": 2107
+    },
+    {
+      "epoch": 1.007168030583597,
+      "grad_norm": 0.5494826948922337,
+      "learning_rate": 1.5862929541807247e-05,
+      "loss": 0.3187,
+      "step": 2108
+    },
+    {
+      "epoch": 1.0076458992891704,
+      "grad_norm": 0.49367247951130727,
+      "learning_rate": 1.5858662171583495e-05,
+      "loss": 0.343,
+      "step": 2109
+    },
+    {
+      "epoch": 1.0081237679947435,
+      "grad_norm": 0.5867003697639408,
+      "learning_rate": 1.5854393176270205e-05,
+      "loss": 0.3222,
+      "step": 2110
+    },
+    {
+      "epoch": 1.0086016367003166,
+      "grad_norm": 0.5202276726676689,
+      "learning_rate": 1.585012255705152e-05,
+      "loss": 0.3395,
+      "step": 2111
+    },
+    {
+      "epoch": 1.0090795054058896,
+      "grad_norm": 0.5331521180511377,
+      "learning_rate": 1.5845850315112025e-05,
+      "loss": 0.304,
+      "step": 2112
+    },
+    {
+      "epoch": 1.009557374111463,
+      "grad_norm": 0.5838843873042685,
+      "learning_rate": 1.5841576451636777e-05,
+      "loss": 0.3392,
+      "step": 2113
+    },
+    {
+      "epoch": 1.010035242817036,
+      "grad_norm": 0.5162331359042389,
+      "learning_rate": 1.5837300967811258e-05,
+      "loss": 0.343,
+      "step": 2114
+    },
+    {
+      "epoch": 1.0105131115226091,
+      "grad_norm": 0.6115190735592178,
+      "learning_rate": 1.5833023864821427e-05,
+      "loss": 0.321,
+      "step": 2115
+    },
+    {
+      "epoch": 1.0109909802281822,
+      "grad_norm": 0.5925028855716654,
+      "learning_rate": 1.5828745143853665e-05,
+      "loss": 0.3393,
+      "step": 2116
+    },
+    {
+      "epoch": 1.0114688489337555,
+      "grad_norm": 0.6132879239893447,
+      "learning_rate": 1.5824464806094817e-05,
+      "loss": 0.3292,
+      "step": 2117
+    },
+    {
+      "epoch": 1.0119467176393286,
+      "grad_norm": 0.5969110143142136,
+      "learning_rate": 1.5820182852732177e-05,
+      "loss": 0.3234,
+      "step": 2118
+    },
+    {
+      "epoch": 1.0124245863449017,
+      "grad_norm": 0.6131274616744073,
+      "learning_rate": 1.5815899284953477e-05,
+      "loss": 0.3437,
+      "step": 2119
+    },
+    {
+      "epoch": 1.0129024550504748,
+      "grad_norm": 0.5161682873500958,
+      "learning_rate": 1.5811614103946905e-05,
+      "loss": 0.3366,
+      "step": 2120
+    },
+    {
+      "epoch": 1.013380323756048,
+      "grad_norm": 0.5311306907505919,
+      "learning_rate": 1.5807327310901096e-05,
+      "loss": 0.338,
+      "step": 2121
+    },
+    {
+      "epoch": 1.0138581924616212,
+      "grad_norm": 0.5252428625482294,
+      "learning_rate": 1.580303890700513e-05,
+      "loss": 0.3215,
+      "step": 2122
+    },
+    {
+      "epoch": 1.0143360611671943,
+      "grad_norm": 0.5329298201469379,
+      "learning_rate": 1.579874889344854e-05,
+      "loss": 0.3231,
+      "step": 2123
+    },
+    {
+      "epoch": 1.0148139298727674,
+      "grad_norm": 0.8494224841188093,
+      "learning_rate": 1.579445727142129e-05,
+      "loss": 0.3209,
+      "step": 2124
+    },
+    {
+      "epoch": 1.0152917985783405,
+      "grad_norm": 0.575938313407013,
+      "learning_rate": 1.5790164042113805e-05,
+      "loss": 0.314,
+      "step": 2125
+    },
+    {
+      "epoch": 1.0157696672839138,
+      "grad_norm": 0.658256081712634,
+      "learning_rate": 1.5785869206716957e-05,
+      "loss": 0.3192,
+      "step": 2126
+    },
+    {
+      "epoch": 1.016247535989487,
+      "grad_norm": 0.7068653729035548,
+      "learning_rate": 1.578157276642205e-05,
+      "loss": 0.3306,
+      "step": 2127
+    },
+    {
+      "epoch": 1.01672540469506,
+      "grad_norm": 0.5411767755265546,
+      "learning_rate": 1.577727472242084e-05,
+      "loss": 0.3233,
+      "step": 2128
+    },
+    {
+      "epoch": 1.017203273400633,
+      "grad_norm": 0.6005936988154575,
+      "learning_rate": 1.577297507590553e-05,
+      "loss": 0.3284,
+      "step": 2129
+    },
+    {
+      "epoch": 1.0176811421062064,
+      "grad_norm": 0.5499614192701662,
+      "learning_rate": 1.576867382806877e-05,
+      "loss": 0.32,
+      "step": 2130
+    },
+    {
+      "epoch": 1.0181590108117795,
+      "grad_norm": 0.5390243829826628,
+      "learning_rate": 1.5764370980103652e-05,
+      "loss": 0.3355,
+      "step": 2131
+    },
+    {
+      "epoch": 1.0186368795173526,
+      "grad_norm": 0.5867439844608485,
+      "learning_rate": 1.57600665332037e-05,
+      "loss": 0.3323,
+      "step": 2132
+    },
+    {
+      "epoch": 1.0191147482229257,
+      "grad_norm": 0.567986991755531,
+      "learning_rate": 1.5755760488562898e-05,
+      "loss": 0.336,
+      "step": 2133
+    },
+    {
+      "epoch": 1.0195926169284988,
+      "grad_norm": 0.5275519852089237,
+      "learning_rate": 1.575145284737567e-05,
+      "loss": 0.3272,
+      "step": 2134
+    },
+    {
+      "epoch": 1.020070485634072,
+      "grad_norm": 0.5116807564883972,
+      "learning_rate": 1.5747143610836873e-05,
+      "loss": 0.3287,
+      "step": 2135
+    },
+    {
+      "epoch": 1.0205483543396452,
+      "grad_norm": 0.5309021872208404,
+      "learning_rate": 1.5742832780141816e-05,
+      "loss": 0.3517,
+      "step": 2136
+    },
+    {
+      "epoch": 1.0210262230452183,
+      "grad_norm": 0.5056039849696519,
+      "learning_rate": 1.573852035648625e-05,
+      "loss": 0.3317,
+      "step": 2137
+    },
+    {
+      "epoch": 1.0215040917507914,
+      "grad_norm": 0.5168796985969982,
+      "learning_rate": 1.5734206341066363e-05,
+      "loss": 0.3287,
+      "step": 2138
+    },
+    {
+      "epoch": 1.0219819604563647,
+      "grad_norm": 0.4971819258293254,
+      "learning_rate": 1.5729890735078782e-05,
+      "loss": 0.3465,
+      "step": 2139
+    },
+    {
+      "epoch": 1.0224598291619378,
+      "grad_norm": 0.5362308867880675,
+      "learning_rate": 1.5725573539720592e-05,
+      "loss": 0.3321,
+      "step": 2140
+    },
+    {
+      "epoch": 1.0229376978675109,
+      "grad_norm": 0.5409439946440039,
+      "learning_rate": 1.5721254756189293e-05,
+      "loss": 0.3385,
+      "step": 2141
+    },
+    {
+      "epoch": 1.023415566573084,
+      "grad_norm": 0.5435540704296682,
+      "learning_rate": 1.5716934385682847e-05,
+      "loss": 0.3117,
+      "step": 2142
+    },
+    {
+      "epoch": 1.0238934352786573,
+      "grad_norm": 0.5353171577649578,
+      "learning_rate": 1.5712612429399648e-05,
+      "loss": 0.3213,
+      "step": 2143
+    },
+    {
+      "epoch": 1.0243713039842304,
+      "grad_norm": 0.5572982587837408,
+      "learning_rate": 1.570828888853853e-05,
+      "loss": 0.325,
+      "step": 2144
+    },
+    {
+      "epoch": 1.0248491726898035,
+      "grad_norm": 0.5032264715461846,
+      "learning_rate": 1.570396376429877e-05,
+      "loss": 0.3253,
+      "step": 2145
+    },
+    {
+      "epoch": 1.0253270413953766,
+      "grad_norm": 0.5157482930173531,
+      "learning_rate": 1.569963705788007e-05,
+      "loss": 0.3426,
+      "step": 2146
+    },
+    {
+      "epoch": 1.0258049101009497,
+      "grad_norm": 0.5008263600349416,
+      "learning_rate": 1.56953087704826e-05,
+      "loss": 0.343,
+      "step": 2147
+    },
+    {
+      "epoch": 1.026282778806523,
+      "grad_norm": 0.5392376218710371,
+      "learning_rate": 1.5690978903306936e-05,
+      "loss": 0.333,
+      "step": 2148
+    },
+    {
+      "epoch": 1.026760647512096,
+      "grad_norm": 0.49909224648803696,
+      "learning_rate": 1.5686647457554108e-05,
+      "loss": 0.3223,
+      "step": 2149
+    },
+    {
+      "epoch": 1.0272385162176692,
+      "grad_norm": 0.5514337816765721,
+      "learning_rate": 1.5682314434425593e-05,
+      "loss": 0.3257,
+      "step": 2150
+    },
+    {
+      "epoch": 1.0277163849232422,
+      "grad_norm": 0.5560498032013678,
+      "learning_rate": 1.5677979835123282e-05,
+      "loss": 0.3172,
+      "step": 2151
+    },
+    {
+      "epoch": 1.0281942536288156,
+      "grad_norm": 0.497633056500836,
+      "learning_rate": 1.5673643660849525e-05,
+      "loss": 0.3472,
+      "step": 2152
+    },
+    {
+      "epoch": 1.0286721223343886,
+      "grad_norm": 0.5090101129547484,
+      "learning_rate": 1.5669305912807095e-05,
+      "loss": 0.3431,
+      "step": 2153
+    },
+    {
+      "epoch": 1.0291499910399617,
+      "grad_norm": 0.5115208530872307,
+      "learning_rate": 1.5664966592199213e-05,
+      "loss": 0.3179,
+      "step": 2154
+    },
+    {
+      "epoch": 1.0296278597455348,
+      "grad_norm": 0.4941887725586954,
+      "learning_rate": 1.5660625700229526e-05,
+      "loss": 0.3199,
+      "step": 2155
+    },
+    {
+      "epoch": 1.0301057284511081,
+      "grad_norm": 0.5051050314513967,
+      "learning_rate": 1.5656283238102125e-05,
+      "loss": 0.3503,
+      "step": 2156
+    },
+    {
+      "epoch": 1.0305835971566812,
+      "grad_norm": 0.47274872720440475,
+      "learning_rate": 1.5651939207021522e-05,
+      "loss": 0.3417,
+      "step": 2157
+    },
+    {
+      "epoch": 1.0310614658622543,
+      "grad_norm": 0.5026973690751055,
+      "learning_rate": 1.5647593608192685e-05,
+      "loss": 0.3307,
+      "step": 2158
+    },
+    {
+      "epoch": 1.0315393345678274,
+      "grad_norm": 0.506764397545058,
+      "learning_rate": 1.5643246442821004e-05,
+      "loss": 0.3446,
+      "step": 2159
+    },
+    {
+      "epoch": 1.0320172032734007,
+      "grad_norm": 0.5727967952694278,
+      "learning_rate": 1.5638897712112303e-05,
+      "loss": 0.3459,
+      "step": 2160
+    },
+    {
+      "epoch": 1.0324950719789738,
+      "grad_norm": 0.59780425526287,
+      "learning_rate": 1.5634547417272847e-05,
+      "loss": 0.3292,
+      "step": 2161
+    },
+    {
+      "epoch": 1.032972940684547,
+      "grad_norm": 0.5022372242667101,
+      "learning_rate": 1.5630195559509326e-05,
+      "loss": 0.3218,
+      "step": 2162
+    },
+    {
+      "epoch": 1.03345080939012,
+      "grad_norm": 0.5536484199445086,
+      "learning_rate": 1.562584214002887e-05,
+      "loss": 0.3037,
+      "step": 2163
+    },
+    {
+      "epoch": 1.033928678095693,
+      "grad_norm": 0.5211875586311786,
+      "learning_rate": 1.562148716003905e-05,
+      "loss": 0.347,
+      "step": 2164
+    },
+    {
+      "epoch": 1.0344065468012664,
+      "grad_norm": 0.5504469873275528,
+      "learning_rate": 1.561713062074785e-05,
+      "loss": 0.3189,
+      "step": 2165
+    },
+    {
+      "epoch": 1.0348844155068395,
+      "grad_norm": 0.5578520136556673,
+      "learning_rate": 1.56127725233637e-05,
+      "loss": 0.3397,
+      "step": 2166
+    },
+    {
+      "epoch": 1.0353622842124126,
+      "grad_norm": 0.547653966471447,
+      "learning_rate": 1.560841286909546e-05,
+      "loss": 0.3355,
+      "step": 2167
+    },
+    {
+      "epoch": 1.0358401529179857,
+      "grad_norm": 0.5470544888308998,
+      "learning_rate": 1.5604051659152418e-05,
+      "loss": 0.3302,
+      "step": 2168
+    },
+    {
+      "epoch": 1.036318021623559,
+      "grad_norm": 0.5805024891652473,
+      "learning_rate": 1.5599688894744304e-05,
+      "loss": 0.3335,
+      "step": 2169
+    },
+    {
+      "epoch": 1.036795890329132,
+      "grad_norm": 0.5136197878985662,
+      "learning_rate": 1.5595324577081265e-05,
+      "loss": 0.3295,
+      "step": 2170
+    },
+    {
+      "epoch": 1.0372737590347052,
+      "grad_norm": 0.56863395023198,
+      "learning_rate": 1.5590958707373886e-05,
+      "loss": 0.3175,
+      "step": 2171
+    },
+    {
+      "epoch": 1.0377516277402783,
+      "grad_norm": 0.5502967372343314,
+      "learning_rate": 1.558659128683319e-05,
+      "loss": 0.337,
+      "step": 2172
+    },
+    {
+      "epoch": 1.0382294964458516,
+      "grad_norm": 0.5135053364771395,
+      "learning_rate": 1.558222231667061e-05,
+      "loss": 0.3324,
+      "step": 2173
+    },
+    {
+      "epoch": 1.0387073651514247,
+      "grad_norm": 0.537644706467191,
+      "learning_rate": 1.5577851798098032e-05,
+      "loss": 0.3261,
+      "step": 2174
+    },
+    {
+      "epoch": 1.0391852338569978,
+      "grad_norm": 0.5382510505587826,
+      "learning_rate": 1.5573479732327758e-05,
+      "loss": 0.338,
+      "step": 2175
+    },
+    {
+      "epoch": 1.0396631025625709,
+      "grad_norm": 0.49204890495941894,
+      "learning_rate": 1.556910612057252e-05,
+      "loss": 0.3477,
+      "step": 2176
+    },
+    {
+      "epoch": 1.040140971268144,
+      "grad_norm": 0.523151317823713,
+      "learning_rate": 1.5564730964045476e-05,
+      "loss": 0.352,
+      "step": 2177
+    },
+    {
+      "epoch": 1.0406188399737173,
+      "grad_norm": 0.4894794188961839,
+      "learning_rate": 1.556035426396023e-05,
+      "loss": 0.3135,
+      "step": 2178
+    },
+    {
+      "epoch": 1.0410967086792904,
+      "grad_norm": 0.6266176917586536,
+      "learning_rate": 1.555597602153079e-05,
+      "loss": 0.3143,
+      "step": 2179
+    },
+    {
+      "epoch": 1.0415745773848635,
+      "grad_norm": 0.5309556876141732,
+      "learning_rate": 1.555159623797161e-05,
+      "loss": 0.3403,
+      "step": 2180
+    },
+    {
+      "epoch": 1.0420524460904366,
+      "grad_norm": 0.4931662085144718,
+      "learning_rate": 1.554721491449756e-05,
+      "loss": 0.3397,
+      "step": 2181
+    },
+    {
+      "epoch": 1.0425303147960099,
+      "grad_norm": 0.5645039752023514,
+      "learning_rate": 1.5542832052323943e-05,
+      "loss": 0.3396,
+      "step": 2182
+    },
+    {
+      "epoch": 1.043008183501583,
+      "grad_norm": 0.49990356604600616,
+      "learning_rate": 1.553844765266649e-05,
+      "loss": 0.339,
+      "step": 2183
+    },
+    {
+      "epoch": 1.043486052207156,
+      "grad_norm": 0.5468317171758479,
+      "learning_rate": 1.5534061716741358e-05,
+      "loss": 0.3219,
+      "step": 2184
+    },
+    {
+      "epoch": 1.0439639209127292,
+      "grad_norm": 1.455991273998016,
+      "learning_rate": 1.552967424576512e-05,
+      "loss": 0.3444,
+      "step": 2185
+    },
+    {
+      "epoch": 1.0444417896183025,
+      "grad_norm": 0.5138914378146582,
+      "learning_rate": 1.5525285240954793e-05,
+      "loss": 0.3437,
+      "step": 2186
+    },
+    {
+      "epoch": 1.0449196583238756,
+      "grad_norm": 0.4771708845133923,
+      "learning_rate": 1.55208947035278e-05,
+      "loss": 0.3401,
+      "step": 2187
+    },
+    {
+      "epoch": 1.0453975270294487,
+      "grad_norm": 17.211331468923103,
+      "learning_rate": 1.5516502634702003e-05,
+      "loss": 0.3233,
+      "step": 2188
+    },
+    {
+      "epoch": 1.0458753957350218,
+      "grad_norm": 0.6438223472017501,
+      "learning_rate": 1.5512109035695688e-05,
+      "loss": 0.3215,
+      "step": 2189
+    },
+    {
+      "epoch": 1.0463532644405948,
+      "grad_norm": 0.5071979054947092,
+      "learning_rate": 1.5507713907727557e-05,
+      "loss": 0.3381,
+      "step": 2190
+    },
+    {
+      "epoch": 1.0468311331461682,
+      "grad_norm": 0.5360652181387802,
+      "learning_rate": 1.550331725201674e-05,
+      "loss": 0.3309,
+      "step": 2191
+    },
+    {
+      "epoch": 1.0473090018517413,
+      "grad_norm": 0.5811188677891611,
+      "learning_rate": 1.54989190697828e-05,
+      "loss": 0.3271,
+      "step": 2192
+    },
+    {
+      "epoch": 1.0477868705573143,
+      "grad_norm": 0.49109998644777375,
+      "learning_rate": 1.5494519362245702e-05,
+      "loss": 0.3217,
+      "step": 2193
+    },
+    {
+      "epoch": 1.0482647392628874,
+      "grad_norm": 0.5249085574500518,
+      "learning_rate": 1.549011813062586e-05,
+      "loss": 0.3363,
+      "step": 2194
+    },
+    {
+      "epoch": 1.0487426079684607,
+      "grad_norm": 0.5450121620506873,
+      "learning_rate": 1.5485715376144087e-05,
+      "loss": 0.3368,
+      "step": 2195
+    },
+    {
+      "epoch": 1.0492204766740338,
+      "grad_norm": 0.5333106429783062,
+      "learning_rate": 1.5481311100021642e-05,
+      "loss": 0.3482,
+      "step": 2196
+    },
+    {
+      "epoch": 1.049698345379607,
+      "grad_norm": 0.5132325689081423,
+      "learning_rate": 1.5476905303480183e-05,
+      "loss": 0.3235,
+      "step": 2197
+    },
+    {
+      "epoch": 1.05017621408518,
+      "grad_norm": 0.5603270076262434,
+      "learning_rate": 1.5472497987741803e-05,
+      "loss": 0.3245,
+      "step": 2198
+    },
+    {
+      "epoch": 1.0506540827907533,
+      "grad_norm": 0.5083807526273509,
+      "learning_rate": 1.5468089154029016e-05,
+      "loss": 0.3382,
+      "step": 2199
+    },
+    {
+      "epoch": 1.0511319514963264,
+      "grad_norm": 0.5338317283874571,
+      "learning_rate": 1.5463678803564753e-05,
+      "loss": 0.3225,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0516098202018995,
+      "grad_norm": 0.5268204769201611,
+      "learning_rate": 1.5459266937572367e-05,
+      "loss": 0.3419,
+      "step": 2201
+    },
+    {
+      "epoch": 1.0520876889074726,
+      "grad_norm": 0.4795433149023876,
+      "learning_rate": 1.5454853557275632e-05,
+      "loss": 0.3464,
+      "step": 2202
+    },
+    {
+      "epoch": 1.0525655576130457,
+      "grad_norm": 0.5808022832179441,
+      "learning_rate": 1.5450438663898743e-05,
+      "loss": 0.3177,
+      "step": 2203
+    },
+    {
+      "epoch": 1.053043426318619,
+      "grad_norm": 0.5145172026904812,
+      "learning_rate": 1.5446022258666313e-05,
+      "loss": 0.3361,
+      "step": 2204
+    },
+    {
+      "epoch": 1.0535212950241921,
+      "grad_norm": 0.47386368364154696,
+      "learning_rate": 1.5441604342803374e-05,
+      "loss": 0.3188,
+      "step": 2205
+    },
+    {
+      "epoch": 1.0539991637297652,
+      "grad_norm": 0.6038585929683551,
+      "learning_rate": 1.5437184917535377e-05,
+      "loss": 0.3415,
+      "step": 2206
+    },
+    {
+      "epoch": 1.0544770324353383,
+      "grad_norm": 0.51095983032144,
+      "learning_rate": 1.5432763984088195e-05,
+      "loss": 0.328,
+      "step": 2207
+    },
+    {
+      "epoch": 1.0549549011409116,
+      "grad_norm": 0.49778348332654354,
+      "learning_rate": 1.5428341543688116e-05,
+      "loss": 0.3249,
+      "step": 2208
+    },
+    {
+      "epoch": 1.0554327698464847,
+      "grad_norm": 0.5813291087757343,
+      "learning_rate": 1.542391759756185e-05,
+      "loss": 0.3324,
+      "step": 2209
+    },
+    {
+      "epoch": 1.0559106385520578,
+      "grad_norm": 0.5194247870224709,
+      "learning_rate": 1.5419492146936518e-05,
+      "loss": 0.3412,
+      "step": 2210
+    },
+    {
+      "epoch": 1.056388507257631,
+      "grad_norm": 0.49251753974846363,
+      "learning_rate": 1.5415065193039658e-05,
+      "loss": 0.3357,
+      "step": 2211
+    },
+    {
+      "epoch": 1.0568663759632042,
+      "grad_norm": 0.5290839901414063,
+      "learning_rate": 1.5410636737099238e-05,
+      "loss": 0.327,
+      "step": 2212
+    },
+    {
+      "epoch": 1.0573442446687773,
+      "grad_norm": 0.5447923991162895,
+      "learning_rate": 1.5406206780343626e-05,
+      "loss": 0.3332,
+      "step": 2213
+    },
+    {
+      "epoch": 1.0578221133743504,
+      "grad_norm": 0.563508010318956,
+      "learning_rate": 1.540177532400162e-05,
+      "loss": 0.3398,
+      "step": 2214
+    },
+    {
+      "epoch": 1.0582999820799235,
+      "grad_norm": 0.5608591424899783,
+      "learning_rate": 1.5397342369302425e-05,
+      "loss": 0.3121,
+      "step": 2215
+    },
+    {
+      "epoch": 1.0587778507854968,
+      "grad_norm": 0.5390541825320886,
+      "learning_rate": 1.5392907917475662e-05,
+      "loss": 0.3263,
+      "step": 2216
+    },
+    {
+      "epoch": 1.05925571949107,
+      "grad_norm": 0.6375454291641146,
+      "learning_rate": 1.5388471969751373e-05,
+      "loss": 0.3286,
+      "step": 2217
+    },
+    {
+      "epoch": 1.059733588196643,
+      "grad_norm": 0.6197563874990852,
+      "learning_rate": 1.538403452736001e-05,
+      "loss": 0.3375,
+      "step": 2218
+    },
+    {
+      "epoch": 1.060211456902216,
+      "grad_norm": 0.6485799859247667,
+      "learning_rate": 1.5379595591532442e-05,
+      "loss": 0.3498,
+      "step": 2219
+    },
+    {
+      "epoch": 1.0606893256077892,
+      "grad_norm": 0.5016640951997636,
+      "learning_rate": 1.5375155163499953e-05,
+      "loss": 0.3103,
+      "step": 2220
+    },
+    {
+      "epoch": 1.0611671943133625,
+      "grad_norm": 0.505528792203374,
+      "learning_rate": 1.5370713244494235e-05,
+      "loss": 0.3398,
+      "step": 2221
+    },
+    {
+      "epoch": 1.0616450630189356,
+      "grad_norm": 0.6607061590143123,
+      "learning_rate": 1.53662698357474e-05,
+      "loss": 0.3382,
+      "step": 2222
+    },
+    {
+      "epoch": 1.0621229317245087,
+      "grad_norm": 0.5195925433144004,
+      "learning_rate": 1.536182493849198e-05,
+      "loss": 0.3231,
+      "step": 2223
+    },
+    {
+      "epoch": 1.0626008004300818,
+      "grad_norm": 0.5789827056376714,
+      "learning_rate": 1.5357378553960892e-05,
+      "loss": 0.3313,
+      "step": 2224
+    },
+    {
+      "epoch": 1.063078669135655,
+      "grad_norm": 0.5365461420949971,
+      "learning_rate": 1.5352930683387502e-05,
+      "loss": 0.3299,
+      "step": 2225
+    },
+    {
+      "epoch": 1.0635565378412282,
+      "grad_norm": 0.512379234100911,
+      "learning_rate": 1.5348481328005566e-05,
+      "loss": 0.321,
+      "step": 2226
+    },
+    {
+      "epoch": 1.0640344065468013,
+      "grad_norm": 0.5335894256345767,
+      "learning_rate": 1.534403048904925e-05,
+      "loss": 0.3451,
+      "step": 2227
+    },
+    {
+      "epoch": 1.0645122752523744,
+      "grad_norm": 0.5464542705568193,
+      "learning_rate": 1.5339578167753145e-05,
+      "loss": 0.3389,
+      "step": 2228
+    },
+    {
+      "epoch": 1.0649901439579477,
+      "grad_norm": 0.5232741311541103,
+      "learning_rate": 1.5335124365352246e-05,
+      "loss": 0.3342,
+      "step": 2229
+    },
+    {
+      "epoch": 1.0654680126635208,
+      "grad_norm": 0.5900396821697615,
+      "learning_rate": 1.5330669083081956e-05,
+      "loss": 0.3391,
+      "step": 2230
+    },
+    {
+      "epoch": 1.0659458813690939,
+      "grad_norm": 0.5093924605652755,
+      "learning_rate": 1.5326212322178097e-05,
+      "loss": 0.3286,
+      "step": 2231
+    },
+    {
+      "epoch": 1.066423750074667,
+      "grad_norm": 0.5716841100121078,
+      "learning_rate": 1.5321754083876893e-05,
+      "loss": 0.323,
+      "step": 2232
+    },
+    {
+      "epoch": 1.06690161878024,
+      "grad_norm": 0.4842785242623549,
+      "learning_rate": 1.5317294369414975e-05,
+      "loss": 0.3408,
+      "step": 2233
+    },
+    {
+      "epoch": 1.0673794874858133,
+      "grad_norm": 0.513994095767651,
+      "learning_rate": 1.53128331800294e-05,
+      "loss": 0.3466,
+      "step": 2234
+    },
+    {
+      "epoch": 1.0678573561913864,
+      "grad_norm": 0.5758288986174495,
+      "learning_rate": 1.5308370516957617e-05,
+      "loss": 0.3217,
+      "step": 2235
+    },
+    {
+      "epoch": 1.0683352248969595,
+      "grad_norm": 0.5200971785900195,
+      "learning_rate": 1.5303906381437487e-05,
+      "loss": 0.3251,
+      "step": 2236
+    },
+    {
+      "epoch": 1.0688130936025326,
+      "grad_norm": 0.5033891597260092,
+      "learning_rate": 1.529944077470729e-05,
+      "loss": 0.3236,
+      "step": 2237
+    },
+    {
+      "epoch": 1.069290962308106,
+      "grad_norm": 0.49510942051055734,
+      "learning_rate": 1.52949736980057e-05,
+      "loss": 0.3232,
+      "step": 2238
+    },
+    {
+      "epoch": 1.069768831013679,
+      "grad_norm": 0.49918931484754825,
+      "learning_rate": 1.529050515257181e-05,
+      "loss": 0.3292,
+      "step": 2239
+    },
+    {
+      "epoch": 1.0702466997192521,
+      "grad_norm": 0.5263649316343764,
+      "learning_rate": 1.528603513964511e-05,
+      "loss": 0.3186,
+      "step": 2240
+    },
+    {
+      "epoch": 1.0707245684248252,
+      "grad_norm": 0.506050997373646,
+      "learning_rate": 1.528156366046551e-05,
+      "loss": 0.3386,
+      "step": 2241
+    },
+    {
+      "epoch": 1.0712024371303985,
+      "grad_norm": 0.4759187913918262,
+      "learning_rate": 1.5277090716273313e-05,
+      "loss": 0.3244,
+      "step": 2242
+    },
+    {
+      "epoch": 1.0716803058359716,
+      "grad_norm": 0.5170077950073158,
+      "learning_rate": 1.527261630830924e-05,
+      "loss": 0.34,
+      "step": 2243
+    },
+    {
+      "epoch": 1.0721581745415447,
+      "grad_norm": 0.5371316049130486,
+      "learning_rate": 1.526814043781441e-05,
+      "loss": 0.3139,
+      "step": 2244
+    },
+    {
+      "epoch": 1.0726360432471178,
+      "grad_norm": 0.4661135714534529,
+      "learning_rate": 1.5263663106030347e-05,
+      "loss": 0.3431,
+      "step": 2245
+    },
+    {
+      "epoch": 1.073113911952691,
+      "grad_norm": 0.5050579846579514,
+      "learning_rate": 1.5259184314198995e-05,
+      "loss": 0.3182,
+      "step": 2246
+    },
+    {
+      "epoch": 1.0735917806582642,
+      "grad_norm": 0.49161171778636,
+      "learning_rate": 1.5254704063562678e-05,
+      "loss": 0.3662,
+      "step": 2247
+    },
+    {
+      "epoch": 1.0740696493638373,
+      "grad_norm": 0.6153694810824738,
+      "learning_rate": 1.5250222355364149e-05,
+      "loss": 0.3144,
+      "step": 2248
+    },
+    {
+      "epoch": 1.0745475180694104,
+      "grad_norm": 0.4888202707715523,
+      "learning_rate": 1.5245739190846549e-05,
+      "loss": 0.3398,
+      "step": 2249
+    },
+    {
+      "epoch": 1.0750253867749835,
+      "grad_norm": 0.5124355341785777,
+      "learning_rate": 1.5241254571253433e-05,
+      "loss": 0.3548,
+      "step": 2250
+    },
+    {
+      "epoch": 1.0755032554805568,
+      "grad_norm": 0.5077616262970712,
+      "learning_rate": 1.5236768497828753e-05,
+      "loss": 0.3502,
+      "step": 2251
+    },
+    {
+      "epoch": 1.07598112418613,
+      "grad_norm": 0.53688126358671,
+      "learning_rate": 1.5232280971816864e-05,
+      "loss": 0.3308,
+      "step": 2252
+    },
+    {
+      "epoch": 1.076458992891703,
+      "grad_norm": 0.49623504126840645,
+      "learning_rate": 1.5227791994462529e-05,
+      "loss": 0.3179,
+      "step": 2253
+    },
+    {
+      "epoch": 1.076936861597276,
+      "grad_norm": 0.5575295987434887,
+      "learning_rate": 1.5223301567010916e-05,
+      "loss": 0.321,
+      "step": 2254
+    },
+    {
+      "epoch": 1.0774147303028494,
+      "grad_norm": 0.5298944158079708,
+      "learning_rate": 1.5218809690707583e-05,
+      "loss": 0.3341,
+      "step": 2255
+    },
+    {
+      "epoch": 1.0778925990084225,
+      "grad_norm": 0.5077709934710473,
+      "learning_rate": 1.5214316366798498e-05,
+      "loss": 0.3521,
+      "step": 2256
+    },
+    {
+      "epoch": 1.0783704677139956,
+      "grad_norm": 0.5439955998090592,
+      "learning_rate": 1.5209821596530035e-05,
+      "loss": 0.3184,
+      "step": 2257
+    },
+    {
+      "epoch": 1.0788483364195687,
+      "grad_norm": 0.5084561135875706,
+      "learning_rate": 1.5205325381148958e-05,
+      "loss": 0.345,
+      "step": 2258
+    },
+    {
+      "epoch": 1.0793262051251418,
+      "grad_norm": 0.5025281601561041,
+      "learning_rate": 1.5200827721902443e-05,
+      "loss": 0.3134,
+      "step": 2259
+    },
+    {
+      "epoch": 1.079804073830715,
+      "grad_norm": 0.5229250055944893,
+      "learning_rate": 1.5196328620038059e-05,
+      "loss": 0.3437,
+      "step": 2260
+    },
+    {
+      "epoch": 1.0802819425362882,
+      "grad_norm": 0.5010024267654524,
+      "learning_rate": 1.5191828076803776e-05,
+      "loss": 0.3161,
+      "step": 2261
+    },
+    {
+      "epoch": 1.0807598112418613,
+      "grad_norm": 0.5752421274767535,
+      "learning_rate": 1.5187326093447965e-05,
+      "loss": 0.346,
+      "step": 2262
+    },
+    {
+      "epoch": 1.0812376799474344,
+      "grad_norm": 0.5513552600942755,
+      "learning_rate": 1.5182822671219404e-05,
+      "loss": 0.3368,
+      "step": 2263
+    },
+    {
+      "epoch": 1.0817155486530077,
+      "grad_norm": 0.49596162792066967,
+      "learning_rate": 1.5178317811367254e-05,
+      "loss": 0.325,
+      "step": 2264
+    },
+    {
+      "epoch": 1.0821934173585808,
+      "grad_norm": 0.5479700565884935,
+      "learning_rate": 1.5173811515141083e-05,
+      "loss": 0.3346,
+      "step": 2265
+    },
+    {
+      "epoch": 1.0826712860641539,
+      "grad_norm": 0.6256543187747687,
+      "learning_rate": 1.516930378379087e-05,
+      "loss": 0.3251,
+      "step": 2266
+    },
+    {
+      "epoch": 1.083149154769727,
+      "grad_norm": 0.5689155161140599,
+      "learning_rate": 1.516479461856697e-05,
+      "loss": 0.3363,
+      "step": 2267
+    },
+    {
+      "epoch": 1.0836270234753003,
+      "grad_norm": 0.7023942415790342,
+      "learning_rate": 1.5160284020720144e-05,
+      "loss": 0.3256,
+      "step": 2268
+    },
+    {
+      "epoch": 1.0841048921808734,
+      "grad_norm": 0.4944938967568079,
+      "learning_rate": 1.5155771991501562e-05,
+      "loss": 0.3397,
+      "step": 2269
+    },
+    {
+      "epoch": 1.0845827608864465,
+      "grad_norm": 0.7545407502316881,
+      "learning_rate": 1.5151258532162771e-05,
+      "loss": 0.3239,
+      "step": 2270
+    },
+    {
+      "epoch": 1.0850606295920195,
+      "grad_norm": 0.5318090108055559,
+      "learning_rate": 1.5146743643955732e-05,
+      "loss": 0.3243,
+      "step": 2271
+    },
+    {
+      "epoch": 1.0855384982975926,
+      "grad_norm": 0.5221147872004993,
+      "learning_rate": 1.5142227328132797e-05,
+      "loss": 0.3364,
+      "step": 2272
+    },
+    {
+      "epoch": 1.086016367003166,
+      "grad_norm": 0.5704820936647907,
+      "learning_rate": 1.5137709585946705e-05,
+      "loss": 0.3423,
+      "step": 2273
+    },
+    {
+      "epoch": 1.086494235708739,
+      "grad_norm": 0.5013932264970669,
+      "learning_rate": 1.51331904186506e-05,
+      "loss": 0.3258,
+      "step": 2274
+    },
+    {
+      "epoch": 1.0869721044143121,
+      "grad_norm": 0.5257762249096876,
+      "learning_rate": 1.5128669827498024e-05,
+      "loss": 0.3246,
+      "step": 2275
+    },
+    {
+      "epoch": 1.0874499731198852,
+      "grad_norm": 0.5139561282311831,
+      "learning_rate": 1.5124147813742904e-05,
+      "loss": 0.3293,
+      "step": 2276
+    },
+    {
+      "epoch": 1.0879278418254585,
+      "grad_norm": 0.543726884049005,
+      "learning_rate": 1.5119624378639568e-05,
+      "loss": 0.3555,
+      "step": 2277
+    },
+    {
+      "epoch": 1.0884057105310316,
+      "grad_norm": 0.512178392507077,
+      "learning_rate": 1.511509952344274e-05,
+      "loss": 0.3254,
+      "step": 2278
+    },
+    {
+      "epoch": 1.0888835792366047,
+      "grad_norm": 0.5176231019242892,
+      "learning_rate": 1.511057324940753e-05,
+      "loss": 0.3253,
+      "step": 2279
+    },
+    {
+      "epoch": 1.0893614479421778,
+      "grad_norm": 0.5278862653717663,
+      "learning_rate": 1.5106045557789453e-05,
+      "loss": 0.3346,
+      "step": 2280
+    },
+    {
+      "epoch": 1.0898393166477511,
+      "grad_norm": 0.47608912937489073,
+      "learning_rate": 1.5101516449844407e-05,
+      "loss": 0.3325,
+      "step": 2281
+    },
+    {
+      "epoch": 1.0903171853533242,
+      "grad_norm": 0.5309476770511793,
+      "learning_rate": 1.5096985926828684e-05,
+      "loss": 0.351,
+      "step": 2282
+    },
+    {
+      "epoch": 1.0907950540588973,
+      "grad_norm": 0.5133648196697228,
+      "learning_rate": 1.5092453989998976e-05,
+      "loss": 0.3369,
+      "step": 2283
+    },
+    {
+      "epoch": 1.0912729227644704,
+      "grad_norm": 0.5748784259519858,
+      "learning_rate": 1.5087920640612361e-05,
+      "loss": 0.342,
+      "step": 2284
+    },
+    {
+      "epoch": 1.0917507914700435,
+      "grad_norm": 0.51205359618062,
+      "learning_rate": 1.5083385879926309e-05,
+      "loss": 0.3304,
+      "step": 2285
+    },
+    {
+      "epoch": 1.0922286601756168,
+      "grad_norm": 0.5640393858567604,
+      "learning_rate": 1.5078849709198687e-05,
+      "loss": 0.3529,
+      "step": 2286
+    },
+    {
+      "epoch": 1.09270652888119,
+      "grad_norm": 0.5141124022477992,
+      "learning_rate": 1.5074312129687741e-05,
+      "loss": 0.3425,
+      "step": 2287
+    },
+    {
+      "epoch": 1.093184397586763,
+      "grad_norm": 0.5187378970830254,
+      "learning_rate": 1.5069773142652119e-05,
+      "loss": 0.3378,
+      "step": 2288
+    },
+    {
+      "epoch": 1.093662266292336,
+      "grad_norm": 0.520566150228239,
+      "learning_rate": 1.506523274935086e-05,
+      "loss": 0.3461,
+      "step": 2289
+    },
+    {
+      "epoch": 1.0941401349979094,
+      "grad_norm": 0.5296392819928428,
+      "learning_rate": 1.5060690951043385e-05,
+      "loss": 0.3417,
+      "step": 2290
+    },
+    {
+      "epoch": 1.0946180037034825,
+      "grad_norm": 0.5195069878439671,
+      "learning_rate": 1.5056147748989505e-05,
+      "loss": 0.3234,
+      "step": 2291
+    },
+    {
+      "epoch": 1.0950958724090556,
+      "grad_norm": 0.5634053365773759,
+      "learning_rate": 1.5051603144449431e-05,
+      "loss": 0.3476,
+      "step": 2292
+    },
+    {
+      "epoch": 1.0955737411146287,
+      "grad_norm": 0.5457090574500167,
+      "learning_rate": 1.5047057138683753e-05,
+      "loss": 0.3214,
+      "step": 2293
+    },
+    {
+      "epoch": 1.096051609820202,
+      "grad_norm": 0.49911077929816133,
+      "learning_rate": 1.5042509732953454e-05,
+      "loss": 0.3414,
+      "step": 2294
+    },
+    {
+      "epoch": 1.096529478525775,
+      "grad_norm": 2.702449341602149,
+      "learning_rate": 1.5037960928519902e-05,
+      "loss": 0.3377,
+      "step": 2295
+    },
+    {
+      "epoch": 1.0970073472313482,
+      "grad_norm": 0.5565135853631065,
+      "learning_rate": 1.5033410726644859e-05,
+      "loss": 0.3343,
+      "step": 2296
+    },
+    {
+      "epoch": 1.0974852159369213,
+      "grad_norm": 0.5073307379974941,
+      "learning_rate": 1.5028859128590468e-05,
+      "loss": 0.327,
+      "step": 2297
+    },
+    {
+      "epoch": 1.0979630846424944,
+      "grad_norm": 0.5160051255941388,
+      "learning_rate": 1.502430613561926e-05,
+      "loss": 0.344,
+      "step": 2298
+    },
+    {
+      "epoch": 1.0984409533480677,
+      "grad_norm": 0.5348498573334795,
+      "learning_rate": 1.5019751748994158e-05,
+      "loss": 0.3494,
+      "step": 2299
+    },
+    {
+      "epoch": 1.0989188220536408,
+      "grad_norm": 0.4865115046264663,
+      "learning_rate": 1.501519596997847e-05,
+      "loss": 0.3435,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0993966907592139,
+      "grad_norm": 0.5216309305575126,
+      "learning_rate": 1.5010638799835884e-05,
+      "loss": 0.33,
+      "step": 2301
+    },
+    {
+      "epoch": 1.099874559464787,
+      "grad_norm": 0.539223989848975,
+      "learning_rate": 1.5006080239830483e-05,
+      "loss": 0.3361,
+      "step": 2302
+    },
+    {
+      "epoch": 1.1003524281703603,
+      "grad_norm": 0.5001901728249805,
+      "learning_rate": 1.5001520291226727e-05,
+      "loss": 0.3236,
+      "step": 2303
+    },
+    {
+      "epoch": 1.1008302968759334,
+      "grad_norm": 0.515833509744522,
+      "learning_rate": 1.499695895528947e-05,
+      "loss": 0.3476,
+      "step": 2304
+    },
+    {
+      "epoch": 1.1013081655815065,
+      "grad_norm": 0.5136798035818348,
+      "learning_rate": 1.499239623328394e-05,
+      "loss": 0.346,
+      "step": 2305
+    },
+    {
+      "epoch": 1.1017860342870796,
+      "grad_norm": 0.5181371149654749,
+      "learning_rate": 1.4987832126475763e-05,
+      "loss": 0.3427,
+      "step": 2306
+    },
+    {
+      "epoch": 1.1022639029926529,
+      "grad_norm": 0.5071439491474284,
+      "learning_rate": 1.4983266636130935e-05,
+      "loss": 0.3404,
+      "step": 2307
+    },
+    {
+      "epoch": 1.102741771698226,
+      "grad_norm": 0.5024865933279972,
+      "learning_rate": 1.4978699763515848e-05,
+      "loss": 0.3408,
+      "step": 2308
+    },
+    {
+      "epoch": 1.103219640403799,
+      "grad_norm": 0.6901876193242537,
+      "learning_rate": 1.4974131509897269e-05,
+      "loss": 0.3246,
+      "step": 2309
+    },
+    {
+      "epoch": 1.1036975091093721,
+      "grad_norm": 0.5498690262019805,
+      "learning_rate": 1.4969561876542348e-05,
+      "loss": 0.339,
+      "step": 2310
+    },
+    {
+      "epoch": 1.1041753778149452,
+      "grad_norm": 0.5305378391568346,
+      "learning_rate": 1.4964990864718627e-05,
+      "loss": 0.324,
+      "step": 2311
+    },
+    {
+      "epoch": 1.1046532465205186,
+      "grad_norm": 0.6920996981731619,
+      "learning_rate": 1.496041847569402e-05,
+      "loss": 0.327,
+      "step": 2312
+    },
+    {
+      "epoch": 1.1051311152260916,
+      "grad_norm": 0.5051871088425068,
+      "learning_rate": 1.4955844710736829e-05,
+      "loss": 0.324,
+      "step": 2313
+    },
+    {
+      "epoch": 1.1056089839316647,
+      "grad_norm": 0.5231668443052977,
+      "learning_rate": 1.4951269571115735e-05,
+      "loss": 0.3192,
+      "step": 2314
+    },
+    {
+      "epoch": 1.1060868526372378,
+      "grad_norm": 0.5107337355608277,
+      "learning_rate": 1.4946693058099802e-05,
+      "loss": 0.3187,
+      "step": 2315
+    },
+    {
+      "epoch": 1.1065647213428111,
+      "grad_norm": 0.504994234754607,
+      "learning_rate": 1.494211517295847e-05,
+      "loss": 0.3274,
+      "step": 2316
+    },
+    {
+      "epoch": 1.1070425900483842,
+      "grad_norm": 0.6160822307451901,
+      "learning_rate": 1.4937535916961568e-05,
+      "loss": 0.3296,
+      "step": 2317
+    },
+    {
+      "epoch": 1.1075204587539573,
+      "grad_norm": 0.4975876476425877,
+      "learning_rate": 1.49329552913793e-05,
+      "loss": 0.3389,
+      "step": 2318
+    },
+    {
+      "epoch": 1.1079983274595304,
+      "grad_norm": 0.5403670488971715,
+      "learning_rate": 1.4928373297482249e-05,
+      "loss": 0.3455,
+      "step": 2319
+    },
+    {
+      "epoch": 1.1084761961651037,
+      "grad_norm": 0.529726563264269,
+      "learning_rate": 1.4923789936541378e-05,
+      "loss": 0.3406,
+      "step": 2320
+    },
+    {
+      "epoch": 1.1089540648706768,
+      "grad_norm": 2.35075832261799,
+      "learning_rate": 1.4919205209828037e-05,
+      "loss": 0.3349,
+      "step": 2321
+    },
+    {
+      "epoch": 1.10943193357625,
+      "grad_norm": 0.5425366907719269,
+      "learning_rate": 1.4914619118613942e-05,
+      "loss": 0.3317,
+      "step": 2322
+    },
+    {
+      "epoch": 1.109909802281823,
+      "grad_norm": 0.5093459744924752,
+      "learning_rate": 1.4910031664171195e-05,
+      "loss": 0.3486,
+      "step": 2323
+    },
+    {
+      "epoch": 1.110387670987396,
+      "grad_norm": 0.5160824382283055,
+      "learning_rate": 1.4905442847772278e-05,
+      "loss": 0.3339,
+      "step": 2324
+    },
+    {
+      "epoch": 1.1108655396929694,
+      "grad_norm": 0.5457445322763627,
+      "learning_rate": 1.4900852670690044e-05,
+      "loss": 0.3391,
+      "step": 2325
+    },
+    {
+      "epoch": 1.1113434083985425,
+      "grad_norm": 0.5181476585227104,
+      "learning_rate": 1.4896261134197729e-05,
+      "loss": 0.3454,
+      "step": 2326
+    },
+    {
+      "epoch": 1.1118212771041156,
+      "grad_norm": 0.5151616532814238,
+      "learning_rate": 1.4891668239568943e-05,
+      "loss": 0.3323,
+      "step": 2327
+    },
+    {
+      "epoch": 1.1122991458096887,
+      "grad_norm": 0.48685718879651513,
+      "learning_rate": 1.4887073988077678e-05,
+      "loss": 0.3325,
+      "step": 2328
+    },
+    {
+      "epoch": 1.112777014515262,
+      "grad_norm": 0.6039961072105209,
+      "learning_rate": 1.4882478380998291e-05,
+      "loss": 0.3354,
+      "step": 2329
+    },
+    {
+      "epoch": 1.113254883220835,
+      "grad_norm": 0.49114939066967817,
+      "learning_rate": 1.487788141960553e-05,
+      "loss": 0.3262,
+      "step": 2330
+    },
+    {
+      "epoch": 1.1137327519264082,
+      "grad_norm": 0.46862190557531513,
+      "learning_rate": 1.4873283105174504e-05,
+      "loss": 0.3366,
+      "step": 2331
+    },
+    {
+      "epoch": 1.1142106206319813,
+      "grad_norm": 0.47383978484154615,
+      "learning_rate": 1.4868683438980714e-05,
+      "loss": 0.3241,
+      "step": 2332
+    },
+    {
+      "epoch": 1.1146884893375546,
+      "grad_norm": 0.48438365346141576,
+      "learning_rate": 1.4864082422300015e-05,
+      "loss": 0.3239,
+      "step": 2333
+    },
+    {
+      "epoch": 1.1151663580431277,
+      "grad_norm": 0.49902567827730593,
+      "learning_rate": 1.4859480056408653e-05,
+      "loss": 0.3343,
+      "step": 2334
+    },
+    {
+      "epoch": 1.1156442267487008,
+      "grad_norm": 0.5240825256150454,
+      "learning_rate": 1.4854876342583246e-05,
+      "loss": 0.3525,
+      "step": 2335
+    },
+    {
+      "epoch": 1.1161220954542739,
+      "grad_norm": 0.6398611930145781,
+      "learning_rate": 1.4850271282100779e-05,
+      "loss": 0.3283,
+      "step": 2336
+    },
+    {
+      "epoch": 1.116599964159847,
+      "grad_norm": 0.5213749367846112,
+      "learning_rate": 1.4845664876238615e-05,
+      "loss": 0.3461,
+      "step": 2337
+    },
+    {
+      "epoch": 1.1170778328654203,
+      "grad_norm": 0.5017747641723227,
+      "learning_rate": 1.4841057126274491e-05,
+      "loss": 0.3476,
+      "step": 2338
+    },
+    {
+      "epoch": 1.1175557015709934,
+      "grad_norm": 0.5121891657183757,
+      "learning_rate": 1.4836448033486513e-05,
+      "loss": 0.3116,
+      "step": 2339
+    },
+    {
+      "epoch": 1.1180335702765665,
+      "grad_norm": 0.490973436785604,
+      "learning_rate": 1.4831837599153165e-05,
+      "loss": 0.3479,
+      "step": 2340
+    },
+    {
+      "epoch": 1.1185114389821396,
+      "grad_norm": 0.4995891989120494,
+      "learning_rate": 1.4827225824553302e-05,
+      "loss": 0.3216,
+      "step": 2341
+    },
+    {
+      "epoch": 1.1189893076877129,
+      "grad_norm": 0.48709022056044177,
+      "learning_rate": 1.4822612710966143e-05,
+      "loss": 0.3325,
+      "step": 2342
+    },
+    {
+      "epoch": 1.119467176393286,
+      "grad_norm": 0.5110741917418703,
+      "learning_rate": 1.4817998259671286e-05,
+      "loss": 0.3411,
+      "step": 2343
+    },
+    {
+      "epoch": 1.119945045098859,
+      "grad_norm": 0.526551969506569,
+      "learning_rate": 1.4813382471948705e-05,
+      "loss": 0.3283,
+      "step": 2344
+    },
+    {
+      "epoch": 1.1204229138044322,
+      "grad_norm": 0.487810166808925,
+      "learning_rate": 1.4808765349078729e-05,
+      "loss": 0.3299,
+      "step": 2345
+    },
+    {
+      "epoch": 1.1209007825100055,
+      "grad_norm": 0.5296951175479421,
+      "learning_rate": 1.4804146892342071e-05,
+      "loss": 0.3379,
+      "step": 2346
+    },
+    {
+      "epoch": 1.1213786512155786,
+      "grad_norm": 0.5101831769474524,
+      "learning_rate": 1.4799527103019808e-05,
+      "loss": 0.3271,
+      "step": 2347
+    },
+    {
+      "epoch": 1.1218565199211517,
+      "grad_norm": 0.4890714139873493,
+      "learning_rate": 1.4794905982393388e-05,
+      "loss": 0.3395,
+      "step": 2348
+    },
+    {
+      "epoch": 1.1223343886267247,
+      "grad_norm": 0.5450992295068559,
+      "learning_rate": 1.4790283531744634e-05,
+      "loss": 0.3446,
+      "step": 2349
+    },
+    {
+      "epoch": 1.1228122573322978,
+      "grad_norm": 0.5031828458579483,
+      "learning_rate": 1.4785659752355724e-05,
+      "loss": 0.3466,
+      "step": 2350
+    },
+    {
+      "epoch": 1.1232901260378712,
+      "grad_norm": 0.4811659005185453,
+      "learning_rate": 1.4781034645509216e-05,
+      "loss": 0.3485,
+      "step": 2351
+    },
+    {
+      "epoch": 1.1237679947434442,
+      "grad_norm": 0.5448333215238423,
+      "learning_rate": 1.4776408212488035e-05,
+      "loss": 0.3263,
+      "step": 2352
+    },
+    {
+      "epoch": 1.1242458634490173,
+      "grad_norm": 0.4953819912341556,
+      "learning_rate": 1.4771780454575468e-05,
+      "loss": 0.3272,
+      "step": 2353
+    },
+    {
+      "epoch": 1.1247237321545904,
+      "grad_norm": 0.5175483590354045,
+      "learning_rate": 1.4767151373055178e-05,
+      "loss": 0.3269,
+      "step": 2354
+    },
+    {
+      "epoch": 1.1252016008601637,
+      "grad_norm": 0.5840296576158932,
+      "learning_rate": 1.4762520969211186e-05,
+      "loss": 0.3311,
+      "step": 2355
+    },
+    {
+      "epoch": 1.1256794695657368,
+      "grad_norm": 0.5464923552621009,
+      "learning_rate": 1.4757889244327888e-05,
+      "loss": 0.3181,
+      "step": 2356
+    },
+    {
+      "epoch": 1.12615733827131,
+      "grad_norm": 0.5188633658259454,
+      "learning_rate": 1.475325619969004e-05,
+      "loss": 0.3477,
+      "step": 2357
+    },
+    {
+      "epoch": 1.126635206976883,
+      "grad_norm": 0.4888981161848857,
+      "learning_rate": 1.4748621836582772e-05,
+      "loss": 0.3253,
+      "step": 2358
+    },
+    {
+      "epoch": 1.1271130756824563,
+      "grad_norm": 0.5231748369329128,
+      "learning_rate": 1.4743986156291568e-05,
+      "loss": 0.3304,
+      "step": 2359
+    },
+    {
+      "epoch": 1.1275909443880294,
+      "grad_norm": 0.4988116880762236,
+      "learning_rate": 1.4739349160102285e-05,
+      "loss": 0.3313,
+      "step": 2360
+    },
+    {
+      "epoch": 1.1280688130936025,
+      "grad_norm": 0.5036551455737499,
+      "learning_rate": 1.4734710849301151e-05,
+      "loss": 0.3258,
+      "step": 2361
+    },
+    {
+      "epoch": 1.1285466817991756,
+      "grad_norm": 0.5138628273300756,
+      "learning_rate": 1.473007122517474e-05,
+      "loss": 0.3305,
+      "step": 2362
+    },
+    {
+      "epoch": 1.1290245505047487,
+      "grad_norm": 0.5138521817190567,
+      "learning_rate": 1.472543028901001e-05,
+      "loss": 0.3438,
+      "step": 2363
+    },
+    {
+      "epoch": 1.129502419210322,
+      "grad_norm": 0.5299139873711323,
+      "learning_rate": 1.4720788042094273e-05,
+      "loss": 0.3243,
+      "step": 2364
+    },
+    {
+      "epoch": 1.129980287915895,
+      "grad_norm": 0.5461978922038623,
+      "learning_rate": 1.4716144485715209e-05,
+      "loss": 0.3236,
+      "step": 2365
+    },
+    {
+      "epoch": 1.1304581566214682,
+      "grad_norm": 0.5311350713847056,
+      "learning_rate": 1.4711499621160855e-05,
+      "loss": 0.3174,
+      "step": 2366
+    },
+    {
+      "epoch": 1.1309360253270415,
+      "grad_norm": 0.5009730559856843,
+      "learning_rate": 1.4706853449719614e-05,
+      "loss": 0.3376,
+      "step": 2367
+    },
+    {
+      "epoch": 1.1314138940326146,
+      "grad_norm": 0.5018126239017102,
+      "learning_rate": 1.4702205972680254e-05,
+      "loss": 0.3328,
+      "step": 2368
+    },
+    {
+      "epoch": 1.1318917627381877,
+      "grad_norm": 0.5242488655503064,
+      "learning_rate": 1.4697557191331902e-05,
+      "loss": 0.3193,
+      "step": 2369
+    },
+    {
+      "epoch": 1.1323696314437608,
+      "grad_norm": 0.5945549331790201,
+      "learning_rate": 1.4692907106964051e-05,
+      "loss": 0.3345,
+      "step": 2370
+    },
+    {
+      "epoch": 1.1328475001493339,
+      "grad_norm": 0.48153137632362025,
+      "learning_rate": 1.468825572086655e-05,
+      "loss": 0.3378,
+      "step": 2371
+    },
+    {
+      "epoch": 1.1333253688549072,
+      "grad_norm": 0.5151533016446908,
+      "learning_rate": 1.4683603034329608e-05,
+      "loss": 0.3168,
+      "step": 2372
+    },
+    {
+      "epoch": 1.1338032375604803,
+      "grad_norm": 0.49149298419228526,
+      "learning_rate": 1.4678949048643806e-05,
+      "loss": 0.3354,
+      "step": 2373
+    },
+    {
+      "epoch": 1.1342811062660534,
+      "grad_norm": 0.48897908673118196,
+      "learning_rate": 1.4674293765100069e-05,
+      "loss": 0.3339,
+      "step": 2374
+    },
+    {
+      "epoch": 1.1347589749716265,
+      "grad_norm": 0.4949702713316802,
+      "learning_rate": 1.4669637184989696e-05,
+      "loss": 0.3257,
+      "step": 2375
+    },
+    {
+      "epoch": 1.1352368436771996,
+      "grad_norm": 0.5058801752873772,
+      "learning_rate": 1.466497930960434e-05,
+      "loss": 0.3519,
+      "step": 2376
+    },
+    {
+      "epoch": 1.1357147123827729,
+      "grad_norm": 0.5237225993986185,
+      "learning_rate": 1.4660320140236014e-05,
+      "loss": 0.3322,
+      "step": 2377
+    },
+    {
+      "epoch": 1.136192581088346,
+      "grad_norm": 0.5177111438994864,
+      "learning_rate": 1.4655659678177084e-05,
+      "loss": 0.336,
+      "step": 2378
+    },
+    {
+      "epoch": 1.136670449793919,
+      "grad_norm": 0.5012469860916707,
+      "learning_rate": 1.4650997924720288e-05,
+      "loss": 0.32,
+      "step": 2379
+    },
+    {
+      "epoch": 1.1371483184994924,
+      "grad_norm": 0.5648557105154629,
+      "learning_rate": 1.4646334881158704e-05,
+      "loss": 0.3446,
+      "step": 2380
+    },
+    {
+      "epoch": 1.1376261872050655,
+      "grad_norm": 0.5492302394701447,
+      "learning_rate": 1.4641670548785788e-05,
+      "loss": 0.3237,
+      "step": 2381
+    },
+    {
+      "epoch": 1.1381040559106386,
+      "grad_norm": 0.5032033816755968,
+      "learning_rate": 1.4637004928895337e-05,
+      "loss": 0.336,
+      "step": 2382
+    },
+    {
+      "epoch": 1.1385819246162117,
+      "grad_norm": 0.5123273009616124,
+      "learning_rate": 1.4632338022781516e-05,
+      "loss": 0.3177,
+      "step": 2383
+    },
+    {
+      "epoch": 1.1390597933217848,
+      "grad_norm": 0.5450557175277998,
+      "learning_rate": 1.462766983173884e-05,
+      "loss": 0.3262,
+      "step": 2384
+    },
+    {
+      "epoch": 1.139537662027358,
+      "grad_norm": 0.5720917143183127,
+      "learning_rate": 1.4623000357062184e-05,
+      "loss": 0.3367,
+      "step": 2385
+    },
+    {
+      "epoch": 1.1400155307329312,
+      "grad_norm": 0.5807405509549664,
+      "learning_rate": 1.4618329600046774e-05,
+      "loss": 0.347,
+      "step": 2386
+    },
+    {
+      "epoch": 1.1404933994385043,
+      "grad_norm": 0.5038498790994034,
+      "learning_rate": 1.4613657561988203e-05,
+      "loss": 0.3368,
+      "step": 2387
+    },
+    {
+      "epoch": 1.1409712681440773,
+      "grad_norm": 0.4737365464414105,
+      "learning_rate": 1.4608984244182407e-05,
+      "loss": 0.3223,
+      "step": 2388
+    },
+    {
+      "epoch": 1.1414491368496504,
+      "grad_norm": 0.5476800314149111,
+      "learning_rate": 1.4604309647925683e-05,
+      "loss": 0.335,
+      "step": 2389
+    },
+    {
+      "epoch": 1.1419270055552238,
+      "grad_norm": 0.522067736414712,
+      "learning_rate": 1.459963377451468e-05,
+      "loss": 0.3407,
+      "step": 2390
+    },
+    {
+      "epoch": 1.1424048742607968,
+      "grad_norm": 0.5362139688122035,
+      "learning_rate": 1.4594956625246406e-05,
+      "loss": 0.3258,
+      "step": 2391
+    },
+    {
+      "epoch": 1.14288274296637,
+      "grad_norm": 0.5547169113461832,
+      "learning_rate": 1.4590278201418216e-05,
+      "loss": 0.3184,
+      "step": 2392
+    },
+    {
+      "epoch": 1.1433606116719432,
+      "grad_norm": 0.5163474508287502,
+      "learning_rate": 1.4585598504327824e-05,
+      "loss": 0.3507,
+      "step": 2393
+    },
+    {
+      "epoch": 1.1438384803775163,
+      "grad_norm": 0.5023941568737598,
+      "learning_rate": 1.4580917535273295e-05,
+      "loss": 0.3541,
+      "step": 2394
+    },
+    {
+      "epoch": 1.1443163490830894,
+      "grad_norm": 0.5253487591862723,
+      "learning_rate": 1.457623529555305e-05,
+      "loss": 0.3262,
+      "step": 2395
+    },
+    {
+      "epoch": 1.1447942177886625,
+      "grad_norm": 0.534817117306971,
+      "learning_rate": 1.4571551786465857e-05,
+      "loss": 0.3324,
+      "step": 2396
+    },
+    {
+      "epoch": 1.1452720864942356,
+      "grad_norm": 0.48543838148344,
+      "learning_rate": 1.4566867009310834e-05,
+      "loss": 0.3483,
+      "step": 2397
+    },
+    {
+      "epoch": 1.145749955199809,
+      "grad_norm": 0.5361413462770422,
+      "learning_rate": 1.4562180965387466e-05,
+      "loss": 0.3236,
+      "step": 2398
+    },
+    {
+      "epoch": 1.146227823905382,
+      "grad_norm": 0.5551669716622398,
+      "learning_rate": 1.4557493655995574e-05,
+      "loss": 0.3553,
+      "step": 2399
+    },
+    {
+      "epoch": 1.1467056926109551,
+      "grad_norm": 0.5044244521639527,
+      "learning_rate": 1.4552805082435333e-05,
+      "loss": 0.3353,
+      "step": 2400
+    },
+    {
+      "epoch": 1.1471835613165282,
+      "grad_norm": 0.5363754449521199,
+      "learning_rate": 1.4548115246007274e-05,
+      "loss": 0.3359,
+      "step": 2401
+    },
+    {
+      "epoch": 1.1476614300221013,
+      "grad_norm": 0.513128864486388,
+      "learning_rate": 1.4543424148012271e-05,
+      "loss": 0.3385,
+      "step": 2402
+    },
+    {
+      "epoch": 1.1481392987276746,
+      "grad_norm": 0.46287451620253967,
+      "learning_rate": 1.4538731789751553e-05,
+      "loss": 0.3286,
+      "step": 2403
+    },
+    {
+      "epoch": 1.1486171674332477,
+      "grad_norm": 0.5195791493585902,
+      "learning_rate": 1.45340381725267e-05,
+      "loss": 0.34,
+      "step": 2404
+    },
+    {
+      "epoch": 1.1490950361388208,
+      "grad_norm": 0.48538977574421593,
+      "learning_rate": 1.4529343297639638e-05,
+      "loss": 0.3307,
+      "step": 2405
+    },
+    {
+      "epoch": 1.1495729048443941,
+      "grad_norm": 0.5095807086855236,
+      "learning_rate": 1.452464716639264e-05,
+      "loss": 0.3502,
+      "step": 2406
+    },
+    {
+      "epoch": 1.1500507735499672,
+      "grad_norm": 0.5122639691694912,
+      "learning_rate": 1.4519949780088334e-05,
+      "loss": 0.3292,
+      "step": 2407
+    },
+    {
+      "epoch": 1.1505286422555403,
+      "grad_norm": 0.5380110437402467,
+      "learning_rate": 1.4515251140029687e-05,
+      "loss": 0.3302,
+      "step": 2408
+    },
+    {
+      "epoch": 1.1510065109611134,
+      "grad_norm": 0.5213594922561937,
+      "learning_rate": 1.451055124752002e-05,
+      "loss": 0.3154,
+      "step": 2409
+    },
+    {
+      "epoch": 1.1514843796666865,
+      "grad_norm": 0.5161145005085588,
+      "learning_rate": 1.4505850103863007e-05,
+      "loss": 0.3373,
+      "step": 2410
+    },
+    {
+      "epoch": 1.1519622483722598,
+      "grad_norm": 0.5206093606448915,
+      "learning_rate": 1.4501147710362653e-05,
+      "loss": 0.3371,
+      "step": 2411
+    },
+    {
+      "epoch": 1.152440117077833,
+      "grad_norm": 0.5091973446696731,
+      "learning_rate": 1.4496444068323322e-05,
+      "loss": 0.3327,
+      "step": 2412
+    },
+    {
+      "epoch": 1.152917985783406,
+      "grad_norm": 0.5085178563346524,
+      "learning_rate": 1.4491739179049725e-05,
+      "loss": 0.3239,
+      "step": 2413
+    },
+    {
+      "epoch": 1.153395854488979,
+      "grad_norm": 0.5268216105789499,
+      "learning_rate": 1.448703304384691e-05,
+      "loss": 0.3334,
+      "step": 2414
+    },
+    {
+      "epoch": 1.1538737231945524,
+      "grad_norm": 0.5068318386596699,
+      "learning_rate": 1.448232566402028e-05,
+      "loss": 0.3527,
+      "step": 2415
+    },
+    {
+      "epoch": 1.1543515919001255,
+      "grad_norm": 0.5258368149602686,
+      "learning_rate": 1.4477617040875577e-05,
+      "loss": 0.3258,
+      "step": 2416
+    },
+    {
+      "epoch": 1.1548294606056986,
+      "grad_norm": 0.47858102510009803,
+      "learning_rate": 1.4472907175718893e-05,
+      "loss": 0.3402,
+      "step": 2417
+    },
+    {
+      "epoch": 1.1553073293112717,
+      "grad_norm": 0.5821829343093783,
+      "learning_rate": 1.4468196069856658e-05,
+      "loss": 0.3197,
+      "step": 2418
+    },
+    {
+      "epoch": 1.155785198016845,
+      "grad_norm": 0.473803724054973,
+      "learning_rate": 1.4463483724595651e-05,
+      "loss": 0.3324,
+      "step": 2419
+    },
+    {
+      "epoch": 1.156263066722418,
+      "grad_norm": 0.512993585884921,
+      "learning_rate": 1.4458770141242992e-05,
+      "loss": 0.3341,
+      "step": 2420
+    },
+    {
+      "epoch": 1.1567409354279912,
+      "grad_norm": 0.49790696160613324,
+      "learning_rate": 1.4454055321106148e-05,
+      "loss": 0.3195,
+      "step": 2421
+    },
+    {
+      "epoch": 1.1572188041335643,
+      "grad_norm": 0.4852347254650604,
+      "learning_rate": 1.4449339265492927e-05,
+      "loss": 0.3296,
+      "step": 2422
+    },
+    {
+      "epoch": 1.1576966728391374,
+      "grad_norm": 0.5307482887330455,
+      "learning_rate": 1.4444621975711477e-05,
+      "loss": 0.3257,
+      "step": 2423
+    },
+    {
+      "epoch": 1.1581745415447107,
+      "grad_norm": 0.5158352775082153,
+      "learning_rate": 1.4439903453070294e-05,
+      "loss": 0.3439,
+      "step": 2424
+    },
+    {
+      "epoch": 1.1586524102502838,
+      "grad_norm": 0.5214779865598925,
+      "learning_rate": 1.4435183698878212e-05,
+      "loss": 0.3232,
+      "step": 2425
+    },
+    {
+      "epoch": 1.1591302789558569,
+      "grad_norm": 0.5149532717689103,
+      "learning_rate": 1.4430462714444406e-05,
+      "loss": 0.3374,
+      "step": 2426
+    },
+    {
+      "epoch": 1.15960814766143,
+      "grad_norm": 0.5313361793801107,
+      "learning_rate": 1.4425740501078396e-05,
+      "loss": 0.3196,
+      "step": 2427
+    },
+    {
+      "epoch": 1.1600860163670033,
+      "grad_norm": 0.49763038284494304,
+      "learning_rate": 1.4421017060090041e-05,
+      "loss": 0.3389,
+      "step": 2428
+    },
+    {
+      "epoch": 1.1605638850725764,
+      "grad_norm": 0.5899500716656553,
+      "learning_rate": 1.4416292392789538e-05,
+      "loss": 0.3331,
+      "step": 2429
+    },
+    {
+      "epoch": 1.1610417537781494,
+      "grad_norm": 0.4872995509813868,
+      "learning_rate": 1.4411566500487425e-05,
+      "loss": 0.3466,
+      "step": 2430
+    },
+    {
+      "epoch": 1.1615196224837225,
+      "grad_norm": 0.4954045549900169,
+      "learning_rate": 1.4406839384494585e-05,
+      "loss": 0.3374,
+      "step": 2431
+    },
+    {
+      "epoch": 1.1619974911892959,
+      "grad_norm": 0.4851398103962068,
+      "learning_rate": 1.4402111046122234e-05,
+      "loss": 0.3233,
+      "step": 2432
+    },
+    {
+      "epoch": 1.162475359894869,
+      "grad_norm": 0.49511298286491195,
+      "learning_rate": 1.4397381486681931e-05,
+      "loss": 0.3317,
+      "step": 2433
+    },
+    {
+      "epoch": 1.162953228600442,
+      "grad_norm": 0.4953595652477772,
+      "learning_rate": 1.4392650707485574e-05,
+      "loss": 0.32,
+      "step": 2434
+    },
+    {
+      "epoch": 1.1634310973060151,
+      "grad_norm": 0.5090056906997068,
+      "learning_rate": 1.4387918709845395e-05,
+      "loss": 0.3406,
+      "step": 2435
+    },
+    {
+      "epoch": 1.1639089660115882,
+      "grad_norm": 0.47175952406629523,
+      "learning_rate": 1.4383185495073968e-05,
+      "loss": 0.351,
+      "step": 2436
+    },
+    {
+      "epoch": 1.1643868347171615,
+      "grad_norm": 0.5195708641390865,
+      "learning_rate": 1.43784510644842e-05,
+      "loss": 0.327,
+      "step": 2437
+    },
+    {
+      "epoch": 1.1648647034227346,
+      "grad_norm": 0.49425130257319605,
+      "learning_rate": 1.4373715419389345e-05,
+      "loss": 0.3203,
+      "step": 2438
+    },
+    {
+      "epoch": 1.1653425721283077,
+      "grad_norm": 0.49907031699241045,
+      "learning_rate": 1.4368978561102982e-05,
+      "loss": 0.3254,
+      "step": 2439
+    },
+    {
+      "epoch": 1.1658204408338808,
+      "grad_norm": 0.5690059412126197,
+      "learning_rate": 1.4364240490939032e-05,
+      "loss": 0.339,
+      "step": 2440
+    },
+    {
+      "epoch": 1.1662983095394541,
+      "grad_norm": 0.48782220485318945,
+      "learning_rate": 1.4359501210211754e-05,
+      "loss": 0.3175,
+      "step": 2441
+    },
+    {
+      "epoch": 1.1667761782450272,
+      "grad_norm": 0.48282271375944663,
+      "learning_rate": 1.4354760720235743e-05,
+      "loss": 0.3276,
+      "step": 2442
+    },
+    {
+      "epoch": 1.1672540469506003,
+      "grad_norm": 0.4907377583967225,
+      "learning_rate": 1.4350019022325925e-05,
+      "loss": 0.332,
+      "step": 2443
+    },
+    {
+      "epoch": 1.1677319156561734,
+      "grad_norm": 0.5201773113156611,
+      "learning_rate": 1.434527611779756e-05,
+      "loss": 0.3322,
+      "step": 2444
+    },
+    {
+      "epoch": 1.1682097843617467,
+      "grad_norm": 0.5073601879220666,
+      "learning_rate": 1.4340532007966252e-05,
+      "loss": 0.344,
+      "step": 2445
+    },
+    {
+      "epoch": 1.1686876530673198,
+      "grad_norm": 0.4966658318824344,
+      "learning_rate": 1.4335786694147931e-05,
+      "loss": 0.3286,
+      "step": 2446
+    },
+    {
+      "epoch": 1.169165521772893,
+      "grad_norm": 0.47795836779748363,
+      "learning_rate": 1.4331040177658859e-05,
+      "loss": 0.3203,
+      "step": 2447
+    },
+    {
+      "epoch": 1.169643390478466,
+      "grad_norm": 0.5167223624760666,
+      "learning_rate": 1.4326292459815642e-05,
+      "loss": 0.3289,
+      "step": 2448
+    },
+    {
+      "epoch": 1.170121259184039,
+      "grad_norm": 0.4943415537304743,
+      "learning_rate": 1.4321543541935213e-05,
+      "loss": 0.3412,
+      "step": 2449
+    },
+    {
+      "epoch": 1.1705991278896124,
+      "grad_norm": 0.5447813509352554,
+      "learning_rate": 1.4316793425334836e-05,
+      "loss": 0.3317,
+      "step": 2450
+    },
+    {
+      "epoch": 1.1710769965951855,
+      "grad_norm": 0.6040250815392916,
+      "learning_rate": 1.4312042111332108e-05,
+      "loss": 0.319,
+      "step": 2451
+    },
+    {
+      "epoch": 1.1715548653007586,
+      "grad_norm": 0.4897011714819888,
+      "learning_rate": 1.4307289601244963e-05,
+      "loss": 0.3253,
+      "step": 2452
+    },
+    {
+      "epoch": 1.1720327340063317,
+      "grad_norm": 0.519171989422915,
+      "learning_rate": 1.4302535896391661e-05,
+      "loss": 0.3387,
+      "step": 2453
+    },
+    {
+      "epoch": 1.172510602711905,
+      "grad_norm": 0.5310376803286037,
+      "learning_rate": 1.4297780998090795e-05,
+      "loss": 0.3306,
+      "step": 2454
+    },
+    {
+      "epoch": 1.172988471417478,
+      "grad_norm": 0.5546851814204874,
+      "learning_rate": 1.4293024907661295e-05,
+      "loss": 0.3302,
+      "step": 2455
+    },
+    {
+      "epoch": 1.1734663401230512,
+      "grad_norm": 0.5438187932879726,
+      "learning_rate": 1.4288267626422411e-05,
+      "loss": 0.3306,
+      "step": 2456
+    },
+    {
+      "epoch": 1.1739442088286243,
+      "grad_norm": 0.5999235748445584,
+      "learning_rate": 1.4283509155693734e-05,
+      "loss": 0.3299,
+      "step": 2457
+    },
+    {
+      "epoch": 1.1744220775341976,
+      "grad_norm": 0.5267230789014691,
+      "learning_rate": 1.4278749496795174e-05,
+      "loss": 0.3385,
+      "step": 2458
+    },
+    {
+      "epoch": 1.1748999462397707,
+      "grad_norm": 0.49465625934833485,
+      "learning_rate": 1.4273988651046982e-05,
+      "loss": 0.3446,
+      "step": 2459
+    },
+    {
+      "epoch": 1.1753778149453438,
+      "grad_norm": 0.520908355310508,
+      "learning_rate": 1.4269226619769727e-05,
+      "loss": 0.3317,
+      "step": 2460
+    },
+    {
+      "epoch": 1.1758556836509169,
+      "grad_norm": 0.49990608969106165,
+      "learning_rate": 1.4264463404284317e-05,
+      "loss": 0.3267,
+      "step": 2461
+    },
+    {
+      "epoch": 1.17633355235649,
+      "grad_norm": 0.49260831125565,
+      "learning_rate": 1.4259699005911984e-05,
+      "loss": 0.334,
+      "step": 2462
+    },
+    {
+      "epoch": 1.1768114210620633,
+      "grad_norm": 0.5685639467668431,
+      "learning_rate": 1.4254933425974284e-05,
+      "loss": 0.3169,
+      "step": 2463
+    },
+    {
+      "epoch": 1.1772892897676364,
+      "grad_norm": 0.5053854894890485,
+      "learning_rate": 1.4250166665793106e-05,
+      "loss": 0.3172,
+      "step": 2464
+    },
+    {
+      "epoch": 1.1777671584732095,
+      "grad_norm": 0.5094103880275894,
+      "learning_rate": 1.424539872669067e-05,
+      "loss": 0.3371,
+      "step": 2465
+    },
+    {
+      "epoch": 1.1782450271787825,
+      "grad_norm": 0.4946974997604784,
+      "learning_rate": 1.4240629609989513e-05,
+      "loss": 0.3334,
+      "step": 2466
+    },
+    {
+      "epoch": 1.1787228958843559,
+      "grad_norm": 0.5018625213670579,
+      "learning_rate": 1.4235859317012506e-05,
+      "loss": 0.321,
+      "step": 2467
+    },
+    {
+      "epoch": 1.179200764589929,
+      "grad_norm": 0.4709704159413237,
+      "learning_rate": 1.4231087849082848e-05,
+      "loss": 0.3285,
+      "step": 2468
+    },
+    {
+      "epoch": 1.179678633295502,
+      "grad_norm": 0.49926378629859847,
+      "learning_rate": 1.4226315207524049e-05,
+      "loss": 0.3313,
+      "step": 2469
+    },
+    {
+      "epoch": 1.1801565020010751,
+      "grad_norm": 0.512829518968713,
+      "learning_rate": 1.4221541393659966e-05,
+      "loss": 0.3374,
+      "step": 2470
+    },
+    {
+      "epoch": 1.1806343707066485,
+      "grad_norm": 0.49421509200107594,
+      "learning_rate": 1.4216766408814766e-05,
+      "loss": 0.3326,
+      "step": 2471
+    },
+    {
+      "epoch": 1.1811122394122215,
+      "grad_norm": 0.5693138964760737,
+      "learning_rate": 1.4211990254312948e-05,
+      "loss": 0.3413,
+      "step": 2472
+    },
+    {
+      "epoch": 1.1815901081177946,
+      "grad_norm": 0.4758930284278045,
+      "learning_rate": 1.4207212931479331e-05,
+      "loss": 0.3227,
+      "step": 2473
+    },
+    {
+      "epoch": 1.1820679768233677,
+      "grad_norm": 0.5414563707616173,
+      "learning_rate": 1.4202434441639061e-05,
+      "loss": 0.3565,
+      "step": 2474
+    },
+    {
+      "epoch": 1.1825458455289408,
+      "grad_norm": 0.47934944618315223,
+      "learning_rate": 1.4197654786117604e-05,
+      "loss": 0.3228,
+      "step": 2475
+    },
+    {
+      "epoch": 1.1830237142345141,
+      "grad_norm": 0.6167228683042635,
+      "learning_rate": 1.419287396624076e-05,
+      "loss": 0.3194,
+      "step": 2476
+    },
+    {
+      "epoch": 1.1835015829400872,
+      "grad_norm": 0.6097773368996701,
+      "learning_rate": 1.4188091983334636e-05,
+      "loss": 0.3344,
+      "step": 2477
+    },
+    {
+      "epoch": 1.1839794516456603,
+      "grad_norm": 0.5410843300407218,
+      "learning_rate": 1.4183308838725669e-05,
+      "loss": 0.3372,
+      "step": 2478
+    },
+    {
+      "epoch": 1.1844573203512334,
+      "grad_norm": 0.48327603901696636,
+      "learning_rate": 1.4178524533740628e-05,
+      "loss": 0.3308,
+      "step": 2479
+    },
+    {
+      "epoch": 1.1849351890568067,
+      "grad_norm": 0.5268010648750129,
+      "learning_rate": 1.4173739069706586e-05,
+      "loss": 0.3396,
+      "step": 2480
+    },
+    {
+      "epoch": 1.1854130577623798,
+      "grad_norm": 0.5058799850737847,
+      "learning_rate": 1.4168952447950948e-05,
+      "loss": 0.3488,
+      "step": 2481
+    },
+    {
+      "epoch": 1.185890926467953,
+      "grad_norm": 0.4793089155470465,
+      "learning_rate": 1.4164164669801444e-05,
+      "loss": 0.3202,
+      "step": 2482
+    },
+    {
+      "epoch": 1.186368795173526,
+      "grad_norm": 0.537347786961874,
+      "learning_rate": 1.4159375736586114e-05,
+      "loss": 0.323,
+      "step": 2483
+    },
+    {
+      "epoch": 1.1868466638790993,
+      "grad_norm": 0.5043366860217638,
+      "learning_rate": 1.4154585649633324e-05,
+      "loss": 0.3332,
+      "step": 2484
+    },
+    {
+      "epoch": 1.1873245325846724,
+      "grad_norm": 0.6175384664642016,
+      "learning_rate": 1.414979441027176e-05,
+      "loss": 0.3298,
+      "step": 2485
+    },
+    {
+      "epoch": 1.1878024012902455,
+      "grad_norm": 0.5172614792354202,
+      "learning_rate": 1.414500201983043e-05,
+      "loss": 0.3299,
+      "step": 2486
+    },
+    {
+      "epoch": 1.1882802699958186,
+      "grad_norm": 0.5074099118761909,
+      "learning_rate": 1.4140208479638653e-05,
+      "loss": 0.3314,
+      "step": 2487
+    },
+    {
+      "epoch": 1.1887581387013917,
+      "grad_norm": 0.49819077409033413,
+      "learning_rate": 1.4135413791026081e-05,
+      "loss": 0.3451,
+      "step": 2488
+    },
+    {
+      "epoch": 1.189236007406965,
+      "grad_norm": 0.5176243373050009,
+      "learning_rate": 1.4130617955322665e-05,
+      "loss": 0.3221,
+      "step": 2489
+    },
+    {
+      "epoch": 1.189713876112538,
+      "grad_norm": 0.473185476669963,
+      "learning_rate": 1.4125820973858693e-05,
+      "loss": 0.3231,
+      "step": 2490
+    },
+    {
+      "epoch": 1.1901917448181112,
+      "grad_norm": 0.48907220673787455,
+      "learning_rate": 1.4121022847964762e-05,
+      "loss": 0.3408,
+      "step": 2491
+    },
+    {
+      "epoch": 1.1906696135236843,
+      "grad_norm": 1.177648905452567,
+      "learning_rate": 1.4116223578971787e-05,
+      "loss": 0.332,
+      "step": 2492
+    },
+    {
+      "epoch": 1.1911474822292576,
+      "grad_norm": 0.4828061304183026,
+      "learning_rate": 1.4111423168210999e-05,
+      "loss": 0.319,
+      "step": 2493
+    },
+    {
+      "epoch": 1.1916253509348307,
+      "grad_norm": 0.5135930311328842,
+      "learning_rate": 1.410662161701395e-05,
+      "loss": 0.3381,
+      "step": 2494
+    },
+    {
+      "epoch": 1.1921032196404038,
+      "grad_norm": 0.526643592475081,
+      "learning_rate": 1.41018189267125e-05,
+      "loss": 0.3262,
+      "step": 2495
+    },
+    {
+      "epoch": 1.1925810883459769,
+      "grad_norm": 0.5090734041849178,
+      "learning_rate": 1.4097015098638838e-05,
+      "loss": 0.3276,
+      "step": 2496
+    },
+    {
+      "epoch": 1.1930589570515502,
+      "grad_norm": 0.5174675269837884,
+      "learning_rate": 1.4092210134125458e-05,
+      "loss": 0.3263,
+      "step": 2497
+    },
+    {
+      "epoch": 1.1935368257571233,
+      "grad_norm": 0.5214389237769186,
+      "learning_rate": 1.4087404034505167e-05,
+      "loss": 0.3375,
+      "step": 2498
+    },
+    {
+      "epoch": 1.1940146944626964,
+      "grad_norm": 0.4846029928861073,
+      "learning_rate": 1.4082596801111104e-05,
+      "loss": 0.3388,
+      "step": 2499
+    },
+    {
+      "epoch": 1.1944925631682695,
+      "grad_norm": 0.5209998258997901,
+      "learning_rate": 1.4077788435276701e-05,
+      "loss": 0.3373,
+      "step": 2500
+    },
+    {
+      "epoch": 1.1949704318738426,
+      "grad_norm": 0.5407837704720175,
+      "learning_rate": 1.4072978938335717e-05,
+      "loss": 0.3242,
+      "step": 2501
+    },
+    {
+      "epoch": 1.1954483005794159,
+      "grad_norm": 0.4792112154512138,
+      "learning_rate": 1.4068168311622223e-05,
+      "loss": 0.3262,
+      "step": 2502
+    },
+    {
+      "epoch": 1.195926169284989,
+      "grad_norm": 0.5012097821642677,
+      "learning_rate": 1.40633565564706e-05,
+      "loss": 0.3299,
+      "step": 2503
+    },
+    {
+      "epoch": 1.196404037990562,
+      "grad_norm": 0.5011100101721819,
+      "learning_rate": 1.4058543674215543e-05,
+      "loss": 0.3303,
+      "step": 2504
+    },
+    {
+      "epoch": 1.1968819066961351,
+      "grad_norm": 0.5072036552663972,
+      "learning_rate": 1.4053729666192067e-05,
+      "loss": 0.3034,
+      "step": 2505
+    },
+    {
+      "epoch": 1.1973597754017085,
+      "grad_norm": 0.49966051888354573,
+      "learning_rate": 1.4048914533735482e-05,
+      "loss": 0.3185,
+      "step": 2506
+    },
+    {
+      "epoch": 1.1978376441072816,
+      "grad_norm": 0.6201559448981543,
+      "learning_rate": 1.404409827818143e-05,
+      "loss": 0.3081,
+      "step": 2507
+    },
+    {
+      "epoch": 1.1983155128128546,
+      "grad_norm": 0.4892341169267986,
+      "learning_rate": 1.4039280900865855e-05,
+      "loss": 0.3347,
+      "step": 2508
+    },
+    {
+      "epoch": 1.1987933815184277,
+      "grad_norm": 0.4894821104819882,
+      "learning_rate": 1.4034462403125004e-05,
+      "loss": 0.319,
+      "step": 2509
+    },
+    {
+      "epoch": 1.199271250224001,
+      "grad_norm": 0.529046836458218,
+      "learning_rate": 1.4029642786295452e-05,
+      "loss": 0.3228,
+      "step": 2510
+    },
+    {
+      "epoch": 1.1997491189295741,
+      "grad_norm": 0.5119754528953527,
+      "learning_rate": 1.4024822051714075e-05,
+      "loss": 0.3259,
+      "step": 2511
+    },
+    {
+      "epoch": 1.2002269876351472,
+      "grad_norm": 0.5166507305196261,
+      "learning_rate": 1.4020000200718053e-05,
+      "loss": 0.347,
+      "step": 2512
+    },
+    {
+      "epoch": 1.2007048563407203,
+      "grad_norm": 0.485424117273842,
+      "learning_rate": 1.401517723464489e-05,
+      "loss": 0.3433,
+      "step": 2513
+    },
+    {
+      "epoch": 1.2011827250462934,
+      "grad_norm": 0.49766272512833354,
+      "learning_rate": 1.4010353154832388e-05,
+      "loss": 0.336,
+      "step": 2514
+    },
+    {
+      "epoch": 1.2016605937518667,
+      "grad_norm": 0.6389939051707664,
+      "learning_rate": 1.400552796261866e-05,
+      "loss": 0.3151,
+      "step": 2515
+    },
+    {
+      "epoch": 1.2021384624574398,
+      "grad_norm": 0.4842050986179893,
+      "learning_rate": 1.4000701659342136e-05,
+      "loss": 0.3315,
+      "step": 2516
+    },
+    {
+      "epoch": 1.202616331163013,
+      "grad_norm": 0.5199065736138204,
+      "learning_rate": 1.3995874246341542e-05,
+      "loss": 0.329,
+      "step": 2517
+    },
+    {
+      "epoch": 1.203094199868586,
+      "grad_norm": 0.46844026904175,
+      "learning_rate": 1.3991045724955915e-05,
+      "loss": 0.3358,
+      "step": 2518
+    },
+    {
+      "epoch": 1.2035720685741593,
+      "grad_norm": 0.5097819606674651,
+      "learning_rate": 1.3986216096524606e-05,
+      "loss": 0.3282,
+      "step": 2519
+    },
+    {
+      "epoch": 1.2040499372797324,
+      "grad_norm": 0.49611533463068996,
+      "learning_rate": 1.3981385362387268e-05,
+      "loss": 0.3237,
+      "step": 2520
+    },
+    {
+      "epoch": 1.2045278059853055,
+      "grad_norm": 0.5427021170944505,
+      "learning_rate": 1.397655352388386e-05,
+      "loss": 0.3174,
+      "step": 2521
+    },
+    {
+      "epoch": 1.2050056746908786,
+      "grad_norm": 0.5177994962109214,
+      "learning_rate": 1.397172058235465e-05,
+      "loss": 0.3266,
+      "step": 2522
+    },
+    {
+      "epoch": 1.205483543396452,
+      "grad_norm": 0.5151693584485367,
+      "learning_rate": 1.3966886539140212e-05,
+      "loss": 0.343,
+      "step": 2523
+    },
+    {
+      "epoch": 1.205961412102025,
+      "grad_norm": 0.4877067051638373,
+      "learning_rate": 1.396205139558142e-05,
+      "loss": 0.3289,
+      "step": 2524
+    },
+    {
+      "epoch": 1.206439280807598,
+      "grad_norm": 0.4986624439516977,
+      "learning_rate": 1.3957215153019463e-05,
+      "loss": 0.3187,
+      "step": 2525
+    },
+    {
+      "epoch": 1.2069171495131712,
+      "grad_norm": 0.5098275657839579,
+      "learning_rate": 1.3952377812795826e-05,
+      "loss": 0.3306,
+      "step": 2526
+    },
+    {
+      "epoch": 1.2073950182187443,
+      "grad_norm": 0.4851080171025919,
+      "learning_rate": 1.3947539376252301e-05,
+      "loss": 0.3326,
+      "step": 2527
+    },
+    {
+      "epoch": 1.2078728869243176,
+      "grad_norm": 0.5181654078258601,
+      "learning_rate": 1.3942699844730986e-05,
+      "loss": 0.3275,
+      "step": 2528
+    },
+    {
+      "epoch": 1.2083507556298907,
+      "grad_norm": 0.517517079824198,
+      "learning_rate": 1.3937859219574286e-05,
+      "loss": 0.308,
+      "step": 2529
+    },
+    {
+      "epoch": 1.2088286243354638,
+      "grad_norm": 0.5359336855688807,
+      "learning_rate": 1.3933017502124897e-05,
+      "loss": 0.3184,
+      "step": 2530
+    },
+    {
+      "epoch": 1.2093064930410369,
+      "grad_norm": 0.5177636108490216,
+      "learning_rate": 1.392817469372583e-05,
+      "loss": 0.3318,
+      "step": 2531
+    },
+    {
+      "epoch": 1.2097843617466102,
+      "grad_norm": 0.5049686036498979,
+      "learning_rate": 1.3923330795720396e-05,
+      "loss": 0.3379,
+      "step": 2532
+    },
+    {
+      "epoch": 1.2102622304521833,
+      "grad_norm": 0.5034062129917729,
+      "learning_rate": 1.3918485809452204e-05,
+      "loss": 0.3297,
+      "step": 2533
+    },
+    {
+      "epoch": 1.2107400991577564,
+      "grad_norm": 0.5046311056670907,
+      "learning_rate": 1.3913639736265175e-05,
+      "loss": 0.3385,
+      "step": 2534
+    },
+    {
+      "epoch": 1.2112179678633295,
+      "grad_norm": 0.7815693194085443,
+      "learning_rate": 1.3908792577503514e-05,
+      "loss": 0.3385,
+      "step": 2535
+    },
+    {
+      "epoch": 1.2116958365689028,
+      "grad_norm": 0.47743692428948703,
+      "learning_rate": 1.3903944334511744e-05,
+      "loss": 0.3177,
+      "step": 2536
+    },
+    {
+      "epoch": 1.2121737052744759,
+      "grad_norm": 0.49090736922505357,
+      "learning_rate": 1.3899095008634678e-05,
+      "loss": 0.3218,
+      "step": 2537
+    },
+    {
+      "epoch": 1.212651573980049,
+      "grad_norm": 0.49841642881825793,
+      "learning_rate": 1.3894244601217435e-05,
+      "loss": 0.3346,
+      "step": 2538
+    },
+    {
+      "epoch": 1.213129442685622,
+      "grad_norm": 0.510969551833143,
+      "learning_rate": 1.3889393113605433e-05,
+      "loss": 0.3552,
+      "step": 2539
+    },
+    {
+      "epoch": 1.2136073113911952,
+      "grad_norm": 0.5523292379752209,
+      "learning_rate": 1.3884540547144393e-05,
+      "loss": 0.3211,
+      "step": 2540
+    },
+    {
+      "epoch": 1.2140851800967685,
+      "grad_norm": 0.5051371714386153,
+      "learning_rate": 1.3879686903180326e-05,
+      "loss": 0.3393,
+      "step": 2541
+    },
+    {
+      "epoch": 1.2145630488023416,
+      "grad_norm": 0.5044231031807489,
+      "learning_rate": 1.3874832183059545e-05,
+      "loss": 0.3157,
+      "step": 2542
+    },
+    {
+      "epoch": 1.2150409175079147,
+      "grad_norm": 0.4730004046440912,
+      "learning_rate": 1.3869976388128672e-05,
+      "loss": 0.3501,
+      "step": 2543
+    },
+    {
+      "epoch": 1.215518786213488,
+      "grad_norm": 0.5105649664830538,
+      "learning_rate": 1.3865119519734611e-05,
+      "loss": 0.3328,
+      "step": 2544
+    },
+    {
+      "epoch": 1.215996654919061,
+      "grad_norm": 0.5704906525223905,
+      "learning_rate": 1.3860261579224574e-05,
+      "loss": 0.3322,
+      "step": 2545
+    },
+    {
+      "epoch": 1.2164745236246342,
+      "grad_norm": 0.525856706643172,
+      "learning_rate": 1.3855402567946072e-05,
+      "loss": 0.3202,
+      "step": 2546
+    },
+    {
+      "epoch": 1.2169523923302072,
+      "grad_norm": 0.47546063015364237,
+      "learning_rate": 1.3850542487246903e-05,
+      "loss": 0.3377,
+      "step": 2547
+    },
+    {
+      "epoch": 1.2174302610357803,
+      "grad_norm": 0.5451603336435644,
+      "learning_rate": 1.384568133847517e-05,
+      "loss": 0.3452,
+      "step": 2548
+    },
+    {
+      "epoch": 1.2179081297413537,
+      "grad_norm": 0.5361634977822047,
+      "learning_rate": 1.3840819122979272e-05,
+      "loss": 0.3131,
+      "step": 2549
+    },
+    {
+      "epoch": 1.2183859984469267,
+      "grad_norm": 0.4922706268092114,
+      "learning_rate": 1.3835955842107897e-05,
+      "loss": 0.3306,
+      "step": 2550
+    },
+    {
+      "epoch": 1.2188638671524998,
+      "grad_norm": 0.4894912959757622,
+      "learning_rate": 1.3831091497210043e-05,
+      "loss": 0.322,
+      "step": 2551
+    },
+    {
+      "epoch": 1.219341735858073,
+      "grad_norm": 0.5551732852060824,
+      "learning_rate": 1.3826226089634982e-05,
+      "loss": 0.3223,
+      "step": 2552
+    },
+    {
+      "epoch": 1.219819604563646,
+      "grad_norm": 0.47984089277387154,
+      "learning_rate": 1.3821359620732297e-05,
+      "loss": 0.3399,
+      "step": 2553
+    },
+    {
+      "epoch": 1.2202974732692193,
+      "grad_norm": 0.469484790257942,
+      "learning_rate": 1.3816492091851865e-05,
+      "loss": 0.3439,
+      "step": 2554
+    },
+    {
+      "epoch": 1.2207753419747924,
+      "grad_norm": 0.49706698283821926,
+      "learning_rate": 1.3811623504343845e-05,
+      "loss": 0.3425,
+      "step": 2555
+    },
+    {
+      "epoch": 1.2212532106803655,
+      "grad_norm": 0.4864721219524991,
+      "learning_rate": 1.3806753859558702e-05,
+      "loss": 0.3333,
+      "step": 2556
+    },
+    {
+      "epoch": 1.2217310793859388,
+      "grad_norm": 0.531371918260641,
+      "learning_rate": 1.380188315884719e-05,
+      "loss": 0.3262,
+      "step": 2557
+    },
+    {
+      "epoch": 1.222208948091512,
+      "grad_norm": 0.508703916265606,
+      "learning_rate": 1.3797011403560349e-05,
+      "loss": 0.3339,
+      "step": 2558
+    },
+    {
+      "epoch": 1.222686816797085,
+      "grad_norm": 0.5519217880427334,
+      "learning_rate": 1.3792138595049526e-05,
+      "loss": 0.3313,
+      "step": 2559
+    },
+    {
+      "epoch": 1.2231646855026581,
+      "grad_norm": 0.49067395021756355,
+      "learning_rate": 1.378726473466635e-05,
+      "loss": 0.3301,
+      "step": 2560
+    },
+    {
+      "epoch": 1.2236425542082312,
+      "grad_norm": 0.527772112286325,
+      "learning_rate": 1.378238982376274e-05,
+      "loss": 0.3121,
+      "step": 2561
+    },
+    {
+      "epoch": 1.2241204229138045,
+      "grad_norm": 0.5044285212896721,
+      "learning_rate": 1.3777513863690914e-05,
+      "loss": 0.3276,
+      "step": 2562
+    },
+    {
+      "epoch": 1.2245982916193776,
+      "grad_norm": 0.5706596519616292,
+      "learning_rate": 1.3772636855803378e-05,
+      "loss": 0.3331,
+      "step": 2563
+    },
+    {
+      "epoch": 1.2250761603249507,
+      "grad_norm": 0.4736938143718017,
+      "learning_rate": 1.3767758801452926e-05,
+      "loss": 0.3358,
+      "step": 2564
+    },
+    {
+      "epoch": 1.2255540290305238,
+      "grad_norm": 0.5097229175872106,
+      "learning_rate": 1.3762879701992642e-05,
+      "loss": 0.3303,
+      "step": 2565
+    },
+    {
+      "epoch": 1.2260318977360969,
+      "grad_norm": 0.5296217110279442,
+      "learning_rate": 1.3757999558775907e-05,
+      "loss": 0.3316,
+      "step": 2566
+    },
+    {
+      "epoch": 1.2265097664416702,
+      "grad_norm": 0.4965806245245254,
+      "learning_rate": 1.3753118373156382e-05,
+      "loss": 0.311,
+      "step": 2567
+    },
+    {
+      "epoch": 1.2269876351472433,
+      "grad_norm": 0.49384330392321485,
+      "learning_rate": 1.3748236146488028e-05,
+      "loss": 0.3394,
+      "step": 2568
+    },
+    {
+      "epoch": 1.2274655038528164,
+      "grad_norm": 0.5159965075461034,
+      "learning_rate": 1.3743352880125083e-05,
+      "loss": 0.315,
+      "step": 2569
+    },
+    {
+      "epoch": 1.2279433725583897,
+      "grad_norm": 0.5382950937633958,
+      "learning_rate": 1.373846857542208e-05,
+      "loss": 0.3495,
+      "step": 2570
+    },
+    {
+      "epoch": 1.2284212412639628,
+      "grad_norm": 0.48821357626356915,
+      "learning_rate": 1.3733583233733843e-05,
+      "loss": 0.3366,
+      "step": 2571
+    },
+    {
+      "epoch": 1.2288991099695359,
+      "grad_norm": 0.4912616599041392,
+      "learning_rate": 1.372869685641547e-05,
+      "loss": 0.3233,
+      "step": 2572
+    },
+    {
+      "epoch": 1.229376978675109,
+      "grad_norm": 0.49906610048849775,
+      "learning_rate": 1.3723809444822366e-05,
+      "loss": 0.3223,
+      "step": 2573
+    },
+    {
+      "epoch": 1.229854847380682,
+      "grad_norm": 0.4900856446449379,
+      "learning_rate": 1.371892100031021e-05,
+      "loss": 0.3358,
+      "step": 2574
+    },
+    {
+      "epoch": 1.2303327160862554,
+      "grad_norm": 0.5218820319587112,
+      "learning_rate": 1.3714031524234965e-05,
+      "loss": 0.3134,
+      "step": 2575
+    },
+    {
+      "epoch": 1.2308105847918285,
+      "grad_norm": 0.4782986698985612,
+      "learning_rate": 1.3709141017952893e-05,
+      "loss": 0.3125,
+      "step": 2576
+    },
+    {
+      "epoch": 1.2312884534974016,
+      "grad_norm": 0.5293380049327159,
+      "learning_rate": 1.370424948282053e-05,
+      "loss": 0.3269,
+      "step": 2577
+    },
+    {
+      "epoch": 1.2317663222029747,
+      "grad_norm": 0.48617546343080253,
+      "learning_rate": 1.3699356920194702e-05,
+      "loss": 0.3262,
+      "step": 2578
+    },
+    {
+      "epoch": 1.2322441909085478,
+      "grad_norm": 0.489962199133281,
+      "learning_rate": 1.3694463331432521e-05,
+      "loss": 0.3145,
+      "step": 2579
+    },
+    {
+      "epoch": 1.232722059614121,
+      "grad_norm": 0.5271714061634383,
+      "learning_rate": 1.3689568717891381e-05,
+      "loss": 0.305,
+      "step": 2580
+    },
+    {
+      "epoch": 1.2331999283196942,
+      "grad_norm": 0.476057076836738,
+      "learning_rate": 1.368467308092896e-05,
+      "loss": 0.3246,
+      "step": 2581
+    },
+    {
+      "epoch": 1.2336777970252673,
+      "grad_norm": 0.5037616861616993,
+      "learning_rate": 1.3679776421903222e-05,
+      "loss": 0.3157,
+      "step": 2582
+    },
+    {
+      "epoch": 1.2341556657308406,
+      "grad_norm": 0.5029866004003386,
+      "learning_rate": 1.3674878742172415e-05,
+      "loss": 0.3424,
+      "step": 2583
+    },
+    {
+      "epoch": 1.2346335344364137,
+      "grad_norm": 0.49776455157060684,
+      "learning_rate": 1.366998004309507e-05,
+      "loss": 0.3221,
+      "step": 2584
+    },
+    {
+      "epoch": 1.2351114031419868,
+      "grad_norm": 0.5159003230211062,
+      "learning_rate": 1.3665080326029997e-05,
+      "loss": 0.3369,
+      "step": 2585
+    },
+    {
+      "epoch": 1.2355892718475598,
+      "grad_norm": 0.4945390309105578,
+      "learning_rate": 1.3660179592336296e-05,
+      "loss": 0.324,
+      "step": 2586
+    },
+    {
+      "epoch": 1.236067140553133,
+      "grad_norm": 0.5028457586094409,
+      "learning_rate": 1.3655277843373338e-05,
+      "loss": 0.3276,
+      "step": 2587
+    },
+    {
+      "epoch": 1.2365450092587063,
+      "grad_norm": 0.4924865463215211,
+      "learning_rate": 1.3650375080500784e-05,
+      "loss": 0.3102,
+      "step": 2588
+    },
+    {
+      "epoch": 1.2370228779642793,
+      "grad_norm": 0.4786677818333961,
+      "learning_rate": 1.3645471305078575e-05,
+      "loss": 0.3336,
+      "step": 2589
+    },
+    {
+      "epoch": 1.2375007466698524,
+      "grad_norm": 0.5645830343314627,
+      "learning_rate": 1.364056651846693e-05,
+      "loss": 0.3441,
+      "step": 2590
+    },
+    {
+      "epoch": 1.2379786153754255,
+      "grad_norm": 0.4772358981883002,
+      "learning_rate": 1.3635660722026352e-05,
+      "loss": 0.3442,
+      "step": 2591
+    },
+    {
+      "epoch": 1.2384564840809986,
+      "grad_norm": 0.5656570549920833,
+      "learning_rate": 1.3630753917117624e-05,
+      "loss": 0.3081,
+      "step": 2592
+    },
+    {
+      "epoch": 1.238934352786572,
+      "grad_norm": 0.4751884801558594,
+      "learning_rate": 1.3625846105101801e-05,
+      "loss": 0.3142,
+      "step": 2593
+    },
+    {
+      "epoch": 1.239412221492145,
+      "grad_norm": 2.4050639525172075,
+      "learning_rate": 1.3620937287340228e-05,
+      "loss": 0.3273,
+      "step": 2594
+    },
+    {
+      "epoch": 1.2398900901977181,
+      "grad_norm": 0.7444744255219318,
+      "learning_rate": 1.3616027465194525e-05,
+      "loss": 0.3273,
+      "step": 2595
+    },
+    {
+      "epoch": 1.2403679589032914,
+      "grad_norm": 0.4944733158259898,
+      "learning_rate": 1.3611116640026589e-05,
+      "loss": 0.3399,
+      "step": 2596
+    },
+    {
+      "epoch": 1.2408458276088645,
+      "grad_norm": 0.4818780698812894,
+      "learning_rate": 1.3606204813198593e-05,
+      "loss": 0.3213,
+      "step": 2597
+    },
+    {
+      "epoch": 1.2413236963144376,
+      "grad_norm": 0.48772958226512186,
+      "learning_rate": 1.3601291986072999e-05,
+      "loss": 0.3328,
+      "step": 2598
+    },
+    {
+      "epoch": 1.2418015650200107,
+      "grad_norm": 0.49554621709972824,
+      "learning_rate": 1.3596378160012529e-05,
+      "loss": 0.3458,
+      "step": 2599
+    },
+    {
+      "epoch": 1.2422794337255838,
+      "grad_norm": 0.473500813505667,
+      "learning_rate": 1.35914633363802e-05,
+      "loss": 0.3203,
+      "step": 2600
+    },
+    {
+      "epoch": 1.2427573024311571,
+      "grad_norm": 0.48856018275808843,
+      "learning_rate": 1.358654751653929e-05,
+      "loss": 0.3171,
+      "step": 2601
+    },
+    {
+      "epoch": 1.2432351711367302,
+      "grad_norm": 0.5132783506070369,
+      "learning_rate": 1.3581630701853368e-05,
+      "loss": 0.3246,
+      "step": 2602
+    },
+    {
+      "epoch": 1.2437130398423033,
+      "grad_norm": 0.4879512811328692,
+      "learning_rate": 1.3576712893686268e-05,
+      "loss": 0.2965,
+      "step": 2603
+    },
+    {
+      "epoch": 1.2441909085478764,
+      "grad_norm": 0.5692497626729707,
+      "learning_rate": 1.3571794093402103e-05,
+      "loss": 0.3554,
+      "step": 2604
+    },
+    {
+      "epoch": 1.2446687772534497,
+      "grad_norm": 0.5262501553177124,
+      "learning_rate": 1.356687430236526e-05,
+      "loss": 0.3426,
+      "step": 2605
+    },
+    {
+      "epoch": 1.2451466459590228,
+      "grad_norm": 0.5101017072346472,
+      "learning_rate": 1.3561953521940408e-05,
+      "loss": 0.3384,
+      "step": 2606
+    },
+    {
+      "epoch": 1.245624514664596,
+      "grad_norm": 0.4855194753569897,
+      "learning_rate": 1.3557031753492477e-05,
+      "loss": 0.3314,
+      "step": 2607
+    },
+    {
+      "epoch": 1.246102383370169,
+      "grad_norm": 0.5165611849608196,
+      "learning_rate": 1.3552108998386683e-05,
+      "loss": 0.3408,
+      "step": 2608
+    },
+    {
+      "epoch": 1.2465802520757423,
+      "grad_norm": 0.4974036775097272,
+      "learning_rate": 1.3547185257988513e-05,
+      "loss": 0.3221,
+      "step": 2609
+    },
+    {
+      "epoch": 1.2470581207813154,
+      "grad_norm": 0.4947570822821488,
+      "learning_rate": 1.3542260533663723e-05,
+      "loss": 0.3427,
+      "step": 2610
+    },
+    {
+      "epoch": 1.2475359894868885,
+      "grad_norm": 0.48247105943498364,
+      "learning_rate": 1.3537334826778343e-05,
+      "loss": 0.33,
+      "step": 2611
+    },
+    {
+      "epoch": 1.2480138581924616,
+      "grad_norm": 0.5029602456053662,
+      "learning_rate": 1.3532408138698685e-05,
+      "loss": 0.337,
+      "step": 2612
+    },
+    {
+      "epoch": 1.2484917268980347,
+      "grad_norm": 0.5072947162416663,
+      "learning_rate": 1.3527480470791314e-05,
+      "loss": 0.3383,
+      "step": 2613
+    },
+    {
+      "epoch": 1.248969595603608,
+      "grad_norm": 0.49966515091084346,
+      "learning_rate": 1.3522551824423088e-05,
+      "loss": 0.3276,
+      "step": 2614
+    },
+    {
+      "epoch": 1.249447464309181,
+      "grad_norm": 0.4845192026566661,
+      "learning_rate": 1.351762220096112e-05,
+      "loss": 0.3403,
+      "step": 2615
+    },
+    {
+      "epoch": 1.2499253330147542,
+      "grad_norm": 0.48326585792046045,
+      "learning_rate": 1.3512691601772803e-05,
+      "loss": 0.3138,
+      "step": 2616
+    },
+    {
+      "epoch": 1.2504032017203273,
+      "grad_norm": 0.500541614774048,
+      "learning_rate": 1.3507760028225801e-05,
+      "loss": 0.3183,
+      "step": 2617
+    },
+    {
+      "epoch": 1.2508810704259004,
+      "grad_norm": 0.46031252587482197,
+      "learning_rate": 1.3502827481688041e-05,
+      "loss": 0.3235,
+      "step": 2618
+    },
+    {
+      "epoch": 1.2513589391314737,
+      "grad_norm": 0.4621265173598907,
+      "learning_rate": 1.3497893963527729e-05,
+      "loss": 0.3192,
+      "step": 2619
+    },
+    {
+      "epoch": 1.2518368078370468,
+      "grad_norm": 0.509128957862185,
+      "learning_rate": 1.3492959475113332e-05,
+      "loss": 0.343,
+      "step": 2620
+    },
+    {
+      "epoch": 1.2523146765426199,
+      "grad_norm": 0.5017135977140528,
+      "learning_rate": 1.348802401781359e-05,
+      "loss": 0.3434,
+      "step": 2621
+    },
+    {
+      "epoch": 1.2527925452481932,
+      "grad_norm": 0.48983060466434775,
+      "learning_rate": 1.3483087592997513e-05,
+      "loss": 0.3393,
+      "step": 2622
+    },
+    {
+      "epoch": 1.2532704139537663,
+      "grad_norm": 0.4917900123369827,
+      "learning_rate": 1.347815020203438e-05,
+      "loss": 0.334,
+      "step": 2623
+    },
+    {
+      "epoch": 1.2537482826593394,
+      "grad_norm": 0.48250587974454096,
+      "learning_rate": 1.3473211846293735e-05,
+      "loss": 0.331,
+      "step": 2624
+    },
+    {
+      "epoch": 1.2542261513649124,
+      "grad_norm": 0.5248984409083738,
+      "learning_rate": 1.3468272527145388e-05,
+      "loss": 0.3289,
+      "step": 2625
+    },
+    {
+      "epoch": 1.2547040200704855,
+      "grad_norm": 0.7447735031783237,
+      "learning_rate": 1.3463332245959424e-05,
+      "loss": 0.3237,
+      "step": 2626
+    },
+    {
+      "epoch": 1.2551818887760589,
+      "grad_norm": 0.4813371324594707,
+      "learning_rate": 1.3458391004106184e-05,
+      "loss": 0.341,
+      "step": 2627
+    },
+    {
+      "epoch": 1.255659757481632,
+      "grad_norm": 0.5512321025879464,
+      "learning_rate": 1.3453448802956285e-05,
+      "loss": 0.3284,
+      "step": 2628
+    },
+    {
+      "epoch": 1.256137626187205,
+      "grad_norm": 0.5150834661725827,
+      "learning_rate": 1.344850564388061e-05,
+      "loss": 0.3277,
+      "step": 2629
+    },
+    {
+      "epoch": 1.2566154948927781,
+      "grad_norm": 0.5120299966158544,
+      "learning_rate": 1.3443561528250295e-05,
+      "loss": 0.3403,
+      "step": 2630
+    },
+    {
+      "epoch": 1.2570933635983512,
+      "grad_norm": 0.4814484918217983,
+      "learning_rate": 1.3438616457436758e-05,
+      "loss": 0.3525,
+      "step": 2631
+    },
+    {
+      "epoch": 1.2575712323039245,
+      "grad_norm": 0.5120493174867615,
+      "learning_rate": 1.343367043281167e-05,
+      "loss": 0.3409,
+      "step": 2632
+    },
+    {
+      "epoch": 1.2580491010094976,
+      "grad_norm": 0.47147647454359,
+      "learning_rate": 1.3428723455746972e-05,
+      "loss": 0.3457,
+      "step": 2633
+    },
+    {
+      "epoch": 1.2585269697150707,
+      "grad_norm": 0.4898796593893291,
+      "learning_rate": 1.3423775527614871e-05,
+      "loss": 0.3284,
+      "step": 2634
+    },
+    {
+      "epoch": 1.259004838420644,
+      "grad_norm": 0.48315639065037624,
+      "learning_rate": 1.3418826649787834e-05,
+      "loss": 0.319,
+      "step": 2635
+    },
+    {
+      "epoch": 1.2594827071262171,
+      "grad_norm": 0.4518418312294045,
+      "learning_rate": 1.341387682363859e-05,
+      "loss": 0.3303,
+      "step": 2636
+    },
+    {
+      "epoch": 1.2599605758317902,
+      "grad_norm": 0.4850719475140083,
+      "learning_rate": 1.3408926050540134e-05,
+      "loss": 0.3323,
+      "step": 2637
+    },
+    {
+      "epoch": 1.2604384445373633,
+      "grad_norm": 0.4608901894307272,
+      "learning_rate": 1.3403974331865728e-05,
+      "loss": 0.3441,
+      "step": 2638
+    },
+    {
+      "epoch": 1.2609163132429364,
+      "grad_norm": 0.5002804502797722,
+      "learning_rate": 1.3399021668988882e-05,
+      "loss": 0.3091,
+      "step": 2639
+    },
+    {
+      "epoch": 1.2613941819485097,
+      "grad_norm": 0.49469049914409363,
+      "learning_rate": 1.3394068063283387e-05,
+      "loss": 0.3302,
+      "step": 2640
+    },
+    {
+      "epoch": 1.2618720506540828,
+      "grad_norm": 0.6390540371820533,
+      "learning_rate": 1.3389113516123283e-05,
+      "loss": 0.3219,
+      "step": 2641
+    },
+    {
+      "epoch": 1.262349919359656,
+      "grad_norm": 0.5039329426448652,
+      "learning_rate": 1.3384158028882866e-05,
+      "loss": 0.3264,
+      "step": 2642
+    },
+    {
+      "epoch": 1.262827788065229,
+      "grad_norm": 0.4908786517804285,
+      "learning_rate": 1.337920160293671e-05,
+      "loss": 0.3383,
+      "step": 2643
+    },
+    {
+      "epoch": 1.263305656770802,
+      "grad_norm": 0.475559094015669,
+      "learning_rate": 1.3374244239659641e-05,
+      "loss": 0.335,
+      "step": 2644
+    },
+    {
+      "epoch": 1.2637835254763754,
+      "grad_norm": 0.5008300872967887,
+      "learning_rate": 1.3369285940426737e-05,
+      "loss": 0.3163,
+      "step": 2645
+    },
+    {
+      "epoch": 1.2642613941819485,
+      "grad_norm": 0.5167558213932772,
+      "learning_rate": 1.3364326706613346e-05,
+      "loss": 0.3169,
+      "step": 2646
+    },
+    {
+      "epoch": 1.2647392628875216,
+      "grad_norm": 0.5070324119135403,
+      "learning_rate": 1.3359366539595075e-05,
+      "loss": 0.3486,
+      "step": 2647
+    },
+    {
+      "epoch": 1.265217131593095,
+      "grad_norm": 0.4926151078159255,
+      "learning_rate": 1.3354405440747783e-05,
+      "loss": 0.322,
+      "step": 2648
+    },
+    {
+      "epoch": 1.265695000298668,
+      "grad_norm": 0.531687534308957,
+      "learning_rate": 1.3349443411447591e-05,
+      "loss": 0.3295,
+      "step": 2649
+    },
+    {
+      "epoch": 1.266172869004241,
+      "grad_norm": 0.4865326711170691,
+      "learning_rate": 1.334448045307088e-05,
+      "loss": 0.321,
+      "step": 2650
+    },
+    {
+      "epoch": 1.2666507377098142,
+      "grad_norm": 0.505824257323991,
+      "learning_rate": 1.3339516566994285e-05,
+      "loss": 0.3275,
+      "step": 2651
+    },
+    {
+      "epoch": 1.2671286064153873,
+      "grad_norm": 0.4743225313426183,
+      "learning_rate": 1.3334551754594709e-05,
+      "loss": 0.3338,
+      "step": 2652
+    },
+    {
+      "epoch": 1.2676064751209606,
+      "grad_norm": 0.51248355708598,
+      "learning_rate": 1.3329586017249293e-05,
+      "loss": 0.3227,
+      "step": 2653
+    },
+    {
+      "epoch": 1.2680843438265337,
+      "grad_norm": 0.48662715453553024,
+      "learning_rate": 1.3324619356335446e-05,
+      "loss": 0.3229,
+      "step": 2654
+    },
+    {
+      "epoch": 1.2685622125321068,
+      "grad_norm": 0.5130913956035601,
+      "learning_rate": 1.331965177323084e-05,
+      "loss": 0.3297,
+      "step": 2655
+    },
+    {
+      "epoch": 1.26904008123768,
+      "grad_norm": 0.507745869575156,
+      "learning_rate": 1.3314683269313387e-05,
+      "loss": 0.3303,
+      "step": 2656
+    },
+    {
+      "epoch": 1.269517949943253,
+      "grad_norm": 0.5295327741107102,
+      "learning_rate": 1.3309713845961265e-05,
+      "loss": 0.3353,
+      "step": 2657
+    },
+    {
+      "epoch": 1.2699958186488263,
+      "grad_norm": 0.5362619437276863,
+      "learning_rate": 1.3304743504552906e-05,
+      "loss": 0.3198,
+      "step": 2658
+    },
+    {
+      "epoch": 1.2704736873543994,
+      "grad_norm": 0.7456002921673187,
+      "learning_rate": 1.3299772246466992e-05,
+      "loss": 0.3247,
+      "step": 2659
+    },
+    {
+      "epoch": 1.2709515560599725,
+      "grad_norm": 0.5057040297711566,
+      "learning_rate": 1.3294800073082464e-05,
+      "loss": 0.3366,
+      "step": 2660
+    },
+    {
+      "epoch": 1.2714294247655458,
+      "grad_norm": 0.5114205500351166,
+      "learning_rate": 1.3289826985778515e-05,
+      "loss": 0.3081,
+      "step": 2661
+    },
+    {
+      "epoch": 1.2719072934711189,
+      "grad_norm": 0.5107340207654342,
+      "learning_rate": 1.3284852985934591e-05,
+      "loss": 0.3003,
+      "step": 2662
+    },
+    {
+      "epoch": 1.272385162176692,
+      "grad_norm": 0.5064044347392543,
+      "learning_rate": 1.3279878074930394e-05,
+      "loss": 0.3368,
+      "step": 2663
+    },
+    {
+      "epoch": 1.272863030882265,
+      "grad_norm": 0.5498353953286683,
+      "learning_rate": 1.3274902254145876e-05,
+      "loss": 0.3265,
+      "step": 2664
+    },
+    {
+      "epoch": 1.2733408995878381,
+      "grad_norm": 0.527749650382651,
+      "learning_rate": 1.3269925524961237e-05,
+      "loss": 0.3413,
+      "step": 2665
+    },
+    {
+      "epoch": 1.2738187682934115,
+      "grad_norm": 0.4800823655478267,
+      "learning_rate": 1.326494788875694e-05,
+      "loss": 0.3383,
+      "step": 2666
+    },
+    {
+      "epoch": 1.2742966369989845,
+      "grad_norm": 0.5476129798251085,
+      "learning_rate": 1.3259969346913692e-05,
+      "loss": 0.3114,
+      "step": 2667
+    },
+    {
+      "epoch": 1.2747745057045576,
+      "grad_norm": 0.6895619251669788,
+      "learning_rate": 1.3254989900812452e-05,
+      "loss": 0.3184,
+      "step": 2668
+    },
+    {
+      "epoch": 1.275252374410131,
+      "grad_norm": 0.5299012176977083,
+      "learning_rate": 1.3250009551834431e-05,
+      "loss": 0.3135,
+      "step": 2669
+    },
+    {
+      "epoch": 1.275730243115704,
+      "grad_norm": 0.5236139281767733,
+      "learning_rate": 1.3245028301361086e-05,
+      "loss": 0.3399,
+      "step": 2670
+    },
+    {
+      "epoch": 1.2762081118212771,
+      "grad_norm": 0.5049494339980716,
+      "learning_rate": 1.3240046150774136e-05,
+      "loss": 0.3293,
+      "step": 2671
+    },
+    {
+      "epoch": 1.2766859805268502,
+      "grad_norm": 0.48846700131030796,
+      "learning_rate": 1.3235063101455536e-05,
+      "loss": 0.3272,
+      "step": 2672
+    },
+    {
+      "epoch": 1.2771638492324233,
+      "grad_norm": 0.534758434518732,
+      "learning_rate": 1.3230079154787497e-05,
+      "loss": 0.2966,
+      "step": 2673
+    },
+    {
+      "epoch": 1.2776417179379966,
+      "grad_norm": 0.5290754783911104,
+      "learning_rate": 1.3225094312152478e-05,
+      "loss": 0.336,
+      "step": 2674
+    },
+    {
+      "epoch": 1.2781195866435697,
+      "grad_norm": 0.5164880573517012,
+      "learning_rate": 1.3220108574933185e-05,
+      "loss": 0.3283,
+      "step": 2675
+    },
+    {
+      "epoch": 1.2785974553491428,
+      "grad_norm": 0.5334112157901388,
+      "learning_rate": 1.3215121944512576e-05,
+      "loss": 0.3334,
+      "step": 2676
+    },
+    {
+      "epoch": 1.279075324054716,
+      "grad_norm": 1.0889045439453695,
+      "learning_rate": 1.3210134422273855e-05,
+      "loss": 0.3356,
+      "step": 2677
+    },
+    {
+      "epoch": 1.279553192760289,
+      "grad_norm": 0.5519985462389302,
+      "learning_rate": 1.3205146009600472e-05,
+      "loss": 0.3169,
+      "step": 2678
+    },
+    {
+      "epoch": 1.2800310614658623,
+      "grad_norm": 0.5034434905243013,
+      "learning_rate": 1.320015670787612e-05,
+      "loss": 0.3264,
+      "step": 2679
+    },
+    {
+      "epoch": 1.2805089301714354,
+      "grad_norm": 0.4936078005967817,
+      "learning_rate": 1.3195166518484748e-05,
+      "loss": 0.3266,
+      "step": 2680
+    },
+    {
+      "epoch": 1.2809867988770085,
+      "grad_norm": 0.48996148603955003,
+      "learning_rate": 1.3190175442810547e-05,
+      "loss": 0.3149,
+      "step": 2681
+    },
+    {
+      "epoch": 1.2814646675825818,
+      "grad_norm": 0.5937307442317561,
+      "learning_rate": 1.3185183482237948e-05,
+      "loss": 0.3398,
+      "step": 2682
+    },
+    {
+      "epoch": 1.281942536288155,
+      "grad_norm": 0.5041569401113013,
+      "learning_rate": 1.318019063815164e-05,
+      "loss": 0.3533,
+      "step": 2683
+    },
+    {
+      "epoch": 1.282420404993728,
+      "grad_norm": 0.49226884478826943,
+      "learning_rate": 1.3175196911936548e-05,
+      "loss": 0.3384,
+      "step": 2684
+    },
+    {
+      "epoch": 1.282898273699301,
+      "grad_norm": 0.5095296957203082,
+      "learning_rate": 1.317020230497784e-05,
+      "loss": 0.3261,
+      "step": 2685
+    },
+    {
+      "epoch": 1.2833761424048742,
+      "grad_norm": 0.5756732067034811,
+      "learning_rate": 1.3165206818660932e-05,
+      "loss": 0.333,
+      "step": 2686
+    },
+    {
+      "epoch": 1.2838540111104475,
+      "grad_norm": 0.5299900729104405,
+      "learning_rate": 1.3160210454371489e-05,
+      "loss": 0.3262,
+      "step": 2687
+    },
+    {
+      "epoch": 1.2843318798160206,
+      "grad_norm": 0.5971638970062076,
+      "learning_rate": 1.315521321349541e-05,
+      "loss": 0.3198,
+      "step": 2688
+    },
+    {
+      "epoch": 1.2848097485215937,
+      "grad_norm": 0.4998008800910582,
+      "learning_rate": 1.3150215097418844e-05,
+      "loss": 0.3317,
+      "step": 2689
+    },
+    {
+      "epoch": 1.2852876172271668,
+      "grad_norm": 0.4930588171021608,
+      "learning_rate": 1.3145216107528178e-05,
+      "loss": 0.3055,
+      "step": 2690
+    },
+    {
+      "epoch": 1.2857654859327399,
+      "grad_norm": 0.5053649686384691,
+      "learning_rate": 1.3140216245210042e-05,
+      "loss": 0.3419,
+      "step": 2691
+    },
+    {
+      "epoch": 1.2862433546383132,
+      "grad_norm": 0.4874790099765457,
+      "learning_rate": 1.3135215511851316e-05,
+      "loss": 0.336,
+      "step": 2692
+    },
+    {
+      "epoch": 1.2867212233438863,
+      "grad_norm": 0.5212769061425828,
+      "learning_rate": 1.313021390883911e-05,
+      "loss": 0.3201,
+      "step": 2693
+    },
+    {
+      "epoch": 1.2871990920494594,
+      "grad_norm": 0.5057814940111431,
+      "learning_rate": 1.312521143756078e-05,
+      "loss": 0.3087,
+      "step": 2694
+    },
+    {
+      "epoch": 1.2876769607550327,
+      "grad_norm": 0.5430285465859317,
+      "learning_rate": 1.3120208099403926e-05,
+      "loss": 0.3223,
+      "step": 2695
+    },
+    {
+      "epoch": 1.2881548294606058,
+      "grad_norm": 0.4973193693517145,
+      "learning_rate": 1.3115203895756387e-05,
+      "loss": 0.3249,
+      "step": 2696
+    },
+    {
+      "epoch": 1.2886326981661789,
+      "grad_norm": 0.48876830415257927,
+      "learning_rate": 1.3110198828006236e-05,
+      "loss": 0.3256,
+      "step": 2697
+    },
+    {
+      "epoch": 1.289110566871752,
+      "grad_norm": 0.5073517633637624,
+      "learning_rate": 1.3105192897541792e-05,
+      "loss": 0.3369,
+      "step": 2698
+    },
+    {
+      "epoch": 1.289588435577325,
+      "grad_norm": 0.5115023299359397,
+      "learning_rate": 1.3100186105751615e-05,
+      "loss": 0.3112,
+      "step": 2699
+    },
+    {
+      "epoch": 1.2900663042828984,
+      "grad_norm": 0.5050875752293276,
+      "learning_rate": 1.3095178454024496e-05,
+      "loss": 0.3355,
+      "step": 2700
+    },
+    {
+      "epoch": 1.2905441729884715,
+      "grad_norm": 0.48487507417501574,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.3494,
+      "step": 2701
+    },
+    {
+      "epoch": 1.2910220416940446,
+      "grad_norm": 0.6454288527132577,
+      "learning_rate": 1.308516057631582e-05,
+      "loss": 0.3378,
+      "step": 2702
+    },
+    {
+      "epoch": 1.2914999103996176,
+      "grad_norm": 0.5860456115191699,
+      "learning_rate": 1.3080150353113044e-05,
+      "loss": 0.3425,
+      "step": 2703
+    },
+    {
+      "epoch": 1.2919777791051907,
+      "grad_norm": 0.45560558853026417,
+      "learning_rate": 1.3075139275530893e-05,
+      "loss": 0.3333,
+      "step": 2704
+    },
+    {
+      "epoch": 1.292455647810764,
+      "grad_norm": 0.5735696204556753,
+      "learning_rate": 1.3070127344959348e-05,
+      "loss": 0.3034,
+      "step": 2705
+    },
+    {
+      "epoch": 1.2929335165163371,
+      "grad_norm": 0.5440263491737388,
+      "learning_rate": 1.3065114562788634e-05,
+      "loss": 0.3148,
+      "step": 2706
+    },
+    {
+      "epoch": 1.2934113852219102,
+      "grad_norm": 0.4894882914596722,
+      "learning_rate": 1.3060100930409211e-05,
+      "loss": 0.3359,
+      "step": 2707
+    },
+    {
+      "epoch": 1.2938892539274836,
+      "grad_norm": 0.5571970234505838,
+      "learning_rate": 1.3055086449211768e-05,
+      "loss": 0.3288,
+      "step": 2708
+    },
+    {
+      "epoch": 1.2943671226330566,
+      "grad_norm": 0.4614963731916626,
+      "learning_rate": 1.3050071120587235e-05,
+      "loss": 0.3191,
+      "step": 2709
+    },
+    {
+      "epoch": 1.2948449913386297,
+      "grad_norm": 0.5415712777984418,
+      "learning_rate": 1.3045054945926775e-05,
+      "loss": 0.3393,
+      "step": 2710
+    },
+    {
+      "epoch": 1.2953228600442028,
+      "grad_norm": 0.5125831058498063,
+      "learning_rate": 1.3040037926621788e-05,
+      "loss": 0.3201,
+      "step": 2711
+    },
+    {
+      "epoch": 1.295800728749776,
+      "grad_norm": 0.5121610378114612,
+      "learning_rate": 1.3035020064063903e-05,
+      "loss": 0.3415,
+      "step": 2712
+    },
+    {
+      "epoch": 1.2962785974553492,
+      "grad_norm": 0.5247492831665274,
+      "learning_rate": 1.3030001359644992e-05,
+      "loss": 0.3252,
+      "step": 2713
+    },
+    {
+      "epoch": 1.2967564661609223,
+      "grad_norm": 0.9578688622282728,
+      "learning_rate": 1.302498181475715e-05,
+      "loss": 0.3308,
+      "step": 2714
+    },
+    {
+      "epoch": 1.2972343348664954,
+      "grad_norm": 0.5081119642289837,
+      "learning_rate": 1.3019961430792711e-05,
+      "loss": 0.3129,
+      "step": 2715
+    },
+    {
+      "epoch": 1.2977122035720685,
+      "grad_norm": 0.486093493880196,
+      "learning_rate": 1.3014940209144246e-05,
+      "loss": 0.3437,
+      "step": 2716
+    },
+    {
+      "epoch": 1.2981900722776416,
+      "grad_norm": 0.5533735834037948,
+      "learning_rate": 1.3009918151204546e-05,
+      "loss": 0.3192,
+      "step": 2717
+    },
+    {
+      "epoch": 1.298667940983215,
+      "grad_norm": 0.5027180369228621,
+      "learning_rate": 1.3004895258366648e-05,
+      "loss": 0.3539,
+      "step": 2718
+    },
+    {
+      "epoch": 1.299145809688788,
+      "grad_norm": 0.48700562710108203,
+      "learning_rate": 1.299987153202381e-05,
+      "loss": 0.3283,
+      "step": 2719
+    },
+    {
+      "epoch": 1.299623678394361,
+      "grad_norm": 0.4947556284065124,
+      "learning_rate": 1.2994846973569524e-05,
+      "loss": 0.3273,
+      "step": 2720
+    },
+    {
+      "epoch": 1.3001015470999344,
+      "grad_norm": 0.5278204218639244,
+      "learning_rate": 1.298982158439752e-05,
+      "loss": 0.3283,
+      "step": 2721
+    },
+    {
+      "epoch": 1.3005794158055075,
+      "grad_norm": 0.5311707177146678,
+      "learning_rate": 1.2984795365901743e-05,
+      "loss": 0.3358,
+      "step": 2722
+    },
+    {
+      "epoch": 1.3010572845110806,
+      "grad_norm": 0.5090844743344694,
+      "learning_rate": 1.2979768319476384e-05,
+      "loss": 0.3265,
+      "step": 2723
+    },
+    {
+      "epoch": 1.3015351532166537,
+      "grad_norm": 0.5069541011918202,
+      "learning_rate": 1.2974740446515858e-05,
+      "loss": 0.3422,
+      "step": 2724
+    },
+    {
+      "epoch": 1.3020130219222268,
+      "grad_norm": 0.48331918054796685,
+      "learning_rate": 1.2969711748414804e-05,
+      "loss": 0.3206,
+      "step": 2725
+    },
+    {
+      "epoch": 1.3024908906278,
+      "grad_norm": 0.5096220010199282,
+      "learning_rate": 1.2964682226568095e-05,
+      "loss": 0.3428,
+      "step": 2726
+    },
+    {
+      "epoch": 1.3029687593333732,
+      "grad_norm": 0.5373918271686221,
+      "learning_rate": 1.2959651882370835e-05,
+      "loss": 0.3182,
+      "step": 2727
+    },
+    {
+      "epoch": 1.3034466280389463,
+      "grad_norm": 0.48152585122740144,
+      "learning_rate": 1.2954620717218344e-05,
+      "loss": 0.3366,
+      "step": 2728
+    },
+    {
+      "epoch": 1.3039244967445194,
+      "grad_norm": 0.5135718903681161,
+      "learning_rate": 1.2949588732506191e-05,
+      "loss": 0.3185,
+      "step": 2729
+    },
+    {
+      "epoch": 1.3044023654500925,
+      "grad_norm": 0.5112482878597828,
+      "learning_rate": 1.2944555929630152e-05,
+      "loss": 0.3175,
+      "step": 2730
+    },
+    {
+      "epoch": 1.3048802341556658,
+      "grad_norm": 0.5064609836837177,
+      "learning_rate": 1.293952230998624e-05,
+      "loss": 0.3276,
+      "step": 2731
+    },
+    {
+      "epoch": 1.3053581028612389,
+      "grad_norm": 0.5400667483073638,
+      "learning_rate": 1.2934487874970686e-05,
+      "loss": 0.3254,
+      "step": 2732
+    },
+    {
+      "epoch": 1.305835971566812,
+      "grad_norm": 0.8384923478396311,
+      "learning_rate": 1.2929452625979966e-05,
+      "loss": 0.3323,
+      "step": 2733
+    },
+    {
+      "epoch": 1.3063138402723853,
+      "grad_norm": 0.5025088436289159,
+      "learning_rate": 1.2924416564410755e-05,
+      "loss": 0.3337,
+      "step": 2734
+    },
+    {
+      "epoch": 1.3067917089779584,
+      "grad_norm": 0.5521316915977825,
+      "learning_rate": 1.2919379691659979e-05,
+      "loss": 0.3338,
+      "step": 2735
+    },
+    {
+      "epoch": 1.3072695776835315,
+      "grad_norm": 0.5297751581493867,
+      "learning_rate": 1.2914342009124777e-05,
+      "loss": 0.3064,
+      "step": 2736
+    },
+    {
+      "epoch": 1.3077474463891046,
+      "grad_norm": 0.5163194639773859,
+      "learning_rate": 1.2909303518202502e-05,
+      "loss": 0.3227,
+      "step": 2737
+    },
+    {
+      "epoch": 1.3082253150946777,
+      "grad_norm": 0.48736656477750007,
+      "learning_rate": 1.2904264220290755e-05,
+      "loss": 0.3385,
+      "step": 2738
+    },
+    {
+      "epoch": 1.308703183800251,
+      "grad_norm": 0.5131870252317862,
+      "learning_rate": 1.2899224116787345e-05,
+      "loss": 0.3316,
+      "step": 2739
+    },
+    {
+      "epoch": 1.309181052505824,
+      "grad_norm": 0.5101460923469608,
+      "learning_rate": 1.2894183209090304e-05,
+      "loss": 0.3302,
+      "step": 2740
+    },
+    {
+      "epoch": 1.3096589212113972,
+      "grad_norm": 0.6483849369195238,
+      "learning_rate": 1.2889141498597893e-05,
+      "loss": 0.3144,
+      "step": 2741
+    },
+    {
+      "epoch": 1.3101367899169702,
+      "grad_norm": 0.4884644846621732,
+      "learning_rate": 1.2884098986708598e-05,
+      "loss": 0.3326,
+      "step": 2742
+    },
+    {
+      "epoch": 1.3106146586225433,
+      "grad_norm": 0.49766916739509637,
+      "learning_rate": 1.2879055674821112e-05,
+      "loss": 0.346,
+      "step": 2743
+    },
+    {
+      "epoch": 1.3110925273281167,
+      "grad_norm": 0.505704966484551,
+      "learning_rate": 1.2874011564334372e-05,
+      "loss": 0.3242,
+      "step": 2744
+    },
+    {
+      "epoch": 1.3115703960336897,
+      "grad_norm": 0.49985444305963106,
+      "learning_rate": 1.2868966656647522e-05,
+      "loss": 0.3169,
+      "step": 2745
+    },
+    {
+      "epoch": 1.3120482647392628,
+      "grad_norm": 0.4722113638853128,
+      "learning_rate": 1.2863920953159925e-05,
+      "loss": 0.3359,
+      "step": 2746
+    },
+    {
+      "epoch": 1.3125261334448362,
+      "grad_norm": 0.5155161827129077,
+      "learning_rate": 1.2858874455271175e-05,
+      "loss": 0.3321,
+      "step": 2747
+    },
+    {
+      "epoch": 1.3130040021504092,
+      "grad_norm": 0.5850743401779861,
+      "learning_rate": 1.2853827164381083e-05,
+      "loss": 0.3404,
+      "step": 2748
+    },
+    {
+      "epoch": 1.3134818708559823,
+      "grad_norm": 0.47900054565538824,
+      "learning_rate": 1.2848779081889675e-05,
+      "loss": 0.3299,
+      "step": 2749
+    },
+    {
+      "epoch": 1.3139597395615554,
+      "grad_norm": 0.4846014694563064,
+      "learning_rate": 1.2843730209197203e-05,
+      "loss": 0.317,
+      "step": 2750
+    },
+    {
+      "epoch": 1.3144376082671285,
+      "grad_norm": 0.46319015268026265,
+      "learning_rate": 1.283868054770413e-05,
+      "loss": 0.3649,
+      "step": 2751
+    },
+    {
+      "epoch": 1.3149154769727018,
+      "grad_norm": 0.492209610537854,
+      "learning_rate": 1.2833630098811148e-05,
+      "loss": 0.3354,
+      "step": 2752
+    },
+    {
+      "epoch": 1.315393345678275,
+      "grad_norm": 0.5469966096817243,
+      "learning_rate": 1.2828578863919163e-05,
+      "loss": 0.3318,
+      "step": 2753
+    },
+    {
+      "epoch": 1.315871214383848,
+      "grad_norm": 0.7017701345315566,
+      "learning_rate": 1.2823526844429295e-05,
+      "loss": 0.343,
+      "step": 2754
+    },
+    {
+      "epoch": 1.3163490830894211,
+      "grad_norm": 0.4870050694221161,
+      "learning_rate": 1.2818474041742885e-05,
+      "loss": 0.3287,
+      "step": 2755
+    },
+    {
+      "epoch": 1.3168269517949942,
+      "grad_norm": 0.48028268862938356,
+      "learning_rate": 1.2813420457261497e-05,
+      "loss": 0.325,
+      "step": 2756
+    },
+    {
+      "epoch": 1.3173048205005675,
+      "grad_norm": 0.47555681810939554,
+      "learning_rate": 1.2808366092386896e-05,
+      "loss": 0.3383,
+      "step": 2757
+    },
+    {
+      "epoch": 1.3177826892061406,
+      "grad_norm": 0.47978915398997984,
+      "learning_rate": 1.2803310948521083e-05,
+      "loss": 0.3344,
+      "step": 2758
+    },
+    {
+      "epoch": 1.3182605579117137,
+      "grad_norm": 0.5676291606710551,
+      "learning_rate": 1.2798255027066265e-05,
+      "loss": 0.3367,
+      "step": 2759
+    },
+    {
+      "epoch": 1.318738426617287,
+      "grad_norm": 0.47898495928629903,
+      "learning_rate": 1.2793198329424858e-05,
+      "loss": 0.344,
+      "step": 2760
+    },
+    {
+      "epoch": 1.31921629532286,
+      "grad_norm": 0.6291553763138783,
+      "learning_rate": 1.278814085699951e-05,
+      "loss": 0.3288,
+      "step": 2761
+    },
+    {
+      "epoch": 1.3196941640284332,
+      "grad_norm": 0.5128423783465226,
+      "learning_rate": 1.2783082611193068e-05,
+      "loss": 0.3326,
+      "step": 2762
+    },
+    {
+      "epoch": 1.3201720327340063,
+      "grad_norm": 0.4966607745937759,
+      "learning_rate": 1.2778023593408601e-05,
+      "loss": 0.335,
+      "step": 2763
+    },
+    {
+      "epoch": 1.3206499014395794,
+      "grad_norm": 0.4621870832599053,
+      "learning_rate": 1.2772963805049395e-05,
+      "loss": 0.3169,
+      "step": 2764
+    },
+    {
+      "epoch": 1.3211277701451527,
+      "grad_norm": 0.5336568579759654,
+      "learning_rate": 1.2767903247518945e-05,
+      "loss": 0.3197,
+      "step": 2765
+    },
+    {
+      "epoch": 1.3216056388507258,
+      "grad_norm": 0.4998758991079805,
+      "learning_rate": 1.2762841922220956e-05,
+      "loss": 0.3101,
+      "step": 2766
+    },
+    {
+      "epoch": 1.3220835075562989,
+      "grad_norm": 0.5061171116591409,
+      "learning_rate": 1.2757779830559353e-05,
+      "loss": 0.3191,
+      "step": 2767
+    },
+    {
+      "epoch": 1.322561376261872,
+      "grad_norm": 0.5779265310597284,
+      "learning_rate": 1.2752716973938272e-05,
+      "loss": 0.3102,
+      "step": 2768
+    },
+    {
+      "epoch": 1.323039244967445,
+      "grad_norm": 0.4806329769254405,
+      "learning_rate": 1.274765335376206e-05,
+      "loss": 0.3441,
+      "step": 2769
+    },
+    {
+      "epoch": 1.3235171136730184,
+      "grad_norm": 0.5262517952171601,
+      "learning_rate": 1.2742588971435276e-05,
+      "loss": 0.3289,
+      "step": 2770
+    },
+    {
+      "epoch": 1.3239949823785915,
+      "grad_norm": 0.5469131458946188,
+      "learning_rate": 1.273752382836269e-05,
+      "loss": 0.3281,
+      "step": 2771
+    },
+    {
+      "epoch": 1.3244728510841646,
+      "grad_norm": 0.4981852013260321,
+      "learning_rate": 1.2732457925949282e-05,
+      "loss": 0.3339,
+      "step": 2772
+    },
+    {
+      "epoch": 1.3249507197897379,
+      "grad_norm": 0.4995218099055462,
+      "learning_rate": 1.2727391265600248e-05,
+      "loss": 0.3217,
+      "step": 2773
+    },
+    {
+      "epoch": 1.325428588495311,
+      "grad_norm": 0.48442166594011293,
+      "learning_rate": 1.2722323848720985e-05,
+      "loss": 0.3198,
+      "step": 2774
+    },
+    {
+      "epoch": 1.325906457200884,
+      "grad_norm": 0.5288357240858039,
+      "learning_rate": 1.2717255676717106e-05,
+      "loss": 0.3265,
+      "step": 2775
+    },
+    {
+      "epoch": 1.3263843259064572,
+      "grad_norm": 0.47363656909947727,
+      "learning_rate": 1.2712186750994437e-05,
+      "loss": 0.3144,
+      "step": 2776
+    },
+    {
+      "epoch": 1.3268621946120303,
+      "grad_norm": 0.5621473769140592,
+      "learning_rate": 1.2707117072959004e-05,
+      "loss": 0.3201,
+      "step": 2777
+    },
+    {
+      "epoch": 1.3273400633176036,
+      "grad_norm": 0.516657131750651,
+      "learning_rate": 1.2702046644017045e-05,
+      "loss": 0.3407,
+      "step": 2778
+    },
+    {
+      "epoch": 1.3278179320231767,
+      "grad_norm": 0.529619985203502,
+      "learning_rate": 1.2696975465575016e-05,
+      "loss": 0.3256,
+      "step": 2779
+    },
+    {
+      "epoch": 1.3282958007287498,
+      "grad_norm": 0.5434444774186435,
+      "learning_rate": 1.2691903539039563e-05,
+      "loss": 0.3228,
+      "step": 2780
+    },
+    {
+      "epoch": 1.3287736694343228,
+      "grad_norm": 0.565815798825105,
+      "learning_rate": 1.2686830865817552e-05,
+      "loss": 0.3327,
+      "step": 2781
+    },
+    {
+      "epoch": 1.329251538139896,
+      "grad_norm": 0.5113159211723257,
+      "learning_rate": 1.2681757447316057e-05,
+      "loss": 0.336,
+      "step": 2782
+    },
+    {
+      "epoch": 1.3297294068454693,
+      "grad_norm": 0.48646729323156335,
+      "learning_rate": 1.2676683284942348e-05,
+      "loss": 0.3377,
+      "step": 2783
+    },
+    {
+      "epoch": 1.3302072755510423,
+      "grad_norm": 0.4730669634348989,
+      "learning_rate": 1.267160838010391e-05,
+      "loss": 0.3511,
+      "step": 2784
+    },
+    {
+      "epoch": 1.3306851442566154,
+      "grad_norm": 0.6446175620584933,
+      "learning_rate": 1.2666532734208437e-05,
+      "loss": 0.3382,
+      "step": 2785
+    },
+    {
+      "epoch": 1.3311630129621888,
+      "grad_norm": 0.5040172790962374,
+      "learning_rate": 1.2661456348663822e-05,
+      "loss": 0.3291,
+      "step": 2786
+    },
+    {
+      "epoch": 1.3316408816677618,
+      "grad_norm": 0.6517602921578377,
+      "learning_rate": 1.265637922487816e-05,
+      "loss": 0.3216,
+      "step": 2787
+    },
+    {
+      "epoch": 1.332118750373335,
+      "grad_norm": 0.4648517246826999,
+      "learning_rate": 1.2651301364259761e-05,
+      "loss": 0.3355,
+      "step": 2788
+    },
+    {
+      "epoch": 1.332596619078908,
+      "grad_norm": 0.48877805021759047,
+      "learning_rate": 1.2646222768217129e-05,
+      "loss": 0.3331,
+      "step": 2789
+    },
+    {
+      "epoch": 1.3330744877844811,
+      "grad_norm": 0.4820080738115575,
+      "learning_rate": 1.264114343815898e-05,
+      "loss": 0.3148,
+      "step": 2790
+    },
+    {
+      "epoch": 1.3335523564900544,
+      "grad_norm": 0.4842200928813744,
+      "learning_rate": 1.2636063375494233e-05,
+      "loss": 0.3271,
+      "step": 2791
+    },
+    {
+      "epoch": 1.3340302251956275,
+      "grad_norm": 0.48858773566706376,
+      "learning_rate": 1.2630982581632003e-05,
+      "loss": 0.3306,
+      "step": 2792
+    },
+    {
+      "epoch": 1.3345080939012006,
+      "grad_norm": 0.49277764185033907,
+      "learning_rate": 1.2625901057981613e-05,
+      "loss": 0.3288,
+      "step": 2793
+    },
+    {
+      "epoch": 1.3349859626067737,
+      "grad_norm": 0.5015471955508418,
+      "learning_rate": 1.2620818805952595e-05,
+      "loss": 0.315,
+      "step": 2794
+    },
+    {
+      "epoch": 1.3354638313123468,
+      "grad_norm": 0.4703720797913821,
+      "learning_rate": 1.2615735826954664e-05,
+      "loss": 0.3377,
+      "step": 2795
+    },
+    {
+      "epoch": 1.3359417000179201,
+      "grad_norm": 0.6095651211109612,
+      "learning_rate": 1.2610652122397762e-05,
+      "loss": 0.3285,
+      "step": 2796
+    },
+    {
+      "epoch": 1.3364195687234932,
+      "grad_norm": 0.4528886888495506,
+      "learning_rate": 1.260556769369201e-05,
+      "loss": 0.3222,
+      "step": 2797
+    },
+    {
+      "epoch": 1.3368974374290663,
+      "grad_norm": 0.46151024325444917,
+      "learning_rate": 1.2600482542247738e-05,
+      "loss": 0.3311,
+      "step": 2798
+    },
+    {
+      "epoch": 1.3373753061346396,
+      "grad_norm": 0.5144167233653287,
+      "learning_rate": 1.2595396669475486e-05,
+      "loss": 0.3292,
+      "step": 2799
+    },
+    {
+      "epoch": 1.3378531748402127,
+      "grad_norm": 0.5408339927194712,
+      "learning_rate": 1.2590310076785974e-05,
+      "loss": 0.3179,
+      "step": 2800
+    },
+    {
+      "epoch": 1.3383310435457858,
+      "grad_norm": 0.47650637067616053,
+      "learning_rate": 1.258522276559014e-05,
+      "loss": 0.3355,
+      "step": 2801
+    },
+    {
+      "epoch": 1.338808912251359,
+      "grad_norm": 0.525246628210698,
+      "learning_rate": 1.2580134737299117e-05,
+      "loss": 0.3481,
+      "step": 2802
+    },
+    {
+      "epoch": 1.339286780956932,
+      "grad_norm": 0.47691152981429813,
+      "learning_rate": 1.2575045993324227e-05,
+      "loss": 0.3122,
+      "step": 2803
+    },
+    {
+      "epoch": 1.3397646496625053,
+      "grad_norm": 0.5092137433283072,
+      "learning_rate": 1.2569956535077004e-05,
+      "loss": 0.3324,
+      "step": 2804
+    },
+    {
+      "epoch": 1.3402425183680784,
+      "grad_norm": 0.4952198893971536,
+      "learning_rate": 1.256486636396917e-05,
+      "loss": 0.338,
+      "step": 2805
+    },
+    {
+      "epoch": 1.3407203870736515,
+      "grad_norm": 0.49148265729272733,
+      "learning_rate": 1.255977548141265e-05,
+      "loss": 0.3133,
+      "step": 2806
+    },
+    {
+      "epoch": 1.3411982557792246,
+      "grad_norm": 0.5396008187377243,
+      "learning_rate": 1.2554683888819565e-05,
+      "loss": 0.3187,
+      "step": 2807
+    },
+    {
+      "epoch": 1.3416761244847977,
+      "grad_norm": 0.4824817484621732,
+      "learning_rate": 1.2549591587602237e-05,
+      "loss": 0.3304,
+      "step": 2808
+    },
+    {
+      "epoch": 1.342153993190371,
+      "grad_norm": 0.5081977637354367,
+      "learning_rate": 1.2544498579173172e-05,
+      "loss": 0.3142,
+      "step": 2809
+    },
+    {
+      "epoch": 1.342631861895944,
+      "grad_norm": 0.5058206945057618,
+      "learning_rate": 1.2539404864945087e-05,
+      "loss": 0.317,
+      "step": 2810
+    },
+    {
+      "epoch": 1.3431097306015172,
+      "grad_norm": 0.5153353255064529,
+      "learning_rate": 1.2534310446330888e-05,
+      "loss": 0.3176,
+      "step": 2811
+    },
+    {
+      "epoch": 1.3435875993070905,
+      "grad_norm": 0.5318910708535983,
+      "learning_rate": 1.2529215324743673e-05,
+      "loss": 0.308,
+      "step": 2812
+    },
+    {
+      "epoch": 1.3440654680126636,
+      "grad_norm": 0.496994923710053,
+      "learning_rate": 1.2524119501596743e-05,
+      "loss": 0.3226,
+      "step": 2813
+    },
+    {
+      "epoch": 1.3445433367182367,
+      "grad_norm": 0.5165301727279924,
+      "learning_rate": 1.251902297830359e-05,
+      "loss": 0.336,
+      "step": 2814
+    },
+    {
+      "epoch": 1.3450212054238098,
+      "grad_norm": 0.5340709701111669,
+      "learning_rate": 1.2513925756277894e-05,
+      "loss": 0.3234,
+      "step": 2815
+    },
+    {
+      "epoch": 1.3454990741293829,
+      "grad_norm": 0.4828618104957075,
+      "learning_rate": 1.250882783693354e-05,
+      "loss": 0.3006,
+      "step": 2816
+    },
+    {
+      "epoch": 1.3459769428349562,
+      "grad_norm": 0.4785810307178594,
+      "learning_rate": 1.25037292216846e-05,
+      "loss": 0.3301,
+      "step": 2817
+    },
+    {
+      "epoch": 1.3464548115405293,
+      "grad_norm": 0.5032661037822173,
+      "learning_rate": 1.2498629911945333e-05,
+      "loss": 0.3262,
+      "step": 2818
+    },
+    {
+      "epoch": 1.3469326802461024,
+      "grad_norm": 0.48799683567616237,
+      "learning_rate": 1.249352990913021e-05,
+      "loss": 0.3166,
+      "step": 2819
+    },
+    {
+      "epoch": 1.3474105489516754,
+      "grad_norm": 0.5225514388182845,
+      "learning_rate": 1.2488429214653871e-05,
+      "loss": 0.3338,
+      "step": 2820
+    },
+    {
+      "epoch": 1.3478884176572485,
+      "grad_norm": 0.4688356836854529,
+      "learning_rate": 1.2483327829931167e-05,
+      "loss": 0.3117,
+      "step": 2821
+    },
+    {
+      "epoch": 1.3483662863628219,
+      "grad_norm": 0.5582134330183454,
+      "learning_rate": 1.2478225756377127e-05,
+      "loss": 0.3239,
+      "step": 2822
+    },
+    {
+      "epoch": 1.348844155068395,
+      "grad_norm": 0.4909856374763454,
+      "learning_rate": 1.2473122995406976e-05,
+      "loss": 0.3374,
+      "step": 2823
+    },
+    {
+      "epoch": 1.349322023773968,
+      "grad_norm": 0.5121950978077848,
+      "learning_rate": 1.2468019548436132e-05,
+      "loss": 0.3232,
+      "step": 2824
+    },
+    {
+      "epoch": 1.3497998924795414,
+      "grad_norm": 0.5347242883406231,
+      "learning_rate": 1.24629154168802e-05,
+      "loss": 0.3201,
+      "step": 2825
+    },
+    {
+      "epoch": 1.3502777611851144,
+      "grad_norm": 0.7886377488771786,
+      "learning_rate": 1.2457810602154975e-05,
+      "loss": 0.3208,
+      "step": 2826
+    },
+    {
+      "epoch": 1.3507556298906875,
+      "grad_norm": 0.48634722610155756,
+      "learning_rate": 1.2452705105676448e-05,
+      "loss": 0.3164,
+      "step": 2827
+    },
+    {
+      "epoch": 1.3512334985962606,
+      "grad_norm": 0.49071957483260065,
+      "learning_rate": 1.244759892886079e-05,
+      "loss": 0.3365,
+      "step": 2828
+    },
+    {
+      "epoch": 1.3517113673018337,
+      "grad_norm": 0.4953873498833253,
+      "learning_rate": 1.2442492073124359e-05,
+      "loss": 0.321,
+      "step": 2829
+    },
+    {
+      "epoch": 1.352189236007407,
+      "grad_norm": 0.5251226282353867,
+      "learning_rate": 1.2437384539883715e-05,
+      "loss": 0.3327,
+      "step": 2830
+    },
+    {
+      "epoch": 1.3526671047129801,
+      "grad_norm": 0.5013267329927095,
+      "learning_rate": 1.2432276330555592e-05,
+      "loss": 0.3445,
+      "step": 2831
+    },
+    {
+      "epoch": 1.3531449734185532,
+      "grad_norm": 0.8028493838704872,
+      "learning_rate": 1.2427167446556922e-05,
+      "loss": 0.3166,
+      "step": 2832
+    },
+    {
+      "epoch": 1.3536228421241263,
+      "grad_norm": 0.6378917954070072,
+      "learning_rate": 1.2422057889304814e-05,
+      "loss": 0.3396,
+      "step": 2833
+    },
+    {
+      "epoch": 1.3541007108296994,
+      "grad_norm": 0.4859271357599526,
+      "learning_rate": 1.2416947660216576e-05,
+      "loss": 0.3371,
+      "step": 2834
+    },
+    {
+      "epoch": 1.3545785795352727,
+      "grad_norm": 0.5227282184131874,
+      "learning_rate": 1.2411836760709686e-05,
+      "loss": 0.327,
+      "step": 2835
+    },
+    {
+      "epoch": 1.3550564482408458,
+      "grad_norm": 0.5208799708860445,
+      "learning_rate": 1.2406725192201828e-05,
+      "loss": 0.3254,
+      "step": 2836
+    },
+    {
+      "epoch": 1.355534316946419,
+      "grad_norm": 0.4795223980334988,
+      "learning_rate": 1.2401612956110853e-05,
+      "loss": 0.3111,
+      "step": 2837
+    },
+    {
+      "epoch": 1.3560121856519922,
+      "grad_norm": 0.4963240280920323,
+      "learning_rate": 1.2396500053854808e-05,
+      "loss": 0.3259,
+      "step": 2838
+    },
+    {
+      "epoch": 1.3564900543575653,
+      "grad_norm": 0.5039449229996626,
+      "learning_rate": 1.2391386486851922e-05,
+      "loss": 0.3197,
+      "step": 2839
+    },
+    {
+      "epoch": 1.3569679230631384,
+      "grad_norm": 0.5317251552116927,
+      "learning_rate": 1.2386272256520606e-05,
+      "loss": 0.3197,
+      "step": 2840
+    },
+    {
+      "epoch": 1.3574457917687115,
+      "grad_norm": 0.48744894671494793,
+      "learning_rate": 1.2381157364279462e-05,
+      "loss": 0.3035,
+      "step": 2841
+    },
+    {
+      "epoch": 1.3579236604742846,
+      "grad_norm": 0.5025271430359258,
+      "learning_rate": 1.2376041811547268e-05,
+      "loss": 0.3314,
+      "step": 2842
+    },
+    {
+      "epoch": 1.358401529179858,
+      "grad_norm": 0.5167243594092309,
+      "learning_rate": 1.2370925599742987e-05,
+      "loss": 0.3226,
+      "step": 2843
+    },
+    {
+      "epoch": 1.358879397885431,
+      "grad_norm": 0.525633231146885,
+      "learning_rate": 1.2365808730285764e-05,
+      "loss": 0.3359,
+      "step": 2844
+    },
+    {
+      "epoch": 1.359357266591004,
+      "grad_norm": 0.4991777969329675,
+      "learning_rate": 1.2360691204594937e-05,
+      "loss": 0.3209,
+      "step": 2845
+    },
+    {
+      "epoch": 1.3598351352965774,
+      "grad_norm": 0.5286964966712626,
+      "learning_rate": 1.2355573024090009e-05,
+      "loss": 0.3243,
+      "step": 2846
+    },
+    {
+      "epoch": 1.3603130040021503,
+      "grad_norm": 0.48540137645110326,
+      "learning_rate": 1.2350454190190675e-05,
+      "loss": 0.3146,
+      "step": 2847
+    },
+    {
+      "epoch": 1.3607908727077236,
+      "grad_norm": 0.526573346210916,
+      "learning_rate": 1.2345334704316811e-05,
+      "loss": 0.3272,
+      "step": 2848
+    },
+    {
+      "epoch": 1.3612687414132967,
+      "grad_norm": 0.5248630946951779,
+      "learning_rate": 1.2340214567888472e-05,
+      "loss": 0.3318,
+      "step": 2849
+    },
+    {
+      "epoch": 1.3617466101188698,
+      "grad_norm": 0.49262488475200295,
+      "learning_rate": 1.2335093782325889e-05,
+      "loss": 0.3335,
+      "step": 2850
+    },
+    {
+      "epoch": 1.362224478824443,
+      "grad_norm": 0.47664312818269816,
+      "learning_rate": 1.2329972349049481e-05,
+      "loss": 0.3357,
+      "step": 2851
+    },
+    {
+      "epoch": 1.3627023475300162,
+      "grad_norm": 0.49333473606642414,
+      "learning_rate": 1.2324850269479847e-05,
+      "loss": 0.3193,
+      "step": 2852
+    },
+    {
+      "epoch": 1.3631802162355893,
+      "grad_norm": 0.5352064541201343,
+      "learning_rate": 1.2319727545037753e-05,
+      "loss": 0.3372,
+      "step": 2853
+    },
+    {
+      "epoch": 1.3636580849411624,
+      "grad_norm": 0.5155080301298953,
+      "learning_rate": 1.2314604177144164e-05,
+      "loss": 0.3165,
+      "step": 2854
+    },
+    {
+      "epoch": 1.3641359536467355,
+      "grad_norm": 0.5556617290050194,
+      "learning_rate": 1.2309480167220203e-05,
+      "loss": 0.3062,
+      "step": 2855
+    },
+    {
+      "epoch": 1.3646138223523088,
+      "grad_norm": 0.5075291890940449,
+      "learning_rate": 1.230435551668718e-05,
+      "loss": 0.3345,
+      "step": 2856
+    },
+    {
+      "epoch": 1.3650916910578819,
+      "grad_norm": 0.5139412287417211,
+      "learning_rate": 1.2299230226966587e-05,
+      "loss": 0.3351,
+      "step": 2857
+    },
+    {
+      "epoch": 1.365569559763455,
+      "grad_norm": 0.5335170451249204,
+      "learning_rate": 1.2294104299480085e-05,
+      "loss": 0.3316,
+      "step": 2858
+    },
+    {
+      "epoch": 1.3660474284690283,
+      "grad_norm": 0.5623627174173891,
+      "learning_rate": 1.2288977735649518e-05,
+      "loss": 0.3182,
+      "step": 2859
+    },
+    {
+      "epoch": 1.3665252971746014,
+      "grad_norm": 0.48937860284820955,
+      "learning_rate": 1.2283850536896907e-05,
+      "loss": 0.3301,
+      "step": 2860
+    },
+    {
+      "epoch": 1.3670031658801745,
+      "grad_norm": 0.4993453513384397,
+      "learning_rate": 1.2278722704644439e-05,
+      "loss": 0.3349,
+      "step": 2861
+    },
+    {
+      "epoch": 1.3674810345857475,
+      "grad_norm": 0.5215921358568462,
+      "learning_rate": 1.227359424031449e-05,
+      "loss": 0.3166,
+      "step": 2862
+    },
+    {
+      "epoch": 1.3679589032913206,
+      "grad_norm": 0.473824073344398,
+      "learning_rate": 1.2268465145329607e-05,
+      "loss": 0.329,
+      "step": 2863
+    },
+    {
+      "epoch": 1.368436771996894,
+      "grad_norm": 0.5100010802249643,
+      "learning_rate": 1.2263335421112505e-05,
+      "loss": 0.3178,
+      "step": 2864
+    },
+    {
+      "epoch": 1.368914640702467,
+      "grad_norm": 0.49849508667435044,
+      "learning_rate": 1.2258205069086082e-05,
+      "loss": 0.3331,
+      "step": 2865
+    },
+    {
+      "epoch": 1.3693925094080401,
+      "grad_norm": 0.5169260084417553,
+      "learning_rate": 1.2253074090673408e-05,
+      "loss": 0.3419,
+      "step": 2866
+    },
+    {
+      "epoch": 1.3698703781136132,
+      "grad_norm": 0.4756763753549606,
+      "learning_rate": 1.2247942487297724e-05,
+      "loss": 0.3263,
+      "step": 2867
+    },
+    {
+      "epoch": 1.3703482468191863,
+      "grad_norm": 0.492096668882402,
+      "learning_rate": 1.2242810260382446e-05,
+      "loss": 0.3301,
+      "step": 2868
+    },
+    {
+      "epoch": 1.3708261155247596,
+      "grad_norm": 0.49548035638201127,
+      "learning_rate": 1.2237677411351165e-05,
+      "loss": 0.3235,
+      "step": 2869
+    },
+    {
+      "epoch": 1.3713039842303327,
+      "grad_norm": 0.5095279037269734,
+      "learning_rate": 1.2232543941627641e-05,
+      "loss": 0.3342,
+      "step": 2870
+    },
+    {
+      "epoch": 1.3717818529359058,
+      "grad_norm": 0.45819233635275786,
+      "learning_rate": 1.2227409852635811e-05,
+      "loss": 0.3265,
+      "step": 2871
+    },
+    {
+      "epoch": 1.3722597216414791,
+      "grad_norm": 0.4923000662072581,
+      "learning_rate": 1.2222275145799778e-05,
+      "loss": 0.3217,
+      "step": 2872
+    },
+    {
+      "epoch": 1.3727375903470522,
+      "grad_norm": 0.5112507834075313,
+      "learning_rate": 1.2217139822543819e-05,
+      "loss": 0.3066,
+      "step": 2873
+    },
+    {
+      "epoch": 1.3732154590526253,
+      "grad_norm": 0.4628910419474185,
+      "learning_rate": 1.2212003884292388e-05,
+      "loss": 0.316,
+      "step": 2874
+    },
+    {
+      "epoch": 1.3736933277581984,
+      "grad_norm": 0.45732279246199825,
+      "learning_rate": 1.2206867332470091e-05,
+      "loss": 0.3165,
+      "step": 2875
+    },
+    {
+      "epoch": 1.3741711964637715,
+      "grad_norm": 0.4936353320658925,
+      "learning_rate": 1.2201730168501729e-05,
+      "loss": 0.3251,
+      "step": 2876
+    },
+    {
+      "epoch": 1.3746490651693448,
+      "grad_norm": 0.48905812990241215,
+      "learning_rate": 1.2196592393812257e-05,
+      "loss": 0.3374,
+      "step": 2877
+    },
+    {
+      "epoch": 1.375126933874918,
+      "grad_norm": 0.45753365931472645,
+      "learning_rate": 1.2191454009826798e-05,
+      "loss": 0.3242,
+      "step": 2878
+    },
+    {
+      "epoch": 1.375604802580491,
+      "grad_norm": 0.5299641397547704,
+      "learning_rate": 1.2186315017970656e-05,
+      "loss": 0.3065,
+      "step": 2879
+    },
+    {
+      "epoch": 1.376082671286064,
+      "grad_norm": 0.4839322939919366,
+      "learning_rate": 1.2181175419669293e-05,
+      "loss": 0.317,
+      "step": 2880
+    },
+    {
+      "epoch": 1.3765605399916372,
+      "grad_norm": 0.48836787664932607,
+      "learning_rate": 1.2176035216348345e-05,
+      "loss": 0.3293,
+      "step": 2881
+    },
+    {
+      "epoch": 1.3770384086972105,
+      "grad_norm": 0.49187662653450037,
+      "learning_rate": 1.2170894409433612e-05,
+      "loss": 0.3438,
+      "step": 2882
+    },
+    {
+      "epoch": 1.3775162774027836,
+      "grad_norm": 0.4818132985229131,
+      "learning_rate": 1.2165753000351064e-05,
+      "loss": 0.3338,
+      "step": 2883
+    },
+    {
+      "epoch": 1.3779941461083567,
+      "grad_norm": 0.4742707254068882,
+      "learning_rate": 1.2160610990526836e-05,
+      "loss": 0.3123,
+      "step": 2884
+    },
+    {
+      "epoch": 1.37847201481393,
+      "grad_norm": 0.475050796561632,
+      "learning_rate": 1.215546838138723e-05,
+      "loss": 0.3223,
+      "step": 2885
+    },
+    {
+      "epoch": 1.378949883519503,
+      "grad_norm": 0.5011042293446867,
+      "learning_rate": 1.215032517435872e-05,
+      "loss": 0.3261,
+      "step": 2886
+    },
+    {
+      "epoch": 1.3794277522250762,
+      "grad_norm": 0.483985527438474,
+      "learning_rate": 1.2145181370867936e-05,
+      "loss": 0.3186,
+      "step": 2887
+    },
+    {
+      "epoch": 1.3799056209306493,
+      "grad_norm": 0.4541289680253298,
+      "learning_rate": 1.2140036972341683e-05,
+      "loss": 0.3343,
+      "step": 2888
+    },
+    {
+      "epoch": 1.3803834896362224,
+      "grad_norm": 0.5073229100491303,
+      "learning_rate": 1.213489198020692e-05,
+      "loss": 0.331,
+      "step": 2889
+    },
+    {
+      "epoch": 1.3808613583417957,
+      "grad_norm": 0.5173324784175866,
+      "learning_rate": 1.212974639589078e-05,
+      "loss": 0.3313,
+      "step": 2890
+    },
+    {
+      "epoch": 1.3813392270473688,
+      "grad_norm": 0.4874640856604241,
+      "learning_rate": 1.2124600220820562e-05,
+      "loss": 0.3372,
+      "step": 2891
+    },
+    {
+      "epoch": 1.3818170957529419,
+      "grad_norm": 0.4805367156153807,
+      "learning_rate": 1.2119453456423718e-05,
+      "loss": 0.3448,
+      "step": 2892
+    },
+    {
+      "epoch": 1.382294964458515,
+      "grad_norm": 0.48556515258960525,
+      "learning_rate": 1.211430610412787e-05,
+      "loss": 0.3233,
+      "step": 2893
+    },
+    {
+      "epoch": 1.382772833164088,
+      "grad_norm": 0.5315828725085151,
+      "learning_rate": 1.2109158165360805e-05,
+      "loss": 0.3142,
+      "step": 2894
+    },
+    {
+      "epoch": 1.3832507018696614,
+      "grad_norm": 0.48141420283737696,
+      "learning_rate": 1.2104009641550472e-05,
+      "loss": 0.3218,
+      "step": 2895
+    },
+    {
+      "epoch": 1.3837285705752345,
+      "grad_norm": 0.4990446188800693,
+      "learning_rate": 1.2098860534124976e-05,
+      "loss": 0.3286,
+      "step": 2896
+    },
+    {
+      "epoch": 1.3842064392808076,
+      "grad_norm": 0.49905032499820146,
+      "learning_rate": 1.2093710844512594e-05,
+      "loss": 0.3273,
+      "step": 2897
+    },
+    {
+      "epoch": 1.3846843079863809,
+      "grad_norm": 0.5206266774252051,
+      "learning_rate": 1.2088560574141754e-05,
+      "loss": 0.3233,
+      "step": 2898
+    },
+    {
+      "epoch": 1.385162176691954,
+      "grad_norm": 0.47976992992949313,
+      "learning_rate": 1.2083409724441054e-05,
+      "loss": 0.3318,
+      "step": 2899
+    },
+    {
+      "epoch": 1.385640045397527,
+      "grad_norm": 0.4858512164454944,
+      "learning_rate": 1.2078258296839245e-05,
+      "loss": 0.3497,
+      "step": 2900
+    },
+    {
+      "epoch": 1.3861179141031001,
+      "grad_norm": 0.5034404806016024,
+      "learning_rate": 1.2073106292765247e-05,
+      "loss": 0.3279,
+      "step": 2901
+    },
+    {
+      "epoch": 1.3865957828086732,
+      "grad_norm": 0.5079287528998049,
+      "learning_rate": 1.2067953713648126e-05,
+      "loss": 0.3344,
+      "step": 2902
+    },
+    {
+      "epoch": 1.3870736515142466,
+      "grad_norm": 0.47886319969478675,
+      "learning_rate": 1.206280056091713e-05,
+      "loss": 0.3384,
+      "step": 2903
+    },
+    {
+      "epoch": 1.3875515202198196,
+      "grad_norm": 0.5187096003859187,
+      "learning_rate": 1.2057646836001641e-05,
+      "loss": 0.3281,
+      "step": 2904
+    },
+    {
+      "epoch": 1.3880293889253927,
+      "grad_norm": 0.4615577488871791,
+      "learning_rate": 1.2052492540331218e-05,
+      "loss": 0.3235,
+      "step": 2905
+    },
+    {
+      "epoch": 1.3885072576309658,
+      "grad_norm": 0.48313478967101564,
+      "learning_rate": 1.2047337675335571e-05,
+      "loss": 0.3167,
+      "step": 2906
+    },
+    {
+      "epoch": 1.388985126336539,
+      "grad_norm": 0.48859241370609713,
+      "learning_rate": 1.2042182242444567e-05,
+      "loss": 0.3285,
+      "step": 2907
+    },
+    {
+      "epoch": 1.3894629950421122,
+      "grad_norm": 0.494594083600188,
+      "learning_rate": 1.203702624308823e-05,
+      "loss": 0.3207,
+      "step": 2908
+    },
+    {
+      "epoch": 1.3899408637476853,
+      "grad_norm": 0.46974833583616593,
+      "learning_rate": 1.2031869678696748e-05,
+      "loss": 0.3247,
+      "step": 2909
+    },
+    {
+      "epoch": 1.3904187324532584,
+      "grad_norm": 0.5040204897201126,
+      "learning_rate": 1.2026712550700457e-05,
+      "loss": 0.3303,
+      "step": 2910
+    },
+    {
+      "epoch": 1.3908966011588317,
+      "grad_norm": 0.47910736507397716,
+      "learning_rate": 1.2021554860529856e-05,
+      "loss": 0.314,
+      "step": 2911
+    },
+    {
+      "epoch": 1.3913744698644048,
+      "grad_norm": 0.4908900810546827,
+      "learning_rate": 1.2016396609615597e-05,
+      "loss": 0.3191,
+      "step": 2912
+    },
+    {
+      "epoch": 1.391852338569978,
+      "grad_norm": 0.4902328178847632,
+      "learning_rate": 1.2011237799388486e-05,
+      "loss": 0.3063,
+      "step": 2913
+    },
+    {
+      "epoch": 1.392330207275551,
+      "grad_norm": 0.48448162975602765,
+      "learning_rate": 1.2006078431279486e-05,
+      "loss": 0.334,
+      "step": 2914
+    },
+    {
+      "epoch": 1.392808075981124,
+      "grad_norm": 0.4975553113261312,
+      "learning_rate": 1.200091850671972e-05,
+      "loss": 0.3086,
+      "step": 2915
+    },
+    {
+      "epoch": 1.3932859446866974,
+      "grad_norm": 0.47682406789705983,
+      "learning_rate": 1.1995758027140451e-05,
+      "loss": 0.3267,
+      "step": 2916
+    },
+    {
+      "epoch": 1.3937638133922705,
+      "grad_norm": 0.4874223745730435,
+      "learning_rate": 1.1990596993973112e-05,
+      "loss": 0.3119,
+      "step": 2917
+    },
+    {
+      "epoch": 1.3942416820978436,
+      "grad_norm": 0.4853795113192791,
+      "learning_rate": 1.1985435408649281e-05,
+      "loss": 0.3346,
+      "step": 2918
+    },
+    {
+      "epoch": 1.3947195508034167,
+      "grad_norm": 0.46269214125213426,
+      "learning_rate": 1.1980273272600687e-05,
+      "loss": 0.3341,
+      "step": 2919
+    },
+    {
+      "epoch": 1.3951974195089898,
+      "grad_norm": 0.5053092201197434,
+      "learning_rate": 1.1975110587259222e-05,
+      "loss": 0.3229,
+      "step": 2920
+    },
+    {
+      "epoch": 1.395675288214563,
+      "grad_norm": 0.4944091898145227,
+      "learning_rate": 1.1969947354056918e-05,
+      "loss": 0.3285,
+      "step": 2921
+    },
+    {
+      "epoch": 1.3961531569201362,
+      "grad_norm": 0.49386779198974684,
+      "learning_rate": 1.1964783574425969e-05,
+      "loss": 0.3212,
+      "step": 2922
+    },
+    {
+      "epoch": 1.3966310256257093,
+      "grad_norm": 0.5269873184728784,
+      "learning_rate": 1.1959619249798717e-05,
+      "loss": 0.3294,
+      "step": 2923
+    },
+    {
+      "epoch": 1.3971088943312826,
+      "grad_norm": 0.46298505676389556,
+      "learning_rate": 1.1954454381607648e-05,
+      "loss": 0.3138,
+      "step": 2924
+    },
+    {
+      "epoch": 1.3975867630368557,
+      "grad_norm": 0.5451763604619123,
+      "learning_rate": 1.1949288971285411e-05,
+      "loss": 0.3109,
+      "step": 2925
+    },
+    {
+      "epoch": 1.3980646317424288,
+      "grad_norm": 0.47048163898437384,
+      "learning_rate": 1.19441230202648e-05,
+      "loss": 0.3332,
+      "step": 2926
+    },
+    {
+      "epoch": 1.3985425004480019,
+      "grad_norm": 0.5134522375456477,
+      "learning_rate": 1.1938956529978754e-05,
+      "loss": 0.3209,
+      "step": 2927
+    },
+    {
+      "epoch": 1.399020369153575,
+      "grad_norm": 0.5437669027850316,
+      "learning_rate": 1.1933789501860371e-05,
+      "loss": 0.3169,
+      "step": 2928
+    },
+    {
+      "epoch": 1.3994982378591483,
+      "grad_norm": 0.4756199727019128,
+      "learning_rate": 1.192862193734289e-05,
+      "loss": 0.3175,
+      "step": 2929
+    },
+    {
+      "epoch": 1.3999761065647214,
+      "grad_norm": 0.5356791744713957,
+      "learning_rate": 1.1923453837859706e-05,
+      "loss": 0.3342,
+      "step": 2930
+    },
+    {
+      "epoch": 1.4004539752702945,
+      "grad_norm": 0.5132793736491843,
+      "learning_rate": 1.1918285204844355e-05,
+      "loss": 0.3226,
+      "step": 2931
+    },
+    {
+      "epoch": 1.4009318439758676,
+      "grad_norm": 0.4654901065559778,
+      "learning_rate": 1.1913116039730528e-05,
+      "loss": 0.3326,
+      "step": 2932
+    },
+    {
+      "epoch": 1.4014097126814407,
+      "grad_norm": 0.5200095543826778,
+      "learning_rate": 1.1907946343952057e-05,
+      "loss": 0.321,
+      "step": 2933
+    },
+    {
+      "epoch": 1.401887581387014,
+      "grad_norm": 0.4772540680937219,
+      "learning_rate": 1.1902776118942924e-05,
+      "loss": 0.3287,
+      "step": 2934
+    },
+    {
+      "epoch": 1.402365450092587,
+      "grad_norm": 0.7737307275643527,
+      "learning_rate": 1.1897605366137264e-05,
+      "loss": 0.3153,
+      "step": 2935
+    },
+    {
+      "epoch": 1.4028433187981602,
+      "grad_norm": 0.5148598861162687,
+      "learning_rate": 1.1892434086969343e-05,
+      "loss": 0.3237,
+      "step": 2936
+    },
+    {
+      "epoch": 1.4033211875037335,
+      "grad_norm": 0.5176901401278625,
+      "learning_rate": 1.1887262282873593e-05,
+      "loss": 0.3143,
+      "step": 2937
+    },
+    {
+      "epoch": 1.4037990562093066,
+      "grad_norm": 0.5251202606650733,
+      "learning_rate": 1.1882089955284575e-05,
+      "loss": 0.3325,
+      "step": 2938
+    },
+    {
+      "epoch": 1.4042769249148797,
+      "grad_norm": 0.5736546704799379,
+      "learning_rate": 1.1876917105637e-05,
+      "loss": 0.317,
+      "step": 2939
+    },
+    {
+      "epoch": 1.4047547936204527,
+      "grad_norm": 0.5121472986243492,
+      "learning_rate": 1.1871743735365735e-05,
+      "loss": 0.33,
+      "step": 2940
+    },
+    {
+      "epoch": 1.4052326623260258,
+      "grad_norm": 0.51293687234706,
+      "learning_rate": 1.186656984590577e-05,
+      "loss": 0.3366,
+      "step": 2941
+    },
+    {
+      "epoch": 1.4057105310315992,
+      "grad_norm": 0.5289581413398837,
+      "learning_rate": 1.1861395438692256e-05,
+      "loss": 0.338,
+      "step": 2942
+    },
+    {
+      "epoch": 1.4061883997371722,
+      "grad_norm": 0.7060472699769851,
+      "learning_rate": 1.1856220515160483e-05,
+      "loss": 0.328,
+      "step": 2943
+    },
+    {
+      "epoch": 1.4066662684427453,
+      "grad_norm": 0.49082212031808403,
+      "learning_rate": 1.185104507674588e-05,
+      "loss": 0.3218,
+      "step": 2944
+    },
+    {
+      "epoch": 1.4071441371483184,
+      "grad_norm": 0.5266564439767433,
+      "learning_rate": 1.1845869124884027e-05,
+      "loss": 0.3254,
+      "step": 2945
+    },
+    {
+      "epoch": 1.4076220058538915,
+      "grad_norm": 0.4967432730788497,
+      "learning_rate": 1.1840692661010639e-05,
+      "loss": 0.3141,
+      "step": 2946
+    },
+    {
+      "epoch": 1.4080998745594648,
+      "grad_norm": 0.49284978510750543,
+      "learning_rate": 1.1835515686561574e-05,
+      "loss": 0.3151,
+      "step": 2947
+    },
+    {
+      "epoch": 1.408577743265038,
+      "grad_norm": 0.4584745987957341,
+      "learning_rate": 1.1830338202972838e-05,
+      "loss": 0.3367,
+      "step": 2948
+    },
+    {
+      "epoch": 1.409055611970611,
+      "grad_norm": 0.5227828009244597,
+      "learning_rate": 1.1825160211680571e-05,
+      "loss": 0.3316,
+      "step": 2949
+    },
+    {
+      "epoch": 1.4095334806761843,
+      "grad_norm": 1.0116286136942643,
+      "learning_rate": 1.1819981714121054e-05,
+      "loss": 0.3457,
+      "step": 2950
+    },
+    {
+      "epoch": 1.4100113493817574,
+      "grad_norm": 0.508515991330586,
+      "learning_rate": 1.1814802711730714e-05,
+      "loss": 0.3234,
+      "step": 2951
+    },
+    {
+      "epoch": 1.4104892180873305,
+      "grad_norm": 0.4806606845844688,
+      "learning_rate": 1.1809623205946116e-05,
+      "loss": 0.3171,
+      "step": 2952
+    },
+    {
+      "epoch": 1.4109670867929036,
+      "grad_norm": 0.5469751195822524,
+      "learning_rate": 1.180444319820396e-05,
+      "loss": 0.3236,
+      "step": 2953
+    },
+    {
+      "epoch": 1.4114449554984767,
+      "grad_norm": 0.5181364422493109,
+      "learning_rate": 1.179926268994109e-05,
+      "loss": 0.33,
+      "step": 2954
+    },
+    {
+      "epoch": 1.41192282420405,
+      "grad_norm": 0.4780046219803632,
+      "learning_rate": 1.1794081682594491e-05,
+      "loss": 0.3224,
+      "step": 2955
+    },
+    {
+      "epoch": 1.4124006929096231,
+      "grad_norm": 0.4912369040242893,
+      "learning_rate": 1.178890017760128e-05,
+      "loss": 0.3153,
+      "step": 2956
+    },
+    {
+      "epoch": 1.4128785616151962,
+      "grad_norm": 0.4775890106833153,
+      "learning_rate": 1.1783718176398716e-05,
+      "loss": 0.3332,
+      "step": 2957
+    },
+    {
+      "epoch": 1.4133564303207693,
+      "grad_norm": 0.525737532682014,
+      "learning_rate": 1.1778535680424192e-05,
+      "loss": 0.314,
+      "step": 2958
+    },
+    {
+      "epoch": 1.4138342990263424,
+      "grad_norm": 0.47808599069608393,
+      "learning_rate": 1.1773352691115246e-05,
+      "loss": 0.3091,
+      "step": 2959
+    },
+    {
+      "epoch": 1.4143121677319157,
+      "grad_norm": 0.5514083626189912,
+      "learning_rate": 1.176816920990954e-05,
+      "loss": 0.3335,
+      "step": 2960
+    },
+    {
+      "epoch": 1.4147900364374888,
+      "grad_norm": 0.5288775984235308,
+      "learning_rate": 1.176298523824489e-05,
+      "loss": 0.3496,
+      "step": 2961
+    },
+    {
+      "epoch": 1.4152679051430619,
+      "grad_norm": 0.5325892186828176,
+      "learning_rate": 1.1757800777559232e-05,
+      "loss": 0.303,
+      "step": 2962
+    },
+    {
+      "epoch": 1.4157457738486352,
+      "grad_norm": 0.48239236208741576,
+      "learning_rate": 1.1752615829290644e-05,
+      "loss": 0.3328,
+      "step": 2963
+    },
+    {
+      "epoch": 1.4162236425542083,
+      "grad_norm": 0.5050757806977875,
+      "learning_rate": 1.1747430394877342e-05,
+      "loss": 0.3391,
+      "step": 2964
+    },
+    {
+      "epoch": 1.4167015112597814,
+      "grad_norm": 0.5076547010445559,
+      "learning_rate": 1.174224447575767e-05,
+      "loss": 0.319,
+      "step": 2965
+    },
+    {
+      "epoch": 1.4171793799653545,
+      "grad_norm": 0.5138942551079375,
+      "learning_rate": 1.1737058073370116e-05,
+      "loss": 0.329,
+      "step": 2966
+    },
+    {
+      "epoch": 1.4176572486709276,
+      "grad_norm": 0.7215604590019389,
+      "learning_rate": 1.1731871189153295e-05,
+      "loss": 0.3423,
+      "step": 2967
+    },
+    {
+      "epoch": 1.4181351173765009,
+      "grad_norm": 0.4880111433749174,
+      "learning_rate": 1.1726683824545953e-05,
+      "loss": 0.3194,
+      "step": 2968
+    },
+    {
+      "epoch": 1.418612986082074,
+      "grad_norm": 0.49438163754649106,
+      "learning_rate": 1.1721495980986975e-05,
+      "loss": 0.3308,
+      "step": 2969
+    },
+    {
+      "epoch": 1.419090854787647,
+      "grad_norm": 0.5137865619807299,
+      "learning_rate": 1.171630765991538e-05,
+      "loss": 0.3181,
+      "step": 2970
+    },
+    {
+      "epoch": 1.4195687234932202,
+      "grad_norm": 0.5058423220393458,
+      "learning_rate": 1.1711118862770314e-05,
+      "loss": 0.3299,
+      "step": 2971
+    },
+    {
+      "epoch": 1.4200465921987933,
+      "grad_norm": 0.5134808993321874,
+      "learning_rate": 1.1705929590991062e-05,
+      "loss": 0.3328,
+      "step": 2972
+    },
+    {
+      "epoch": 1.4205244609043666,
+      "grad_norm": 0.5311157334978709,
+      "learning_rate": 1.1700739846017033e-05,
+      "loss": 0.3254,
+      "step": 2973
+    },
+    {
+      "epoch": 1.4210023296099397,
+      "grad_norm": 0.47075019738009893,
+      "learning_rate": 1.169554962928777e-05,
+      "loss": 0.3283,
+      "step": 2974
+    },
+    {
+      "epoch": 1.4214801983155128,
+      "grad_norm": 0.5260956963173294,
+      "learning_rate": 1.169035894224295e-05,
+      "loss": 0.3304,
+      "step": 2975
+    },
+    {
+      "epoch": 1.421958067021086,
+      "grad_norm": 0.5540899904553104,
+      "learning_rate": 1.1685167786322375e-05,
+      "loss": 0.3338,
+      "step": 2976
+    },
+    {
+      "epoch": 1.4224359357266592,
+      "grad_norm": 0.5415508885164177,
+      "learning_rate": 1.1679976162965984e-05,
+      "loss": 0.3244,
+      "step": 2977
+    },
+    {
+      "epoch": 1.4229138044322323,
+      "grad_norm": 0.47480665610205475,
+      "learning_rate": 1.1674784073613841e-05,
+      "loss": 0.3329,
+      "step": 2978
+    },
+    {
+      "epoch": 1.4233916731378053,
+      "grad_norm": 0.5732469730028841,
+      "learning_rate": 1.1669591519706134e-05,
+      "loss": 0.3413,
+      "step": 2979
+    },
+    {
+      "epoch": 1.4238695418433784,
+      "grad_norm": 0.49708436697419595,
+      "learning_rate": 1.1664398502683194e-05,
+      "loss": 0.3129,
+      "step": 2980
+    },
+    {
+      "epoch": 1.4243474105489518,
+      "grad_norm": 0.46717071543365596,
+      "learning_rate": 1.165920502398547e-05,
+      "loss": 0.3372,
+      "step": 2981
+    },
+    {
+      "epoch": 1.4248252792545248,
+      "grad_norm": 0.4770094789534314,
+      "learning_rate": 1.1654011085053537e-05,
+      "loss": 0.3341,
+      "step": 2982
+    },
+    {
+      "epoch": 1.425303147960098,
+      "grad_norm": 0.4957722154844253,
+      "learning_rate": 1.1648816687328104e-05,
+      "loss": 0.3112,
+      "step": 2983
+    },
+    {
+      "epoch": 1.425781016665671,
+      "grad_norm": 0.4954214821075703,
+      "learning_rate": 1.1643621832250012e-05,
+      "loss": 0.3202,
+      "step": 2984
+    },
+    {
+      "epoch": 1.4262588853712441,
+      "grad_norm": 0.560090241257771,
+      "learning_rate": 1.1638426521260211e-05,
+      "loss": 0.3387,
+      "step": 2985
+    },
+    {
+      "epoch": 1.4267367540768174,
+      "grad_norm": 0.5088929436420992,
+      "learning_rate": 1.1633230755799799e-05,
+      "loss": 0.3337,
+      "step": 2986
+    },
+    {
+      "epoch": 1.4272146227823905,
+      "grad_norm": 0.5423381244933303,
+      "learning_rate": 1.162803453730998e-05,
+      "loss": 0.3151,
+      "step": 2987
+    },
+    {
+      "epoch": 1.4276924914879636,
+      "grad_norm": 0.49395978930167594,
+      "learning_rate": 1.1622837867232102e-05,
+      "loss": 0.3265,
+      "step": 2988
+    },
+    {
+      "epoch": 1.428170360193537,
+      "grad_norm": 0.4718130842189427,
+      "learning_rate": 1.1617640747007626e-05,
+      "loss": 0.3246,
+      "step": 2989
+    },
+    {
+      "epoch": 1.42864822889911,
+      "grad_norm": 0.4980105057115814,
+      "learning_rate": 1.1612443178078138e-05,
+      "loss": 0.3472,
+      "step": 2990
+    },
+    {
+      "epoch": 1.4291260976046831,
+      "grad_norm": 0.5938682319676443,
+      "learning_rate": 1.1607245161885358e-05,
+      "loss": 0.3105,
+      "step": 2991
+    },
+    {
+      "epoch": 1.4296039663102562,
+      "grad_norm": 0.48289231319482784,
+      "learning_rate": 1.1602046699871126e-05,
+      "loss": 0.3224,
+      "step": 2992
+    },
+    {
+      "epoch": 1.4300818350158293,
+      "grad_norm": 0.5006882224163078,
+      "learning_rate": 1.1596847793477393e-05,
+      "loss": 0.333,
+      "step": 2993
+    },
+    {
+      "epoch": 1.4305597037214026,
+      "grad_norm": 0.5004173467454556,
+      "learning_rate": 1.1591648444146251e-05,
+      "loss": 0.3128,
+      "step": 2994
+    },
+    {
+      "epoch": 1.4310375724269757,
+      "grad_norm": 0.49393923359090247,
+      "learning_rate": 1.1586448653319908e-05,
+      "loss": 0.3196,
+      "step": 2995
+    },
+    {
+      "epoch": 1.4315154411325488,
+      "grad_norm": 0.4696580955679868,
+      "learning_rate": 1.1581248422440692e-05,
+      "loss": 0.3249,
+      "step": 2996
+    },
+    {
+      "epoch": 1.431993309838122,
+      "grad_norm": 0.5122849607619827,
+      "learning_rate": 1.1576047752951056e-05,
+      "loss": 0.3295,
+      "step": 2997
+    },
+    {
+      "epoch": 1.432471178543695,
+      "grad_norm": 0.5080265014360067,
+      "learning_rate": 1.157084664629358e-05,
+      "loss": 0.318,
+      "step": 2998
+    },
+    {
+      "epoch": 1.4329490472492683,
+      "grad_norm": 0.5283144934239468,
+      "learning_rate": 1.1565645103910945e-05,
+      "loss": 0.3259,
+      "step": 2999
+    },
+    {
+      "epoch": 1.4334269159548414,
+      "grad_norm": 0.5254577539682926,
+      "learning_rate": 1.156044312724598e-05,
+      "loss": 0.3244,
+      "step": 3000
+    },
+    {
+      "epoch": 1.4339047846604145,
+      "grad_norm": 0.4926424316230558,
+      "learning_rate": 1.1555240717741618e-05,
+      "loss": 0.3417,
+      "step": 3001
+    },
+    {
+      "epoch": 1.4343826533659878,
+      "grad_norm": 0.4796896772668644,
+      "learning_rate": 1.1550037876840913e-05,
+      "loss": 0.3166,
+      "step": 3002
+    },
+    {
+      "epoch": 1.434860522071561,
+      "grad_norm": 0.4773964347570398,
+      "learning_rate": 1.1544834605987042e-05,
+      "loss": 0.3325,
+      "step": 3003
+    },
+    {
+      "epoch": 1.435338390777134,
+      "grad_norm": 0.5159550135561818,
+      "learning_rate": 1.1539630906623305e-05,
+      "loss": 0.3169,
+      "step": 3004
+    },
+    {
+      "epoch": 1.435816259482707,
+      "grad_norm": 0.4972490804066091,
+      "learning_rate": 1.153442678019311e-05,
+      "loss": 0.3306,
+      "step": 3005
+    },
+    {
+      "epoch": 1.4362941281882802,
+      "grad_norm": 0.48572512446743665,
+      "learning_rate": 1.1529222228139993e-05,
+      "loss": 0.3196,
+      "step": 3006
+    },
+    {
+      "epoch": 1.4367719968938535,
+      "grad_norm": 0.4956598068058492,
+      "learning_rate": 1.1524017251907609e-05,
+      "loss": 0.3233,
+      "step": 3007
+    },
+    {
+      "epoch": 1.4372498655994266,
+      "grad_norm": 0.47225254602078437,
+      "learning_rate": 1.151881185293972e-05,
+      "loss": 0.322,
+      "step": 3008
+    },
+    {
+      "epoch": 1.4377277343049997,
+      "grad_norm": 0.4994314156150159,
+      "learning_rate": 1.1513606032680214e-05,
+      "loss": 0.33,
+      "step": 3009
+    },
+    {
+      "epoch": 1.4382056030105728,
+      "grad_norm": 0.4932853873998178,
+      "learning_rate": 1.1508399792573095e-05,
+      "loss": 0.3316,
+      "step": 3010
+    },
+    {
+      "epoch": 1.4386834717161459,
+      "grad_norm": 0.4692505918945855,
+      "learning_rate": 1.150319313406248e-05,
+      "loss": 0.3068,
+      "step": 3011
+    },
+    {
+      "epoch": 1.4391613404217192,
+      "grad_norm": 0.5056071919354226,
+      "learning_rate": 1.1497986058592607e-05,
+      "loss": 0.3485,
+      "step": 3012
+    },
+    {
+      "epoch": 1.4396392091272923,
+      "grad_norm": 0.5537829323649248,
+      "learning_rate": 1.1492778567607826e-05,
+      "loss": 0.2966,
+      "step": 3013
+    },
+    {
+      "epoch": 1.4401170778328654,
+      "grad_norm": 0.49754562492925625,
+      "learning_rate": 1.1487570662552601e-05,
+      "loss": 0.3409,
+      "step": 3014
+    },
+    {
+      "epoch": 1.4405949465384387,
+      "grad_norm": 0.5261937593332422,
+      "learning_rate": 1.1482362344871514e-05,
+      "loss": 0.3353,
+      "step": 3015
+    },
+    {
+      "epoch": 1.4410728152440118,
+      "grad_norm": 0.44720642661167864,
+      "learning_rate": 1.1477153616009262e-05,
+      "loss": 0.3287,
+      "step": 3016
+    },
+    {
+      "epoch": 1.4415506839495849,
+      "grad_norm": 0.46950785439169124,
+      "learning_rate": 1.1471944477410652e-05,
+      "loss": 0.3198,
+      "step": 3017
+    },
+    {
+      "epoch": 1.442028552655158,
+      "grad_norm": 0.5086685855274399,
+      "learning_rate": 1.1466734930520609e-05,
+      "loss": 0.309,
+      "step": 3018
+    },
+    {
+      "epoch": 1.442506421360731,
+      "grad_norm": 0.47131182541150224,
+      "learning_rate": 1.1461524976784172e-05,
+      "loss": 0.3426,
+      "step": 3019
+    },
+    {
+      "epoch": 1.4429842900663044,
+      "grad_norm": 0.4858076737245921,
+      "learning_rate": 1.1456314617646482e-05,
+      "loss": 0.3335,
+      "step": 3020
+    },
+    {
+      "epoch": 1.4434621587718774,
+      "grad_norm": 0.47177866647330885,
+      "learning_rate": 1.145110385455281e-05,
+      "loss": 0.3117,
+      "step": 3021
+    },
+    {
+      "epoch": 1.4439400274774505,
+      "grad_norm": 0.49146463401072915,
+      "learning_rate": 1.1445892688948525e-05,
+      "loss": 0.3288,
+      "step": 3022
+    },
+    {
+      "epoch": 1.4444178961830239,
+      "grad_norm": 0.5021205966554858,
+      "learning_rate": 1.1440681122279113e-05,
+      "loss": 0.3244,
+      "step": 3023
+    },
+    {
+      "epoch": 1.4448957648885967,
+      "grad_norm": 0.4819803103747588,
+      "learning_rate": 1.1435469155990171e-05,
+      "loss": 0.3103,
+      "step": 3024
+    },
+    {
+      "epoch": 1.44537363359417,
+      "grad_norm": 1.383385618058039,
+      "learning_rate": 1.1430256791527406e-05,
+      "loss": 0.3312,
+      "step": 3025
+    },
+    {
+      "epoch": 1.4458515022997431,
+      "grad_norm": 0.5170297557146272,
+      "learning_rate": 1.1425044030336636e-05,
+      "loss": 0.3127,
+      "step": 3026
+    },
+    {
+      "epoch": 1.4463293710053162,
+      "grad_norm": 0.4832092172218538,
+      "learning_rate": 1.1419830873863792e-05,
+      "loss": 0.3267,
+      "step": 3027
+    },
+    {
+      "epoch": 1.4468072397108895,
+      "grad_norm": 0.5132105382295371,
+      "learning_rate": 1.1414617323554906e-05,
+      "loss": 0.3224,
+      "step": 3028
+    },
+    {
+      "epoch": 1.4472851084164626,
+      "grad_norm": 0.5023592255793703,
+      "learning_rate": 1.1409403380856128e-05,
+      "loss": 0.3197,
+      "step": 3029
+    },
+    {
+      "epoch": 1.4477629771220357,
+      "grad_norm": 0.49076288043984206,
+      "learning_rate": 1.1404189047213716e-05,
+      "loss": 0.3323,
+      "step": 3030
+    },
+    {
+      "epoch": 1.4482408458276088,
+      "grad_norm": 0.49562922226214057,
+      "learning_rate": 1.139897432407403e-05,
+      "loss": 0.333,
+      "step": 3031
+    },
+    {
+      "epoch": 1.448718714533182,
+      "grad_norm": 0.4849984150072255,
+      "learning_rate": 1.1393759212883544e-05,
+      "loss": 0.3169,
+      "step": 3032
+    },
+    {
+      "epoch": 1.4491965832387552,
+      "grad_norm": 0.4706486064594749,
+      "learning_rate": 1.1388543715088838e-05,
+      "loss": 0.2996,
+      "step": 3033
+    },
+    {
+      "epoch": 1.4496744519443283,
+      "grad_norm": 0.5374385570329433,
+      "learning_rate": 1.13833278321366e-05,
+      "loss": 0.3099,
+      "step": 3034
+    },
+    {
+      "epoch": 1.4501523206499014,
+      "grad_norm": 0.5207013029763824,
+      "learning_rate": 1.137811156547362e-05,
+      "loss": 0.3135,
+      "step": 3035
+    },
+    {
+      "epoch": 1.4506301893554747,
+      "grad_norm": 0.5204531377813438,
+      "learning_rate": 1.1372894916546804e-05,
+      "loss": 0.3254,
+      "step": 3036
+    },
+    {
+      "epoch": 1.4511080580610476,
+      "grad_norm": 0.46838194787481924,
+      "learning_rate": 1.1367677886803152e-05,
+      "loss": 0.3311,
+      "step": 3037
+    },
+    {
+      "epoch": 1.451585926766621,
+      "grad_norm": 0.5507044912148897,
+      "learning_rate": 1.1362460477689784e-05,
+      "loss": 0.3367,
+      "step": 3038
+    },
+    {
+      "epoch": 1.452063795472194,
+      "grad_norm": 0.5069990443386853,
+      "learning_rate": 1.1357242690653911e-05,
+      "loss": 0.342,
+      "step": 3039
+    },
+    {
+      "epoch": 1.452541664177767,
+      "grad_norm": 0.48662576527328144,
+      "learning_rate": 1.1352024527142855e-05,
+      "loss": 0.3279,
+      "step": 3040
+    },
+    {
+      "epoch": 1.4530195328833404,
+      "grad_norm": 0.5009454804753579,
+      "learning_rate": 1.1346805988604048e-05,
+      "loss": 0.3205,
+      "step": 3041
+    },
+    {
+      "epoch": 1.4534974015889135,
+      "grad_norm": 0.5033573059665943,
+      "learning_rate": 1.1341587076485015e-05,
+      "loss": 0.3258,
+      "step": 3042
+    },
+    {
+      "epoch": 1.4539752702944866,
+      "grad_norm": 0.48696663644276184,
+      "learning_rate": 1.1336367792233394e-05,
+      "loss": 0.306,
+      "step": 3043
+    },
+    {
+      "epoch": 1.4544531390000597,
+      "grad_norm": 0.4703928054737579,
+      "learning_rate": 1.133114813729692e-05,
+      "loss": 0.3342,
+      "step": 3044
+    },
+    {
+      "epoch": 1.4549310077056328,
+      "grad_norm": 0.5033596475668163,
+      "learning_rate": 1.1325928113123431e-05,
+      "loss": 0.3275,
+      "step": 3045
+    },
+    {
+      "epoch": 1.455408876411206,
+      "grad_norm": 0.4994474514417599,
+      "learning_rate": 1.1320707721160876e-05,
+      "loss": 0.3334,
+      "step": 3046
+    },
+    {
+      "epoch": 1.4558867451167792,
+      "grad_norm": 0.5430329386590091,
+      "learning_rate": 1.1315486962857293e-05,
+      "loss": 0.3116,
+      "step": 3047
+    },
+    {
+      "epoch": 1.4563646138223523,
+      "grad_norm": 0.4838126377080352,
+      "learning_rate": 1.1310265839660835e-05,
+      "loss": 0.3131,
+      "step": 3048
+    },
+    {
+      "epoch": 1.4568424825279256,
+      "grad_norm": 0.6439581398779899,
+      "learning_rate": 1.130504435301974e-05,
+      "loss": 0.3308,
+      "step": 3049
+    },
+    {
+      "epoch": 1.4573203512334987,
+      "grad_norm": 0.4840420029195805,
+      "learning_rate": 1.129982250438237e-05,
+      "loss": 0.3257,
+      "step": 3050
+    },
+    {
+      "epoch": 1.4577982199390718,
+      "grad_norm": 0.47837719127146977,
+      "learning_rate": 1.129460029519716e-05,
+      "loss": 0.3358,
+      "step": 3051
+    },
+    {
+      "epoch": 1.4582760886446449,
+      "grad_norm": 0.4893765918667975,
+      "learning_rate": 1.1289377726912665e-05,
+      "loss": 0.3352,
+      "step": 3052
+    },
+    {
+      "epoch": 1.458753957350218,
+      "grad_norm": 0.4923586108450949,
+      "learning_rate": 1.1284154800977533e-05,
+      "loss": 0.3281,
+      "step": 3053
+    },
+    {
+      "epoch": 1.4592318260557913,
+      "grad_norm": 0.4621191750554461,
+      "learning_rate": 1.127893151884051e-05,
+      "loss": 0.3401,
+      "step": 3054
+    },
+    {
+      "epoch": 1.4597096947613644,
+      "grad_norm": 0.492839774425873,
+      "learning_rate": 1.1273707881950445e-05,
+      "loss": 0.3327,
+      "step": 3055
+    },
+    {
+      "epoch": 1.4601875634669375,
+      "grad_norm": 0.4643479712090378,
+      "learning_rate": 1.1268483891756283e-05,
+      "loss": 0.3292,
+      "step": 3056
+    },
+    {
+      "epoch": 1.4606654321725105,
+      "grad_norm": 0.4671866712639955,
+      "learning_rate": 1.1263259549707063e-05,
+      "loss": 0.3255,
+      "step": 3057
+    },
+    {
+      "epoch": 1.4611433008780836,
+      "grad_norm": 0.493972595273388,
+      "learning_rate": 1.125803485725193e-05,
+      "loss": 0.3326,
+      "step": 3058
+    },
+    {
+      "epoch": 1.461621169583657,
+      "grad_norm": 0.4660787343724186,
+      "learning_rate": 1.1252809815840118e-05,
+      "loss": 0.3128,
+      "step": 3059
+    },
+    {
+      "epoch": 1.46209903828923,
+      "grad_norm": 0.4975144467520985,
+      "learning_rate": 1.1247584426920962e-05,
+      "loss": 0.3134,
+      "step": 3060
+    },
+    {
+      "epoch": 1.4625769069948031,
+      "grad_norm": 0.45843098521936504,
+      "learning_rate": 1.124235869194389e-05,
+      "loss": 0.3154,
+      "step": 3061
+    },
+    {
+      "epoch": 1.4630547757003765,
+      "grad_norm": 0.4818123666317099,
+      "learning_rate": 1.1237132612358436e-05,
+      "loss": 0.3275,
+      "step": 3062
+    },
+    {
+      "epoch": 1.4635326444059495,
+      "grad_norm": 0.5292710710183121,
+      "learning_rate": 1.1231906189614217e-05,
+      "loss": 0.3257,
+      "step": 3063
+    },
+    {
+      "epoch": 1.4640105131115226,
+      "grad_norm": 0.48700887730356873,
+      "learning_rate": 1.1226679425160949e-05,
+      "loss": 0.3329,
+      "step": 3064
+    },
+    {
+      "epoch": 1.4644883818170957,
+      "grad_norm": 0.48240701019079835,
+      "learning_rate": 1.1221452320448449e-05,
+      "loss": 0.3425,
+      "step": 3065
+    },
+    {
+      "epoch": 1.4649662505226688,
+      "grad_norm": 0.48147101653808244,
+      "learning_rate": 1.1216224876926622e-05,
+      "loss": 0.318,
+      "step": 3066
+    },
+    {
+      "epoch": 1.4654441192282421,
+      "grad_norm": 0.467596203651121,
+      "learning_rate": 1.1210997096045466e-05,
+      "loss": 0.3293,
+      "step": 3067
+    },
+    {
+      "epoch": 1.4659219879338152,
+      "grad_norm": 0.4550529019067923,
+      "learning_rate": 1.1205768979255078e-05,
+      "loss": 0.3158,
+      "step": 3068
+    },
+    {
+      "epoch": 1.4663998566393883,
+      "grad_norm": 0.48042440869684455,
+      "learning_rate": 1.1200540528005645e-05,
+      "loss": 0.3298,
+      "step": 3069
+    },
+    {
+      "epoch": 1.4668777253449614,
+      "grad_norm": 0.5277816140348442,
+      "learning_rate": 1.1195311743747445e-05,
+      "loss": 0.3363,
+      "step": 3070
+    },
+    {
+      "epoch": 1.4673555940505345,
+      "grad_norm": 0.4888123296530078,
+      "learning_rate": 1.1190082627930854e-05,
+      "loss": 0.3226,
+      "step": 3071
+    },
+    {
+      "epoch": 1.4678334627561078,
+      "grad_norm": 0.5015585568421738,
+      "learning_rate": 1.1184853182006332e-05,
+      "loss": 0.3172,
+      "step": 3072
+    },
+    {
+      "epoch": 1.468311331461681,
+      "grad_norm": 0.47745495211196265,
+      "learning_rate": 1.1179623407424442e-05,
+      "loss": 0.3181,
+      "step": 3073
+    },
+    {
+      "epoch": 1.468789200167254,
+      "grad_norm": 0.5013475163036912,
+      "learning_rate": 1.1174393305635825e-05,
+      "loss": 0.3083,
+      "step": 3074
+    },
+    {
+      "epoch": 1.4692670688728273,
+      "grad_norm": 0.521653074954796,
+      "learning_rate": 1.116916287809122e-05,
+      "loss": 0.3178,
+      "step": 3075
+    },
+    {
+      "epoch": 1.4697449375784004,
+      "grad_norm": 0.4953449613595978,
+      "learning_rate": 1.116393212624146e-05,
+      "loss": 0.3375,
+      "step": 3076
+    },
+    {
+      "epoch": 1.4702228062839735,
+      "grad_norm": 0.4704912345390417,
+      "learning_rate": 1.1158701051537455e-05,
+      "loss": 0.3199,
+      "step": 3077
+    },
+    {
+      "epoch": 1.4707006749895466,
+      "grad_norm": 0.5066215917086822,
+      "learning_rate": 1.1153469655430218e-05,
+      "loss": 0.3028,
+      "step": 3078
+    },
+    {
+      "epoch": 1.4711785436951197,
+      "grad_norm": 0.49299038538236867,
+      "learning_rate": 1.1148237939370847e-05,
+      "loss": 0.3013,
+      "step": 3079
+    },
+    {
+      "epoch": 1.471656412400693,
+      "grad_norm": 0.503131186431184,
+      "learning_rate": 1.1143005904810527e-05,
+      "loss": 0.3228,
+      "step": 3080
+    },
+    {
+      "epoch": 1.472134281106266,
+      "grad_norm": 0.48541708299393943,
+      "learning_rate": 1.1137773553200528e-05,
+      "loss": 0.3247,
+      "step": 3081
+    },
+    {
+      "epoch": 1.4726121498118392,
+      "grad_norm": 0.4667618110206343,
+      "learning_rate": 1.1132540885992221e-05,
+      "loss": 0.3302,
+      "step": 3082
+    },
+    {
+      "epoch": 1.4730900185174123,
+      "grad_norm": 0.4966223048730619,
+      "learning_rate": 1.1127307904637044e-05,
+      "loss": 0.3098,
+      "step": 3083
+    },
+    {
+      "epoch": 1.4735678872229854,
+      "grad_norm": 0.49913335938844805,
+      "learning_rate": 1.1122074610586541e-05,
+      "loss": 0.3338,
+      "step": 3084
+    },
+    {
+      "epoch": 1.4740457559285587,
+      "grad_norm": 0.5787027610611666,
+      "learning_rate": 1.1116841005292339e-05,
+      "loss": 0.3233,
+      "step": 3085
+    },
+    {
+      "epoch": 1.4745236246341318,
+      "grad_norm": 0.4988518285658534,
+      "learning_rate": 1.1111607090206135e-05,
+      "loss": 0.3451,
+      "step": 3086
+    },
+    {
+      "epoch": 1.4750014933397049,
+      "grad_norm": 0.48890728691122937,
+      "learning_rate": 1.1106372866779738e-05,
+      "loss": 0.3226,
+      "step": 3087
+    },
+    {
+      "epoch": 1.4754793620452782,
+      "grad_norm": 0.4743490756030402,
+      "learning_rate": 1.110113833646502e-05,
+      "loss": 0.3174,
+      "step": 3088
+    },
+    {
+      "epoch": 1.4759572307508513,
+      "grad_norm": 0.496094343763048,
+      "learning_rate": 1.1095903500713953e-05,
+      "loss": 0.3114,
+      "step": 3089
+    },
+    {
+      "epoch": 1.4764350994564244,
+      "grad_norm": 0.5359507037011488,
+      "learning_rate": 1.1090668360978589e-05,
+      "loss": 0.3144,
+      "step": 3090
+    },
+    {
+      "epoch": 1.4769129681619975,
+      "grad_norm": 0.5615293353766108,
+      "learning_rate": 1.1085432918711059e-05,
+      "loss": 0.3209,
+      "step": 3091
+    },
+    {
+      "epoch": 1.4773908368675706,
+      "grad_norm": 0.512584855813972,
+      "learning_rate": 1.1080197175363584e-05,
+      "loss": 0.3242,
+      "step": 3092
+    },
+    {
+      "epoch": 1.4778687055731439,
+      "grad_norm": 0.5265949611457437,
+      "learning_rate": 1.1074961132388466e-05,
+      "loss": 0.3275,
+      "step": 3093
+    },
+    {
+      "epoch": 1.478346574278717,
+      "grad_norm": 0.49105786313448796,
+      "learning_rate": 1.1069724791238092e-05,
+      "loss": 0.3118,
+      "step": 3094
+    },
+    {
+      "epoch": 1.47882444298429,
+      "grad_norm": 0.47704097499158915,
+      "learning_rate": 1.106448815336493e-05,
+      "loss": 0.3289,
+      "step": 3095
+    },
+    {
+      "epoch": 1.4793023116898631,
+      "grad_norm": 0.5232471767406344,
+      "learning_rate": 1.1059251220221534e-05,
+      "loss": 0.3302,
+      "step": 3096
+    },
+    {
+      "epoch": 1.4797801803954362,
+      "grad_norm": 0.5228817184494445,
+      "learning_rate": 1.1054013993260533e-05,
+      "loss": 0.3201,
+      "step": 3097
+    },
+    {
+      "epoch": 1.4802580491010096,
+      "grad_norm": 0.4936116345958341,
+      "learning_rate": 1.1048776473934642e-05,
+      "loss": 0.32,
+      "step": 3098
+    },
+    {
+      "epoch": 1.4807359178065826,
+      "grad_norm": 0.4930123354861475,
+      "learning_rate": 1.1043538663696658e-05,
+      "loss": 0.319,
+      "step": 3099
+    },
+    {
+      "epoch": 1.4812137865121557,
+      "grad_norm": 0.5490307592266955,
+      "learning_rate": 1.1038300563999455e-05,
+      "loss": 0.3392,
+      "step": 3100
+    },
+    {
+      "epoch": 1.481691655217729,
+      "grad_norm": 0.5030549664101823,
+      "learning_rate": 1.1033062176295992e-05,
+      "loss": 0.3256,
+      "step": 3101
+    },
+    {
+      "epoch": 1.4821695239233021,
+      "grad_norm": 0.4693206453740588,
+      "learning_rate": 1.1027823502039307e-05,
+      "loss": 0.3357,
+      "step": 3102
+    },
+    {
+      "epoch": 1.4826473926288752,
+      "grad_norm": 0.49980642230471584,
+      "learning_rate": 1.1022584542682508e-05,
+      "loss": 0.3224,
+      "step": 3103
+    },
+    {
+      "epoch": 1.4831252613344483,
+      "grad_norm": 0.4872985953951889,
+      "learning_rate": 1.1017345299678797e-05,
+      "loss": 0.32,
+      "step": 3104
+    },
+    {
+      "epoch": 1.4836031300400214,
+      "grad_norm": 0.5155764862652535,
+      "learning_rate": 1.1012105774481446e-05,
+      "loss": 0.3255,
+      "step": 3105
+    },
+    {
+      "epoch": 1.4840809987455947,
+      "grad_norm": 0.49092199999070874,
+      "learning_rate": 1.1006865968543805e-05,
+      "loss": 0.3022,
+      "step": 3106
+    },
+    {
+      "epoch": 1.4845588674511678,
+      "grad_norm": 0.4638968403292164,
+      "learning_rate": 1.1001625883319307e-05,
+      "loss": 0.3284,
+      "step": 3107
+    },
+    {
+      "epoch": 1.485036736156741,
+      "grad_norm": 0.47749399048412333,
+      "learning_rate": 1.0996385520261457e-05,
+      "loss": 0.3171,
+      "step": 3108
+    },
+    {
+      "epoch": 1.485514604862314,
+      "grad_norm": 0.571045799710566,
+      "learning_rate": 1.0991144880823838e-05,
+      "loss": 0.3236,
+      "step": 3109
+    },
+    {
+      "epoch": 1.485992473567887,
+      "grad_norm": 0.49745529315284953,
+      "learning_rate": 1.0985903966460115e-05,
+      "loss": 0.3439,
+      "step": 3110
+    },
+    {
+      "epoch": 1.4864703422734604,
+      "grad_norm": 0.5308655028324103,
+      "learning_rate": 1.0980662778624023e-05,
+      "loss": 0.329,
+      "step": 3111
+    },
+    {
+      "epoch": 1.4869482109790335,
+      "grad_norm": 0.5029668660735324,
+      "learning_rate": 1.0975421318769373e-05,
+      "loss": 0.332,
+      "step": 3112
+    },
+    {
+      "epoch": 1.4874260796846066,
+      "grad_norm": 0.5166152056859417,
+      "learning_rate": 1.0970179588350054e-05,
+      "loss": 0.3311,
+      "step": 3113
+    },
+    {
+      "epoch": 1.48790394839018,
+      "grad_norm": 0.5380965879327918,
+      "learning_rate": 1.0964937588820036e-05,
+      "loss": 0.3368,
+      "step": 3114
+    },
+    {
+      "epoch": 1.488381817095753,
+      "grad_norm": 0.4866414171234857,
+      "learning_rate": 1.0959695321633346e-05,
+      "loss": 0.3182,
+      "step": 3115
+    },
+    {
+      "epoch": 1.488859685801326,
+      "grad_norm": 0.5061118269564608,
+      "learning_rate": 1.0954452788244106e-05,
+      "loss": 0.3256,
+      "step": 3116
+    },
+    {
+      "epoch": 1.4893375545068992,
+      "grad_norm": 0.4940241531258727,
+      "learning_rate": 1.0949209990106497e-05,
+      "loss": 0.3083,
+      "step": 3117
+    },
+    {
+      "epoch": 1.4898154232124723,
+      "grad_norm": 0.4543349967578536,
+      "learning_rate": 1.0943966928674783e-05,
+      "loss": 0.3339,
+      "step": 3118
+    },
+    {
+      "epoch": 1.4902932919180456,
+      "grad_norm": 0.5120304669937483,
+      "learning_rate": 1.093872360540329e-05,
+      "loss": 0.3219,
+      "step": 3119
+    },
+    {
+      "epoch": 1.4907711606236187,
+      "grad_norm": 0.4662796909068133,
+      "learning_rate": 1.0933480021746432e-05,
+      "loss": 0.327,
+      "step": 3120
+    },
+    {
+      "epoch": 1.4912490293291918,
+      "grad_norm": 0.4846749119742991,
+      "learning_rate": 1.0928236179158678e-05,
+      "loss": 0.3087,
+      "step": 3121
+    },
+    {
+      "epoch": 1.4917268980347649,
+      "grad_norm": 0.48452861211325327,
+      "learning_rate": 1.0922992079094588e-05,
+      "loss": 0.3141,
+      "step": 3122
+    },
+    {
+      "epoch": 1.492204766740338,
+      "grad_norm": 0.499123257716151,
+      "learning_rate": 1.0917747723008771e-05,
+      "loss": 0.3193,
+      "step": 3123
+    },
+    {
+      "epoch": 1.4926826354459113,
+      "grad_norm": 0.46393153800632286,
+      "learning_rate": 1.0912503112355926e-05,
+      "loss": 0.3166,
+      "step": 3124
+    },
+    {
+      "epoch": 1.4931605041514844,
+      "grad_norm": 0.5298869777156393,
+      "learning_rate": 1.0907258248590816e-05,
+      "loss": 0.3258,
+      "step": 3125
+    },
+    {
+      "epoch": 1.4936383728570575,
+      "grad_norm": 0.47615348856877815,
+      "learning_rate": 1.0902013133168267e-05,
+      "loss": 0.296,
+      "step": 3126
+    },
+    {
+      "epoch": 1.4941162415626308,
+      "grad_norm": 0.4861605658720827,
+      "learning_rate": 1.089676776754319e-05,
+      "loss": 0.3092,
+      "step": 3127
+    },
+    {
+      "epoch": 1.4945941102682039,
+      "grad_norm": 0.5061227916241133,
+      "learning_rate": 1.0891522153170553e-05,
+      "loss": 0.3258,
+      "step": 3128
+    },
+    {
+      "epoch": 1.495071978973777,
+      "grad_norm": 0.44470015995888657,
+      "learning_rate": 1.0886276291505395e-05,
+      "loss": 0.3264,
+      "step": 3129
+    },
+    {
+      "epoch": 1.49554984767935,
+      "grad_norm": 0.4956355952980061,
+      "learning_rate": 1.0881030184002827e-05,
+      "loss": 0.3124,
+      "step": 3130
+    },
+    {
+      "epoch": 1.4960277163849232,
+      "grad_norm": 0.47672673270267124,
+      "learning_rate": 1.0875783832118032e-05,
+      "loss": 0.3075,
+      "step": 3131
+    },
+    {
+      "epoch": 1.4965055850904965,
+      "grad_norm": 0.48537882960310214,
+      "learning_rate": 1.0870537237306245e-05,
+      "loss": 0.3264,
+      "step": 3132
+    },
+    {
+      "epoch": 1.4969834537960696,
+      "grad_norm": 0.5110345138782598,
+      "learning_rate": 1.0865290401022785e-05,
+      "loss": 0.325,
+      "step": 3133
+    },
+    {
+      "epoch": 1.4974613225016427,
+      "grad_norm": 0.4626387908464134,
+      "learning_rate": 1.0860043324723035e-05,
+      "loss": 0.315,
+      "step": 3134
+    },
+    {
+      "epoch": 1.4979391912072157,
+      "grad_norm": 0.4425903321792646,
+      "learning_rate": 1.0854796009862434e-05,
+      "loss": 0.3349,
+      "step": 3135
+    },
+    {
+      "epoch": 1.4984170599127888,
+      "grad_norm": 0.5005573549699335,
+      "learning_rate": 1.0849548457896499e-05,
+      "loss": 0.3229,
+      "step": 3136
+    },
+    {
+      "epoch": 1.4988949286183622,
+      "grad_norm": 0.4629721762379501,
+      "learning_rate": 1.0844300670280809e-05,
+      "loss": 0.3098,
+      "step": 3137
+    },
+    {
+      "epoch": 1.4993727973239352,
+      "grad_norm": 0.45157919259053075,
+      "learning_rate": 1.0839052648471002e-05,
+      "loss": 0.3338,
+      "step": 3138
+    },
+    {
+      "epoch": 1.4998506660295083,
+      "grad_norm": 0.49218452155784925,
+      "learning_rate": 1.0833804393922796e-05,
+      "loss": 0.3279,
+      "step": 3139
+    },
+    {
+      "epoch": 1.5003285347350817,
+      "grad_norm": 0.46697736877733986,
+      "learning_rate": 1.0828555908091958e-05,
+      "loss": 0.3209,
+      "step": 3140
+    },
+    {
+      "epoch": 1.5008064034406545,
+      "grad_norm": 0.502846029719294,
+      "learning_rate": 1.0823307192434325e-05,
+      "loss": 0.3211,
+      "step": 3141
+    },
+    {
+      "epoch": 1.5012842721462278,
+      "grad_norm": 0.5308861043661305,
+      "learning_rate": 1.08180582484058e-05,
+      "loss": 0.317,
+      "step": 3142
+    },
+    {
+      "epoch": 1.501762140851801,
+      "grad_norm": 0.9950929077061546,
+      "learning_rate": 1.0812809077462348e-05,
+      "loss": 0.3308,
+      "step": 3143
+    },
+    {
+      "epoch": 1.502240009557374,
+      "grad_norm": 0.4748834877132264,
+      "learning_rate": 1.0807559681059993e-05,
+      "loss": 0.3225,
+      "step": 3144
+    },
+    {
+      "epoch": 1.5027178782629473,
+      "grad_norm": 0.48714653402335817,
+      "learning_rate": 1.0802310060654832e-05,
+      "loss": 0.3294,
+      "step": 3145
+    },
+    {
+      "epoch": 1.5031957469685204,
+      "grad_norm": 0.48089986557496106,
+      "learning_rate": 1.079706021770301e-05,
+      "loss": 0.3175,
+      "step": 3146
+    },
+    {
+      "epoch": 1.5036736156740935,
+      "grad_norm": 0.5517783451396178,
+      "learning_rate": 1.0791810153660745e-05,
+      "loss": 0.3294,
+      "step": 3147
+    },
+    {
+      "epoch": 1.5041514843796668,
+      "grad_norm": 0.5650570568015437,
+      "learning_rate": 1.078655986998431e-05,
+      "loss": 0.3312,
+      "step": 3148
+    },
+    {
+      "epoch": 1.5046293530852397,
+      "grad_norm": 0.45637472471019414,
+      "learning_rate": 1.0781309368130042e-05,
+      "loss": 0.3262,
+      "step": 3149
+    },
+    {
+      "epoch": 1.505107221790813,
+      "grad_norm": 0.4653008253186181,
+      "learning_rate": 1.0776058649554336e-05,
+      "loss": 0.3049,
+      "step": 3150
+    },
+    {
+      "epoch": 1.5055850904963861,
+      "grad_norm": 0.545056353096684,
+      "learning_rate": 1.0770807715713651e-05,
+      "loss": 0.3362,
+      "step": 3151
+    },
+    {
+      "epoch": 1.5060629592019592,
+      "grad_norm": 0.4902414971923998,
+      "learning_rate": 1.0765556568064503e-05,
+      "loss": 0.3204,
+      "step": 3152
+    },
+    {
+      "epoch": 1.5065408279075325,
+      "grad_norm": 0.4608071782816651,
+      "learning_rate": 1.0760305208063467e-05,
+      "loss": 0.3208,
+      "step": 3153
+    },
+    {
+      "epoch": 1.5070186966131054,
+      "grad_norm": 0.4808143098839568,
+      "learning_rate": 1.0755053637167178e-05,
+      "loss": 0.3328,
+      "step": 3154
+    },
+    {
+      "epoch": 1.5074965653186787,
+      "grad_norm": 0.48834994164881984,
+      "learning_rate": 1.0749801856832325e-05,
+      "loss": 0.3175,
+      "step": 3155
+    },
+    {
+      "epoch": 1.5079744340242518,
+      "grad_norm": 0.49638350606323217,
+      "learning_rate": 1.0744549868515667e-05,
+      "loss": 0.3324,
+      "step": 3156
+    },
+    {
+      "epoch": 1.508452302729825,
+      "grad_norm": 0.49222615340506876,
+      "learning_rate": 1.073929767367401e-05,
+      "loss": 0.322,
+      "step": 3157
+    },
+    {
+      "epoch": 1.5089301714353982,
+      "grad_norm": 0.4564826854769966,
+      "learning_rate": 1.0734045273764217e-05,
+      "loss": 0.3301,
+      "step": 3158
+    },
+    {
+      "epoch": 1.5094080401409713,
+      "grad_norm": 0.4941738943168986,
+      "learning_rate": 1.0728792670243215e-05,
+      "loss": 0.3282,
+      "step": 3159
+    },
+    {
+      "epoch": 1.5098859088465444,
+      "grad_norm": 0.4565157299448364,
+      "learning_rate": 1.0723539864567983e-05,
+      "loss": 0.3139,
+      "step": 3160
+    },
+    {
+      "epoch": 1.5103637775521177,
+      "grad_norm": 0.46731577882122116,
+      "learning_rate": 1.0718286858195553e-05,
+      "loss": 0.3094,
+      "step": 3161
+    },
+    {
+      "epoch": 1.5108416462576906,
+      "grad_norm": 0.5380454191031079,
+      "learning_rate": 1.071303365258302e-05,
+      "loss": 0.3274,
+      "step": 3162
+    },
+    {
+      "epoch": 1.5113195149632639,
+      "grad_norm": 0.4504131152594778,
+      "learning_rate": 1.070778024918753e-05,
+      "loss": 0.3404,
+      "step": 3163
+    },
+    {
+      "epoch": 1.511797383668837,
+      "grad_norm": 0.47414368841241944,
+      "learning_rate": 1.0702526649466282e-05,
+      "loss": 0.3326,
+      "step": 3164
+    },
+    {
+      "epoch": 1.51227525237441,
+      "grad_norm": 0.48484243615521744,
+      "learning_rate": 1.0697272854876537e-05,
+      "loss": 0.3149,
+      "step": 3165
+    },
+    {
+      "epoch": 1.5127531210799834,
+      "grad_norm": 0.4926823715355633,
+      "learning_rate": 1.0692018866875598e-05,
+      "loss": 0.3274,
+      "step": 3166
+    },
+    {
+      "epoch": 1.5132309897855563,
+      "grad_norm": 0.47882967616119354,
+      "learning_rate": 1.0686764686920834e-05,
+      "loss": 0.3292,
+      "step": 3167
+    },
+    {
+      "epoch": 1.5137088584911296,
+      "grad_norm": 0.4626903520160237,
+      "learning_rate": 1.0681510316469661e-05,
+      "loss": 0.3185,
+      "step": 3168
+    },
+    {
+      "epoch": 1.5141867271967027,
+      "grad_norm": 0.560531953606567,
+      "learning_rate": 1.0676255756979548e-05,
+      "loss": 0.3301,
+      "step": 3169
+    },
+    {
+      "epoch": 1.5146645959022758,
+      "grad_norm": 0.48792270052341974,
+      "learning_rate": 1.0671001009908015e-05,
+      "loss": 0.3126,
+      "step": 3170
+    },
+    {
+      "epoch": 1.515142464607849,
+      "grad_norm": 0.5129007440119009,
+      "learning_rate": 1.066574607671264e-05,
+      "loss": 0.3181,
+      "step": 3171
+    },
+    {
+      "epoch": 1.5156203333134222,
+      "grad_norm": 0.4711038320320487,
+      "learning_rate": 1.0660490958851044e-05,
+      "loss": 0.3178,
+      "step": 3172
+    },
+    {
+      "epoch": 1.5160982020189953,
+      "grad_norm": 0.5092552024671185,
+      "learning_rate": 1.0655235657780906e-05,
+      "loss": 0.3289,
+      "step": 3173
+    },
+    {
+      "epoch": 1.5165760707245686,
+      "grad_norm": 0.5953972215800041,
+      "learning_rate": 1.0649980174959961e-05,
+      "loss": 0.3102,
+      "step": 3174
+    },
+    {
+      "epoch": 1.5170539394301414,
+      "grad_norm": 0.4703123203138388,
+      "learning_rate": 1.0644724511845976e-05,
+      "loss": 0.3143,
+      "step": 3175
+    },
+    {
+      "epoch": 1.5175318081357148,
+      "grad_norm": 0.5165416055915609,
+      "learning_rate": 1.0639468669896787e-05,
+      "loss": 0.3115,
+      "step": 3176
+    },
+    {
+      "epoch": 1.5180096768412878,
+      "grad_norm": 0.582450596746319,
+      "learning_rate": 1.0634212650570269e-05,
+      "loss": 0.3316,
+      "step": 3177
+    },
+    {
+      "epoch": 1.518487545546861,
+      "grad_norm": 0.4535700489101072,
+      "learning_rate": 1.0628956455324347e-05,
+      "loss": 0.3307,
+      "step": 3178
+    },
+    {
+      "epoch": 1.5189654142524343,
+      "grad_norm": 0.48156284886715084,
+      "learning_rate": 1.0623700085616999e-05,
+      "loss": 0.327,
+      "step": 3179
+    },
+    {
+      "epoch": 1.5194432829580071,
+      "grad_norm": 0.4838090723343779,
+      "learning_rate": 1.0618443542906251e-05,
+      "loss": 0.3302,
+      "step": 3180
+    },
+    {
+      "epoch": 1.5199211516635804,
+      "grad_norm": 0.5079130746721264,
+      "learning_rate": 1.0613186828650171e-05,
+      "loss": 0.3202,
+      "step": 3181
+    },
+    {
+      "epoch": 1.5203990203691535,
+      "grad_norm": 0.4986317378339555,
+      "learning_rate": 1.0607929944306883e-05,
+      "loss": 0.308,
+      "step": 3182
+    },
+    {
+      "epoch": 1.5208768890747266,
+      "grad_norm": 0.5275576649032065,
+      "learning_rate": 1.0602672891334552e-05,
+      "loss": 0.3332,
+      "step": 3183
+    },
+    {
+      "epoch": 1.5213547577803,
+      "grad_norm": 0.4731647767150495,
+      "learning_rate": 1.0597415671191391e-05,
+      "loss": 0.3509,
+      "step": 3184
+    },
+    {
+      "epoch": 1.521832626485873,
+      "grad_norm": 0.4763797455280031,
+      "learning_rate": 1.059215828533566e-05,
+      "loss": 0.3286,
+      "step": 3185
+    },
+    {
+      "epoch": 1.5223104951914461,
+      "grad_norm": 0.47355536256007635,
+      "learning_rate": 1.0586900735225669e-05,
+      "loss": 0.3239,
+      "step": 3186
+    },
+    {
+      "epoch": 1.5227883638970194,
+      "grad_norm": 0.4674306668056484,
+      "learning_rate": 1.0581643022319765e-05,
+      "loss": 0.3065,
+      "step": 3187
+    },
+    {
+      "epoch": 1.5232662326025923,
+      "grad_norm": 0.5012159446084873,
+      "learning_rate": 1.0576385148076346e-05,
+      "loss": 0.3288,
+      "step": 3188
+    },
+    {
+      "epoch": 1.5237441013081656,
+      "grad_norm": 0.4926801554454291,
+      "learning_rate": 1.0571127113953855e-05,
+      "loss": 0.321,
+      "step": 3189
+    },
+    {
+      "epoch": 1.5242219700137387,
+      "grad_norm": 0.5352440355066403,
+      "learning_rate": 1.0565868921410776e-05,
+      "loss": 0.3326,
+      "step": 3190
+    },
+    {
+      "epoch": 1.5246998387193118,
+      "grad_norm": 0.5707201595977832,
+      "learning_rate": 1.0560610571905642e-05,
+      "loss": 0.3247,
+      "step": 3191
+    },
+    {
+      "epoch": 1.5251777074248851,
+      "grad_norm": 0.46143438621243094,
+      "learning_rate": 1.0555352066897025e-05,
+      "loss": 0.335,
+      "step": 3192
+    },
+    {
+      "epoch": 1.525655576130458,
+      "grad_norm": 0.46914561506550345,
+      "learning_rate": 1.0550093407843538e-05,
+      "loss": 0.3035,
+      "step": 3193
+    },
+    {
+      "epoch": 1.5261334448360313,
+      "grad_norm": 0.4745615418585816,
+      "learning_rate": 1.0544834596203846e-05,
+      "loss": 0.3192,
+      "step": 3194
+    },
+    {
+      "epoch": 1.5266113135416044,
+      "grad_norm": 0.48717953885939197,
+      "learning_rate": 1.0539575633436645e-05,
+      "loss": 0.3333,
+      "step": 3195
+    },
+    {
+      "epoch": 1.5270891822471775,
+      "grad_norm": 0.4735843171248546,
+      "learning_rate": 1.0534316521000683e-05,
+      "loss": 0.3256,
+      "step": 3196
+    },
+    {
+      "epoch": 1.5275670509527508,
+      "grad_norm": 0.4796146078480026,
+      "learning_rate": 1.0529057260354744e-05,
+      "loss": 0.311,
+      "step": 3197
+    },
+    {
+      "epoch": 1.528044919658324,
+      "grad_norm": 0.4964693345022835,
+      "learning_rate": 1.052379785295765e-05,
+      "loss": 0.3328,
+      "step": 3198
+    },
+    {
+      "epoch": 1.528522788363897,
+      "grad_norm": 0.4551957925753305,
+      "learning_rate": 1.0518538300268275e-05,
+      "loss": 0.3267,
+      "step": 3199
+    },
+    {
+      "epoch": 1.5290006570694703,
+      "grad_norm": 0.4833164973926133,
+      "learning_rate": 1.0513278603745523e-05,
+      "loss": 0.3346,
+      "step": 3200
+    },
+    {
+      "epoch": 1.5294785257750432,
+      "grad_norm": 0.4976783327391536,
+      "learning_rate": 1.0508018764848336e-05,
+      "loss": 0.3214,
+      "step": 3201
+    },
+    {
+      "epoch": 1.5299563944806165,
+      "grad_norm": 0.5163458481272856,
+      "learning_rate": 1.0502758785035708e-05,
+      "loss": 0.3276,
+      "step": 3202
+    },
+    {
+      "epoch": 1.5304342631861896,
+      "grad_norm": 0.5188646785189226,
+      "learning_rate": 1.0497498665766662e-05,
+      "loss": 0.3228,
+      "step": 3203
+    },
+    {
+      "epoch": 1.5309121318917627,
+      "grad_norm": 0.4865309249115541,
+      "learning_rate": 1.049223840850026e-05,
+      "loss": 0.3232,
+      "step": 3204
+    },
+    {
+      "epoch": 1.531390000597336,
+      "grad_norm": 0.508377716195505,
+      "learning_rate": 1.0486978014695606e-05,
+      "loss": 0.3224,
+      "step": 3205
+    },
+    {
+      "epoch": 1.531867869302909,
+      "grad_norm": 0.4890171119416868,
+      "learning_rate": 1.048171748581184e-05,
+      "loss": 0.321,
+      "step": 3206
+    },
+    {
+      "epoch": 1.5323457380084822,
+      "grad_norm": 0.49225304466391234,
+      "learning_rate": 1.0476456823308144e-05,
+      "loss": 0.295,
+      "step": 3207
+    },
+    {
+      "epoch": 1.5328236067140553,
+      "grad_norm": 0.533313040699636,
+      "learning_rate": 1.0471196028643728e-05,
+      "loss": 0.3171,
+      "step": 3208
+    },
+    {
+      "epoch": 1.5333014754196284,
+      "grad_norm": 0.5656534777964286,
+      "learning_rate": 1.0465935103277845e-05,
+      "loss": 0.324,
+      "step": 3209
+    },
+    {
+      "epoch": 1.5337793441252017,
+      "grad_norm": 0.49695878757528583,
+      "learning_rate": 1.0460674048669783e-05,
+      "loss": 0.3218,
+      "step": 3210
+    },
+    {
+      "epoch": 1.5342572128307748,
+      "grad_norm": 0.5362838974016912,
+      "learning_rate": 1.0455412866278868e-05,
+      "loss": 0.3242,
+      "step": 3211
+    },
+    {
+      "epoch": 1.5347350815363479,
+      "grad_norm": 0.5144518785758332,
+      "learning_rate": 1.0450151557564457e-05,
+      "loss": 0.3334,
+      "step": 3212
+    },
+    {
+      "epoch": 1.5352129502419212,
+      "grad_norm": 0.5068918594348978,
+      "learning_rate": 1.0444890123985942e-05,
+      "loss": 0.3299,
+      "step": 3213
+    },
+    {
+      "epoch": 1.535690818947494,
+      "grad_norm": 0.4780919059955396,
+      "learning_rate": 1.043962856700276e-05,
+      "loss": 0.3199,
+      "step": 3214
+    },
+    {
+      "epoch": 1.5361686876530674,
+      "grad_norm": 0.5127686326304569,
+      "learning_rate": 1.0434366888074363e-05,
+      "loss": 0.3193,
+      "step": 3215
+    },
+    {
+      "epoch": 1.5366465563586404,
+      "grad_norm": 0.49589298134321574,
+      "learning_rate": 1.0429105088660253e-05,
+      "loss": 0.3226,
+      "step": 3216
+    },
+    {
+      "epoch": 1.5371244250642135,
+      "grad_norm": 0.4860233798195414,
+      "learning_rate": 1.0423843170219966e-05,
+      "loss": 0.3365,
+      "step": 3217
+    },
+    {
+      "epoch": 1.5376022937697869,
+      "grad_norm": 0.4714518886662182,
+      "learning_rate": 1.0418581134213055e-05,
+      "loss": 0.3244,
+      "step": 3218
+    },
+    {
+      "epoch": 1.53808016247536,
+      "grad_norm": 0.4955762103566397,
+      "learning_rate": 1.0413318982099124e-05,
+      "loss": 0.3226,
+      "step": 3219
+    },
+    {
+      "epoch": 1.538558031180933,
+      "grad_norm": 0.4973793360007407,
+      "learning_rate": 1.0408056715337797e-05,
+      "loss": 0.3154,
+      "step": 3220
+    },
+    {
+      "epoch": 1.5390358998865061,
+      "grad_norm": 0.487987592402328,
+      "learning_rate": 1.0402794335388733e-05,
+      "loss": 0.3219,
+      "step": 3221
+    },
+    {
+      "epoch": 1.5395137685920792,
+      "grad_norm": 0.5013231822647967,
+      "learning_rate": 1.0397531843711626e-05,
+      "loss": 0.3144,
+      "step": 3222
+    },
+    {
+      "epoch": 1.5399916372976525,
+      "grad_norm": 0.5122546611335848,
+      "learning_rate": 1.0392269241766199e-05,
+      "loss": 0.3161,
+      "step": 3223
+    },
+    {
+      "epoch": 1.5404695060032256,
+      "grad_norm": 0.5023537156876484,
+      "learning_rate": 1.0387006531012204e-05,
+      "loss": 0.3307,
+      "step": 3224
+    },
+    {
+      "epoch": 1.5409473747087987,
+      "grad_norm": 0.46724605937313596,
+      "learning_rate": 1.0381743712909424e-05,
+      "loss": 0.3229,
+      "step": 3225
+    },
+    {
+      "epoch": 1.541425243414372,
+      "grad_norm": 0.47093375402216126,
+      "learning_rate": 1.0376480788917676e-05,
+      "loss": 0.3279,
+      "step": 3226
+    },
+    {
+      "epoch": 1.541903112119945,
+      "grad_norm": 0.46994586921571396,
+      "learning_rate": 1.0371217760496792e-05,
+      "loss": 0.3236,
+      "step": 3227
+    },
+    {
+      "epoch": 1.5423809808255182,
+      "grad_norm": 0.4843800423622691,
+      "learning_rate": 1.0365954629106652e-05,
+      "loss": 0.3085,
+      "step": 3228
+    },
+    {
+      "epoch": 1.5428588495310913,
+      "grad_norm": 0.4772582302201957,
+      "learning_rate": 1.0360691396207155e-05,
+      "loss": 0.3076,
+      "step": 3229
+    },
+    {
+      "epoch": 1.5433367182366644,
+      "grad_norm": 0.481322355383063,
+      "learning_rate": 1.0355428063258224e-05,
+      "loss": 0.3068,
+      "step": 3230
+    },
+    {
+      "epoch": 1.5438145869422377,
+      "grad_norm": 0.511101779977841,
+      "learning_rate": 1.0350164631719816e-05,
+      "loss": 0.3217,
+      "step": 3231
+    },
+    {
+      "epoch": 1.5442924556478108,
+      "grad_norm": 0.45963843948522115,
+      "learning_rate": 1.0344901103051923e-05,
+      "loss": 0.343,
+      "step": 3232
+    },
+    {
+      "epoch": 1.544770324353384,
+      "grad_norm": 0.48757882063659436,
+      "learning_rate": 1.033963747871454e-05,
+      "loss": 0.3232,
+      "step": 3233
+    },
+    {
+      "epoch": 1.545248193058957,
+      "grad_norm": 0.5055497495839371,
+      "learning_rate": 1.0334373760167718e-05,
+      "loss": 0.3234,
+      "step": 3234
+    },
+    {
+      "epoch": 1.54572606176453,
+      "grad_norm": 0.49927220968524727,
+      "learning_rate": 1.0329109948871512e-05,
+      "loss": 0.3197,
+      "step": 3235
+    },
+    {
+      "epoch": 1.5462039304701034,
+      "grad_norm": 0.48763244928090066,
+      "learning_rate": 1.032384604628601e-05,
+      "loss": 0.3222,
+      "step": 3236
+    },
+    {
+      "epoch": 1.5466817991756765,
+      "grad_norm": 0.464310330481023,
+      "learning_rate": 1.0318582053871326e-05,
+      "loss": 0.319,
+      "step": 3237
+    },
+    {
+      "epoch": 1.5471596678812496,
+      "grad_norm": 0.4970739536764959,
+      "learning_rate": 1.0313317973087603e-05,
+      "loss": 0.3121,
+      "step": 3238
+    },
+    {
+      "epoch": 1.547637536586823,
+      "grad_norm": 0.5697514900420939,
+      "learning_rate": 1.0308053805394998e-05,
+      "loss": 0.3153,
+      "step": 3239
+    },
+    {
+      "epoch": 1.5481154052923958,
+      "grad_norm": 0.48345598443017196,
+      "learning_rate": 1.0302789552253702e-05,
+      "loss": 0.3194,
+      "step": 3240
+    },
+    {
+      "epoch": 1.548593273997969,
+      "grad_norm": 0.4761641613916091,
+      "learning_rate": 1.0297525215123927e-05,
+      "loss": 0.3175,
+      "step": 3241
+    },
+    {
+      "epoch": 1.5490711427035422,
+      "grad_norm": 0.4899651453697106,
+      "learning_rate": 1.0292260795465905e-05,
+      "loss": 0.3289,
+      "step": 3242
+    },
+    {
+      "epoch": 1.5495490114091153,
+      "grad_norm": 0.5535534354531657,
+      "learning_rate": 1.0286996294739895e-05,
+      "loss": 0.3321,
+      "step": 3243
+    },
+    {
+      "epoch": 1.5500268801146886,
+      "grad_norm": 0.49382296783652835,
+      "learning_rate": 1.0281731714406172e-05,
+      "loss": 0.3274,
+      "step": 3244
+    },
+    {
+      "epoch": 1.5505047488202617,
+      "grad_norm": 0.48868429365584,
+      "learning_rate": 1.0276467055925044e-05,
+      "loss": 0.315,
+      "step": 3245
+    },
+    {
+      "epoch": 1.5509826175258348,
+      "grad_norm": 0.4900357280511654,
+      "learning_rate": 1.027120232075683e-05,
+      "loss": 0.3273,
+      "step": 3246
+    },
+    {
+      "epoch": 1.551460486231408,
+      "grad_norm": 0.44278332935205783,
+      "learning_rate": 1.0265937510361876e-05,
+      "loss": 0.3363,
+      "step": 3247
+    },
+    {
+      "epoch": 1.551938354936981,
+      "grad_norm": 0.5025588525393508,
+      "learning_rate": 1.0260672626200548e-05,
+      "loss": 0.3467,
+      "step": 3248
+    },
+    {
+      "epoch": 1.5524162236425543,
+      "grad_norm": 0.48485895773001003,
+      "learning_rate": 1.0255407669733235e-05,
+      "loss": 0.3179,
+      "step": 3249
+    },
+    {
+      "epoch": 1.5528940923481274,
+      "grad_norm": 0.48681474256092666,
+      "learning_rate": 1.0250142642420335e-05,
+      "loss": 0.3131,
+      "step": 3250
+    },
+    {
+      "epoch": 1.5533719610537005,
+      "grad_norm": 0.5072938805616622,
+      "learning_rate": 1.024487754572228e-05,
+      "loss": 0.3048,
+      "step": 3251
+    },
+    {
+      "epoch": 1.5538498297592738,
+      "grad_norm": 0.464960127721567,
+      "learning_rate": 1.0239612381099515e-05,
+      "loss": 0.3081,
+      "step": 3252
+    },
+    {
+      "epoch": 1.5543276984648466,
+      "grad_norm": 0.5029051775399344,
+      "learning_rate": 1.02343471500125e-05,
+      "loss": 0.326,
+      "step": 3253
+    },
+    {
+      "epoch": 1.55480556717042,
+      "grad_norm": 0.4760931121940165,
+      "learning_rate": 1.0229081853921719e-05,
+      "loss": 0.3373,
+      "step": 3254
+    },
+    {
+      "epoch": 1.555283435875993,
+      "grad_norm": 0.5211628472885691,
+      "learning_rate": 1.0223816494287675e-05,
+      "loss": 0.3194,
+      "step": 3255
+    },
+    {
+      "epoch": 1.5557613045815661,
+      "grad_norm": 4.00869479916537,
+      "learning_rate": 1.021855107257088e-05,
+      "loss": 0.3147,
+      "step": 3256
+    },
+    {
+      "epoch": 1.5562391732871395,
+      "grad_norm": 0.5278043616111543,
+      "learning_rate": 1.0213285590231877e-05,
+      "loss": 0.2986,
+      "step": 3257
+    },
+    {
+      "epoch": 1.5567170419927125,
+      "grad_norm": 0.48164426331213295,
+      "learning_rate": 1.020802004873121e-05,
+      "loss": 0.3253,
+      "step": 3258
+    },
+    {
+      "epoch": 1.5571949106982856,
+      "grad_norm": 0.5500585812101135,
+      "learning_rate": 1.0202754449529453e-05,
+      "loss": 0.3209,
+      "step": 3259
+    },
+    {
+      "epoch": 1.557672779403859,
+      "grad_norm": 0.49446254990048405,
+      "learning_rate": 1.0197488794087188e-05,
+      "loss": 0.3198,
+      "step": 3260
+    },
+    {
+      "epoch": 1.5581506481094318,
+      "grad_norm": 0.4657507878684054,
+      "learning_rate": 1.0192223083865013e-05,
+      "loss": 0.3256,
+      "step": 3261
+    },
+    {
+      "epoch": 1.5586285168150051,
+      "grad_norm": 0.4477764129200038,
+      "learning_rate": 1.0186957320323547e-05,
+      "loss": 0.3271,
+      "step": 3262
+    },
+    {
+      "epoch": 1.5591063855205782,
+      "grad_norm": 0.48278945630965814,
+      "learning_rate": 1.0181691504923421e-05,
+      "loss": 0.307,
+      "step": 3263
+    },
+    {
+      "epoch": 1.5595842542261513,
+      "grad_norm": 1.1355856675651534,
+      "learning_rate": 1.0176425639125273e-05,
+      "loss": 0.3271,
+      "step": 3264
+    },
+    {
+      "epoch": 1.5600621229317246,
+      "grad_norm": 0.4819769035656966,
+      "learning_rate": 1.0171159724389766e-05,
+      "loss": 0.3178,
+      "step": 3265
+    },
+    {
+      "epoch": 1.5605399916372975,
+      "grad_norm": 0.4786394343649937,
+      "learning_rate": 1.016589376217757e-05,
+      "loss": 0.3382,
+      "step": 3266
+    },
+    {
+      "epoch": 1.5610178603428708,
+      "grad_norm": 0.48736910260272115,
+      "learning_rate": 1.016062775394937e-05,
+      "loss": 0.3196,
+      "step": 3267
+    },
+    {
+      "epoch": 1.561495729048444,
+      "grad_norm": 0.46448638746562965,
+      "learning_rate": 1.0155361701165867e-05,
+      "loss": 0.3148,
+      "step": 3268
+    },
+    {
+      "epoch": 1.561973597754017,
+      "grad_norm": 0.49431236575933374,
+      "learning_rate": 1.0150095605287768e-05,
+      "loss": 0.3093,
+      "step": 3269
+    },
+    {
+      "epoch": 1.5624514664595903,
+      "grad_norm": 0.4686420840220977,
+      "learning_rate": 1.0144829467775794e-05,
+      "loss": 0.3168,
+      "step": 3270
+    },
+    {
+      "epoch": 1.5629293351651634,
+      "grad_norm": 0.44817236471431465,
+      "learning_rate": 1.0139563290090679e-05,
+      "loss": 0.3154,
+      "step": 3271
+    },
+    {
+      "epoch": 1.5634072038707365,
+      "grad_norm": 0.46921012393643297,
+      "learning_rate": 1.0134297073693173e-05,
+      "loss": 0.3257,
+      "step": 3272
+    },
+    {
+      "epoch": 1.5638850725763098,
+      "grad_norm": 0.46434620917562563,
+      "learning_rate": 1.0129030820044024e-05,
+      "loss": 0.3303,
+      "step": 3273
+    },
+    {
+      "epoch": 1.5643629412818827,
+      "grad_norm": 0.6236012194460561,
+      "learning_rate": 1.0123764530604003e-05,
+      "loss": 0.3127,
+      "step": 3274
+    },
+    {
+      "epoch": 1.564840809987456,
+      "grad_norm": 0.4944268701162515,
+      "learning_rate": 1.0118498206833886e-05,
+      "loss": 0.3269,
+      "step": 3275
+    },
+    {
+      "epoch": 1.565318678693029,
+      "grad_norm": 0.5192897092562184,
+      "learning_rate": 1.0113231850194455e-05,
+      "loss": 0.3332,
+      "step": 3276
+    },
+    {
+      "epoch": 1.5657965473986022,
+      "grad_norm": 0.4841696394389553,
+      "learning_rate": 1.0107965462146507e-05,
+      "loss": 0.3369,
+      "step": 3277
+    },
+    {
+      "epoch": 1.5662744161041755,
+      "grad_norm": 0.5059546085514659,
+      "learning_rate": 1.0102699044150845e-05,
+      "loss": 0.3233,
+      "step": 3278
+    },
+    {
+      "epoch": 1.5667522848097484,
+      "grad_norm": 0.4818467825525027,
+      "learning_rate": 1.0097432597668279e-05,
+      "loss": 0.3099,
+      "step": 3279
+    },
+    {
+      "epoch": 1.5672301535153217,
+      "grad_norm": 0.4889420234246756,
+      "learning_rate": 1.0092166124159628e-05,
+      "loss": 0.3182,
+      "step": 3280
+    },
+    {
+      "epoch": 1.5677080222208948,
+      "grad_norm": 0.4826052928580796,
+      "learning_rate": 1.0086899625085725e-05,
+      "loss": 0.3247,
+      "step": 3281
+    },
+    {
+      "epoch": 1.5681858909264679,
+      "grad_norm": 0.49036739531543294,
+      "learning_rate": 1.0081633101907393e-05,
+      "loss": 0.329,
+      "step": 3282
+    },
+    {
+      "epoch": 1.5686637596320412,
+      "grad_norm": 0.4574906703009941,
+      "learning_rate": 1.007636655608548e-05,
+      "loss": 0.3157,
+      "step": 3283
+    },
+    {
+      "epoch": 1.5691416283376143,
+      "grad_norm": 0.4738908087609072,
+      "learning_rate": 1.0071099989080833e-05,
+      "loss": 0.3433,
+      "step": 3284
+    },
+    {
+      "epoch": 1.5696194970431874,
+      "grad_norm": 0.46547191964568313,
+      "learning_rate": 1.0065833402354302e-05,
+      "loss": 0.3208,
+      "step": 3285
+    },
+    {
+      "epoch": 1.5700973657487607,
+      "grad_norm": 0.5839094375309312,
+      "learning_rate": 1.0060566797366744e-05,
+      "loss": 0.3255,
+      "step": 3286
+    },
+    {
+      "epoch": 1.5705752344543336,
+      "grad_norm": 0.4947521545103725,
+      "learning_rate": 1.005530017557903e-05,
+      "loss": 0.3208,
+      "step": 3287
+    },
+    {
+      "epoch": 1.5710531031599069,
+      "grad_norm": 0.5025064223155531,
+      "learning_rate": 1.005003353845202e-05,
+      "loss": 0.3076,
+      "step": 3288
+    },
+    {
+      "epoch": 1.57153097186548,
+      "grad_norm": 0.4669404424154083,
+      "learning_rate": 1.0044766887446586e-05,
+      "loss": 0.3276,
+      "step": 3289
+    },
+    {
+      "epoch": 1.572008840571053,
+      "grad_norm": 0.5009595930368698,
+      "learning_rate": 1.003950022402361e-05,
+      "loss": 0.3106,
+      "step": 3290
+    },
+    {
+      "epoch": 1.5724867092766264,
+      "grad_norm": 0.4948274057807226,
+      "learning_rate": 1.0034233549643969e-05,
+      "loss": 0.3424,
+      "step": 3291
+    },
+    {
+      "epoch": 1.5729645779821992,
+      "grad_norm": 0.45361579014587694,
+      "learning_rate": 1.0028966865768546e-05,
+      "loss": 0.3258,
+      "step": 3292
+    },
+    {
+      "epoch": 1.5734424466877726,
+      "grad_norm": 0.7899603895576185,
+      "learning_rate": 1.0023700173858224e-05,
+      "loss": 0.3189,
+      "step": 3293
+    },
+    {
+      "epoch": 1.5739203153933456,
+      "grad_norm": 0.45700787119484876,
+      "learning_rate": 1.0018433475373891e-05,
+      "loss": 0.3257,
+      "step": 3294
+    },
+    {
+      "epoch": 1.5743981840989187,
+      "grad_norm": 0.4721568980282006,
+      "learning_rate": 1.0013166771776441e-05,
+      "loss": 0.3356,
+      "step": 3295
+    },
+    {
+      "epoch": 1.574876052804492,
+      "grad_norm": 0.4854731194781878,
+      "learning_rate": 1.0007900064526756e-05,
+      "loss": 0.3142,
+      "step": 3296
+    },
+    {
+      "epoch": 1.5753539215100651,
+      "grad_norm": 0.4880060979457883,
+      "learning_rate": 1.0002633355085734e-05,
+      "loss": 0.3223,
+      "step": 3297
+    },
+    {
+      "epoch": 1.5758317902156382,
+      "grad_norm": 0.4561283926612506,
+      "learning_rate": 9.997366644914266e-06,
+      "loss": 0.3206,
+      "step": 3298
+    },
+    {
+      "epoch": 1.5763096589212116,
+      "grad_norm": 0.47632004228079955,
+      "learning_rate": 9.992099935473244e-06,
+      "loss": 0.3128,
+      "step": 3299
+    },
+    {
+      "epoch": 1.5767875276267844,
+      "grad_norm": 0.4847083170328396,
+      "learning_rate": 9.986833228223562e-06,
+      "loss": 0.2984,
+      "step": 3300
+    },
+    {
+      "epoch": 1.5772653963323577,
+      "grad_norm": 0.473252193657243,
+      "learning_rate": 9.98156652462611e-06,
+      "loss": 0.3223,
+      "step": 3301
+    },
+    {
+      "epoch": 1.5777432650379308,
+      "grad_norm": 0.5012398605644248,
+      "learning_rate": 9.976299826141776e-06,
+      "loss": 0.3186,
+      "step": 3302
+    },
+    {
+      "epoch": 1.578221133743504,
+      "grad_norm": 0.5394851324327755,
+      "learning_rate": 9.971033134231458e-06,
+      "loss": 0.3228,
+      "step": 3303
+    },
+    {
+      "epoch": 1.5786990024490772,
+      "grad_norm": 0.46951070560264985,
+      "learning_rate": 9.965766450356031e-06,
+      "loss": 0.3249,
+      "step": 3304
+    },
+    {
+      "epoch": 1.57917687115465,
+      "grad_norm": 0.47934334873980533,
+      "learning_rate": 9.96049977597639e-06,
+      "loss": 0.3197,
+      "step": 3305
+    },
+    {
+      "epoch": 1.5796547398602234,
+      "grad_norm": 0.514993715660872,
+      "learning_rate": 9.955233112553416e-06,
+      "loss": 0.3037,
+      "step": 3306
+    },
+    {
+      "epoch": 1.5801326085657965,
+      "grad_norm": 0.4812618916635242,
+      "learning_rate": 9.949966461547984e-06,
+      "loss": 0.3224,
+      "step": 3307
+    },
+    {
+      "epoch": 1.5806104772713696,
+      "grad_norm": 0.49269064197863277,
+      "learning_rate": 9.944699824420973e-06,
+      "loss": 0.3335,
+      "step": 3308
+    },
+    {
+      "epoch": 1.581088345976943,
+      "grad_norm": 0.4888765058606419,
+      "learning_rate": 9.939433202633258e-06,
+      "loss": 0.3188,
+      "step": 3309
+    },
+    {
+      "epoch": 1.581566214682516,
+      "grad_norm": 0.47130242909704306,
+      "learning_rate": 9.934166597645703e-06,
+      "loss": 0.327,
+      "step": 3310
+    },
+    {
+      "epoch": 1.582044083388089,
+      "grad_norm": 0.4980212169856602,
+      "learning_rate": 9.92890001091917e-06,
+      "loss": 0.3144,
+      "step": 3311
+    },
+    {
+      "epoch": 1.5825219520936624,
+      "grad_norm": 0.5047705095751753,
+      "learning_rate": 9.923633443914522e-06,
+      "loss": 0.3098,
+      "step": 3312
+    },
+    {
+      "epoch": 1.5829998207992353,
+      "grad_norm": 0.5055307123773655,
+      "learning_rate": 9.91836689809261e-06,
+      "loss": 0.3249,
+      "step": 3313
+    },
+    {
+      "epoch": 1.5834776895048086,
+      "grad_norm": 0.45564093127252564,
+      "learning_rate": 9.913100374914279e-06,
+      "loss": 0.3027,
+      "step": 3314
+    },
+    {
+      "epoch": 1.5839555582103817,
+      "grad_norm": 0.48312085794596293,
+      "learning_rate": 9.907833875840374e-06,
+      "loss": 0.3143,
+      "step": 3315
+    },
+    {
+      "epoch": 1.5844334269159548,
+      "grad_norm": 0.4737622360232326,
+      "learning_rate": 9.902567402331723e-06,
+      "loss": 0.3171,
+      "step": 3316
+    },
+    {
+      "epoch": 1.584911295621528,
+      "grad_norm": 0.46646546545393947,
+      "learning_rate": 9.897300955849157e-06,
+      "loss": 0.306,
+      "step": 3317
+    },
+    {
+      "epoch": 1.585389164327101,
+      "grad_norm": 0.5382467045040225,
+      "learning_rate": 9.892034537853495e-06,
+      "loss": 0.3118,
+      "step": 3318
+    },
+    {
+      "epoch": 1.5858670330326743,
+      "grad_norm": 0.46491097933901154,
+      "learning_rate": 9.886768149805546e-06,
+      "loss": 0.3236,
+      "step": 3319
+    },
+    {
+      "epoch": 1.5863449017382474,
+      "grad_norm": 0.48990510634701584,
+      "learning_rate": 9.881501793166117e-06,
+      "loss": 0.32,
+      "step": 3320
+    },
+    {
+      "epoch": 1.5868227704438205,
+      "grad_norm": 0.4872711839238837,
+      "learning_rate": 9.876235469395999e-06,
+      "loss": 0.3242,
+      "step": 3321
+    },
+    {
+      "epoch": 1.5873006391493938,
+      "grad_norm": 0.46588753435922464,
+      "learning_rate": 9.870969179955978e-06,
+      "loss": 0.3045,
+      "step": 3322
+    },
+    {
+      "epoch": 1.5877785078549669,
+      "grad_norm": 0.4365913398780867,
+      "learning_rate": 9.86570292630683e-06,
+      "loss": 0.31,
+      "step": 3323
+    },
+    {
+      "epoch": 1.58825637656054,
+      "grad_norm": 0.5319535297106422,
+      "learning_rate": 9.860436709909324e-06,
+      "loss": 0.3159,
+      "step": 3324
+    },
+    {
+      "epoch": 1.5887342452661133,
+      "grad_norm": 0.49154790591120767,
+      "learning_rate": 9.85517053222421e-06,
+      "loss": 0.3555,
+      "step": 3325
+    },
+    {
+      "epoch": 1.5892121139716862,
+      "grad_norm": 0.4843832531975595,
+      "learning_rate": 9.849904394712237e-06,
+      "loss": 0.312,
+      "step": 3326
+    },
+    {
+      "epoch": 1.5896899826772595,
+      "grad_norm": 0.466629508182958,
+      "learning_rate": 9.84463829883414e-06,
+      "loss": 0.3207,
+      "step": 3327
+    },
+    {
+      "epoch": 1.5901678513828326,
+      "grad_norm": 0.5132633803194183,
+      "learning_rate": 9.839372246050633e-06,
+      "loss": 0.3106,
+      "step": 3328
+    },
+    {
+      "epoch": 1.5906457200884057,
+      "grad_norm": 0.47341023027177437,
+      "learning_rate": 9.834106237822434e-06,
+      "loss": 0.3185,
+      "step": 3329
+    },
+    {
+      "epoch": 1.591123588793979,
+      "grad_norm": 0.49348692682013234,
+      "learning_rate": 9.82884027561024e-06,
+      "loss": 0.3286,
+      "step": 3330
+    },
+    {
+      "epoch": 1.5916014574995518,
+      "grad_norm": 0.4663364393993305,
+      "learning_rate": 9.823574360874732e-06,
+      "loss": 0.3146,
+      "step": 3331
+    },
+    {
+      "epoch": 1.5920793262051252,
+      "grad_norm": 0.5075761152310269,
+      "learning_rate": 9.818308495076582e-06,
+      "loss": 0.3231,
+      "step": 3332
+    },
+    {
+      "epoch": 1.5925571949106982,
+      "grad_norm": 0.4901694360126208,
+      "learning_rate": 9.813042679676453e-06,
+      "loss": 0.3078,
+      "step": 3333
+    },
+    {
+      "epoch": 1.5930350636162713,
+      "grad_norm": 0.47515746573089945,
+      "learning_rate": 9.807776916134985e-06,
+      "loss": 0.3177,
+      "step": 3334
+    },
+    {
+      "epoch": 1.5935129323218447,
+      "grad_norm": 0.45885158737432963,
+      "learning_rate": 9.802511205912815e-06,
+      "loss": 0.333,
+      "step": 3335
+    },
+    {
+      "epoch": 1.5939908010274177,
+      "grad_norm": 0.4932981943331556,
+      "learning_rate": 9.797245550470549e-06,
+      "loss": 0.2977,
+      "step": 3336
+    },
+    {
+      "epoch": 1.5944686697329908,
+      "grad_norm": 0.4566237024471887,
+      "learning_rate": 9.791979951268791e-06,
+      "loss": 0.3147,
+      "step": 3337
+    },
+    {
+      "epoch": 1.5949465384385642,
+      "grad_norm": 0.4767957332887537,
+      "learning_rate": 9.786714409768127e-06,
+      "loss": 0.3114,
+      "step": 3338
+    },
+    {
+      "epoch": 1.595424407144137,
+      "grad_norm": 0.583636525688097,
+      "learning_rate": 9.78144892742912e-06,
+      "loss": 0.3029,
+      "step": 3339
+    },
+    {
+      "epoch": 1.5959022758497103,
+      "grad_norm": 0.5062035780498737,
+      "learning_rate": 9.776183505712327e-06,
+      "loss": 0.3161,
+      "step": 3340
+    },
+    {
+      "epoch": 1.5963801445552834,
+      "grad_norm": 0.4838854040827197,
+      "learning_rate": 9.770918146078283e-06,
+      "loss": 0.3555,
+      "step": 3341
+    },
+    {
+      "epoch": 1.5968580132608565,
+      "grad_norm": 0.49830687795336376,
+      "learning_rate": 9.765652849987504e-06,
+      "loss": 0.3151,
+      "step": 3342
+    },
+    {
+      "epoch": 1.5973358819664298,
+      "grad_norm": 0.49241625719381726,
+      "learning_rate": 9.760387618900488e-06,
+      "loss": 0.3171,
+      "step": 3343
+    },
+    {
+      "epoch": 1.5978137506720027,
+      "grad_norm": 0.5523584158121128,
+      "learning_rate": 9.755122454277723e-06,
+      "loss": 0.3274,
+      "step": 3344
+    },
+    {
+      "epoch": 1.598291619377576,
+      "grad_norm": 0.4845078185424916,
+      "learning_rate": 9.749857357579667e-06,
+      "loss": 0.3149,
+      "step": 3345
+    },
+    {
+      "epoch": 1.5987694880831491,
+      "grad_norm": 0.5024697204031072,
+      "learning_rate": 9.744592330266769e-06,
+      "loss": 0.314,
+      "step": 3346
+    },
+    {
+      "epoch": 1.5992473567887222,
+      "grad_norm": 0.45934959722680274,
+      "learning_rate": 9.739327373799454e-06,
+      "loss": 0.3142,
+      "step": 3347
+    },
+    {
+      "epoch": 1.5997252254942955,
+      "grad_norm": 0.464086055829073,
+      "learning_rate": 9.734062489638127e-06,
+      "loss": 0.323,
+      "step": 3348
+    },
+    {
+      "epoch": 1.6002030941998686,
+      "grad_norm": 0.4831840810555116,
+      "learning_rate": 9.728797679243172e-06,
+      "loss": 0.3306,
+      "step": 3349
+    },
+    {
+      "epoch": 1.6006809629054417,
+      "grad_norm": 0.4764342274703041,
+      "learning_rate": 9.723532944074961e-06,
+      "loss": 0.3201,
+      "step": 3350
+    },
+    {
+      "epoch": 1.601158831611015,
+      "grad_norm": 0.48070401161567705,
+      "learning_rate": 9.71826828559383e-06,
+      "loss": 0.3281,
+      "step": 3351
+    },
+    {
+      "epoch": 1.601636700316588,
+      "grad_norm": 0.4708155298765893,
+      "learning_rate": 9.71300370526011e-06,
+      "loss": 0.322,
+      "step": 3352
+    },
+    {
+      "epoch": 1.6021145690221612,
+      "grad_norm": 0.4767194553320532,
+      "learning_rate": 9.7077392045341e-06,
+      "loss": 0.3258,
+      "step": 3353
+    },
+    {
+      "epoch": 1.6025924377277343,
+      "grad_norm": 0.479523768892733,
+      "learning_rate": 9.702474784876075e-06,
+      "loss": 0.3199,
+      "step": 3354
+    },
+    {
+      "epoch": 1.6030703064333074,
+      "grad_norm": 0.5236359911668674,
+      "learning_rate": 9.6972104477463e-06,
+      "loss": 0.3237,
+      "step": 3355
+    },
+    {
+      "epoch": 1.6035481751388807,
+      "grad_norm": 0.46485496717447494,
+      "learning_rate": 9.691946194605007e-06,
+      "loss": 0.3084,
+      "step": 3356
+    },
+    {
+      "epoch": 1.6040260438444536,
+      "grad_norm": 0.48134675159435425,
+      "learning_rate": 9.686682026912402e-06,
+      "loss": 0.3084,
+      "step": 3357
+    },
+    {
+      "epoch": 1.604503912550027,
+      "grad_norm": 0.49397380566853843,
+      "learning_rate": 9.681417946128677e-06,
+      "loss": 0.2973,
+      "step": 3358
+    },
+    {
+      "epoch": 1.6049817812556,
+      "grad_norm": 0.48607625348138517,
+      "learning_rate": 9.676153953713996e-06,
+      "loss": 0.3212,
+      "step": 3359
+    },
+    {
+      "epoch": 1.605459649961173,
+      "grad_norm": 0.487640811201148,
+      "learning_rate": 9.670890051128493e-06,
+      "loss": 0.3181,
+      "step": 3360
+    },
+    {
+      "epoch": 1.6059375186667464,
+      "grad_norm": 0.47749175931112464,
+      "learning_rate": 9.665626239832286e-06,
+      "loss": 0.3211,
+      "step": 3361
+    },
+    {
+      "epoch": 1.6064153873723195,
+      "grad_norm": 0.7933182151405382,
+      "learning_rate": 9.660362521285463e-06,
+      "loss": 0.3277,
+      "step": 3362
+    },
+    {
+      "epoch": 1.6068932560778926,
+      "grad_norm": 0.47250312135112554,
+      "learning_rate": 9.655098896948083e-06,
+      "loss": 0.3167,
+      "step": 3363
+    },
+    {
+      "epoch": 1.6073711247834659,
+      "grad_norm": 0.4693883987261192,
+      "learning_rate": 9.649835368280186e-06,
+      "loss": 0.3241,
+      "step": 3364
+    },
+    {
+      "epoch": 1.6078489934890388,
+      "grad_norm": 0.47939105139177257,
+      "learning_rate": 9.644571936741778e-06,
+      "loss": 0.3253,
+      "step": 3365
+    },
+    {
+      "epoch": 1.608326862194612,
+      "grad_norm": 0.471799876196094,
+      "learning_rate": 9.639308603792847e-06,
+      "loss": 0.2978,
+      "step": 3366
+    },
+    {
+      "epoch": 1.6088047309001852,
+      "grad_norm": 0.5037783606345849,
+      "learning_rate": 9.634045370893348e-06,
+      "loss": 0.3178,
+      "step": 3367
+    },
+    {
+      "epoch": 1.6092825996057583,
+      "grad_norm": 0.5003960430326154,
+      "learning_rate": 9.628782239503208e-06,
+      "loss": 0.3416,
+      "step": 3368
+    },
+    {
+      "epoch": 1.6097604683113316,
+      "grad_norm": 0.48869672104414125,
+      "learning_rate": 9.623519211082325e-06,
+      "loss": 0.3125,
+      "step": 3369
+    },
+    {
+      "epoch": 1.6102383370169044,
+      "grad_norm": 0.49864918116539536,
+      "learning_rate": 9.618256287090576e-06,
+      "loss": 0.3244,
+      "step": 3370
+    },
+    {
+      "epoch": 1.6107162057224778,
+      "grad_norm": 0.5019807545385679,
+      "learning_rate": 9.612993468987796e-06,
+      "loss": 0.3138,
+      "step": 3371
+    },
+    {
+      "epoch": 1.6111940744280508,
+      "grad_norm": 0.4683993877429079,
+      "learning_rate": 9.6077307582338e-06,
+      "loss": 0.332,
+      "step": 3372
+    },
+    {
+      "epoch": 1.611671943133624,
+      "grad_norm": 0.4605238593535419,
+      "learning_rate": 9.602468156288374e-06,
+      "loss": 0.3209,
+      "step": 3373
+    },
+    {
+      "epoch": 1.6121498118391973,
+      "grad_norm": 0.47258089848935597,
+      "learning_rate": 9.597205664611269e-06,
+      "loss": 0.3218,
+      "step": 3374
+    },
+    {
+      "epoch": 1.6126276805447703,
+      "grad_norm": 0.48987083989083485,
+      "learning_rate": 9.591943284662206e-06,
+      "loss": 0.3216,
+      "step": 3375
+    },
+    {
+      "epoch": 1.6131055492503434,
+      "grad_norm": 0.45509362905511785,
+      "learning_rate": 9.586681017900881e-06,
+      "loss": 0.3151,
+      "step": 3376
+    },
+    {
+      "epoch": 1.6135834179559168,
+      "grad_norm": 0.5226274148001115,
+      "learning_rate": 9.581418865786948e-06,
+      "loss": 0.296,
+      "step": 3377
+    },
+    {
+      "epoch": 1.6140612866614896,
+      "grad_norm": 0.4972753917887101,
+      "learning_rate": 9.576156829780038e-06,
+      "loss": 0.3083,
+      "step": 3378
+    },
+    {
+      "epoch": 1.614539155367063,
+      "grad_norm": 0.5066525482936579,
+      "learning_rate": 9.570894911339748e-06,
+      "loss": 0.3135,
+      "step": 3379
+    },
+    {
+      "epoch": 1.615017024072636,
+      "grad_norm": 0.4594810626516822,
+      "learning_rate": 9.56563311192564e-06,
+      "loss": 0.3315,
+      "step": 3380
+    },
+    {
+      "epoch": 1.6154948927782091,
+      "grad_norm": 0.5042615947206254,
+      "learning_rate": 9.560371432997244e-06,
+      "loss": 0.3119,
+      "step": 3381
+    },
+    {
+      "epoch": 1.6159727614837824,
+      "grad_norm": 0.4914519937789319,
+      "learning_rate": 9.55510987601406e-06,
+      "loss": 0.3276,
+      "step": 3382
+    },
+    {
+      "epoch": 1.6164506301893553,
+      "grad_norm": 0.45163236705607995,
+      "learning_rate": 9.549848442435547e-06,
+      "loss": 0.3236,
+      "step": 3383
+    },
+    {
+      "epoch": 1.6169284988949286,
+      "grad_norm": 0.4853699956930337,
+      "learning_rate": 9.544587133721133e-06,
+      "loss": 0.3262,
+      "step": 3384
+    },
+    {
+      "epoch": 1.6174063676005017,
+      "grad_norm": 0.5027837659953032,
+      "learning_rate": 9.53932595133022e-06,
+      "loss": 0.3159,
+      "step": 3385
+    },
+    {
+      "epoch": 1.6178842363060748,
+      "grad_norm": 0.5154740574278834,
+      "learning_rate": 9.534064896722157e-06,
+      "loss": 0.3022,
+      "step": 3386
+    },
+    {
+      "epoch": 1.6183621050116481,
+      "grad_norm": 0.4571584542414456,
+      "learning_rate": 9.528803971356275e-06,
+      "loss": 0.3106,
+      "step": 3387
+    },
+    {
+      "epoch": 1.6188399737172212,
+      "grad_norm": 0.5686619283974957,
+      "learning_rate": 9.523543176691861e-06,
+      "loss": 0.3067,
+      "step": 3388
+    },
+    {
+      "epoch": 1.6193178424227943,
+      "grad_norm": 0.5191198661164662,
+      "learning_rate": 9.518282514188163e-06,
+      "loss": 0.3071,
+      "step": 3389
+    },
+    {
+      "epoch": 1.6197957111283676,
+      "grad_norm": 0.46818165462038835,
+      "learning_rate": 9.513021985304399e-06,
+      "loss": 0.3116,
+      "step": 3390
+    },
+    {
+      "epoch": 1.6202735798339405,
+      "grad_norm": 0.4948050354348871,
+      "learning_rate": 9.507761591499747e-06,
+      "loss": 0.3252,
+      "step": 3391
+    },
+    {
+      "epoch": 1.6207514485395138,
+      "grad_norm": 0.5123067692293698,
+      "learning_rate": 9.502501334233343e-06,
+      "loss": 0.312,
+      "step": 3392
+    },
+    {
+      "epoch": 1.621229317245087,
+      "grad_norm": 0.4591469898065541,
+      "learning_rate": 9.497241214964297e-06,
+      "loss": 0.3208,
+      "step": 3393
+    },
+    {
+      "epoch": 1.62170718595066,
+      "grad_norm": 0.4756275676531446,
+      "learning_rate": 9.491981235151669e-06,
+      "loss": 0.3121,
+      "step": 3394
+    },
+    {
+      "epoch": 1.6221850546562333,
+      "grad_norm": 0.4827825826648765,
+      "learning_rate": 9.486721396254484e-06,
+      "loss": 0.3049,
+      "step": 3395
+    },
+    {
+      "epoch": 1.6226629233618064,
+      "grad_norm": 0.4915126644495333,
+      "learning_rate": 9.48146169973173e-06,
+      "loss": 0.3192,
+      "step": 3396
+    },
+    {
+      "epoch": 1.6231407920673795,
+      "grad_norm": 0.4610798399908291,
+      "learning_rate": 9.476202147042354e-06,
+      "loss": 0.3269,
+      "step": 3397
+    },
+    {
+      "epoch": 1.6236186607729526,
+      "grad_norm": 0.5599323558551276,
+      "learning_rate": 9.47094273964526e-06,
+      "loss": 0.335,
+      "step": 3398
+    },
+    {
+      "epoch": 1.6240965294785257,
+      "grad_norm": 0.5255208607648829,
+      "learning_rate": 9.465683478999319e-06,
+      "loss": 0.3291,
+      "step": 3399
+    },
+    {
+      "epoch": 1.624574398184099,
+      "grad_norm": 0.4924211708443086,
+      "learning_rate": 9.460424366563355e-06,
+      "loss": 0.3217,
+      "step": 3400
+    },
+    {
+      "epoch": 1.625052266889672,
+      "grad_norm": 0.476959287524869,
+      "learning_rate": 9.455165403796157e-06,
+      "loss": 0.3091,
+      "step": 3401
+    },
+    {
+      "epoch": 1.6255301355952452,
+      "grad_norm": 0.4672843679938778,
+      "learning_rate": 9.449906592156463e-06,
+      "loss": 0.3176,
+      "step": 3402
+    },
+    {
+      "epoch": 1.6260080043008185,
+      "grad_norm": 0.5009267665364574,
+      "learning_rate": 9.444647933102977e-06,
+      "loss": 0.3306,
+      "step": 3403
+    },
+    {
+      "epoch": 1.6264858730063914,
+      "grad_norm": 0.49020359549356995,
+      "learning_rate": 9.43938942809436e-06,
+      "loss": 0.3016,
+      "step": 3404
+    },
+    {
+      "epoch": 1.6269637417119647,
+      "grad_norm": 0.4871954472988403,
+      "learning_rate": 9.434131078589224e-06,
+      "loss": 0.3118,
+      "step": 3405
+    },
+    {
+      "epoch": 1.6274416104175378,
+      "grad_norm": 0.47764178365700216,
+      "learning_rate": 9.428872886046145e-06,
+      "loss": 0.3277,
+      "step": 3406
+    },
+    {
+      "epoch": 1.6279194791231109,
+      "grad_norm": 0.49682895676211,
+      "learning_rate": 9.423614851923657e-06,
+      "loss": 0.3052,
+      "step": 3407
+    },
+    {
+      "epoch": 1.6283973478286842,
+      "grad_norm": 0.4664914179892642,
+      "learning_rate": 9.418356977680238e-06,
+      "loss": 0.3182,
+      "step": 3408
+    },
+    {
+      "epoch": 1.6288752165342573,
+      "grad_norm": 0.46865718754503594,
+      "learning_rate": 9.413099264774334e-06,
+      "loss": 0.303,
+      "step": 3409
+    },
+    {
+      "epoch": 1.6293530852398304,
+      "grad_norm": 0.487385856422879,
+      "learning_rate": 9.407841714664343e-06,
+      "loss": 0.3303,
+      "step": 3410
+    },
+    {
+      "epoch": 1.6298309539454034,
+      "grad_norm": 0.5465692345398058,
+      "learning_rate": 9.402584328808614e-06,
+      "loss": 0.3158,
+      "step": 3411
+    },
+    {
+      "epoch": 1.6303088226509765,
+      "grad_norm": 0.47198199422777404,
+      "learning_rate": 9.39732710866545e-06,
+      "loss": 0.3182,
+      "step": 3412
+    },
+    {
+      "epoch": 1.6307866913565499,
+      "grad_norm": 0.46116344865689524,
+      "learning_rate": 9.392070055693122e-06,
+      "loss": 0.3393,
+      "step": 3413
+    },
+    {
+      "epoch": 1.631264560062123,
+      "grad_norm": 0.4763679490362691,
+      "learning_rate": 9.38681317134983e-06,
+      "loss": 0.3161,
+      "step": 3414
+    },
+    {
+      "epoch": 1.631742428767696,
+      "grad_norm": 0.46624869036989713,
+      "learning_rate": 9.381556457093752e-06,
+      "loss": 0.3285,
+      "step": 3415
+    },
+    {
+      "epoch": 1.6322202974732694,
+      "grad_norm": 0.46060193625574064,
+      "learning_rate": 9.376299914383004e-06,
+      "loss": 0.3169,
+      "step": 3416
+    },
+    {
+      "epoch": 1.6326981661788422,
+      "grad_norm": 0.48033327662418424,
+      "learning_rate": 9.371043544675656e-06,
+      "loss": 0.3169,
+      "step": 3417
+    },
+    {
+      "epoch": 1.6331760348844155,
+      "grad_norm": 0.4802037904197471,
+      "learning_rate": 9.365787349429734e-06,
+      "loss": 0.325,
+      "step": 3418
+    },
+    {
+      "epoch": 1.6336539035899886,
+      "grad_norm": 0.5496693935684847,
+      "learning_rate": 9.360531330103218e-06,
+      "loss": 0.3025,
+      "step": 3419
+    },
+    {
+      "epoch": 1.6341317722955617,
+      "grad_norm": 0.4824651576505537,
+      "learning_rate": 9.355275488154025e-06,
+      "loss": 0.3251,
+      "step": 3420
+    },
+    {
+      "epoch": 1.634609641001135,
+      "grad_norm": 0.49309460242445563,
+      "learning_rate": 9.350019825040042e-06,
+      "loss": 0.3023,
+      "step": 3421
+    },
+    {
+      "epoch": 1.6350875097067081,
+      "grad_norm": 0.471529984373171,
+      "learning_rate": 9.344764342219096e-06,
+      "loss": 0.3212,
+      "step": 3422
+    },
+    {
+      "epoch": 1.6355653784122812,
+      "grad_norm": 0.47065857771956116,
+      "learning_rate": 9.33950904114896e-06,
+      "loss": 0.3107,
+      "step": 3423
+    },
+    {
+      "epoch": 1.6360432471178543,
+      "grad_norm": 0.47686302451072865,
+      "learning_rate": 9.334253923287364e-06,
+      "loss": 0.3156,
+      "step": 3424
+    },
+    {
+      "epoch": 1.6365211158234274,
+      "grad_norm": 0.5127539654610185,
+      "learning_rate": 9.32899899009199e-06,
+      "loss": 0.3162,
+      "step": 3425
+    },
+    {
+      "epoch": 1.6369989845290007,
+      "grad_norm": 0.4747555573267503,
+      "learning_rate": 9.323744243020458e-06,
+      "loss": 0.3094,
+      "step": 3426
+    },
+    {
+      "epoch": 1.6374768532345738,
+      "grad_norm": 2.954360238976603,
+      "learning_rate": 9.318489683530342e-06,
+      "loss": 0.2972,
+      "step": 3427
+    },
+    {
+      "epoch": 1.637954721940147,
+      "grad_norm": 0.5656096550756227,
+      "learning_rate": 9.313235313079171e-06,
+      "loss": 0.321,
+      "step": 3428
+    },
+    {
+      "epoch": 1.6384325906457202,
+      "grad_norm": 0.5225575379322978,
+      "learning_rate": 9.307981133124407e-06,
+      "loss": 0.3252,
+      "step": 3429
+    },
+    {
+      "epoch": 1.638910459351293,
+      "grad_norm": 0.5091222365768304,
+      "learning_rate": 9.30272714512347e-06,
+      "loss": 0.3249,
+      "step": 3430
+    },
+    {
+      "epoch": 1.6393883280568664,
+      "grad_norm": 0.4992499483033707,
+      "learning_rate": 9.297473350533723e-06,
+      "loss": 0.3172,
+      "step": 3431
+    },
+    {
+      "epoch": 1.6398661967624395,
+      "grad_norm": 0.4831370776686128,
+      "learning_rate": 9.292219750812475e-06,
+      "loss": 0.3244,
+      "step": 3432
+    },
+    {
+      "epoch": 1.6403440654680126,
+      "grad_norm": 0.4476338694579795,
+      "learning_rate": 9.286966347416982e-06,
+      "loss": 0.3148,
+      "step": 3433
+    },
+    {
+      "epoch": 1.640821934173586,
+      "grad_norm": 0.43830875865078245,
+      "learning_rate": 9.281713141804449e-06,
+      "loss": 0.3182,
+      "step": 3434
+    },
+    {
+      "epoch": 1.641299802879159,
+      "grad_norm": 0.4809168285139566,
+      "learning_rate": 9.276460135432019e-06,
+      "loss": 0.3241,
+      "step": 3435
+    },
+    {
+      "epoch": 1.641777671584732,
+      "grad_norm": 0.4754134644489622,
+      "learning_rate": 9.271207329756787e-06,
+      "loss": 0.2952,
+      "step": 3436
+    },
+    {
+      "epoch": 1.6422555402903054,
+      "grad_norm": 0.48655404908499644,
+      "learning_rate": 9.265954726235783e-06,
+      "loss": 0.312,
+      "step": 3437
+    },
+    {
+      "epoch": 1.6427334089958783,
+      "grad_norm": 0.4683670052071872,
+      "learning_rate": 9.26070232632599e-06,
+      "loss": 0.324,
+      "step": 3438
+    },
+    {
+      "epoch": 1.6432112777014516,
+      "grad_norm": 0.4890631847154369,
+      "learning_rate": 9.255450131484334e-06,
+      "loss": 0.3092,
+      "step": 3439
+    },
+    {
+      "epoch": 1.6436891464070247,
+      "grad_norm": 0.5288231867382709,
+      "learning_rate": 9.250198143167675e-06,
+      "loss": 0.3033,
+      "step": 3440
+    },
+    {
+      "epoch": 1.6441670151125978,
+      "grad_norm": 0.48338257041337923,
+      "learning_rate": 9.244946362832825e-06,
+      "loss": 0.3129,
+      "step": 3441
+    },
+    {
+      "epoch": 1.644644883818171,
+      "grad_norm": 0.4614188770486306,
+      "learning_rate": 9.239694791936536e-06,
+      "loss": 0.3121,
+      "step": 3442
+    },
+    {
+      "epoch": 1.645122752523744,
+      "grad_norm": 0.48639433398453336,
+      "learning_rate": 9.234443431935498e-06,
+      "loss": 0.3359,
+      "step": 3443
+    },
+    {
+      "epoch": 1.6456006212293173,
+      "grad_norm": 0.454810660207841,
+      "learning_rate": 9.22919228428635e-06,
+      "loss": 0.3307,
+      "step": 3444
+    },
+    {
+      "epoch": 1.6460784899348904,
+      "grad_norm": 0.4813257235091246,
+      "learning_rate": 9.223941350445666e-06,
+      "loss": 0.3322,
+      "step": 3445
+    },
+    {
+      "epoch": 1.6465563586404635,
+      "grad_norm": 0.466364711126299,
+      "learning_rate": 9.218690631869961e-06,
+      "loss": 0.3192,
+      "step": 3446
+    },
+    {
+      "epoch": 1.6470342273460368,
+      "grad_norm": 0.5066591698413345,
+      "learning_rate": 9.213440130015692e-06,
+      "loss": 0.3273,
+      "step": 3447
+    },
+    {
+      "epoch": 1.6475120960516099,
+      "grad_norm": 0.49776839266177914,
+      "learning_rate": 9.208189846339259e-06,
+      "loss": 0.3338,
+      "step": 3448
+    },
+    {
+      "epoch": 1.647989964757183,
+      "grad_norm": 0.480237463551773,
+      "learning_rate": 9.202939782296992e-06,
+      "loss": 0.3249,
+      "step": 3449
+    },
+    {
+      "epoch": 1.6484678334627563,
+      "grad_norm": 0.46449483591155677,
+      "learning_rate": 9.19768993934517e-06,
+      "loss": 0.2952,
+      "step": 3450
+    },
+    {
+      "epoch": 1.6489457021683291,
+      "grad_norm": 0.46253453287395435,
+      "learning_rate": 9.192440318940009e-06,
+      "loss": 0.3152,
+      "step": 3451
+    },
+    {
+      "epoch": 1.6494235708739025,
+      "grad_norm": 0.8458069073772643,
+      "learning_rate": 9.187190922537654e-06,
+      "loss": 0.3168,
+      "step": 3452
+    },
+    {
+      "epoch": 1.6499014395794755,
+      "grad_norm": 0.6969051435404784,
+      "learning_rate": 9.181941751594203e-06,
+      "loss": 0.3276,
+      "step": 3453
+    },
+    {
+      "epoch": 1.6503793082850486,
+      "grad_norm": 0.461496229769572,
+      "learning_rate": 9.176692807565679e-06,
+      "loss": 0.326,
+      "step": 3454
+    },
+    {
+      "epoch": 1.650857176990622,
+      "grad_norm": 0.483293334225472,
+      "learning_rate": 9.171444091908046e-06,
+      "loss": 0.3164,
+      "step": 3455
+    },
+    {
+      "epoch": 1.6513350456961948,
+      "grad_norm": 0.4776240975371244,
+      "learning_rate": 9.166195606077205e-06,
+      "loss": 0.3153,
+      "step": 3456
+    },
+    {
+      "epoch": 1.6518129144017681,
+      "grad_norm": 0.47229526671977295,
+      "learning_rate": 9.160947351529001e-06,
+      "loss": 0.3199,
+      "step": 3457
+    },
+    {
+      "epoch": 1.6522907831073412,
+      "grad_norm": 0.5146895918341302,
+      "learning_rate": 9.155699329719196e-06,
+      "loss": 0.3286,
+      "step": 3458
+    },
+    {
+      "epoch": 1.6527686518129143,
+      "grad_norm": 0.4384463475806529,
+      "learning_rate": 9.150451542103505e-06,
+      "loss": 0.3301,
+      "step": 3459
+    },
+    {
+      "epoch": 1.6532465205184876,
+      "grad_norm": 0.4703968423103728,
+      "learning_rate": 9.145203990137571e-06,
+      "loss": 0.3096,
+      "step": 3460
+    },
+    {
+      "epoch": 1.6537243892240607,
+      "grad_norm": 0.4793369058069929,
+      "learning_rate": 9.13995667527697e-06,
+      "loss": 0.3284,
+      "step": 3461
+    },
+    {
+      "epoch": 1.6542022579296338,
+      "grad_norm": 0.45236867666956726,
+      "learning_rate": 9.134709598977218e-06,
+      "loss": 0.3234,
+      "step": 3462
+    },
+    {
+      "epoch": 1.6546801266352071,
+      "grad_norm": 0.47207424023467076,
+      "learning_rate": 9.129462762693759e-06,
+      "loss": 0.3096,
+      "step": 3463
+    },
+    {
+      "epoch": 1.65515799534078,
+      "grad_norm": 0.4750207016386637,
+      "learning_rate": 9.124216167881974e-06,
+      "loss": 0.3025,
+      "step": 3464
+    },
+    {
+      "epoch": 1.6556358640463533,
+      "grad_norm": 0.45011316729860085,
+      "learning_rate": 9.118969815997174e-06,
+      "loss": 0.3117,
+      "step": 3465
+    },
+    {
+      "epoch": 1.6561137327519264,
+      "grad_norm": 0.45982543099231327,
+      "learning_rate": 9.11372370849461e-06,
+      "loss": 0.338,
+      "step": 3466
+    },
+    {
+      "epoch": 1.6565916014574995,
+      "grad_norm": 0.47099246855087673,
+      "learning_rate": 9.108477846829447e-06,
+      "loss": 0.313,
+      "step": 3467
+    },
+    {
+      "epoch": 1.6570694701630728,
+      "grad_norm": 0.4872573696860828,
+      "learning_rate": 9.103232232456812e-06,
+      "loss": 0.3257,
+      "step": 3468
+    },
+    {
+      "epoch": 1.6575473388686457,
+      "grad_norm": 0.44856712782419816,
+      "learning_rate": 9.097986866831733e-06,
+      "loss": 0.321,
+      "step": 3469
+    },
+    {
+      "epoch": 1.658025207574219,
+      "grad_norm": 0.4420962792302268,
+      "learning_rate": 9.092741751409186e-06,
+      "loss": 0.3273,
+      "step": 3470
+    },
+    {
+      "epoch": 1.658503076279792,
+      "grad_norm": 0.4687035556845388,
+      "learning_rate": 9.087496887644075e-06,
+      "loss": 0.3249,
+      "step": 3471
+    },
+    {
+      "epoch": 1.6589809449853652,
+      "grad_norm": 0.49495908912270503,
+      "learning_rate": 9.08225227699123e-06,
+      "loss": 0.3054,
+      "step": 3472
+    },
+    {
+      "epoch": 1.6594588136909385,
+      "grad_norm": 0.48042565301421647,
+      "learning_rate": 9.077007920905413e-06,
+      "loss": 0.3207,
+      "step": 3473
+    },
+    {
+      "epoch": 1.6599366823965116,
+      "grad_norm": 0.46644981695972604,
+      "learning_rate": 9.071763820841322e-06,
+      "loss": 0.3226,
+      "step": 3474
+    },
+    {
+      "epoch": 1.6604145511020847,
+      "grad_norm": 0.4810549358732364,
+      "learning_rate": 9.06651997825357e-06,
+      "loss": 0.3275,
+      "step": 3475
+    },
+    {
+      "epoch": 1.660892419807658,
+      "grad_norm": 0.4889810661635855,
+      "learning_rate": 9.061276394596712e-06,
+      "loss": 0.3238,
+      "step": 3476
+    },
+    {
+      "epoch": 1.6613702885132309,
+      "grad_norm": 0.4429594397309242,
+      "learning_rate": 9.05603307132522e-06,
+      "loss": 0.3114,
+      "step": 3477
+    },
+    {
+      "epoch": 1.6618481572188042,
+      "grad_norm": 0.46379425469030755,
+      "learning_rate": 9.050790009893505e-06,
+      "loss": 0.313,
+      "step": 3478
+    },
+    {
+      "epoch": 1.6623260259243773,
+      "grad_norm": 0.5433136463867893,
+      "learning_rate": 9.0455472117559e-06,
+      "loss": 0.3061,
+      "step": 3479
+    },
+    {
+      "epoch": 1.6628038946299504,
+      "grad_norm": 0.4514939829237272,
+      "learning_rate": 9.040304678366658e-06,
+      "loss": 0.3147,
+      "step": 3480
+    },
+    {
+      "epoch": 1.6632817633355237,
+      "grad_norm": 0.471795348976342,
+      "learning_rate": 9.03506241117997e-06,
+      "loss": 0.3277,
+      "step": 3481
+    },
+    {
+      "epoch": 1.6637596320410966,
+      "grad_norm": 0.480378485496424,
+      "learning_rate": 9.02982041164995e-06,
+      "loss": 0.3062,
+      "step": 3482
+    },
+    {
+      "epoch": 1.6642375007466699,
+      "grad_norm": 0.44413418724919074,
+      "learning_rate": 9.024578681230632e-06,
+      "loss": 0.3231,
+      "step": 3483
+    },
+    {
+      "epoch": 1.664715369452243,
+      "grad_norm": 0.495040097592052,
+      "learning_rate": 9.01933722137598e-06,
+      "loss": 0.3113,
+      "step": 3484
+    },
+    {
+      "epoch": 1.665193238157816,
+      "grad_norm": 0.5093870178221196,
+      "learning_rate": 9.014096033539889e-06,
+      "loss": 0.3206,
+      "step": 3485
+    },
+    {
+      "epoch": 1.6656711068633894,
+      "grad_norm": 0.5133513477654141,
+      "learning_rate": 9.008855119176165e-06,
+      "loss": 0.3057,
+      "step": 3486
+    },
+    {
+      "epoch": 1.6661489755689625,
+      "grad_norm": 0.4699567002694254,
+      "learning_rate": 9.003614479738544e-06,
+      "loss": 0.3287,
+      "step": 3487
+    },
+    {
+      "epoch": 1.6666268442745356,
+      "grad_norm": 0.4892249229390645,
+      "learning_rate": 8.998374116680697e-06,
+      "loss": 0.3274,
+      "step": 3488
+    },
+    {
+      "epoch": 1.6671047129801089,
+      "grad_norm": 0.4946263872577261,
+      "learning_rate": 8.993134031456198e-06,
+      "loss": 0.3179,
+      "step": 3489
+    },
+    {
+      "epoch": 1.6675825816856817,
+      "grad_norm": 0.48264814521862553,
+      "learning_rate": 8.987894225518556e-06,
+      "loss": 0.3097,
+      "step": 3490
+    },
+    {
+      "epoch": 1.668060450391255,
+      "grad_norm": 0.4605926310411199,
+      "learning_rate": 8.982654700321207e-06,
+      "loss": 0.3218,
+      "step": 3491
+    },
+    {
+      "epoch": 1.6685383190968281,
+      "grad_norm": 0.4807584069153299,
+      "learning_rate": 8.977415457317495e-06,
+      "loss": 0.3295,
+      "step": 3492
+    },
+    {
+      "epoch": 1.6690161878024012,
+      "grad_norm": 0.5126134343679848,
+      "learning_rate": 8.972176497960698e-06,
+      "loss": 0.3023,
+      "step": 3493
+    },
+    {
+      "epoch": 1.6694940565079746,
+      "grad_norm": 0.4866870303998035,
+      "learning_rate": 8.966937823704013e-06,
+      "loss": 0.3246,
+      "step": 3494
+    },
+    {
+      "epoch": 1.6699719252135474,
+      "grad_norm": 0.4701808785486881,
+      "learning_rate": 8.961699436000548e-06,
+      "loss": 0.3138,
+      "step": 3495
+    },
+    {
+      "epoch": 1.6704497939191207,
+      "grad_norm": 0.48437050470284443,
+      "learning_rate": 8.956461336303345e-06,
+      "loss": 0.3309,
+      "step": 3496
+    },
+    {
+      "epoch": 1.6709276626246938,
+      "grad_norm": 0.4915042033611595,
+      "learning_rate": 8.951223526065363e-06,
+      "loss": 0.3056,
+      "step": 3497
+    },
+    {
+      "epoch": 1.671405531330267,
+      "grad_norm": 0.48853602825271614,
+      "learning_rate": 8.945986006739472e-06,
+      "loss": 0.3398,
+      "step": 3498
+    },
+    {
+      "epoch": 1.6718834000358402,
+      "grad_norm": 0.47166819435850216,
+      "learning_rate": 8.94074877977847e-06,
+      "loss": 0.32,
+      "step": 3499
+    },
+    {
+      "epoch": 1.6723612687414133,
+      "grad_norm": 0.4848661547844431,
+      "learning_rate": 8.93551184663507e-06,
+      "loss": 0.3163,
+      "step": 3500
+    },
+    {
+      "epoch": 1.6728391374469864,
+      "grad_norm": 0.5013003475484796,
+      "learning_rate": 8.930275208761908e-06,
+      "loss": 0.3187,
+      "step": 3501
+    },
+    {
+      "epoch": 1.6733170061525597,
+      "grad_norm": 0.5469587538744264,
+      "learning_rate": 8.925038867611536e-06,
+      "loss": 0.3181,
+      "step": 3502
+    },
+    {
+      "epoch": 1.6737948748581326,
+      "grad_norm": 0.4669973541608491,
+      "learning_rate": 8.919802824636418e-06,
+      "loss": 0.3089,
+      "step": 3503
+    },
+    {
+      "epoch": 1.674272743563706,
+      "grad_norm": 0.6631093786446882,
+      "learning_rate": 8.914567081288943e-06,
+      "loss": 0.3196,
+      "step": 3504
+    },
+    {
+      "epoch": 1.674750612269279,
+      "grad_norm": 0.4870905723140869,
+      "learning_rate": 8.909331639021414e-06,
+      "loss": 0.3002,
+      "step": 3505
+    },
+    {
+      "epoch": 1.675228480974852,
+      "grad_norm": 0.5011009424411035,
+      "learning_rate": 8.904096499286047e-06,
+      "loss": 0.3197,
+      "step": 3506
+    },
+    {
+      "epoch": 1.6757063496804254,
+      "grad_norm": 0.49155633759059203,
+      "learning_rate": 8.89886166353498e-06,
+      "loss": 0.3081,
+      "step": 3507
+    },
+    {
+      "epoch": 1.6761842183859983,
+      "grad_norm": 0.49765551877135294,
+      "learning_rate": 8.893627133220266e-06,
+      "loss": 0.3117,
+      "step": 3508
+    },
+    {
+      "epoch": 1.6766620870915716,
+      "grad_norm": 0.4911618547326789,
+      "learning_rate": 8.888392909793866e-06,
+      "loss": 0.3098,
+      "step": 3509
+    },
+    {
+      "epoch": 1.6771399557971447,
+      "grad_norm": 0.5063806765110067,
+      "learning_rate": 8.883158994707666e-06,
+      "loss": 0.3335,
+      "step": 3510
+    },
+    {
+      "epoch": 1.6776178245027178,
+      "grad_norm": 0.5734707096493096,
+      "learning_rate": 8.87792538941346e-06,
+      "loss": 0.3082,
+      "step": 3511
+    },
+    {
+      "epoch": 1.678095693208291,
+      "grad_norm": 0.5067693296298389,
+      "learning_rate": 8.872692095362957e-06,
+      "loss": 0.3061,
+      "step": 3512
+    },
+    {
+      "epoch": 1.6785735619138642,
+      "grad_norm": 0.5479709311181276,
+      "learning_rate": 8.867459114007784e-06,
+      "loss": 0.3164,
+      "step": 3513
+    },
+    {
+      "epoch": 1.6790514306194373,
+      "grad_norm": 0.5287134929809657,
+      "learning_rate": 8.862226446799474e-06,
+      "loss": 0.3307,
+      "step": 3514
+    },
+    {
+      "epoch": 1.6795292993250106,
+      "grad_norm": 0.4649713213931607,
+      "learning_rate": 8.856994095189477e-06,
+      "loss": 0.3155,
+      "step": 3515
+    },
+    {
+      "epoch": 1.6800071680305835,
+      "grad_norm": 0.49222649083051145,
+      "learning_rate": 8.851762060629155e-06,
+      "loss": 0.3309,
+      "step": 3516
+    },
+    {
+      "epoch": 1.6804850367361568,
+      "grad_norm": 0.5143659759951493,
+      "learning_rate": 8.846530344569785e-06,
+      "loss": 0.3245,
+      "step": 3517
+    },
+    {
+      "epoch": 1.6809629054417299,
+      "grad_norm": 0.49180603565206343,
+      "learning_rate": 8.841298948462548e-06,
+      "loss": 0.319,
+      "step": 3518
+    },
+    {
+      "epoch": 1.681440774147303,
+      "grad_norm": 0.4904305397252043,
+      "learning_rate": 8.836067873758544e-06,
+      "loss": 0.3076,
+      "step": 3519
+    },
+    {
+      "epoch": 1.6819186428528763,
+      "grad_norm": 0.5069444940408103,
+      "learning_rate": 8.830837121908783e-06,
+      "loss": 0.3142,
+      "step": 3520
+    },
+    {
+      "epoch": 1.6823965115584492,
+      "grad_norm": 0.47329267540124315,
+      "learning_rate": 8.825606694364178e-06,
+      "loss": 0.3116,
+      "step": 3521
+    },
+    {
+      "epoch": 1.6828743802640225,
+      "grad_norm": 0.5502456613469378,
+      "learning_rate": 8.820376592575562e-06,
+      "loss": 0.3266,
+      "step": 3522
+    },
+    {
+      "epoch": 1.6833522489695956,
+      "grad_norm": 0.5242294700558009,
+      "learning_rate": 8.81514681799367e-06,
+      "loss": 0.3171,
+      "step": 3523
+    },
+    {
+      "epoch": 1.6838301176751687,
+      "grad_norm": 0.4587826429514902,
+      "learning_rate": 8.80991737206915e-06,
+      "loss": 0.3128,
+      "step": 3524
+    },
+    {
+      "epoch": 1.684307986380742,
+      "grad_norm": 0.5210196583488915,
+      "learning_rate": 8.804688256252557e-06,
+      "loss": 0.3201,
+      "step": 3525
+    },
+    {
+      "epoch": 1.684785855086315,
+      "grad_norm": 0.511732360451873,
+      "learning_rate": 8.79945947199436e-06,
+      "loss": 0.3213,
+      "step": 3526
+    },
+    {
+      "epoch": 1.6852637237918882,
+      "grad_norm": 0.6266517581717058,
+      "learning_rate": 8.794231020744926e-06,
+      "loss": 0.308,
+      "step": 3527
+    },
+    {
+      "epoch": 1.6857415924974615,
+      "grad_norm": 0.4306673990518024,
+      "learning_rate": 8.789002903954538e-06,
+      "loss": 0.3062,
+      "step": 3528
+    },
+    {
+      "epoch": 1.6862194612030343,
+      "grad_norm": 0.5201991430967379,
+      "learning_rate": 8.783775123073383e-06,
+      "loss": 0.2972,
+      "step": 3529
+    },
+    {
+      "epoch": 1.6866973299086077,
+      "grad_norm": 0.4789566466601809,
+      "learning_rate": 8.778547679551555e-06,
+      "loss": 0.3287,
+      "step": 3530
+    },
+    {
+      "epoch": 1.6871751986141807,
+      "grad_norm": 0.4503867654809924,
+      "learning_rate": 8.773320574839055e-06,
+      "loss": 0.3158,
+      "step": 3531
+    },
+    {
+      "epoch": 1.6876530673197538,
+      "grad_norm": 0.5592132809330164,
+      "learning_rate": 8.76809381038579e-06,
+      "loss": 0.2928,
+      "step": 3532
+    },
+    {
+      "epoch": 1.6881309360253272,
+      "grad_norm": 0.5248394969953822,
+      "learning_rate": 8.762867387641569e-06,
+      "loss": 0.3192,
+      "step": 3533
+    },
+    {
+      "epoch": 1.6886088047309,
+      "grad_norm": 0.473584910793579,
+      "learning_rate": 8.757641308056111e-06,
+      "loss": 0.315,
+      "step": 3534
+    },
+    {
+      "epoch": 1.6890866734364733,
+      "grad_norm": 0.5181876158952626,
+      "learning_rate": 8.75241557307904e-06,
+      "loss": 0.3034,
+      "step": 3535
+    },
+    {
+      "epoch": 1.6895645421420464,
+      "grad_norm": 0.4871622666683337,
+      "learning_rate": 8.747190184159883e-06,
+      "loss": 0.3042,
+      "step": 3536
+    },
+    {
+      "epoch": 1.6900424108476195,
+      "grad_norm": 0.4857165458002748,
+      "learning_rate": 8.741965142748072e-06,
+      "loss": 0.3125,
+      "step": 3537
+    },
+    {
+      "epoch": 1.6905202795531928,
+      "grad_norm": 0.47357991306741637,
+      "learning_rate": 8.736740450292937e-06,
+      "loss": 0.3216,
+      "step": 3538
+    },
+    {
+      "epoch": 1.690998148258766,
+      "grad_norm": 0.45787328331002924,
+      "learning_rate": 8.731516108243717e-06,
+      "loss": 0.3252,
+      "step": 3539
+    },
+    {
+      "epoch": 1.691476016964339,
+      "grad_norm": 0.479337722950027,
+      "learning_rate": 8.726292118049555e-06,
+      "loss": 0.3217,
+      "step": 3540
+    },
+    {
+      "epoch": 1.6919538856699123,
+      "grad_norm": 0.4770400698452951,
+      "learning_rate": 8.721068481159491e-06,
+      "loss": 0.3332,
+      "step": 3541
+    },
+    {
+      "epoch": 1.6924317543754852,
+      "grad_norm": 0.5242710945344221,
+      "learning_rate": 8.715845199022468e-06,
+      "loss": 0.3065,
+      "step": 3542
+    },
+    {
+      "epoch": 1.6929096230810585,
+      "grad_norm": 0.47188838291316537,
+      "learning_rate": 8.710622273087338e-06,
+      "loss": 0.3181,
+      "step": 3543
+    },
+    {
+      "epoch": 1.6933874917866316,
+      "grad_norm": 0.4916479471672269,
+      "learning_rate": 8.705399704802844e-06,
+      "loss": 0.3124,
+      "step": 3544
+    },
+    {
+      "epoch": 1.6938653604922047,
+      "grad_norm": 0.4864876674585418,
+      "learning_rate": 8.700177495617635e-06,
+      "loss": 0.303,
+      "step": 3545
+    },
+    {
+      "epoch": 1.694343229197778,
+      "grad_norm": 0.48143706434962735,
+      "learning_rate": 8.694955646980261e-06,
+      "loss": 0.3244,
+      "step": 3546
+    },
+    {
+      "epoch": 1.694821097903351,
+      "grad_norm": 0.5024128714258727,
+      "learning_rate": 8.68973416033917e-06,
+      "loss": 0.3214,
+      "step": 3547
+    },
+    {
+      "epoch": 1.6952989666089242,
+      "grad_norm": 0.4701290161733491,
+      "learning_rate": 8.684513037142708e-06,
+      "loss": 0.319,
+      "step": 3548
+    },
+    {
+      "epoch": 1.6957768353144973,
+      "grad_norm": 0.5196310295903178,
+      "learning_rate": 8.67929227883913e-06,
+      "loss": 0.3304,
+      "step": 3549
+    },
+    {
+      "epoch": 1.6962547040200704,
+      "grad_norm": 0.4470857047416331,
+      "learning_rate": 8.674071886876572e-06,
+      "loss": 0.3116,
+      "step": 3550
+    },
+    {
+      "epoch": 1.6967325727256437,
+      "grad_norm": 0.49139633093819357,
+      "learning_rate": 8.668851862703084e-06,
+      "loss": 0.3147,
+      "step": 3551
+    },
+    {
+      "epoch": 1.6972104414312168,
+      "grad_norm": 0.4533593202968756,
+      "learning_rate": 8.66363220776661e-06,
+      "loss": 0.3117,
+      "step": 3552
+    },
+    {
+      "epoch": 1.69768831013679,
+      "grad_norm": 0.5016957872077247,
+      "learning_rate": 8.658412923514987e-06,
+      "loss": 0.3058,
+      "step": 3553
+    },
+    {
+      "epoch": 1.6981661788423632,
+      "grad_norm": 0.49422695943685363,
+      "learning_rate": 8.653194011395955e-06,
+      "loss": 0.3166,
+      "step": 3554
+    },
+    {
+      "epoch": 1.698644047547936,
+      "grad_norm": 0.5041371009037742,
+      "learning_rate": 8.647975472857148e-06,
+      "loss": 0.3058,
+      "step": 3555
+    },
+    {
+      "epoch": 1.6991219162535094,
+      "grad_norm": 0.46401191058629204,
+      "learning_rate": 8.642757309346092e-06,
+      "loss": 0.2981,
+      "step": 3556
+    },
+    {
+      "epoch": 1.6995997849590825,
+      "grad_norm": 0.47462647906742783,
+      "learning_rate": 8.637539522310219e-06,
+      "loss": 0.3158,
+      "step": 3557
+    },
+    {
+      "epoch": 1.7000776536646556,
+      "grad_norm": 0.48560418398414035,
+      "learning_rate": 8.63232211319685e-06,
+      "loss": 0.3101,
+      "step": 3558
+    },
+    {
+      "epoch": 1.700555522370229,
+      "grad_norm": 0.4924892083480203,
+      "learning_rate": 8.6271050834532e-06,
+      "loss": 0.3175,
+      "step": 3559
+    },
+    {
+      "epoch": 1.7010333910758018,
+      "grad_norm": 0.4739306925770209,
+      "learning_rate": 8.621888434526382e-06,
+      "loss": 0.3055,
+      "step": 3560
+    },
+    {
+      "epoch": 1.701511259781375,
+      "grad_norm": 0.5668938530252869,
+      "learning_rate": 8.616672167863406e-06,
+      "loss": 0.3001,
+      "step": 3561
+    },
+    {
+      "epoch": 1.7019891284869482,
+      "grad_norm": 0.5258055525735766,
+      "learning_rate": 8.611456284911167e-06,
+      "loss": 0.3317,
+      "step": 3562
+    },
+    {
+      "epoch": 1.7024669971925213,
+      "grad_norm": 0.4731593091810082,
+      "learning_rate": 8.606240787116459e-06,
+      "loss": 0.305,
+      "step": 3563
+    },
+    {
+      "epoch": 1.7029448658980946,
+      "grad_norm": 0.47298302795830727,
+      "learning_rate": 8.601025675925976e-06,
+      "loss": 0.3518,
+      "step": 3564
+    },
+    {
+      "epoch": 1.7034227346036677,
+      "grad_norm": 0.5044823125447081,
+      "learning_rate": 8.595810952786289e-06,
+      "loss": 0.316,
+      "step": 3565
+    },
+    {
+      "epoch": 1.7039006033092408,
+      "grad_norm": 0.48834528388610515,
+      "learning_rate": 8.590596619143874e-06,
+      "loss": 0.3185,
+      "step": 3566
+    },
+    {
+      "epoch": 1.704378472014814,
+      "grad_norm": 0.47503311635279094,
+      "learning_rate": 8.585382676445099e-06,
+      "loss": 0.3122,
+      "step": 3567
+    },
+    {
+      "epoch": 1.704856340720387,
+      "grad_norm": 0.4779679121742256,
+      "learning_rate": 8.580169126136211e-06,
+      "loss": 0.3024,
+      "step": 3568
+    },
+    {
+      "epoch": 1.7053342094259603,
+      "grad_norm": 0.4881996436437521,
+      "learning_rate": 8.574955969663364e-06,
+      "loss": 0.308,
+      "step": 3569
+    },
+    {
+      "epoch": 1.7058120781315333,
+      "grad_norm": 0.49243529307480594,
+      "learning_rate": 8.569743208472594e-06,
+      "loss": 0.3211,
+      "step": 3570
+    },
+    {
+      "epoch": 1.7062899468371064,
+      "grad_norm": 0.47769625174462077,
+      "learning_rate": 8.56453084400983e-06,
+      "loss": 0.3022,
+      "step": 3571
+    },
+    {
+      "epoch": 1.7067678155426798,
+      "grad_norm": 0.4687274182843692,
+      "learning_rate": 8.559318877720889e-06,
+      "loss": 0.3026,
+      "step": 3572
+    },
+    {
+      "epoch": 1.7072456842482526,
+      "grad_norm": 0.48608349498390174,
+      "learning_rate": 8.554107311051477e-06,
+      "loss": 0.3178,
+      "step": 3573
+    },
+    {
+      "epoch": 1.707723552953826,
+      "grad_norm": 0.48723365794182133,
+      "learning_rate": 8.548896145447191e-06,
+      "loss": 0.3041,
+      "step": 3574
+    },
+    {
+      "epoch": 1.708201421659399,
+      "grad_norm": 0.5215881012424272,
+      "learning_rate": 8.543685382353518e-06,
+      "loss": 0.3147,
+      "step": 3575
+    },
+    {
+      "epoch": 1.7086792903649721,
+      "grad_norm": 0.46461361060388295,
+      "learning_rate": 8.538475023215831e-06,
+      "loss": 0.3296,
+      "step": 3576
+    },
+    {
+      "epoch": 1.7091571590705454,
+      "grad_norm": 0.48227283989606085,
+      "learning_rate": 8.533265069479393e-06,
+      "loss": 0.3143,
+      "step": 3577
+    },
+    {
+      "epoch": 1.7096350277761185,
+      "grad_norm": 0.4427755116826597,
+      "learning_rate": 8.52805552258935e-06,
+      "loss": 0.2946,
+      "step": 3578
+    },
+    {
+      "epoch": 1.7101128964816916,
+      "grad_norm": 0.4557912224968937,
+      "learning_rate": 8.52284638399074e-06,
+      "loss": 0.3117,
+      "step": 3579
+    },
+    {
+      "epoch": 1.710590765187265,
+      "grad_norm": 0.48947216925097453,
+      "learning_rate": 8.51763765512849e-06,
+      "loss": 0.3135,
+      "step": 3580
+    },
+    {
+      "epoch": 1.7110686338928378,
+      "grad_norm": 0.48750224853053237,
+      "learning_rate": 8.512429337447404e-06,
+      "loss": 0.3152,
+      "step": 3581
+    },
+    {
+      "epoch": 1.7115465025984111,
+      "grad_norm": 0.5424080934153999,
+      "learning_rate": 8.507221432392177e-06,
+      "loss": 0.3185,
+      "step": 3582
+    },
+    {
+      "epoch": 1.7120243713039842,
+      "grad_norm": 0.5243992703363829,
+      "learning_rate": 8.502013941407396e-06,
+      "loss": 0.2851,
+      "step": 3583
+    },
+    {
+      "epoch": 1.7125022400095573,
+      "grad_norm": 0.47093855520770667,
+      "learning_rate": 8.496806865937523e-06,
+      "loss": 0.3141,
+      "step": 3584
+    },
+    {
+      "epoch": 1.7129801087151306,
+      "grad_norm": 0.5163527516479693,
+      "learning_rate": 8.491600207426907e-06,
+      "loss": 0.3048,
+      "step": 3585
+    },
+    {
+      "epoch": 1.7134579774207037,
+      "grad_norm": 0.46552143826305653,
+      "learning_rate": 8.48639396731979e-06,
+      "loss": 0.3002,
+      "step": 3586
+    },
+    {
+      "epoch": 1.7139358461262768,
+      "grad_norm": 0.47841277412531025,
+      "learning_rate": 8.481188147060283e-06,
+      "loss": 0.3095,
+      "step": 3587
+    },
+    {
+      "epoch": 1.71441371483185,
+      "grad_norm": 0.5911896029327872,
+      "learning_rate": 8.475982748092395e-06,
+      "loss": 0.3257,
+      "step": 3588
+    },
+    {
+      "epoch": 1.714891583537423,
+      "grad_norm": 0.4716246559506084,
+      "learning_rate": 8.470777771860009e-06,
+      "loss": 0.3108,
+      "step": 3589
+    },
+    {
+      "epoch": 1.7153694522429963,
+      "grad_norm": 0.4544478697048045,
+      "learning_rate": 8.465573219806893e-06,
+      "loss": 0.3144,
+      "step": 3590
+    },
+    {
+      "epoch": 1.7158473209485694,
+      "grad_norm": 0.48108714707741906,
+      "learning_rate": 8.460369093376698e-06,
+      "loss": 0.2957,
+      "step": 3591
+    },
+    {
+      "epoch": 1.7163251896541425,
+      "grad_norm": 0.8635235335475422,
+      "learning_rate": 8.455165394012962e-06,
+      "loss": 0.3218,
+      "step": 3592
+    },
+    {
+      "epoch": 1.7168030583597158,
+      "grad_norm": 0.5054277355892234,
+      "learning_rate": 8.449962123159092e-06,
+      "loss": 0.3072,
+      "step": 3593
+    },
+    {
+      "epoch": 1.7172809270652887,
+      "grad_norm": 0.4829413421685776,
+      "learning_rate": 8.444759282258387e-06,
+      "loss": 0.3052,
+      "step": 3594
+    },
+    {
+      "epoch": 1.717758795770862,
+      "grad_norm": 0.4494055048503989,
+      "learning_rate": 8.439556872754025e-06,
+      "loss": 0.3111,
+      "step": 3595
+    },
+    {
+      "epoch": 1.718236664476435,
+      "grad_norm": 0.4736205047237309,
+      "learning_rate": 8.434354896089058e-06,
+      "loss": 0.3211,
+      "step": 3596
+    },
+    {
+      "epoch": 1.7187145331820082,
+      "grad_norm": 0.47081905982842837,
+      "learning_rate": 8.429153353706427e-06,
+      "loss": 0.3341,
+      "step": 3597
+    },
+    {
+      "epoch": 1.7191924018875815,
+      "grad_norm": 0.4748503377753521,
+      "learning_rate": 8.423952247048948e-06,
+      "loss": 0.3215,
+      "step": 3598
+    },
+    {
+      "epoch": 1.7196702705931546,
+      "grad_norm": 0.4944887369523441,
+      "learning_rate": 8.418751577559313e-06,
+      "loss": 0.3099,
+      "step": 3599
+    },
+    {
+      "epoch": 1.7201481392987277,
+      "grad_norm": 0.47254363092097545,
+      "learning_rate": 8.413551346680095e-06,
+      "loss": 0.3067,
+      "step": 3600
+    },
+    {
+      "epoch": 1.7206260080043008,
+      "grad_norm": 0.47647885914472093,
+      "learning_rate": 8.40835155585375e-06,
+      "loss": 0.3129,
+      "step": 3601
+    },
+    {
+      "epoch": 1.7211038767098739,
+      "grad_norm": 0.48835317263293404,
+      "learning_rate": 8.403152206522607e-06,
+      "loss": 0.3107,
+      "step": 3602
+    },
+    {
+      "epoch": 1.7215817454154472,
+      "grad_norm": 0.4596088980636602,
+      "learning_rate": 8.397953300128879e-06,
+      "loss": 0.3157,
+      "step": 3603
+    },
+    {
+      "epoch": 1.7220596141210203,
+      "grad_norm": 0.49182267500267884,
+      "learning_rate": 8.39275483811464e-06,
+      "loss": 0.3254,
+      "step": 3604
+    },
+    {
+      "epoch": 1.7225374828265934,
+      "grad_norm": 0.48328627094269644,
+      "learning_rate": 8.38755682192186e-06,
+      "loss": 0.3112,
+      "step": 3605
+    },
+    {
+      "epoch": 1.7230153515321667,
+      "grad_norm": 0.47304365223535483,
+      "learning_rate": 8.382359252992377e-06,
+      "loss": 0.3069,
+      "step": 3606
+    },
+    {
+      "epoch": 1.7234932202377395,
+      "grad_norm": 0.4813058108065857,
+      "learning_rate": 8.3771621327679e-06,
+      "loss": 0.3198,
+      "step": 3607
+    },
+    {
+      "epoch": 1.7239710889433129,
+      "grad_norm": 0.47699476225663495,
+      "learning_rate": 8.371965462690021e-06,
+      "loss": 0.32,
+      "step": 3608
+    },
+    {
+      "epoch": 1.724448957648886,
+      "grad_norm": 0.48478741000916475,
+      "learning_rate": 8.366769244200206e-06,
+      "loss": 0.3123,
+      "step": 3609
+    },
+    {
+      "epoch": 1.724926826354459,
+      "grad_norm": 0.4912926079455751,
+      "learning_rate": 8.36157347873979e-06,
+      "loss": 0.3116,
+      "step": 3610
+    },
+    {
+      "epoch": 1.7254046950600324,
+      "grad_norm": 0.5174639765845415,
+      "learning_rate": 8.356378167749993e-06,
+      "loss": 0.298,
+      "step": 3611
+    },
+    {
+      "epoch": 1.7258825637656054,
+      "grad_norm": 0.4793465538100503,
+      "learning_rate": 8.351183312671898e-06,
+      "loss": 0.31,
+      "step": 3612
+    },
+    {
+      "epoch": 1.7263604324711785,
+      "grad_norm": 0.4814025970072797,
+      "learning_rate": 8.345988914946467e-06,
+      "loss": 0.3103,
+      "step": 3613
+    },
+    {
+      "epoch": 1.7268383011767516,
+      "grad_norm": 0.46966946318183,
+      "learning_rate": 8.340794976014535e-06,
+      "loss": 0.3055,
+      "step": 3614
+    },
+    {
+      "epoch": 1.7273161698823247,
+      "grad_norm": 0.48421924623191237,
+      "learning_rate": 8.335601497316809e-06,
+      "loss": 0.3147,
+      "step": 3615
+    },
+    {
+      "epoch": 1.727794038587898,
+      "grad_norm": 0.47415453106644484,
+      "learning_rate": 8.330408480293867e-06,
+      "loss": 0.3189,
+      "step": 3616
+    },
+    {
+      "epoch": 1.7282719072934711,
+      "grad_norm": 0.5311247241943203,
+      "learning_rate": 8.325215926386164e-06,
+      "loss": 0.3081,
+      "step": 3617
+    },
+    {
+      "epoch": 1.7287497759990442,
+      "grad_norm": 0.45880552685400533,
+      "learning_rate": 8.32002383703402e-06,
+      "loss": 0.3079,
+      "step": 3618
+    },
+    {
+      "epoch": 1.7292276447046175,
+      "grad_norm": 0.7527896953526857,
+      "learning_rate": 8.314832213677627e-06,
+      "loss": 0.3276,
+      "step": 3619
+    },
+    {
+      "epoch": 1.7297055134101904,
+      "grad_norm": 0.47577023953198666,
+      "learning_rate": 8.309641057757052e-06,
+      "loss": 0.3193,
+      "step": 3620
+    },
+    {
+      "epoch": 1.7301833821157637,
+      "grad_norm": 0.4600219319465241,
+      "learning_rate": 8.304450370712234e-06,
+      "loss": 0.3247,
+      "step": 3621
+    },
+    {
+      "epoch": 1.7306612508213368,
+      "grad_norm": 0.49059880424437236,
+      "learning_rate": 8.299260153982969e-06,
+      "loss": 0.3199,
+      "step": 3622
+    },
+    {
+      "epoch": 1.73113911952691,
+      "grad_norm": 0.4882276209446351,
+      "learning_rate": 8.29407040900894e-06,
+      "loss": 0.3041,
+      "step": 3623
+    },
+    {
+      "epoch": 1.7316169882324832,
+      "grad_norm": 0.44643071797150663,
+      "learning_rate": 8.288881137229687e-06,
+      "loss": 0.3095,
+      "step": 3624
+    },
+    {
+      "epoch": 1.7320948569380563,
+      "grad_norm": 0.46790133259886285,
+      "learning_rate": 8.283692340084623e-06,
+      "loss": 0.3146,
+      "step": 3625
+    },
+    {
+      "epoch": 1.7325727256436294,
+      "grad_norm": 0.48231263905937954,
+      "learning_rate": 8.278504019013026e-06,
+      "loss": 0.3126,
+      "step": 3626
+    },
+    {
+      "epoch": 1.7330505943492027,
+      "grad_norm": 0.4774119989895905,
+      "learning_rate": 8.273316175454052e-06,
+      "loss": 0.3261,
+      "step": 3627
+    },
+    {
+      "epoch": 1.7335284630547756,
+      "grad_norm": 0.45700207528874304,
+      "learning_rate": 8.26812881084671e-06,
+      "loss": 0.3016,
+      "step": 3628
+    },
+    {
+      "epoch": 1.734006331760349,
+      "grad_norm": 0.8002514747178684,
+      "learning_rate": 8.262941926629888e-06,
+      "loss": 0.311,
+      "step": 3629
+    },
+    {
+      "epoch": 1.734484200465922,
+      "grad_norm": 0.4762774044807064,
+      "learning_rate": 8.257755524242333e-06,
+      "loss": 0.3037,
+      "step": 3630
+    },
+    {
+      "epoch": 1.734962069171495,
+      "grad_norm": 0.4694744914517388,
+      "learning_rate": 8.252569605122662e-06,
+      "loss": 0.3129,
+      "step": 3631
+    },
+    {
+      "epoch": 1.7354399378770684,
+      "grad_norm": 0.4991347812581403,
+      "learning_rate": 8.24738417070936e-06,
+      "loss": 0.3395,
+      "step": 3632
+    },
+    {
+      "epoch": 1.7359178065826413,
+      "grad_norm": 0.504295162908251,
+      "learning_rate": 8.242199222440773e-06,
+      "loss": 0.3125,
+      "step": 3633
+    },
+    {
+      "epoch": 1.7363956752882146,
+      "grad_norm": 0.6877943879571605,
+      "learning_rate": 8.237014761755116e-06,
+      "loss": 0.3096,
+      "step": 3634
+    },
+    {
+      "epoch": 1.7368735439937877,
+      "grad_norm": 0.4672144021586577,
+      "learning_rate": 8.231830790090461e-06,
+      "loss": 0.3082,
+      "step": 3635
+    },
+    {
+      "epoch": 1.7373514126993608,
+      "grad_norm": 0.44504887155749057,
+      "learning_rate": 8.226647308884756e-06,
+      "loss": 0.3228,
+      "step": 3636
+    },
+    {
+      "epoch": 1.737829281404934,
+      "grad_norm": 0.4640809118349916,
+      "learning_rate": 8.221464319575808e-06,
+      "loss": 0.3251,
+      "step": 3637
+    },
+    {
+      "epoch": 1.7383071501105072,
+      "grad_norm": 0.46673343701161357,
+      "learning_rate": 8.216281823601286e-06,
+      "loss": 0.3146,
+      "step": 3638
+    },
+    {
+      "epoch": 1.7387850188160803,
+      "grad_norm": 0.4820552462095371,
+      "learning_rate": 8.211099822398721e-06,
+      "loss": 0.3118,
+      "step": 3639
+    },
+    {
+      "epoch": 1.7392628875216536,
+      "grad_norm": 0.49498233493294336,
+      "learning_rate": 8.205918317405508e-06,
+      "loss": 0.3249,
+      "step": 3640
+    },
+    {
+      "epoch": 1.7397407562272265,
+      "grad_norm": 0.48843945985854864,
+      "learning_rate": 8.20073731005891e-06,
+      "loss": 0.3164,
+      "step": 3641
+    },
+    {
+      "epoch": 1.7402186249327998,
+      "grad_norm": 0.4814877105665985,
+      "learning_rate": 8.195556801796041e-06,
+      "loss": 0.3267,
+      "step": 3642
+    },
+    {
+      "epoch": 1.7406964936383729,
+      "grad_norm": 0.4683984421194578,
+      "learning_rate": 8.190376794053886e-06,
+      "loss": 0.3048,
+      "step": 3643
+    },
+    {
+      "epoch": 1.741174362343946,
+      "grad_norm": 0.4713542128746563,
+      "learning_rate": 8.185197288269289e-06,
+      "loss": 0.3114,
+      "step": 3644
+    },
+    {
+      "epoch": 1.7416522310495193,
+      "grad_norm": 0.5033578491109322,
+      "learning_rate": 8.18001828587895e-06,
+      "loss": 0.3356,
+      "step": 3645
+    },
+    {
+      "epoch": 1.7421300997550921,
+      "grad_norm": 0.48757140141481903,
+      "learning_rate": 8.174839788319432e-06,
+      "loss": 0.2978,
+      "step": 3646
+    },
+    {
+      "epoch": 1.7426079684606655,
+      "grad_norm": 0.5198704975496817,
+      "learning_rate": 8.169661797027167e-06,
+      "loss": 0.3215,
+      "step": 3647
+    },
+    {
+      "epoch": 1.7430858371662385,
+      "grad_norm": 0.5077562055422693,
+      "learning_rate": 8.164484313438427e-06,
+      "loss": 0.3096,
+      "step": 3648
+    },
+    {
+      "epoch": 1.7435637058718116,
+      "grad_norm": 0.476581737348631,
+      "learning_rate": 8.159307338989364e-06,
+      "loss": 0.3266,
+      "step": 3649
+    },
+    {
+      "epoch": 1.744041574577385,
+      "grad_norm": 0.474540506795891,
+      "learning_rate": 8.154130875115978e-06,
+      "loss": 0.3041,
+      "step": 3650
+    },
+    {
+      "epoch": 1.744519443282958,
+      "grad_norm": 0.4801751996496636,
+      "learning_rate": 8.148954923254122e-06,
+      "loss": 0.3127,
+      "step": 3651
+    },
+    {
+      "epoch": 1.7449973119885311,
+      "grad_norm": 0.4852373646526146,
+      "learning_rate": 8.14377948483952e-06,
+      "loss": 0.3295,
+      "step": 3652
+    },
+    {
+      "epoch": 1.7454751806941045,
+      "grad_norm": 0.47761777592380195,
+      "learning_rate": 8.138604561307748e-06,
+      "loss": 0.3239,
+      "step": 3653
+    },
+    {
+      "epoch": 1.7459530493996773,
+      "grad_norm": 0.5234441445204633,
+      "learning_rate": 8.133430154094232e-06,
+      "loss": 0.3189,
+      "step": 3654
+    },
+    {
+      "epoch": 1.7464309181052506,
+      "grad_norm": 0.49427462068157,
+      "learning_rate": 8.12825626463427e-06,
+      "loss": 0.3126,
+      "step": 3655
+    },
+    {
+      "epoch": 1.7469087868108237,
+      "grad_norm": 0.4470937189715777,
+      "learning_rate": 8.123082894363001e-06,
+      "loss": 0.3129,
+      "step": 3656
+    },
+    {
+      "epoch": 1.7473866555163968,
+      "grad_norm": 0.5522576075572359,
+      "learning_rate": 8.11791004471543e-06,
+      "loss": 0.3012,
+      "step": 3657
+    },
+    {
+      "epoch": 1.7478645242219701,
+      "grad_norm": 0.49371091975472053,
+      "learning_rate": 8.11273771712641e-06,
+      "loss": 0.3192,
+      "step": 3658
+    },
+    {
+      "epoch": 1.748342392927543,
+      "grad_norm": 0.45307522023914715,
+      "learning_rate": 8.10756591303066e-06,
+      "loss": 0.3204,
+      "step": 3659
+    },
+    {
+      "epoch": 1.7488202616331163,
+      "grad_norm": 0.46977495936479996,
+      "learning_rate": 8.102394633862743e-06,
+      "loss": 0.3327,
+      "step": 3660
+    },
+    {
+      "epoch": 1.7492981303386894,
+      "grad_norm": 0.49827574100114275,
+      "learning_rate": 8.097223881057079e-06,
+      "loss": 0.3308,
+      "step": 3661
+    },
+    {
+      "epoch": 1.7497759990442625,
+      "grad_norm": 0.46928949793764024,
+      "learning_rate": 8.09205365604795e-06,
+      "loss": 0.3085,
+      "step": 3662
+    },
+    {
+      "epoch": 1.7502538677498358,
+      "grad_norm": 0.46261108160255604,
+      "learning_rate": 8.086883960269477e-06,
+      "loss": 0.3281,
+      "step": 3663
+    },
+    {
+      "epoch": 1.750731736455409,
+      "grad_norm": 0.45709628622401643,
+      "learning_rate": 8.081714795155648e-06,
+      "loss": 0.3195,
+      "step": 3664
+    },
+    {
+      "epoch": 1.751209605160982,
+      "grad_norm": 0.5065744246500246,
+      "learning_rate": 8.0765461621403e-06,
+      "loss": 0.3302,
+      "step": 3665
+    },
+    {
+      "epoch": 1.7516874738665553,
+      "grad_norm": 0.46058472348695123,
+      "learning_rate": 8.071378062657114e-06,
+      "loss": 0.3293,
+      "step": 3666
+    },
+    {
+      "epoch": 1.7521653425721282,
+      "grad_norm": 0.5009271488564606,
+      "learning_rate": 8.066210498139632e-06,
+      "loss": 0.3049,
+      "step": 3667
+    },
+    {
+      "epoch": 1.7526432112777015,
+      "grad_norm": 0.5227592677426378,
+      "learning_rate": 8.061043470021251e-06,
+      "loss": 0.3135,
+      "step": 3668
+    },
+    {
+      "epoch": 1.7531210799832746,
+      "grad_norm": 0.5018329300564572,
+      "learning_rate": 8.055876979735203e-06,
+      "loss": 0.3071,
+      "step": 3669
+    },
+    {
+      "epoch": 1.7535989486888477,
+      "grad_norm": 0.4680917060285861,
+      "learning_rate": 8.050711028714589e-06,
+      "loss": 0.2993,
+      "step": 3670
+    },
+    {
+      "epoch": 1.754076817394421,
+      "grad_norm": 0.5167520768333849,
+      "learning_rate": 8.045545618392352e-06,
+      "loss": 0.3084,
+      "step": 3671
+    },
+    {
+      "epoch": 1.7545546860999939,
+      "grad_norm": 0.4654107230776164,
+      "learning_rate": 8.040380750201286e-06,
+      "loss": 0.3149,
+      "step": 3672
+    },
+    {
+      "epoch": 1.7550325548055672,
+      "grad_norm": 0.5273472331111888,
+      "learning_rate": 8.035216425574031e-06,
+      "loss": 0.3195,
+      "step": 3673
+    },
+    {
+      "epoch": 1.7555104235111403,
+      "grad_norm": 0.4671145108043513,
+      "learning_rate": 8.03005264594308e-06,
+      "loss": 0.3249,
+      "step": 3674
+    },
+    {
+      "epoch": 1.7559882922167134,
+      "grad_norm": 0.6794461125069341,
+      "learning_rate": 8.02488941274078e-06,
+      "loss": 0.3101,
+      "step": 3675
+    },
+    {
+      "epoch": 1.7564661609222867,
+      "grad_norm": 0.46888345908111245,
+      "learning_rate": 8.019726727399313e-06,
+      "loss": 0.3028,
+      "step": 3676
+    },
+    {
+      "epoch": 1.7569440296278598,
+      "grad_norm": 0.4913169627797939,
+      "learning_rate": 8.01456459135072e-06,
+      "loss": 0.3357,
+      "step": 3677
+    },
+    {
+      "epoch": 1.7574218983334329,
+      "grad_norm": 0.475145978227504,
+      "learning_rate": 8.00940300602689e-06,
+      "loss": 0.3149,
+      "step": 3678
+    },
+    {
+      "epoch": 1.7578997670390062,
+      "grad_norm": 0.44803279971539967,
+      "learning_rate": 8.004241972859552e-06,
+      "loss": 0.3347,
+      "step": 3679
+    },
+    {
+      "epoch": 1.758377635744579,
+      "grad_norm": 0.4787438847451073,
+      "learning_rate": 7.999081493280283e-06,
+      "loss": 0.3168,
+      "step": 3680
+    },
+    {
+      "epoch": 1.7588555044501524,
+      "grad_norm": 0.47005813353883635,
+      "learning_rate": 7.993921568720515e-06,
+      "loss": 0.3322,
+      "step": 3681
+    },
+    {
+      "epoch": 1.7593333731557255,
+      "grad_norm": 0.4661035524219349,
+      "learning_rate": 7.988762200611517e-06,
+      "loss": 0.322,
+      "step": 3682
+    },
+    {
+      "epoch": 1.7598112418612986,
+      "grad_norm": 0.45866488888688306,
+      "learning_rate": 7.983603390384405e-06,
+      "loss": 0.3227,
+      "step": 3683
+    },
+    {
+      "epoch": 1.7602891105668719,
+      "grad_norm": 0.5137297024673525,
+      "learning_rate": 7.978445139470147e-06,
+      "loss": 0.323,
+      "step": 3684
+    },
+    {
+      "epoch": 1.7607669792724447,
+      "grad_norm": 0.9472143703913082,
+      "learning_rate": 7.973287449299545e-06,
+      "loss": 0.3266,
+      "step": 3685
+    },
+    {
+      "epoch": 1.761244847978018,
+      "grad_norm": 0.46102852447781484,
+      "learning_rate": 7.968130321303254e-06,
+      "loss": 0.3103,
+      "step": 3686
+    },
+    {
+      "epoch": 1.7617227166835912,
+      "grad_norm": 0.5664763306205914,
+      "learning_rate": 7.962973756911773e-06,
+      "loss": 0.3173,
+      "step": 3687
+    },
+    {
+      "epoch": 1.7622005853891642,
+      "grad_norm": 0.48417061898824687,
+      "learning_rate": 7.957817757555438e-06,
+      "loss": 0.3044,
+      "step": 3688
+    },
+    {
+      "epoch": 1.7626784540947376,
+      "grad_norm": 0.4891312388302578,
+      "learning_rate": 7.95266232466443e-06,
+      "loss": 0.3075,
+      "step": 3689
+    },
+    {
+      "epoch": 1.7631563228003106,
+      "grad_norm": 0.4839119830817586,
+      "learning_rate": 7.947507459668784e-06,
+      "loss": 0.3227,
+      "step": 3690
+    },
+    {
+      "epoch": 1.7636341915058837,
+      "grad_norm": 0.5122896682596106,
+      "learning_rate": 7.94235316399836e-06,
+      "loss": 0.3229,
+      "step": 3691
+    },
+    {
+      "epoch": 1.764112060211457,
+      "grad_norm": 0.4895385220427247,
+      "learning_rate": 7.937199439082874e-06,
+      "loss": 0.2932,
+      "step": 3692
+    },
+    {
+      "epoch": 1.76458992891703,
+      "grad_norm": 0.47779205504554345,
+      "learning_rate": 7.932046286351877e-06,
+      "loss": 0.3098,
+      "step": 3693
+    },
+    {
+      "epoch": 1.7650677976226032,
+      "grad_norm": 0.46886659878011405,
+      "learning_rate": 7.92689370723476e-06,
+      "loss": 0.3195,
+      "step": 3694
+    },
+    {
+      "epoch": 1.7655456663281763,
+      "grad_norm": 0.47588844827515836,
+      "learning_rate": 7.921741703160758e-06,
+      "loss": 0.3178,
+      "step": 3695
+    },
+    {
+      "epoch": 1.7660235350337494,
+      "grad_norm": 0.48249496323108887,
+      "learning_rate": 7.916590275558953e-06,
+      "loss": 0.3135,
+      "step": 3696
+    },
+    {
+      "epoch": 1.7665014037393227,
+      "grad_norm": 0.5528175204761951,
+      "learning_rate": 7.91143942585825e-06,
+      "loss": 0.3119,
+      "step": 3697
+    },
+    {
+      "epoch": 1.7669792724448956,
+      "grad_norm": 0.49703375415055845,
+      "learning_rate": 7.90628915548741e-06,
+      "loss": 0.319,
+      "step": 3698
+    },
+    {
+      "epoch": 1.767457141150469,
+      "grad_norm": 0.4797540216428928,
+      "learning_rate": 7.901139465875029e-06,
+      "loss": 0.3216,
+      "step": 3699
+    },
+    {
+      "epoch": 1.767935009856042,
+      "grad_norm": 0.4685398215123699,
+      "learning_rate": 7.895990358449533e-06,
+      "loss": 0.2973,
+      "step": 3700
+    },
+    {
+      "epoch": 1.768412878561615,
+      "grad_norm": 0.47072717607559234,
+      "learning_rate": 7.890841834639198e-06,
+      "loss": 0.3204,
+      "step": 3701
+    },
+    {
+      "epoch": 1.7688907472671884,
+      "grad_norm": 0.47672448748601165,
+      "learning_rate": 7.88569389587213e-06,
+      "loss": 0.3201,
+      "step": 3702
+    },
+    {
+      "epoch": 1.7693686159727615,
+      "grad_norm": 0.4417548032639278,
+      "learning_rate": 7.880546543576283e-06,
+      "loss": 0.3278,
+      "step": 3703
+    },
+    {
+      "epoch": 1.7698464846783346,
+      "grad_norm": 0.5159968343841332,
+      "learning_rate": 7.875399779179442e-06,
+      "loss": 0.3204,
+      "step": 3704
+    },
+    {
+      "epoch": 1.770324353383908,
+      "grad_norm": 0.5117769103897961,
+      "learning_rate": 7.87025360410922e-06,
+      "loss": 0.3129,
+      "step": 3705
+    },
+    {
+      "epoch": 1.7708022220894808,
+      "grad_norm": 0.514433283638377,
+      "learning_rate": 7.865108019793082e-06,
+      "loss": 0.3174,
+      "step": 3706
+    },
+    {
+      "epoch": 1.771280090795054,
+      "grad_norm": 0.58396258202816,
+      "learning_rate": 7.859963027658322e-06,
+      "loss": 0.3238,
+      "step": 3707
+    },
+    {
+      "epoch": 1.7717579595006272,
+      "grad_norm": 0.4542920457799816,
+      "learning_rate": 7.854818629132065e-06,
+      "loss": 0.3169,
+      "step": 3708
+    },
+    {
+      "epoch": 1.7722358282062003,
+      "grad_norm": 0.49838845584475117,
+      "learning_rate": 7.849674825641282e-06,
+      "loss": 0.323,
+      "step": 3709
+    },
+    {
+      "epoch": 1.7727136969117736,
+      "grad_norm": 0.4707270211274528,
+      "learning_rate": 7.844531618612772e-06,
+      "loss": 0.3196,
+      "step": 3710
+    },
+    {
+      "epoch": 1.7731915656173465,
+      "grad_norm": 0.45205689040938446,
+      "learning_rate": 7.839389009473167e-06,
+      "loss": 0.3276,
+      "step": 3711
+    },
+    {
+      "epoch": 1.7736694343229198,
+      "grad_norm": 0.4717638415399161,
+      "learning_rate": 7.83424699964894e-06,
+      "loss": 0.3403,
+      "step": 3712
+    },
+    {
+      "epoch": 1.7741473030284929,
+      "grad_norm": 0.5086223507316545,
+      "learning_rate": 7.829105590566393e-06,
+      "loss": 0.3124,
+      "step": 3713
+    },
+    {
+      "epoch": 1.774625171734066,
+      "grad_norm": 0.49314322583865455,
+      "learning_rate": 7.823964783651659e-06,
+      "loss": 0.3089,
+      "step": 3714
+    },
+    {
+      "epoch": 1.7751030404396393,
+      "grad_norm": 0.4623851784080253,
+      "learning_rate": 7.81882458033071e-06,
+      "loss": 0.3048,
+      "step": 3715
+    },
+    {
+      "epoch": 1.7755809091452124,
+      "grad_norm": 0.5172646679609034,
+      "learning_rate": 7.813684982029347e-06,
+      "loss": 0.3164,
+      "step": 3716
+    },
+    {
+      "epoch": 1.7760587778507855,
+      "grad_norm": 0.4947812594178152,
+      "learning_rate": 7.808545990173204e-06,
+      "loss": 0.3134,
+      "step": 3717
+    },
+    {
+      "epoch": 1.7765366465563588,
+      "grad_norm": 0.4779432224978824,
+      "learning_rate": 7.803407606187748e-06,
+      "loss": 0.3098,
+      "step": 3718
+    },
+    {
+      "epoch": 1.7770145152619317,
+      "grad_norm": 0.5012122088224422,
+      "learning_rate": 7.798269831498275e-06,
+      "loss": 0.3096,
+      "step": 3719
+    },
+    {
+      "epoch": 1.777492383967505,
+      "grad_norm": 0.47084003336822555,
+      "learning_rate": 7.79313266752991e-06,
+      "loss": 0.312,
+      "step": 3720
+    },
+    {
+      "epoch": 1.777970252673078,
+      "grad_norm": 0.47286383043551417,
+      "learning_rate": 7.787996115707617e-06,
+      "loss": 0.3197,
+      "step": 3721
+    },
+    {
+      "epoch": 1.7784481213786512,
+      "grad_norm": 0.4780990261307311,
+      "learning_rate": 7.782860177456183e-06,
+      "loss": 0.3041,
+      "step": 3722
+    },
+    {
+      "epoch": 1.7789259900842245,
+      "grad_norm": 0.5239288813334391,
+      "learning_rate": 7.777724854200224e-06,
+      "loss": 0.3107,
+      "step": 3723
+    },
+    {
+      "epoch": 1.7794038587897973,
+      "grad_norm": 0.47039194395335954,
+      "learning_rate": 7.77259014736419e-06,
+      "loss": 0.3097,
+      "step": 3724
+    },
+    {
+      "epoch": 1.7798817274953707,
+      "grad_norm": 0.45970590084025953,
+      "learning_rate": 7.767456058372362e-06,
+      "loss": 0.3211,
+      "step": 3725
+    },
+    {
+      "epoch": 1.7803595962009438,
+      "grad_norm": 0.4734707205616713,
+      "learning_rate": 7.762322588648839e-06,
+      "loss": 0.3074,
+      "step": 3726
+    },
+    {
+      "epoch": 1.7808374649065168,
+      "grad_norm": 0.5154833143648376,
+      "learning_rate": 7.757189739617556e-06,
+      "loss": 0.3048,
+      "step": 3727
+    },
+    {
+      "epoch": 1.7813153336120902,
+      "grad_norm": 0.47648033493201797,
+      "learning_rate": 7.75205751270228e-06,
+      "loss": 0.325,
+      "step": 3728
+    },
+    {
+      "epoch": 1.7817932023176632,
+      "grad_norm": 0.46731827189098973,
+      "learning_rate": 7.746925909326597e-06,
+      "loss": 0.3175,
+      "step": 3729
+    },
+    {
+      "epoch": 1.7822710710232363,
+      "grad_norm": 0.4600389521064629,
+      "learning_rate": 7.741794930913922e-06,
+      "loss": 0.3058,
+      "step": 3730
+    },
+    {
+      "epoch": 1.7827489397288097,
+      "grad_norm": 0.4471951076152281,
+      "learning_rate": 7.7366645788875e-06,
+      "loss": 0.3226,
+      "step": 3731
+    },
+    {
+      "epoch": 1.7832268084343825,
+      "grad_norm": 0.45824019393467236,
+      "learning_rate": 7.731534854670398e-06,
+      "loss": 0.3184,
+      "step": 3732
+    },
+    {
+      "epoch": 1.7837046771399558,
+      "grad_norm": 0.5039947063633161,
+      "learning_rate": 7.726405759685512e-06,
+      "loss": 0.3294,
+      "step": 3733
+    },
+    {
+      "epoch": 1.784182545845529,
+      "grad_norm": 0.48151111030694355,
+      "learning_rate": 7.721277295355566e-06,
+      "loss": 0.3135,
+      "step": 3734
+    },
+    {
+      "epoch": 1.784660414551102,
+      "grad_norm": 0.47276411179310734,
+      "learning_rate": 7.7161494631031e-06,
+      "loss": 0.3087,
+      "step": 3735
+    },
+    {
+      "epoch": 1.7851382832566753,
+      "grad_norm": 0.5222088309394991,
+      "learning_rate": 7.711022264350483e-06,
+      "loss": 0.3032,
+      "step": 3736
+    },
+    {
+      "epoch": 1.7856161519622482,
+      "grad_norm": 0.45961959884253906,
+      "learning_rate": 7.705895700519915e-06,
+      "loss": 0.324,
+      "step": 3737
+    },
+    {
+      "epoch": 1.7860940206678215,
+      "grad_norm": 0.45367078593289256,
+      "learning_rate": 7.700769773033414e-06,
+      "loss": 0.312,
+      "step": 3738
+    },
+    {
+      "epoch": 1.7865718893733946,
+      "grad_norm": 0.48592430357796496,
+      "learning_rate": 7.69564448331282e-06,
+      "loss": 0.3064,
+      "step": 3739
+    },
+    {
+      "epoch": 1.7870497580789677,
+      "grad_norm": 0.49401805489694345,
+      "learning_rate": 7.690519832779799e-06,
+      "loss": 0.3158,
+      "step": 3740
+    },
+    {
+      "epoch": 1.787527626784541,
+      "grad_norm": 0.4920841853592163,
+      "learning_rate": 7.685395822855837e-06,
+      "loss": 0.3193,
+      "step": 3741
+    },
+    {
+      "epoch": 1.7880054954901141,
+      "grad_norm": 0.48732165094819957,
+      "learning_rate": 7.680272454962245e-06,
+      "loss": 0.3163,
+      "step": 3742
+    },
+    {
+      "epoch": 1.7884833641956872,
+      "grad_norm": 0.49308527781981676,
+      "learning_rate": 7.675149730520155e-06,
+      "loss": 0.3173,
+      "step": 3743
+    },
+    {
+      "epoch": 1.7889612329012605,
+      "grad_norm": 0.48955038846140736,
+      "learning_rate": 7.670027650950519e-06,
+      "loss": 0.3266,
+      "step": 3744
+    },
+    {
+      "epoch": 1.7894391016068334,
+      "grad_norm": 0.4710450744507011,
+      "learning_rate": 7.664906217674115e-06,
+      "loss": 0.309,
+      "step": 3745
+    },
+    {
+      "epoch": 1.7899169703124067,
+      "grad_norm": 0.4738846393248144,
+      "learning_rate": 7.659785432111533e-06,
+      "loss": 0.3105,
+      "step": 3746
+    },
+    {
+      "epoch": 1.7903948390179798,
+      "grad_norm": 0.48221267868714024,
+      "learning_rate": 7.654665295683192e-06,
+      "loss": 0.3188,
+      "step": 3747
+    },
+    {
+      "epoch": 1.790872707723553,
+      "grad_norm": 0.46375474845886444,
+      "learning_rate": 7.649545809809329e-06,
+      "loss": 0.2975,
+      "step": 3748
+    },
+    {
+      "epoch": 1.7913505764291262,
+      "grad_norm": 0.47721809669352233,
+      "learning_rate": 7.644426975909995e-06,
+      "loss": 0.3113,
+      "step": 3749
+    },
+    {
+      "epoch": 1.791828445134699,
+      "grad_norm": 4.805146189592445,
+      "learning_rate": 7.639308795405066e-06,
+      "loss": 0.3143,
+      "step": 3750
+    },
+    {
+      "epoch": 1.7923063138402724,
+      "grad_norm": 0.5455951254365826,
+      "learning_rate": 7.634191269714238e-06,
+      "loss": 0.2981,
+      "step": 3751
+    },
+    {
+      "epoch": 1.7927841825458455,
+      "grad_norm": 0.4744407650385994,
+      "learning_rate": 7.6290744002570176e-06,
+      "loss": 0.3249,
+      "step": 3752
+    },
+    {
+      "epoch": 1.7932620512514186,
+      "grad_norm": 0.45267368874034103,
+      "learning_rate": 7.6239581884527354e-06,
+      "loss": 0.3169,
+      "step": 3753
+    },
+    {
+      "epoch": 1.793739919956992,
+      "grad_norm": 0.46081369771470404,
+      "learning_rate": 7.618842635720542e-06,
+      "loss": 0.3122,
+      "step": 3754
+    },
+    {
+      "epoch": 1.794217788662565,
+      "grad_norm": 0.46409053301581027,
+      "learning_rate": 7.613727743479395e-06,
+      "loss": 0.317,
+      "step": 3755
+    },
+    {
+      "epoch": 1.794695657368138,
+      "grad_norm": 0.45315984581264734,
+      "learning_rate": 7.608613513148081e-06,
+      "loss": 0.3253,
+      "step": 3756
+    },
+    {
+      "epoch": 1.7951735260737114,
+      "grad_norm": 0.44114117490408994,
+      "learning_rate": 7.6034999461451956e-06,
+      "loss": 0.3272,
+      "step": 3757
+    },
+    {
+      "epoch": 1.7956513947792843,
+      "grad_norm": 1.5565443010930862,
+      "learning_rate": 7.5983870438891505e-06,
+      "loss": 0.3197,
+      "step": 3758
+    },
+    {
+      "epoch": 1.7961292634848576,
+      "grad_norm": 0.4911118507351176,
+      "learning_rate": 7.593274807798175e-06,
+      "loss": 0.3245,
+      "step": 3759
+    },
+    {
+      "epoch": 1.7966071321904307,
+      "grad_norm": 0.44995132281699324,
+      "learning_rate": 7.588163239290316e-06,
+      "loss": 0.3186,
+      "step": 3760
+    },
+    {
+      "epoch": 1.7970850008960038,
+      "grad_norm": 0.46456788831578155,
+      "learning_rate": 7.583052339783428e-06,
+      "loss": 0.3172,
+      "step": 3761
+    },
+    {
+      "epoch": 1.797562869601577,
+      "grad_norm": 0.4734482643803076,
+      "learning_rate": 7.5779421106951874e-06,
+      "loss": 0.3086,
+      "step": 3762
+    },
+    {
+      "epoch": 1.7980407383071502,
+      "grad_norm": 0.46205250286552624,
+      "learning_rate": 7.572832553443083e-06,
+      "loss": 0.309,
+      "step": 3763
+    },
+    {
+      "epoch": 1.7985186070127233,
+      "grad_norm": 0.5474167526568322,
+      "learning_rate": 7.567723669444411e-06,
+      "loss": 0.3263,
+      "step": 3764
+    },
+    {
+      "epoch": 1.7989964757182964,
+      "grad_norm": 0.4807279141914673,
+      "learning_rate": 7.562615460116289e-06,
+      "loss": 0.3114,
+      "step": 3765
+    },
+    {
+      "epoch": 1.7994743444238694,
+      "grad_norm": 0.47642414056800825,
+      "learning_rate": 7.557507926875646e-06,
+      "loss": 0.3141,
+      "step": 3766
+    },
+    {
+      "epoch": 1.7999522131294428,
+      "grad_norm": 0.44512012977708126,
+      "learning_rate": 7.552401071139217e-06,
+      "loss": 0.3202,
+      "step": 3767
+    },
+    {
+      "epoch": 1.8004300818350158,
+      "grad_norm": 0.48382096788679013,
+      "learning_rate": 7.547294894323556e-06,
+      "loss": 0.3183,
+      "step": 3768
+    },
+    {
+      "epoch": 1.800907950540589,
+      "grad_norm": 0.4631932373502014,
+      "learning_rate": 7.542189397845028e-06,
+      "loss": 0.3163,
+      "step": 3769
+    },
+    {
+      "epoch": 1.8013858192461623,
+      "grad_norm": 0.4661580950852657,
+      "learning_rate": 7.537084583119802e-06,
+      "loss": 0.3009,
+      "step": 3770
+    },
+    {
+      "epoch": 1.8018636879517351,
+      "grad_norm": 0.44890347173646833,
+      "learning_rate": 7.531980451563869e-06,
+      "loss": 0.324,
+      "step": 3771
+    },
+    {
+      "epoch": 1.8023415566573084,
+      "grad_norm": 0.46211335360678135,
+      "learning_rate": 7.5268770045930255e-06,
+      "loss": 0.3063,
+      "step": 3772
+    },
+    {
+      "epoch": 1.8028194253628815,
+      "grad_norm": 0.47324725173384913,
+      "learning_rate": 7.521774243622875e-06,
+      "loss": 0.3232,
+      "step": 3773
+    },
+    {
+      "epoch": 1.8032972940684546,
+      "grad_norm": 0.4769086546795592,
+      "learning_rate": 7.516672170068835e-06,
+      "loss": 0.3081,
+      "step": 3774
+    },
+    {
+      "epoch": 1.803775162774028,
+      "grad_norm": 0.4987810815938432,
+      "learning_rate": 7.511570785346129e-06,
+      "loss": 0.3331,
+      "step": 3775
+    },
+    {
+      "epoch": 1.804253031479601,
+      "grad_norm": 0.515528221050649,
+      "learning_rate": 7.506470090869793e-06,
+      "loss": 0.3081,
+      "step": 3776
+    },
+    {
+      "epoch": 1.8047309001851741,
+      "grad_norm": 0.48258513179828666,
+      "learning_rate": 7.501370088054667e-06,
+      "loss": 0.3253,
+      "step": 3777
+    },
+    {
+      "epoch": 1.8052087688907472,
+      "grad_norm": 0.4687979828089521,
+      "learning_rate": 7.496270778315404e-06,
+      "loss": 0.3038,
+      "step": 3778
+    },
+    {
+      "epoch": 1.8056866375963203,
+      "grad_norm": 0.4675700871564693,
+      "learning_rate": 7.4911721630664644e-06,
+      "loss": 0.3131,
+      "step": 3779
+    },
+    {
+      "epoch": 1.8061645063018936,
+      "grad_norm": 0.4488321146024742,
+      "learning_rate": 7.486074243722109e-06,
+      "loss": 0.3097,
+      "step": 3780
+    },
+    {
+      "epoch": 1.8066423750074667,
+      "grad_norm": 0.4731548586712247,
+      "learning_rate": 7.480977021696414e-06,
+      "loss": 0.3036,
+      "step": 3781
+    },
+    {
+      "epoch": 1.8071202437130398,
+      "grad_norm": 0.5469397778562964,
+      "learning_rate": 7.475880498403261e-06,
+      "loss": 0.32,
+      "step": 3782
+    },
+    {
+      "epoch": 1.8075981124186131,
+      "grad_norm": 0.45764608523635797,
+      "learning_rate": 7.470784675256329e-06,
+      "loss": 0.3237,
+      "step": 3783
+    },
+    {
+      "epoch": 1.808075981124186,
+      "grad_norm": 0.4850459958523287,
+      "learning_rate": 7.4656895536691154e-06,
+      "loss": 0.3132,
+      "step": 3784
+    },
+    {
+      "epoch": 1.8085538498297593,
+      "grad_norm": 0.45406496016389747,
+      "learning_rate": 7.460595135054916e-06,
+      "loss": 0.3124,
+      "step": 3785
+    },
+    {
+      "epoch": 1.8090317185353324,
+      "grad_norm": 0.45950621992569934,
+      "learning_rate": 7.455501420826831e-06,
+      "loss": 0.3218,
+      "step": 3786
+    },
+    {
+      "epoch": 1.8095095872409055,
+      "grad_norm": 0.4591525516581785,
+      "learning_rate": 7.450408412397767e-06,
+      "loss": 0.3206,
+      "step": 3787
+    },
+    {
+      "epoch": 1.8099874559464788,
+      "grad_norm": 0.508129994813578,
+      "learning_rate": 7.445316111180436e-06,
+      "loss": 0.3265,
+      "step": 3788
+    },
+    {
+      "epoch": 1.810465324652052,
+      "grad_norm": 0.4487238152879312,
+      "learning_rate": 7.440224518587353e-06,
+      "loss": 0.317,
+      "step": 3789
+    },
+    {
+      "epoch": 1.810943193357625,
+      "grad_norm": 0.44328561901531993,
+      "learning_rate": 7.435133636030831e-06,
+      "loss": 0.3091,
+      "step": 3790
+    },
+    {
+      "epoch": 1.811421062063198,
+      "grad_norm": 0.4386069882959427,
+      "learning_rate": 7.430043464923e-06,
+      "loss": 0.33,
+      "step": 3791
+    },
+    {
+      "epoch": 1.8118989307687712,
+      "grad_norm": 0.4737344270055403,
+      "learning_rate": 7.424954006675775e-06,
+      "loss": 0.3075,
+      "step": 3792
+    },
+    {
+      "epoch": 1.8123767994743445,
+      "grad_norm": 0.46437704276405334,
+      "learning_rate": 7.419865262700887e-06,
+      "loss": 0.3163,
+      "step": 3793
+    },
+    {
+      "epoch": 1.8128546681799176,
+      "grad_norm": 0.4582800308955872,
+      "learning_rate": 7.414777234409863e-06,
+      "loss": 0.336,
+      "step": 3794
+    },
+    {
+      "epoch": 1.8133325368854907,
+      "grad_norm": 0.4639778390365935,
+      "learning_rate": 7.4096899232140295e-06,
+      "loss": 0.3253,
+      "step": 3795
+    },
+    {
+      "epoch": 1.813810405591064,
+      "grad_norm": 0.47191555510498095,
+      "learning_rate": 7.40460333052452e-06,
+      "loss": 0.3164,
+      "step": 3796
+    },
+    {
+      "epoch": 1.8142882742966369,
+      "grad_norm": 0.8754820249606956,
+      "learning_rate": 7.399517457752266e-06,
+      "loss": 0.3068,
+      "step": 3797
+    },
+    {
+      "epoch": 1.8147661430022102,
+      "grad_norm": 0.4647598211303201,
+      "learning_rate": 7.394432306307997e-06,
+      "loss": 0.3236,
+      "step": 3798
+    },
+    {
+      "epoch": 1.8152440117077833,
+      "grad_norm": 0.8061057528954818,
+      "learning_rate": 7.389347877602242e-06,
+      "loss": 0.3204,
+      "step": 3799
+    },
+    {
+      "epoch": 1.8157218804133564,
+      "grad_norm": 0.4832339467960314,
+      "learning_rate": 7.384264173045339e-06,
+      "loss": 0.3068,
+      "step": 3800
+    },
+    {
+      "epoch": 1.8161997491189297,
+      "grad_norm": 0.48537677126279627,
+      "learning_rate": 7.379181194047412e-06,
+      "loss": 0.3044,
+      "step": 3801
+    },
+    {
+      "epoch": 1.8166776178245028,
+      "grad_norm": 0.4370504478039206,
+      "learning_rate": 7.374098942018388e-06,
+      "loss": 0.3103,
+      "step": 3802
+    },
+    {
+      "epoch": 1.8171554865300759,
+      "grad_norm": 0.4540403617689952,
+      "learning_rate": 7.3690174183680015e-06,
+      "loss": 0.293,
+      "step": 3803
+    },
+    {
+      "epoch": 1.817633355235649,
+      "grad_norm": 0.44834138741434454,
+      "learning_rate": 7.363936624505767e-06,
+      "loss": 0.3038,
+      "step": 3804
+    },
+    {
+      "epoch": 1.818111223941222,
+      "grad_norm": 0.5071526227296054,
+      "learning_rate": 7.358856561841021e-06,
+      "loss": 0.33,
+      "step": 3805
+    },
+    {
+      "epoch": 1.8185890926467954,
+      "grad_norm": 0.45524356000949795,
+      "learning_rate": 7.353777231782873e-06,
+      "loss": 0.3037,
+      "step": 3806
+    },
+    {
+      "epoch": 1.8190669613523685,
+      "grad_norm": 0.4323954513417616,
+      "learning_rate": 7.3486986357402414e-06,
+      "loss": 0.2984,
+      "step": 3807
+    },
+    {
+      "epoch": 1.8195448300579415,
+      "grad_norm": 0.4748222174160758,
+      "learning_rate": 7.343620775121842e-06,
+      "loss": 0.3197,
+      "step": 3808
+    },
+    {
+      "epoch": 1.8200226987635149,
+      "grad_norm": 0.5434867730758338,
+      "learning_rate": 7.338543651336181e-06,
+      "loss": 0.3048,
+      "step": 3809
+    },
+    {
+      "epoch": 1.8205005674690877,
+      "grad_norm": 0.5261663632567056,
+      "learning_rate": 7.333467265791563e-06,
+      "loss": 0.3269,
+      "step": 3810
+    },
+    {
+      "epoch": 1.820978436174661,
+      "grad_norm": 0.4668224192467844,
+      "learning_rate": 7.328391619896092e-06,
+      "loss": 0.3256,
+      "step": 3811
+    },
+    {
+      "epoch": 1.8214563048802341,
+      "grad_norm": 0.47668206195798435,
+      "learning_rate": 7.3233167150576554e-06,
+      "loss": 0.3268,
+      "step": 3812
+    },
+    {
+      "epoch": 1.8219341735858072,
+      "grad_norm": 0.5070836849852632,
+      "learning_rate": 7.318242552683948e-06,
+      "loss": 0.321,
+      "step": 3813
+    },
+    {
+      "epoch": 1.8224120422913805,
+      "grad_norm": 0.494039037408748,
+      "learning_rate": 7.3131691341824515e-06,
+      "loss": 0.3083,
+      "step": 3814
+    },
+    {
+      "epoch": 1.8228899109969536,
+      "grad_norm": 0.4701617746326885,
+      "learning_rate": 7.308096460960441e-06,
+      "loss": 0.3036,
+      "step": 3815
+    },
+    {
+      "epoch": 1.8233677797025267,
+      "grad_norm": 0.48182454253938445,
+      "learning_rate": 7.303024534424987e-06,
+      "loss": 0.3191,
+      "step": 3816
+    },
+    {
+      "epoch": 1.8238456484081,
+      "grad_norm": 0.4968254147423399,
+      "learning_rate": 7.297953355982956e-06,
+      "loss": 0.3228,
+      "step": 3817
+    },
+    {
+      "epoch": 1.824323517113673,
+      "grad_norm": 0.48879138443250414,
+      "learning_rate": 7.292882927040999e-06,
+      "loss": 0.3083,
+      "step": 3818
+    },
+    {
+      "epoch": 1.8248013858192462,
+      "grad_norm": 0.47334912056885264,
+      "learning_rate": 7.287813249005565e-06,
+      "loss": 0.3111,
+      "step": 3819
+    },
+    {
+      "epoch": 1.8252792545248193,
+      "grad_norm": 0.4884790884097843,
+      "learning_rate": 7.282744323282895e-06,
+      "loss": 0.3256,
+      "step": 3820
+    },
+    {
+      "epoch": 1.8257571232303924,
+      "grad_norm": 0.5022837404885262,
+      "learning_rate": 7.277676151279019e-06,
+      "loss": 0.3076,
+      "step": 3821
+    },
+    {
+      "epoch": 1.8262349919359657,
+      "grad_norm": 0.47479766132527035,
+      "learning_rate": 7.272608734399754e-06,
+      "loss": 0.3033,
+      "step": 3822
+    },
+    {
+      "epoch": 1.8267128606415386,
+      "grad_norm": 1.3020109975876173,
+      "learning_rate": 7.26754207405072e-06,
+      "loss": 0.3079,
+      "step": 3823
+    },
+    {
+      "epoch": 1.827190729347112,
+      "grad_norm": 0.4919024454500918,
+      "learning_rate": 7.262476171637311e-06,
+      "loss": 0.3095,
+      "step": 3824
+    },
+    {
+      "epoch": 1.827668598052685,
+      "grad_norm": 0.44665992779272773,
+      "learning_rate": 7.2574110285647244e-06,
+      "loss": 0.3103,
+      "step": 3825
+    },
+    {
+      "epoch": 1.828146466758258,
+      "grad_norm": 0.44838920186117176,
+      "learning_rate": 7.252346646237942e-06,
+      "loss": 0.3092,
+      "step": 3826
+    },
+    {
+      "epoch": 1.8286243354638314,
+      "grad_norm": 0.7199341984258678,
+      "learning_rate": 7.24728302606173e-06,
+      "loss": 0.3233,
+      "step": 3827
+    },
+    {
+      "epoch": 1.8291022041694045,
+      "grad_norm": 0.4594979265995351,
+      "learning_rate": 7.242220169440649e-06,
+      "loss": 0.3203,
+      "step": 3828
+    },
+    {
+      "epoch": 1.8295800728749776,
+      "grad_norm": 0.46434411184296537,
+      "learning_rate": 7.2371580777790494e-06,
+      "loss": 0.3297,
+      "step": 3829
+    },
+    {
+      "epoch": 1.830057941580551,
+      "grad_norm": 0.4661421339585899,
+      "learning_rate": 7.232096752481061e-06,
+      "loss": 0.31,
+      "step": 3830
+    },
+    {
+      "epoch": 1.8305358102861238,
+      "grad_norm": 0.4796291545667761,
+      "learning_rate": 7.2270361949506075e-06,
+      "loss": 0.3196,
+      "step": 3831
+    },
+    {
+      "epoch": 1.831013678991697,
+      "grad_norm": 0.46983415973458353,
+      "learning_rate": 7.2219764065914024e-06,
+      "loss": 0.317,
+      "step": 3832
+    },
+    {
+      "epoch": 1.8314915476972702,
+      "grad_norm": 0.5194541122421759,
+      "learning_rate": 7.216917388806936e-06,
+      "loss": 0.323,
+      "step": 3833
+    },
+    {
+      "epoch": 1.8319694164028433,
+      "grad_norm": 0.483395653309714,
+      "learning_rate": 7.211859143000495e-06,
+      "loss": 0.3084,
+      "step": 3834
+    },
+    {
+      "epoch": 1.8324472851084166,
+      "grad_norm": 0.4802816137671634,
+      "learning_rate": 7.206801670575145e-06,
+      "loss": 0.3213,
+      "step": 3835
+    },
+    {
+      "epoch": 1.8329251538139895,
+      "grad_norm": 0.46676119411435246,
+      "learning_rate": 7.2017449729337396e-06,
+      "loss": 0.2915,
+      "step": 3836
+    },
+    {
+      "epoch": 1.8334030225195628,
+      "grad_norm": 0.4751615080956812,
+      "learning_rate": 7.196689051478917e-06,
+      "loss": 0.3172,
+      "step": 3837
+    },
+    {
+      "epoch": 1.8338808912251359,
+      "grad_norm": 0.4518355207103295,
+      "learning_rate": 7.191633907613103e-06,
+      "loss": 0.3037,
+      "step": 3838
+    },
+    {
+      "epoch": 1.834358759930709,
+      "grad_norm": 0.4964989199148571,
+      "learning_rate": 7.186579542738507e-06,
+      "loss": 0.3241,
+      "step": 3839
+    },
+    {
+      "epoch": 1.8348366286362823,
+      "grad_norm": 0.45654528244066295,
+      "learning_rate": 7.181525958257116e-06,
+      "loss": 0.3127,
+      "step": 3840
+    },
+    {
+      "epoch": 1.8353144973418554,
+      "grad_norm": 0.46918542594424667,
+      "learning_rate": 7.176473155570707e-06,
+      "loss": 0.3119,
+      "step": 3841
+    },
+    {
+      "epoch": 1.8357923660474285,
+      "grad_norm": 0.4507255821310943,
+      "learning_rate": 7.171421136080841e-06,
+      "loss": 0.3096,
+      "step": 3842
+    },
+    {
+      "epoch": 1.8362702347530018,
+      "grad_norm": 0.4345435682976803,
+      "learning_rate": 7.1663699011888524e-06,
+      "loss": 0.3327,
+      "step": 3843
+    },
+    {
+      "epoch": 1.8367481034585746,
+      "grad_norm": 0.46364207609499564,
+      "learning_rate": 7.1613194522958705e-06,
+      "loss": 0.3321,
+      "step": 3844
+    },
+    {
+      "epoch": 1.837225972164148,
+      "grad_norm": 1.3350277825716896,
+      "learning_rate": 7.156269790802801e-06,
+      "loss": 0.3215,
+      "step": 3845
+    },
+    {
+      "epoch": 1.837703840869721,
+      "grad_norm": 0.45482610525005635,
+      "learning_rate": 7.151220918110326e-06,
+      "loss": 0.3122,
+      "step": 3846
+    },
+    {
+      "epoch": 1.8381817095752941,
+      "grad_norm": 0.4799378572791614,
+      "learning_rate": 7.146172835618919e-06,
+      "loss": 0.3202,
+      "step": 3847
+    },
+    {
+      "epoch": 1.8386595782808675,
+      "grad_norm": 0.4560594548210913,
+      "learning_rate": 7.1411255447288266e-06,
+      "loss": 0.3178,
+      "step": 3848
+    },
+    {
+      "epoch": 1.8391374469864403,
+      "grad_norm": 0.42536499972455255,
+      "learning_rate": 7.136079046840078e-06,
+      "loss": 0.3098,
+      "step": 3849
+    },
+    {
+      "epoch": 1.8396153156920136,
+      "grad_norm": 0.4728281050415246,
+      "learning_rate": 7.131033343352483e-06,
+      "loss": 0.3162,
+      "step": 3850
+    },
+    {
+      "epoch": 1.8400931843975867,
+      "grad_norm": 0.45137859761216986,
+      "learning_rate": 7.125988435665632e-06,
+      "loss": 0.3258,
+      "step": 3851
+    },
+    {
+      "epoch": 1.8405710531031598,
+      "grad_norm": 0.4305757718398639,
+      "learning_rate": 7.120944325178889e-06,
+      "loss": 0.3034,
+      "step": 3852
+    },
+    {
+      "epoch": 1.8410489218087331,
+      "grad_norm": 0.46513614861758734,
+      "learning_rate": 7.1159010132914065e-06,
+      "loss": 0.3233,
+      "step": 3853
+    },
+    {
+      "epoch": 1.8415267905143062,
+      "grad_norm": 0.4516843311861331,
+      "learning_rate": 7.1108585014021095e-06,
+      "loss": 0.3123,
+      "step": 3854
+    },
+    {
+      "epoch": 1.8420046592198793,
+      "grad_norm": 0.44599301993670454,
+      "learning_rate": 7.105816790909699e-06,
+      "loss": 0.303,
+      "step": 3855
+    },
+    {
+      "epoch": 1.8424825279254526,
+      "grad_norm": 0.45407721798005035,
+      "learning_rate": 7.100775883212658e-06,
+      "loss": 0.2951,
+      "step": 3856
+    },
+    {
+      "epoch": 1.8429603966310255,
+      "grad_norm": 0.6330334384592656,
+      "learning_rate": 7.095735779709248e-06,
+      "loss": 0.3187,
+      "step": 3857
+    },
+    {
+      "epoch": 1.8434382653365988,
+      "grad_norm": 0.45481681430926496,
+      "learning_rate": 7.0906964817974984e-06,
+      "loss": 0.3068,
+      "step": 3858
+    },
+    {
+      "epoch": 1.843916134042172,
+      "grad_norm": 0.4552031236149239,
+      "learning_rate": 7.085657990875227e-06,
+      "loss": 0.3038,
+      "step": 3859
+    },
+    {
+      "epoch": 1.844394002747745,
+      "grad_norm": 0.4834807925984325,
+      "learning_rate": 7.080620308340024e-06,
+      "loss": 0.3047,
+      "step": 3860
+    },
+    {
+      "epoch": 1.8448718714533183,
+      "grad_norm": 0.46556812455338153,
+      "learning_rate": 7.075583435589248e-06,
+      "loss": 0.3085,
+      "step": 3861
+    },
+    {
+      "epoch": 1.8453497401588912,
+      "grad_norm": 0.43412355998395485,
+      "learning_rate": 7.07054737402004e-06,
+      "loss": 0.297,
+      "step": 3862
+    },
+    {
+      "epoch": 1.8458276088644645,
+      "grad_norm": 0.4346076047173381,
+      "learning_rate": 7.065512125029318e-06,
+      "loss": 0.3116,
+      "step": 3863
+    },
+    {
+      "epoch": 1.8463054775700376,
+      "grad_norm": 0.4604136030349345,
+      "learning_rate": 7.060477690013767e-06,
+      "loss": 0.3257,
+      "step": 3864
+    },
+    {
+      "epoch": 1.8467833462756107,
+      "grad_norm": 0.4900954658380756,
+      "learning_rate": 7.055444070369852e-06,
+      "loss": 0.3204,
+      "step": 3865
+    },
+    {
+      "epoch": 1.847261214981184,
+      "grad_norm": 0.4813520085061644,
+      "learning_rate": 7.050411267493815e-06,
+      "loss": 0.3139,
+      "step": 3866
+    },
+    {
+      "epoch": 1.847739083686757,
+      "grad_norm": 0.4567730976405721,
+      "learning_rate": 7.045379282781659e-06,
+      "loss": 0.3125,
+      "step": 3867
+    },
+    {
+      "epoch": 1.8482169523923302,
+      "grad_norm": 0.474550906247094,
+      "learning_rate": 7.040348117629172e-06,
+      "loss": 0.3149,
+      "step": 3868
+    },
+    {
+      "epoch": 1.8486948210979035,
+      "grad_norm": 0.4432230508973788,
+      "learning_rate": 7.035317773431911e-06,
+      "loss": 0.3055,
+      "step": 3869
+    },
+    {
+      "epoch": 1.8491726898034764,
+      "grad_norm": 0.5522942357232046,
+      "learning_rate": 7.0302882515852025e-06,
+      "loss": 0.3366,
+      "step": 3870
+    },
+    {
+      "epoch": 1.8496505585090497,
+      "grad_norm": 0.4752790460554777,
+      "learning_rate": 7.025259553484145e-06,
+      "loss": 0.3324,
+      "step": 3871
+    },
+    {
+      "epoch": 1.8501284272146228,
+      "grad_norm": 0.4643972132496278,
+      "learning_rate": 7.020231680523616e-06,
+      "loss": 0.3093,
+      "step": 3872
+    },
+    {
+      "epoch": 1.8506062959201959,
+      "grad_norm": 0.4554789111097097,
+      "learning_rate": 7.015204634098256e-06,
+      "loss": 0.3319,
+      "step": 3873
+    },
+    {
+      "epoch": 1.8510841646257692,
+      "grad_norm": 0.45867714400719684,
+      "learning_rate": 7.010178415602485e-06,
+      "loss": 0.3133,
+      "step": 3874
+    },
+    {
+      "epoch": 1.851562033331342,
+      "grad_norm": 0.45519094876743393,
+      "learning_rate": 7.005153026430476e-06,
+      "loss": 0.311,
+      "step": 3875
+    },
+    {
+      "epoch": 1.8520399020369154,
+      "grad_norm": 0.4456318199658115,
+      "learning_rate": 7.00012846797619e-06,
+      "loss": 0.3298,
+      "step": 3876
+    },
+    {
+      "epoch": 1.8525177707424885,
+      "grad_norm": 0.4877288837711452,
+      "learning_rate": 6.995104741633354e-06,
+      "loss": 0.302,
+      "step": 3877
+    },
+    {
+      "epoch": 1.8529956394480616,
+      "grad_norm": 0.45490188442789403,
+      "learning_rate": 6.990081848795453e-06,
+      "loss": 0.3225,
+      "step": 3878
+    },
+    {
+      "epoch": 1.8534735081536349,
+      "grad_norm": 0.46268349832494854,
+      "learning_rate": 6.985059790855755e-06,
+      "loss": 0.31,
+      "step": 3879
+    },
+    {
+      "epoch": 1.853951376859208,
+      "grad_norm": 0.47114477836168356,
+      "learning_rate": 6.980038569207291e-06,
+      "loss": 0.3215,
+      "step": 3880
+    },
+    {
+      "epoch": 1.854429245564781,
+      "grad_norm": 0.4627024705877978,
+      "learning_rate": 6.975018185242852e-06,
+      "loss": 0.322,
+      "step": 3881
+    },
+    {
+      "epoch": 1.8549071142703544,
+      "grad_norm": 0.4403644932273325,
+      "learning_rate": 6.969998640355011e-06,
+      "loss": 0.3075,
+      "step": 3882
+    },
+    {
+      "epoch": 1.8553849829759272,
+      "grad_norm": 0.4734619620341228,
+      "learning_rate": 6.9649799359361e-06,
+      "loss": 0.3153,
+      "step": 3883
+    },
+    {
+      "epoch": 1.8558628516815006,
+      "grad_norm": 0.45177620940516144,
+      "learning_rate": 6.959962073378216e-06,
+      "loss": 0.3014,
+      "step": 3884
+    },
+    {
+      "epoch": 1.8563407203870737,
+      "grad_norm": 0.4898115665474695,
+      "learning_rate": 6.954945054073228e-06,
+      "loss": 0.3159,
+      "step": 3885
+    },
+    {
+      "epoch": 1.8568185890926467,
+      "grad_norm": 0.49590283970632837,
+      "learning_rate": 6.949928879412768e-06,
+      "loss": 0.3122,
+      "step": 3886
+    },
+    {
+      "epoch": 1.85729645779822,
+      "grad_norm": 0.483933886138119,
+      "learning_rate": 6.944913550788235e-06,
+      "loss": 0.3134,
+      "step": 3887
+    },
+    {
+      "epoch": 1.857774326503793,
+      "grad_norm": 0.503142035037706,
+      "learning_rate": 6.939899069590791e-06,
+      "loss": 0.327,
+      "step": 3888
+    },
+    {
+      "epoch": 1.8582521952093662,
+      "grad_norm": 0.47557368760354807,
+      "learning_rate": 6.934885437211367e-06,
+      "loss": 0.3218,
+      "step": 3889
+    },
+    {
+      "epoch": 1.8587300639149393,
+      "grad_norm": 0.4620961694190902,
+      "learning_rate": 6.929872655040655e-06,
+      "loss": 0.2842,
+      "step": 3890
+    },
+    {
+      "epoch": 1.8592079326205124,
+      "grad_norm": 0.4673591975296846,
+      "learning_rate": 6.924860724469111e-06,
+      "loss": 0.315,
+      "step": 3891
+    },
+    {
+      "epoch": 1.8596858013260857,
+      "grad_norm": 0.46945354634024067,
+      "learning_rate": 6.9198496468869605e-06,
+      "loss": 0.3005,
+      "step": 3892
+    },
+    {
+      "epoch": 1.8601636700316588,
+      "grad_norm": 0.4851381015060144,
+      "learning_rate": 6.914839423684183e-06,
+      "loss": 0.3195,
+      "step": 3893
+    },
+    {
+      "epoch": 1.860641538737232,
+      "grad_norm": 0.5201132587365642,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.3026,
+      "step": 3894
+    },
+    {
+      "epoch": 1.8611194074428052,
+      "grad_norm": 0.5134723697086168,
+      "learning_rate": 6.904821545975507e-06,
+      "loss": 0.3245,
+      "step": 3895
+    },
+    {
+      "epoch": 1.861597276148378,
+      "grad_norm": 0.4518202226167905,
+      "learning_rate": 6.899813894248388e-06,
+      "loss": 0.3088,
+      "step": 3896
+    },
+    {
+      "epoch": 1.8620751448539514,
+      "grad_norm": 0.4683222226158601,
+      "learning_rate": 6.894807102458211e-06,
+      "loss": 0.3239,
+      "step": 3897
+    },
+    {
+      "epoch": 1.8625530135595245,
+      "grad_norm": 0.48007891114431994,
+      "learning_rate": 6.889801171993769e-06,
+      "loss": 0.2836,
+      "step": 3898
+    },
+    {
+      "epoch": 1.8630308822650976,
+      "grad_norm": 0.4772065271771579,
+      "learning_rate": 6.8847961042436185e-06,
+      "loss": 0.3176,
+      "step": 3899
+    },
+    {
+      "epoch": 1.863508750970671,
+      "grad_norm": 0.45177407887015353,
+      "learning_rate": 6.879791900596077e-06,
+      "loss": 0.3081,
+      "step": 3900
+    },
+    {
+      "epoch": 1.8639866196762438,
+      "grad_norm": 0.4729590065660319,
+      "learning_rate": 6.874788562439225e-06,
+      "loss": 0.3102,
+      "step": 3901
+    },
+    {
+      "epoch": 1.864464488381817,
+      "grad_norm": 0.49303248211065703,
+      "learning_rate": 6.869786091160895e-06,
+      "loss": 0.3371,
+      "step": 3902
+    },
+    {
+      "epoch": 1.8649423570873902,
+      "grad_norm": 0.4798724268373024,
+      "learning_rate": 6.864784488148688e-06,
+      "loss": 0.3172,
+      "step": 3903
+    },
+    {
+      "epoch": 1.8654202257929633,
+      "grad_norm": 0.4566461806579496,
+      "learning_rate": 6.859783754789962e-06,
+      "loss": 0.3159,
+      "step": 3904
+    },
+    {
+      "epoch": 1.8658980944985366,
+      "grad_norm": 0.45266075225313984,
+      "learning_rate": 6.854783892471823e-06,
+      "loss": 0.3336,
+      "step": 3905
+    },
+    {
+      "epoch": 1.8663759632041097,
+      "grad_norm": 0.47703082895473686,
+      "learning_rate": 6.849784902581158e-06,
+      "loss": 0.2831,
+      "step": 3906
+    },
+    {
+      "epoch": 1.8668538319096828,
+      "grad_norm": 0.5127241356341654,
+      "learning_rate": 6.8447867865045905e-06,
+      "loss": 0.3055,
+      "step": 3907
+    },
+    {
+      "epoch": 1.867331700615256,
+      "grad_norm": 0.46983255513156474,
+      "learning_rate": 6.83978954562851e-06,
+      "loss": 0.3053,
+      "step": 3908
+    },
+    {
+      "epoch": 1.867809569320829,
+      "grad_norm": 0.44944233352974683,
+      "learning_rate": 6.834793181339068e-06,
+      "loss": 0.312,
+      "step": 3909
+    },
+    {
+      "epoch": 1.8682874380264023,
+      "grad_norm": 0.48300744589768124,
+      "learning_rate": 6.829797695022163e-06,
+      "loss": 0.2879,
+      "step": 3910
+    },
+    {
+      "epoch": 1.8687653067319754,
+      "grad_norm": 0.49195148134310035,
+      "learning_rate": 6.824803088063454e-06,
+      "loss": 0.3086,
+      "step": 3911
+    },
+    {
+      "epoch": 1.8692431754375485,
+      "grad_norm": 0.48226235431323466,
+      "learning_rate": 6.819809361848362e-06,
+      "loss": 0.3161,
+      "step": 3912
+    },
+    {
+      "epoch": 1.8697210441431218,
+      "grad_norm": 0.4673032858019702,
+      "learning_rate": 6.814816517762053e-06,
+      "loss": 0.3009,
+      "step": 3913
+    },
+    {
+      "epoch": 1.8701989128486947,
+      "grad_norm": 0.5120888668650547,
+      "learning_rate": 6.809824557189456e-06,
+      "loss": 0.3047,
+      "step": 3914
+    },
+    {
+      "epoch": 1.870676781554268,
+      "grad_norm": 0.5188704505048501,
+      "learning_rate": 6.804833481515256e-06,
+      "loss": 0.2996,
+      "step": 3915
+    },
+    {
+      "epoch": 1.871154650259841,
+      "grad_norm": 0.4586738347773163,
+      "learning_rate": 6.799843292123883e-06,
+      "loss": 0.3216,
+      "step": 3916
+    },
+    {
+      "epoch": 1.8716325189654142,
+      "grad_norm": 0.4514115914431411,
+      "learning_rate": 6.794853990399533e-06,
+      "loss": 0.3157,
+      "step": 3917
+    },
+    {
+      "epoch": 1.8721103876709875,
+      "grad_norm": 0.4567742118666483,
+      "learning_rate": 6.7898655777261494e-06,
+      "loss": 0.3228,
+      "step": 3918
+    },
+    {
+      "epoch": 1.8725882563765606,
+      "grad_norm": 0.5130560266812821,
+      "learning_rate": 6.784878055487425e-06,
+      "loss": 0.3059,
+      "step": 3919
+    },
+    {
+      "epoch": 1.8730661250821337,
+      "grad_norm": 0.471031405151132,
+      "learning_rate": 6.7798914250668154e-06,
+      "loss": 0.3208,
+      "step": 3920
+    },
+    {
+      "epoch": 1.873543993787707,
+      "grad_norm": 0.4488894073471434,
+      "learning_rate": 6.774905687847526e-06,
+      "loss": 0.3082,
+      "step": 3921
+    },
+    {
+      "epoch": 1.8740218624932798,
+      "grad_norm": 0.4837073839553914,
+      "learning_rate": 6.769920845212506e-06,
+      "loss": 0.3114,
+      "step": 3922
+    },
+    {
+      "epoch": 1.8744997311988532,
+      "grad_norm": 0.45927942054259496,
+      "learning_rate": 6.764936898544466e-06,
+      "loss": 0.3162,
+      "step": 3923
+    },
+    {
+      "epoch": 1.8749775999044263,
+      "grad_norm": 0.517741398437042,
+      "learning_rate": 6.759953849225867e-06,
+      "loss": 0.3065,
+      "step": 3924
+    },
+    {
+      "epoch": 1.8754554686099993,
+      "grad_norm": 0.4830767260559844,
+      "learning_rate": 6.7549716986389146e-06,
+      "loss": 0.3142,
+      "step": 3925
+    },
+    {
+      "epoch": 1.8759333373155727,
+      "grad_norm": 0.4686254460663425,
+      "learning_rate": 6.749990448165572e-06,
+      "loss": 0.3139,
+      "step": 3926
+    },
+    {
+      "epoch": 1.8764112060211455,
+      "grad_norm": 0.4460888626954049,
+      "learning_rate": 6.745010099187552e-06,
+      "loss": 0.3315,
+      "step": 3927
+    },
+    {
+      "epoch": 1.8768890747267188,
+      "grad_norm": 0.4643706696696984,
+      "learning_rate": 6.740030653086311e-06,
+      "loss": 0.2954,
+      "step": 3928
+    },
+    {
+      "epoch": 1.877366943432292,
+      "grad_norm": 0.5033306648528414,
+      "learning_rate": 6.735052111243061e-06,
+      "loss": 0.298,
+      "step": 3929
+    },
+    {
+      "epoch": 1.877844812137865,
+      "grad_norm": 0.44669150190220186,
+      "learning_rate": 6.730074475038766e-06,
+      "loss": 0.3019,
+      "step": 3930
+    },
+    {
+      "epoch": 1.8783226808434383,
+      "grad_norm": 0.4653710125052024,
+      "learning_rate": 6.72509774585413e-06,
+      "loss": 0.307,
+      "step": 3931
+    },
+    {
+      "epoch": 1.8788005495490114,
+      "grad_norm": 0.48207706455379085,
+      "learning_rate": 6.720121925069609e-06,
+      "loss": 0.3321,
+      "step": 3932
+    },
+    {
+      "epoch": 1.8792784182545845,
+      "grad_norm": 0.446573686847004,
+      "learning_rate": 6.715147014065413e-06,
+      "loss": 0.3118,
+      "step": 3933
+    },
+    {
+      "epoch": 1.8797562869601578,
+      "grad_norm": 0.4360305190566346,
+      "learning_rate": 6.710173014221489e-06,
+      "loss": 0.3305,
+      "step": 3934
+    },
+    {
+      "epoch": 1.8802341556657307,
+      "grad_norm": 0.4928534869463769,
+      "learning_rate": 6.7051999269175405e-06,
+      "loss": 0.3335,
+      "step": 3935
+    },
+    {
+      "epoch": 1.880712024371304,
+      "grad_norm": 0.5334772505119653,
+      "learning_rate": 6.700227753533013e-06,
+      "loss": 0.3203,
+      "step": 3936
+    },
+    {
+      "epoch": 1.8811898930768771,
+      "grad_norm": 0.49357554615411325,
+      "learning_rate": 6.695256495447099e-06,
+      "loss": 0.3086,
+      "step": 3937
+    },
+    {
+      "epoch": 1.8816677617824502,
+      "grad_norm": 0.46109123255173334,
+      "learning_rate": 6.690286154038736e-06,
+      "loss": 0.3013,
+      "step": 3938
+    },
+    {
+      "epoch": 1.8821456304880235,
+      "grad_norm": 0.449128708951584,
+      "learning_rate": 6.685316730686614e-06,
+      "loss": 0.3243,
+      "step": 3939
+    },
+    {
+      "epoch": 1.8826234991935964,
+      "grad_norm": 0.46653691376464573,
+      "learning_rate": 6.680348226769162e-06,
+      "loss": 0.3299,
+      "step": 3940
+    },
+    {
+      "epoch": 1.8831013678991697,
+      "grad_norm": 0.4647819337168596,
+      "learning_rate": 6.675380643664553e-06,
+      "loss": 0.2945,
+      "step": 3941
+    },
+    {
+      "epoch": 1.8835792366047428,
+      "grad_norm": 0.4644806300385637,
+      "learning_rate": 6.670413982750709e-06,
+      "loss": 0.2997,
+      "step": 3942
+    },
+    {
+      "epoch": 1.884057105310316,
+      "grad_norm": 0.46165409315562544,
+      "learning_rate": 6.6654482454052936e-06,
+      "loss": 0.2979,
+      "step": 3943
+    },
+    {
+      "epoch": 1.8845349740158892,
+      "grad_norm": 0.4604824497984356,
+      "learning_rate": 6.660483433005714e-06,
+      "loss": 0.3179,
+      "step": 3944
+    },
+    {
+      "epoch": 1.8850128427214623,
+      "grad_norm": 0.46168391343184206,
+      "learning_rate": 6.655519546929121e-06,
+      "loss": 0.303,
+      "step": 3945
+    },
+    {
+      "epoch": 1.8854907114270354,
+      "grad_norm": 0.4678840236583196,
+      "learning_rate": 6.650556588552413e-06,
+      "loss": 0.3107,
+      "step": 3946
+    },
+    {
+      "epoch": 1.8859685801326087,
+      "grad_norm": 0.4594155584535815,
+      "learning_rate": 6.64559455925222e-06,
+      "loss": 0.3178,
+      "step": 3947
+    },
+    {
+      "epoch": 1.8864464488381816,
+      "grad_norm": 0.45341402313704277,
+      "learning_rate": 6.640633460404927e-06,
+      "loss": 0.3231,
+      "step": 3948
+    },
+    {
+      "epoch": 1.886924317543755,
+      "grad_norm": 0.5076384285562465,
+      "learning_rate": 6.635673293386656e-06,
+      "loss": 0.3197,
+      "step": 3949
+    },
+    {
+      "epoch": 1.887402186249328,
+      "grad_norm": 0.4643275428624748,
+      "learning_rate": 6.630714059573267e-06,
+      "loss": 0.3217,
+      "step": 3950
+    },
+    {
+      "epoch": 1.887880054954901,
+      "grad_norm": 0.49068283754529113,
+      "learning_rate": 6.625755760340362e-06,
+      "loss": 0.2889,
+      "step": 3951
+    },
+    {
+      "epoch": 1.8883579236604744,
+      "grad_norm": 0.47460107283399794,
+      "learning_rate": 6.620798397063291e-06,
+      "loss": 0.3271,
+      "step": 3952
+    },
+    {
+      "epoch": 1.8888357923660475,
+      "grad_norm": 0.48812107695821716,
+      "learning_rate": 6.615841971117136e-06,
+      "loss": 0.3138,
+      "step": 3953
+    },
+    {
+      "epoch": 1.8893136610716206,
+      "grad_norm": 0.5734196274573817,
+      "learning_rate": 6.610886483876721e-06,
+      "loss": 0.3044,
+      "step": 3954
+    },
+    {
+      "epoch": 1.8897915297771937,
+      "grad_norm": 0.4617227039696627,
+      "learning_rate": 6.6059319367166165e-06,
+      "loss": 0.3175,
+      "step": 3955
+    },
+    {
+      "epoch": 1.8902693984827668,
+      "grad_norm": 0.6100117271945517,
+      "learning_rate": 6.600978331011118e-06,
+      "loss": 0.3084,
+      "step": 3956
+    },
+    {
+      "epoch": 1.89074726718834,
+      "grad_norm": 0.46971941183014404,
+      "learning_rate": 6.596025668134276e-06,
+      "loss": 0.2983,
+      "step": 3957
+    },
+    {
+      "epoch": 1.8912251358939132,
+      "grad_norm": 0.461607197188674,
+      "learning_rate": 6.5910739494598675e-06,
+      "loss": 0.314,
+      "step": 3958
+    },
+    {
+      "epoch": 1.8917030045994863,
+      "grad_norm": 0.4654393182079575,
+      "learning_rate": 6.586123176361412e-06,
+      "loss": 0.3093,
+      "step": 3959
+    },
+    {
+      "epoch": 1.8921808733050596,
+      "grad_norm": 0.49908066713941035,
+      "learning_rate": 6.581173350212169e-06,
+      "loss": 0.3159,
+      "step": 3960
+    },
+    {
+      "epoch": 1.8926587420106324,
+      "grad_norm": 0.4512238380490708,
+      "learning_rate": 6.576224472385132e-06,
+      "loss": 0.3226,
+      "step": 3961
+    },
+    {
+      "epoch": 1.8931366107162058,
+      "grad_norm": 0.4496277486152005,
+      "learning_rate": 6.5712765442530305e-06,
+      "loss": 0.3107,
+      "step": 3962
+    },
+    {
+      "epoch": 1.8936144794217789,
+      "grad_norm": 0.4489167401241315,
+      "learning_rate": 6.566329567188334e-06,
+      "loss": 0.3026,
+      "step": 3963
+    },
+    {
+      "epoch": 1.894092348127352,
+      "grad_norm": 0.45699132086998806,
+      "learning_rate": 6.5613835425632475e-06,
+      "loss": 0.3145,
+      "step": 3964
+    },
+    {
+      "epoch": 1.8945702168329253,
+      "grad_norm": 0.45425664515009867,
+      "learning_rate": 6.55643847174971e-06,
+      "loss": 0.3174,
+      "step": 3965
+    },
+    {
+      "epoch": 1.8950480855384984,
+      "grad_norm": 0.500523026242252,
+      "learning_rate": 6.551494356119395e-06,
+      "loss": 0.3115,
+      "step": 3966
+    },
+    {
+      "epoch": 1.8955259542440714,
+      "grad_norm": 0.44931814929891095,
+      "learning_rate": 6.546551197043719e-06,
+      "loss": 0.3027,
+      "step": 3967
+    },
+    {
+      "epoch": 1.8960038229496445,
+      "grad_norm": 0.4626590116957062,
+      "learning_rate": 6.54160899589382e-06,
+      "loss": 0.2975,
+      "step": 3968
+    },
+    {
+      "epoch": 1.8964816916552176,
+      "grad_norm": 0.6053375872302748,
+      "learning_rate": 6.536667754040581e-06,
+      "loss": 0.3172,
+      "step": 3969
+    },
+    {
+      "epoch": 1.896959560360791,
+      "grad_norm": 0.46082649951002486,
+      "learning_rate": 6.531727472854617e-06,
+      "loss": 0.3041,
+      "step": 3970
+    },
+    {
+      "epoch": 1.897437429066364,
+      "grad_norm": 0.4647416997290529,
+      "learning_rate": 6.52678815370627e-06,
+      "loss": 0.312,
+      "step": 3971
+    },
+    {
+      "epoch": 1.8979152977719371,
+      "grad_norm": 0.8690617380399219,
+      "learning_rate": 6.521849797965623e-06,
+      "loss": 0.3076,
+      "step": 3972
+    },
+    {
+      "epoch": 1.8983931664775104,
+      "grad_norm": 0.44236729928797086,
+      "learning_rate": 6.516912407002487e-06,
+      "loss": 0.3154,
+      "step": 3973
+    },
+    {
+      "epoch": 1.8988710351830833,
+      "grad_norm": 0.48835977208241177,
+      "learning_rate": 6.511975982186412e-06,
+      "loss": 0.3152,
+      "step": 3974
+    },
+    {
+      "epoch": 1.8993489038886566,
+      "grad_norm": 0.44854340512081925,
+      "learning_rate": 6.507040524886672e-06,
+      "loss": 0.305,
+      "step": 3975
+    },
+    {
+      "epoch": 1.8998267725942297,
+      "grad_norm": 0.46881260069389524,
+      "learning_rate": 6.502106036472274e-06,
+      "loss": 0.3059,
+      "step": 3976
+    },
+    {
+      "epoch": 1.9003046412998028,
+      "grad_norm": 0.4534629224226534,
+      "learning_rate": 6.4971725183119596e-06,
+      "loss": 0.3133,
+      "step": 3977
+    },
+    {
+      "epoch": 1.9007825100053761,
+      "grad_norm": 0.4727954464086862,
+      "learning_rate": 6.492239971774201e-06,
+      "loss": 0.3112,
+      "step": 3978
+    },
+    {
+      "epoch": 1.9012603787109492,
+      "grad_norm": 0.4905488487960794,
+      "learning_rate": 6.487308398227198e-06,
+      "loss": 0.3084,
+      "step": 3979
+    },
+    {
+      "epoch": 1.9017382474165223,
+      "grad_norm": 0.45969932453025913,
+      "learning_rate": 6.482377799038882e-06,
+      "loss": 0.3074,
+      "step": 3980
+    },
+    {
+      "epoch": 1.9022161161220954,
+      "grad_norm": 0.4503140751242902,
+      "learning_rate": 6.477448175576917e-06,
+      "loss": 0.3112,
+      "step": 3981
+    },
+    {
+      "epoch": 1.9026939848276685,
+      "grad_norm": 0.5162485390503506,
+      "learning_rate": 6.472519529208688e-06,
+      "loss": 0.3116,
+      "step": 3982
+    },
+    {
+      "epoch": 1.9031718535332418,
+      "grad_norm": 0.4393300520762453,
+      "learning_rate": 6.467591861301319e-06,
+      "loss": 0.3186,
+      "step": 3983
+    },
+    {
+      "epoch": 1.903649722238815,
+      "grad_norm": 0.4404302931504677,
+      "learning_rate": 6.462665173221658e-06,
+      "loss": 0.3027,
+      "step": 3984
+    },
+    {
+      "epoch": 1.904127590944388,
+      "grad_norm": 0.4513739623364187,
+      "learning_rate": 6.45773946633628e-06,
+      "loss": 0.3191,
+      "step": 3985
+    },
+    {
+      "epoch": 1.9046054596499613,
+      "grad_norm": 0.43862099527361414,
+      "learning_rate": 6.45281474201149e-06,
+      "loss": 0.3186,
+      "step": 3986
+    },
+    {
+      "epoch": 1.9050833283555342,
+      "grad_norm": 0.48035117229015456,
+      "learning_rate": 6.44789100161332e-06,
+      "loss": 0.2902,
+      "step": 3987
+    },
+    {
+      "epoch": 1.9055611970611075,
+      "grad_norm": 0.46306152743712947,
+      "learning_rate": 6.442968246507526e-06,
+      "loss": 0.3228,
+      "step": 3988
+    },
+    {
+      "epoch": 1.9060390657666806,
+      "grad_norm": 0.4547005976928245,
+      "learning_rate": 6.438046478059597e-06,
+      "loss": 0.3081,
+      "step": 3989
+    },
+    {
+      "epoch": 1.9065169344722537,
+      "grad_norm": 0.455512028698371,
+      "learning_rate": 6.4331256976347434e-06,
+      "loss": 0.3136,
+      "step": 3990
+    },
+    {
+      "epoch": 1.906994803177827,
+      "grad_norm": 0.45184250650864355,
+      "learning_rate": 6.4282059065979e-06,
+      "loss": 0.3234,
+      "step": 3991
+    },
+    {
+      "epoch": 1.9074726718834,
+      "grad_norm": 0.4589810256657876,
+      "learning_rate": 6.423287106313734e-06,
+      "loss": 0.3139,
+      "step": 3992
+    },
+    {
+      "epoch": 1.9079505405889732,
+      "grad_norm": 0.4469761574276576,
+      "learning_rate": 6.4183692981466354e-06,
+      "loss": 0.3178,
+      "step": 3993
+    },
+    {
+      "epoch": 1.9084284092945463,
+      "grad_norm": 0.4819073592214982,
+      "learning_rate": 6.413452483460712e-06,
+      "loss": 0.3078,
+      "step": 3994
+    },
+    {
+      "epoch": 1.9089062780001194,
+      "grad_norm": 0.473783423003341,
+      "learning_rate": 6.408536663619803e-06,
+      "loss": 0.3232,
+      "step": 3995
+    },
+    {
+      "epoch": 1.9093841467056927,
+      "grad_norm": 0.46190041854892633,
+      "learning_rate": 6.403621839987475e-06,
+      "loss": 0.2947,
+      "step": 3996
+    },
+    {
+      "epoch": 1.9098620154112658,
+      "grad_norm": 0.4853262011211858,
+      "learning_rate": 6.398708013927006e-06,
+      "loss": 0.3191,
+      "step": 3997
+    },
+    {
+      "epoch": 1.9103398841168389,
+      "grad_norm": 0.46613152697547133,
+      "learning_rate": 6.393795186801408e-06,
+      "loss": 0.3148,
+      "step": 3998
+    },
+    {
+      "epoch": 1.9108177528224122,
+      "grad_norm": 0.478468525695969,
+      "learning_rate": 6.3888833599734164e-06,
+      "loss": 0.3088,
+      "step": 3999
+    },
+    {
+      "epoch": 1.911295621527985,
+      "grad_norm": 0.4807025101803059,
+      "learning_rate": 6.383972534805478e-06,
+      "loss": 0.3137,
+      "step": 4000
+    },
+    {
+      "epoch": 1.9117734902335584,
+      "grad_norm": 0.4699617352574663,
+      "learning_rate": 6.379062712659775e-06,
+      "loss": 0.3151,
+      "step": 4001
+    },
+    {
+      "epoch": 1.9122513589391315,
+      "grad_norm": 0.4432981759624267,
+      "learning_rate": 6.374153894898204e-06,
+      "loss": 0.301,
+      "step": 4002
+    },
+    {
+      "epoch": 1.9127292276447045,
+      "grad_norm": 0.4591365732275299,
+      "learning_rate": 6.369246082882381e-06,
+      "loss": 0.3183,
+      "step": 4003
+    },
+    {
+      "epoch": 1.9132070963502779,
+      "grad_norm": 0.4866953393750346,
+      "learning_rate": 6.36433927797365e-06,
+      "loss": 0.3077,
+      "step": 4004
+    },
+    {
+      "epoch": 1.913684965055851,
+      "grad_norm": 0.4474609520754289,
+      "learning_rate": 6.359433481533074e-06,
+      "loss": 0.314,
+      "step": 4005
+    },
+    {
+      "epoch": 1.914162833761424,
+      "grad_norm": 0.48773485797623917,
+      "learning_rate": 6.3545286949214245e-06,
+      "loss": 0.3329,
+      "step": 4006
+    },
+    {
+      "epoch": 1.9146407024669974,
+      "grad_norm": 0.47016604797082867,
+      "learning_rate": 6.349624919499218e-06,
+      "loss": 0.302,
+      "step": 4007
+    },
+    {
+      "epoch": 1.9151185711725702,
+      "grad_norm": 0.4760285819352319,
+      "learning_rate": 6.344722156626663e-06,
+      "loss": 0.2967,
+      "step": 4008
+    },
+    {
+      "epoch": 1.9155964398781435,
+      "grad_norm": 0.4759007373191583,
+      "learning_rate": 6.3398204076637035e-06,
+      "loss": 0.3126,
+      "step": 4009
+    },
+    {
+      "epoch": 1.9160743085837166,
+      "grad_norm": 0.45794278531506144,
+      "learning_rate": 6.3349196739700024e-06,
+      "loss": 0.3035,
+      "step": 4010
+    },
+    {
+      "epoch": 1.9165521772892897,
+      "grad_norm": 0.4692457739911771,
+      "learning_rate": 6.3300199569049305e-06,
+      "loss": 0.3215,
+      "step": 4011
+    },
+    {
+      "epoch": 1.917030045994863,
+      "grad_norm": 0.4630100738986396,
+      "learning_rate": 6.325121257827584e-06,
+      "loss": 0.2977,
+      "step": 4012
+    },
+    {
+      "epoch": 1.917507914700436,
+      "grad_norm": 0.44301606576563307,
+      "learning_rate": 6.32022357809678e-06,
+      "loss": 0.3033,
+      "step": 4013
+    },
+    {
+      "epoch": 1.9179857834060092,
+      "grad_norm": 0.4878278270456394,
+      "learning_rate": 6.3153269190710435e-06,
+      "loss": 0.3061,
+      "step": 4014
+    },
+    {
+      "epoch": 1.9184636521115823,
+      "grad_norm": 0.46093038914319545,
+      "learning_rate": 6.310431282108622e-06,
+      "loss": 0.3094,
+      "step": 4015
+    },
+    {
+      "epoch": 1.9189415208171554,
+      "grad_norm": 0.4342745654994972,
+      "learning_rate": 6.305536668567482e-06,
+      "loss": 0.2909,
+      "step": 4016
+    },
+    {
+      "epoch": 1.9194193895227287,
+      "grad_norm": 0.46848755594806835,
+      "learning_rate": 6.3006430798053e-06,
+      "loss": 0.317,
+      "step": 4017
+    },
+    {
+      "epoch": 1.9198972582283018,
+      "grad_norm": 0.45778293002207393,
+      "learning_rate": 6.295750517179471e-06,
+      "loss": 0.3253,
+      "step": 4018
+    },
+    {
+      "epoch": 1.920375126933875,
+      "grad_norm": 0.48174882670851016,
+      "learning_rate": 6.29085898204711e-06,
+      "loss": 0.2912,
+      "step": 4019
+    },
+    {
+      "epoch": 1.9208529956394482,
+      "grad_norm": 0.44572761526104454,
+      "learning_rate": 6.2859684757650365e-06,
+      "loss": 0.3104,
+      "step": 4020
+    },
+    {
+      "epoch": 1.921330864345021,
+      "grad_norm": 0.4477457574600347,
+      "learning_rate": 6.281078999689794e-06,
+      "loss": 0.3135,
+      "step": 4021
+    },
+    {
+      "epoch": 1.9218087330505944,
+      "grad_norm": 0.4612358982666258,
+      "learning_rate": 6.276190555177637e-06,
+      "loss": 0.3121,
+      "step": 4022
+    },
+    {
+      "epoch": 1.9222866017561675,
+      "grad_norm": 0.4575159149820703,
+      "learning_rate": 6.271303143584532e-06,
+      "loss": 0.3016,
+      "step": 4023
+    },
+    {
+      "epoch": 1.9227644704617406,
+      "grad_norm": 0.4482684891108602,
+      "learning_rate": 6.266416766266161e-06,
+      "loss": 0.3073,
+      "step": 4024
+    },
+    {
+      "epoch": 1.923242339167314,
+      "grad_norm": 0.4520583335071011,
+      "learning_rate": 6.261531424577923e-06,
+      "loss": 0.2965,
+      "step": 4025
+    },
+    {
+      "epoch": 1.9237202078728868,
+      "grad_norm": 0.47269951843491825,
+      "learning_rate": 6.256647119874919e-06,
+      "loss": 0.3121,
+      "step": 4026
+    },
+    {
+      "epoch": 1.92419807657846,
+      "grad_norm": 0.4758823780817171,
+      "learning_rate": 6.251763853511974e-06,
+      "loss": 0.315,
+      "step": 4027
+    },
+    {
+      "epoch": 1.9246759452840332,
+      "grad_norm": 0.4471003165029847,
+      "learning_rate": 6.24688162684362e-06,
+      "loss": 0.308,
+      "step": 4028
+    },
+    {
+      "epoch": 1.9251538139896063,
+      "grad_norm": 0.4842554500937279,
+      "learning_rate": 6.242000441224096e-06,
+      "loss": 0.3197,
+      "step": 4029
+    },
+    {
+      "epoch": 1.9256316826951796,
+      "grad_norm": 0.4766633160283862,
+      "learning_rate": 6.2371202980073596e-06,
+      "loss": 0.3098,
+      "step": 4030
+    },
+    {
+      "epoch": 1.9261095514007527,
+      "grad_norm": 0.4753968836250539,
+      "learning_rate": 6.23224119854708e-06,
+      "loss": 0.3108,
+      "step": 4031
+    },
+    {
+      "epoch": 1.9265874201063258,
+      "grad_norm": 0.5527062509603616,
+      "learning_rate": 6.227363144196625e-06,
+      "loss": 0.2904,
+      "step": 4032
+    },
+    {
+      "epoch": 1.927065288811899,
+      "grad_norm": 0.4689169847484964,
+      "learning_rate": 6.222486136309087e-06,
+      "loss": 0.2986,
+      "step": 4033
+    },
+    {
+      "epoch": 1.927543157517472,
+      "grad_norm": 0.45259176341385676,
+      "learning_rate": 6.217610176237263e-06,
+      "loss": 0.3121,
+      "step": 4034
+    },
+    {
+      "epoch": 1.9280210262230453,
+      "grad_norm": 0.45564864703780517,
+      "learning_rate": 6.212735265333655e-06,
+      "loss": 0.3016,
+      "step": 4035
+    },
+    {
+      "epoch": 1.9284988949286184,
+      "grad_norm": 0.4559492608859359,
+      "learning_rate": 6.207861404950477e-06,
+      "loss": 0.3143,
+      "step": 4036
+    },
+    {
+      "epoch": 1.9289767636341915,
+      "grad_norm": 0.4451863912439992,
+      "learning_rate": 6.2029885964396544e-06,
+      "loss": 0.3317,
+      "step": 4037
+    },
+    {
+      "epoch": 1.9294546323397648,
+      "grad_norm": 0.4535151095832485,
+      "learning_rate": 6.198116841152816e-06,
+      "loss": 0.3132,
+      "step": 4038
+    },
+    {
+      "epoch": 1.9299325010453376,
+      "grad_norm": 0.4459931426857373,
+      "learning_rate": 6.1932461404412994e-06,
+      "loss": 0.3109,
+      "step": 4039
+    },
+    {
+      "epoch": 1.930410369750911,
+      "grad_norm": 0.5319828701956664,
+      "learning_rate": 6.188376495656156e-06,
+      "loss": 0.3035,
+      "step": 4040
+    },
+    {
+      "epoch": 1.930888238456484,
+      "grad_norm": 0.46969694056011513,
+      "learning_rate": 6.183507908148137e-06,
+      "loss": 0.3163,
+      "step": 4041
+    },
+    {
+      "epoch": 1.9313661071620571,
+      "grad_norm": 0.45244458398728504,
+      "learning_rate": 6.178640379267702e-06,
+      "loss": 0.3025,
+      "step": 4042
+    },
+    {
+      "epoch": 1.9318439758676305,
+      "grad_norm": 0.45648090587965856,
+      "learning_rate": 6.173773910365018e-06,
+      "loss": 0.2901,
+      "step": 4043
+    },
+    {
+      "epoch": 1.9323218445732036,
+      "grad_norm": 0.4610212804634707,
+      "learning_rate": 6.168908502789961e-06,
+      "loss": 0.3086,
+      "step": 4044
+    },
+    {
+      "epoch": 1.9327997132787766,
+      "grad_norm": 0.46276231755271585,
+      "learning_rate": 6.164044157892102e-06,
+      "loss": 0.3272,
+      "step": 4045
+    },
+    {
+      "epoch": 1.93327758198435,
+      "grad_norm": 0.4494641795334654,
+      "learning_rate": 6.15918087702073e-06,
+      "loss": 0.3214,
+      "step": 4046
+    },
+    {
+      "epoch": 1.9337554506899228,
+      "grad_norm": 0.4495205402291461,
+      "learning_rate": 6.154318661524832e-06,
+      "loss": 0.3209,
+      "step": 4047
+    },
+    {
+      "epoch": 1.9342333193954961,
+      "grad_norm": 0.4707004184620095,
+      "learning_rate": 6.149457512753101e-06,
+      "loss": 0.3053,
+      "step": 4048
+    },
+    {
+      "epoch": 1.9347111881010692,
+      "grad_norm": 0.5357951441151997,
+      "learning_rate": 6.144597432053932e-06,
+      "loss": 0.3119,
+      "step": 4049
+    },
+    {
+      "epoch": 1.9351890568066423,
+      "grad_norm": 0.4535876834445935,
+      "learning_rate": 6.13973842077543e-06,
+      "loss": 0.3023,
+      "step": 4050
+    },
+    {
+      "epoch": 1.9356669255122156,
+      "grad_norm": 0.48376924175039243,
+      "learning_rate": 6.134880480265393e-06,
+      "loss": 0.3059,
+      "step": 4051
+    },
+    {
+      "epoch": 1.9361447942177885,
+      "grad_norm": 0.49754708572204154,
+      "learning_rate": 6.130023611871332e-06,
+      "loss": 0.3139,
+      "step": 4052
+    },
+    {
+      "epoch": 1.9366226629233618,
+      "grad_norm": 0.6147957148377768,
+      "learning_rate": 6.1251678169404585e-06,
+      "loss": 0.3017,
+      "step": 4053
+    },
+    {
+      "epoch": 1.937100531628935,
+      "grad_norm": 0.4597512390463644,
+      "learning_rate": 6.120313096819679e-06,
+      "loss": 0.3184,
+      "step": 4054
+    },
+    {
+      "epoch": 1.937578400334508,
+      "grad_norm": 0.4559692746363391,
+      "learning_rate": 6.11545945285561e-06,
+      "loss": 0.3085,
+      "step": 4055
+    },
+    {
+      "epoch": 1.9380562690400813,
+      "grad_norm": 0.47922658966713283,
+      "learning_rate": 6.110606886394568e-06,
+      "loss": 0.3254,
+      "step": 4056
+    },
+    {
+      "epoch": 1.9385341377456544,
+      "grad_norm": 0.4534671409403072,
+      "learning_rate": 6.105755398782567e-06,
+      "loss": 0.324,
+      "step": 4057
+    },
+    {
+      "epoch": 1.9390120064512275,
+      "grad_norm": 0.4545090220212905,
+      "learning_rate": 6.100904991365324e-06,
+      "loss": 0.3079,
+      "step": 4058
+    },
+    {
+      "epoch": 1.9394898751568008,
+      "grad_norm": 0.4762232060665426,
+      "learning_rate": 6.096055665488261e-06,
+      "loss": 0.3051,
+      "step": 4059
+    },
+    {
+      "epoch": 1.9399677438623737,
+      "grad_norm": 0.5066682145376573,
+      "learning_rate": 6.091207422496489e-06,
+      "loss": 0.3159,
+      "step": 4060
+    },
+    {
+      "epoch": 1.940445612567947,
+      "grad_norm": 0.44386055788270473,
+      "learning_rate": 6.0863602637348284e-06,
+      "loss": 0.3119,
+      "step": 4061
+    },
+    {
+      "epoch": 1.94092348127352,
+      "grad_norm": 0.43903421860827463,
+      "learning_rate": 6.081514190547797e-06,
+      "loss": 0.3078,
+      "step": 4062
+    },
+    {
+      "epoch": 1.9414013499790932,
+      "grad_norm": 0.47742457334875266,
+      "learning_rate": 6.076669204279606e-06,
+      "loss": 0.3176,
+      "step": 4063
+    },
+    {
+      "epoch": 1.9418792186846665,
+      "grad_norm": 0.4665930593611481,
+      "learning_rate": 6.071825306274173e-06,
+      "loss": 0.3036,
+      "step": 4064
+    },
+    {
+      "epoch": 1.9423570873902394,
+      "grad_norm": 0.5253515300391597,
+      "learning_rate": 6.066982497875109e-06,
+      "loss": 0.3198,
+      "step": 4065
+    },
+    {
+      "epoch": 1.9428349560958127,
+      "grad_norm": 0.5049405090302596,
+      "learning_rate": 6.0621407804257205e-06,
+      "loss": 0.3048,
+      "step": 4066
+    },
+    {
+      "epoch": 1.9433128248013858,
+      "grad_norm": 0.4700459384897942,
+      "learning_rate": 6.057300155269017e-06,
+      "loss": 0.3106,
+      "step": 4067
+    },
+    {
+      "epoch": 1.9437906935069589,
+      "grad_norm": 0.4434734524599051,
+      "learning_rate": 6.052460623747705e-06,
+      "loss": 0.3237,
+      "step": 4068
+    },
+    {
+      "epoch": 1.9442685622125322,
+      "grad_norm": 0.46014030921295707,
+      "learning_rate": 6.0476221872041794e-06,
+      "loss": 0.3223,
+      "step": 4069
+    },
+    {
+      "epoch": 1.9447464309181053,
+      "grad_norm": 0.468750131948572,
+      "learning_rate": 6.042784846980542e-06,
+      "loss": 0.2991,
+      "step": 4070
+    },
+    {
+      "epoch": 1.9452242996236784,
+      "grad_norm": 0.47243399384340334,
+      "learning_rate": 6.037948604418584e-06,
+      "loss": 0.3229,
+      "step": 4071
+    },
+    {
+      "epoch": 1.9457021683292517,
+      "grad_norm": 0.48404534391010834,
+      "learning_rate": 6.033113460859794e-06,
+      "loss": 0.3042,
+      "step": 4072
+    },
+    {
+      "epoch": 1.9461800370348246,
+      "grad_norm": 0.459036713881152,
+      "learning_rate": 6.028279417645351e-06,
+      "loss": 0.3125,
+      "step": 4073
+    },
+    {
+      "epoch": 1.9466579057403979,
+      "grad_norm": 0.7112099246861842,
+      "learning_rate": 6.023446476116141e-06,
+      "loss": 0.3112,
+      "step": 4074
+    },
+    {
+      "epoch": 1.947135774445971,
+      "grad_norm": 0.4643938545098871,
+      "learning_rate": 6.018614637612733e-06,
+      "loss": 0.2963,
+      "step": 4075
+    },
+    {
+      "epoch": 1.947613643151544,
+      "grad_norm": 0.44678142004677796,
+      "learning_rate": 6.013783903475396e-06,
+      "loss": 0.3037,
+      "step": 4076
+    },
+    {
+      "epoch": 1.9480915118571174,
+      "grad_norm": 0.5656950978702013,
+      "learning_rate": 6.008954275044088e-06,
+      "loss": 0.2962,
+      "step": 4077
+    },
+    {
+      "epoch": 1.9485693805626902,
+      "grad_norm": 0.47423542716615724,
+      "learning_rate": 6.004125753658461e-06,
+      "loss": 0.3106,
+      "step": 4078
+    },
+    {
+      "epoch": 1.9490472492682636,
+      "grad_norm": 0.45972148442045957,
+      "learning_rate": 5.9992983406578666e-06,
+      "loss": 0.3016,
+      "step": 4079
+    },
+    {
+      "epoch": 1.9495251179738367,
+      "grad_norm": 0.48436905917421313,
+      "learning_rate": 5.99447203738134e-06,
+      "loss": 0.3036,
+      "step": 4080
+    },
+    {
+      "epoch": 1.9500029866794097,
+      "grad_norm": 0.44553011226551475,
+      "learning_rate": 5.989646845167614e-06,
+      "loss": 0.2932,
+      "step": 4081
+    },
+    {
+      "epoch": 1.950480855384983,
+      "grad_norm": 0.4607346503392114,
+      "learning_rate": 5.984822765355113e-06,
+      "loss": 0.3358,
+      "step": 4082
+    },
+    {
+      "epoch": 1.9509587240905562,
+      "grad_norm": 0.4952981445916362,
+      "learning_rate": 5.979999799281948e-06,
+      "loss": 0.3151,
+      "step": 4083
+    },
+    {
+      "epoch": 1.9514365927961292,
+      "grad_norm": 0.4675121455424015,
+      "learning_rate": 5.975177948285929e-06,
+      "loss": 0.3148,
+      "step": 4084
+    },
+    {
+      "epoch": 1.9519144615017026,
+      "grad_norm": 0.45369885707604624,
+      "learning_rate": 5.9703572137045495e-06,
+      "loss": 0.3034,
+      "step": 4085
+    },
+    {
+      "epoch": 1.9523923302072754,
+      "grad_norm": 0.4579342776702539,
+      "learning_rate": 5.965537596874997e-06,
+      "loss": 0.313,
+      "step": 4086
+    },
+    {
+      "epoch": 1.9528701989128487,
+      "grad_norm": 0.6942835585119591,
+      "learning_rate": 5.960719099134149e-06,
+      "loss": 0.317,
+      "step": 4087
+    },
+    {
+      "epoch": 1.9533480676184218,
+      "grad_norm": 0.46918542083573794,
+      "learning_rate": 5.9559017218185724e-06,
+      "loss": 0.309,
+      "step": 4088
+    },
+    {
+      "epoch": 1.953825936323995,
+      "grad_norm": 0.45277876411651685,
+      "learning_rate": 5.951085466264519e-06,
+      "loss": 0.3032,
+      "step": 4089
+    },
+    {
+      "epoch": 1.9543038050295682,
+      "grad_norm": 0.4690874120536212,
+      "learning_rate": 5.946270333807937e-06,
+      "loss": 0.3039,
+      "step": 4090
+    },
+    {
+      "epoch": 1.9547816737351411,
+      "grad_norm": 0.4507823330535709,
+      "learning_rate": 5.94145632578446e-06,
+      "loss": 0.3077,
+      "step": 4091
+    },
+    {
+      "epoch": 1.9552595424407144,
+      "grad_norm": 0.44925892593784034,
+      "learning_rate": 5.9366434435294026e-06,
+      "loss": 0.3155,
+      "step": 4092
+    },
+    {
+      "epoch": 1.9557374111462875,
+      "grad_norm": 0.48240656700569656,
+      "learning_rate": 5.9318316883777795e-06,
+      "loss": 0.3208,
+      "step": 4093
+    },
+    {
+      "epoch": 1.9562152798518606,
+      "grad_norm": 0.4604534040175604,
+      "learning_rate": 5.927021061664287e-06,
+      "loss": 0.2998,
+      "step": 4094
+    },
+    {
+      "epoch": 1.956693148557434,
+      "grad_norm": 0.4526034134761243,
+      "learning_rate": 5.922211564723302e-06,
+      "loss": 0.2946,
+      "step": 4095
+    },
+    {
+      "epoch": 1.957171017263007,
+      "grad_norm": 0.46416284670050006,
+      "learning_rate": 5.9174031988888995e-06,
+      "loss": 0.3118,
+      "step": 4096
+    },
+    {
+      "epoch": 1.95764888596858,
+      "grad_norm": 0.45215047247388573,
+      "learning_rate": 5.912595965494835e-06,
+      "loss": 0.3182,
+      "step": 4097
+    },
+    {
+      "epoch": 1.9581267546741534,
+      "grad_norm": 0.5546875852085521,
+      "learning_rate": 5.907789865874547e-06,
+      "loss": 0.3064,
+      "step": 4098
+    },
+    {
+      "epoch": 1.9586046233797263,
+      "grad_norm": 0.4752323586160744,
+      "learning_rate": 5.902984901361166e-06,
+      "loss": 0.3278,
+      "step": 4099
+    },
+    {
+      "epoch": 1.9590824920852996,
+      "grad_norm": 0.452399975076052,
+      "learning_rate": 5.898181073287504e-06,
+      "loss": 0.3028,
+      "step": 4100
+    },
+    {
+      "epoch": 1.9595603607908727,
+      "grad_norm": 0.4578845518507173,
+      "learning_rate": 5.893378382986057e-06,
+      "loss": 0.2949,
+      "step": 4101
+    },
+    {
+      "epoch": 1.9600382294964458,
+      "grad_norm": 0.4390145249730691,
+      "learning_rate": 5.8885768317890054e-06,
+      "loss": 0.3016,
+      "step": 4102
+    },
+    {
+      "epoch": 1.960516098202019,
+      "grad_norm": 0.9622036180396034,
+      "learning_rate": 5.883776421028219e-06,
+      "loss": 0.3221,
+      "step": 4103
+    },
+    {
+      "epoch": 1.960993966907592,
+      "grad_norm": 0.45023967591580155,
+      "learning_rate": 5.878977152035243e-06,
+      "loss": 0.3117,
+      "step": 4104
+    },
+    {
+      "epoch": 1.9614718356131653,
+      "grad_norm": 0.44945352983330045,
+      "learning_rate": 5.87417902614131e-06,
+      "loss": 0.2991,
+      "step": 4105
+    },
+    {
+      "epoch": 1.9619497043187384,
+      "grad_norm": 0.4507418181158918,
+      "learning_rate": 5.869382044677341e-06,
+      "loss": 0.3181,
+      "step": 4106
+    },
+    {
+      "epoch": 1.9624275730243115,
+      "grad_norm": 0.44350954473838944,
+      "learning_rate": 5.8645862089739215e-06,
+      "loss": 0.3151,
+      "step": 4107
+    },
+    {
+      "epoch": 1.9629054417298848,
+      "grad_norm": 0.5341361831039657,
+      "learning_rate": 5.859791520361348e-06,
+      "loss": 0.3072,
+      "step": 4108
+    },
+    {
+      "epoch": 1.9633833104354579,
+      "grad_norm": 0.46709165414809023,
+      "learning_rate": 5.854997980169572e-06,
+      "loss": 0.3031,
+      "step": 4109
+    },
+    {
+      "epoch": 1.963861179141031,
+      "grad_norm": 0.4280650147245011,
+      "learning_rate": 5.850205589728239e-06,
+      "loss": 0.3127,
+      "step": 4110
+    },
+    {
+      "epoch": 1.9643390478466043,
+      "grad_norm": 0.44278861997127844,
+      "learning_rate": 5.845414350366679e-06,
+      "loss": 0.3177,
+      "step": 4111
+    },
+    {
+      "epoch": 1.9648169165521772,
+      "grad_norm": 0.43260620835576497,
+      "learning_rate": 5.8406242634138875e-06,
+      "loss": 0.3133,
+      "step": 4112
+    },
+    {
+      "epoch": 1.9652947852577505,
+      "grad_norm": 0.46599928724656164,
+      "learning_rate": 5.835835330198558e-06,
+      "loss": 0.317,
+      "step": 4113
+    },
+    {
+      "epoch": 1.9657726539633236,
+      "grad_norm": 0.4868101146764809,
+      "learning_rate": 5.83104755204905e-06,
+      "loss": 0.3072,
+      "step": 4114
+    },
+    {
+      "epoch": 1.9662505226688967,
+      "grad_norm": 0.43340141877422556,
+      "learning_rate": 5.826260930293417e-06,
+      "loss": 0.3041,
+      "step": 4115
+    },
+    {
+      "epoch": 1.96672839137447,
+      "grad_norm": 0.4737663310328414,
+      "learning_rate": 5.8214754662593765e-06,
+      "loss": 0.3196,
+      "step": 4116
+    },
+    {
+      "epoch": 1.9672062600800428,
+      "grad_norm": 0.4510243883089674,
+      "learning_rate": 5.81669116127433e-06,
+      "loss": 0.2846,
+      "step": 4117
+    },
+    {
+      "epoch": 1.9676841287856162,
+      "grad_norm": 0.4721590552134255,
+      "learning_rate": 5.811908016665369e-06,
+      "loss": 0.2991,
+      "step": 4118
+    },
+    {
+      "epoch": 1.9681619974911893,
+      "grad_norm": 0.4485701829219417,
+      "learning_rate": 5.807126033759245e-06,
+      "loss": 0.3201,
+      "step": 4119
+    },
+    {
+      "epoch": 1.9686398661967623,
+      "grad_norm": 0.42971050420170925,
+      "learning_rate": 5.802345213882396e-06,
+      "loss": 0.2848,
+      "step": 4120
+    },
+    {
+      "epoch": 1.9691177349023357,
+      "grad_norm": 0.5060676790854636,
+      "learning_rate": 5.797565558360943e-06,
+      "loss": 0.3189,
+      "step": 4121
+    },
+    {
+      "epoch": 1.9695956036079088,
+      "grad_norm": 0.5368619185936723,
+      "learning_rate": 5.792787068520674e-06,
+      "loss": 0.3272,
+      "step": 4122
+    },
+    {
+      "epoch": 1.9700734723134818,
+      "grad_norm": 0.4969544005475237,
+      "learning_rate": 5.788009745687053e-06,
+      "loss": 0.3044,
+      "step": 4123
+    },
+    {
+      "epoch": 1.9705513410190552,
+      "grad_norm": 0.5192317905793113,
+      "learning_rate": 5.7832335911852355e-06,
+      "loss": 0.3073,
+      "step": 4124
+    },
+    {
+      "epoch": 1.971029209724628,
+      "grad_norm": 0.4584957853344697,
+      "learning_rate": 5.778458606340037e-06,
+      "loss": 0.3182,
+      "step": 4125
+    },
+    {
+      "epoch": 1.9715070784302013,
+      "grad_norm": 0.4826482249055619,
+      "learning_rate": 5.7736847924759505e-06,
+      "loss": 0.2936,
+      "step": 4126
+    },
+    {
+      "epoch": 1.9719849471357744,
+      "grad_norm": 0.4565375353417649,
+      "learning_rate": 5.7689121509171564e-06,
+      "loss": 0.3002,
+      "step": 4127
+    },
+    {
+      "epoch": 1.9724628158413475,
+      "grad_norm": 0.4524472919502672,
+      "learning_rate": 5.764140682987496e-06,
+      "loss": 0.3091,
+      "step": 4128
+    },
+    {
+      "epoch": 1.9729406845469208,
+      "grad_norm": 0.47905618351860524,
+      "learning_rate": 5.759370390010487e-06,
+      "loss": 0.3123,
+      "step": 4129
+    },
+    {
+      "epoch": 1.9734185532524937,
+      "grad_norm": 0.4588694980027561,
+      "learning_rate": 5.754601273309333e-06,
+      "loss": 0.3194,
+      "step": 4130
+    },
+    {
+      "epoch": 1.973896421958067,
+      "grad_norm": 0.45707604771801996,
+      "learning_rate": 5.749833334206897e-06,
+      "loss": 0.3094,
+      "step": 4131
+    },
+    {
+      "epoch": 1.9743742906636401,
+      "grad_norm": 0.4587355771749119,
+      "learning_rate": 5.745066574025718e-06,
+      "loss": 0.302,
+      "step": 4132
+    },
+    {
+      "epoch": 1.9748521593692132,
+      "grad_norm": 0.4740061736075553,
+      "learning_rate": 5.740300994088022e-06,
+      "loss": 0.3304,
+      "step": 4133
+    },
+    {
+      "epoch": 1.9753300280747865,
+      "grad_norm": 0.4566089421929211,
+      "learning_rate": 5.735536595715687e-06,
+      "loss": 0.3113,
+      "step": 4134
+    },
+    {
+      "epoch": 1.9758078967803596,
+      "grad_norm": 0.4663581459744624,
+      "learning_rate": 5.730773380230276e-06,
+      "loss": 0.2976,
+      "step": 4135
+    },
+    {
+      "epoch": 1.9762857654859327,
+      "grad_norm": 0.44354111986279265,
+      "learning_rate": 5.726011348953023e-06,
+      "loss": 0.303,
+      "step": 4136
+    },
+    {
+      "epoch": 1.976763634191506,
+      "grad_norm": 0.43956310394463377,
+      "learning_rate": 5.7212505032048315e-06,
+      "loss": 0.3212,
+      "step": 4137
+    },
+    {
+      "epoch": 1.977241502897079,
+      "grad_norm": 0.43772616497346545,
+      "learning_rate": 5.716490844306271e-06,
+      "loss": 0.3071,
+      "step": 4138
+    },
+    {
+      "epoch": 1.9777193716026522,
+      "grad_norm": 0.5091670239004066,
+      "learning_rate": 5.711732373577592e-06,
+      "loss": 0.3224,
+      "step": 4139
+    },
+    {
+      "epoch": 1.9781972403082253,
+      "grad_norm": 0.47000699212819913,
+      "learning_rate": 5.70697509233871e-06,
+      "loss": 0.29,
+      "step": 4140
+    },
+    {
+      "epoch": 1.9786751090137984,
+      "grad_norm": 0.4704139528256369,
+      "learning_rate": 5.702219001909206e-06,
+      "loss": 0.2997,
+      "step": 4141
+    },
+    {
+      "epoch": 1.9791529777193717,
+      "grad_norm": 0.4614015610752338,
+      "learning_rate": 5.697464103608339e-06,
+      "loss": 0.3103,
+      "step": 4142
+    },
+    {
+      "epoch": 1.9796308464249448,
+      "grad_norm": 0.45164231657830634,
+      "learning_rate": 5.692710398755039e-06,
+      "loss": 0.3092,
+      "step": 4143
+    },
+    {
+      "epoch": 1.980108715130518,
+      "grad_norm": 0.47238767638550566,
+      "learning_rate": 5.687957888667894e-06,
+      "loss": 0.3087,
+      "step": 4144
+    },
+    {
+      "epoch": 1.980586583836091,
+      "grad_norm": 0.4513243311661097,
+      "learning_rate": 5.683206574665165e-06,
+      "loss": 0.3104,
+      "step": 4145
+    },
+    {
+      "epoch": 1.981064452541664,
+      "grad_norm": 0.4721048544965712,
+      "learning_rate": 5.678456458064788e-06,
+      "loss": 0.306,
+      "step": 4146
+    },
+    {
+      "epoch": 1.9815423212472374,
+      "grad_norm": 0.5125862122325965,
+      "learning_rate": 5.673707540184359e-06,
+      "loss": 0.2986,
+      "step": 4147
+    },
+    {
+      "epoch": 1.9820201899528105,
+      "grad_norm": 0.4389466347186533,
+      "learning_rate": 5.66895982234114e-06,
+      "loss": 0.3081,
+      "step": 4148
+    },
+    {
+      "epoch": 1.9824980586583836,
+      "grad_norm": 0.4407487284192881,
+      "learning_rate": 5.664213305852073e-06,
+      "loss": 0.3238,
+      "step": 4149
+    },
+    {
+      "epoch": 1.982975927363957,
+      "grad_norm": 0.47364655963546964,
+      "learning_rate": 5.6594679920337514e-06,
+      "loss": 0.3118,
+      "step": 4150
+    },
+    {
+      "epoch": 1.9834537960695298,
+      "grad_norm": 0.4304623563483301,
+      "learning_rate": 5.6547238822024395e-06,
+      "loss": 0.3229,
+      "step": 4151
+    },
+    {
+      "epoch": 1.983931664775103,
+      "grad_norm": 0.45174215596394307,
+      "learning_rate": 5.649980977674079e-06,
+      "loss": 0.3041,
+      "step": 4152
+    },
+    {
+      "epoch": 1.9844095334806762,
+      "grad_norm": 0.45178499648893194,
+      "learning_rate": 5.6452392797642605e-06,
+      "loss": 0.3035,
+      "step": 4153
+    },
+    {
+      "epoch": 1.9848874021862493,
+      "grad_norm": 0.4521203750930736,
+      "learning_rate": 5.640498789788246e-06,
+      "loss": 0.3094,
+      "step": 4154
+    },
+    {
+      "epoch": 1.9853652708918226,
+      "grad_norm": 0.47602827989218915,
+      "learning_rate": 5.635759509060969e-06,
+      "loss": 0.3116,
+      "step": 4155
+    },
+    {
+      "epoch": 1.9858431395973957,
+      "grad_norm": 0.4801852786233523,
+      "learning_rate": 5.631021438897023e-06,
+      "loss": 0.3241,
+      "step": 4156
+    },
+    {
+      "epoch": 1.9863210083029688,
+      "grad_norm": 0.44827145290236947,
+      "learning_rate": 5.626284580610657e-06,
+      "loss": 0.3137,
+      "step": 4157
+    },
+    {
+      "epoch": 1.9867988770085419,
+      "grad_norm": 0.520461667707988,
+      "learning_rate": 5.621548935515801e-06,
+      "loss": 0.3203,
+      "step": 4158
+    },
+    {
+      "epoch": 1.987276745714115,
+      "grad_norm": 0.4662181204209844,
+      "learning_rate": 5.616814504926037e-06,
+      "loss": 0.3011,
+      "step": 4159
+    },
+    {
+      "epoch": 1.9877546144196883,
+      "grad_norm": 0.4531917716602397,
+      "learning_rate": 5.612081290154607e-06,
+      "loss": 0.3324,
+      "step": 4160
+    },
+    {
+      "epoch": 1.9882324831252614,
+      "grad_norm": 0.47179765398325363,
+      "learning_rate": 5.607349292514429e-06,
+      "loss": 0.3167,
+      "step": 4161
+    },
+    {
+      "epoch": 1.9887103518308344,
+      "grad_norm": 0.7180803991243624,
+      "learning_rate": 5.602618513318072e-06,
+      "loss": 0.3052,
+      "step": 4162
+    },
+    {
+      "epoch": 1.9891882205364078,
+      "grad_norm": 0.4606351292564438,
+      "learning_rate": 5.597888953877768e-06,
+      "loss": 0.3148,
+      "step": 4163
+    },
+    {
+      "epoch": 1.9896660892419806,
+      "grad_norm": 0.44908296173819723,
+      "learning_rate": 5.5931606155054195e-06,
+      "loss": 0.3084,
+      "step": 4164
+    },
+    {
+      "epoch": 1.990143957947554,
+      "grad_norm": 0.5095275810168982,
+      "learning_rate": 5.58843349951258e-06,
+      "loss": 0.2959,
+      "step": 4165
+    },
+    {
+      "epoch": 1.990621826653127,
+      "grad_norm": 0.46103156732406936,
+      "learning_rate": 5.583707607210467e-06,
+      "loss": 0.3,
+      "step": 4166
+    },
+    {
+      "epoch": 1.9910996953587001,
+      "grad_norm": 0.45327848630075884,
+      "learning_rate": 5.578982939909965e-06,
+      "loss": 0.3107,
+      "step": 4167
+    },
+    {
+      "epoch": 1.9915775640642734,
+      "grad_norm": 0.47076365416712834,
+      "learning_rate": 5.574259498921608e-06,
+      "loss": 0.292,
+      "step": 4168
+    },
+    {
+      "epoch": 1.9920554327698465,
+      "grad_norm": 0.4667403906050238,
+      "learning_rate": 5.569537285555596e-06,
+      "loss": 0.3196,
+      "step": 4169
+    },
+    {
+      "epoch": 1.9925333014754196,
+      "grad_norm": 0.4926455741619648,
+      "learning_rate": 5.564816301121792e-06,
+      "loss": 0.3176,
+      "step": 4170
+    },
+    {
+      "epoch": 1.9930111701809927,
+      "grad_norm": 0.4940948150038382,
+      "learning_rate": 5.5600965469297105e-06,
+      "loss": 0.3132,
+      "step": 4171
+    },
+    {
+      "epoch": 1.9934890388865658,
+      "grad_norm": 0.4516008067632424,
+      "learning_rate": 5.555378024288525e-06,
+      "loss": 0.3259,
+      "step": 4172
+    },
+    {
+      "epoch": 1.9939669075921391,
+      "grad_norm": 0.45739246251961,
+      "learning_rate": 5.550660734507077e-06,
+      "loss": 0.3084,
+      "step": 4173
+    },
+    {
+      "epoch": 1.9944447762977122,
+      "grad_norm": 0.4526553226283521,
+      "learning_rate": 5.545944678893853e-06,
+      "loss": 0.3158,
+      "step": 4174
+    },
+    {
+      "epoch": 1.9949226450032853,
+      "grad_norm": 0.4659888379257696,
+      "learning_rate": 5.541229858757011e-06,
+      "loss": 0.2973,
+      "step": 4175
+    },
+    {
+      "epoch": 1.9954005137088586,
+      "grad_norm": 0.45374769534588466,
+      "learning_rate": 5.53651627540435e-06,
+      "loss": 0.3078,
+      "step": 4176
+    },
+    {
+      "epoch": 1.9958783824144315,
+      "grad_norm": 0.47368493531667,
+      "learning_rate": 5.531803930143345e-06,
+      "loss": 0.3152,
+      "step": 4177
+    },
+    {
+      "epoch": 1.9963562511200048,
+      "grad_norm": 0.470660248933249,
+      "learning_rate": 5.527092824281111e-06,
+      "loss": 0.2873,
+      "step": 4178
+    },
+    {
+      "epoch": 1.996834119825578,
+      "grad_norm": 0.4557800815216673,
+      "learning_rate": 5.522382959124422e-06,
+      "loss": 0.2972,
+      "step": 4179
+    },
+    {
+      "epoch": 1.997311988531151,
+      "grad_norm": 0.4748327062670451,
+      "learning_rate": 5.517674335979721e-06,
+      "loss": 0.2962,
+      "step": 4180
+    },
+    {
+      "epoch": 1.9977898572367243,
+      "grad_norm": 0.4718260147834405,
+      "learning_rate": 5.512966956153093e-06,
+      "loss": 0.3072,
+      "step": 4181
+    },
+    {
+      "epoch": 1.9982677259422974,
+      "grad_norm": 0.4766665255326838,
+      "learning_rate": 5.508260820950278e-06,
+      "loss": 0.3006,
+      "step": 4182
+    },
+    {
+      "epoch": 1.9987455946478705,
+      "grad_norm": 0.43482332106111293,
+      "learning_rate": 5.503555931676681e-06,
+      "loss": 0.3002,
+      "step": 4183
+    },
+    {
+      "epoch": 1.9992234633534436,
+      "grad_norm": 0.5879689150189751,
+      "learning_rate": 5.498852289637353e-06,
+      "loss": 0.3064,
+      "step": 4184
+    },
+    {
+      "epoch": 1.9997013320590167,
+      "grad_norm": 0.6039537333215967,
+      "learning_rate": 5.494149896136998e-06,
+      "loss": 0.2986,
+      "step": 4185
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.507643331231026,
+      "learning_rate": 5.489448752479982e-06,
+      "loss": 0.3016,
+      "step": 4186
+    },
+    {
+      "epoch": 2.0004778687055733,
+      "grad_norm": 0.5557131392016892,
+      "learning_rate": 5.484748859970319e-06,
+      "loss": 0.2706,
+      "step": 4187
+    },
+    {
+      "epoch": 2.000955737411146,
+      "grad_norm": 0.5129666479233886,
+      "learning_rate": 5.48005021991167e-06,
+      "loss": 0.2874,
+      "step": 4188
+    },
+    {
+      "epoch": 2.0014336061167195,
+      "grad_norm": 0.48363044424144763,
+      "learning_rate": 5.475352833607363e-06,
+      "loss": 0.2666,
+      "step": 4189
+    },
+    {
+      "epoch": 2.0019114748222924,
+      "grad_norm": 0.4571679864809817,
+      "learning_rate": 5.470656702360367e-06,
+      "loss": 0.2944,
+      "step": 4190
+    },
+    {
+      "epoch": 2.0023893435278657,
+      "grad_norm": 0.49923351769769775,
+      "learning_rate": 5.4659618274733e-06,
+      "loss": 0.2708,
+      "step": 4191
+    },
+    {
+      "epoch": 2.002867212233439,
+      "grad_norm": 0.5726085889940755,
+      "learning_rate": 5.461268210248449e-06,
+      "loss": 0.2778,
+      "step": 4192
+    },
+    {
+      "epoch": 2.003345080939012,
+      "grad_norm": 0.5151146352983149,
+      "learning_rate": 5.4565758519877354e-06,
+      "loss": 0.2768,
+      "step": 4193
+    },
+    {
+      "epoch": 2.003822949644585,
+      "grad_norm": 0.5408118208195342,
+      "learning_rate": 5.45188475399273e-06,
+      "loss": 0.2789,
+      "step": 4194
+    },
+    {
+      "epoch": 2.0043008183501585,
+      "grad_norm": 0.5218303636526377,
+      "learning_rate": 5.447194917564671e-06,
+      "loss": 0.283,
+      "step": 4195
+    },
+    {
+      "epoch": 2.0047786870557314,
+      "grad_norm": 0.48975172353959573,
+      "learning_rate": 5.442506344004433e-06,
+      "loss": 0.2583,
+      "step": 4196
+    },
+    {
+      "epoch": 2.0052565557613047,
+      "grad_norm": 0.48789372594335206,
+      "learning_rate": 5.437819034612536e-06,
+      "loss": 0.2891,
+      "step": 4197
+    },
+    {
+      "epoch": 2.0057344244668776,
+      "grad_norm": 0.5173557034209996,
+      "learning_rate": 5.433132990689168e-06,
+      "loss": 0.2718,
+      "step": 4198
+    },
+    {
+      "epoch": 2.006212293172451,
+      "grad_norm": 0.4979488229723115,
+      "learning_rate": 5.42844821353415e-06,
+      "loss": 0.2763,
+      "step": 4199
+    },
+    {
+      "epoch": 2.006690161878024,
+      "grad_norm": 0.46766644811447605,
+      "learning_rate": 5.423764704446954e-06,
+      "loss": 0.2718,
+      "step": 4200
+    },
+    {
+      "epoch": 2.007168030583597,
+      "grad_norm": 0.4858229963672474,
+      "learning_rate": 5.41908246472671e-06,
+      "loss": 0.2717,
+      "step": 4201
+    },
+    {
+      "epoch": 2.0076458992891704,
+      "grad_norm": 0.4871157460335423,
+      "learning_rate": 5.414401495672183e-06,
+      "loss": 0.2479,
+      "step": 4202
+    },
+    {
+      "epoch": 2.0081237679947432,
+      "grad_norm": 0.4719329157303939,
+      "learning_rate": 5.4097217985817885e-06,
+      "loss": 0.2707,
+      "step": 4203
+    },
+    {
+      "epoch": 2.0086016367003166,
+      "grad_norm": 0.4784620215856789,
+      "learning_rate": 5.4050433747536e-06,
+      "loss": 0.2793,
+      "step": 4204
+    },
+    {
+      "epoch": 2.00907950540589,
+      "grad_norm": 0.4480022553241313,
+      "learning_rate": 5.400366225485326e-06,
+      "loss": 0.2551,
+      "step": 4205
+    },
+    {
+      "epoch": 2.0095573741114627,
+      "grad_norm": 0.4862355352173036,
+      "learning_rate": 5.395690352074321e-06,
+      "loss": 0.277,
+      "step": 4206
+    },
+    {
+      "epoch": 2.010035242817036,
+      "grad_norm": 0.4832048787344954,
+      "learning_rate": 5.391015755817597e-06,
+      "loss": 0.2741,
+      "step": 4207
+    },
+    {
+      "epoch": 2.0105131115226094,
+      "grad_norm": 0.5260287493937732,
+      "learning_rate": 5.386342438011798e-06,
+      "loss": 0.2737,
+      "step": 4208
+    },
+    {
+      "epoch": 2.0109909802281822,
+      "grad_norm": 0.46713974738731096,
+      "learning_rate": 5.3816703999532225e-06,
+      "loss": 0.2707,
+      "step": 4209
+    },
+    {
+      "epoch": 2.0114688489337555,
+      "grad_norm": 0.48820679807463374,
+      "learning_rate": 5.376999642937817e-06,
+      "loss": 0.271,
+      "step": 4210
+    },
+    {
+      "epoch": 2.0119467176393284,
+      "grad_norm": 0.4546420853003914,
+      "learning_rate": 5.372330168261162e-06,
+      "loss": 0.266,
+      "step": 4211
+    },
+    {
+      "epoch": 2.0124245863449017,
+      "grad_norm": 0.5092550048537878,
+      "learning_rate": 5.367661977218484e-06,
+      "loss": 0.2718,
+      "step": 4212
+    },
+    {
+      "epoch": 2.012902455050475,
+      "grad_norm": 0.6128690560615392,
+      "learning_rate": 5.362995071104664e-06,
+      "loss": 0.2576,
+      "step": 4213
+    },
+    {
+      "epoch": 2.013380323756048,
+      "grad_norm": 0.45952890722819323,
+      "learning_rate": 5.358329451214215e-06,
+      "loss": 0.2868,
+      "step": 4214
+    },
+    {
+      "epoch": 2.0138581924616212,
+      "grad_norm": 0.48006938509230707,
+      "learning_rate": 5.353665118841296e-06,
+      "loss": 0.2816,
+      "step": 4215
+    },
+    {
+      "epoch": 2.014336061167194,
+      "grad_norm": 0.45754420825780606,
+      "learning_rate": 5.3490020752797165e-06,
+      "loss": 0.2846,
+      "step": 4216
+    },
+    {
+      "epoch": 2.0148139298727674,
+      "grad_norm": 0.48929584524947306,
+      "learning_rate": 5.344340321822919e-06,
+      "loss": 0.2877,
+      "step": 4217
+    },
+    {
+      "epoch": 2.0152917985783407,
+      "grad_norm": 0.4735902467266419,
+      "learning_rate": 5.339679859763988e-06,
+      "loss": 0.2622,
+      "step": 4218
+    },
+    {
+      "epoch": 2.0157696672839136,
+      "grad_norm": 0.4723270046896756,
+      "learning_rate": 5.33502069039566e-06,
+      "loss": 0.283,
+      "step": 4219
+    },
+    {
+      "epoch": 2.016247535989487,
+      "grad_norm": 0.47606202201064185,
+      "learning_rate": 5.330362815010306e-06,
+      "loss": 0.2811,
+      "step": 4220
+    },
+    {
+      "epoch": 2.0167254046950602,
+      "grad_norm": 0.47024447912354794,
+      "learning_rate": 5.325706234899931e-06,
+      "loss": 0.2724,
+      "step": 4221
+    },
+    {
+      "epoch": 2.017203273400633,
+      "grad_norm": 0.46214107281210676,
+      "learning_rate": 5.321050951356197e-06,
+      "loss": 0.2671,
+      "step": 4222
+    },
+    {
+      "epoch": 2.0176811421062064,
+      "grad_norm": 0.4685381460638152,
+      "learning_rate": 5.316396965670394e-06,
+      "loss": 0.2763,
+      "step": 4223
+    },
+    {
+      "epoch": 2.0181590108117793,
+      "grad_norm": 0.45819421363781176,
+      "learning_rate": 5.3117442791334514e-06,
+      "loss": 0.264,
+      "step": 4224
+    },
+    {
+      "epoch": 2.0186368795173526,
+      "grad_norm": 0.45413644290919253,
+      "learning_rate": 5.307092893035951e-06,
+      "loss": 0.285,
+      "step": 4225
+    },
+    {
+      "epoch": 2.019114748222926,
+      "grad_norm": 0.45027926657460016,
+      "learning_rate": 5.3024428086681e-06,
+      "loss": 0.2687,
+      "step": 4226
+    },
+    {
+      "epoch": 2.019592616928499,
+      "grad_norm": 0.4656522036888355,
+      "learning_rate": 5.297794027319747e-06,
+      "loss": 0.2688,
+      "step": 4227
+    },
+    {
+      "epoch": 2.020070485634072,
+      "grad_norm": 0.4763467893183739,
+      "learning_rate": 5.293146550280388e-06,
+      "loss": 0.2758,
+      "step": 4228
+    },
+    {
+      "epoch": 2.020548354339645,
+      "grad_norm": 0.46374013514301876,
+      "learning_rate": 5.28850037883915e-06,
+      "loss": 0.2621,
+      "step": 4229
+    },
+    {
+      "epoch": 2.0210262230452183,
+      "grad_norm": 0.48224561494719265,
+      "learning_rate": 5.2838555142847925e-06,
+      "loss": 0.2731,
+      "step": 4230
+    },
+    {
+      "epoch": 2.0215040917507916,
+      "grad_norm": 0.6679479212555328,
+      "learning_rate": 5.2792119579057275e-06,
+      "loss": 0.2578,
+      "step": 4231
+    },
+    {
+      "epoch": 2.0219819604563645,
+      "grad_norm": 0.49348773245256794,
+      "learning_rate": 5.274569710989994e-06,
+      "loss": 0.2573,
+      "step": 4232
+    },
+    {
+      "epoch": 2.022459829161938,
+      "grad_norm": 0.47356875200053883,
+      "learning_rate": 5.269928774825261e-06,
+      "loss": 0.2815,
+      "step": 4233
+    },
+    {
+      "epoch": 2.022937697867511,
+      "grad_norm": 0.531779784720976,
+      "learning_rate": 5.265289150698855e-06,
+      "loss": 0.286,
+      "step": 4234
+    },
+    {
+      "epoch": 2.023415566573084,
+      "grad_norm": 0.46128402209008207,
+      "learning_rate": 5.260650839897719e-06,
+      "loss": 0.2742,
+      "step": 4235
+    },
+    {
+      "epoch": 2.0238934352786573,
+      "grad_norm": 0.5084306437023877,
+      "learning_rate": 5.256013843708435e-06,
+      "loss": 0.3016,
+      "step": 4236
+    },
+    {
+      "epoch": 2.02437130398423,
+      "grad_norm": 0.5186816576776002,
+      "learning_rate": 5.251378163417232e-06,
+      "loss": 0.2646,
+      "step": 4237
+    },
+    {
+      "epoch": 2.0248491726898035,
+      "grad_norm": 0.7432820985970882,
+      "learning_rate": 5.246743800309964e-06,
+      "loss": 0.2893,
+      "step": 4238
+    },
+    {
+      "epoch": 2.025327041395377,
+      "grad_norm": 0.5031043114302443,
+      "learning_rate": 5.242110755672114e-06,
+      "loss": 0.2574,
+      "step": 4239
+    },
+    {
+      "epoch": 2.0258049101009497,
+      "grad_norm": 0.47460643773168915,
+      "learning_rate": 5.237479030788817e-06,
+      "loss": 0.2619,
+      "step": 4240
+    },
+    {
+      "epoch": 2.026282778806523,
+      "grad_norm": 0.483227882308964,
+      "learning_rate": 5.232848626944827e-06,
+      "loss": 0.2796,
+      "step": 4241
+    },
+    {
+      "epoch": 2.026760647512096,
+      "grad_norm": 0.5588184535169874,
+      "learning_rate": 5.228219545424533e-06,
+      "loss": 0.2732,
+      "step": 4242
+    },
+    {
+      "epoch": 2.027238516217669,
+      "grad_norm": 0.4734462488933155,
+      "learning_rate": 5.2235917875119656e-06,
+      "loss": 0.2708,
+      "step": 4243
+    },
+    {
+      "epoch": 2.0277163849232425,
+      "grad_norm": 0.4781528576455179,
+      "learning_rate": 5.218965354490786e-06,
+      "loss": 0.2753,
+      "step": 4244
+    },
+    {
+      "epoch": 2.0281942536288153,
+      "grad_norm": 0.5011498208217005,
+      "learning_rate": 5.214340247644278e-06,
+      "loss": 0.2675,
+      "step": 4245
+    },
+    {
+      "epoch": 2.0286721223343886,
+      "grad_norm": 0.48552354280981097,
+      "learning_rate": 5.209716468255367e-06,
+      "loss": 0.274,
+      "step": 4246
+    },
+    {
+      "epoch": 2.029149991039962,
+      "grad_norm": 0.5513420020431447,
+      "learning_rate": 5.205094017606611e-06,
+      "loss": 0.2806,
+      "step": 4247
+    },
+    {
+      "epoch": 2.029627859745535,
+      "grad_norm": 0.7169421131443054,
+      "learning_rate": 5.2004728969801945e-06,
+      "loss": 0.281,
+      "step": 4248
+    },
+    {
+      "epoch": 2.030105728451108,
+      "grad_norm": 0.48421888141015607,
+      "learning_rate": 5.19585310765793e-06,
+      "loss": 0.2649,
+      "step": 4249
+    },
+    {
+      "epoch": 2.030583597156681,
+      "grad_norm": 0.4750848932286133,
+      "learning_rate": 5.191234650921273e-06,
+      "loss": 0.2617,
+      "step": 4250
+    },
+    {
+      "epoch": 2.0310614658622543,
+      "grad_norm": 0.45867486142995845,
+      "learning_rate": 5.1866175280513e-06,
+      "loss": 0.2652,
+      "step": 4251
+    },
+    {
+      "epoch": 2.0315393345678276,
+      "grad_norm": 0.4993525355930482,
+      "learning_rate": 5.182001740328713e-06,
+      "loss": 0.2535,
+      "step": 4252
+    },
+    {
+      "epoch": 2.0320172032734005,
+      "grad_norm": 0.478090275533228,
+      "learning_rate": 5.17738728903386e-06,
+      "loss": 0.2761,
+      "step": 4253
+    },
+    {
+      "epoch": 2.032495071978974,
+      "grad_norm": 0.47697353288833244,
+      "learning_rate": 5.172774175446703e-06,
+      "loss": 0.2733,
+      "step": 4254
+    },
+    {
+      "epoch": 2.0329729406845467,
+      "grad_norm": 0.46412526746517374,
+      "learning_rate": 5.168162400846835e-06,
+      "loss": 0.2818,
+      "step": 4255
+    },
+    {
+      "epoch": 2.03345080939012,
+      "grad_norm": 0.4740087446649736,
+      "learning_rate": 5.16355196651349e-06,
+      "loss": 0.2696,
+      "step": 4256
+    },
+    {
+      "epoch": 2.0339286780956933,
+      "grad_norm": 0.4895373321168128,
+      "learning_rate": 5.158942873725514e-06,
+      "loss": 0.2909,
+      "step": 4257
+    },
+    {
+      "epoch": 2.034406546801266,
+      "grad_norm": 0.48823594256633335,
+      "learning_rate": 5.154335123761387e-06,
+      "loss": 0.2603,
+      "step": 4258
+    },
+    {
+      "epoch": 2.0348844155068395,
+      "grad_norm": 0.4403350386861437,
+      "learning_rate": 5.149728717899225e-06,
+      "loss": 0.2739,
+      "step": 4259
+    },
+    {
+      "epoch": 2.035362284212413,
+      "grad_norm": 0.4744290808806091,
+      "learning_rate": 5.145123657416759e-06,
+      "loss": 0.2665,
+      "step": 4260
+    },
+    {
+      "epoch": 2.0358401529179857,
+      "grad_norm": 0.49559161901762544,
+      "learning_rate": 5.140519943591348e-06,
+      "loss": 0.2802,
+      "step": 4261
+    },
+    {
+      "epoch": 2.036318021623559,
+      "grad_norm": 0.4945068985838198,
+      "learning_rate": 5.135917577699988e-06,
+      "loss": 0.2627,
+      "step": 4262
+    },
+    {
+      "epoch": 2.036795890329132,
+      "grad_norm": 0.47314478267085935,
+      "learning_rate": 5.131316561019293e-06,
+      "loss": 0.2787,
+      "step": 4263
+    },
+    {
+      "epoch": 2.037273759034705,
+      "grad_norm": 0.48612430158367564,
+      "learning_rate": 5.126716894825496e-06,
+      "loss": 0.2784,
+      "step": 4264
+    },
+    {
+      "epoch": 2.0377516277402785,
+      "grad_norm": 0.4685580006629703,
+      "learning_rate": 5.122118580394473e-06,
+      "loss": 0.2604,
+      "step": 4265
+    },
+    {
+      "epoch": 2.0382294964458514,
+      "grad_norm": 0.4876824775843487,
+      "learning_rate": 5.117521619001713e-06,
+      "loss": 0.2737,
+      "step": 4266
+    },
+    {
+      "epoch": 2.0387073651514247,
+      "grad_norm": 0.4533392736697577,
+      "learning_rate": 5.112926011922326e-06,
+      "loss": 0.2571,
+      "step": 4267
+    },
+    {
+      "epoch": 2.0391852338569976,
+      "grad_norm": 0.48519263547086117,
+      "learning_rate": 5.10833176043106e-06,
+      "loss": 0.2742,
+      "step": 4268
+    },
+    {
+      "epoch": 2.039663102562571,
+      "grad_norm": 0.4562442726047049,
+      "learning_rate": 5.103738865802277e-06,
+      "loss": 0.2658,
+      "step": 4269
+    },
+    {
+      "epoch": 2.040140971268144,
+      "grad_norm": 0.4672534849486868,
+      "learning_rate": 5.099147329309959e-06,
+      "loss": 0.256,
+      "step": 4270
+    },
+    {
+      "epoch": 2.040618839973717,
+      "grad_norm": 0.4772694807476173,
+      "learning_rate": 5.0945571522277255e-06,
+      "loss": 0.2663,
+      "step": 4271
+    },
+    {
+      "epoch": 2.0410967086792904,
+      "grad_norm": 0.5706905661498267,
+      "learning_rate": 5.08996833582881e-06,
+      "loss": 0.2859,
+      "step": 4272
+    },
+    {
+      "epoch": 2.0415745773848637,
+      "grad_norm": 0.4947667148137507,
+      "learning_rate": 5.0853808813860616e-06,
+      "loss": 0.3067,
+      "step": 4273
+    },
+    {
+      "epoch": 2.0420524460904366,
+      "grad_norm": 0.4851005503979229,
+      "learning_rate": 5.080794790171968e-06,
+      "loss": 0.2716,
+      "step": 4274
+    },
+    {
+      "epoch": 2.04253031479601,
+      "grad_norm": 0.48079461477386043,
+      "learning_rate": 5.076210063458622e-06,
+      "loss": 0.2688,
+      "step": 4275
+    },
+    {
+      "epoch": 2.0430081835015828,
+      "grad_norm": 0.4673487498716285,
+      "learning_rate": 5.071626702517756e-06,
+      "loss": 0.2632,
+      "step": 4276
+    },
+    {
+      "epoch": 2.043486052207156,
+      "grad_norm": 0.4653754050455356,
+      "learning_rate": 5.067044708620702e-06,
+      "loss": 0.2827,
+      "step": 4277
+    },
+    {
+      "epoch": 2.0439639209127294,
+      "grad_norm": 0.4697586401370185,
+      "learning_rate": 5.062464083038434e-06,
+      "loss": 0.2667,
+      "step": 4278
+    },
+    {
+      "epoch": 2.0444417896183023,
+      "grad_norm": 0.5260281606453924,
+      "learning_rate": 5.057884827041533e-06,
+      "loss": 0.2515,
+      "step": 4279
+    },
+    {
+      "epoch": 2.0449196583238756,
+      "grad_norm": 0.4994642856059735,
+      "learning_rate": 5.0533069419002e-06,
+      "loss": 0.2671,
+      "step": 4280
+    },
+    {
+      "epoch": 2.045397527029449,
+      "grad_norm": 0.48493301210780937,
+      "learning_rate": 5.048730428884268e-06,
+      "loss": 0.2727,
+      "step": 4281
+    },
+    {
+      "epoch": 2.0458753957350218,
+      "grad_norm": 0.46353129132177423,
+      "learning_rate": 5.044155289263174e-06,
+      "loss": 0.2765,
+      "step": 4282
+    },
+    {
+      "epoch": 2.046353264440595,
+      "grad_norm": 0.5018313449986281,
+      "learning_rate": 5.03958152430598e-06,
+      "loss": 0.2727,
+      "step": 4283
+    },
+    {
+      "epoch": 2.046831133146168,
+      "grad_norm": 0.4851187115521681,
+      "learning_rate": 5.035009135281375e-06,
+      "loss": 0.2652,
+      "step": 4284
+    },
+    {
+      "epoch": 2.0473090018517413,
+      "grad_norm": 0.4723372968025927,
+      "learning_rate": 5.030438123457655e-06,
+      "loss": 0.2729,
+      "step": 4285
+    },
+    {
+      "epoch": 2.0477868705573146,
+      "grad_norm": 0.47152065145636834,
+      "learning_rate": 5.025868490102734e-06,
+      "loss": 0.2724,
+      "step": 4286
+    },
+    {
+      "epoch": 2.0482647392628874,
+      "grad_norm": 0.5260680519772822,
+      "learning_rate": 5.021300236484156e-06,
+      "loss": 0.2648,
+      "step": 4287
+    },
+    {
+      "epoch": 2.0487426079684607,
+      "grad_norm": 0.47329119567837263,
+      "learning_rate": 5.016733363869068e-06,
+      "loss": 0.2713,
+      "step": 4288
+    },
+    {
+      "epoch": 2.0492204766740336,
+      "grad_norm": 0.50642114847147,
+      "learning_rate": 5.01216787352424e-06,
+      "loss": 0.2638,
+      "step": 4289
+    },
+    {
+      "epoch": 2.049698345379607,
+      "grad_norm": 0.4705941501131214,
+      "learning_rate": 5.007603766716063e-06,
+      "loss": 0.2643,
+      "step": 4290
+    },
+    {
+      "epoch": 2.0501762140851802,
+      "grad_norm": 0.4803945099496459,
+      "learning_rate": 5.003041044710536e-06,
+      "loss": 0.2717,
+      "step": 4291
+    },
+    {
+      "epoch": 2.050654082790753,
+      "grad_norm": 0.4612851976830831,
+      "learning_rate": 4.998479708773275e-06,
+      "loss": 0.2805,
+      "step": 4292
+    },
+    {
+      "epoch": 2.0511319514963264,
+      "grad_norm": 0.4759211681133958,
+      "learning_rate": 4.993919760169521e-06,
+      "loss": 0.2659,
+      "step": 4293
+    },
+    {
+      "epoch": 2.0516098202018993,
+      "grad_norm": 0.4979267001248353,
+      "learning_rate": 4.98936120016412e-06,
+      "loss": 0.2764,
+      "step": 4294
+    },
+    {
+      "epoch": 2.0520876889074726,
+      "grad_norm": 0.5214662750542031,
+      "learning_rate": 4.984804030021533e-06,
+      "loss": 0.2707,
+      "step": 4295
+    },
+    {
+      "epoch": 2.052565557613046,
+      "grad_norm": 0.4913707086962388,
+      "learning_rate": 4.9802482510058445e-06,
+      "loss": 0.2687,
+      "step": 4296
+    },
+    {
+      "epoch": 2.053043426318619,
+      "grad_norm": 0.5027711346349333,
+      "learning_rate": 4.975693864380744e-06,
+      "loss": 0.2613,
+      "step": 4297
+    },
+    {
+      "epoch": 2.053521295024192,
+      "grad_norm": 0.5024104113465125,
+      "learning_rate": 4.971140871409536e-06,
+      "loss": 0.2741,
+      "step": 4298
+    },
+    {
+      "epoch": 2.0539991637297654,
+      "grad_norm": 0.4811287363006302,
+      "learning_rate": 4.966589273355144e-06,
+      "loss": 0.2689,
+      "step": 4299
+    },
+    {
+      "epoch": 2.0544770324353383,
+      "grad_norm": 0.4644489936036945,
+      "learning_rate": 4.962039071480102e-06,
+      "loss": 0.2791,
+      "step": 4300
+    },
+    {
+      "epoch": 2.0549549011409116,
+      "grad_norm": 0.5336660306893748,
+      "learning_rate": 4.957490267046549e-06,
+      "loss": 0.2803,
+      "step": 4301
+    },
+    {
+      "epoch": 2.0554327698464845,
+      "grad_norm": 0.47071851786985547,
+      "learning_rate": 4.95294286131625e-06,
+      "loss": 0.2673,
+      "step": 4302
+    },
+    {
+      "epoch": 2.055910638552058,
+      "grad_norm": 0.7828065837615993,
+      "learning_rate": 4.948396855550575e-06,
+      "loss": 0.2633,
+      "step": 4303
+    },
+    {
+      "epoch": 2.056388507257631,
+      "grad_norm": 0.45464028072528234,
+      "learning_rate": 4.943852251010498e-06,
+      "loss": 0.2793,
+      "step": 4304
+    },
+    {
+      "epoch": 2.056866375963204,
+      "grad_norm": 0.6373445627548838,
+      "learning_rate": 4.939309048956622e-06,
+      "loss": 0.2825,
+      "step": 4305
+    },
+    {
+      "epoch": 2.0573442446687773,
+      "grad_norm": 0.4671574195509635,
+      "learning_rate": 4.934767250649146e-06,
+      "loss": 0.2825,
+      "step": 4306
+    },
+    {
+      "epoch": 2.0578221133743506,
+      "grad_norm": 0.5249600277106131,
+      "learning_rate": 4.9302268573478825e-06,
+      "loss": 0.2798,
+      "step": 4307
+    },
+    {
+      "epoch": 2.0582999820799235,
+      "grad_norm": 0.532751022686781,
+      "learning_rate": 4.925687870312263e-06,
+      "loss": 0.2704,
+      "step": 4308
+    },
+    {
+      "epoch": 2.058777850785497,
+      "grad_norm": 0.4936173702678726,
+      "learning_rate": 4.921150290801316e-06,
+      "loss": 0.2791,
+      "step": 4309
+    },
+    {
+      "epoch": 2.0592557194910697,
+      "grad_norm": 0.46696117374431717,
+      "learning_rate": 4.9166141200736885e-06,
+      "loss": 0.2717,
+      "step": 4310
+    },
+    {
+      "epoch": 2.059733588196643,
+      "grad_norm": 0.4612537280249499,
+      "learning_rate": 4.912079359387638e-06,
+      "loss": 0.2712,
+      "step": 4311
+    },
+    {
+      "epoch": 2.0602114569022163,
+      "grad_norm": 0.44406796450445546,
+      "learning_rate": 4.907546010001026e-06,
+      "loss": 0.2772,
+      "step": 4312
+    },
+    {
+      "epoch": 2.060689325607789,
+      "grad_norm": 0.4685017422306754,
+      "learning_rate": 4.903014073171315e-06,
+      "loss": 0.2689,
+      "step": 4313
+    },
+    {
+      "epoch": 2.0611671943133625,
+      "grad_norm": 0.49168518163616726,
+      "learning_rate": 4.898483550155595e-06,
+      "loss": 0.2685,
+      "step": 4314
+    },
+    {
+      "epoch": 2.0616450630189354,
+      "grad_norm": 0.47506846437092554,
+      "learning_rate": 4.89395444221055e-06,
+      "loss": 0.2636,
+      "step": 4315
+    },
+    {
+      "epoch": 2.0621229317245087,
+      "grad_norm": 0.44749028284198716,
+      "learning_rate": 4.889426750592469e-06,
+      "loss": 0.273,
+      "step": 4316
+    },
+    {
+      "epoch": 2.062600800430082,
+      "grad_norm": 0.4643039012469354,
+      "learning_rate": 4.884900476557263e-06,
+      "loss": 0.2843,
+      "step": 4317
+    },
+    {
+      "epoch": 2.063078669135655,
+      "grad_norm": 0.524850579695799,
+      "learning_rate": 4.880375621360435e-06,
+      "loss": 0.2732,
+      "step": 4318
+    },
+    {
+      "epoch": 2.063556537841228,
+      "grad_norm": 0.4937697172955831,
+      "learning_rate": 4.8758521862570975e-06,
+      "loss": 0.2877,
+      "step": 4319
+    },
+    {
+      "epoch": 2.0640344065468015,
+      "grad_norm": 0.4568437096178462,
+      "learning_rate": 4.871330172501979e-06,
+      "loss": 0.2591,
+      "step": 4320
+    },
+    {
+      "epoch": 2.0645122752523744,
+      "grad_norm": 0.4818485682691764,
+      "learning_rate": 4.866809581349403e-06,
+      "loss": 0.2842,
+      "step": 4321
+    },
+    {
+      "epoch": 2.0649901439579477,
+      "grad_norm": 0.500279053318082,
+      "learning_rate": 4.862290414053296e-06,
+      "loss": 0.2689,
+      "step": 4322
+    },
+    {
+      "epoch": 2.0654680126635205,
+      "grad_norm": 0.4796867445554122,
+      "learning_rate": 4.857772671867206e-06,
+      "loss": 0.261,
+      "step": 4323
+    },
+    {
+      "epoch": 2.065945881369094,
+      "grad_norm": 0.45209725252949634,
+      "learning_rate": 4.853256356044269e-06,
+      "loss": 0.263,
+      "step": 4324
+    },
+    {
+      "epoch": 2.066423750074667,
+      "grad_norm": 0.5262595142130551,
+      "learning_rate": 4.848741467837228e-06,
+      "loss": 0.3065,
+      "step": 4325
+    },
+    {
+      "epoch": 2.06690161878024,
+      "grad_norm": 0.5027222558313732,
+      "learning_rate": 4.844228008498441e-06,
+      "loss": 0.2771,
+      "step": 4326
+    },
+    {
+      "epoch": 2.0673794874858133,
+      "grad_norm": 0.5140923745008394,
+      "learning_rate": 4.839715979279857e-06,
+      "loss": 0.2673,
+      "step": 4327
+    },
+    {
+      "epoch": 2.067857356191386,
+      "grad_norm": 0.6688922617742407,
+      "learning_rate": 4.835205381433033e-06,
+      "loss": 0.2759,
+      "step": 4328
+    },
+    {
+      "epoch": 2.0683352248969595,
+      "grad_norm": 0.5106868534593813,
+      "learning_rate": 4.830696216209133e-06,
+      "loss": 0.2707,
+      "step": 4329
+    },
+    {
+      "epoch": 2.068813093602533,
+      "grad_norm": 0.5125392579142816,
+      "learning_rate": 4.826188484858918e-06,
+      "loss": 0.2796,
+      "step": 4330
+    },
+    {
+      "epoch": 2.0692909623081057,
+      "grad_norm": 0.4731305023489138,
+      "learning_rate": 4.821682188632749e-06,
+      "loss": 0.2698,
+      "step": 4331
+    },
+    {
+      "epoch": 2.069768831013679,
+      "grad_norm": 0.537078693052272,
+      "learning_rate": 4.8171773287806e-06,
+      "loss": 0.2763,
+      "step": 4332
+    },
+    {
+      "epoch": 2.0702466997192523,
+      "grad_norm": 0.48986368836511185,
+      "learning_rate": 4.812673906552038e-06,
+      "loss": 0.2682,
+      "step": 4333
+    },
+    {
+      "epoch": 2.070724568424825,
+      "grad_norm": 0.49970157825319683,
+      "learning_rate": 4.808171923196227e-06,
+      "loss": 0.2745,
+      "step": 4334
+    },
+    {
+      "epoch": 2.0712024371303985,
+      "grad_norm": 0.46279135492922463,
+      "learning_rate": 4.803671379961945e-06,
+      "loss": 0.2732,
+      "step": 4335
+    },
+    {
+      "epoch": 2.0716803058359714,
+      "grad_norm": 0.5228404024805013,
+      "learning_rate": 4.7991722780975614e-06,
+      "loss": 0.2569,
+      "step": 4336
+    },
+    {
+      "epoch": 2.0721581745415447,
+      "grad_norm": 0.47489759603230547,
+      "learning_rate": 4.794674618851044e-06,
+      "loss": 0.2702,
+      "step": 4337
+    },
+    {
+      "epoch": 2.072636043247118,
+      "grad_norm": 0.45711910658365396,
+      "learning_rate": 4.7901784034699695e-06,
+      "loss": 0.2581,
+      "step": 4338
+    },
+    {
+      "epoch": 2.073113911952691,
+      "grad_norm": 0.5062614569880117,
+      "learning_rate": 4.785683633201507e-06,
+      "loss": 0.2494,
+      "step": 4339
+    },
+    {
+      "epoch": 2.073591780658264,
+      "grad_norm": 0.4838987562625956,
+      "learning_rate": 4.781190309292421e-06,
+      "loss": 0.2666,
+      "step": 4340
+    },
+    {
+      "epoch": 2.074069649363837,
+      "grad_norm": 0.48244525999987964,
+      "learning_rate": 4.776698432989089e-06,
+      "loss": 0.2641,
+      "step": 4341
+    },
+    {
+      "epoch": 2.0745475180694104,
+      "grad_norm": 0.481389690872325,
+      "learning_rate": 4.7722080055374745e-06,
+      "loss": 0.2766,
+      "step": 4342
+    },
+    {
+      "epoch": 2.0750253867749837,
+      "grad_norm": 0.547344894388519,
+      "learning_rate": 4.767719028183139e-06,
+      "loss": 0.2779,
+      "step": 4343
+    },
+    {
+      "epoch": 2.0755032554805566,
+      "grad_norm": 0.4969679278468139,
+      "learning_rate": 4.7632315021712494e-06,
+      "loss": 0.2767,
+      "step": 4344
+    },
+    {
+      "epoch": 2.07598112418613,
+      "grad_norm": 0.4969616199208339,
+      "learning_rate": 4.758745428746569e-06,
+      "loss": 0.2609,
+      "step": 4345
+    },
+    {
+      "epoch": 2.076458992891703,
+      "grad_norm": 0.46420623564379426,
+      "learning_rate": 4.754260809153453e-06,
+      "loss": 0.2606,
+      "step": 4346
+    },
+    {
+      "epoch": 2.076936861597276,
+      "grad_norm": 0.4624803013184737,
+      "learning_rate": 4.749777644635851e-06,
+      "loss": 0.2658,
+      "step": 4347
+    },
+    {
+      "epoch": 2.0774147303028494,
+      "grad_norm": 0.4564411009275747,
+      "learning_rate": 4.745295936437323e-06,
+      "loss": 0.2893,
+      "step": 4348
+    },
+    {
+      "epoch": 2.0778925990084223,
+      "grad_norm": 0.45522303982023976,
+      "learning_rate": 4.74081568580101e-06,
+      "loss": 0.2888,
+      "step": 4349
+    },
+    {
+      "epoch": 2.0783704677139956,
+      "grad_norm": 0.45649179256830386,
+      "learning_rate": 4.736336893969652e-06,
+      "loss": 0.2627,
+      "step": 4350
+    },
+    {
+      "epoch": 2.078848336419569,
+      "grad_norm": 0.8720177385632242,
+      "learning_rate": 4.731859562185593e-06,
+      "loss": 0.2504,
+      "step": 4351
+    },
+    {
+      "epoch": 2.0793262051251418,
+      "grad_norm": 0.464019520768861,
+      "learning_rate": 4.727383691690765e-06,
+      "loss": 0.2803,
+      "step": 4352
+    },
+    {
+      "epoch": 2.079804073830715,
+      "grad_norm": 0.4829753872971073,
+      "learning_rate": 4.722909283726687e-06,
+      "loss": 0.2747,
+      "step": 4353
+    },
+    {
+      "epoch": 2.080281942536288,
+      "grad_norm": 0.49238958382928866,
+      "learning_rate": 4.718436339534493e-06,
+      "loss": 0.2696,
+      "step": 4354
+    },
+    {
+      "epoch": 2.0807598112418613,
+      "grad_norm": 0.5331681056474048,
+      "learning_rate": 4.7139648603548925e-06,
+      "loss": 0.2803,
+      "step": 4355
+    },
+    {
+      "epoch": 2.0812376799474346,
+      "grad_norm": 0.5073464839867855,
+      "learning_rate": 4.709494847428193e-06,
+      "loss": 0.2841,
+      "step": 4356
+    },
+    {
+      "epoch": 2.0817155486530075,
+      "grad_norm": 0.45403717199200416,
+      "learning_rate": 4.7050263019943035e-06,
+      "loss": 0.2693,
+      "step": 4357
+    },
+    {
+      "epoch": 2.0821934173585808,
+      "grad_norm": 0.4790899870289159,
+      "learning_rate": 4.700559225292714e-06,
+      "loss": 0.2723,
+      "step": 4358
+    },
+    {
+      "epoch": 2.082671286064154,
+      "grad_norm": 0.5479563382822494,
+      "learning_rate": 4.696093618562514e-06,
+      "loss": 0.2827,
+      "step": 4359
+    },
+    {
+      "epoch": 2.083149154769727,
+      "grad_norm": 0.4645988394382689,
+      "learning_rate": 4.691629483042387e-06,
+      "loss": 0.2834,
+      "step": 4360
+    },
+    {
+      "epoch": 2.0836270234753003,
+      "grad_norm": 0.48819600396553203,
+      "learning_rate": 4.687166819970605e-06,
+      "loss": 0.2806,
+      "step": 4361
+    },
+    {
+      "epoch": 2.084104892180873,
+      "grad_norm": 0.48443281966006396,
+      "learning_rate": 4.682705630585024e-06,
+      "loss": 0.2777,
+      "step": 4362
+    },
+    {
+      "epoch": 2.0845827608864465,
+      "grad_norm": 0.4840904657426456,
+      "learning_rate": 4.678245916123111e-06,
+      "loss": 0.2632,
+      "step": 4363
+    },
+    {
+      "epoch": 2.0850606295920198,
+      "grad_norm": 0.4817924591446688,
+      "learning_rate": 4.673787677821906e-06,
+      "loss": 0.2571,
+      "step": 4364
+    },
+    {
+      "epoch": 2.0855384982975926,
+      "grad_norm": 0.4525499762733216,
+      "learning_rate": 4.669330916918043e-06,
+      "loss": 0.2837,
+      "step": 4365
+    },
+    {
+      "epoch": 2.086016367003166,
+      "grad_norm": 0.4568315340280084,
+      "learning_rate": 4.664875634647756e-06,
+      "loss": 0.2785,
+      "step": 4366
+    },
+    {
+      "epoch": 2.086494235708739,
+      "grad_norm": 0.5183117984656025,
+      "learning_rate": 4.660421832246858e-06,
+      "loss": 0.2936,
+      "step": 4367
+    },
+    {
+      "epoch": 2.086972104414312,
+      "grad_norm": 0.4872870778331395,
+      "learning_rate": 4.655969510950752e-06,
+      "loss": 0.2733,
+      "step": 4368
+    },
+    {
+      "epoch": 2.0874499731198854,
+      "grad_norm": 0.46677867691861946,
+      "learning_rate": 4.65151867199444e-06,
+      "loss": 0.2678,
+      "step": 4369
+    },
+    {
+      "epoch": 2.0879278418254583,
+      "grad_norm": 0.46455148899464205,
+      "learning_rate": 4.647069316612502e-06,
+      "loss": 0.2687,
+      "step": 4370
+    },
+    {
+      "epoch": 2.0884057105310316,
+      "grad_norm": 1.2838537701376698,
+      "learning_rate": 4.6426214460391095e-06,
+      "loss": 0.2844,
+      "step": 4371
+    },
+    {
+      "epoch": 2.088883579236605,
+      "grad_norm": 0.47029125249903925,
+      "learning_rate": 4.6381750615080275e-06,
+      "loss": 0.2618,
+      "step": 4372
+    },
+    {
+      "epoch": 2.089361447942178,
+      "grad_norm": 0.4888788479772587,
+      "learning_rate": 4.633730164252603e-06,
+      "loss": 0.2643,
+      "step": 4373
+    },
+    {
+      "epoch": 2.089839316647751,
+      "grad_norm": 0.4591781777504566,
+      "learning_rate": 4.629286755505768e-06,
+      "loss": 0.2754,
+      "step": 4374
+    },
+    {
+      "epoch": 2.090317185353324,
+      "grad_norm": 0.5110989494551736,
+      "learning_rate": 4.624844836500052e-06,
+      "loss": 0.2586,
+      "step": 4375
+    },
+    {
+      "epoch": 2.0907950540588973,
+      "grad_norm": 0.4951473466562376,
+      "learning_rate": 4.620404408467559e-06,
+      "loss": 0.2701,
+      "step": 4376
+    },
+    {
+      "epoch": 2.0912729227644706,
+      "grad_norm": 0.45662970354071386,
+      "learning_rate": 4.615965472639992e-06,
+      "loss": 0.2649,
+      "step": 4377
+    },
+    {
+      "epoch": 2.0917507914700435,
+      "grad_norm": 0.46953730734656046,
+      "learning_rate": 4.611528030248629e-06,
+      "loss": 0.2712,
+      "step": 4378
+    },
+    {
+      "epoch": 2.092228660175617,
+      "grad_norm": 0.47897826667944887,
+      "learning_rate": 4.607092082524341e-06,
+      "loss": 0.2544,
+      "step": 4379
+    },
+    {
+      "epoch": 2.0927065288811897,
+      "grad_norm": 0.45754772942931315,
+      "learning_rate": 4.60265763069758e-06,
+      "loss": 0.2739,
+      "step": 4380
+    },
+    {
+      "epoch": 2.093184397586763,
+      "grad_norm": 0.47343432218709175,
+      "learning_rate": 4.598224675998381e-06,
+      "loss": 0.2632,
+      "step": 4381
+    },
+    {
+      "epoch": 2.0936622662923363,
+      "grad_norm": 0.5079217054437215,
+      "learning_rate": 4.593793219656375e-06,
+      "loss": 0.2884,
+      "step": 4382
+    },
+    {
+      "epoch": 2.094140134997909,
+      "grad_norm": 0.4755846137336544,
+      "learning_rate": 4.589363262900767e-06,
+      "loss": 0.2769,
+      "step": 4383
+    },
+    {
+      "epoch": 2.0946180037034825,
+      "grad_norm": 0.4624399910521351,
+      "learning_rate": 4.5849348069603424e-06,
+      "loss": 0.2833,
+      "step": 4384
+    },
+    {
+      "epoch": 2.095095872409056,
+      "grad_norm": 0.47379072504138925,
+      "learning_rate": 4.580507853063487e-06,
+      "loss": 0.2723,
+      "step": 4385
+    },
+    {
+      "epoch": 2.0955737411146287,
+      "grad_norm": 0.5404425798815802,
+      "learning_rate": 4.5760824024381545e-06,
+      "loss": 0.2836,
+      "step": 4386
+    },
+    {
+      "epoch": 2.096051609820202,
+      "grad_norm": 0.4744131279778523,
+      "learning_rate": 4.571658456311885e-06,
+      "loss": 0.2646,
+      "step": 4387
+    },
+    {
+      "epoch": 2.096529478525775,
+      "grad_norm": 0.5257445500444607,
+      "learning_rate": 4.567236015911808e-06,
+      "loss": 0.2756,
+      "step": 4388
+    },
+    {
+      "epoch": 2.097007347231348,
+      "grad_norm": 0.46379662230318247,
+      "learning_rate": 4.562815082464628e-06,
+      "loss": 0.2577,
+      "step": 4389
+    },
+    {
+      "epoch": 2.0974852159369215,
+      "grad_norm": 0.4711856915631454,
+      "learning_rate": 4.5583956571966295e-06,
+      "loss": 0.2695,
+      "step": 4390
+    },
+    {
+      "epoch": 2.0979630846424944,
+      "grad_norm": 0.4668862612467058,
+      "learning_rate": 4.5539777413336916e-06,
+      "loss": 0.2811,
+      "step": 4391
+    },
+    {
+      "epoch": 2.0984409533480677,
+      "grad_norm": 0.8250470635919427,
+      "learning_rate": 4.549561336101263e-06,
+      "loss": 0.2681,
+      "step": 4392
+    },
+    {
+      "epoch": 2.0989188220536406,
+      "grad_norm": 0.5474390912842495,
+      "learning_rate": 4.545146442724371e-06,
+      "loss": 0.278,
+      "step": 4393
+    },
+    {
+      "epoch": 2.099396690759214,
+      "grad_norm": 0.45864322699427934,
+      "learning_rate": 4.540733062427637e-06,
+      "loss": 0.2572,
+      "step": 4394
+    },
+    {
+      "epoch": 2.099874559464787,
+      "grad_norm": 0.4642926001377059,
+      "learning_rate": 4.5363211964352524e-06,
+      "loss": 0.2754,
+      "step": 4395
+    },
+    {
+      "epoch": 2.10035242817036,
+      "grad_norm": 1.2708133595604119,
+      "learning_rate": 4.531910845970986e-06,
+      "loss": 0.2515,
+      "step": 4396
+    },
+    {
+      "epoch": 2.1008302968759334,
+      "grad_norm": 0.46107312847431264,
+      "learning_rate": 4.527502012258201e-06,
+      "loss": 0.2798,
+      "step": 4397
+    },
+    {
+      "epoch": 2.1013081655815067,
+      "grad_norm": 0.4654804719208443,
+      "learning_rate": 4.523094696519822e-06,
+      "loss": 0.2599,
+      "step": 4398
+    },
+    {
+      "epoch": 2.1017860342870796,
+      "grad_norm": 0.5035051987740295,
+      "learning_rate": 4.5186888999783604e-06,
+      "loss": 0.2796,
+      "step": 4399
+    },
+    {
+      "epoch": 2.102263902992653,
+      "grad_norm": 0.4639728245767181,
+      "learning_rate": 4.514284623855915e-06,
+      "loss": 0.2711,
+      "step": 4400
+    },
+    {
+      "epoch": 2.1027417716982257,
+      "grad_norm": 0.4628009973817066,
+      "learning_rate": 4.509881869374146e-06,
+      "loss": 0.2749,
+      "step": 4401
+    },
+    {
+      "epoch": 2.103219640403799,
+      "grad_norm": 0.46310930729244,
+      "learning_rate": 4.5054806377543e-06,
+      "loss": 0.2766,
+      "step": 4402
+    },
+    {
+      "epoch": 2.1036975091093724,
+      "grad_norm": 0.4815051304539407,
+      "learning_rate": 4.501080930217206e-06,
+      "loss": 0.2686,
+      "step": 4403
+    },
+    {
+      "epoch": 2.1041753778149452,
+      "grad_norm": 0.452776883071896,
+      "learning_rate": 4.4966827479832645e-06,
+      "loss": 0.2815,
+      "step": 4404
+    },
+    {
+      "epoch": 2.1046532465205186,
+      "grad_norm": 0.5572260774768857,
+      "learning_rate": 4.4922860922724466e-06,
+      "loss": 0.2795,
+      "step": 4405
+    },
+    {
+      "epoch": 2.1051311152260914,
+      "grad_norm": 0.46536041364576214,
+      "learning_rate": 4.487890964304317e-06,
+      "loss": 0.2607,
+      "step": 4406
+    },
+    {
+      "epoch": 2.1056089839316647,
+      "grad_norm": 0.4671149161280602,
+      "learning_rate": 4.483497365298001e-06,
+      "loss": 0.2656,
+      "step": 4407
+    },
+    {
+      "epoch": 2.106086852637238,
+      "grad_norm": 0.47754698203735124,
+      "learning_rate": 4.479105296472204e-06,
+      "loss": 0.2636,
+      "step": 4408
+    },
+    {
+      "epoch": 2.106564721342811,
+      "grad_norm": 0.4691739905835083,
+      "learning_rate": 4.474714759045213e-06,
+      "loss": 0.2797,
+      "step": 4409
+    },
+    {
+      "epoch": 2.1070425900483842,
+      "grad_norm": 0.5429458246931929,
+      "learning_rate": 4.470325754234881e-06,
+      "loss": 0.2727,
+      "step": 4410
+    },
+    {
+      "epoch": 2.1075204587539575,
+      "grad_norm": 0.5405366505121745,
+      "learning_rate": 4.465938283258643e-06,
+      "loss": 0.2734,
+      "step": 4411
+    },
+    {
+      "epoch": 2.1079983274595304,
+      "grad_norm": 0.7314425834233769,
+      "learning_rate": 4.461552347333509e-06,
+      "loss": 0.2515,
+      "step": 4412
+    },
+    {
+      "epoch": 2.1084761961651037,
+      "grad_norm": 0.6100608642443537,
+      "learning_rate": 4.457167947676058e-06,
+      "loss": 0.2697,
+      "step": 4413
+    },
+    {
+      "epoch": 2.1089540648706766,
+      "grad_norm": 0.4837731698853667,
+      "learning_rate": 4.45278508550244e-06,
+      "loss": 0.2907,
+      "step": 4414
+    },
+    {
+      "epoch": 2.10943193357625,
+      "grad_norm": 0.47440995597138375,
+      "learning_rate": 4.448403762028391e-06,
+      "loss": 0.2709,
+      "step": 4415
+    },
+    {
+      "epoch": 2.1099098022818232,
+      "grad_norm": 0.5691216197594825,
+      "learning_rate": 4.444023978469212e-06,
+      "loss": 0.2778,
+      "step": 4416
+    },
+    {
+      "epoch": 2.110387670987396,
+      "grad_norm": 0.4528201720944284,
+      "learning_rate": 4.4396457360397704e-06,
+      "loss": 0.2619,
+      "step": 4417
+    },
+    {
+      "epoch": 2.1108655396929694,
+      "grad_norm": 0.7031077512368775,
+      "learning_rate": 4.435269035954523e-06,
+      "loss": 0.2658,
+      "step": 4418
+    },
+    {
+      "epoch": 2.1113434083985423,
+      "grad_norm": 0.4719207523909015,
+      "learning_rate": 4.430893879427486e-06,
+      "loss": 0.2622,
+      "step": 4419
+    },
+    {
+      "epoch": 2.1118212771041156,
+      "grad_norm": 0.4711222028375983,
+      "learning_rate": 4.426520267672244e-06,
+      "loss": 0.2908,
+      "step": 4420
+    },
+    {
+      "epoch": 2.112299145809689,
+      "grad_norm": 0.5717573571560005,
+      "learning_rate": 4.422148201901969e-06,
+      "loss": 0.272,
+      "step": 4421
+    },
+    {
+      "epoch": 2.112777014515262,
+      "grad_norm": 0.48469162216982337,
+      "learning_rate": 4.4177776833293915e-06,
+      "loss": 0.2789,
+      "step": 4422
+    },
+    {
+      "epoch": 2.113254883220835,
+      "grad_norm": 0.4591193788359184,
+      "learning_rate": 4.4134087131668135e-06,
+      "loss": 0.2855,
+      "step": 4423
+    },
+    {
+      "epoch": 2.1137327519264084,
+      "grad_norm": 0.451459503417534,
+      "learning_rate": 4.409041292626115e-06,
+      "loss": 0.2718,
+      "step": 4424
+    },
+    {
+      "epoch": 2.1142106206319813,
+      "grad_norm": 0.5180358351813875,
+      "learning_rate": 4.40467542291874e-06,
+      "loss": 0.2669,
+      "step": 4425
+    },
+    {
+      "epoch": 2.1146884893375546,
+      "grad_norm": 0.48205075176726636,
+      "learning_rate": 4.400311105255698e-06,
+      "loss": 0.2659,
+      "step": 4426
+    },
+    {
+      "epoch": 2.1151663580431275,
+      "grad_norm": 0.4544529548772736,
+      "learning_rate": 4.395948340847584e-06,
+      "loss": 0.2739,
+      "step": 4427
+    },
+    {
+      "epoch": 2.115644226748701,
+      "grad_norm": 0.44998143432924953,
+      "learning_rate": 4.391587130904544e-06,
+      "loss": 0.2568,
+      "step": 4428
+    },
+    {
+      "epoch": 2.116122095454274,
+      "grad_norm": 0.4857935888324327,
+      "learning_rate": 4.387227476636301e-06,
+      "loss": 0.2564,
+      "step": 4429
+    },
+    {
+      "epoch": 2.116599964159847,
+      "grad_norm": 0.482388126357244,
+      "learning_rate": 4.382869379252152e-06,
+      "loss": 0.269,
+      "step": 4430
+    },
+    {
+      "epoch": 2.1170778328654203,
+      "grad_norm": 0.46675426655921626,
+      "learning_rate": 4.378512839960953e-06,
+      "loss": 0.2701,
+      "step": 4431
+    },
+    {
+      "epoch": 2.1175557015709936,
+      "grad_norm": 0.4567298710073206,
+      "learning_rate": 4.374157859971127e-06,
+      "loss": 0.2582,
+      "step": 4432
+    },
+    {
+      "epoch": 2.1180335702765665,
+      "grad_norm": 1.0653818376694562,
+      "learning_rate": 4.369804440490676e-06,
+      "loss": 0.2658,
+      "step": 4433
+    },
+    {
+      "epoch": 2.11851143898214,
+      "grad_norm": 0.4691784307308584,
+      "learning_rate": 4.3654525827271576e-06,
+      "loss": 0.2797,
+      "step": 4434
+    },
+    {
+      "epoch": 2.1189893076877127,
+      "grad_norm": 1.179538881720667,
+      "learning_rate": 4.361102287887698e-06,
+      "loss": 0.2675,
+      "step": 4435
+    },
+    {
+      "epoch": 2.119467176393286,
+      "grad_norm": 0.470987292017048,
+      "learning_rate": 4.356753557178999e-06,
+      "loss": 0.2782,
+      "step": 4436
+    },
+    {
+      "epoch": 2.1199450450988593,
+      "grad_norm": 0.46682100572447127,
+      "learning_rate": 4.352406391807318e-06,
+      "loss": 0.2718,
+      "step": 4437
+    },
+    {
+      "epoch": 2.120422913804432,
+      "grad_norm": 0.5044197345551845,
+      "learning_rate": 4.348060792978479e-06,
+      "loss": 0.2655,
+      "step": 4438
+    },
+    {
+      "epoch": 2.1209007825100055,
+      "grad_norm": 0.4681899555394538,
+      "learning_rate": 4.34371676189788e-06,
+      "loss": 0.2658,
+      "step": 4439
+    },
+    {
+      "epoch": 2.1213786512155783,
+      "grad_norm": 0.4703827383388374,
+      "learning_rate": 4.339374299770477e-06,
+      "loss": 0.2728,
+      "step": 4440
+    },
+    {
+      "epoch": 2.1218565199211517,
+      "grad_norm": 0.48211199249212094,
+      "learning_rate": 4.335033407800787e-06,
+      "loss": 0.2863,
+      "step": 4441
+    },
+    {
+      "epoch": 2.122334388626725,
+      "grad_norm": 0.48843268326528816,
+      "learning_rate": 4.330694087192906e-06,
+      "loss": 0.2813,
+      "step": 4442
+    },
+    {
+      "epoch": 2.122812257332298,
+      "grad_norm": 0.5164345291433518,
+      "learning_rate": 4.32635633915048e-06,
+      "loss": 0.2712,
+      "step": 4443
+    },
+    {
+      "epoch": 2.123290126037871,
+      "grad_norm": 0.46371527898883846,
+      "learning_rate": 4.322020164876722e-06,
+      "loss": 0.2639,
+      "step": 4444
+    },
+    {
+      "epoch": 2.123767994743444,
+      "grad_norm": 0.48627299210477126,
+      "learning_rate": 4.31768556557441e-06,
+      "loss": 0.2661,
+      "step": 4445
+    },
+    {
+      "epoch": 2.1242458634490173,
+      "grad_norm": 0.5076108512131333,
+      "learning_rate": 4.313352542445892e-06,
+      "loss": 0.2763,
+      "step": 4446
+    },
+    {
+      "epoch": 2.1247237321545906,
+      "grad_norm": 0.4982314792290385,
+      "learning_rate": 4.309021096693069e-06,
+      "loss": 0.2623,
+      "step": 4447
+    },
+    {
+      "epoch": 2.1252016008601635,
+      "grad_norm": 0.5098348084219093,
+      "learning_rate": 4.3046912295174015e-06,
+      "loss": 0.2598,
+      "step": 4448
+    },
+    {
+      "epoch": 2.125679469565737,
+      "grad_norm": 0.4748754501419402,
+      "learning_rate": 4.300362942119929e-06,
+      "loss": 0.2707,
+      "step": 4449
+    },
+    {
+      "epoch": 2.12615733827131,
+      "grad_norm": 0.47655181277542424,
+      "learning_rate": 4.296036235701235e-06,
+      "loss": 0.2738,
+      "step": 4450
+    },
+    {
+      "epoch": 2.126635206976883,
+      "grad_norm": 0.7574458511584494,
+      "learning_rate": 4.29171111146147e-06,
+      "loss": 0.2991,
+      "step": 4451
+    },
+    {
+      "epoch": 2.1271130756824563,
+      "grad_norm": 0.47222019908644236,
+      "learning_rate": 4.2873875706003535e-06,
+      "loss": 0.2535,
+      "step": 4452
+    },
+    {
+      "epoch": 2.127590944388029,
+      "grad_norm": 0.45799953680867805,
+      "learning_rate": 4.283065614317156e-06,
+      "loss": 0.2612,
+      "step": 4453
+    },
+    {
+      "epoch": 2.1280688130936025,
+      "grad_norm": 0.522281047897022,
+      "learning_rate": 4.278745243810709e-06,
+      "loss": 0.2844,
+      "step": 4454
+    },
+    {
+      "epoch": 2.128546681799176,
+      "grad_norm": 0.4660306281172418,
+      "learning_rate": 4.274426460279412e-06,
+      "loss": 0.2466,
+      "step": 4455
+    },
+    {
+      "epoch": 2.1290245505047487,
+      "grad_norm": 0.4785082179915365,
+      "learning_rate": 4.270109264921221e-06,
+      "loss": 0.2797,
+      "step": 4456
+    },
+    {
+      "epoch": 2.129502419210322,
+      "grad_norm": 0.48055449069442646,
+      "learning_rate": 4.26579365893364e-06,
+      "loss": 0.2984,
+      "step": 4457
+    },
+    {
+      "epoch": 2.1299802879158953,
+      "grad_norm": 0.5098722373675344,
+      "learning_rate": 4.261479643513753e-06,
+      "loss": 0.26,
+      "step": 4458
+    },
+    {
+      "epoch": 2.130458156621468,
+      "grad_norm": 0.49399059503460835,
+      "learning_rate": 4.257167219858187e-06,
+      "loss": 0.2683,
+      "step": 4459
+    },
+    {
+      "epoch": 2.1309360253270415,
+      "grad_norm": 0.5080365898455398,
+      "learning_rate": 4.252856389163128e-06,
+      "loss": 0.2754,
+      "step": 4460
+    },
+    {
+      "epoch": 2.1314138940326144,
+      "grad_norm": 0.5554438469805086,
+      "learning_rate": 4.248547152624334e-06,
+      "loss": 0.2601,
+      "step": 4461
+    },
+    {
+      "epoch": 2.1318917627381877,
+      "grad_norm": 0.48601767105637644,
+      "learning_rate": 4.244239511437105e-06,
+      "loss": 0.2734,
+      "step": 4462
+    },
+    {
+      "epoch": 2.132369631443761,
+      "grad_norm": 0.46636918648450326,
+      "learning_rate": 4.239933466796301e-06,
+      "loss": 0.2573,
+      "step": 4463
+    },
+    {
+      "epoch": 2.132847500149334,
+      "grad_norm": 0.4882887183192489,
+      "learning_rate": 4.235629019896352e-06,
+      "loss": 0.2736,
+      "step": 4464
+    },
+    {
+      "epoch": 2.133325368854907,
+      "grad_norm": 0.4776658176673478,
+      "learning_rate": 4.231326171931231e-06,
+      "loss": 0.2692,
+      "step": 4465
+    },
+    {
+      "epoch": 2.13380323756048,
+      "grad_norm": 0.4866846981727982,
+      "learning_rate": 4.227024924094469e-06,
+      "loss": 0.2577,
+      "step": 4466
+    },
+    {
+      "epoch": 2.1342811062660534,
+      "grad_norm": 0.4980447283271286,
+      "learning_rate": 4.222725277579164e-06,
+      "loss": 0.2876,
+      "step": 4467
+    },
+    {
+      "epoch": 2.1347589749716267,
+      "grad_norm": 0.4623292009049184,
+      "learning_rate": 4.218427233577956e-06,
+      "loss": 0.2736,
+      "step": 4468
+    },
+    {
+      "epoch": 2.1352368436771996,
+      "grad_norm": 0.48084141681866016,
+      "learning_rate": 4.214130793283046e-06,
+      "loss": 0.2801,
+      "step": 4469
+    },
+    {
+      "epoch": 2.135714712382773,
+      "grad_norm": 0.4649177248019152,
+      "learning_rate": 4.209835957886196e-06,
+      "loss": 0.2729,
+      "step": 4470
+    },
+    {
+      "epoch": 2.1361925810883458,
+      "grad_norm": 0.47322585157208374,
+      "learning_rate": 4.205542728578714e-06,
+      "loss": 0.2757,
+      "step": 4471
+    },
+    {
+      "epoch": 2.136670449793919,
+      "grad_norm": 0.4583877785505501,
+      "learning_rate": 4.2012511065514636e-06,
+      "loss": 0.2689,
+      "step": 4472
+    },
+    {
+      "epoch": 2.1371483184994924,
+      "grad_norm": 0.4870399136505855,
+      "learning_rate": 4.196961092994871e-06,
+      "loss": 0.2805,
+      "step": 4473
+    },
+    {
+      "epoch": 2.1376261872050653,
+      "grad_norm": 0.4685135597479858,
+      "learning_rate": 4.192672689098908e-06,
+      "loss": 0.2648,
+      "step": 4474
+    },
+    {
+      "epoch": 2.1381040559106386,
+      "grad_norm": 0.47535234788643843,
+      "learning_rate": 4.188385896053098e-06,
+      "loss": 0.2791,
+      "step": 4475
+    },
+    {
+      "epoch": 2.138581924616212,
+      "grad_norm": 0.5302293391229079,
+      "learning_rate": 4.184100715046529e-06,
+      "loss": 0.2731,
+      "step": 4476
+    },
+    {
+      "epoch": 2.1390597933217848,
+      "grad_norm": 0.5240248943881992,
+      "learning_rate": 4.179817147267829e-06,
+      "loss": 0.2638,
+      "step": 4477
+    },
+    {
+      "epoch": 2.139537662027358,
+      "grad_norm": 0.4619458564887166,
+      "learning_rate": 4.1755351939051845e-06,
+      "loss": 0.2643,
+      "step": 4478
+    },
+    {
+      "epoch": 2.140015530732931,
+      "grad_norm": 0.4702047131010919,
+      "learning_rate": 4.171254856146335e-06,
+      "loss": 0.2615,
+      "step": 4479
+    },
+    {
+      "epoch": 2.1404933994385043,
+      "grad_norm": 0.4773661172330248,
+      "learning_rate": 4.166976135178575e-06,
+      "loss": 0.2865,
+      "step": 4480
+    },
+    {
+      "epoch": 2.1409712681440776,
+      "grad_norm": 0.47727793467693286,
+      "learning_rate": 4.1626990321887425e-06,
+      "loss": 0.2621,
+      "step": 4481
+    },
+    {
+      "epoch": 2.1414491368496504,
+      "grad_norm": 0.4871383170567948,
+      "learning_rate": 4.1584235483632265e-06,
+      "loss": 0.2651,
+      "step": 4482
+    },
+    {
+      "epoch": 2.1419270055552238,
+      "grad_norm": 0.4714614885059131,
+      "learning_rate": 4.154149684887977e-06,
+      "loss": 0.2794,
+      "step": 4483
+    },
+    {
+      "epoch": 2.142404874260797,
+      "grad_norm": 0.4969507858614288,
+      "learning_rate": 4.149877442948486e-06,
+      "loss": 0.2622,
+      "step": 4484
+    },
+    {
+      "epoch": 2.14288274296637,
+      "grad_norm": 0.507286683219254,
+      "learning_rate": 4.1456068237297964e-06,
+      "loss": 0.2664,
+      "step": 4485
+    },
+    {
+      "epoch": 2.1433606116719432,
+      "grad_norm": 0.4810756637158575,
+      "learning_rate": 4.1413378284165065e-06,
+      "loss": 0.2584,
+      "step": 4486
+    },
+    {
+      "epoch": 2.143838480377516,
+      "grad_norm": 0.49276173044738886,
+      "learning_rate": 4.1370704581927575e-06,
+      "loss": 0.2645,
+      "step": 4487
+    },
+    {
+      "epoch": 2.1443163490830894,
+      "grad_norm": 0.5062183592646868,
+      "learning_rate": 4.13280471424224e-06,
+      "loss": 0.2681,
+      "step": 4488
+    },
+    {
+      "epoch": 2.1447942177886627,
+      "grad_norm": 0.49099788891531865,
+      "learning_rate": 4.128540597748203e-06,
+      "loss": 0.2761,
+      "step": 4489
+    },
+    {
+      "epoch": 2.1452720864942356,
+      "grad_norm": 0.518176466847488,
+      "learning_rate": 4.124278109893432e-06,
+      "loss": 0.2663,
+      "step": 4490
+    },
+    {
+      "epoch": 2.145749955199809,
+      "grad_norm": 0.49131335284040245,
+      "learning_rate": 4.120017251860266e-06,
+      "loss": 0.2871,
+      "step": 4491
+    },
+    {
+      "epoch": 2.146227823905382,
+      "grad_norm": 0.5072033097189093,
+      "learning_rate": 4.115758024830595e-06,
+      "loss": 0.2814,
+      "step": 4492
+    },
+    {
+      "epoch": 2.146705692610955,
+      "grad_norm": 0.4876564849105318,
+      "learning_rate": 4.111500429985853e-06,
+      "loss": 0.2759,
+      "step": 4493
+    },
+    {
+      "epoch": 2.1471835613165284,
+      "grad_norm": 0.4421916536653529,
+      "learning_rate": 4.1072444685070155e-06,
+      "loss": 0.2618,
+      "step": 4494
+    },
+    {
+      "epoch": 2.1476614300221013,
+      "grad_norm": 0.46331905163767734,
+      "learning_rate": 4.10299014157462e-06,
+      "loss": 0.2763,
+      "step": 4495
+    },
+    {
+      "epoch": 2.1481392987276746,
+      "grad_norm": 0.5158897747210883,
+      "learning_rate": 4.098737450368738e-06,
+      "loss": 0.2773,
+      "step": 4496
+    },
+    {
+      "epoch": 2.1486171674332475,
+      "grad_norm": 0.45891199284426043,
+      "learning_rate": 4.094486396068987e-06,
+      "loss": 0.2774,
+      "step": 4497
+    },
+    {
+      "epoch": 2.149095036138821,
+      "grad_norm": 0.47523435146742227,
+      "learning_rate": 4.0902369798545426e-06,
+      "loss": 0.2551,
+      "step": 4498
+    },
+    {
+      "epoch": 2.149572904844394,
+      "grad_norm": 0.44933432383147964,
+      "learning_rate": 4.085989202904113e-06,
+      "loss": 0.2606,
+      "step": 4499
+    },
+    {
+      "epoch": 2.150050773549967,
+      "grad_norm": 0.47676276634177017,
+      "learning_rate": 4.0817430663959536e-06,
+      "loss": 0.2785,
+      "step": 4500
+    },
+    {
+      "epoch": 2.1505286422555403,
+      "grad_norm": 0.45939776538275545,
+      "learning_rate": 4.077498571507874e-06,
+      "loss": 0.2714,
+      "step": 4501
+    },
+    {
+      "epoch": 2.1510065109611136,
+      "grad_norm": 0.4702734224108815,
+      "learning_rate": 4.073255719417221e-06,
+      "loss": 0.2615,
+      "step": 4502
+    },
+    {
+      "epoch": 2.1514843796666865,
+      "grad_norm": 0.4910412013612873,
+      "learning_rate": 4.0690145113008815e-06,
+      "loss": 0.2566,
+      "step": 4503
+    },
+    {
+      "epoch": 2.15196224837226,
+      "grad_norm": 0.48456487516342267,
+      "learning_rate": 4.064774948335299e-06,
+      "loss": 0.2683,
+      "step": 4504
+    },
+    {
+      "epoch": 2.1524401170778327,
+      "grad_norm": 0.4873630350755374,
+      "learning_rate": 4.06053703169645e-06,
+      "loss": 0.2628,
+      "step": 4505
+    },
+    {
+      "epoch": 2.152917985783406,
+      "grad_norm": 0.512860605368886,
+      "learning_rate": 4.056300762559855e-06,
+      "loss": 0.267,
+      "step": 4506
+    },
+    {
+      "epoch": 2.1533958544889793,
+      "grad_norm": 0.5806471029367954,
+      "learning_rate": 4.052066142100587e-06,
+      "loss": 0.2632,
+      "step": 4507
+    },
+    {
+      "epoch": 2.153873723194552,
+      "grad_norm": 0.4748345500291511,
+      "learning_rate": 4.047833171493251e-06,
+      "loss": 0.2567,
+      "step": 4508
+    },
+    {
+      "epoch": 2.1543515919001255,
+      "grad_norm": 0.4789611805358055,
+      "learning_rate": 4.043601851911996e-06,
+      "loss": 0.2696,
+      "step": 4509
+    },
+    {
+      "epoch": 2.154829460605699,
+      "grad_norm": 0.49413914811960363,
+      "learning_rate": 4.039372184530521e-06,
+      "loss": 0.2575,
+      "step": 4510
+    },
+    {
+      "epoch": 2.1553073293112717,
+      "grad_norm": 0.5070625267068553,
+      "learning_rate": 4.035144170522055e-06,
+      "loss": 0.2765,
+      "step": 4511
+    },
+    {
+      "epoch": 2.155785198016845,
+      "grad_norm": 0.45705266978001025,
+      "learning_rate": 4.030917811059378e-06,
+      "loss": 0.2705,
+      "step": 4512
+    },
+    {
+      "epoch": 2.156263066722418,
+      "grad_norm": 1.5569910667354723,
+      "learning_rate": 4.02669310731481e-06,
+      "loss": 0.2803,
+      "step": 4513
+    },
+    {
+      "epoch": 2.156740935427991,
+      "grad_norm": 0.49066820686495577,
+      "learning_rate": 4.0224700604602085e-06,
+      "loss": 0.2786,
+      "step": 4514
+    },
+    {
+      "epoch": 2.1572188041335645,
+      "grad_norm": 0.47471065389215006,
+      "learning_rate": 4.0182486716669656e-06,
+      "loss": 0.2566,
+      "step": 4515
+    },
+    {
+      "epoch": 2.1576966728391374,
+      "grad_norm": 0.46201303993122544,
+      "learning_rate": 4.014028942106028e-06,
+      "loss": 0.2644,
+      "step": 4516
+    },
+    {
+      "epoch": 2.1581745415447107,
+      "grad_norm": 0.4810938505072536,
+      "learning_rate": 4.009810872947873e-06,
+      "loss": 0.2723,
+      "step": 4517
+    },
+    {
+      "epoch": 2.1586524102502835,
+      "grad_norm": 0.45055790887900427,
+      "learning_rate": 4.005594465362512e-06,
+      "loss": 0.2731,
+      "step": 4518
+    },
+    {
+      "epoch": 2.159130278955857,
+      "grad_norm": 0.5023239987249768,
+      "learning_rate": 4.00137972051951e-06,
+      "loss": 0.2721,
+      "step": 4519
+    },
+    {
+      "epoch": 2.15960814766143,
+      "grad_norm": 0.507580098147782,
+      "learning_rate": 3.9971666395879605e-06,
+      "loss": 0.2635,
+      "step": 4520
+    },
+    {
+      "epoch": 2.160086016367003,
+      "grad_norm": 0.47302338046197384,
+      "learning_rate": 3.992955223736493e-06,
+      "loss": 0.2621,
+      "step": 4521
+    },
+    {
+      "epoch": 2.1605638850725764,
+      "grad_norm": 0.4426581792894162,
+      "learning_rate": 3.9887454741332874e-06,
+      "loss": 0.2738,
+      "step": 4522
+    },
+    {
+      "epoch": 2.161041753778149,
+      "grad_norm": 0.5038924092723644,
+      "learning_rate": 3.984537391946051e-06,
+      "loss": 0.2859,
+      "step": 4523
+    },
+    {
+      "epoch": 2.1615196224837225,
+      "grad_norm": 0.5097699500822241,
+      "learning_rate": 3.980330978342027e-06,
+      "loss": 0.2836,
+      "step": 4524
+    },
+    {
+      "epoch": 2.161997491189296,
+      "grad_norm": 0.4997990150539291,
+      "learning_rate": 3.9761262344880096e-06,
+      "loss": 0.2862,
+      "step": 4525
+    },
+    {
+      "epoch": 2.1624753598948687,
+      "grad_norm": 0.5068009705170106,
+      "learning_rate": 3.971923161550314e-06,
+      "loss": 0.2617,
+      "step": 4526
+    },
+    {
+      "epoch": 2.162953228600442,
+      "grad_norm": 0.5121923973953294,
+      "learning_rate": 3.967721760694796e-06,
+      "loss": 0.2677,
+      "step": 4527
+    },
+    {
+      "epoch": 2.1634310973060153,
+      "grad_norm": 0.4586590918299566,
+      "learning_rate": 3.963522033086858e-06,
+      "loss": 0.2812,
+      "step": 4528
+    },
+    {
+      "epoch": 2.163908966011588,
+      "grad_norm": 0.48386944735308773,
+      "learning_rate": 3.959323979891427e-06,
+      "loss": 0.2657,
+      "step": 4529
+    },
+    {
+      "epoch": 2.1643868347171615,
+      "grad_norm": 0.48660859734621076,
+      "learning_rate": 3.9551276022729644e-06,
+      "loss": 0.292,
+      "step": 4530
+    },
+    {
+      "epoch": 2.1648647034227344,
+      "grad_norm": 0.571916496126486,
+      "learning_rate": 3.9509329013954775e-06,
+      "loss": 0.2669,
+      "step": 4531
+    },
+    {
+      "epoch": 2.1653425721283077,
+      "grad_norm": 0.4731522622117423,
+      "learning_rate": 3.946739878422502e-06,
+      "loss": 0.2674,
+      "step": 4532
+    },
+    {
+      "epoch": 2.165820440833881,
+      "grad_norm": 0.4585592677843948,
+      "learning_rate": 3.942548534517102e-06,
+      "loss": 0.2754,
+      "step": 4533
+    },
+    {
+      "epoch": 2.166298309539454,
+      "grad_norm": 0.45992994288587097,
+      "learning_rate": 3.938358870841891e-06,
+      "loss": 0.2842,
+      "step": 4534
+    },
+    {
+      "epoch": 2.166776178245027,
+      "grad_norm": 0.48157396171772915,
+      "learning_rate": 3.9341708885590034e-06,
+      "loss": 0.2721,
+      "step": 4535
+    },
+    {
+      "epoch": 2.1672540469506005,
+      "grad_norm": 0.46636131218315996,
+      "learning_rate": 3.9299845888301084e-06,
+      "loss": 0.2545,
+      "step": 4536
+    },
+    {
+      "epoch": 2.1677319156561734,
+      "grad_norm": 0.48593748291916344,
+      "learning_rate": 3.925799972816419e-06,
+      "loss": 0.2478,
+      "step": 4537
+    },
+    {
+      "epoch": 2.1682097843617467,
+      "grad_norm": 0.5044885208397369,
+      "learning_rate": 3.921617041678669e-06,
+      "loss": 0.2679,
+      "step": 4538
+    },
+    {
+      "epoch": 2.1686876530673196,
+      "grad_norm": 0.4667942698450788,
+      "learning_rate": 3.917435796577128e-06,
+      "loss": 0.2522,
+      "step": 4539
+    },
+    {
+      "epoch": 2.169165521772893,
+      "grad_norm": 0.4949207913026006,
+      "learning_rate": 3.913256238671607e-06,
+      "loss": 0.2694,
+      "step": 4540
+    },
+    {
+      "epoch": 2.169643390478466,
+      "grad_norm": 0.4922718382704214,
+      "learning_rate": 3.909078369121435e-06,
+      "loss": 0.272,
+      "step": 4541
+    },
+    {
+      "epoch": 2.170121259184039,
+      "grad_norm": 0.5568367688738352,
+      "learning_rate": 3.904902189085479e-06,
+      "loss": 0.2744,
+      "step": 4542
+    },
+    {
+      "epoch": 2.1705991278896124,
+      "grad_norm": 0.4687165591772829,
+      "learning_rate": 3.900727699722144e-06,
+      "loss": 0.2758,
+      "step": 4543
+    },
+    {
+      "epoch": 2.1710769965951853,
+      "grad_norm": 0.45872513093731365,
+      "learning_rate": 3.896554902189355e-06,
+      "loss": 0.2723,
+      "step": 4544
+    },
+    {
+      "epoch": 2.1715548653007586,
+      "grad_norm": 0.4637298698793799,
+      "learning_rate": 3.89238379764457e-06,
+      "loss": 0.2448,
+      "step": 4545
+    },
+    {
+      "epoch": 2.172032734006332,
+      "grad_norm": 0.48363038021481214,
+      "learning_rate": 3.888214387244783e-06,
+      "loss": 0.2657,
+      "step": 4546
+    },
+    {
+      "epoch": 2.1725106027119048,
+      "grad_norm": 0.4550871624285722,
+      "learning_rate": 3.884046672146518e-06,
+      "loss": 0.2854,
+      "step": 4547
+    },
+    {
+      "epoch": 2.172988471417478,
+      "grad_norm": 0.495376118420438,
+      "learning_rate": 3.879880653505824e-06,
+      "loss": 0.2633,
+      "step": 4548
+    },
+    {
+      "epoch": 2.1734663401230514,
+      "grad_norm": 0.46098656670137583,
+      "learning_rate": 3.875716332478275e-06,
+      "loss": 0.2657,
+      "step": 4549
+    },
+    {
+      "epoch": 2.1739442088286243,
+      "grad_norm": 0.4751772768586887,
+      "learning_rate": 3.871553710218988e-06,
+      "loss": 0.2837,
+      "step": 4550
+    },
+    {
+      "epoch": 2.1744220775341976,
+      "grad_norm": 0.47152321412549136,
+      "learning_rate": 3.867392787882599e-06,
+      "loss": 0.2638,
+      "step": 4551
+    },
+    {
+      "epoch": 2.1748999462397705,
+      "grad_norm": 0.44704792893751294,
+      "learning_rate": 3.8632335666232686e-06,
+      "loss": 0.2602,
+      "step": 4552
+    },
+    {
+      "epoch": 2.1753778149453438,
+      "grad_norm": 0.6555651445287016,
+      "learning_rate": 3.859076047594701e-06,
+      "loss": 0.2518,
+      "step": 4553
+    },
+    {
+      "epoch": 2.175855683650917,
+      "grad_norm": 0.4869920295121113,
+      "learning_rate": 3.854920231950113e-06,
+      "loss": 0.2821,
+      "step": 4554
+    },
+    {
+      "epoch": 2.17633355235649,
+      "grad_norm": 0.45671533130726394,
+      "learning_rate": 3.850766120842252e-06,
+      "loss": 0.2801,
+      "step": 4555
+    },
+    {
+      "epoch": 2.1768114210620633,
+      "grad_norm": 0.4608951885566162,
+      "learning_rate": 3.846613715423402e-06,
+      "loss": 0.2658,
+      "step": 4556
+    },
+    {
+      "epoch": 2.177289289767636,
+      "grad_norm": 0.46418435619628584,
+      "learning_rate": 3.842463016845362e-06,
+      "loss": 0.2735,
+      "step": 4557
+    },
+    {
+      "epoch": 2.1777671584732095,
+      "grad_norm": 0.47932679730203265,
+      "learning_rate": 3.838314026259462e-06,
+      "loss": 0.25,
+      "step": 4558
+    },
+    {
+      "epoch": 2.1782450271787828,
+      "grad_norm": 0.47237247496031814,
+      "learning_rate": 3.8341667448165645e-06,
+      "loss": 0.262,
+      "step": 4559
+    },
+    {
+      "epoch": 2.1787228958843556,
+      "grad_norm": 0.5186757703706331,
+      "learning_rate": 3.830021173667048e-06,
+      "loss": 0.2884,
+      "step": 4560
+    },
+    {
+      "epoch": 2.179200764589929,
+      "grad_norm": 0.44583744351159743,
+      "learning_rate": 3.8258773139608185e-06,
+      "loss": 0.2725,
+      "step": 4561
+    },
+    {
+      "epoch": 2.1796786332955023,
+      "grad_norm": 0.4772180156318942,
+      "learning_rate": 3.821735166847316e-06,
+      "loss": 0.2641,
+      "step": 4562
+    },
+    {
+      "epoch": 2.180156502001075,
+      "grad_norm": 0.5094816309115656,
+      "learning_rate": 3.817594733475494e-06,
+      "loss": 0.2657,
+      "step": 4563
+    },
+    {
+      "epoch": 2.1806343707066485,
+      "grad_norm": 0.4456377533344565,
+      "learning_rate": 3.813456014993835e-06,
+      "loss": 0.2637,
+      "step": 4564
+    },
+    {
+      "epoch": 2.1811122394122213,
+      "grad_norm": 0.47505222662407803,
+      "learning_rate": 3.809319012550352e-06,
+      "loss": 0.2713,
+      "step": 4565
+    },
+    {
+      "epoch": 2.1815901081177946,
+      "grad_norm": 0.5173681589868041,
+      "learning_rate": 3.8051837272925728e-06,
+      "loss": 0.2813,
+      "step": 4566
+    },
+    {
+      "epoch": 2.182067976823368,
+      "grad_norm": 0.5015740347460304,
+      "learning_rate": 3.80105016036755e-06,
+      "loss": 0.2708,
+      "step": 4567
+    },
+    {
+      "epoch": 2.182545845528941,
+      "grad_norm": 0.45211083979868544,
+      "learning_rate": 3.796918312921868e-06,
+      "loss": 0.2774,
+      "step": 4568
+    },
+    {
+      "epoch": 2.183023714234514,
+      "grad_norm": 0.4580673412053882,
+      "learning_rate": 3.792788186101626e-06,
+      "loss": 0.2786,
+      "step": 4569
+    },
+    {
+      "epoch": 2.183501582940087,
+      "grad_norm": 0.483142068598242,
+      "learning_rate": 3.788659781052444e-06,
+      "loss": 0.2662,
+      "step": 4570
+    },
+    {
+      "epoch": 2.1839794516456603,
+      "grad_norm": 0.4585071424628377,
+      "learning_rate": 3.7845330989194762e-06,
+      "loss": 0.2537,
+      "step": 4571
+    },
+    {
+      "epoch": 2.1844573203512336,
+      "grad_norm": 0.46364154074230174,
+      "learning_rate": 3.780408140847387e-06,
+      "loss": 0.2703,
+      "step": 4572
+    },
+    {
+      "epoch": 2.1849351890568065,
+      "grad_norm": 0.47912865130428883,
+      "learning_rate": 3.7762849079803654e-06,
+      "loss": 0.2599,
+      "step": 4573
+    },
+    {
+      "epoch": 2.18541305776238,
+      "grad_norm": 0.486225553510062,
+      "learning_rate": 3.772163401462129e-06,
+      "loss": 0.2707,
+      "step": 4574
+    },
+    {
+      "epoch": 2.185890926467953,
+      "grad_norm": 0.5017350180630475,
+      "learning_rate": 3.7680436224359084e-06,
+      "loss": 0.278,
+      "step": 4575
+    },
+    {
+      "epoch": 2.186368795173526,
+      "grad_norm": 0.44633892621886945,
+      "learning_rate": 3.7639255720444532e-06,
+      "loss": 0.2564,
+      "step": 4576
+    },
+    {
+      "epoch": 2.1868466638790993,
+      "grad_norm": 0.47140634322880026,
+      "learning_rate": 3.7598092514300456e-06,
+      "loss": 0.2706,
+      "step": 4577
+    },
+    {
+      "epoch": 2.187324532584672,
+      "grad_norm": 0.47775263124381995,
+      "learning_rate": 3.7556946617344757e-06,
+      "loss": 0.2768,
+      "step": 4578
+    },
+    {
+      "epoch": 2.1878024012902455,
+      "grad_norm": 0.4662918039512788,
+      "learning_rate": 3.751581804099056e-06,
+      "loss": 0.2702,
+      "step": 4579
+    },
+    {
+      "epoch": 2.188280269995819,
+      "grad_norm": 0.4526279489034841,
+      "learning_rate": 3.747470679664624e-06,
+      "loss": 0.27,
+      "step": 4580
+    },
+    {
+      "epoch": 2.1887581387013917,
+      "grad_norm": 0.461833024668242,
+      "learning_rate": 3.7433612895715356e-06,
+      "loss": 0.2525,
+      "step": 4581
+    },
+    {
+      "epoch": 2.189236007406965,
+      "grad_norm": 0.45967111066950117,
+      "learning_rate": 3.739253634959661e-06,
+      "loss": 0.2755,
+      "step": 4582
+    },
+    {
+      "epoch": 2.1897138761125383,
+      "grad_norm": 0.4888680597465754,
+      "learning_rate": 3.735147716968386e-06,
+      "loss": 0.2578,
+      "step": 4583
+    },
+    {
+      "epoch": 2.190191744818111,
+      "grad_norm": 0.4986060150836118,
+      "learning_rate": 3.731043536736628e-06,
+      "loss": 0.2694,
+      "step": 4584
+    },
+    {
+      "epoch": 2.1906696135236845,
+      "grad_norm": 0.4543474936433639,
+      "learning_rate": 3.7269410954028107e-06,
+      "loss": 0.2592,
+      "step": 4585
+    },
+    {
+      "epoch": 2.1911474822292574,
+      "grad_norm": 1.2235953943030329,
+      "learning_rate": 3.7228403941048753e-06,
+      "loss": 0.2595,
+      "step": 4586
+    },
+    {
+      "epoch": 2.1916253509348307,
+      "grad_norm": 0.48594813503110623,
+      "learning_rate": 3.7187414339802906e-06,
+      "loss": 0.2774,
+      "step": 4587
+    },
+    {
+      "epoch": 2.192103219640404,
+      "grad_norm": 0.4911464829074915,
+      "learning_rate": 3.7146442161660336e-06,
+      "loss": 0.2636,
+      "step": 4588
+    },
+    {
+      "epoch": 2.192581088345977,
+      "grad_norm": 0.47380456620012534,
+      "learning_rate": 3.710548741798594e-06,
+      "loss": 0.265,
+      "step": 4589
+    },
+    {
+      "epoch": 2.19305895705155,
+      "grad_norm": 0.4706478201898707,
+      "learning_rate": 3.706455012013994e-06,
+      "loss": 0.2758,
+      "step": 4590
+    },
+    {
+      "epoch": 2.193536825757123,
+      "grad_norm": 0.4635842650188313,
+      "learning_rate": 3.702363027947757e-06,
+      "loss": 0.2703,
+      "step": 4591
+    },
+    {
+      "epoch": 2.1940146944626964,
+      "grad_norm": 0.4655317804125815,
+      "learning_rate": 3.6982727907349247e-06,
+      "loss": 0.2718,
+      "step": 4592
+    },
+    {
+      "epoch": 2.1944925631682697,
+      "grad_norm": 0.4495126115308564,
+      "learning_rate": 3.694184301510063e-06,
+      "loss": 0.2737,
+      "step": 4593
+    },
+    {
+      "epoch": 2.1949704318738426,
+      "grad_norm": 0.4535524871952706,
+      "learning_rate": 3.6900975614072433e-06,
+      "loss": 0.2621,
+      "step": 4594
+    },
+    {
+      "epoch": 2.195448300579416,
+      "grad_norm": 0.44992841504040854,
+      "learning_rate": 3.6860125715600513e-06,
+      "loss": 0.2805,
+      "step": 4595
+    },
+    {
+      "epoch": 2.1959261692849887,
+      "grad_norm": 0.45869083287966966,
+      "learning_rate": 3.6819293331015993e-06,
+      "loss": 0.2865,
+      "step": 4596
+    },
+    {
+      "epoch": 2.196404037990562,
+      "grad_norm": 0.46107098925354134,
+      "learning_rate": 3.6778478471645008e-06,
+      "loss": 0.2624,
+      "step": 4597
+    },
+    {
+      "epoch": 2.1968819066961354,
+      "grad_norm": 0.44462298567559727,
+      "learning_rate": 3.6737681148808855e-06,
+      "loss": 0.2602,
+      "step": 4598
+    },
+    {
+      "epoch": 2.1973597754017082,
+      "grad_norm": 0.4596846501814117,
+      "learning_rate": 3.6696901373824056e-06,
+      "loss": 0.2721,
+      "step": 4599
+    },
+    {
+      "epoch": 2.1978376441072816,
+      "grad_norm": 0.6638891345781536,
+      "learning_rate": 3.665613915800217e-06,
+      "loss": 0.261,
+      "step": 4600
+    },
+    {
+      "epoch": 2.198315512812855,
+      "grad_norm": 0.49610959186848946,
+      "learning_rate": 3.6615394512649884e-06,
+      "loss": 0.2774,
+      "step": 4601
+    },
+    {
+      "epoch": 2.1987933815184277,
+      "grad_norm": 0.5056345462489438,
+      "learning_rate": 3.65746674490691e-06,
+      "loss": 0.2835,
+      "step": 4602
+    },
+    {
+      "epoch": 2.199271250224001,
+      "grad_norm": 0.44612162373312525,
+      "learning_rate": 3.6533957978556777e-06,
+      "loss": 0.2798,
+      "step": 4603
+    },
+    {
+      "epoch": 2.199749118929574,
+      "grad_norm": 0.4722815796388612,
+      "learning_rate": 3.6493266112404947e-06,
+      "loss": 0.2571,
+      "step": 4604
+    },
+    {
+      "epoch": 2.2002269876351472,
+      "grad_norm": 0.4719966235892821,
+      "learning_rate": 3.6452591861900886e-06,
+      "loss": 0.2504,
+      "step": 4605
+    },
+    {
+      "epoch": 2.2007048563407205,
+      "grad_norm": 0.4738688790143464,
+      "learning_rate": 3.641193523832689e-06,
+      "loss": 0.2642,
+      "step": 4606
+    },
+    {
+      "epoch": 2.2011827250462934,
+      "grad_norm": 0.5033486532463676,
+      "learning_rate": 3.637129625296035e-06,
+      "loss": 0.2634,
+      "step": 4607
+    },
+    {
+      "epoch": 2.2016605937518667,
+      "grad_norm": 0.4960105505883145,
+      "learning_rate": 3.633067491707387e-06,
+      "loss": 0.2978,
+      "step": 4608
+    },
+    {
+      "epoch": 2.20213846245744,
+      "grad_norm": 0.47252895928191396,
+      "learning_rate": 3.6290071241935067e-06,
+      "loss": 0.2712,
+      "step": 4609
+    },
+    {
+      "epoch": 2.202616331163013,
+      "grad_norm": 0.46564634687060924,
+      "learning_rate": 3.6249485238806637e-06,
+      "loss": 0.2749,
+      "step": 4610
+    },
+    {
+      "epoch": 2.2030941998685862,
+      "grad_norm": 0.47341686338285643,
+      "learning_rate": 3.62089169189465e-06,
+      "loss": 0.2481,
+      "step": 4611
+    },
+    {
+      "epoch": 2.203572068574159,
+      "grad_norm": 0.4813130606312593,
+      "learning_rate": 3.6168366293607526e-06,
+      "loss": 0.2686,
+      "step": 4612
+    },
+    {
+      "epoch": 2.2040499372797324,
+      "grad_norm": 0.4572672198704176,
+      "learning_rate": 3.612783337403776e-06,
+      "loss": 0.2702,
+      "step": 4613
+    },
+    {
+      "epoch": 2.2045278059853057,
+      "grad_norm": 0.4506090325494477,
+      "learning_rate": 3.6087318171480368e-06,
+      "loss": 0.2668,
+      "step": 4614
+    },
+    {
+      "epoch": 2.2050056746908786,
+      "grad_norm": 0.47410093157277494,
+      "learning_rate": 3.6046820697173514e-06,
+      "loss": 0.2699,
+      "step": 4615
+    },
+    {
+      "epoch": 2.205483543396452,
+      "grad_norm": 0.46061091915894664,
+      "learning_rate": 3.600634096235046e-06,
+      "loss": 0.2567,
+      "step": 4616
+    },
+    {
+      "epoch": 2.205961412102025,
+      "grad_norm": 0.4868564505945191,
+      "learning_rate": 3.596587897823962e-06,
+      "loss": 0.2779,
+      "step": 4617
+    },
+    {
+      "epoch": 2.206439280807598,
+      "grad_norm": 0.4832363834974607,
+      "learning_rate": 3.59254347560644e-06,
+      "loss": 0.2638,
+      "step": 4618
+    },
+    {
+      "epoch": 2.2069171495131714,
+      "grad_norm": 0.5648718684599673,
+      "learning_rate": 3.58850083070433e-06,
+      "loss": 0.274,
+      "step": 4619
+    },
+    {
+      "epoch": 2.2073950182187443,
+      "grad_norm": 0.4940852949638152,
+      "learning_rate": 3.5844599642389965e-06,
+      "loss": 0.2675,
+      "step": 4620
+    },
+    {
+      "epoch": 2.2078728869243176,
+      "grad_norm": 0.4787783856445962,
+      "learning_rate": 3.5804208773313e-06,
+      "loss": 0.2678,
+      "step": 4621
+    },
+    {
+      "epoch": 2.2083507556298905,
+      "grad_norm": 0.4985708934791662,
+      "learning_rate": 3.576383571101609e-06,
+      "loss": 0.2607,
+      "step": 4622
+    },
+    {
+      "epoch": 2.208828624335464,
+      "grad_norm": 0.4467068890356679,
+      "learning_rate": 3.572348046669809e-06,
+      "loss": 0.2782,
+      "step": 4623
+    },
+    {
+      "epoch": 2.209306493041037,
+      "grad_norm": 0.624896596323991,
+      "learning_rate": 3.5683143051552784e-06,
+      "loss": 0.2758,
+      "step": 4624
+    },
+    {
+      "epoch": 2.20978436174661,
+      "grad_norm": 0.49994003409060306,
+      "learning_rate": 3.564282347676903e-06,
+      "loss": 0.2788,
+      "step": 4625
+    },
+    {
+      "epoch": 2.2102622304521833,
+      "grad_norm": 0.47742176535286496,
+      "learning_rate": 3.560252175353084e-06,
+      "loss": 0.266,
+      "step": 4626
+    },
+    {
+      "epoch": 2.2107400991577566,
+      "grad_norm": 0.47656971421522326,
+      "learning_rate": 3.556223789301716e-06,
+      "loss": 0.2681,
+      "step": 4627
+    },
+    {
+      "epoch": 2.2112179678633295,
+      "grad_norm": 0.569244955801473,
+      "learning_rate": 3.552197190640203e-06,
+      "loss": 0.2668,
+      "step": 4628
+    },
+    {
+      "epoch": 2.211695836568903,
+      "grad_norm": 0.4642389811727502,
+      "learning_rate": 3.5481723804854485e-06,
+      "loss": 0.2682,
+      "step": 4629
+    },
+    {
+      "epoch": 2.2121737052744757,
+      "grad_norm": 0.5017601977034839,
+      "learning_rate": 3.54414935995387e-06,
+      "loss": 0.257,
+      "step": 4630
+    },
+    {
+      "epoch": 2.212651573980049,
+      "grad_norm": 0.4986978616999572,
+      "learning_rate": 3.540128130161381e-06,
+      "loss": 0.2618,
+      "step": 4631
+    },
+    {
+      "epoch": 2.2131294426856223,
+      "grad_norm": 0.6552711808635082,
+      "learning_rate": 3.5361086922233944e-06,
+      "loss": 0.2771,
+      "step": 4632
+    },
+    {
+      "epoch": 2.213607311391195,
+      "grad_norm": 0.5079540763737515,
+      "learning_rate": 3.53209104725484e-06,
+      "loss": 0.2715,
+      "step": 4633
+    },
+    {
+      "epoch": 2.2140851800967685,
+      "grad_norm": 0.5478379669178205,
+      "learning_rate": 3.5280751963701356e-06,
+      "loss": 0.253,
+      "step": 4634
+    },
+    {
+      "epoch": 2.214563048802342,
+      "grad_norm": 0.481691741150192,
+      "learning_rate": 3.524061140683206e-06,
+      "loss": 0.2822,
+      "step": 4635
+    },
+    {
+      "epoch": 2.2150409175079147,
+      "grad_norm": 0.46183826966564606,
+      "learning_rate": 3.520048881307486e-06,
+      "loss": 0.277,
+      "step": 4636
+    },
+    {
+      "epoch": 2.215518786213488,
+      "grad_norm": 0.6334146180967442,
+      "learning_rate": 3.5160384193559017e-06,
+      "loss": 0.2745,
+      "step": 4637
+    },
+    {
+      "epoch": 2.215996654919061,
+      "grad_norm": 0.4842285182462256,
+      "learning_rate": 3.512029755940882e-06,
+      "loss": 0.2682,
+      "step": 4638
+    },
+    {
+      "epoch": 2.216474523624634,
+      "grad_norm": 0.47466868378671784,
+      "learning_rate": 3.5080228921743653e-06,
+      "loss": 0.2902,
+      "step": 4639
+    },
+    {
+      "epoch": 2.2169523923302075,
+      "grad_norm": 0.51214418083174,
+      "learning_rate": 3.5040178291677816e-06,
+      "loss": 0.2522,
+      "step": 4640
+    },
+    {
+      "epoch": 2.2174302610357803,
+      "grad_norm": 0.4665166989520757,
+      "learning_rate": 3.5000145680320617e-06,
+      "loss": 0.265,
+      "step": 4641
+    },
+    {
+      "epoch": 2.2179081297413537,
+      "grad_norm": 0.48603526531215135,
+      "learning_rate": 3.496013109877646e-06,
+      "loss": 0.2772,
+      "step": 4642
+    },
+    {
+      "epoch": 2.2183859984469265,
+      "grad_norm": 0.5575762979046887,
+      "learning_rate": 3.4920134558144645e-06,
+      "loss": 0.2589,
+      "step": 4643
+    },
+    {
+      "epoch": 2.2188638671525,
+      "grad_norm": 0.4895518470618773,
+      "learning_rate": 3.48801560695195e-06,
+      "loss": 0.2759,
+      "step": 4644
+    },
+    {
+      "epoch": 2.219341735858073,
+      "grad_norm": 0.6734863253510791,
+      "learning_rate": 3.4840195643990383e-06,
+      "loss": 0.2593,
+      "step": 4645
+    },
+    {
+      "epoch": 2.219819604563646,
+      "grad_norm": 0.48164159307131443,
+      "learning_rate": 3.4800253292641574e-06,
+      "loss": 0.2496,
+      "step": 4646
+    },
+    {
+      "epoch": 2.2202974732692193,
+      "grad_norm": 0.5194894835045092,
+      "learning_rate": 3.476032902655239e-06,
+      "loss": 0.2731,
+      "step": 4647
+    },
+    {
+      "epoch": 2.220775341974792,
+      "grad_norm": 0.5048689424410697,
+      "learning_rate": 3.4720422856797163e-06,
+      "loss": 0.272,
+      "step": 4648
+    },
+    {
+      "epoch": 2.2212532106803655,
+      "grad_norm": 0.4555324657809056,
+      "learning_rate": 3.468053479444512e-06,
+      "loss": 0.275,
+      "step": 4649
+    },
+    {
+      "epoch": 2.221731079385939,
+      "grad_norm": 0.49021751307121847,
+      "learning_rate": 3.464066485056048e-06,
+      "loss": 0.2744,
+      "step": 4650
+    },
+    {
+      "epoch": 2.2222089480915117,
+      "grad_norm": 0.4410281797101959,
+      "learning_rate": 3.460081303620252e-06,
+      "loss": 0.2757,
+      "step": 4651
+    },
+    {
+      "epoch": 2.222686816797085,
+      "grad_norm": 0.5014942315206841,
+      "learning_rate": 3.4560979362425406e-06,
+      "loss": 0.2797,
+      "step": 4652
+    },
+    {
+      "epoch": 2.2231646855026583,
+      "grad_norm": 0.45504543451237,
+      "learning_rate": 3.452116384027826e-06,
+      "loss": 0.2573,
+      "step": 4653
+    },
+    {
+      "epoch": 2.223642554208231,
+      "grad_norm": 0.4803760937380114,
+      "learning_rate": 3.4481366480805266e-06,
+      "loss": 0.2483,
+      "step": 4654
+    },
+    {
+      "epoch": 2.2241204229138045,
+      "grad_norm": 0.457505611780426,
+      "learning_rate": 3.444158729504549e-06,
+      "loss": 0.2645,
+      "step": 4655
+    },
+    {
+      "epoch": 2.2245982916193774,
+      "grad_norm": 0.49598384831058834,
+      "learning_rate": 3.4401826294032924e-06,
+      "loss": 0.2738,
+      "step": 4656
+    },
+    {
+      "epoch": 2.2250761603249507,
+      "grad_norm": 0.47321887795699297,
+      "learning_rate": 3.436208348879665e-06,
+      "loss": 0.2852,
+      "step": 4657
+    },
+    {
+      "epoch": 2.225554029030524,
+      "grad_norm": 0.47386541058260834,
+      "learning_rate": 3.4322358890360586e-06,
+      "loss": 0.2809,
+      "step": 4658
+    },
+    {
+      "epoch": 2.226031897736097,
+      "grad_norm": 0.4403468594295977,
+      "learning_rate": 3.4282652509743596e-06,
+      "loss": 0.2586,
+      "step": 4659
+    },
+    {
+      "epoch": 2.22650976644167,
+      "grad_norm": 0.48721058760605696,
+      "learning_rate": 3.4242964357959597e-06,
+      "loss": 0.277,
+      "step": 4660
+    },
+    {
+      "epoch": 2.2269876351472435,
+      "grad_norm": 0.5068739043457748,
+      "learning_rate": 3.4203294446017354e-06,
+      "loss": 0.2519,
+      "step": 4661
+    },
+    {
+      "epoch": 2.2274655038528164,
+      "grad_norm": 0.4420653696705306,
+      "learning_rate": 3.416364278492057e-06,
+      "loss": 0.2831,
+      "step": 4662
+    },
+    {
+      "epoch": 2.2279433725583897,
+      "grad_norm": 0.486948816480161,
+      "learning_rate": 3.4124009385667967e-06,
+      "loss": 0.2734,
+      "step": 4663
+    },
+    {
+      "epoch": 2.2284212412639626,
+      "grad_norm": 0.49182635386072504,
+      "learning_rate": 3.408439425925313e-06,
+      "loss": 0.2615,
+      "step": 4664
+    },
+    {
+      "epoch": 2.228899109969536,
+      "grad_norm": 0.4713713292607333,
+      "learning_rate": 3.4044797416664564e-06,
+      "loss": 0.2789,
+      "step": 4665
+    },
+    {
+      "epoch": 2.229376978675109,
+      "grad_norm": 0.4553911870292443,
+      "learning_rate": 3.4005218868885794e-06,
+      "loss": 0.2865,
+      "step": 4666
+    },
+    {
+      "epoch": 2.229854847380682,
+      "grad_norm": 0.48295582603522347,
+      "learning_rate": 3.396565862689518e-06,
+      "loss": 0.2564,
+      "step": 4667
+    },
+    {
+      "epoch": 2.2303327160862554,
+      "grad_norm": 0.5092986006165036,
+      "learning_rate": 3.3926116701666013e-06,
+      "loss": 0.2805,
+      "step": 4668
+    },
+    {
+      "epoch": 2.2308105847918283,
+      "grad_norm": 0.45858410229406726,
+      "learning_rate": 3.3886593104166575e-06,
+      "loss": 0.2815,
+      "step": 4669
+    },
+    {
+      "epoch": 2.2312884534974016,
+      "grad_norm": 0.4947175505998691,
+      "learning_rate": 3.3847087845359996e-06,
+      "loss": 0.2752,
+      "step": 4670
+    },
+    {
+      "epoch": 2.231766322202975,
+      "grad_norm": 0.4978122665367534,
+      "learning_rate": 3.38076009362043e-06,
+      "loss": 0.2818,
+      "step": 4671
+    },
+    {
+      "epoch": 2.2322441909085478,
+      "grad_norm": 0.5749495630859802,
+      "learning_rate": 3.376813238765252e-06,
+      "loss": 0.2616,
+      "step": 4672
+    },
+    {
+      "epoch": 2.232722059614121,
+      "grad_norm": 0.46541839278207525,
+      "learning_rate": 3.3728682210652497e-06,
+      "loss": 0.2738,
+      "step": 4673
+    },
+    {
+      "epoch": 2.233199928319694,
+      "grad_norm": 0.4822660477000455,
+      "learning_rate": 3.3689250416147e-06,
+      "loss": 0.2846,
+      "step": 4674
+    },
+    {
+      "epoch": 2.2336777970252673,
+      "grad_norm": 0.4569814026626145,
+      "learning_rate": 3.364983701507376e-06,
+      "loss": 0.2602,
+      "step": 4675
+    },
+    {
+      "epoch": 2.2341556657308406,
+      "grad_norm": 0.4445416419282075,
+      "learning_rate": 3.361044201836534e-06,
+      "loss": 0.2695,
+      "step": 4676
+    },
+    {
+      "epoch": 2.2346335344364134,
+      "grad_norm": 0.45506836119647587,
+      "learning_rate": 3.357106543694918e-06,
+      "loss": 0.2778,
+      "step": 4677
+    },
+    {
+      "epoch": 2.2351114031419868,
+      "grad_norm": 0.4574261555122098,
+      "learning_rate": 3.3531707281747717e-06,
+      "loss": 0.2693,
+      "step": 4678
+    },
+    {
+      "epoch": 2.23558927184756,
+      "grad_norm": 0.5141167452823477,
+      "learning_rate": 3.3492367563678173e-06,
+      "loss": 0.2765,
+      "step": 4679
+    },
+    {
+      "epoch": 2.236067140553133,
+      "grad_norm": 0.6521885973282003,
+      "learning_rate": 3.3453046293652657e-06,
+      "loss": 0.2638,
+      "step": 4680
+    },
+    {
+      "epoch": 2.2365450092587063,
+      "grad_norm": 0.492272086211248,
+      "learning_rate": 3.3413743482578233e-06,
+      "loss": 0.2608,
+      "step": 4681
+    },
+    {
+      "epoch": 2.237022877964279,
+      "grad_norm": 0.4635845305006642,
+      "learning_rate": 3.337445914135684e-06,
+      "loss": 0.2617,
+      "step": 4682
+    },
+    {
+      "epoch": 2.2375007466698524,
+      "grad_norm": 0.5023980255782854,
+      "learning_rate": 3.3335193280885215e-06,
+      "loss": 0.2761,
+      "step": 4683
+    },
+    {
+      "epoch": 2.2379786153754258,
+      "grad_norm": 0.5343058309706397,
+      "learning_rate": 3.3295945912055006e-06,
+      "loss": 0.2845,
+      "step": 4684
+    },
+    {
+      "epoch": 2.2384564840809986,
+      "grad_norm": 0.45734291677162836,
+      "learning_rate": 3.3256717045752794e-06,
+      "loss": 0.2629,
+      "step": 4685
+    },
+    {
+      "epoch": 2.238934352786572,
+      "grad_norm": 0.4585226946873842,
+      "learning_rate": 3.3217506692859937e-06,
+      "loss": 0.2684,
+      "step": 4686
+    },
+    {
+      "epoch": 2.2394122214921452,
+      "grad_norm": 0.48321193724337996,
+      "learning_rate": 3.317831486425267e-06,
+      "loss": 0.2785,
+      "step": 4687
+    },
+    {
+      "epoch": 2.239890090197718,
+      "grad_norm": 0.49773006035607,
+      "learning_rate": 3.313914157080218e-06,
+      "loss": 0.2723,
+      "step": 4688
+    },
+    {
+      "epoch": 2.2403679589032914,
+      "grad_norm": 0.4908921121030998,
+      "learning_rate": 3.3099986823374407e-06,
+      "loss": 0.264,
+      "step": 4689
+    },
+    {
+      "epoch": 2.2408458276088643,
+      "grad_norm": 0.5288018646857777,
+      "learning_rate": 3.3060850632830167e-06,
+      "loss": 0.2614,
+      "step": 4690
+    },
+    {
+      "epoch": 2.2413236963144376,
+      "grad_norm": 0.4422685276222856,
+      "learning_rate": 3.3021733010025203e-06,
+      "loss": 0.268,
+      "step": 4691
+    },
+    {
+      "epoch": 2.241801565020011,
+      "grad_norm": 0.4989573893125483,
+      "learning_rate": 3.298263396581003e-06,
+      "loss": 0.266,
+      "step": 4692
+    },
+    {
+      "epoch": 2.242279433725584,
+      "grad_norm": 0.48611829997172,
+      "learning_rate": 3.294355351102999e-06,
+      "loss": 0.2527,
+      "step": 4693
+    },
+    {
+      "epoch": 2.242757302431157,
+      "grad_norm": 0.4975978815944788,
+      "learning_rate": 3.2904491656525396e-06,
+      "loss": 0.2634,
+      "step": 4694
+    },
+    {
+      "epoch": 2.24323517113673,
+      "grad_norm": 0.47584888029570965,
+      "learning_rate": 3.286544841313126e-06,
+      "loss": 0.2589,
+      "step": 4695
+    },
+    {
+      "epoch": 2.2437130398423033,
+      "grad_norm": 0.4684430376922013,
+      "learning_rate": 3.2826423791677475e-06,
+      "loss": 0.2665,
+      "step": 4696
+    },
+    {
+      "epoch": 2.2441909085478766,
+      "grad_norm": 0.4697921966643509,
+      "learning_rate": 3.278741780298883e-06,
+      "loss": 0.2684,
+      "step": 4697
+    },
+    {
+      "epoch": 2.2446687772534495,
+      "grad_norm": 0.4471854069336229,
+      "learning_rate": 3.2748430457884883e-06,
+      "loss": 0.2858,
+      "step": 4698
+    },
+    {
+      "epoch": 2.245146645959023,
+      "grad_norm": 0.5125951863846561,
+      "learning_rate": 3.2709461767180007e-06,
+      "loss": 0.2806,
+      "step": 4699
+    },
+    {
+      "epoch": 2.2456245146645957,
+      "grad_norm": 0.5015064985997865,
+      "learning_rate": 3.2670511741683475e-06,
+      "loss": 0.2919,
+      "step": 4700
+    },
+    {
+      "epoch": 2.246102383370169,
+      "grad_norm": 0.441527020343577,
+      "learning_rate": 3.2631580392199316e-06,
+      "loss": 0.281,
+      "step": 4701
+    },
+    {
+      "epoch": 2.2465802520757423,
+      "grad_norm": 0.48081974361271224,
+      "learning_rate": 3.259266772952636e-06,
+      "loss": 0.2814,
+      "step": 4702
+    },
+    {
+      "epoch": 2.247058120781315,
+      "grad_norm": 0.5288631768506852,
+      "learning_rate": 3.2553773764458374e-06,
+      "loss": 0.2626,
+      "step": 4703
+    },
+    {
+      "epoch": 2.2475359894868885,
+      "grad_norm": 0.46254313690849835,
+      "learning_rate": 3.251489850778381e-06,
+      "loss": 0.2711,
+      "step": 4704
+    },
+    {
+      "epoch": 2.248013858192462,
+      "grad_norm": 0.4737518642755567,
+      "learning_rate": 3.2476041970285945e-06,
+      "loss": 0.2625,
+      "step": 4705
+    },
+    {
+      "epoch": 2.2484917268980347,
+      "grad_norm": 0.44502937263508435,
+      "learning_rate": 3.2437204162742975e-06,
+      "loss": 0.2571,
+      "step": 4706
+    },
+    {
+      "epoch": 2.248969595603608,
+      "grad_norm": 0.45119405560525677,
+      "learning_rate": 3.2398385095927775e-06,
+      "loss": 0.2877,
+      "step": 4707
+    },
+    {
+      "epoch": 2.249447464309181,
+      "grad_norm": 0.45027149302862923,
+      "learning_rate": 3.2359584780608055e-06,
+      "loss": 0.2709,
+      "step": 4708
+    },
+    {
+      "epoch": 2.249925333014754,
+      "grad_norm": 0.4419939110436344,
+      "learning_rate": 3.232080322754638e-06,
+      "loss": 0.2712,
+      "step": 4709
+    },
+    {
+      "epoch": 2.2504032017203275,
+      "grad_norm": 0.45427067027017565,
+      "learning_rate": 3.2282040447500063e-06,
+      "loss": 0.282,
+      "step": 4710
+    },
+    {
+      "epoch": 2.2508810704259004,
+      "grad_norm": 0.45850633742518837,
+      "learning_rate": 3.2243296451221164e-06,
+      "loss": 0.2502,
+      "step": 4711
+    },
+    {
+      "epoch": 2.2513589391314737,
+      "grad_norm": 0.4575443548721124,
+      "learning_rate": 3.220457124945665e-06,
+      "loss": 0.2625,
+      "step": 4712
+    },
+    {
+      "epoch": 2.251836807837047,
+      "grad_norm": 0.4657370801081544,
+      "learning_rate": 3.2165864852948147e-06,
+      "loss": 0.2373,
+      "step": 4713
+    },
+    {
+      "epoch": 2.25231467654262,
+      "grad_norm": 0.45836793153138744,
+      "learning_rate": 3.21271772724322e-06,
+      "loss": 0.2762,
+      "step": 4714
+    },
+    {
+      "epoch": 2.252792545248193,
+      "grad_norm": 0.4946186941630403,
+      "learning_rate": 3.208850851863998e-06,
+      "loss": 0.2888,
+      "step": 4715
+    },
+    {
+      "epoch": 2.253270413953766,
+      "grad_norm": 0.45248472166716014,
+      "learning_rate": 3.20498586022976e-06,
+      "loss": 0.2702,
+      "step": 4716
+    },
+    {
+      "epoch": 2.2537482826593394,
+      "grad_norm": 0.4594724499710993,
+      "learning_rate": 3.201122753412582e-06,
+      "loss": 0.2632,
+      "step": 4717
+    },
+    {
+      "epoch": 2.2542261513649127,
+      "grad_norm": 0.48222462052144954,
+      "learning_rate": 3.1972615324840197e-06,
+      "loss": 0.2553,
+      "step": 4718
+    },
+    {
+      "epoch": 2.2547040200704855,
+      "grad_norm": 0.4616647404690517,
+      "learning_rate": 3.193402198515112e-06,
+      "loss": 0.2704,
+      "step": 4719
+    },
+    {
+      "epoch": 2.255181888776059,
+      "grad_norm": 0.46141169216599737,
+      "learning_rate": 3.189544752576369e-06,
+      "loss": 0.2547,
+      "step": 4720
+    },
+    {
+      "epoch": 2.2556597574816317,
+      "grad_norm": 0.46118690700708775,
+      "learning_rate": 3.1856891957377735e-06,
+      "loss": 0.2678,
+      "step": 4721
+    },
+    {
+      "epoch": 2.256137626187205,
+      "grad_norm": 0.45651170157809823,
+      "learning_rate": 3.1818355290687962e-06,
+      "loss": 0.2782,
+      "step": 4722
+    },
+    {
+      "epoch": 2.2566154948927784,
+      "grad_norm": 0.49019149614604685,
+      "learning_rate": 3.177983753638373e-06,
+      "loss": 0.2498,
+      "step": 4723
+    },
+    {
+      "epoch": 2.257093363598351,
+      "grad_norm": 0.47354007893571354,
+      "learning_rate": 3.174133870514914e-06,
+      "loss": 0.2607,
+      "step": 4724
+    },
+    {
+      "epoch": 2.2575712323039245,
+      "grad_norm": 0.4752144096947728,
+      "learning_rate": 3.1702858807663175e-06,
+      "loss": 0.2769,
+      "step": 4725
+    },
+    {
+      "epoch": 2.2580491010094974,
+      "grad_norm": 0.48451667423176453,
+      "learning_rate": 3.166439785459943e-06,
+      "loss": 0.2708,
+      "step": 4726
+    },
+    {
+      "epoch": 2.2585269697150707,
+      "grad_norm": 0.4570766080552371,
+      "learning_rate": 3.1625955856626267e-06,
+      "loss": 0.2646,
+      "step": 4727
+    },
+    {
+      "epoch": 2.259004838420644,
+      "grad_norm": 0.4445274144500832,
+      "learning_rate": 3.1587532824406887e-06,
+      "loss": 0.2777,
+      "step": 4728
+    },
+    {
+      "epoch": 2.259482707126217,
+      "grad_norm": 0.4703985338457953,
+      "learning_rate": 3.1549128768599123e-06,
+      "loss": 0.2724,
+      "step": 4729
+    },
+    {
+      "epoch": 2.25996057583179,
+      "grad_norm": 0.46555737939351904,
+      "learning_rate": 3.151074369985556e-06,
+      "loss": 0.2716,
+      "step": 4730
+    },
+    {
+      "epoch": 2.2604384445373635,
+      "grad_norm": 0.45652113924751747,
+      "learning_rate": 3.147237762882359e-06,
+      "loss": 0.2695,
+      "step": 4731
+    },
+    {
+      "epoch": 2.2609163132429364,
+      "grad_norm": 0.4598715452430841,
+      "learning_rate": 3.143403056614527e-06,
+      "loss": 0.2631,
+      "step": 4732
+    },
+    {
+      "epoch": 2.2613941819485097,
+      "grad_norm": 0.5304735402801358,
+      "learning_rate": 3.139570252245734e-06,
+      "loss": 0.2731,
+      "step": 4733
+    },
+    {
+      "epoch": 2.261872050654083,
+      "grad_norm": 0.47206865977241064,
+      "learning_rate": 3.135739350839141e-06,
+      "loss": 0.2752,
+      "step": 4734
+    },
+    {
+      "epoch": 2.262349919359656,
+      "grad_norm": 0.4527194547413837,
+      "learning_rate": 3.131910353457369e-06,
+      "loss": 0.2663,
+      "step": 4735
+    },
+    {
+      "epoch": 2.262827788065229,
+      "grad_norm": 0.4742204418152278,
+      "learning_rate": 3.1280832611625112e-06,
+      "loss": 0.2727,
+      "step": 4736
+    },
+    {
+      "epoch": 2.263305656770802,
+      "grad_norm": 0.4698999007147011,
+      "learning_rate": 3.12425807501614e-06,
+      "loss": 0.2672,
+      "step": 4737
+    },
+    {
+      "epoch": 2.2637835254763754,
+      "grad_norm": 0.4507660759859787,
+      "learning_rate": 3.1204347960792935e-06,
+      "loss": 0.3011,
+      "step": 4738
+    },
+    {
+      "epoch": 2.2642613941819487,
+      "grad_norm": 0.47815020250201823,
+      "learning_rate": 3.116613425412478e-06,
+      "loss": 0.2707,
+      "step": 4739
+    },
+    {
+      "epoch": 2.2647392628875216,
+      "grad_norm": 0.466778767224418,
+      "learning_rate": 3.112793964075681e-06,
+      "loss": 0.2742,
+      "step": 4740
+    },
+    {
+      "epoch": 2.265217131593095,
+      "grad_norm": 0.46764545840736804,
+      "learning_rate": 3.1089764131283497e-06,
+      "loss": 0.2618,
+      "step": 4741
+    },
+    {
+      "epoch": 2.2656950002986678,
+      "grad_norm": 0.4598354811835029,
+      "learning_rate": 3.105160773629402e-06,
+      "loss": 0.2727,
+      "step": 4742
+    },
+    {
+      "epoch": 2.266172869004241,
+      "grad_norm": 0.45939402218354186,
+      "learning_rate": 3.1013470466372373e-06,
+      "loss": 0.2617,
+      "step": 4743
+    },
+    {
+      "epoch": 2.2666507377098144,
+      "grad_norm": 0.689169049474016,
+      "learning_rate": 3.0975352332097107e-06,
+      "loss": 0.2719,
+      "step": 4744
+    },
+    {
+      "epoch": 2.2671286064153873,
+      "grad_norm": 0.4457399095001578,
+      "learning_rate": 3.0937253344041507e-06,
+      "loss": 0.2641,
+      "step": 4745
+    },
+    {
+      "epoch": 2.2676064751209606,
+      "grad_norm": 0.5931896862224231,
+      "learning_rate": 3.0899173512773607e-06,
+      "loss": 0.271,
+      "step": 4746
+    },
+    {
+      "epoch": 2.2680843438265335,
+      "grad_norm": 0.44667772647454984,
+      "learning_rate": 3.0861112848856024e-06,
+      "loss": 0.2756,
+      "step": 4747
+    },
+    {
+      "epoch": 2.2685622125321068,
+      "grad_norm": 0.4870112961699444,
+      "learning_rate": 3.082307136284616e-06,
+      "loss": 0.2586,
+      "step": 4748
+    },
+    {
+      "epoch": 2.26904008123768,
+      "grad_norm": 1.0214438039442433,
+      "learning_rate": 3.0785049065296057e-06,
+      "loss": 0.2663,
+      "step": 4749
+    },
+    {
+      "epoch": 2.269517949943253,
+      "grad_norm": 0.5664811753038376,
+      "learning_rate": 3.074704596675242e-06,
+      "loss": 0.2616,
+      "step": 4750
+    },
+    {
+      "epoch": 2.2699958186488263,
+      "grad_norm": 0.4869676974104964,
+      "learning_rate": 3.07090620777566e-06,
+      "loss": 0.2865,
+      "step": 4751
+    },
+    {
+      "epoch": 2.270473687354399,
+      "grad_norm": 0.5088607622709265,
+      "learning_rate": 3.067109740884472e-06,
+      "loss": 0.268,
+      "step": 4752
+    },
+    {
+      "epoch": 2.2709515560599725,
+      "grad_norm": 0.4469599957069902,
+      "learning_rate": 3.063315197054747e-06,
+      "loss": 0.2721,
+      "step": 4753
+    },
+    {
+      "epoch": 2.2714294247655458,
+      "grad_norm": 0.4524405576091524,
+      "learning_rate": 3.0595225773390225e-06,
+      "loss": 0.2709,
+      "step": 4754
+    },
+    {
+      "epoch": 2.2719072934711186,
+      "grad_norm": 0.5863151040314684,
+      "learning_rate": 3.055731882789311e-06,
+      "loss": 0.2697,
+      "step": 4755
+    },
+    {
+      "epoch": 2.272385162176692,
+      "grad_norm": 0.4626038538364683,
+      "learning_rate": 3.05194311445708e-06,
+      "loss": 0.2539,
+      "step": 4756
+    },
+    {
+      "epoch": 2.2728630308822653,
+      "grad_norm": 0.45285520691551345,
+      "learning_rate": 3.0481562733932647e-06,
+      "loss": 0.2605,
+      "step": 4757
+    },
+    {
+      "epoch": 2.273340899587838,
+      "grad_norm": 0.48766103070408395,
+      "learning_rate": 3.0443713606482727e-06,
+      "loss": 0.2584,
+      "step": 4758
+    },
+    {
+      "epoch": 2.2738187682934115,
+      "grad_norm": 0.49630506462100404,
+      "learning_rate": 3.0405883772719715e-06,
+      "loss": 0.2783,
+      "step": 4759
+    },
+    {
+      "epoch": 2.2742966369989848,
+      "grad_norm": 0.4608594610607203,
+      "learning_rate": 3.0368073243136874e-06,
+      "loss": 0.2561,
+      "step": 4760
+    },
+    {
+      "epoch": 2.2747745057045576,
+      "grad_norm": 0.4530972395926978,
+      "learning_rate": 3.033028202822228e-06,
+      "loss": 0.2554,
+      "step": 4761
+    },
+    {
+      "epoch": 2.275252374410131,
+      "grad_norm": 0.45511832040091293,
+      "learning_rate": 3.029251013845849e-06,
+      "loss": 0.2577,
+      "step": 4762
+    },
+    {
+      "epoch": 2.275730243115704,
+      "grad_norm": 0.9922747694397978,
+      "learning_rate": 3.0254757584322736e-06,
+      "loss": 0.2623,
+      "step": 4763
+    },
+    {
+      "epoch": 2.276208111821277,
+      "grad_norm": 0.48096765381573897,
+      "learning_rate": 3.0217024376286984e-06,
+      "loss": 0.2493,
+      "step": 4764
+    },
+    {
+      "epoch": 2.2766859805268505,
+      "grad_norm": 0.47667943487788295,
+      "learning_rate": 3.0179310524817707e-06,
+      "loss": 0.2859,
+      "step": 4765
+    },
+    {
+      "epoch": 2.2771638492324233,
+      "grad_norm": 0.43920123160920244,
+      "learning_rate": 3.0141616040376052e-06,
+      "loss": 0.2761,
+      "step": 4766
+    },
+    {
+      "epoch": 2.2776417179379966,
+      "grad_norm": 0.485107810636819,
+      "learning_rate": 3.010394093341785e-06,
+      "loss": 0.2865,
+      "step": 4767
+    },
+    {
+      "epoch": 2.2781195866435695,
+      "grad_norm": 0.4621289185973684,
+      "learning_rate": 3.00662852143935e-06,
+      "loss": 0.2788,
+      "step": 4768
+    },
+    {
+      "epoch": 2.278597455349143,
+      "grad_norm": 0.4850819067216184,
+      "learning_rate": 3.002864889374798e-06,
+      "loss": 0.2655,
+      "step": 4769
+    },
+    {
+      "epoch": 2.279075324054716,
+      "grad_norm": 0.47125925955991355,
+      "learning_rate": 2.9991031981921026e-06,
+      "loss": 0.2511,
+      "step": 4770
+    },
+    {
+      "epoch": 2.279553192760289,
+      "grad_norm": 1.1105889455924962,
+      "learning_rate": 2.9953434489346856e-06,
+      "loss": 0.2551,
+      "step": 4771
+    },
+    {
+      "epoch": 2.2800310614658623,
+      "grad_norm": 0.4773544991225756,
+      "learning_rate": 2.9915856426454324e-06,
+      "loss": 0.2583,
+      "step": 4772
+    },
+    {
+      "epoch": 2.280508930171435,
+      "grad_norm": 0.49707858432136753,
+      "learning_rate": 2.987829780366699e-06,
+      "loss": 0.2675,
+      "step": 4773
+    },
+    {
+      "epoch": 2.2809867988770085,
+      "grad_norm": 0.47711687874889647,
+      "learning_rate": 2.984075863140292e-06,
+      "loss": 0.2658,
+      "step": 4774
+    },
+    {
+      "epoch": 2.281464667582582,
+      "grad_norm": 0.585688509831411,
+      "learning_rate": 2.9803238920074784e-06,
+      "loss": 0.2547,
+      "step": 4775
+    },
+    {
+      "epoch": 2.2819425362881547,
+      "grad_norm": 0.46857304420027696,
+      "learning_rate": 2.976573868008995e-06,
+      "loss": 0.2658,
+      "step": 4776
+    },
+    {
+      "epoch": 2.282420404993728,
+      "grad_norm": 0.4558766138524307,
+      "learning_rate": 2.9728257921850302e-06,
+      "loss": 0.2586,
+      "step": 4777
+    },
+    {
+      "epoch": 2.282898273699301,
+      "grad_norm": 0.45756901243901504,
+      "learning_rate": 2.9690796655752306e-06,
+      "loss": 0.2689,
+      "step": 4778
+    },
+    {
+      "epoch": 2.283376142404874,
+      "grad_norm": 0.47832703138164223,
+      "learning_rate": 2.965335489218711e-06,
+      "loss": 0.2786,
+      "step": 4779
+    },
+    {
+      "epoch": 2.2838540111104475,
+      "grad_norm": 0.45406741476852475,
+      "learning_rate": 2.961593264154038e-06,
+      "loss": 0.2628,
+      "step": 4780
+    },
+    {
+      "epoch": 2.2843318798160204,
+      "grad_norm": 0.47436505180344063,
+      "learning_rate": 2.9578529914192342e-06,
+      "loss": 0.2603,
+      "step": 4781
+    },
+    {
+      "epoch": 2.2848097485215937,
+      "grad_norm": 0.45974865789697694,
+      "learning_rate": 2.954114672051789e-06,
+      "loss": 0.2602,
+      "step": 4782
+    },
+    {
+      "epoch": 2.285287617227167,
+      "grad_norm": 0.5087537120467954,
+      "learning_rate": 2.9503783070886504e-06,
+      "loss": 0.2638,
+      "step": 4783
+    },
+    {
+      "epoch": 2.28576548593274,
+      "grad_norm": 0.4593223370985407,
+      "learning_rate": 2.946643897566216e-06,
+      "loss": 0.26,
+      "step": 4784
+    },
+    {
+      "epoch": 2.286243354638313,
+      "grad_norm": 0.4739581626972536,
+      "learning_rate": 2.9429114445203423e-06,
+      "loss": 0.2834,
+      "step": 4785
+    },
+    {
+      "epoch": 2.2867212233438865,
+      "grad_norm": 0.4473066092125534,
+      "learning_rate": 2.939180948986352e-06,
+      "loss": 0.258,
+      "step": 4786
+    },
+    {
+      "epoch": 2.2871990920494594,
+      "grad_norm": 0.7178714404426283,
+      "learning_rate": 2.9354524119990156e-06,
+      "loss": 0.264,
+      "step": 4787
+    },
+    {
+      "epoch": 2.2876769607550327,
+      "grad_norm": 0.4518483998022226,
+      "learning_rate": 2.9317258345925603e-06,
+      "loss": 0.2552,
+      "step": 4788
+    },
+    {
+      "epoch": 2.2881548294606056,
+      "grad_norm": 0.4592657516604581,
+      "learning_rate": 2.92800121780068e-06,
+      "loss": 0.2758,
+      "step": 4789
+    },
+    {
+      "epoch": 2.288632698166179,
+      "grad_norm": 0.703821050907462,
+      "learning_rate": 2.924278562656514e-06,
+      "loss": 0.2712,
+      "step": 4790
+    },
+    {
+      "epoch": 2.289110566871752,
+      "grad_norm": 0.5327889383144262,
+      "learning_rate": 2.9205578701926575e-06,
+      "loss": 0.2548,
+      "step": 4791
+    },
+    {
+      "epoch": 2.289588435577325,
+      "grad_norm": 0.4542697805884633,
+      "learning_rate": 2.916839141441172e-06,
+      "loss": 0.2648,
+      "step": 4792
+    },
+    {
+      "epoch": 2.2900663042828984,
+      "grad_norm": 0.4471368664844579,
+      "learning_rate": 2.913122377433564e-06,
+      "loss": 0.2701,
+      "step": 4793
+    },
+    {
+      "epoch": 2.2905441729884712,
+      "grad_norm": 0.49515537031904927,
+      "learning_rate": 2.9094075792007948e-06,
+      "loss": 0.2612,
+      "step": 4794
+    },
+    {
+      "epoch": 2.2910220416940446,
+      "grad_norm": 0.5023092878736718,
+      "learning_rate": 2.90569474777329e-06,
+      "loss": 0.2731,
+      "step": 4795
+    },
+    {
+      "epoch": 2.291499910399618,
+      "grad_norm": 0.5203729973321249,
+      "learning_rate": 2.901983884180921e-06,
+      "loss": 0.2693,
+      "step": 4796
+    },
+    {
+      "epoch": 2.2919777791051907,
+      "grad_norm": 0.458762113548676,
+      "learning_rate": 2.8982749894530128e-06,
+      "loss": 0.2528,
+      "step": 4797
+    },
+    {
+      "epoch": 2.292455647810764,
+      "grad_norm": 0.4454688102007361,
+      "learning_rate": 2.8945680646183527e-06,
+      "loss": 0.2642,
+      "step": 4798
+    },
+    {
+      "epoch": 2.292933516516337,
+      "grad_norm": 0.47272093877831456,
+      "learning_rate": 2.8908631107051743e-06,
+      "loss": 0.2845,
+      "step": 4799
+    },
+    {
+      "epoch": 2.2934113852219102,
+      "grad_norm": 0.44051662539793884,
+      "learning_rate": 2.8871601287411634e-06,
+      "loss": 0.2814,
+      "step": 4800
+    },
+    {
+      "epoch": 2.2938892539274836,
+      "grad_norm": 0.44919987478542955,
+      "learning_rate": 2.8834591197534668e-06,
+      "loss": 0.2616,
+      "step": 4801
+    },
+    {
+      "epoch": 2.2943671226330564,
+      "grad_norm": 0.4655075873793121,
+      "learning_rate": 2.879760084768677e-06,
+      "loss": 0.2598,
+      "step": 4802
+    },
+    {
+      "epoch": 2.2948449913386297,
+      "grad_norm": 0.44695659849319586,
+      "learning_rate": 2.8760630248128374e-06,
+      "loss": 0.2809,
+      "step": 4803
+    },
+    {
+      "epoch": 2.2953228600442026,
+      "grad_norm": 0.46112658531258754,
+      "learning_rate": 2.8723679409114536e-06,
+      "loss": 0.2625,
+      "step": 4804
+    },
+    {
+      "epoch": 2.295800728749776,
+      "grad_norm": 0.4656636436014579,
+      "learning_rate": 2.8686748340894744e-06,
+      "loss": 0.2731,
+      "step": 4805
+    },
+    {
+      "epoch": 2.2962785974553492,
+      "grad_norm": 0.5844196537588953,
+      "learning_rate": 2.864983705371298e-06,
+      "loss": 0.2663,
+      "step": 4806
+    },
+    {
+      "epoch": 2.296756466160922,
+      "grad_norm": 0.46998718606495676,
+      "learning_rate": 2.861294555780786e-06,
+      "loss": 0.2574,
+      "step": 4807
+    },
+    {
+      "epoch": 2.2972343348664954,
+      "grad_norm": 0.4823897817406597,
+      "learning_rate": 2.8576073863412402e-06,
+      "loss": 0.2823,
+      "step": 4808
+    },
+    {
+      "epoch": 2.2977122035720687,
+      "grad_norm": 0.4531753739973854,
+      "learning_rate": 2.8539221980754115e-06,
+      "loss": 0.2541,
+      "step": 4809
+    },
+    {
+      "epoch": 2.2981900722776416,
+      "grad_norm": 0.4579423043984502,
+      "learning_rate": 2.850238992005514e-06,
+      "loss": 0.2737,
+      "step": 4810
+    },
+    {
+      "epoch": 2.298667940983215,
+      "grad_norm": 0.44211975158230343,
+      "learning_rate": 2.8465577691532e-06,
+      "loss": 0.2716,
+      "step": 4811
+    },
+    {
+      "epoch": 2.2991458096887882,
+      "grad_norm": 0.4480696060954269,
+      "learning_rate": 2.8428785305395733e-06,
+      "loss": 0.261,
+      "step": 4812
+    },
+    {
+      "epoch": 2.299623678394361,
+      "grad_norm": 0.4605681190004609,
+      "learning_rate": 2.8392012771851963e-06,
+      "loss": 0.2693,
+      "step": 4813
+    },
+    {
+      "epoch": 2.3001015470999344,
+      "grad_norm": 0.6621094950095469,
+      "learning_rate": 2.83552601011007e-06,
+      "loss": 0.2694,
+      "step": 4814
+    },
+    {
+      "epoch": 2.3005794158055073,
+      "grad_norm": 0.4783897161449604,
+      "learning_rate": 2.8318527303336465e-06,
+      "loss": 0.2756,
+      "step": 4815
+    },
+    {
+      "epoch": 2.3010572845110806,
+      "grad_norm": 0.4512383810963222,
+      "learning_rate": 2.828181438874832e-06,
+      "loss": 0.2712,
+      "step": 4816
+    },
+    {
+      "epoch": 2.301535153216654,
+      "grad_norm": 0.4712970827373998,
+      "learning_rate": 2.8245121367519812e-06,
+      "loss": 0.2603,
+      "step": 4817
+    },
+    {
+      "epoch": 2.302013021922227,
+      "grad_norm": 0.47038727743638786,
+      "learning_rate": 2.820844824982889e-06,
+      "loss": 0.2601,
+      "step": 4818
+    },
+    {
+      "epoch": 2.3024908906278,
+      "grad_norm": 0.4593763213754035,
+      "learning_rate": 2.817179504584802e-06,
+      "loss": 0.2699,
+      "step": 4819
+    },
+    {
+      "epoch": 2.302968759333373,
+      "grad_norm": 0.4753946259917545,
+      "learning_rate": 2.81351617657442e-06,
+      "loss": 0.2676,
+      "step": 4820
+    },
+    {
+      "epoch": 2.3034466280389463,
+      "grad_norm": 0.46869051732758144,
+      "learning_rate": 2.8098548419678838e-06,
+      "loss": 0.2838,
+      "step": 4821
+    },
+    {
+      "epoch": 2.3039244967445196,
+      "grad_norm": 0.45543892367550015,
+      "learning_rate": 2.8061955017807797e-06,
+      "loss": 0.2707,
+      "step": 4822
+    },
+    {
+      "epoch": 2.3044023654500925,
+      "grad_norm": 0.48105929755291854,
+      "learning_rate": 2.8025381570281495e-06,
+      "loss": 0.2735,
+      "step": 4823
+    },
+    {
+      "epoch": 2.304880234155666,
+      "grad_norm": 0.4562110398780488,
+      "learning_rate": 2.7988828087244735e-06,
+      "loss": 0.2584,
+      "step": 4824
+    },
+    {
+      "epoch": 2.3053581028612387,
+      "grad_norm": 0.4708490596777413,
+      "learning_rate": 2.795229457883678e-06,
+      "loss": 0.2521,
+      "step": 4825
+    },
+    {
+      "epoch": 2.305835971566812,
+      "grad_norm": 0.47823032227536855,
+      "learning_rate": 2.7915781055191437e-06,
+      "loss": 0.2446,
+      "step": 4826
+    },
+    {
+      "epoch": 2.3063138402723853,
+      "grad_norm": 0.4583497787749354,
+      "learning_rate": 2.7879287526436884e-06,
+      "loss": 0.2682,
+      "step": 4827
+    },
+    {
+      "epoch": 2.306791708977958,
+      "grad_norm": 0.5195107891663974,
+      "learning_rate": 2.784281400269575e-06,
+      "loss": 0.2639,
+      "step": 4828
+    },
+    {
+      "epoch": 2.3072695776835315,
+      "grad_norm": 0.44700898622956803,
+      "learning_rate": 2.7806360494085218e-06,
+      "loss": 0.2704,
+      "step": 4829
+    },
+    {
+      "epoch": 2.307747446389105,
+      "grad_norm": 0.45593935022847376,
+      "learning_rate": 2.7769927010716814e-06,
+      "loss": 0.2561,
+      "step": 4830
+    },
+    {
+      "epoch": 2.3082253150946777,
+      "grad_norm": 0.4779945808794087,
+      "learning_rate": 2.77335135626965e-06,
+      "loss": 0.2612,
+      "step": 4831
+    },
+    {
+      "epoch": 2.308703183800251,
+      "grad_norm": 0.4650534484982984,
+      "learning_rate": 2.76971201601248e-06,
+      "loss": 0.2704,
+      "step": 4832
+    },
+    {
+      "epoch": 2.309181052505824,
+      "grad_norm": 0.45968865169832307,
+      "learning_rate": 2.7660746813096575e-06,
+      "loss": 0.2788,
+      "step": 4833
+    },
+    {
+      "epoch": 2.309658921211397,
+      "grad_norm": 0.5586814813353537,
+      "learning_rate": 2.76243935317011e-06,
+      "loss": 0.2586,
+      "step": 4834
+    },
+    {
+      "epoch": 2.3101367899169705,
+      "grad_norm": 0.46957361443859275,
+      "learning_rate": 2.7588060326022205e-06,
+      "loss": 0.2636,
+      "step": 4835
+    },
+    {
+      "epoch": 2.3106146586225433,
+      "grad_norm": 0.5799118470767503,
+      "learning_rate": 2.755174720613806e-06,
+      "loss": 0.2903,
+      "step": 4836
+    },
+    {
+      "epoch": 2.3110925273281167,
+      "grad_norm": 0.48771811958787314,
+      "learning_rate": 2.7515454182121238e-06,
+      "loss": 0.2648,
+      "step": 4837
+    },
+    {
+      "epoch": 2.31157039603369,
+      "grad_norm": 0.4454160846341557,
+      "learning_rate": 2.7479181264038847e-06,
+      "loss": 0.2673,
+      "step": 4838
+    },
+    {
+      "epoch": 2.312048264739263,
+      "grad_norm": 0.48485972776665004,
+      "learning_rate": 2.7442928461952333e-06,
+      "loss": 0.2792,
+      "step": 4839
+    },
+    {
+      "epoch": 2.312526133444836,
+      "grad_norm": 0.44389446972461244,
+      "learning_rate": 2.740669578591755e-06,
+      "loss": 0.2575,
+      "step": 4840
+    },
+    {
+      "epoch": 2.313004002150409,
+      "grad_norm": 0.5119507485701839,
+      "learning_rate": 2.7370483245984857e-06,
+      "loss": 0.2579,
+      "step": 4841
+    },
+    {
+      "epoch": 2.3134818708559823,
+      "grad_norm": 0.479630718784202,
+      "learning_rate": 2.733429085219895e-06,
+      "loss": 0.2533,
+      "step": 4842
+    },
+    {
+      "epoch": 2.3139597395615557,
+      "grad_norm": 0.45539893096166484,
+      "learning_rate": 2.7298118614598934e-06,
+      "loss": 0.2874,
+      "step": 4843
+    },
+    {
+      "epoch": 2.3144376082671285,
+      "grad_norm": 0.4597008048466559,
+      "learning_rate": 2.726196654321841e-06,
+      "loss": 0.2775,
+      "step": 4844
+    },
+    {
+      "epoch": 2.314915476972702,
+      "grad_norm": 0.4884323297495012,
+      "learning_rate": 2.7225834648085282e-06,
+      "loss": 0.2512,
+      "step": 4845
+    },
+    {
+      "epoch": 2.3153933456782747,
+      "grad_norm": 0.4765028408676871,
+      "learning_rate": 2.7189722939221875e-06,
+      "loss": 0.2589,
+      "step": 4846
+    },
+    {
+      "epoch": 2.315871214383848,
+      "grad_norm": 0.4660406532791925,
+      "learning_rate": 2.715363142664501e-06,
+      "loss": 0.2648,
+      "step": 4847
+    },
+    {
+      "epoch": 2.3163490830894213,
+      "grad_norm": 0.4747962483625691,
+      "learning_rate": 2.711756012036577e-06,
+      "loss": 0.2778,
+      "step": 4848
+    },
+    {
+      "epoch": 2.316826951794994,
+      "grad_norm": 0.4632488608800758,
+      "learning_rate": 2.708150903038972e-06,
+      "loss": 0.275,
+      "step": 4849
+    },
+    {
+      "epoch": 2.3173048205005675,
+      "grad_norm": 0.5453668545195846,
+      "learning_rate": 2.7045478166716843e-06,
+      "loss": 0.2755,
+      "step": 4850
+    },
+    {
+      "epoch": 2.3177826892061404,
+      "grad_norm": 0.5001628937077411,
+      "learning_rate": 2.7009467539341426e-06,
+      "loss": 0.2691,
+      "step": 4851
+    },
+    {
+      "epoch": 2.3182605579117137,
+      "grad_norm": 0.4553954046591676,
+      "learning_rate": 2.6973477158252146e-06,
+      "loss": 0.2655,
+      "step": 4852
+    },
+    {
+      "epoch": 2.318738426617287,
+      "grad_norm": 0.4767807917673003,
+      "learning_rate": 2.6937507033432177e-06,
+      "loss": 0.2806,
+      "step": 4853
+    },
+    {
+      "epoch": 2.31921629532286,
+      "grad_norm": 0.45179451190892383,
+      "learning_rate": 2.690155717485895e-06,
+      "loss": 0.2696,
+      "step": 4854
+    },
+    {
+      "epoch": 2.319694164028433,
+      "grad_norm": 0.45905304275528375,
+      "learning_rate": 2.6865627592504295e-06,
+      "loss": 0.2672,
+      "step": 4855
+    },
+    {
+      "epoch": 2.3201720327340065,
+      "grad_norm": 0.4375085321749962,
+      "learning_rate": 2.6829718296334516e-06,
+      "loss": 0.2562,
+      "step": 4856
+    },
+    {
+      "epoch": 2.3206499014395794,
+      "grad_norm": 0.45149139878521694,
+      "learning_rate": 2.6793829296310183e-06,
+      "loss": 0.2903,
+      "step": 4857
+    },
+    {
+      "epoch": 2.3211277701451527,
+      "grad_norm": 0.4575764223416466,
+      "learning_rate": 2.6757960602386223e-06,
+      "loss": 0.2534,
+      "step": 4858
+    },
+    {
+      "epoch": 2.3216056388507256,
+      "grad_norm": 0.4537222729437179,
+      "learning_rate": 2.6722112224512063e-06,
+      "loss": 0.2518,
+      "step": 4859
+    },
+    {
+      "epoch": 2.322083507556299,
+      "grad_norm": 0.45191436383379496,
+      "learning_rate": 2.668628417263137e-06,
+      "loss": 0.2737,
+      "step": 4860
+    },
+    {
+      "epoch": 2.322561376261872,
+      "grad_norm": 0.4670433808850828,
+      "learning_rate": 2.6650476456682195e-06,
+      "loss": 0.2645,
+      "step": 4861
+    },
+    {
+      "epoch": 2.323039244967445,
+      "grad_norm": 0.46184013428909193,
+      "learning_rate": 2.661468908659701e-06,
+      "loss": 0.2703,
+      "step": 4862
+    },
+    {
+      "epoch": 2.3235171136730184,
+      "grad_norm": 0.46573908878240283,
+      "learning_rate": 2.6578922072302572e-06,
+      "loss": 0.2681,
+      "step": 4863
+    },
+    {
+      "epoch": 2.3239949823785917,
+      "grad_norm": 0.4495053674176749,
+      "learning_rate": 2.6543175423720004e-06,
+      "loss": 0.2613,
+      "step": 4864
+    },
+    {
+      "epoch": 2.3244728510841646,
+      "grad_norm": 0.48067303236066894,
+      "learning_rate": 2.6507449150764852e-06,
+      "loss": 0.2743,
+      "step": 4865
+    },
+    {
+      "epoch": 2.324950719789738,
+      "grad_norm": 0.450687819523822,
+      "learning_rate": 2.6471743263346903e-06,
+      "loss": 0.2762,
+      "step": 4866
+    },
+    {
+      "epoch": 2.3254285884953108,
+      "grad_norm": 0.5034337207294447,
+      "learning_rate": 2.643605777137034e-06,
+      "loss": 0.2749,
+      "step": 4867
+    },
+    {
+      "epoch": 2.325906457200884,
+      "grad_norm": 0.5112858649436341,
+      "learning_rate": 2.6400392684733735e-06,
+      "loss": 0.2726,
+      "step": 4868
+    },
+    {
+      "epoch": 2.3263843259064574,
+      "grad_norm": 0.447531179096655,
+      "learning_rate": 2.636474801332992e-06,
+      "loss": 0.2798,
+      "step": 4869
+    },
+    {
+      "epoch": 2.3268621946120303,
+      "grad_norm": 0.47053088394428416,
+      "learning_rate": 2.632912376704607e-06,
+      "loss": 0.2772,
+      "step": 4870
+    },
+    {
+      "epoch": 2.3273400633176036,
+      "grad_norm": 0.4892476562482173,
+      "learning_rate": 2.629351995576379e-06,
+      "loss": 0.2728,
+      "step": 4871
+    },
+    {
+      "epoch": 2.3278179320231764,
+      "grad_norm": 0.4496411171519612,
+      "learning_rate": 2.6257936589358914e-06,
+      "loss": 0.2715,
+      "step": 4872
+    },
+    {
+      "epoch": 2.3282958007287498,
+      "grad_norm": 0.4706505012171985,
+      "learning_rate": 2.6222373677701607e-06,
+      "loss": 0.2611,
+      "step": 4873
+    },
+    {
+      "epoch": 2.328773669434323,
+      "grad_norm": 0.46352164932756873,
+      "learning_rate": 2.618683123065646e-06,
+      "loss": 0.2802,
+      "step": 4874
+    },
+    {
+      "epoch": 2.329251538139896,
+      "grad_norm": 0.4668367012503617,
+      "learning_rate": 2.615130925808228e-06,
+      "loss": 0.255,
+      "step": 4875
+    },
+    {
+      "epoch": 2.3297294068454693,
+      "grad_norm": 0.4526742078440504,
+      "learning_rate": 2.6115807769832226e-06,
+      "loss": 0.2543,
+      "step": 4876
+    },
+    {
+      "epoch": 2.330207275551042,
+      "grad_norm": 0.48303918462763046,
+      "learning_rate": 2.6080326775753816e-06,
+      "loss": 0.2816,
+      "step": 4877
+    },
+    {
+      "epoch": 2.3306851442566154,
+      "grad_norm": 0.5333148870731564,
+      "learning_rate": 2.604486628568885e-06,
+      "loss": 0.274,
+      "step": 4878
+    },
+    {
+      "epoch": 2.3311630129621888,
+      "grad_norm": 0.4581104638202767,
+      "learning_rate": 2.6009426309473397e-06,
+      "loss": 0.2565,
+      "step": 4879
+    },
+    {
+      "epoch": 2.3316408816677616,
+      "grad_norm": 0.4665861539340606,
+      "learning_rate": 2.597400685693795e-06,
+      "loss": 0.2801,
+      "step": 4880
+    },
+    {
+      "epoch": 2.332118750373335,
+      "grad_norm": 0.5767159313567876,
+      "learning_rate": 2.59386079379072e-06,
+      "loss": 0.2744,
+      "step": 4881
+    },
+    {
+      "epoch": 2.3325966190789083,
+      "grad_norm": 0.6636916277687075,
+      "learning_rate": 2.590322956220015e-06,
+      "loss": 0.2714,
+      "step": 4882
+    },
+    {
+      "epoch": 2.333074487784481,
+      "grad_norm": 0.4707596869750743,
+      "learning_rate": 2.586787173963019e-06,
+      "loss": 0.274,
+      "step": 4883
+    },
+    {
+      "epoch": 2.3335523564900544,
+      "grad_norm": 0.5135378298995066,
+      "learning_rate": 2.5832534480004955e-06,
+      "loss": 0.2703,
+      "step": 4884
+    },
+    {
+      "epoch": 2.3340302251956278,
+      "grad_norm": 0.4520964774142493,
+      "learning_rate": 2.5797217793126373e-06,
+      "loss": 0.2639,
+      "step": 4885
+    },
+    {
+      "epoch": 2.3345080939012006,
+      "grad_norm": 0.5034516755934882,
+      "learning_rate": 2.5761921688790635e-06,
+      "loss": 0.2665,
+      "step": 4886
+    },
+    {
+      "epoch": 2.334985962606774,
+      "grad_norm": 0.44517562515923137,
+      "learning_rate": 2.5726646176788307e-06,
+      "loss": 0.2736,
+      "step": 4887
+    },
+    {
+      "epoch": 2.335463831312347,
+      "grad_norm": 0.5542169246110865,
+      "learning_rate": 2.5691391266904165e-06,
+      "loss": 0.2665,
+      "step": 4888
+    },
+    {
+      "epoch": 2.33594170001792,
+      "grad_norm": 0.4735315168725527,
+      "learning_rate": 2.5656156968917277e-06,
+      "loss": 0.257,
+      "step": 4889
+    },
+    {
+      "epoch": 2.3364195687234934,
+      "grad_norm": 0.48423296811822936,
+      "learning_rate": 2.5620943292601074e-06,
+      "loss": 0.2784,
+      "step": 4890
+    },
+    {
+      "epoch": 2.3368974374290663,
+      "grad_norm": 0.4515981436184163,
+      "learning_rate": 2.5585750247723183e-06,
+      "loss": 0.2705,
+      "step": 4891
+    },
+    {
+      "epoch": 2.3373753061346396,
+      "grad_norm": 0.4733626892725105,
+      "learning_rate": 2.5550577844045498e-06,
+      "loss": 0.2558,
+      "step": 4892
+    },
+    {
+      "epoch": 2.3378531748402125,
+      "grad_norm": 0.5043562480358438,
+      "learning_rate": 2.551542609132428e-06,
+      "loss": 0.2436,
+      "step": 4893
+    },
+    {
+      "epoch": 2.338331043545786,
+      "grad_norm": 0.5000834206474044,
+      "learning_rate": 2.548029499930997e-06,
+      "loss": 0.2566,
+      "step": 4894
+    },
+    {
+      "epoch": 2.338808912251359,
+      "grad_norm": 0.4478341988025251,
+      "learning_rate": 2.5445184577747305e-06,
+      "loss": 0.2586,
+      "step": 4895
+    },
+    {
+      "epoch": 2.339286780956932,
+      "grad_norm": 0.5257945834666825,
+      "learning_rate": 2.5410094836375343e-06,
+      "loss": 0.2644,
+      "step": 4896
+    },
+    {
+      "epoch": 2.3397646496625053,
+      "grad_norm": 0.4531122219133908,
+      "learning_rate": 2.537502578492733e-06,
+      "loss": 0.2785,
+      "step": 4897
+    },
+    {
+      "epoch": 2.340242518368078,
+      "grad_norm": 0.4863425466939351,
+      "learning_rate": 2.533997743313077e-06,
+      "loss": 0.2512,
+      "step": 4898
+    },
+    {
+      "epoch": 2.3407203870736515,
+      "grad_norm": 0.47019660570137406,
+      "learning_rate": 2.5304949790707512e-06,
+      "loss": 0.257,
+      "step": 4899
+    },
+    {
+      "epoch": 2.341198255779225,
+      "grad_norm": 0.46196723454827476,
+      "learning_rate": 2.52699428673736e-06,
+      "loss": 0.2603,
+      "step": 4900
+    },
+    {
+      "epoch": 2.3416761244847977,
+      "grad_norm": 0.5103312090564901,
+      "learning_rate": 2.5234956672839273e-06,
+      "loss": 0.2668,
+      "step": 4901
+    },
+    {
+      "epoch": 2.342153993190371,
+      "grad_norm": 0.442174404107808,
+      "learning_rate": 2.519999121680917e-06,
+      "loss": 0.2504,
+      "step": 4902
+    },
+    {
+      "epoch": 2.342631861895944,
+      "grad_norm": 0.46898344435430483,
+      "learning_rate": 2.516504650898206e-06,
+      "loss": 0.269,
+      "step": 4903
+    },
+    {
+      "epoch": 2.343109730601517,
+      "grad_norm": 0.5615098737605004,
+      "learning_rate": 2.513012255905095e-06,
+      "loss": 0.2708,
+      "step": 4904
+    },
+    {
+      "epoch": 2.3435875993070905,
+      "grad_norm": 0.4488077245674206,
+      "learning_rate": 2.5095219376703183e-06,
+      "loss": 0.2668,
+      "step": 4905
+    },
+    {
+      "epoch": 2.3440654680126634,
+      "grad_norm": 0.4667442735851483,
+      "learning_rate": 2.5060336971620268e-06,
+      "loss": 0.2774,
+      "step": 4906
+    },
+    {
+      "epoch": 2.3445433367182367,
+      "grad_norm": 0.45038553396302394,
+      "learning_rate": 2.5025475353477933e-06,
+      "loss": 0.2595,
+      "step": 4907
+    },
+    {
+      "epoch": 2.34502120542381,
+      "grad_norm": 0.45570886034637664,
+      "learning_rate": 2.4990634531946247e-06,
+      "loss": 0.2662,
+      "step": 4908
+    },
+    {
+      "epoch": 2.345499074129383,
+      "grad_norm": 0.4532263527919922,
+      "learning_rate": 2.495581451668938e-06,
+      "loss": 0.2712,
+      "step": 4909
+    },
+    {
+      "epoch": 2.345976942834956,
+      "grad_norm": 0.5644370739027057,
+      "learning_rate": 2.4921015317365794e-06,
+      "loss": 0.2745,
+      "step": 4910
+    },
+    {
+      "epoch": 2.3464548115405295,
+      "grad_norm": 0.45617320912995984,
+      "learning_rate": 2.488623694362822e-06,
+      "loss": 0.254,
+      "step": 4911
+    },
+    {
+      "epoch": 2.3469326802461024,
+      "grad_norm": 0.743987369148091,
+      "learning_rate": 2.4851479405123524e-06,
+      "loss": 0.2797,
+      "step": 4912
+    },
+    {
+      "epoch": 2.3474105489516757,
+      "grad_norm": 0.4687580573243133,
+      "learning_rate": 2.4816742711492813e-06,
+      "loss": 0.2646,
+      "step": 4913
+    },
+    {
+      "epoch": 2.3478884176572485,
+      "grad_norm": 0.4689346718000839,
+      "learning_rate": 2.47820268723715e-06,
+      "loss": 0.2733,
+      "step": 4914
+    },
+    {
+      "epoch": 2.348366286362822,
+      "grad_norm": 0.4490192177022896,
+      "learning_rate": 2.4747331897389103e-06,
+      "loss": 0.2665,
+      "step": 4915
+    },
+    {
+      "epoch": 2.348844155068395,
+      "grad_norm": 0.49311803151688366,
+      "learning_rate": 2.471265779616938e-06,
+      "loss": 0.2461,
+      "step": 4916
+    },
+    {
+      "epoch": 2.349322023773968,
+      "grad_norm": 0.467178488370915,
+      "learning_rate": 2.467800457833034e-06,
+      "loss": 0.2606,
+      "step": 4917
+    },
+    {
+      "epoch": 2.3497998924795414,
+      "grad_norm": 0.4596377441591877,
+      "learning_rate": 2.46433722534842e-06,
+      "loss": 0.2535,
+      "step": 4918
+    },
+    {
+      "epoch": 2.3502777611851142,
+      "grad_norm": 0.4694260851272706,
+      "learning_rate": 2.460876083123733e-06,
+      "loss": 0.289,
+      "step": 4919
+    },
+    {
+      "epoch": 2.3507556298906875,
+      "grad_norm": 0.45269147527252934,
+      "learning_rate": 2.4574170321190305e-06,
+      "loss": 0.2706,
+      "step": 4920
+    },
+    {
+      "epoch": 2.351233498596261,
+      "grad_norm": 0.4770873996963581,
+      "learning_rate": 2.4539600732937964e-06,
+      "loss": 0.2492,
+      "step": 4921
+    },
+    {
+      "epoch": 2.3517113673018337,
+      "grad_norm": 0.4601878264182159,
+      "learning_rate": 2.450505207606928e-06,
+      "loss": 0.2758,
+      "step": 4922
+    },
+    {
+      "epoch": 2.352189236007407,
+      "grad_norm": 0.4716682705387417,
+      "learning_rate": 2.4470524360167413e-06,
+      "loss": 0.2731,
+      "step": 4923
+    },
+    {
+      "epoch": 2.35266710471298,
+      "grad_norm": 0.4939845519900784,
+      "learning_rate": 2.4436017594809804e-06,
+      "loss": 0.2924,
+      "step": 4924
+    },
+    {
+      "epoch": 2.353144973418553,
+      "grad_norm": 0.4919791373245082,
+      "learning_rate": 2.440153178956798e-06,
+      "loss": 0.2826,
+      "step": 4925
+    },
+    {
+      "epoch": 2.3536228421241265,
+      "grad_norm": 0.48187056236329523,
+      "learning_rate": 2.436706695400769e-06,
+      "loss": 0.2796,
+      "step": 4926
+    },
+    {
+      "epoch": 2.3541007108296994,
+      "grad_norm": 0.4641311091931838,
+      "learning_rate": 2.43326230976889e-06,
+      "loss": 0.2629,
+      "step": 4927
+    },
+    {
+      "epoch": 2.3545785795352727,
+      "grad_norm": 0.4670532060941091,
+      "learning_rate": 2.4298200230165713e-06,
+      "loss": 0.2706,
+      "step": 4928
+    },
+    {
+      "epoch": 2.3550564482408456,
+      "grad_norm": 0.4754697383640508,
+      "learning_rate": 2.4263798360986403e-06,
+      "loss": 0.258,
+      "step": 4929
+    },
+    {
+      "epoch": 2.355534316946419,
+      "grad_norm": 0.4958072214830247,
+      "learning_rate": 2.42294174996935e-06,
+      "loss": 0.2731,
+      "step": 4930
+    },
+    {
+      "epoch": 2.356012185651992,
+      "grad_norm": 0.4688304594840339,
+      "learning_rate": 2.4195057655823596e-06,
+      "loss": 0.2762,
+      "step": 4931
+    },
+    {
+      "epoch": 2.356490054357565,
+      "grad_norm": 0.45489501298416457,
+      "learning_rate": 2.4160718838907502e-06,
+      "loss": 0.2682,
+      "step": 4932
+    },
+    {
+      "epoch": 2.3569679230631384,
+      "grad_norm": 0.5108041983410803,
+      "learning_rate": 2.412640105847025e-06,
+      "loss": 0.2641,
+      "step": 4933
+    },
+    {
+      "epoch": 2.3574457917687117,
+      "grad_norm": 0.48773893586445766,
+      "learning_rate": 2.4092104324030952e-06,
+      "loss": 0.2573,
+      "step": 4934
+    },
+    {
+      "epoch": 2.3579236604742846,
+      "grad_norm": 0.48133666696803284,
+      "learning_rate": 2.40578286451029e-06,
+      "loss": 0.2949,
+      "step": 4935
+    },
+    {
+      "epoch": 2.358401529179858,
+      "grad_norm": 0.4912715666573927,
+      "learning_rate": 2.4023574031193607e-06,
+      "loss": 0.2742,
+      "step": 4936
+    },
+    {
+      "epoch": 2.358879397885431,
+      "grad_norm": 0.4842008813394927,
+      "learning_rate": 2.398934049180468e-06,
+      "loss": 0.2797,
+      "step": 4937
+    },
+    {
+      "epoch": 2.359357266591004,
+      "grad_norm": 0.48439100068631125,
+      "learning_rate": 2.395512803643186e-06,
+      "loss": 0.265,
+      "step": 4938
+    },
+    {
+      "epoch": 2.3598351352965774,
+      "grad_norm": 0.47578624590536817,
+      "learning_rate": 2.3920936674565155e-06,
+      "loss": 0.2881,
+      "step": 4939
+    },
+    {
+      "epoch": 2.3603130040021503,
+      "grad_norm": 0.4437343615163969,
+      "learning_rate": 2.38867664156886e-06,
+      "loss": 0.2797,
+      "step": 4940
+    },
+    {
+      "epoch": 2.3607908727077236,
+      "grad_norm": 0.46566862296582356,
+      "learning_rate": 2.38526172692804e-06,
+      "loss": 0.2719,
+      "step": 4941
+    },
+    {
+      "epoch": 2.361268741413297,
+      "grad_norm": 0.5051001842897216,
+      "learning_rate": 2.381848924481297e-06,
+      "loss": 0.2664,
+      "step": 4942
+    },
+    {
+      "epoch": 2.3617466101188698,
+      "grad_norm": 0.4806571977298347,
+      "learning_rate": 2.378438235175281e-06,
+      "loss": 0.2777,
+      "step": 4943
+    },
+    {
+      "epoch": 2.362224478824443,
+      "grad_norm": 0.4537705945452327,
+      "learning_rate": 2.375029659956054e-06,
+      "loss": 0.2692,
+      "step": 4944
+    },
+    {
+      "epoch": 2.362702347530016,
+      "grad_norm": 0.4538521904099215,
+      "learning_rate": 2.3716231997691007e-06,
+      "loss": 0.2653,
+      "step": 4945
+    },
+    {
+      "epoch": 2.3631802162355893,
+      "grad_norm": 0.44790702234683916,
+      "learning_rate": 2.368218855559309e-06,
+      "loss": 0.2547,
+      "step": 4946
+    },
+    {
+      "epoch": 2.3636580849411626,
+      "grad_norm": 0.4411839227865902,
+      "learning_rate": 2.3648166282709806e-06,
+      "loss": 0.2611,
+      "step": 4947
+    },
+    {
+      "epoch": 2.3641359536467355,
+      "grad_norm": 0.4512659481124571,
+      "learning_rate": 2.361416518847841e-06,
+      "loss": 0.2614,
+      "step": 4948
+    },
+    {
+      "epoch": 2.3646138223523088,
+      "grad_norm": 0.45377164683979093,
+      "learning_rate": 2.3580185282330137e-06,
+      "loss": 0.2686,
+      "step": 4949
+    },
+    {
+      "epoch": 2.3650916910578816,
+      "grad_norm": 0.4512279362886925,
+      "learning_rate": 2.3546226573690444e-06,
+      "loss": 0.2661,
+      "step": 4950
+    },
+    {
+      "epoch": 2.365569559763455,
+      "grad_norm": 0.4907151137339124,
+      "learning_rate": 2.351228907197891e-06,
+      "loss": 0.2725,
+      "step": 4951
+    },
+    {
+      "epoch": 2.3660474284690283,
+      "grad_norm": 0.461594964603084,
+      "learning_rate": 2.347837278660917e-06,
+      "loss": 0.274,
+      "step": 4952
+    },
+    {
+      "epoch": 2.366525297174601,
+      "grad_norm": 0.49276481493447555,
+      "learning_rate": 2.3444477726988966e-06,
+      "loss": 0.256,
+      "step": 4953
+    },
+    {
+      "epoch": 2.3670031658801745,
+      "grad_norm": 0.5271634745570248,
+      "learning_rate": 2.3410603902520245e-06,
+      "loss": 0.2734,
+      "step": 4954
+    },
+    {
+      "epoch": 2.3674810345857473,
+      "grad_norm": 0.5046047617256642,
+      "learning_rate": 2.3376751322599e-06,
+      "loss": 0.2665,
+      "step": 4955
+    },
+    {
+      "epoch": 2.3679589032913206,
+      "grad_norm": 0.43935104834974015,
+      "learning_rate": 2.33429199966153e-06,
+      "loss": 0.2542,
+      "step": 4956
+    },
+    {
+      "epoch": 2.368436771996894,
+      "grad_norm": 0.47827957926642184,
+      "learning_rate": 2.330910993395341e-06,
+      "loss": 0.2558,
+      "step": 4957
+    },
+    {
+      "epoch": 2.368914640702467,
+      "grad_norm": 0.46077109730693455,
+      "learning_rate": 2.3275321143991613e-06,
+      "loss": 0.2665,
+      "step": 4958
+    },
+    {
+      "epoch": 2.36939250940804,
+      "grad_norm": 0.4620740363224271,
+      "learning_rate": 2.324155363610231e-06,
+      "loss": 0.281,
+      "step": 4959
+    },
+    {
+      "epoch": 2.3698703781136135,
+      "grad_norm": 0.4743650364517856,
+      "learning_rate": 2.320780741965206e-06,
+      "loss": 0.2514,
+      "step": 4960
+    },
+    {
+      "epoch": 2.3703482468191863,
+      "grad_norm": 0.48881821904437056,
+      "learning_rate": 2.317408250400144e-06,
+      "loss": 0.2916,
+      "step": 4961
+    },
+    {
+      "epoch": 2.3708261155247596,
+      "grad_norm": 0.44779357419192,
+      "learning_rate": 2.3140378898505125e-06,
+      "loss": 0.2567,
+      "step": 4962
+    },
+    {
+      "epoch": 2.371303984230333,
+      "grad_norm": 0.4571272492508665,
+      "learning_rate": 2.3106696612511937e-06,
+      "loss": 0.2689,
+      "step": 4963
+    },
+    {
+      "epoch": 2.371781852935906,
+      "grad_norm": 0.4680515782108428,
+      "learning_rate": 2.307303565536474e-06,
+      "loss": 0.2632,
+      "step": 4964
+    },
+    {
+      "epoch": 2.372259721641479,
+      "grad_norm": 0.50999540376331,
+      "learning_rate": 2.3039396036400463e-06,
+      "loss": 0.2699,
+      "step": 4965
+    },
+    {
+      "epoch": 2.372737590347052,
+      "grad_norm": 0.45711754864038767,
+      "learning_rate": 2.3005777764950187e-06,
+      "loss": 0.2672,
+      "step": 4966
+    },
+    {
+      "epoch": 2.3732154590526253,
+      "grad_norm": 0.46541233157527173,
+      "learning_rate": 2.2972180850339e-06,
+      "loss": 0.2772,
+      "step": 4967
+    },
+    {
+      "epoch": 2.3736933277581986,
+      "grad_norm": 0.45372335105255907,
+      "learning_rate": 2.2938605301886075e-06,
+      "loss": 0.245,
+      "step": 4968
+    },
+    {
+      "epoch": 2.3741711964637715,
+      "grad_norm": 0.49762996831577216,
+      "learning_rate": 2.2905051128904733e-06,
+      "loss": 0.2846,
+      "step": 4969
+    },
+    {
+      "epoch": 2.374649065169345,
+      "grad_norm": 0.45309610239893056,
+      "learning_rate": 2.287151834070226e-06,
+      "loss": 0.2453,
+      "step": 4970
+    },
+    {
+      "epoch": 2.3751269338749177,
+      "grad_norm": 0.45646843129089465,
+      "learning_rate": 2.283800694658006e-06,
+      "loss": 0.2523,
+      "step": 4971
+    },
+    {
+      "epoch": 2.375604802580491,
+      "grad_norm": 0.45497585076759617,
+      "learning_rate": 2.2804516955833645e-06,
+      "loss": 0.2636,
+      "step": 4972
+    },
+    {
+      "epoch": 2.3760826712860643,
+      "grad_norm": 0.4698293678470528,
+      "learning_rate": 2.2771048377752527e-06,
+      "loss": 0.2792,
+      "step": 4973
+    },
+    {
+      "epoch": 2.376560539991637,
+      "grad_norm": 0.4481308379841722,
+      "learning_rate": 2.2737601221620252e-06,
+      "loss": 0.2734,
+      "step": 4974
+    },
+    {
+      "epoch": 2.3770384086972105,
+      "grad_norm": 0.47464422001892004,
+      "learning_rate": 2.2704175496714552e-06,
+      "loss": 0.2653,
+      "step": 4975
+    },
+    {
+      "epoch": 2.3775162774027834,
+      "grad_norm": 0.48442075929966344,
+      "learning_rate": 2.2670771212307087e-06,
+      "loss": 0.2664,
+      "step": 4976
+    },
+    {
+      "epoch": 2.3779941461083567,
+      "grad_norm": 0.46567581192533997,
+      "learning_rate": 2.2637388377663605e-06,
+      "loss": 0.2569,
+      "step": 4977
+    },
+    {
+      "epoch": 2.37847201481393,
+      "grad_norm": 0.4640997153073419,
+      "learning_rate": 2.260402700204395e-06,
+      "loss": 0.2691,
+      "step": 4978
+    },
+    {
+      "epoch": 2.378949883519503,
+      "grad_norm": 0.4635972785067223,
+      "learning_rate": 2.257068709470197e-06,
+      "loss": 0.2618,
+      "step": 4979
+    },
+    {
+      "epoch": 2.379427752225076,
+      "grad_norm": 0.4773585828087478,
+      "learning_rate": 2.2537368664885527e-06,
+      "loss": 0.2754,
+      "step": 4980
+    },
+    {
+      "epoch": 2.379905620930649,
+      "grad_norm": 0.439472956742948,
+      "learning_rate": 2.250407172183664e-06,
+      "loss": 0.2664,
+      "step": 4981
+    },
+    {
+      "epoch": 2.3803834896362224,
+      "grad_norm": 0.4558716330221375,
+      "learning_rate": 2.247079627479124e-06,
+      "loss": 0.2683,
+      "step": 4982
+    },
+    {
+      "epoch": 2.3808613583417957,
+      "grad_norm": 0.4608097977312484,
+      "learning_rate": 2.2437542332979336e-06,
+      "loss": 0.2835,
+      "step": 4983
+    },
+    {
+      "epoch": 2.3813392270473686,
+      "grad_norm": 0.4380401331798129,
+      "learning_rate": 2.240430990562501e-06,
+      "loss": 0.2919,
+      "step": 4984
+    },
+    {
+      "epoch": 2.381817095752942,
+      "grad_norm": 0.5895395029494848,
+      "learning_rate": 2.2371099001946385e-06,
+      "loss": 0.2624,
+      "step": 4985
+    },
+    {
+      "epoch": 2.382294964458515,
+      "grad_norm": 0.46508326131778577,
+      "learning_rate": 2.233790963115554e-06,
+      "loss": 0.2504,
+      "step": 4986
+    },
+    {
+      "epoch": 2.382772833164088,
+      "grad_norm": 0.4477530552158064,
+      "learning_rate": 2.2304741802458606e-06,
+      "loss": 0.2661,
+      "step": 4987
+    },
+    {
+      "epoch": 2.3832507018696614,
+      "grad_norm": 0.45202080179265264,
+      "learning_rate": 2.22715955250558e-06,
+      "loss": 0.2801,
+      "step": 4988
+    },
+    {
+      "epoch": 2.3837285705752347,
+      "grad_norm": 0.46022019801647834,
+      "learning_rate": 2.223847080814129e-06,
+      "loss": 0.2522,
+      "step": 4989
+    },
+    {
+      "epoch": 2.3842064392808076,
+      "grad_norm": 0.46337347162678094,
+      "learning_rate": 2.2205367660903267e-06,
+      "loss": 0.2622,
+      "step": 4990
+    },
+    {
+      "epoch": 2.384684307986381,
+      "grad_norm": 0.532373751607701,
+      "learning_rate": 2.2172286092523998e-06,
+      "loss": 0.253,
+      "step": 4991
+    },
+    {
+      "epoch": 2.3851621766919537,
+      "grad_norm": 0.44622554326728614,
+      "learning_rate": 2.2139226112179713e-06,
+      "loss": 0.2566,
+      "step": 4992
+    },
+    {
+      "epoch": 2.385640045397527,
+      "grad_norm": 0.4538870200700637,
+      "learning_rate": 2.210618772904064e-06,
+      "loss": 0.2646,
+      "step": 4993
+    },
+    {
+      "epoch": 2.3861179141031004,
+      "grad_norm": 0.4585327419151689,
+      "learning_rate": 2.2073170952271085e-06,
+      "loss": 0.2623,
+      "step": 4994
+    },
+    {
+      "epoch": 2.3865957828086732,
+      "grad_norm": 0.46815712443376617,
+      "learning_rate": 2.2040175791029305e-06,
+      "loss": 0.2818,
+      "step": 4995
+    },
+    {
+      "epoch": 2.3870736515142466,
+      "grad_norm": 0.5231070665972924,
+      "learning_rate": 2.200720225446755e-06,
+      "loss": 0.2722,
+      "step": 4996
+    },
+    {
+      "epoch": 2.3875515202198194,
+      "grad_norm": 0.654443625332581,
+      "learning_rate": 2.197425035173215e-06,
+      "loss": 0.2563,
+      "step": 4997
+    },
+    {
+      "epoch": 2.3880293889253927,
+      "grad_norm": 0.45803873330264716,
+      "learning_rate": 2.194132009196335e-06,
+      "loss": 0.2952,
+      "step": 4998
+    },
+    {
+      "epoch": 2.388507257630966,
+      "grad_norm": 0.4521097635127778,
+      "learning_rate": 2.19084114842954e-06,
+      "loss": 0.265,
+      "step": 4999
+    },
+    {
+      "epoch": 2.388985126336539,
+      "grad_norm": 0.4544732994444673,
+      "learning_rate": 2.187552453785662e-06,
+      "loss": 0.2745,
+      "step": 5000
+    },
+    {
+      "epoch": 2.3894629950421122,
+      "grad_norm": 0.4793264716313783,
+      "learning_rate": 2.1842659261769226e-06,
+      "loss": 0.262,
+      "step": 5001
+    },
+    {
+      "epoch": 2.389940863747685,
+      "grad_norm": 0.448019751252878,
+      "learning_rate": 2.180981566514947e-06,
+      "loss": 0.2688,
+      "step": 5002
+    },
+    {
+      "epoch": 2.3904187324532584,
+      "grad_norm": 0.48536454285303543,
+      "learning_rate": 2.177699375710762e-06,
+      "loss": 0.2623,
+      "step": 5003
+    },
+    {
+      "epoch": 2.3908966011588317,
+      "grad_norm": 0.6915355830232118,
+      "learning_rate": 2.174419354674787e-06,
+      "loss": 0.2697,
+      "step": 5004
+    },
+    {
+      "epoch": 2.3913744698644046,
+      "grad_norm": 0.47423927727053033,
+      "learning_rate": 2.1711415043168395e-06,
+      "loss": 0.2806,
+      "step": 5005
+    },
+    {
+      "epoch": 2.391852338569978,
+      "grad_norm": 0.4699229189597673,
+      "learning_rate": 2.1678658255461427e-06,
+      "loss": 0.2764,
+      "step": 5006
+    },
+    {
+      "epoch": 2.3923302072755512,
+      "grad_norm": 0.4366728988564249,
+      "learning_rate": 2.164592319271309e-06,
+      "loss": 0.2626,
+      "step": 5007
+    },
+    {
+      "epoch": 2.392808075981124,
+      "grad_norm": 0.4680895354927544,
+      "learning_rate": 2.16132098640035e-06,
+      "loss": 0.2856,
+      "step": 5008
+    },
+    {
+      "epoch": 2.3932859446866974,
+      "grad_norm": 0.4831506206251398,
+      "learning_rate": 2.1580518278406793e-06,
+      "loss": 0.2588,
+      "step": 5009
+    },
+    {
+      "epoch": 2.3937638133922703,
+      "grad_norm": 0.4699558577692185,
+      "learning_rate": 2.1547848444991025e-06,
+      "loss": 0.2701,
+      "step": 5010
+    },
+    {
+      "epoch": 2.3942416820978436,
+      "grad_norm": 0.4489222989978539,
+      "learning_rate": 2.15152003728182e-06,
+      "loss": 0.2834,
+      "step": 5011
+    },
+    {
+      "epoch": 2.394719550803417,
+      "grad_norm": 0.47665521913868575,
+      "learning_rate": 2.148257407094436e-06,
+      "loss": 0.2583,
+      "step": 5012
+    },
+    {
+      "epoch": 2.39519741950899,
+      "grad_norm": 0.45743260638815375,
+      "learning_rate": 2.1449969548419456e-06,
+      "loss": 0.2677,
+      "step": 5013
+    },
+    {
+      "epoch": 2.395675288214563,
+      "grad_norm": 0.45342448226517174,
+      "learning_rate": 2.141738681428738e-06,
+      "loss": 0.2653,
+      "step": 5014
+    },
+    {
+      "epoch": 2.3961531569201364,
+      "grad_norm": 0.44238225246261725,
+      "learning_rate": 2.138482587758605e-06,
+      "loss": 0.2823,
+      "step": 5015
+    },
+    {
+      "epoch": 2.3966310256257093,
+      "grad_norm": 0.4457574279651664,
+      "learning_rate": 2.1352286747347273e-06,
+      "loss": 0.257,
+      "step": 5016
+    },
+    {
+      "epoch": 2.3971088943312826,
+      "grad_norm": 0.4702199309191923,
+      "learning_rate": 2.1319769432596804e-06,
+      "loss": 0.276,
+      "step": 5017
+    },
+    {
+      "epoch": 2.3975867630368555,
+      "grad_norm": 0.48532027343768513,
+      "learning_rate": 2.1287273942354393e-06,
+      "loss": 0.2615,
+      "step": 5018
+    },
+    {
+      "epoch": 2.398064631742429,
+      "grad_norm": 0.4532406315784221,
+      "learning_rate": 2.125480028563376e-06,
+      "loss": 0.2611,
+      "step": 5019
+    },
+    {
+      "epoch": 2.398542500448002,
+      "grad_norm": 0.43656492650267525,
+      "learning_rate": 2.1222348471442477e-06,
+      "loss": 0.2607,
+      "step": 5020
+    },
+    {
+      "epoch": 2.399020369153575,
+      "grad_norm": 0.4437139617252678,
+      "learning_rate": 2.118991850878209e-06,
+      "loss": 0.2634,
+      "step": 5021
+    },
+    {
+      "epoch": 2.3994982378591483,
+      "grad_norm": 0.4607343235133938,
+      "learning_rate": 2.115751040664815e-06,
+      "loss": 0.2617,
+      "step": 5022
+    },
+    {
+      "epoch": 2.399976106564721,
+      "grad_norm": 0.4509401894862981,
+      "learning_rate": 2.1125124174030066e-06,
+      "loss": 0.2649,
+      "step": 5023
+    },
+    {
+      "epoch": 2.4004539752702945,
+      "grad_norm": 0.4780514985721637,
+      "learning_rate": 2.1092759819911178e-06,
+      "loss": 0.2587,
+      "step": 5024
+    },
+    {
+      "epoch": 2.400931843975868,
+      "grad_norm": 0.47293976576430685,
+      "learning_rate": 2.1060417353268845e-06,
+      "loss": 0.2605,
+      "step": 5025
+    },
+    {
+      "epoch": 2.4014097126814407,
+      "grad_norm": 0.4632747242085837,
+      "learning_rate": 2.102809678307427e-06,
+      "loss": 0.2701,
+      "step": 5026
+    },
+    {
+      "epoch": 2.401887581387014,
+      "grad_norm": 0.4519610118896511,
+      "learning_rate": 2.0995798118292574e-06,
+      "loss": 0.274,
+      "step": 5027
+    },
+    {
+      "epoch": 2.402365450092587,
+      "grad_norm": 0.49007996568127915,
+      "learning_rate": 2.09635213678829e-06,
+      "loss": 0.275,
+      "step": 5028
+    },
+    {
+      "epoch": 2.40284331879816,
+      "grad_norm": 0.4482075654858903,
+      "learning_rate": 2.093126654079822e-06,
+      "loss": 0.2724,
+      "step": 5029
+    },
+    {
+      "epoch": 2.4033211875037335,
+      "grad_norm": 0.46175751639190643,
+      "learning_rate": 2.0899033645985423e-06,
+      "loss": 0.2734,
+      "step": 5030
+    },
+    {
+      "epoch": 2.4037990562093063,
+      "grad_norm": 0.4526383312411215,
+      "learning_rate": 2.0866822692385404e-06,
+      "loss": 0.2845,
+      "step": 5031
+    },
+    {
+      "epoch": 2.4042769249148797,
+      "grad_norm": 0.4713547161425226,
+      "learning_rate": 2.083463368893289e-06,
+      "loss": 0.2838,
+      "step": 5032
+    },
+    {
+      "epoch": 2.404754793620453,
+      "grad_norm": 0.5388592684225536,
+      "learning_rate": 2.0802466644556507e-06,
+      "loss": 0.2581,
+      "step": 5033
+    },
+    {
+      "epoch": 2.405232662326026,
+      "grad_norm": 0.44781970240862895,
+      "learning_rate": 2.0770321568178873e-06,
+      "loss": 0.2552,
+      "step": 5034
+    },
+    {
+      "epoch": 2.405710531031599,
+      "grad_norm": 0.45727753815499,
+      "learning_rate": 2.073819846871646e-06,
+      "loss": 0.2784,
+      "step": 5035
+    },
+    {
+      "epoch": 2.406188399737172,
+      "grad_norm": 0.4523294487636907,
+      "learning_rate": 2.0706097355079614e-06,
+      "loss": 0.2548,
+      "step": 5036
+    },
+    {
+      "epoch": 2.4066662684427453,
+      "grad_norm": 0.4979869120677124,
+      "learning_rate": 2.0674018236172654e-06,
+      "loss": 0.262,
+      "step": 5037
+    },
+    {
+      "epoch": 2.4071441371483187,
+      "grad_norm": 0.4670045351404005,
+      "learning_rate": 2.064196112089376e-06,
+      "loss": 0.2836,
+      "step": 5038
+    },
+    {
+      "epoch": 2.4076220058538915,
+      "grad_norm": 0.44302766764113355,
+      "learning_rate": 2.0609926018134972e-06,
+      "loss": 0.2544,
+      "step": 5039
+    },
+    {
+      "epoch": 2.408099874559465,
+      "grad_norm": 0.4466796302647838,
+      "learning_rate": 2.0577912936782317e-06,
+      "loss": 0.2531,
+      "step": 5040
+    },
+    {
+      "epoch": 2.408577743265038,
+      "grad_norm": 0.4662441229505657,
+      "learning_rate": 2.0545921885715624e-06,
+      "loss": 0.2634,
+      "step": 5041
+    },
+    {
+      "epoch": 2.409055611970611,
+      "grad_norm": 0.4621440744087701,
+      "learning_rate": 2.051395287380864e-06,
+      "loss": 0.2711,
+      "step": 5042
+    },
+    {
+      "epoch": 2.4095334806761843,
+      "grad_norm": 0.4910784013605513,
+      "learning_rate": 2.048200590992904e-06,
+      "loss": 0.2564,
+      "step": 5043
+    },
+    {
+      "epoch": 2.410011349381757,
+      "grad_norm": 0.4768608798682245,
+      "learning_rate": 2.0450081002938316e-06,
+      "loss": 0.2643,
+      "step": 5044
+    },
+    {
+      "epoch": 2.4104892180873305,
+      "grad_norm": 0.45467660225646006,
+      "learning_rate": 2.041817816169187e-06,
+      "loss": 0.2567,
+      "step": 5045
+    },
+    {
+      "epoch": 2.410967086792904,
+      "grad_norm": 0.4570097506050878,
+      "learning_rate": 2.0386297395039023e-06,
+      "loss": 0.2681,
+      "step": 5046
+    },
+    {
+      "epoch": 2.4114449554984767,
+      "grad_norm": 0.4718064608424706,
+      "learning_rate": 2.035443871182292e-06,
+      "loss": 0.2523,
+      "step": 5047
+    },
+    {
+      "epoch": 2.41192282420405,
+      "grad_norm": 0.510560238857416,
+      "learning_rate": 2.0322602120880576e-06,
+      "loss": 0.2596,
+      "step": 5048
+    },
+    {
+      "epoch": 2.412400692909623,
+      "grad_norm": 0.45874298526457225,
+      "learning_rate": 2.0290787631042942e-06,
+      "loss": 0.2654,
+      "step": 5049
+    },
+    {
+      "epoch": 2.412878561615196,
+      "grad_norm": 0.4702898124625645,
+      "learning_rate": 2.025899525113474e-06,
+      "loss": 0.2607,
+      "step": 5050
+    },
+    {
+      "epoch": 2.4133564303207695,
+      "grad_norm": 0.4419385668349205,
+      "learning_rate": 2.022722498997465e-06,
+      "loss": 0.28,
+      "step": 5051
+    },
+    {
+      "epoch": 2.4138342990263424,
+      "grad_norm": 0.49832278835296895,
+      "learning_rate": 2.0195476856375206e-06,
+      "loss": 0.2722,
+      "step": 5052
+    },
+    {
+      "epoch": 2.4143121677319157,
+      "grad_norm": 0.4453378413752286,
+      "learning_rate": 2.016375085914275e-06,
+      "loss": 0.2631,
+      "step": 5053
+    },
+    {
+      "epoch": 2.4147900364374886,
+      "grad_norm": 0.4464873807438464,
+      "learning_rate": 2.0132047007077504e-06,
+      "loss": 0.2617,
+      "step": 5054
+    },
+    {
+      "epoch": 2.415267905143062,
+      "grad_norm": 0.43769077224069114,
+      "learning_rate": 2.010036530897359e-06,
+      "loss": 0.2523,
+      "step": 5055
+    },
+    {
+      "epoch": 2.415745773848635,
+      "grad_norm": 0.44478509964099755,
+      "learning_rate": 2.0068705773618937e-06,
+      "loss": 0.2494,
+      "step": 5056
+    },
+    {
+      "epoch": 2.416223642554208,
+      "grad_norm": 0.4558953495461828,
+      "learning_rate": 2.003706840979531e-06,
+      "loss": 0.2716,
+      "step": 5057
+    },
+    {
+      "epoch": 2.4167015112597814,
+      "grad_norm": 0.7657464247162521,
+      "learning_rate": 2.0005453226278403e-06,
+      "loss": 0.2683,
+      "step": 5058
+    },
+    {
+      "epoch": 2.4171793799653547,
+      "grad_norm": 0.4719546599335128,
+      "learning_rate": 1.9973860231837705e-06,
+      "loss": 0.256,
+      "step": 5059
+    },
+    {
+      "epoch": 2.4176572486709276,
+      "grad_norm": 0.46006978419118905,
+      "learning_rate": 1.9942289435236506e-06,
+      "loss": 0.2772,
+      "step": 5060
+    },
+    {
+      "epoch": 2.418135117376501,
+      "grad_norm": 0.5094940178312064,
+      "learning_rate": 1.9910740845232058e-06,
+      "loss": 0.2713,
+      "step": 5061
+    },
+    {
+      "epoch": 2.4186129860820738,
+      "grad_norm": 0.46698646284000606,
+      "learning_rate": 1.9879214470575347e-06,
+      "loss": 0.274,
+      "step": 5062
+    },
+    {
+      "epoch": 2.419090854787647,
+      "grad_norm": 0.4589782752776786,
+      "learning_rate": 1.9847710320011206e-06,
+      "loss": 0.2499,
+      "step": 5063
+    },
+    {
+      "epoch": 2.4195687234932204,
+      "grad_norm": 0.469003199954616,
+      "learning_rate": 1.9816228402278392e-06,
+      "loss": 0.2681,
+      "step": 5064
+    },
+    {
+      "epoch": 2.4200465921987933,
+      "grad_norm": 0.4481176604472585,
+      "learning_rate": 1.978476872610939e-06,
+      "loss": 0.2577,
+      "step": 5065
+    },
+    {
+      "epoch": 2.4205244609043666,
+      "grad_norm": 0.4659618384134064,
+      "learning_rate": 1.975333130023056e-06,
+      "loss": 0.2647,
+      "step": 5066
+    },
+    {
+      "epoch": 2.42100232960994,
+      "grad_norm": 0.4828500059187709,
+      "learning_rate": 1.972191613336212e-06,
+      "loss": 0.2622,
+      "step": 5067
+    },
+    {
+      "epoch": 2.4214801983155128,
+      "grad_norm": 0.4551712653015207,
+      "learning_rate": 1.969052323421806e-06,
+      "loss": 0.2661,
+      "step": 5068
+    },
+    {
+      "epoch": 2.421958067021086,
+      "grad_norm": 0.46637572776776237,
+      "learning_rate": 1.9659152611506193e-06,
+      "loss": 0.2722,
+      "step": 5069
+    },
+    {
+      "epoch": 2.422435935726659,
+      "grad_norm": 0.4486269710544213,
+      "learning_rate": 1.962780427392823e-06,
+      "loss": 0.2742,
+      "step": 5070
+    },
+    {
+      "epoch": 2.4229138044322323,
+      "grad_norm": 0.4481995215441604,
+      "learning_rate": 1.959647823017963e-06,
+      "loss": 0.2709,
+      "step": 5071
+    },
+    {
+      "epoch": 2.4233916731378056,
+      "grad_norm": 0.4637020847420594,
+      "learning_rate": 1.9565174488949636e-06,
+      "loss": 0.268,
+      "step": 5072
+    },
+    {
+      "epoch": 2.4238695418433784,
+      "grad_norm": 0.4522661575040067,
+      "learning_rate": 1.953389305892143e-06,
+      "loss": 0.2714,
+      "step": 5073
+    },
+    {
+      "epoch": 2.4243474105489518,
+      "grad_norm": 0.5188133426193942,
+      "learning_rate": 1.9502633948771888e-06,
+      "loss": 0.2682,
+      "step": 5074
+    },
+    {
+      "epoch": 2.4248252792545246,
+      "grad_norm": 0.4570248324569066,
+      "learning_rate": 1.9471397167171714e-06,
+      "loss": 0.2652,
+      "step": 5075
+    },
+    {
+      "epoch": 2.425303147960098,
+      "grad_norm": 0.459560178829993,
+      "learning_rate": 1.944018272278548e-06,
+      "loss": 0.2511,
+      "step": 5076
+    },
+    {
+      "epoch": 2.4257810166656713,
+      "grad_norm": 0.466238298627355,
+      "learning_rate": 1.9408990624271516e-06,
+      "loss": 0.2687,
+      "step": 5077
+    },
+    {
+      "epoch": 2.426258885371244,
+      "grad_norm": 0.49223929370166464,
+      "learning_rate": 1.9377820880281928e-06,
+      "loss": 0.2739,
+      "step": 5078
+    },
+    {
+      "epoch": 2.4267367540768174,
+      "grad_norm": 0.46613653698623686,
+      "learning_rate": 1.934667349946271e-06,
+      "loss": 0.2676,
+      "step": 5079
+    },
+    {
+      "epoch": 2.4272146227823903,
+      "grad_norm": 0.44196513176630936,
+      "learning_rate": 1.931554849045355e-06,
+      "loss": 0.2662,
+      "step": 5080
+    },
+    {
+      "epoch": 2.4276924914879636,
+      "grad_norm": 0.45591063669478454,
+      "learning_rate": 1.9284445861887966e-06,
+      "loss": 0.2799,
+      "step": 5081
+    },
+    {
+      "epoch": 2.428170360193537,
+      "grad_norm": 0.47960546577577334,
+      "learning_rate": 1.9253365622393337e-06,
+      "loss": 0.2612,
+      "step": 5082
+    },
+    {
+      "epoch": 2.42864822889911,
+      "grad_norm": 0.49874783389566263,
+      "learning_rate": 1.9222307780590734e-06,
+      "loss": 0.259,
+      "step": 5083
+    },
+    {
+      "epoch": 2.429126097604683,
+      "grad_norm": 0.48649273799607196,
+      "learning_rate": 1.9191272345095025e-06,
+      "loss": 0.2626,
+      "step": 5084
+    },
+    {
+      "epoch": 2.4296039663102564,
+      "grad_norm": 0.46265212095706404,
+      "learning_rate": 1.916025932451493e-06,
+      "loss": 0.2673,
+      "step": 5085
+    },
+    {
+      "epoch": 2.4300818350158293,
+      "grad_norm": 0.44902330400044727,
+      "learning_rate": 1.912926872745294e-06,
+      "loss": 0.2641,
+      "step": 5086
+    },
+    {
+      "epoch": 2.4305597037214026,
+      "grad_norm": 0.48237243869527174,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.2677,
+      "step": 5087
+    },
+    {
+      "epoch": 2.431037572426976,
+      "grad_norm": 0.47922262239457364,
+      "learning_rate": 1.9067354838261908e-06,
+      "loss": 0.239,
+      "step": 5088
+    },
+    {
+      "epoch": 2.431515441132549,
+      "grad_norm": 0.4484227748244973,
+      "learning_rate": 1.9036431563306723e-06,
+      "loss": 0.2676,
+      "step": 5089
+    },
+    {
+      "epoch": 2.431993309838122,
+      "grad_norm": 0.4481486490497434,
+      "learning_rate": 1.9005530746217238e-06,
+      "loss": 0.2612,
+      "step": 5090
+    },
+    {
+      "epoch": 2.432471178543695,
+      "grad_norm": 0.4751868893945265,
+      "learning_rate": 1.8974652395564785e-06,
+      "loss": 0.2658,
+      "step": 5091
+    },
+    {
+      "epoch": 2.4329490472492683,
+      "grad_norm": 0.4910217856403449,
+      "learning_rate": 1.8943796519914525e-06,
+      "loss": 0.2632,
+      "step": 5092
+    },
+    {
+      "epoch": 2.4334269159548416,
+      "grad_norm": 0.4671224266607131,
+      "learning_rate": 1.89129631278253e-06,
+      "loss": 0.2804,
+      "step": 5093
+    },
+    {
+      "epoch": 2.4339047846604145,
+      "grad_norm": 0.5193462465411075,
+      "learning_rate": 1.8882152227849727e-06,
+      "loss": 0.2845,
+      "step": 5094
+    },
+    {
+      "epoch": 2.434382653365988,
+      "grad_norm": 0.5153634268527411,
+      "learning_rate": 1.8851363828534253e-06,
+      "loss": 0.263,
+      "step": 5095
+    },
+    {
+      "epoch": 2.4348605220715607,
+      "grad_norm": 0.45007658573232884,
+      "learning_rate": 1.8820597938419028e-06,
+      "loss": 0.2653,
+      "step": 5096
+    },
+    {
+      "epoch": 2.435338390777134,
+      "grad_norm": 0.4708101712892127,
+      "learning_rate": 1.8789854566037912e-06,
+      "loss": 0.2681,
+      "step": 5097
+    },
+    {
+      "epoch": 2.4358162594827073,
+      "grad_norm": 0.466359009946716,
+      "learning_rate": 1.8759133719918654e-06,
+      "loss": 0.2683,
+      "step": 5098
+    },
+    {
+      "epoch": 2.43629412818828,
+      "grad_norm": 0.5260130493552598,
+      "learning_rate": 1.8728435408582634e-06,
+      "loss": 0.2664,
+      "step": 5099
+    },
+    {
+      "epoch": 2.4367719968938535,
+      "grad_norm": 0.4471972729113941,
+      "learning_rate": 1.869775964054501e-06,
+      "loss": 0.2667,
+      "step": 5100
+    },
+    {
+      "epoch": 2.4372498655994264,
+      "grad_norm": 0.4605895897583309,
+      "learning_rate": 1.866710642431473e-06,
+      "loss": 0.2794,
+      "step": 5101
+    },
+    {
+      "epoch": 2.4377277343049997,
+      "grad_norm": 0.49147254180946887,
+      "learning_rate": 1.8636475768394446e-06,
+      "loss": 0.2812,
+      "step": 5102
+    },
+    {
+      "epoch": 2.438205603010573,
+      "grad_norm": 0.44497897456741325,
+      "learning_rate": 1.860586768128052e-06,
+      "loss": 0.2679,
+      "step": 5103
+    },
+    {
+      "epoch": 2.438683471716146,
+      "grad_norm": 0.45664419543781765,
+      "learning_rate": 1.857528217146317e-06,
+      "loss": 0.2631,
+      "step": 5104
+    },
+    {
+      "epoch": 2.439161340421719,
+      "grad_norm": 0.463599939436042,
+      "learning_rate": 1.8544719247426224e-06,
+      "loss": 0.2664,
+      "step": 5105
+    },
+    {
+      "epoch": 2.439639209127292,
+      "grad_norm": 0.5020393530396908,
+      "learning_rate": 1.8514178917647297e-06,
+      "loss": 0.2751,
+      "step": 5106
+    },
+    {
+      "epoch": 2.4401170778328654,
+      "grad_norm": 0.6334592929322761,
+      "learning_rate": 1.8483661190597778e-06,
+      "loss": 0.273,
+      "step": 5107
+    },
+    {
+      "epoch": 2.4405949465384387,
+      "grad_norm": 0.44949715968087817,
+      "learning_rate": 1.8453166074742723e-06,
+      "loss": 0.2626,
+      "step": 5108
+    },
+    {
+      "epoch": 2.4410728152440115,
+      "grad_norm": 0.4601875206689804,
+      "learning_rate": 1.8422693578540907e-06,
+      "loss": 0.2524,
+      "step": 5109
+    },
+    {
+      "epoch": 2.441550683949585,
+      "grad_norm": 0.4432035611183387,
+      "learning_rate": 1.8392243710444911e-06,
+      "loss": 0.2568,
+      "step": 5110
+    },
+    {
+      "epoch": 2.442028552655158,
+      "grad_norm": 0.4420828252613538,
+      "learning_rate": 1.8361816478900986e-06,
+      "loss": 0.2687,
+      "step": 5111
+    },
+    {
+      "epoch": 2.442506421360731,
+      "grad_norm": 0.4698539634957639,
+      "learning_rate": 1.833141189234907e-06,
+      "loss": 0.2556,
+      "step": 5112
+    },
+    {
+      "epoch": 2.4429842900663044,
+      "grad_norm": 0.4454077363137819,
+      "learning_rate": 1.8301029959222916e-06,
+      "loss": 0.2644,
+      "step": 5113
+    },
+    {
+      "epoch": 2.4434621587718777,
+      "grad_norm": 0.4694638175516713,
+      "learning_rate": 1.8270670687949898e-06,
+      "loss": 0.2725,
+      "step": 5114
+    },
+    {
+      "epoch": 2.4439400274774505,
+      "grad_norm": 0.4608350257915752,
+      "learning_rate": 1.8240334086951117e-06,
+      "loss": 0.2696,
+      "step": 5115
+    },
+    {
+      "epoch": 2.444417896183024,
+      "grad_norm": 0.44781409506792363,
+      "learning_rate": 1.8210020164641483e-06,
+      "loss": 0.2716,
+      "step": 5116
+    },
+    {
+      "epoch": 2.4448957648885967,
+      "grad_norm": 0.4969053175588786,
+      "learning_rate": 1.8179728929429507e-06,
+      "loss": 0.2643,
+      "step": 5117
+    },
+    {
+      "epoch": 2.44537363359417,
+      "grad_norm": 0.4559222255934472,
+      "learning_rate": 1.814946038971741e-06,
+      "loss": 0.2713,
+      "step": 5118
+    },
+    {
+      "epoch": 2.4458515022997434,
+      "grad_norm": 0.44970009381193676,
+      "learning_rate": 1.8119214553901177e-06,
+      "loss": 0.2622,
+      "step": 5119
+    },
+    {
+      "epoch": 2.4463293710053162,
+      "grad_norm": 0.4559369781065068,
+      "learning_rate": 1.8088991430370506e-06,
+      "loss": 0.2666,
+      "step": 5120
+    },
+    {
+      "epoch": 2.4468072397108895,
+      "grad_norm": 0.45234671729475956,
+      "learning_rate": 1.8058791027508726e-06,
+      "loss": 0.2686,
+      "step": 5121
+    },
+    {
+      "epoch": 2.4472851084164624,
+      "grad_norm": 0.44137882355941305,
+      "learning_rate": 1.802861335369287e-06,
+      "loss": 0.2779,
+      "step": 5122
+    },
+    {
+      "epoch": 2.4477629771220357,
+      "grad_norm": 0.48440328157222995,
+      "learning_rate": 1.799845841729375e-06,
+      "loss": 0.283,
+      "step": 5123
+    },
+    {
+      "epoch": 2.448240845827609,
+      "grad_norm": 0.494486208129345,
+      "learning_rate": 1.796832622667578e-06,
+      "loss": 0.2656,
+      "step": 5124
+    },
+    {
+      "epoch": 2.448718714533182,
+      "grad_norm": 0.47590323638368665,
+      "learning_rate": 1.7938216790197071e-06,
+      "loss": 0.2497,
+      "step": 5125
+    },
+    {
+      "epoch": 2.449196583238755,
+      "grad_norm": 0.449129199222899,
+      "learning_rate": 1.790813011620951e-06,
+      "loss": 0.2654,
+      "step": 5126
+    },
+    {
+      "epoch": 2.449674451944328,
+      "grad_norm": 0.4626004740804751,
+      "learning_rate": 1.7878066213058575e-06,
+      "loss": 0.2708,
+      "step": 5127
+    },
+    {
+      "epoch": 2.4501523206499014,
+      "grad_norm": 0.4702626534514859,
+      "learning_rate": 1.7848025089083442e-06,
+      "loss": 0.2782,
+      "step": 5128
+    },
+    {
+      "epoch": 2.4506301893554747,
+      "grad_norm": 0.44737255344314847,
+      "learning_rate": 1.7818006752617034e-06,
+      "loss": 0.2791,
+      "step": 5129
+    },
+    {
+      "epoch": 2.4511080580610476,
+      "grad_norm": 0.454394369273916,
+      "learning_rate": 1.77880112119859e-06,
+      "loss": 0.2579,
+      "step": 5130
+    },
+    {
+      "epoch": 2.451585926766621,
+      "grad_norm": 0.4418070881214252,
+      "learning_rate": 1.7758038475510232e-06,
+      "loss": 0.2703,
+      "step": 5131
+    },
+    {
+      "epoch": 2.4520637954721938,
+      "grad_norm": 0.4336894058264998,
+      "learning_rate": 1.7728088551503986e-06,
+      "loss": 0.2688,
+      "step": 5132
+    },
+    {
+      "epoch": 2.452541664177767,
+      "grad_norm": 0.46001029127787707,
+      "learning_rate": 1.769816144827472e-06,
+      "loss": 0.2532,
+      "step": 5133
+    },
+    {
+      "epoch": 2.4530195328833404,
+      "grad_norm": 0.4507732662463414,
+      "learning_rate": 1.7668257174123672e-06,
+      "loss": 0.2725,
+      "step": 5134
+    },
+    {
+      "epoch": 2.4534974015889133,
+      "grad_norm": 0.5054107742226862,
+      "learning_rate": 1.7638375737345804e-06,
+      "loss": 0.2612,
+      "step": 5135
+    },
+    {
+      "epoch": 2.4539752702944866,
+      "grad_norm": 0.5719525646703848,
+      "learning_rate": 1.7608517146229677e-06,
+      "loss": 0.2622,
+      "step": 5136
+    },
+    {
+      "epoch": 2.45445313900006,
+      "grad_norm": 0.45263029420038425,
+      "learning_rate": 1.7578681409057497e-06,
+      "loss": 0.2571,
+      "step": 5137
+    },
+    {
+      "epoch": 2.4549310077056328,
+      "grad_norm": 0.4469493461520194,
+      "learning_rate": 1.7548868534105234e-06,
+      "loss": 0.2795,
+      "step": 5138
+    },
+    {
+      "epoch": 2.455408876411206,
+      "grad_norm": 0.47460938094773264,
+      "learning_rate": 1.751907852964243e-06,
+      "loss": 0.2535,
+      "step": 5139
+    },
+    {
+      "epoch": 2.4558867451167794,
+      "grad_norm": 0.43572000742600314,
+      "learning_rate": 1.7489311403932274e-06,
+      "loss": 0.2616,
+      "step": 5140
+    },
+    {
+      "epoch": 2.4563646138223523,
+      "grad_norm": 0.49154307546187714,
+      "learning_rate": 1.7459567165231695e-06,
+      "loss": 0.2679,
+      "step": 5141
+    },
+    {
+      "epoch": 2.4568424825279256,
+      "grad_norm": 0.43369075353473246,
+      "learning_rate": 1.7429845821791202e-06,
+      "loss": 0.2641,
+      "step": 5142
+    },
+    {
+      "epoch": 2.4573203512334985,
+      "grad_norm": 0.44569959028444417,
+      "learning_rate": 1.7400147381854936e-06,
+      "loss": 0.2556,
+      "step": 5143
+    },
+    {
+      "epoch": 2.4577982199390718,
+      "grad_norm": 0.44301724734443204,
+      "learning_rate": 1.7370471853660775e-06,
+      "loss": 0.2594,
+      "step": 5144
+    },
+    {
+      "epoch": 2.458276088644645,
+      "grad_norm": 0.4331333804341633,
+      "learning_rate": 1.7340819245440166e-06,
+      "loss": 0.2698,
+      "step": 5145
+    },
+    {
+      "epoch": 2.458753957350218,
+      "grad_norm": 0.44930485499724493,
+      "learning_rate": 1.7311189565418186e-06,
+      "loss": 0.2694,
+      "step": 5146
+    },
+    {
+      "epoch": 2.4592318260557913,
+      "grad_norm": 0.4431936634915009,
+      "learning_rate": 1.7281582821813647e-06,
+      "loss": 0.2539,
+      "step": 5147
+    },
+    {
+      "epoch": 2.459709694761364,
+      "grad_norm": 0.4801540861517632,
+      "learning_rate": 1.7251999022838895e-06,
+      "loss": 0.26,
+      "step": 5148
+    },
+    {
+      "epoch": 2.4601875634669375,
+      "grad_norm": 0.44046656799275724,
+      "learning_rate": 1.722243817669994e-06,
+      "loss": 0.2602,
+      "step": 5149
+    },
+    {
+      "epoch": 2.4606654321725108,
+      "grad_norm": 0.4569712842143844,
+      "learning_rate": 1.7192900291596493e-06,
+      "loss": 0.2677,
+      "step": 5150
+    },
+    {
+      "epoch": 2.4611433008780836,
+      "grad_norm": 0.5414744354026383,
+      "learning_rate": 1.7163385375721819e-06,
+      "loss": 0.2657,
+      "step": 5151
+    },
+    {
+      "epoch": 2.461621169583657,
+      "grad_norm": 0.5030301613387165,
+      "learning_rate": 1.7133893437262771e-06,
+      "loss": 0.2694,
+      "step": 5152
+    },
+    {
+      "epoch": 2.46209903828923,
+      "grad_norm": 0.46022742529581706,
+      "learning_rate": 1.7104424484400006e-06,
+      "loss": 0.2657,
+      "step": 5153
+    },
+    {
+      "epoch": 2.462576906994803,
+      "grad_norm": 0.4480168925969945,
+      "learning_rate": 1.707497852530763e-06,
+      "loss": 0.2678,
+      "step": 5154
+    },
+    {
+      "epoch": 2.4630547757003765,
+      "grad_norm": 0.45217430934005776,
+      "learning_rate": 1.7045555568153415e-06,
+      "loss": 0.2573,
+      "step": 5155
+    },
+    {
+      "epoch": 2.4635326444059493,
+      "grad_norm": 0.4771262260653937,
+      "learning_rate": 1.7016155621098818e-06,
+      "loss": 0.2663,
+      "step": 5156
+    },
+    {
+      "epoch": 2.4640105131115226,
+      "grad_norm": 0.43652614153589214,
+      "learning_rate": 1.6986778692298843e-06,
+      "loss": 0.257,
+      "step": 5157
+    },
+    {
+      "epoch": 2.4644883818170955,
+      "grad_norm": 0.45508857688246773,
+      "learning_rate": 1.695742478990211e-06,
+      "loss": 0.2625,
+      "step": 5158
+    },
+    {
+      "epoch": 2.464966250522669,
+      "grad_norm": 0.44398221371433727,
+      "learning_rate": 1.6928093922050913e-06,
+      "loss": 0.259,
+      "step": 5159
+    },
+    {
+      "epoch": 2.465444119228242,
+      "grad_norm": 0.46877325097045486,
+      "learning_rate": 1.6898786096881104e-06,
+      "loss": 0.2515,
+      "step": 5160
+    },
+    {
+      "epoch": 2.465921987933815,
+      "grad_norm": 0.47517976840848003,
+      "learning_rate": 1.686950132252213e-06,
+      "loss": 0.2668,
+      "step": 5161
+    },
+    {
+      "epoch": 2.4663998566393883,
+      "grad_norm": 0.47013178542557804,
+      "learning_rate": 1.6840239607097109e-06,
+      "loss": 0.273,
+      "step": 5162
+    },
+    {
+      "epoch": 2.4668777253449616,
+      "grad_norm": 0.6731321265086605,
+      "learning_rate": 1.6811000958722713e-06,
+      "loss": 0.2653,
+      "step": 5163
+    },
+    {
+      "epoch": 2.4673555940505345,
+      "grad_norm": 0.500612444710717,
+      "learning_rate": 1.6781785385509197e-06,
+      "loss": 0.2553,
+      "step": 5164
+    },
+    {
+      "epoch": 2.467833462756108,
+      "grad_norm": 0.44968808398968174,
+      "learning_rate": 1.6752592895560493e-06,
+      "loss": 0.2651,
+      "step": 5165
+    },
+    {
+      "epoch": 2.468311331461681,
+      "grad_norm": 0.4625732004203681,
+      "learning_rate": 1.6723423496974057e-06,
+      "loss": 0.2808,
+      "step": 5166
+    },
+    {
+      "epoch": 2.468789200167254,
+      "grad_norm": 0.4489275940956645,
+      "learning_rate": 1.6694277197840947e-06,
+      "loss": 0.2628,
+      "step": 5167
+    },
+    {
+      "epoch": 2.4692670688728273,
+      "grad_norm": 0.535652789404287,
+      "learning_rate": 1.6665154006245888e-06,
+      "loss": 0.272,
+      "step": 5168
+    },
+    {
+      "epoch": 2.4697449375784,
+      "grad_norm": 0.47206952227811455,
+      "learning_rate": 1.6636053930267093e-06,
+      "loss": 0.2618,
+      "step": 5169
+    },
+    {
+      "epoch": 2.4702228062839735,
+      "grad_norm": 0.4583292667257452,
+      "learning_rate": 1.6606976977976408e-06,
+      "loss": 0.2575,
+      "step": 5170
+    },
+    {
+      "epoch": 2.470700674989547,
+      "grad_norm": 0.4784423605188315,
+      "learning_rate": 1.6577923157439302e-06,
+      "loss": 0.2642,
+      "step": 5171
+    },
+    {
+      "epoch": 2.4711785436951197,
+      "grad_norm": 0.4666671362911929,
+      "learning_rate": 1.6548892476714772e-06,
+      "loss": 0.2663,
+      "step": 5172
+    },
+    {
+      "epoch": 2.471656412400693,
+      "grad_norm": 0.4821043795677321,
+      "learning_rate": 1.6519884943855391e-06,
+      "loss": 0.2718,
+      "step": 5173
+    },
+    {
+      "epoch": 2.472134281106266,
+      "grad_norm": 0.43488458866311225,
+      "learning_rate": 1.6490900566907396e-06,
+      "loss": 0.2441,
+      "step": 5174
+    },
+    {
+      "epoch": 2.472612149811839,
+      "grad_norm": 0.4670451983441338,
+      "learning_rate": 1.6461939353910494e-06,
+      "loss": 0.2845,
+      "step": 5175
+    },
+    {
+      "epoch": 2.4730900185174125,
+      "grad_norm": 0.45536226914155875,
+      "learning_rate": 1.643300131289801e-06,
+      "loss": 0.2494,
+      "step": 5176
+    },
+    {
+      "epoch": 2.4735678872229854,
+      "grad_norm": 0.4517133543283046,
+      "learning_rate": 1.6404086451896896e-06,
+      "loss": 0.2609,
+      "step": 5177
+    },
+    {
+      "epoch": 2.4740457559285587,
+      "grad_norm": 0.453282470130923,
+      "learning_rate": 1.6375194778927594e-06,
+      "loss": 0.274,
+      "step": 5178
+    },
+    {
+      "epoch": 2.4745236246341316,
+      "grad_norm": 0.4758223874514832,
+      "learning_rate": 1.6346326302004111e-06,
+      "loss": 0.2682,
+      "step": 5179
+    },
+    {
+      "epoch": 2.475001493339705,
+      "grad_norm": 0.4522472979675584,
+      "learning_rate": 1.631748102913412e-06,
+      "loss": 0.2838,
+      "step": 5180
+    },
+    {
+      "epoch": 2.475479362045278,
+      "grad_norm": 0.4389102055107512,
+      "learning_rate": 1.6288658968318748e-06,
+      "loss": 0.2756,
+      "step": 5181
+    },
+    {
+      "epoch": 2.475957230750851,
+      "grad_norm": 0.4561620443865254,
+      "learning_rate": 1.6259860127552718e-06,
+      "loss": 0.2704,
+      "step": 5182
+    },
+    {
+      "epoch": 2.4764350994564244,
+      "grad_norm": 0.46070242911962683,
+      "learning_rate": 1.6231084514824357e-06,
+      "loss": 0.248,
+      "step": 5183
+    },
+    {
+      "epoch": 2.4769129681619972,
+      "grad_norm": 0.4500952574278607,
+      "learning_rate": 1.6202332138115495e-06,
+      "loss": 0.2714,
+      "step": 5184
+    },
+    {
+      "epoch": 2.4773908368675706,
+      "grad_norm": 0.44423358778364996,
+      "learning_rate": 1.6173603005401505e-06,
+      "loss": 0.2694,
+      "step": 5185
+    },
+    {
+      "epoch": 2.477868705573144,
+      "grad_norm": 0.4463179849081536,
+      "learning_rate": 1.614489712465137e-06,
+      "loss": 0.27,
+      "step": 5186
+    },
+    {
+      "epoch": 2.4783465742787167,
+      "grad_norm": 0.4983501684463973,
+      "learning_rate": 1.6116214503827632e-06,
+      "loss": 0.2621,
+      "step": 5187
+    },
+    {
+      "epoch": 2.47882444298429,
+      "grad_norm": 0.443262010521369,
+      "learning_rate": 1.6087555150886291e-06,
+      "loss": 0.2722,
+      "step": 5188
+    },
+    {
+      "epoch": 2.4793023116898634,
+      "grad_norm": 0.4558884591228606,
+      "learning_rate": 1.6058919073776936e-06,
+      "loss": 0.2629,
+      "step": 5189
+    },
+    {
+      "epoch": 2.4797801803954362,
+      "grad_norm": 0.4495529234567257,
+      "learning_rate": 1.6030306280442764e-06,
+      "loss": 0.2516,
+      "step": 5190
+    },
+    {
+      "epoch": 2.4802580491010096,
+      "grad_norm": 0.4572305669305225,
+      "learning_rate": 1.6001716778820432e-06,
+      "loss": 0.2662,
+      "step": 5191
+    },
+    {
+      "epoch": 2.480735917806583,
+      "grad_norm": 0.4459974662375841,
+      "learning_rate": 1.5973150576840134e-06,
+      "loss": 0.2657,
+      "step": 5192
+    },
+    {
+      "epoch": 2.4812137865121557,
+      "grad_norm": 0.5110956161774068,
+      "learning_rate": 1.5944607682425684e-06,
+      "loss": 0.2771,
+      "step": 5193
+    },
+    {
+      "epoch": 2.481691655217729,
+      "grad_norm": 0.5018127615583312,
+      "learning_rate": 1.5916088103494353e-06,
+      "loss": 0.2706,
+      "step": 5194
+    },
+    {
+      "epoch": 2.482169523923302,
+      "grad_norm": 0.46765637570821916,
+      "learning_rate": 1.588759184795694e-06,
+      "loss": 0.2681,
+      "step": 5195
+    },
+    {
+      "epoch": 2.4826473926288752,
+      "grad_norm": 0.4455104728164743,
+      "learning_rate": 1.5859118923717853e-06,
+      "loss": 0.2543,
+      "step": 5196
+    },
+    {
+      "epoch": 2.4831252613344486,
+      "grad_norm": 0.4943599831926817,
+      "learning_rate": 1.5830669338674953e-06,
+      "loss": 0.2632,
+      "step": 5197
+    },
+    {
+      "epoch": 2.4836031300400214,
+      "grad_norm": 0.45460304300287496,
+      "learning_rate": 1.580224310071964e-06,
+      "loss": 0.24,
+      "step": 5198
+    },
+    {
+      "epoch": 2.4840809987455947,
+      "grad_norm": 0.4470853999975232,
+      "learning_rate": 1.577384021773689e-06,
+      "loss": 0.2559,
+      "step": 5199
+    },
+    {
+      "epoch": 2.4845588674511676,
+      "grad_norm": 0.4435833286013539,
+      "learning_rate": 1.574546069760514e-06,
+      "loss": 0.2511,
+      "step": 5200
+    },
+    {
+      "epoch": 2.485036736156741,
+      "grad_norm": 0.4342162740665016,
+      "learning_rate": 1.571710454819635e-06,
+      "loss": 0.2632,
+      "step": 5201
+    },
+    {
+      "epoch": 2.4855146048623142,
+      "grad_norm": 0.47609127386587247,
+      "learning_rate": 1.5688771777376044e-06,
+      "loss": 0.2887,
+      "step": 5202
+    },
+    {
+      "epoch": 2.485992473567887,
+      "grad_norm": 0.4763885822985041,
+      "learning_rate": 1.5660462393003228e-06,
+      "loss": 0.283,
+      "step": 5203
+    },
+    {
+      "epoch": 2.4864703422734604,
+      "grad_norm": 0.45641941404129543,
+      "learning_rate": 1.56321764029304e-06,
+      "loss": 0.2764,
+      "step": 5204
+    },
+    {
+      "epoch": 2.4869482109790333,
+      "grad_norm": 0.45134971688590414,
+      "learning_rate": 1.5603913815003634e-06,
+      "loss": 0.2781,
+      "step": 5205
+    },
+    {
+      "epoch": 2.4874260796846066,
+      "grad_norm": 0.48230919161187935,
+      "learning_rate": 1.5575674637062465e-06,
+      "loss": 0.2715,
+      "step": 5206
+    },
+    {
+      "epoch": 2.48790394839018,
+      "grad_norm": 0.4498323903981208,
+      "learning_rate": 1.5547458876939902e-06,
+      "loss": 0.2675,
+      "step": 5207
+    },
+    {
+      "epoch": 2.488381817095753,
+      "grad_norm": 0.5874324453501406,
+      "learning_rate": 1.5519266542462552e-06,
+      "loss": 0.2554,
+      "step": 5208
+    },
+    {
+      "epoch": 2.488859685801326,
+      "grad_norm": 0.4541224538296049,
+      "learning_rate": 1.5491097641450448e-06,
+      "loss": 0.2506,
+      "step": 5209
+    },
+    {
+      "epoch": 2.4893375545068994,
+      "grad_norm": 0.4646704460437421,
+      "learning_rate": 1.5462952181717117e-06,
+      "loss": 0.267,
+      "step": 5210
+    },
+    {
+      "epoch": 2.4898154232124723,
+      "grad_norm": 0.462874473285813,
+      "learning_rate": 1.543483017106967e-06,
+      "loss": 0.2682,
+      "step": 5211
+    },
+    {
+      "epoch": 2.4902932919180456,
+      "grad_norm": 0.4602131270975226,
+      "learning_rate": 1.5406731617308635e-06,
+      "loss": 0.2705,
+      "step": 5212
+    },
+    {
+      "epoch": 2.4907711606236185,
+      "grad_norm": 0.48190334430500326,
+      "learning_rate": 1.5378656528228032e-06,
+      "loss": 0.2515,
+      "step": 5213
+    },
+    {
+      "epoch": 2.491249029329192,
+      "grad_norm": 0.4591216279627935,
+      "learning_rate": 1.535060491161542e-06,
+      "loss": 0.2675,
+      "step": 5214
+    },
+    {
+      "epoch": 2.491726898034765,
+      "grad_norm": 0.4619766580382836,
+      "learning_rate": 1.532257677525183e-06,
+      "loss": 0.265,
+      "step": 5215
+    },
+    {
+      "epoch": 2.492204766740338,
+      "grad_norm": 0.46916882001283083,
+      "learning_rate": 1.5294572126911723e-06,
+      "loss": 0.2526,
+      "step": 5216
+    },
+    {
+      "epoch": 2.4926826354459113,
+      "grad_norm": 0.4622413318042114,
+      "learning_rate": 1.526659097436316e-06,
+      "loss": 0.2688,
+      "step": 5217
+    },
+    {
+      "epoch": 2.4931605041514846,
+      "grad_norm": 0.4593352758045286,
+      "learning_rate": 1.5238633325367592e-06,
+      "loss": 0.2721,
+      "step": 5218
+    },
+    {
+      "epoch": 2.4936383728570575,
+      "grad_norm": 0.4453905535036544,
+      "learning_rate": 1.5210699187679945e-06,
+      "loss": 0.2707,
+      "step": 5219
+    },
+    {
+      "epoch": 2.494116241562631,
+      "grad_norm": 0.4360154512554449,
+      "learning_rate": 1.5182788569048689e-06,
+      "loss": 0.2547,
+      "step": 5220
+    },
+    {
+      "epoch": 2.4945941102682037,
+      "grad_norm": 0.4673303368639725,
+      "learning_rate": 1.5154901477215756e-06,
+      "loss": 0.2697,
+      "step": 5221
+    },
+    {
+      "epoch": 2.495071978973777,
+      "grad_norm": 0.9738069702656994,
+      "learning_rate": 1.512703791991651e-06,
+      "loss": 0.2795,
+      "step": 5222
+    },
+    {
+      "epoch": 2.4955498476793503,
+      "grad_norm": 0.47649155832990264,
+      "learning_rate": 1.5099197904879792e-06,
+      "loss": 0.2777,
+      "step": 5223
+    },
+    {
+      "epoch": 2.496027716384923,
+      "grad_norm": 0.454990764062367,
+      "learning_rate": 1.507138143982797e-06,
+      "loss": 0.252,
+      "step": 5224
+    },
+    {
+      "epoch": 2.4965055850904965,
+      "grad_norm": 0.4352195318005461,
+      "learning_rate": 1.5043588532476827e-06,
+      "loss": 0.2558,
+      "step": 5225
+    },
+    {
+      "epoch": 2.4969834537960693,
+      "grad_norm": 0.4444748687649701,
+      "learning_rate": 1.5015819190535586e-06,
+      "loss": 0.2738,
+      "step": 5226
+    },
+    {
+      "epoch": 2.4974613225016427,
+      "grad_norm": 0.45483690731958265,
+      "learning_rate": 1.498807342170704e-06,
+      "loss": 0.2872,
+      "step": 5227
+    },
+    {
+      "epoch": 2.497939191207216,
+      "grad_norm": 0.4561626305870393,
+      "learning_rate": 1.4960351233687342e-06,
+      "loss": 0.2676,
+      "step": 5228
+    },
+    {
+      "epoch": 2.498417059912789,
+      "grad_norm": 0.5047010790116747,
+      "learning_rate": 1.493265263416611e-06,
+      "loss": 0.2802,
+      "step": 5229
+    },
+    {
+      "epoch": 2.498894928618362,
+      "grad_norm": 0.43638203904528666,
+      "learning_rate": 1.49049776308265e-06,
+      "loss": 0.264,
+      "step": 5230
+    },
+    {
+      "epoch": 2.499372797323935,
+      "grad_norm": 0.4343364990309924,
+      "learning_rate": 1.4877326231345046e-06,
+      "loss": 0.2681,
+      "step": 5231
+    },
+    {
+      "epoch": 2.4998506660295083,
+      "grad_norm": 0.8690286716299779,
+      "learning_rate": 1.4849698443391724e-06,
+      "loss": 0.2644,
+      "step": 5232
+    },
+    {
+      "epoch": 2.5003285347350817,
+      "grad_norm": 0.47147128875554356,
+      "learning_rate": 1.4822094274630062e-06,
+      "loss": 0.2457,
+      "step": 5233
+    },
+    {
+      "epoch": 2.5008064034406545,
+      "grad_norm": 0.4673984226297333,
+      "learning_rate": 1.479451373271693e-06,
+      "loss": 0.2765,
+      "step": 5234
+    },
+    {
+      "epoch": 2.501284272146228,
+      "grad_norm": 0.4542815587395272,
+      "learning_rate": 1.476695682530268e-06,
+      "loss": 0.2895,
+      "step": 5235
+    },
+    {
+      "epoch": 2.5017621408518007,
+      "grad_norm": 0.4454228012735007,
+      "learning_rate": 1.473942356003113e-06,
+      "loss": 0.2511,
+      "step": 5236
+    },
+    {
+      "epoch": 2.502240009557374,
+      "grad_norm": 0.47387095366988674,
+      "learning_rate": 1.4711913944539524e-06,
+      "loss": 0.2591,
+      "step": 5237
+    },
+    {
+      "epoch": 2.5027178782629473,
+      "grad_norm": 0.44223677795487265,
+      "learning_rate": 1.4684427986458506e-06,
+      "loss": 0.2741,
+      "step": 5238
+    },
+    {
+      "epoch": 2.5031957469685207,
+      "grad_norm": 0.4744918967505802,
+      "learning_rate": 1.465696569341224e-06,
+      "loss": 0.2664,
+      "step": 5239
+    },
+    {
+      "epoch": 2.5036736156740935,
+      "grad_norm": 0.4452348551981471,
+      "learning_rate": 1.4629527073018267e-06,
+      "loss": 0.2449,
+      "step": 5240
+    },
+    {
+      "epoch": 2.504151484379667,
+      "grad_norm": 0.449269605401987,
+      "learning_rate": 1.4602112132887558e-06,
+      "loss": 0.2602,
+      "step": 5241
+    },
+    {
+      "epoch": 2.5046293530852397,
+      "grad_norm": 0.4633053489713711,
+      "learning_rate": 1.457472088062457e-06,
+      "loss": 0.2636,
+      "step": 5242
+    },
+    {
+      "epoch": 2.505107221790813,
+      "grad_norm": 0.4873950617138543,
+      "learning_rate": 1.4547353323827141e-06,
+      "loss": 0.2725,
+      "step": 5243
+    },
+    {
+      "epoch": 2.5055850904963863,
+      "grad_norm": 0.47479818038514116,
+      "learning_rate": 1.4520009470086505e-06,
+      "loss": 0.2502,
+      "step": 5244
+    },
+    {
+      "epoch": 2.506062959201959,
+      "grad_norm": 0.4685977267735002,
+      "learning_rate": 1.449268932698743e-06,
+      "loss": 0.2608,
+      "step": 5245
+    },
+    {
+      "epoch": 2.5065408279075325,
+      "grad_norm": 0.4536044909575744,
+      "learning_rate": 1.4465392902108011e-06,
+      "loss": 0.2439,
+      "step": 5246
+    },
+    {
+      "epoch": 2.5070186966131054,
+      "grad_norm": 0.4361436758103319,
+      "learning_rate": 1.4438120203019779e-06,
+      "loss": 0.26,
+      "step": 5247
+    },
+    {
+      "epoch": 2.5074965653186787,
+      "grad_norm": 0.47056425580013644,
+      "learning_rate": 1.4410871237287738e-06,
+      "loss": 0.2575,
+      "step": 5248
+    },
+    {
+      "epoch": 2.507974434024252,
+      "grad_norm": 0.48001156924264604,
+      "learning_rate": 1.4383646012470254e-06,
+      "loss": 0.2782,
+      "step": 5249
+    },
+    {
+      "epoch": 2.508452302729825,
+      "grad_norm": 0.4657177084103505,
+      "learning_rate": 1.4356444536119085e-06,
+      "loss": 0.2592,
+      "step": 5250
+    },
+    {
+      "epoch": 2.508930171435398,
+      "grad_norm": 0.4329970043724182,
+      "learning_rate": 1.4329266815779507e-06,
+      "loss": 0.2594,
+      "step": 5251
+    },
+    {
+      "epoch": 2.509408040140971,
+      "grad_norm": 0.4624450792688428,
+      "learning_rate": 1.4302112858990103e-06,
+      "loss": 0.2633,
+      "step": 5252
+    },
+    {
+      "epoch": 2.5098859088465444,
+      "grad_norm": 0.45995574276765083,
+      "learning_rate": 1.4274982673282867e-06,
+      "loss": 0.2872,
+      "step": 5253
+    },
+    {
+      "epoch": 2.5103637775521177,
+      "grad_norm": 0.5483488802392914,
+      "learning_rate": 1.4247876266183314e-06,
+      "loss": 0.2601,
+      "step": 5254
+    },
+    {
+      "epoch": 2.5108416462576906,
+      "grad_norm": 0.5288243133437358,
+      "learning_rate": 1.422079364521024e-06,
+      "loss": 0.2625,
+      "step": 5255
+    },
+    {
+      "epoch": 2.511319514963264,
+      "grad_norm": 0.44122821307838805,
+      "learning_rate": 1.419373481787587e-06,
+      "loss": 0.2498,
+      "step": 5256
+    },
+    {
+      "epoch": 2.5117973836688368,
+      "grad_norm": 0.4805211349693034,
+      "learning_rate": 1.41666997916859e-06,
+      "loss": 0.2691,
+      "step": 5257
+    },
+    {
+      "epoch": 2.51227525237441,
+      "grad_norm": 0.4828655990618,
+      "learning_rate": 1.413968857413932e-06,
+      "loss": 0.2589,
+      "step": 5258
+    },
+    {
+      "epoch": 2.5127531210799834,
+      "grad_norm": 0.45262455510116995,
+      "learning_rate": 1.411270117272856e-06,
+      "loss": 0.268,
+      "step": 5259
+    },
+    {
+      "epoch": 2.5132309897855563,
+      "grad_norm": 0.4517397475555426,
+      "learning_rate": 1.4085737594939497e-06,
+      "loss": 0.2474,
+      "step": 5260
+    },
+    {
+      "epoch": 2.5137088584911296,
+      "grad_norm": 0.4581435924810488,
+      "learning_rate": 1.4058797848251315e-06,
+      "loss": 0.2687,
+      "step": 5261
+    },
+    {
+      "epoch": 2.5141867271967024,
+      "grad_norm": 0.48238520728771184,
+      "learning_rate": 1.4031881940136615e-06,
+      "loss": 0.275,
+      "step": 5262
+    },
+    {
+      "epoch": 2.5146645959022758,
+      "grad_norm": 0.4780046524911139,
+      "learning_rate": 1.4004989878061437e-06,
+      "loss": 0.2711,
+      "step": 5263
+    },
+    {
+      "epoch": 2.515142464607849,
+      "grad_norm": 0.45615713033317606,
+      "learning_rate": 1.3978121669485135e-06,
+      "loss": 0.279,
+      "step": 5264
+    },
+    {
+      "epoch": 2.5156203333134224,
+      "grad_norm": 0.48685067829542655,
+      "learning_rate": 1.3951277321860468e-06,
+      "loss": 0.2517,
+      "step": 5265
+    },
+    {
+      "epoch": 2.5160982020189953,
+      "grad_norm": 0.44894317063915157,
+      "learning_rate": 1.3924456842633615e-06,
+      "loss": 0.267,
+      "step": 5266
+    },
+    {
+      "epoch": 2.5165760707245686,
+      "grad_norm": 0.8685504993927438,
+      "learning_rate": 1.3897660239244093e-06,
+      "loss": 0.266,
+      "step": 5267
+    },
+    {
+      "epoch": 2.5170539394301414,
+      "grad_norm": 0.4609315916656919,
+      "learning_rate": 1.3870887519124777e-06,
+      "loss": 0.2784,
+      "step": 5268
+    },
+    {
+      "epoch": 2.5175318081357148,
+      "grad_norm": 0.4691311012646939,
+      "learning_rate": 1.384413868970199e-06,
+      "loss": 0.2626,
+      "step": 5269
+    },
+    {
+      "epoch": 2.518009676841288,
+      "grad_norm": 0.47785266314222014,
+      "learning_rate": 1.381741375839537e-06,
+      "loss": 0.2656,
+      "step": 5270
+    },
+    {
+      "epoch": 2.518487545546861,
+      "grad_norm": 0.45350612752353825,
+      "learning_rate": 1.3790712732617918e-06,
+      "loss": 0.2717,
+      "step": 5271
+    },
+    {
+      "epoch": 2.5189654142524343,
+      "grad_norm": 0.4465246048292593,
+      "learning_rate": 1.3764035619776062e-06,
+      "loss": 0.2549,
+      "step": 5272
+    },
+    {
+      "epoch": 2.519443282958007,
+      "grad_norm": 0.4413523350209254,
+      "learning_rate": 1.3737382427269551e-06,
+      "loss": 0.2622,
+      "step": 5273
+    },
+    {
+      "epoch": 2.5199211516635804,
+      "grad_norm": 0.4402416126165139,
+      "learning_rate": 1.3710753162491498e-06,
+      "loss": 0.2425,
+      "step": 5274
+    },
+    {
+      "epoch": 2.5203990203691538,
+      "grad_norm": 0.4986151441542075,
+      "learning_rate": 1.3684147832828409e-06,
+      "loss": 0.2519,
+      "step": 5275
+    },
+    {
+      "epoch": 2.5208768890747266,
+      "grad_norm": 0.5457742238724771,
+      "learning_rate": 1.365756644566013e-06,
+      "loss": 0.2611,
+      "step": 5276
+    },
+    {
+      "epoch": 2.5213547577803,
+      "grad_norm": 0.4435897317652216,
+      "learning_rate": 1.3631009008359874e-06,
+      "loss": 0.2667,
+      "step": 5277
+    },
+    {
+      "epoch": 2.521832626485873,
+      "grad_norm": 0.45681886545784145,
+      "learning_rate": 1.360447552829417e-06,
+      "loss": 0.2539,
+      "step": 5278
+    },
+    {
+      "epoch": 2.522310495191446,
+      "grad_norm": 0.45114564080703556,
+      "learning_rate": 1.3577966012822974e-06,
+      "loss": 0.2526,
+      "step": 5279
+    },
+    {
+      "epoch": 2.5227883638970194,
+      "grad_norm": 0.5748577158549483,
+      "learning_rate": 1.355148046929956e-06,
+      "loss": 0.2846,
+      "step": 5280
+    },
+    {
+      "epoch": 2.5232662326025923,
+      "grad_norm": 0.5478505157579419,
+      "learning_rate": 1.352501890507051e-06,
+      "loss": 0.2546,
+      "step": 5281
+    },
+    {
+      "epoch": 2.5237441013081656,
+      "grad_norm": 1.090704314845147,
+      "learning_rate": 1.3498581327475847e-06,
+      "loss": 0.2617,
+      "step": 5282
+    },
+    {
+      "epoch": 2.5242219700137385,
+      "grad_norm": 0.44651134484658483,
+      "learning_rate": 1.3472167743848863e-06,
+      "loss": 0.2797,
+      "step": 5283
+    },
+    {
+      "epoch": 2.524699838719312,
+      "grad_norm": 0.4332379133919312,
+      "learning_rate": 1.344577816151621e-06,
+      "loss": 0.2693,
+      "step": 5284
+    },
+    {
+      "epoch": 2.525177707424885,
+      "grad_norm": 0.4454197457889718,
+      "learning_rate": 1.3419412587797908e-06,
+      "loss": 0.2913,
+      "step": 5285
+    },
+    {
+      "epoch": 2.525655576130458,
+      "grad_norm": 0.5539819066751364,
+      "learning_rate": 1.3393071030007298e-06,
+      "loss": 0.2734,
+      "step": 5286
+    },
+    {
+      "epoch": 2.5261334448360313,
+      "grad_norm": 0.4907491307225207,
+      "learning_rate": 1.3366753495451046e-06,
+      "loss": 0.2559,
+      "step": 5287
+    },
+    {
+      "epoch": 2.526611313541604,
+      "grad_norm": 0.44486691210900325,
+      "learning_rate": 1.3340459991429221e-06,
+      "loss": 0.2717,
+      "step": 5288
+    },
+    {
+      "epoch": 2.5270891822471775,
+      "grad_norm": 0.524341426871989,
+      "learning_rate": 1.3314190525235148e-06,
+      "loss": 0.2785,
+      "step": 5289
+    },
+    {
+      "epoch": 2.527567050952751,
+      "grad_norm": 0.4507961520371896,
+      "learning_rate": 1.3287945104155487e-06,
+      "loss": 0.2692,
+      "step": 5290
+    },
+    {
+      "epoch": 2.528044919658324,
+      "grad_norm": 0.4931269433170495,
+      "learning_rate": 1.32617237354703e-06,
+      "loss": 0.2577,
+      "step": 5291
+    },
+    {
+      "epoch": 2.528522788363897,
+      "grad_norm": 0.44578474408366947,
+      "learning_rate": 1.3235526426452916e-06,
+      "loss": 0.2853,
+      "step": 5292
+    },
+    {
+      "epoch": 2.5290006570694703,
+      "grad_norm": 0.4549319341008648,
+      "learning_rate": 1.3209353184369978e-06,
+      "loss": 0.2599,
+      "step": 5293
+    },
+    {
+      "epoch": 2.529478525775043,
+      "grad_norm": 0.44026422411476585,
+      "learning_rate": 1.318320401648152e-06,
+      "loss": 0.274,
+      "step": 5294
+    },
+    {
+      "epoch": 2.5299563944806165,
+      "grad_norm": 0.44736941554716725,
+      "learning_rate": 1.3157078930040856e-06,
+      "loss": 0.2668,
+      "step": 5295
+    },
+    {
+      "epoch": 2.53043426318619,
+      "grad_norm": 0.4735831103712228,
+      "learning_rate": 1.3130977932294597e-06,
+      "loss": 0.2688,
+      "step": 5296
+    },
+    {
+      "epoch": 2.5309121318917627,
+      "grad_norm": 0.4685212925748311,
+      "learning_rate": 1.310490103048273e-06,
+      "loss": 0.2482,
+      "step": 5297
+    },
+    {
+      "epoch": 2.531390000597336,
+      "grad_norm": 0.4593947890008595,
+      "learning_rate": 1.3078848231838514e-06,
+      "loss": 0.2581,
+      "step": 5298
+    },
+    {
+      "epoch": 2.531867869302909,
+      "grad_norm": 0.4311892224206543,
+      "learning_rate": 1.3052819543588512e-06,
+      "loss": 0.2465,
+      "step": 5299
+    },
+    {
+      "epoch": 2.532345738008482,
+      "grad_norm": 0.46314465945990146,
+      "learning_rate": 1.3026814972952674e-06,
+      "loss": 0.2684,
+      "step": 5300
+    },
+    {
+      "epoch": 2.5328236067140555,
+      "grad_norm": 0.46309469275162923,
+      "learning_rate": 1.300083452714418e-06,
+      "loss": 0.2615,
+      "step": 5301
+    },
+    {
+      "epoch": 2.5333014754196284,
+      "grad_norm": 0.4437569382154475,
+      "learning_rate": 1.2974878213369523e-06,
+      "loss": 0.2639,
+      "step": 5302
+    },
+    {
+      "epoch": 2.5337793441252017,
+      "grad_norm": 0.5664502089285215,
+      "learning_rate": 1.294894603882858e-06,
+      "loss": 0.2615,
+      "step": 5303
+    },
+    {
+      "epoch": 2.5342572128307745,
+      "grad_norm": 0.4592275859545169,
+      "learning_rate": 1.2923038010714451e-06,
+      "loss": 0.2601,
+      "step": 5304
+    },
+    {
+      "epoch": 2.534735081536348,
+      "grad_norm": 0.4309337745483858,
+      "learning_rate": 1.2897154136213542e-06,
+      "loss": 0.2823,
+      "step": 5305
+    },
+    {
+      "epoch": 2.535212950241921,
+      "grad_norm": 0.48862862670470414,
+      "learning_rate": 1.287129442250562e-06,
+      "loss": 0.2745,
+      "step": 5306
+    },
+    {
+      "epoch": 2.535690818947494,
+      "grad_norm": 0.83841385727078,
+      "learning_rate": 1.2845458876763718e-06,
+      "loss": 0.2782,
+      "step": 5307
+    },
+    {
+      "epoch": 2.5361686876530674,
+      "grad_norm": 0.43581783536876467,
+      "learning_rate": 1.281964750615412e-06,
+      "loss": 0.2651,
+      "step": 5308
+    },
+    {
+      "epoch": 2.5366465563586402,
+      "grad_norm": 0.4411137035318371,
+      "learning_rate": 1.2793860317836482e-06,
+      "loss": 0.2884,
+      "step": 5309
+    },
+    {
+      "epoch": 2.5371244250642135,
+      "grad_norm": 0.46287211410540297,
+      "learning_rate": 1.2768097318963701e-06,
+      "loss": 0.2376,
+      "step": 5310
+    },
+    {
+      "epoch": 2.537602293769787,
+      "grad_norm": 0.44652943411212714,
+      "learning_rate": 1.2742358516681963e-06,
+      "loss": 0.2763,
+      "step": 5311
+    },
+    {
+      "epoch": 2.53808016247536,
+      "grad_norm": 0.45278372197354316,
+      "learning_rate": 1.27166439181308e-06,
+      "loss": 0.2682,
+      "step": 5312
+    },
+    {
+      "epoch": 2.538558031180933,
+      "grad_norm": 0.44885245556664793,
+      "learning_rate": 1.2690953530442963e-06,
+      "loss": 0.2417,
+      "step": 5313
+    },
+    {
+      "epoch": 2.539035899886506,
+      "grad_norm": 0.43904932734636665,
+      "learning_rate": 1.2665287360744482e-06,
+      "loss": 0.2647,
+      "step": 5314
+    },
+    {
+      "epoch": 2.5395137685920792,
+      "grad_norm": 0.42657730106507596,
+      "learning_rate": 1.2639645416154744e-06,
+      "loss": 0.2546,
+      "step": 5315
+    },
+    {
+      "epoch": 2.5399916372976525,
+      "grad_norm": 0.5240326634712088,
+      "learning_rate": 1.2614027703786369e-06,
+      "loss": 0.249,
+      "step": 5316
+    },
+    {
+      "epoch": 2.540469506003226,
+      "grad_norm": 0.44736006043482424,
+      "learning_rate": 1.2588434230745228e-06,
+      "loss": 0.2483,
+      "step": 5317
+    },
+    {
+      "epoch": 2.5409473747087987,
+      "grad_norm": 0.4431785975651459,
+      "learning_rate": 1.2562865004130532e-06,
+      "loss": 0.2623,
+      "step": 5318
+    },
+    {
+      "epoch": 2.541425243414372,
+      "grad_norm": 0.4599390657644793,
+      "learning_rate": 1.2537320031034717e-06,
+      "loss": 0.2704,
+      "step": 5319
+    },
+    {
+      "epoch": 2.541903112119945,
+      "grad_norm": 0.4588576873276534,
+      "learning_rate": 1.2511799318543493e-06,
+      "loss": 0.2862,
+      "step": 5320
+    },
+    {
+      "epoch": 2.542380980825518,
+      "grad_norm": 0.4409900007452561,
+      "learning_rate": 1.2486302873735878e-06,
+      "loss": 0.2639,
+      "step": 5321
+    },
+    {
+      "epoch": 2.5428588495310915,
+      "grad_norm": 0.4463947225335007,
+      "learning_rate": 1.2460830703684147e-06,
+      "loss": 0.2563,
+      "step": 5322
+    },
+    {
+      "epoch": 2.5433367182366644,
+      "grad_norm": 0.44554797570869803,
+      "learning_rate": 1.243538281545381e-06,
+      "loss": 0.2618,
+      "step": 5323
+    },
+    {
+      "epoch": 2.5438145869422377,
+      "grad_norm": 0.4366283293412151,
+      "learning_rate": 1.2409959216103651e-06,
+      "loss": 0.2727,
+      "step": 5324
+    },
+    {
+      "epoch": 2.5442924556478106,
+      "grad_norm": 0.6447095546778173,
+      "learning_rate": 1.2384559912685768e-06,
+      "loss": 0.2596,
+      "step": 5325
+    },
+    {
+      "epoch": 2.544770324353384,
+      "grad_norm": 0.46130620935874256,
+      "learning_rate": 1.2359184912245448e-06,
+      "loss": 0.2674,
+      "step": 5326
+    },
+    {
+      "epoch": 2.545248193058957,
+      "grad_norm": 0.45602539579070933,
+      "learning_rate": 1.2333834221821262e-06,
+      "loss": 0.2573,
+      "step": 5327
+    },
+    {
+      "epoch": 2.54572606176453,
+      "grad_norm": 0.4549646150650348,
+      "learning_rate": 1.2308507848445072e-06,
+      "loss": 0.2579,
+      "step": 5328
+    },
+    {
+      "epoch": 2.5462039304701034,
+      "grad_norm": 0.4415752044414283,
+      "learning_rate": 1.228320579914195e-06,
+      "loss": 0.2794,
+      "step": 5329
+    },
+    {
+      "epoch": 2.5466817991756763,
+      "grad_norm": 0.47703375879035975,
+      "learning_rate": 1.2257928080930236e-06,
+      "loss": 0.2729,
+      "step": 5330
+    },
+    {
+      "epoch": 2.5471596678812496,
+      "grad_norm": 0.6782750426427807,
+      "learning_rate": 1.2232674700821535e-06,
+      "loss": 0.2681,
+      "step": 5331
+    },
+    {
+      "epoch": 2.547637536586823,
+      "grad_norm": 0.44620672912194137,
+      "learning_rate": 1.2207445665820695e-06,
+      "loss": 0.2655,
+      "step": 5332
+    },
+    {
+      "epoch": 2.5481154052923958,
+      "grad_norm": 0.49483270413584723,
+      "learning_rate": 1.2182240982925764e-06,
+      "loss": 0.255,
+      "step": 5333
+    },
+    {
+      "epoch": 2.548593273997969,
+      "grad_norm": 0.4760394124784553,
+      "learning_rate": 1.2157060659128128e-06,
+      "loss": 0.261,
+      "step": 5334
+    },
+    {
+      "epoch": 2.549071142703542,
+      "grad_norm": 0.43536619499616924,
+      "learning_rate": 1.2131904701412345e-06,
+      "loss": 0.2699,
+      "step": 5335
+    },
+    {
+      "epoch": 2.5495490114091153,
+      "grad_norm": 0.4567672339365635,
+      "learning_rate": 1.2106773116756198e-06,
+      "loss": 0.2887,
+      "step": 5336
+    },
+    {
+      "epoch": 2.5500268801146886,
+      "grad_norm": 0.44424439879811406,
+      "learning_rate": 1.2081665912130813e-06,
+      "loss": 0.2715,
+      "step": 5337
+    },
+    {
+      "epoch": 2.550504748820262,
+      "grad_norm": 0.47455117509378236,
+      "learning_rate": 1.2056583094500451e-06,
+      "loss": 0.2692,
+      "step": 5338
+    },
+    {
+      "epoch": 2.5509826175258348,
+      "grad_norm": 0.4577068157459303,
+      "learning_rate": 1.2031524670822613e-06,
+      "loss": 0.2527,
+      "step": 5339
+    },
+    {
+      "epoch": 2.551460486231408,
+      "grad_norm": 0.4635419551668722,
+      "learning_rate": 1.2006490648048118e-06,
+      "loss": 0.2616,
+      "step": 5340
+    },
+    {
+      "epoch": 2.551938354936981,
+      "grad_norm": 0.4872478703098568,
+      "learning_rate": 1.1981481033120945e-06,
+      "loss": 0.2779,
+      "step": 5341
+    },
+    {
+      "epoch": 2.5524162236425543,
+      "grad_norm": 0.43900058178403417,
+      "learning_rate": 1.1956495832978289e-06,
+      "loss": 0.2635,
+      "step": 5342
+    },
+    {
+      "epoch": 2.5528940923481276,
+      "grad_norm": 0.48199098433370163,
+      "learning_rate": 1.1931535054550647e-06,
+      "loss": 0.2598,
+      "step": 5343
+    },
+    {
+      "epoch": 2.5533719610537005,
+      "grad_norm": 0.476582985861211,
+      "learning_rate": 1.1906598704761685e-06,
+      "loss": 0.2681,
+      "step": 5344
+    },
+    {
+      "epoch": 2.5538498297592738,
+      "grad_norm": 0.43490218370942263,
+      "learning_rate": 1.1881686790528279e-06,
+      "loss": 0.2522,
+      "step": 5345
+    },
+    {
+      "epoch": 2.5543276984648466,
+      "grad_norm": 0.435352449223695,
+      "learning_rate": 1.1856799318760592e-06,
+      "loss": 0.2582,
+      "step": 5346
+    },
+    {
+      "epoch": 2.55480556717042,
+      "grad_norm": 0.48299189717460383,
+      "learning_rate": 1.1831936296361957e-06,
+      "loss": 0.2582,
+      "step": 5347
+    },
+    {
+      "epoch": 2.5552834358759933,
+      "grad_norm": 0.45700996353152185,
+      "learning_rate": 1.1807097730228912e-06,
+      "loss": 0.2652,
+      "step": 5348
+    },
+    {
+      "epoch": 2.555761304581566,
+      "grad_norm": 0.45128915246269297,
+      "learning_rate": 1.178228362725129e-06,
+      "loss": 0.2566,
+      "step": 5349
+    },
+    {
+      "epoch": 2.5562391732871395,
+      "grad_norm": 0.49437093690101924,
+      "learning_rate": 1.1757493994312052e-06,
+      "loss": 0.2716,
+      "step": 5350
+    },
+    {
+      "epoch": 2.5567170419927123,
+      "grad_norm": 0.5541061982972699,
+      "learning_rate": 1.1732728838287388e-06,
+      "loss": 0.2563,
+      "step": 5351
+    },
+    {
+      "epoch": 2.5571949106982856,
+      "grad_norm": 0.46593343771916595,
+      "learning_rate": 1.1707988166046757e-06,
+      "loss": 0.2553,
+      "step": 5352
+    },
+    {
+      "epoch": 2.557672779403859,
+      "grad_norm": 0.48119739643546555,
+      "learning_rate": 1.168327198445276e-06,
+      "loss": 0.2692,
+      "step": 5353
+    },
+    {
+      "epoch": 2.558150648109432,
+      "grad_norm": 0.44694892042251333,
+      "learning_rate": 1.1658580300361223e-06,
+      "loss": 0.2623,
+      "step": 5354
+    },
+    {
+      "epoch": 2.558628516815005,
+      "grad_norm": 0.7466740830810888,
+      "learning_rate": 1.1633913120621188e-06,
+      "loss": 0.275,
+      "step": 5355
+    },
+    {
+      "epoch": 2.559106385520578,
+      "grad_norm": 0.4510071850128832,
+      "learning_rate": 1.1609270452074917e-06,
+      "loss": 0.2514,
+      "step": 5356
+    },
+    {
+      "epoch": 2.5595842542261513,
+      "grad_norm": 0.4667471409690589,
+      "learning_rate": 1.158465230155784e-06,
+      "loss": 0.2806,
+      "step": 5357
+    },
+    {
+      "epoch": 2.5600621229317246,
+      "grad_norm": 0.4836523364080265,
+      "learning_rate": 1.1560058675898577e-06,
+      "loss": 0.277,
+      "step": 5358
+    },
+    {
+      "epoch": 2.5605399916372975,
+      "grad_norm": 0.4864707078033067,
+      "learning_rate": 1.1535489581919012e-06,
+      "loss": 0.2656,
+      "step": 5359
+    },
+    {
+      "epoch": 2.561017860342871,
+      "grad_norm": 0.43787020641937124,
+      "learning_rate": 1.151094502643414e-06,
+      "loss": 0.2451,
+      "step": 5360
+    },
+    {
+      "epoch": 2.5614957290484437,
+      "grad_norm": 0.456591257908703,
+      "learning_rate": 1.148642501625218e-06,
+      "loss": 0.2525,
+      "step": 5361
+    },
+    {
+      "epoch": 2.561973597754017,
+      "grad_norm": 0.45414715242812,
+      "learning_rate": 1.1461929558174589e-06,
+      "loss": 0.2576,
+      "step": 5362
+    },
+    {
+      "epoch": 2.5624514664595903,
+      "grad_norm": 0.4297472089254829,
+      "learning_rate": 1.1437458658995947e-06,
+      "loss": 0.2565,
+      "step": 5363
+    },
+    {
+      "epoch": 2.5629293351651636,
+      "grad_norm": 0.6242800594293806,
+      "learning_rate": 1.1413012325504048e-06,
+      "loss": 0.252,
+      "step": 5364
+    },
+    {
+      "epoch": 2.5634072038707365,
+      "grad_norm": 0.44685469395106875,
+      "learning_rate": 1.1388590564479895e-06,
+      "loss": 0.2716,
+      "step": 5365
+    },
+    {
+      "epoch": 2.56388507257631,
+      "grad_norm": 0.43936844793345164,
+      "learning_rate": 1.1364193382697642e-06,
+      "loss": 0.2564,
+      "step": 5366
+    },
+    {
+      "epoch": 2.5643629412818827,
+      "grad_norm": 0.4655861680277552,
+      "learning_rate": 1.1339820786924616e-06,
+      "loss": 0.2878,
+      "step": 5367
+    },
+    {
+      "epoch": 2.564840809987456,
+      "grad_norm": 0.46292732997498603,
+      "learning_rate": 1.1315472783921378e-06,
+      "loss": 0.2688,
+      "step": 5368
+    },
+    {
+      "epoch": 2.5653186786930293,
+      "grad_norm": 0.4598254943620399,
+      "learning_rate": 1.1291149380441636e-06,
+      "loss": 0.2673,
+      "step": 5369
+    },
+    {
+      "epoch": 2.565796547398602,
+      "grad_norm": 0.454472063477016,
+      "learning_rate": 1.1266850583232224e-06,
+      "loss": 0.2478,
+      "step": 5370
+    },
+    {
+      "epoch": 2.5662744161041755,
+      "grad_norm": 0.4325026111247085,
+      "learning_rate": 1.1242576399033267e-06,
+      "loss": 0.2562,
+      "step": 5371
+    },
+    {
+      "epoch": 2.5667522848097484,
+      "grad_norm": 0.4453897324065896,
+      "learning_rate": 1.1218326834577953e-06,
+      "loss": 0.2518,
+      "step": 5372
+    },
+    {
+      "epoch": 2.5672301535153217,
+      "grad_norm": 0.4650279923070856,
+      "learning_rate": 1.119410189659268e-06,
+      "loss": 0.2888,
+      "step": 5373
+    },
+    {
+      "epoch": 2.567708022220895,
+      "grad_norm": 0.4475210944515468,
+      "learning_rate": 1.116990159179705e-06,
+      "loss": 0.2607,
+      "step": 5374
+    },
+    {
+      "epoch": 2.568185890926468,
+      "grad_norm": 0.4867286499909107,
+      "learning_rate": 1.1145725926903772e-06,
+      "loss": 0.2759,
+      "step": 5375
+    },
+    {
+      "epoch": 2.568663759632041,
+      "grad_norm": 0.5000116826228374,
+      "learning_rate": 1.112157490861875e-06,
+      "loss": 0.2527,
+      "step": 5376
+    },
+    {
+      "epoch": 2.569141628337614,
+      "grad_norm": 0.4353063417332031,
+      "learning_rate": 1.1097448543641077e-06,
+      "loss": 0.257,
+      "step": 5377
+    },
+    {
+      "epoch": 2.5696194970431874,
+      "grad_norm": 0.5646763109551065,
+      "learning_rate": 1.107334683866297e-06,
+      "loss": 0.2804,
+      "step": 5378
+    },
+    {
+      "epoch": 2.5700973657487607,
+      "grad_norm": 0.4829620257030537,
+      "learning_rate": 1.1049269800369787e-06,
+      "loss": 0.2713,
+      "step": 5379
+    },
+    {
+      "epoch": 2.5705752344543336,
+      "grad_norm": 0.44373746829302196,
+      "learning_rate": 1.1025217435440116e-06,
+      "loss": 0.2708,
+      "step": 5380
+    },
+    {
+      "epoch": 2.571053103159907,
+      "grad_norm": 0.45314693225438013,
+      "learning_rate": 1.1001189750545637e-06,
+      "loss": 0.2661,
+      "step": 5381
+    },
+    {
+      "epoch": 2.5715309718654797,
+      "grad_norm": 0.4396802093044451,
+      "learning_rate": 1.0977186752351187e-06,
+      "loss": 0.2518,
+      "step": 5382
+    },
+    {
+      "epoch": 2.572008840571053,
+      "grad_norm": 0.43552638814311895,
+      "learning_rate": 1.095320844751483e-06,
+      "loss": 0.2547,
+      "step": 5383
+    },
+    {
+      "epoch": 2.5724867092766264,
+      "grad_norm": 0.46000133848749897,
+      "learning_rate": 1.0929254842687676e-06,
+      "loss": 0.2746,
+      "step": 5384
+    },
+    {
+      "epoch": 2.5729645779821992,
+      "grad_norm": 0.514481355649078,
+      "learning_rate": 1.0905325944514034e-06,
+      "loss": 0.2714,
+      "step": 5385
+    },
+    {
+      "epoch": 2.5734424466877726,
+      "grad_norm": 0.44321694655322663,
+      "learning_rate": 1.0881421759631394e-06,
+      "loss": 0.2642,
+      "step": 5386
+    },
+    {
+      "epoch": 2.5739203153933454,
+      "grad_norm": 0.4519160941249428,
+      "learning_rate": 1.085754229467032e-06,
+      "loss": 0.2406,
+      "step": 5387
+    },
+    {
+      "epoch": 2.5743981840989187,
+      "grad_norm": 0.45129680242728143,
+      "learning_rate": 1.0833687556254558e-06,
+      "loss": 0.2617,
+      "step": 5388
+    },
+    {
+      "epoch": 2.574876052804492,
+      "grad_norm": 0.47689305663408194,
+      "learning_rate": 1.0809857551001013e-06,
+      "loss": 0.2532,
+      "step": 5389
+    },
+    {
+      "epoch": 2.5753539215100654,
+      "grad_norm": 0.46337528524382954,
+      "learning_rate": 1.078605228551971e-06,
+      "loss": 0.2718,
+      "step": 5390
+    },
+    {
+      "epoch": 2.5758317902156382,
+      "grad_norm": 0.4470813207728647,
+      "learning_rate": 1.0762271766413768e-06,
+      "loss": 0.243,
+      "step": 5391
+    },
+    {
+      "epoch": 2.5763096589212116,
+      "grad_norm": 0.44289569253985883,
+      "learning_rate": 1.0738516000279542e-06,
+      "loss": 0.2492,
+      "step": 5392
+    },
+    {
+      "epoch": 2.5767875276267844,
+      "grad_norm": 0.4728394742707393,
+      "learning_rate": 1.0714784993706418e-06,
+      "loss": 0.2685,
+      "step": 5393
+    },
+    {
+      "epoch": 2.5772653963323577,
+      "grad_norm": 0.4549229472146639,
+      "learning_rate": 1.0691078753276962e-06,
+      "loss": 0.2512,
+      "step": 5394
+    },
+    {
+      "epoch": 2.577743265037931,
+      "grad_norm": 0.44764612742438686,
+      "learning_rate": 1.0667397285566893e-06,
+      "loss": 0.2549,
+      "step": 5395
+    },
+    {
+      "epoch": 2.578221133743504,
+      "grad_norm": 0.4822527109733988,
+      "learning_rate": 1.0643740597145025e-06,
+      "loss": 0.2631,
+      "step": 5396
+    },
+    {
+      "epoch": 2.5786990024490772,
+      "grad_norm": 0.4572397580448997,
+      "learning_rate": 1.0620108694573272e-06,
+      "loss": 0.2842,
+      "step": 5397
+    },
+    {
+      "epoch": 2.57917687115465,
+      "grad_norm": 0.42473552272997284,
+      "learning_rate": 1.0596501584406749e-06,
+      "loss": 0.2633,
+      "step": 5398
+    },
+    {
+      "epoch": 2.5796547398602234,
+      "grad_norm": 0.4314598541362525,
+      "learning_rate": 1.0572919273193639e-06,
+      "loss": 0.2598,
+      "step": 5399
+    },
+    {
+      "epoch": 2.5801326085657967,
+      "grad_norm": 0.43609204699648757,
+      "learning_rate": 1.0549361767475241e-06,
+      "loss": 0.2537,
+      "step": 5400
+    },
+    {
+      "epoch": 2.5806104772713696,
+      "grad_norm": 0.448934958642449,
+      "learning_rate": 1.052582907378602e-06,
+      "loss": 0.2655,
+      "step": 5401
+    },
+    {
+      "epoch": 2.581088345976943,
+      "grad_norm": 0.44431074903974976,
+      "learning_rate": 1.050232119865352e-06,
+      "loss": 0.266,
+      "step": 5402
+    },
+    {
+      "epoch": 2.581566214682516,
+      "grad_norm": 0.4752805202026781,
+      "learning_rate": 1.047883814859838e-06,
+      "loss": 0.2891,
+      "step": 5403
+    },
+    {
+      "epoch": 2.582044083388089,
+      "grad_norm": 0.4475172447028092,
+      "learning_rate": 1.0455379930134435e-06,
+      "loss": 0.2694,
+      "step": 5404
+    },
+    {
+      "epoch": 2.5825219520936624,
+      "grad_norm": 0.5428975699380926,
+      "learning_rate": 1.0431946549768567e-06,
+      "loss": 0.2692,
+      "step": 5405
+    },
+    {
+      "epoch": 2.5829998207992353,
+      "grad_norm": 0.46361910025632774,
+      "learning_rate": 1.0408538014000747e-06,
+      "loss": 0.2635,
+      "step": 5406
+    },
+    {
+      "epoch": 2.5834776895048086,
+      "grad_norm": 0.4420791110581533,
+      "learning_rate": 1.0385154329324132e-06,
+      "loss": 0.2624,
+      "step": 5407
+    },
+    {
+      "epoch": 2.5839555582103815,
+      "grad_norm": 0.4595527525010734,
+      "learning_rate": 1.0361795502224925e-06,
+      "loss": 0.2608,
+      "step": 5408
+    },
+    {
+      "epoch": 2.584433426915955,
+      "grad_norm": 0.45101202999040235,
+      "learning_rate": 1.0338461539182443e-06,
+      "loss": 0.2618,
+      "step": 5409
+    },
+    {
+      "epoch": 2.584911295621528,
+      "grad_norm": 0.4599073300017213,
+      "learning_rate": 1.0315152446669142e-06,
+      "loss": 0.2703,
+      "step": 5410
+    },
+    {
+      "epoch": 2.585389164327101,
+      "grad_norm": 0.5079598888681588,
+      "learning_rate": 1.0291868231150537e-06,
+      "loss": 0.2681,
+      "step": 5411
+    },
+    {
+      "epoch": 2.5858670330326743,
+      "grad_norm": 0.48499751233901656,
+      "learning_rate": 1.0268608899085241e-06,
+      "loss": 0.2548,
+      "step": 5412
+    },
+    {
+      "epoch": 2.586344901738247,
+      "grad_norm": 0.4422616538627771,
+      "learning_rate": 1.0245374456925029e-06,
+      "loss": 0.2696,
+      "step": 5413
+    },
+    {
+      "epoch": 2.5868227704438205,
+      "grad_norm": 0.4443272586161862,
+      "learning_rate": 1.0222164911114697e-06,
+      "loss": 0.2582,
+      "step": 5414
+    },
+    {
+      "epoch": 2.587300639149394,
+      "grad_norm": 0.44296846825131514,
+      "learning_rate": 1.019898026809214e-06,
+      "loss": 0.2803,
+      "step": 5415
+    },
+    {
+      "epoch": 2.587778507854967,
+      "grad_norm": 0.4517679987221167,
+      "learning_rate": 1.0175820534288416e-06,
+      "loss": 0.2806,
+      "step": 5416
+    },
+    {
+      "epoch": 2.58825637656054,
+      "grad_norm": 0.48611487954550253,
+      "learning_rate": 1.0152685716127598e-06,
+      "loss": 0.2602,
+      "step": 5417
+    },
+    {
+      "epoch": 2.5887342452661133,
+      "grad_norm": 0.4628864103901881,
+      "learning_rate": 1.0129575820026872e-06,
+      "loss": 0.272,
+      "step": 5418
+    },
+    {
+      "epoch": 2.589212113971686,
+      "grad_norm": 0.4649352347539501,
+      "learning_rate": 1.0106490852396544e-06,
+      "loss": 0.2539,
+      "step": 5419
+    },
+    {
+      "epoch": 2.5896899826772595,
+      "grad_norm": 0.43939173718242563,
+      "learning_rate": 1.0083430819639962e-06,
+      "loss": 0.2475,
+      "step": 5420
+    },
+    {
+      "epoch": 2.590167851382833,
+      "grad_norm": 0.44410831432228415,
+      "learning_rate": 1.0060395728153539e-06,
+      "loss": 0.2722,
+      "step": 5421
+    },
+    {
+      "epoch": 2.5906457200884057,
+      "grad_norm": 0.5288794827685146,
+      "learning_rate": 1.0037385584326843e-06,
+      "loss": 0.2566,
+      "step": 5422
+    },
+    {
+      "epoch": 2.591123588793979,
+      "grad_norm": 0.43733087257688036,
+      "learning_rate": 1.0014400394542489e-06,
+      "loss": 0.2572,
+      "step": 5423
+    },
+    {
+      "epoch": 2.591601457499552,
+      "grad_norm": 0.4783741515835818,
+      "learning_rate": 9.991440165176147e-07,
+      "loss": 0.251,
+      "step": 5424
+    },
+    {
+      "epoch": 2.592079326205125,
+      "grad_norm": 0.45766909636052455,
+      "learning_rate": 9.968504902596566e-07,
+      "loss": 0.2809,
+      "step": 5425
+    },
+    {
+      "epoch": 2.5925571949106985,
+      "grad_norm": 0.47258848724591235,
+      "learning_rate": 9.94559461316561e-07,
+      "loss": 0.262,
+      "step": 5426
+    },
+    {
+      "epoch": 2.5930350636162713,
+      "grad_norm": 0.4790702365770836,
+      "learning_rate": 9.922709303238175e-07,
+      "loss": 0.262,
+      "step": 5427
+    },
+    {
+      "epoch": 2.5935129323218447,
+      "grad_norm": 0.676497020285866,
+      "learning_rate": 9.899848979162218e-07,
+      "loss": 0.2711,
+      "step": 5428
+    },
+    {
+      "epoch": 2.5939908010274175,
+      "grad_norm": 0.44747328409752335,
+      "learning_rate": 9.877013647278844e-07,
+      "loss": 0.2647,
+      "step": 5429
+    },
+    {
+      "epoch": 2.594468669732991,
+      "grad_norm": 0.4606135261704647,
+      "learning_rate": 9.85420331392214e-07,
+      "loss": 0.2532,
+      "step": 5430
+    },
+    {
+      "epoch": 2.594946538438564,
+      "grad_norm": 0.43604837240539995,
+      "learning_rate": 9.831417985419278e-07,
+      "loss": 0.2645,
+      "step": 5431
+    },
+    {
+      "epoch": 2.595424407144137,
+      "grad_norm": 0.4373132967875298,
+      "learning_rate": 9.808657668090527e-07,
+      "loss": 0.2579,
+      "step": 5432
+    },
+    {
+      "epoch": 2.5959022758497103,
+      "grad_norm": 0.4573400005148769,
+      "learning_rate": 9.785922368249201e-07,
+      "loss": 0.2649,
+      "step": 5433
+    },
+    {
+      "epoch": 2.596380144555283,
+      "grad_norm": 0.4286328321592479,
+      "learning_rate": 9.763212092201634e-07,
+      "loss": 0.2771,
+      "step": 5434
+    },
+    {
+      "epoch": 2.5968580132608565,
+      "grad_norm": 0.46487186359736915,
+      "learning_rate": 9.74052684624731e-07,
+      "loss": 0.2605,
+      "step": 5435
+    },
+    {
+      "epoch": 2.59733588196643,
+      "grad_norm": 0.45134027488832124,
+      "learning_rate": 9.717866636678685e-07,
+      "loss": 0.2696,
+      "step": 5436
+    },
+    {
+      "epoch": 2.5978137506720027,
+      "grad_norm": 0.4315621650790117,
+      "learning_rate": 9.695231469781285e-07,
+      "loss": 0.2595,
+      "step": 5437
+    },
+    {
+      "epoch": 2.598291619377576,
+      "grad_norm": 0.43086059451199493,
+      "learning_rate": 9.672621351833754e-07,
+      "loss": 0.2563,
+      "step": 5438
+    },
+    {
+      "epoch": 2.598769488083149,
+      "grad_norm": 0.4510589878533883,
+      "learning_rate": 9.6500362891077e-07,
+      "loss": 0.2753,
+      "step": 5439
+    },
+    {
+      "epoch": 2.599247356788722,
+      "grad_norm": 0.47484782844891105,
+      "learning_rate": 9.62747628786782e-07,
+      "loss": 0.2804,
+      "step": 5440
+    },
+    {
+      "epoch": 2.5997252254942955,
+      "grad_norm": 0.44734361221136104,
+      "learning_rate": 9.604941354371899e-07,
+      "loss": 0.2565,
+      "step": 5441
+    },
+    {
+      "epoch": 2.600203094199869,
+      "grad_norm": 0.4523591857593359,
+      "learning_rate": 9.582431494870693e-07,
+      "loss": 0.2537,
+      "step": 5442
+    },
+    {
+      "epoch": 2.6006809629054417,
+      "grad_norm": 0.45376389340573675,
+      "learning_rate": 9.559946715608037e-07,
+      "loss": 0.2473,
+      "step": 5443
+    },
+    {
+      "epoch": 2.601158831611015,
+      "grad_norm": 0.5018534323943927,
+      "learning_rate": 9.537487022820846e-07,
+      "loss": 0.2692,
+      "step": 5444
+    },
+    {
+      "epoch": 2.601636700316588,
+      "grad_norm": 0.571660106743449,
+      "learning_rate": 9.515052422739035e-07,
+      "loss": 0.2762,
+      "step": 5445
+    },
+    {
+      "epoch": 2.602114569022161,
+      "grad_norm": 0.4478057549739928,
+      "learning_rate": 9.492642921585526e-07,
+      "loss": 0.2676,
+      "step": 5446
+    },
+    {
+      "epoch": 2.6025924377277345,
+      "grad_norm": 0.4695545143748728,
+      "learning_rate": 9.47025852557636e-07,
+      "loss": 0.2771,
+      "step": 5447
+    },
+    {
+      "epoch": 2.6030703064333074,
+      "grad_norm": 0.4452493287419494,
+      "learning_rate": 9.447899240920566e-07,
+      "loss": 0.2578,
+      "step": 5448
+    },
+    {
+      "epoch": 2.6035481751388807,
+      "grad_norm": 0.440730727424376,
+      "learning_rate": 9.425565073820198e-07,
+      "loss": 0.3075,
+      "step": 5449
+    },
+    {
+      "epoch": 2.6040260438444536,
+      "grad_norm": 0.44510497848773967,
+      "learning_rate": 9.403256030470386e-07,
+      "loss": 0.2695,
+      "step": 5450
+    },
+    {
+      "epoch": 2.604503912550027,
+      "grad_norm": 0.4519704788457889,
+      "learning_rate": 9.380972117059262e-07,
+      "loss": 0.2568,
+      "step": 5451
+    },
+    {
+      "epoch": 2.6049817812556,
+      "grad_norm": 0.5598137706232549,
+      "learning_rate": 9.358713339767955e-07,
+      "loss": 0.252,
+      "step": 5452
+    },
+    {
+      "epoch": 2.605459649961173,
+      "grad_norm": 0.46188600754569076,
+      "learning_rate": 9.336479704770696e-07,
+      "loss": 0.2665,
+      "step": 5453
+    },
+    {
+      "epoch": 2.6059375186667464,
+      "grad_norm": 0.49788442318547643,
+      "learning_rate": 9.314271218234693e-07,
+      "loss": 0.2529,
+      "step": 5454
+    },
+    {
+      "epoch": 2.6064153873723193,
+      "grad_norm": 0.4345623662569078,
+      "learning_rate": 9.292087886320166e-07,
+      "loss": 0.2615,
+      "step": 5455
+    },
+    {
+      "epoch": 2.6068932560778926,
+      "grad_norm": 0.44139821572916704,
+      "learning_rate": 9.269929715180404e-07,
+      "loss": 0.2822,
+      "step": 5456
+    },
+    {
+      "epoch": 2.607371124783466,
+      "grad_norm": 0.4252933161869911,
+      "learning_rate": 9.247796710961699e-07,
+      "loss": 0.2483,
+      "step": 5457
+    },
+    {
+      "epoch": 2.6078489934890388,
+      "grad_norm": 0.4412914528127249,
+      "learning_rate": 9.225688879803351e-07,
+      "loss": 0.2647,
+      "step": 5458
+    },
+    {
+      "epoch": 2.608326862194612,
+      "grad_norm": 0.4357571583911059,
+      "learning_rate": 9.203606227837658e-07,
+      "loss": 0.2616,
+      "step": 5459
+    },
+    {
+      "epoch": 2.608804730900185,
+      "grad_norm": 0.4994871920703874,
+      "learning_rate": 9.181548761189996e-07,
+      "loss": 0.2688,
+      "step": 5460
+    },
+    {
+      "epoch": 2.6092825996057583,
+      "grad_norm": 0.4351987789797297,
+      "learning_rate": 9.159516485978692e-07,
+      "loss": 0.2691,
+      "step": 5461
+    },
+    {
+      "epoch": 2.6097604683113316,
+      "grad_norm": 0.4430845065598318,
+      "learning_rate": 9.137509408315104e-07,
+      "loss": 0.2593,
+      "step": 5462
+    },
+    {
+      "epoch": 2.6102383370169044,
+      "grad_norm": 0.47947175182055163,
+      "learning_rate": 9.115527534303637e-07,
+      "loss": 0.2645,
+      "step": 5463
+    },
+    {
+      "epoch": 2.6107162057224778,
+      "grad_norm": 0.4428209512787218,
+      "learning_rate": 9.093570870041645e-07,
+      "loss": 0.2655,
+      "step": 5464
+    },
+    {
+      "epoch": 2.6111940744280506,
+      "grad_norm": 0.46538192551551194,
+      "learning_rate": 9.071639421619527e-07,
+      "loss": 0.2834,
+      "step": 5465
+    },
+    {
+      "epoch": 2.611671943133624,
+      "grad_norm": 0.42759421702996847,
+      "learning_rate": 9.049733195120703e-07,
+      "loss": 0.2607,
+      "step": 5466
+    },
+    {
+      "epoch": 2.6121498118391973,
+      "grad_norm": 0.4208547298808128,
+      "learning_rate": 9.027852196621545e-07,
+      "loss": 0.2422,
+      "step": 5467
+    },
+    {
+      "epoch": 2.6126276805447706,
+      "grad_norm": 0.4382513214899705,
+      "learning_rate": 9.005996432191455e-07,
+      "loss": 0.2649,
+      "step": 5468
+    },
+    {
+      "epoch": 2.6131055492503434,
+      "grad_norm": 0.4914340398062526,
+      "learning_rate": 8.984165907892872e-07,
+      "loss": 0.2635,
+      "step": 5469
+    },
+    {
+      "epoch": 2.6135834179559168,
+      "grad_norm": 0.42931017707652125,
+      "learning_rate": 8.962360629781164e-07,
+      "loss": 0.2506,
+      "step": 5470
+    },
+    {
+      "epoch": 2.6140612866614896,
+      "grad_norm": 0.46339088752838625,
+      "learning_rate": 8.940580603904736e-07,
+      "loss": 0.2733,
+      "step": 5471
+    },
+    {
+      "epoch": 2.614539155367063,
+      "grad_norm": 0.48172702720835325,
+      "learning_rate": 8.918825836304989e-07,
+      "loss": 0.2797,
+      "step": 5472
+    },
+    {
+      "epoch": 2.6150170240726363,
+      "grad_norm": 0.4507180023251733,
+      "learning_rate": 8.89709633301632e-07,
+      "loss": 0.2698,
+      "step": 5473
+    },
+    {
+      "epoch": 2.615494892778209,
+      "grad_norm": 0.4542073635472215,
+      "learning_rate": 8.875392100066082e-07,
+      "loss": 0.2796,
+      "step": 5474
+    },
+    {
+      "epoch": 2.6159727614837824,
+      "grad_norm": 0.4512619657743737,
+      "learning_rate": 8.853713143474685e-07,
+      "loss": 0.2457,
+      "step": 5475
+    },
+    {
+      "epoch": 2.6164506301893553,
+      "grad_norm": 0.43470496799876945,
+      "learning_rate": 8.832059469255461e-07,
+      "loss": 0.2638,
+      "step": 5476
+    },
+    {
+      "epoch": 2.6169284988949286,
+      "grad_norm": 0.4696707255099846,
+      "learning_rate": 8.810431083414761e-07,
+      "loss": 0.2794,
+      "step": 5477
+    },
+    {
+      "epoch": 2.617406367600502,
+      "grad_norm": 0.4654756321536549,
+      "learning_rate": 8.788827991951932e-07,
+      "loss": 0.2772,
+      "step": 5478
+    },
+    {
+      "epoch": 2.617884236306075,
+      "grad_norm": 1.1496364863049888,
+      "learning_rate": 8.767250200859278e-07,
+      "loss": 0.2443,
+      "step": 5479
+    },
+    {
+      "epoch": 2.618362105011648,
+      "grad_norm": 0.5038462450910971,
+      "learning_rate": 8.745697716122081e-07,
+      "loss": 0.2584,
+      "step": 5480
+    },
+    {
+      "epoch": 2.618839973717221,
+      "grad_norm": 0.4531843698900754,
+      "learning_rate": 8.724170543718657e-07,
+      "loss": 0.2617,
+      "step": 5481
+    },
+    {
+      "epoch": 2.6193178424227943,
+      "grad_norm": 0.44068312743815813,
+      "learning_rate": 8.702668689620252e-07,
+      "loss": 0.2664,
+      "step": 5482
+    },
+    {
+      "epoch": 2.6197957111283676,
+      "grad_norm": 0.4527311915552218,
+      "learning_rate": 8.681192159791074e-07,
+      "loss": 0.2439,
+      "step": 5483
+    },
+    {
+      "epoch": 2.6202735798339405,
+      "grad_norm": 0.46007711834116877,
+      "learning_rate": 8.659740960188379e-07,
+      "loss": 0.2632,
+      "step": 5484
+    },
+    {
+      "epoch": 2.620751448539514,
+      "grad_norm": 0.4654714237772124,
+      "learning_rate": 8.638315096762318e-07,
+      "loss": 0.264,
+      "step": 5485
+    },
+    {
+      "epoch": 2.6212293172450867,
+      "grad_norm": 0.4468898821749278,
+      "learning_rate": 8.616914575456048e-07,
+      "loss": 0.259,
+      "step": 5486
+    },
+    {
+      "epoch": 2.62170718595066,
+      "grad_norm": 0.45382995132457704,
+      "learning_rate": 8.595539402205711e-07,
+      "loss": 0.26,
+      "step": 5487
+    },
+    {
+      "epoch": 2.6221850546562333,
+      "grad_norm": 0.4466920475534839,
+      "learning_rate": 8.574189582940407e-07,
+      "loss": 0.2509,
+      "step": 5488
+    },
+    {
+      "epoch": 2.6226629233618066,
+      "grad_norm": 0.4340782648580306,
+      "learning_rate": 8.552865123582143e-07,
+      "loss": 0.2648,
+      "step": 5489
+    },
+    {
+      "epoch": 2.6231407920673795,
+      "grad_norm": 0.44864783353774906,
+      "learning_rate": 8.531566030046035e-07,
+      "loss": 0.2591,
+      "step": 5490
+    },
+    {
+      "epoch": 2.6236186607729524,
+      "grad_norm": 0.48371815986744804,
+      "learning_rate": 8.510292308240043e-07,
+      "loss": 0.2819,
+      "step": 5491
+    },
+    {
+      "epoch": 2.6240965294785257,
+      "grad_norm": 0.4732488135933287,
+      "learning_rate": 8.489043964065091e-07,
+      "loss": 0.2735,
+      "step": 5492
+    },
+    {
+      "epoch": 2.624574398184099,
+      "grad_norm": 0.5366303704759061,
+      "learning_rate": 8.467821003415133e-07,
+      "loss": 0.2659,
+      "step": 5493
+    },
+    {
+      "epoch": 2.6250522668896723,
+      "grad_norm": 0.5036292178699638,
+      "learning_rate": 8.446623432177025e-07,
+      "loss": 0.2566,
+      "step": 5494
+    },
+    {
+      "epoch": 2.625530135595245,
+      "grad_norm": 0.4953780048542685,
+      "learning_rate": 8.425451256230588e-07,
+      "loss": 0.2638,
+      "step": 5495
+    },
+    {
+      "epoch": 2.6260080043008185,
+      "grad_norm": 0.4490484978515238,
+      "learning_rate": 8.404304481448644e-07,
+      "loss": 0.2801,
+      "step": 5496
+    },
+    {
+      "epoch": 2.6264858730063914,
+      "grad_norm": 0.48605177710536757,
+      "learning_rate": 8.383183113696914e-07,
+      "loss": 0.2751,
+      "step": 5497
+    },
+    {
+      "epoch": 2.6269637417119647,
+      "grad_norm": 0.42633172550873427,
+      "learning_rate": 8.362087158834087e-07,
+      "loss": 0.266,
+      "step": 5498
+    },
+    {
+      "epoch": 2.627441610417538,
+      "grad_norm": 0.4472570484490212,
+      "learning_rate": 8.341016622711829e-07,
+      "loss": 0.264,
+      "step": 5499
+    },
+    {
+      "epoch": 2.627919479123111,
+      "grad_norm": 0.45968606482345986,
+      "learning_rate": 8.319971511174718e-07,
+      "loss": 0.2736,
+      "step": 5500
+    },
+    {
+      "epoch": 2.628397347828684,
+      "grad_norm": 0.5381860245104173,
+      "learning_rate": 8.298951830060286e-07,
+      "loss": 0.2638,
+      "step": 5501
+    },
+    {
+      "epoch": 2.628875216534257,
+      "grad_norm": 0.4439356147429359,
+      "learning_rate": 8.277957585199059e-07,
+      "loss": 0.2721,
+      "step": 5502
+    },
+    {
+      "epoch": 2.6293530852398304,
+      "grad_norm": 0.46056697797714147,
+      "learning_rate": 8.256988782414454e-07,
+      "loss": 0.2774,
+      "step": 5503
+    },
+    {
+      "epoch": 2.6298309539454037,
+      "grad_norm": 0.4429290965835113,
+      "learning_rate": 8.23604542752281e-07,
+      "loss": 0.2608,
+      "step": 5504
+    },
+    {
+      "epoch": 2.6303088226509765,
+      "grad_norm": 0.46251564537739054,
+      "learning_rate": 8.215127526333499e-07,
+      "loss": 0.2772,
+      "step": 5505
+    },
+    {
+      "epoch": 2.63078669135655,
+      "grad_norm": 0.4896830023615981,
+      "learning_rate": 8.19423508464876e-07,
+      "loss": 0.2595,
+      "step": 5506
+    },
+    {
+      "epoch": 2.6312645600621227,
+      "grad_norm": 0.44334268809073074,
+      "learning_rate": 8.173368108263768e-07,
+      "loss": 0.2604,
+      "step": 5507
+    },
+    {
+      "epoch": 2.631742428767696,
+      "grad_norm": 0.47811781456110813,
+      "learning_rate": 8.15252660296667e-07,
+      "loss": 0.2692,
+      "step": 5508
+    },
+    {
+      "epoch": 2.6322202974732694,
+      "grad_norm": 0.4363044164663179,
+      "learning_rate": 8.131710574538543e-07,
+      "loss": 0.2454,
+      "step": 5509
+    },
+    {
+      "epoch": 2.6326981661788422,
+      "grad_norm": 0.449330353460876,
+      "learning_rate": 8.110920028753355e-07,
+      "loss": 0.277,
+      "step": 5510
+    },
+    {
+      "epoch": 2.6331760348844155,
+      "grad_norm": 0.4961190817267826,
+      "learning_rate": 8.090154971378073e-07,
+      "loss": 0.27,
+      "step": 5511
+    },
+    {
+      "epoch": 2.6336539035899884,
+      "grad_norm": 0.45171261984580413,
+      "learning_rate": 8.069415408172543e-07,
+      "loss": 0.2712,
+      "step": 5512
+    },
+    {
+      "epoch": 2.6341317722955617,
+      "grad_norm": 0.4434767897838476,
+      "learning_rate": 8.048701344889531e-07,
+      "loss": 0.2508,
+      "step": 5513
+    },
+    {
+      "epoch": 2.634609641001135,
+      "grad_norm": 0.4623705120513708,
+      "learning_rate": 8.028012787274786e-07,
+      "loss": 0.2844,
+      "step": 5514
+    },
+    {
+      "epoch": 2.6350875097067084,
+      "grad_norm": 0.4574657819597167,
+      "learning_rate": 8.007349741066939e-07,
+      "loss": 0.2568,
+      "step": 5515
+    },
+    {
+      "epoch": 2.6355653784122812,
+      "grad_norm": 0.4450702051786837,
+      "learning_rate": 7.986712211997538e-07,
+      "loss": 0.2759,
+      "step": 5516
+    },
+    {
+      "epoch": 2.636043247117854,
+      "grad_norm": 0.6165437153757278,
+      "learning_rate": 7.966100205791094e-07,
+      "loss": 0.2445,
+      "step": 5517
+    },
+    {
+      "epoch": 2.6365211158234274,
+      "grad_norm": 0.45144330242103814,
+      "learning_rate": 7.945513728164999e-07,
+      "loss": 0.2495,
+      "step": 5518
+    },
+    {
+      "epoch": 2.6369989845290007,
+      "grad_norm": 0.4564427161692554,
+      "learning_rate": 7.924952784829576e-07,
+      "loss": 0.2617,
+      "step": 5519
+    },
+    {
+      "epoch": 2.637476853234574,
+      "grad_norm": 0.48463913280050425,
+      "learning_rate": 7.904417381488083e-07,
+      "loss": 0.2655,
+      "step": 5520
+    },
+    {
+      "epoch": 2.637954721940147,
+      "grad_norm": 0.4516813676731921,
+      "learning_rate": 7.883907523836676e-07,
+      "loss": 0.2695,
+      "step": 5521
+    },
+    {
+      "epoch": 2.63843259064572,
+      "grad_norm": 0.4604887951307146,
+      "learning_rate": 7.863423217564403e-07,
+      "loss": 0.2529,
+      "step": 5522
+    },
+    {
+      "epoch": 2.638910459351293,
+      "grad_norm": 0.46157386421059765,
+      "learning_rate": 7.842964468353265e-07,
+      "loss": 0.2618,
+      "step": 5523
+    },
+    {
+      "epoch": 2.6393883280568664,
+      "grad_norm": 0.4539399407409506,
+      "learning_rate": 7.822531281878188e-07,
+      "loss": 0.2526,
+      "step": 5524
+    },
+    {
+      "epoch": 2.6398661967624397,
+      "grad_norm": 0.44721471315508016,
+      "learning_rate": 7.802123663806938e-07,
+      "loss": 0.2355,
+      "step": 5525
+    },
+    {
+      "epoch": 2.6403440654680126,
+      "grad_norm": 0.45922430751841153,
+      "learning_rate": 7.781741619800231e-07,
+      "loss": 0.261,
+      "step": 5526
+    },
+    {
+      "epoch": 2.640821934173586,
+      "grad_norm": 0.46053983816343896,
+      "learning_rate": 7.761385155511714e-07,
+      "loss": 0.2486,
+      "step": 5527
+    },
+    {
+      "epoch": 2.6412998028791588,
+      "grad_norm": 0.45385759627956784,
+      "learning_rate": 7.741054276587889e-07,
+      "loss": 0.2623,
+      "step": 5528
+    },
+    {
+      "epoch": 2.641777671584732,
+      "grad_norm": 0.44483157426363334,
+      "learning_rate": 7.72074898866817e-07,
+      "loss": 0.2903,
+      "step": 5529
+    },
+    {
+      "epoch": 2.6422555402903054,
+      "grad_norm": 0.46467024535160295,
+      "learning_rate": 7.700469297384927e-07,
+      "loss": 0.27,
+      "step": 5530
+    },
+    {
+      "epoch": 2.6427334089958783,
+      "grad_norm": 0.46336583237822215,
+      "learning_rate": 7.680215208363362e-07,
+      "loss": 0.2571,
+      "step": 5531
+    },
+    {
+      "epoch": 2.6432112777014516,
+      "grad_norm": 0.4589731194382133,
+      "learning_rate": 7.659986727221591e-07,
+      "loss": 0.2748,
+      "step": 5532
+    },
+    {
+      "epoch": 2.6436891464070245,
+      "grad_norm": 0.4479455221268227,
+      "learning_rate": 7.63978385957066e-07,
+      "loss": 0.2579,
+      "step": 5533
+    },
+    {
+      "epoch": 2.6441670151125978,
+      "grad_norm": 0.45680582982603707,
+      "learning_rate": 7.619606611014485e-07,
+      "loss": 0.2563,
+      "step": 5534
+    },
+    {
+      "epoch": 2.644644883818171,
+      "grad_norm": 0.4341476061003986,
+      "learning_rate": 7.599454987149868e-07,
+      "loss": 0.2633,
+      "step": 5535
+    },
+    {
+      "epoch": 2.645122752523744,
+      "grad_norm": 0.5906143297103831,
+      "learning_rate": 7.579328993566526e-07,
+      "loss": 0.287,
+      "step": 5536
+    },
+    {
+      "epoch": 2.6456006212293173,
+      "grad_norm": 0.5449600193189947,
+      "learning_rate": 7.559228635847049e-07,
+      "loss": 0.267,
+      "step": 5537
+    },
+    {
+      "epoch": 2.64607848993489,
+      "grad_norm": 0.46047021853305603,
+      "learning_rate": 7.5391539195669e-07,
+      "loss": 0.2745,
+      "step": 5538
+    },
+    {
+      "epoch": 2.6465563586404635,
+      "grad_norm": 0.4884388346418383,
+      "learning_rate": 7.5191048502945e-07,
+      "loss": 0.2482,
+      "step": 5539
+    },
+    {
+      "epoch": 2.6470342273460368,
+      "grad_norm": 0.4774890364230652,
+      "learning_rate": 7.499081433591071e-07,
+      "loss": 0.2819,
+      "step": 5540
+    },
+    {
+      "epoch": 2.64751209605161,
+      "grad_norm": 0.43773957361680527,
+      "learning_rate": 7.479083675010746e-07,
+      "loss": 0.2711,
+      "step": 5541
+    },
+    {
+      "epoch": 2.647989964757183,
+      "grad_norm": 0.4400542304427017,
+      "learning_rate": 7.459111580100587e-07,
+      "loss": 0.2584,
+      "step": 5542
+    },
+    {
+      "epoch": 2.6484678334627563,
+      "grad_norm": 0.47571447666450534,
+      "learning_rate": 7.439165154400485e-07,
+      "loss": 0.2619,
+      "step": 5543
+    },
+    {
+      "epoch": 2.648945702168329,
+      "grad_norm": 0.45317629524346903,
+      "learning_rate": 7.419244403443215e-07,
+      "loss": 0.2426,
+      "step": 5544
+    },
+    {
+      "epoch": 2.6494235708739025,
+      "grad_norm": 0.43903772149485587,
+      "learning_rate": 7.399349332754458e-07,
+      "loss": 0.2731,
+      "step": 5545
+    },
+    {
+      "epoch": 2.6499014395794758,
+      "grad_norm": 0.45753347345809436,
+      "learning_rate": 7.379479947852752e-07,
+      "loss": 0.2752,
+      "step": 5546
+    },
+    {
+      "epoch": 2.6503793082850486,
+      "grad_norm": 0.4465731511139684,
+      "learning_rate": 7.359636254249491e-07,
+      "loss": 0.2749,
+      "step": 5547
+    },
+    {
+      "epoch": 2.650857176990622,
+      "grad_norm": 0.45206339925897226,
+      "learning_rate": 7.339818257448994e-07,
+      "loss": 0.2641,
+      "step": 5548
+    },
+    {
+      "epoch": 2.651335045696195,
+      "grad_norm": 0.44301665799534984,
+      "learning_rate": 7.320025962948429e-07,
+      "loss": 0.266,
+      "step": 5549
+    },
+    {
+      "epoch": 2.651812914401768,
+      "grad_norm": 0.4510933254303754,
+      "learning_rate": 7.300259376237795e-07,
+      "loss": 0.2658,
+      "step": 5550
+    },
+    {
+      "epoch": 2.6522907831073415,
+      "grad_norm": 0.45072706019066416,
+      "learning_rate": 7.280518502800027e-07,
+      "loss": 0.276,
+      "step": 5551
+    },
+    {
+      "epoch": 2.6527686518129143,
+      "grad_norm": 0.5069678986700116,
+      "learning_rate": 7.260803348110879e-07,
+      "loss": 0.2749,
+      "step": 5552
+    },
+    {
+      "epoch": 2.6532465205184876,
+      "grad_norm": 0.4204064472652244,
+      "learning_rate": 7.241113917638987e-07,
+      "loss": 0.2482,
+      "step": 5553
+    },
+    {
+      "epoch": 2.6537243892240605,
+      "grad_norm": 0.4606764828883414,
+      "learning_rate": 7.22145021684586e-07,
+      "loss": 0.2686,
+      "step": 5554
+    },
+    {
+      "epoch": 2.654202257929634,
+      "grad_norm": 0.43007497445227916,
+      "learning_rate": 7.201812251185869e-07,
+      "loss": 0.2331,
+      "step": 5555
+    },
+    {
+      "epoch": 2.654680126635207,
+      "grad_norm": 0.43763153678237027,
+      "learning_rate": 7.182200026106201e-07,
+      "loss": 0.2569,
+      "step": 5556
+    },
+    {
+      "epoch": 2.65515799534078,
+      "grad_norm": 0.4720679073397097,
+      "learning_rate": 7.16261354704697e-07,
+      "loss": 0.2645,
+      "step": 5557
+    },
+    {
+      "epoch": 2.6556358640463533,
+      "grad_norm": 0.4342444064007902,
+      "learning_rate": 7.143052819441143e-07,
+      "loss": 0.2586,
+      "step": 5558
+    },
+    {
+      "epoch": 2.656113732751926,
+      "grad_norm": 0.4761061871295211,
+      "learning_rate": 7.123517848714479e-07,
+      "loss": 0.2519,
+      "step": 5559
+    },
+    {
+      "epoch": 2.6565916014574995,
+      "grad_norm": 0.4487477696170997,
+      "learning_rate": 7.104008640285642e-07,
+      "loss": 0.2691,
+      "step": 5560
+    },
+    {
+      "epoch": 2.657069470163073,
+      "grad_norm": 0.43960627882338654,
+      "learning_rate": 7.084525199566172e-07,
+      "loss": 0.2735,
+      "step": 5561
+    },
+    {
+      "epoch": 2.6575473388686457,
+      "grad_norm": 0.43396104580229183,
+      "learning_rate": 7.065067531960412e-07,
+      "loss": 0.2416,
+      "step": 5562
+    },
+    {
+      "epoch": 2.658025207574219,
+      "grad_norm": 0.4262202664604116,
+      "learning_rate": 7.045635642865555e-07,
+      "loss": 0.246,
+      "step": 5563
+    },
+    {
+      "epoch": 2.658503076279792,
+      "grad_norm": 0.4305745545871823,
+      "learning_rate": 7.026229537671692e-07,
+      "loss": 0.2754,
+      "step": 5564
+    },
+    {
+      "epoch": 2.658980944985365,
+      "grad_norm": 0.595294358705173,
+      "learning_rate": 7.006849221761736e-07,
+      "loss": 0.2681,
+      "step": 5565
+    },
+    {
+      "epoch": 2.6594588136909385,
+      "grad_norm": 0.4579698632907277,
+      "learning_rate": 6.987494700511411e-07,
+      "loss": 0.2691,
+      "step": 5566
+    },
+    {
+      "epoch": 2.659936682396512,
+      "grad_norm": 0.42289834526883935,
+      "learning_rate": 6.968165979289365e-07,
+      "loss": 0.2346,
+      "step": 5567
+    },
+    {
+      "epoch": 2.6604145511020847,
+      "grad_norm": 0.43997486144720366,
+      "learning_rate": 6.948863063457023e-07,
+      "loss": 0.2627,
+      "step": 5568
+    },
+    {
+      "epoch": 2.660892419807658,
+      "grad_norm": 0.5245364644243271,
+      "learning_rate": 6.929585958368656e-07,
+      "loss": 0.2744,
+      "step": 5569
+    },
+    {
+      "epoch": 2.661370288513231,
+      "grad_norm": 0.44871184731230485,
+      "learning_rate": 6.910334669371433e-07,
+      "loss": 0.2597,
+      "step": 5570
+    },
+    {
+      "epoch": 2.661848157218804,
+      "grad_norm": 0.4530485747359608,
+      "learning_rate": 6.891109201805291e-07,
+      "loss": 0.289,
+      "step": 5571
+    },
+    {
+      "epoch": 2.6623260259243775,
+      "grad_norm": 0.4459777138982337,
+      "learning_rate": 6.871909561003032e-07,
+      "loss": 0.2773,
+      "step": 5572
+    },
+    {
+      "epoch": 2.6628038946299504,
+      "grad_norm": 0.44624182034655313,
+      "learning_rate": 6.852735752290318e-07,
+      "loss": 0.2535,
+      "step": 5573
+    },
+    {
+      "epoch": 2.6632817633355237,
+      "grad_norm": 0.4408183385200215,
+      "learning_rate": 6.833587780985618e-07,
+      "loss": 0.2609,
+      "step": 5574
+    },
+    {
+      "epoch": 2.6637596320410966,
+      "grad_norm": 0.48483246105806804,
+      "learning_rate": 6.814465652400237e-07,
+      "loss": 0.2797,
+      "step": 5575
+    },
+    {
+      "epoch": 2.66423750074667,
+      "grad_norm": 0.4412804280001876,
+      "learning_rate": 6.795369371838323e-07,
+      "loss": 0.2561,
+      "step": 5576
+    },
+    {
+      "epoch": 2.664715369452243,
+      "grad_norm": 0.4401284623732288,
+      "learning_rate": 6.776298944596849e-07,
+      "loss": 0.2667,
+      "step": 5577
+    },
+    {
+      "epoch": 2.665193238157816,
+      "grad_norm": 0.49218861155591376,
+      "learning_rate": 6.757254375965583e-07,
+      "loss": 0.2639,
+      "step": 5578
+    },
+    {
+      "epoch": 2.6656711068633894,
+      "grad_norm": 0.4452641669010618,
+      "learning_rate": 6.738235671227212e-07,
+      "loss": 0.2782,
+      "step": 5579
+    },
+    {
+      "epoch": 2.6661489755689622,
+      "grad_norm": 0.4385273221330306,
+      "learning_rate": 6.719242835657147e-07,
+      "loss": 0.2631,
+      "step": 5580
+    },
+    {
+      "epoch": 2.6666268442745356,
+      "grad_norm": 0.42111306891766304,
+      "learning_rate": 6.700275874523665e-07,
+      "loss": 0.262,
+      "step": 5581
+    },
+    {
+      "epoch": 2.667104712980109,
+      "grad_norm": 0.4348075279027199,
+      "learning_rate": 6.681334793087879e-07,
+      "loss": 0.2421,
+      "step": 5582
+    },
+    {
+      "epoch": 2.6675825816856817,
+      "grad_norm": 0.5049628037720003,
+      "learning_rate": 6.66241959660372e-07,
+      "loss": 0.2847,
+      "step": 5583
+    },
+    {
+      "epoch": 2.668060450391255,
+      "grad_norm": 0.4683050522187598,
+      "learning_rate": 6.643530290317901e-07,
+      "loss": 0.2568,
+      "step": 5584
+    },
+    {
+      "epoch": 2.668538319096828,
+      "grad_norm": 0.4496767613831931,
+      "learning_rate": 6.62466687947001e-07,
+      "loss": 0.255,
+      "step": 5585
+    },
+    {
+      "epoch": 2.6690161878024012,
+      "grad_norm": 0.4383351210567021,
+      "learning_rate": 6.605829369292427e-07,
+      "loss": 0.2624,
+      "step": 5586
+    },
+    {
+      "epoch": 2.6694940565079746,
+      "grad_norm": 0.4633981201266098,
+      "learning_rate": 6.587017765010306e-07,
+      "loss": 0.2696,
+      "step": 5587
+    },
+    {
+      "epoch": 2.6699719252135474,
+      "grad_norm": 0.43636703634880375,
+      "learning_rate": 6.568232071841695e-07,
+      "loss": 0.2474,
+      "step": 5588
+    },
+    {
+      "epoch": 2.6704497939191207,
+      "grad_norm": 0.4639743358839676,
+      "learning_rate": 6.549472294997405e-07,
+      "loss": 0.2789,
+      "step": 5589
+    },
+    {
+      "epoch": 2.6709276626246936,
+      "grad_norm": 0.48119723931800046,
+      "learning_rate": 6.530738439681017e-07,
+      "loss": 0.2675,
+      "step": 5590
+    },
+    {
+      "epoch": 2.671405531330267,
+      "grad_norm": 0.4541784315517988,
+      "learning_rate": 6.512030511089063e-07,
+      "loss": 0.2663,
+      "step": 5591
+    },
+    {
+      "epoch": 2.6718834000358402,
+      "grad_norm": 0.4386851803015807,
+      "learning_rate": 6.493348514410735e-07,
+      "loss": 0.2664,
+      "step": 5592
+    },
+    {
+      "epoch": 2.6723612687414136,
+      "grad_norm": 0.443049021233755,
+      "learning_rate": 6.474692454828091e-07,
+      "loss": 0.2634,
+      "step": 5593
+    },
+    {
+      "epoch": 2.6728391374469864,
+      "grad_norm": 0.5368940349519489,
+      "learning_rate": 6.456062337516023e-07,
+      "loss": 0.2691,
+      "step": 5594
+    },
+    {
+      "epoch": 2.6733170061525597,
+      "grad_norm": 0.4362459871804461,
+      "learning_rate": 6.437458167642164e-07,
+      "loss": 0.2768,
+      "step": 5595
+    },
+    {
+      "epoch": 2.6737948748581326,
+      "grad_norm": 0.43405302870615214,
+      "learning_rate": 6.418879950366986e-07,
+      "loss": 0.2628,
+      "step": 5596
+    },
+    {
+      "epoch": 2.674272743563706,
+      "grad_norm": 0.4537944478141257,
+      "learning_rate": 6.400327690843777e-07,
+      "loss": 0.2537,
+      "step": 5597
+    },
+    {
+      "epoch": 2.6747506122692792,
+      "grad_norm": 0.44225140880189395,
+      "learning_rate": 6.3818013942186e-07,
+      "loss": 0.2623,
+      "step": 5598
+    },
+    {
+      "epoch": 2.675228480974852,
+      "grad_norm": 0.48940459598044767,
+      "learning_rate": 6.363301065630301e-07,
+      "loss": 0.262,
+      "step": 5599
+    },
+    {
+      "epoch": 2.6757063496804254,
+      "grad_norm": 0.4290257298591847,
+      "learning_rate": 6.344826710210584e-07,
+      "loss": 0.2574,
+      "step": 5600
+    },
+    {
+      "epoch": 2.6761842183859983,
+      "grad_norm": 0.44260393461554365,
+      "learning_rate": 6.326378333083883e-07,
+      "loss": 0.2613,
+      "step": 5601
+    },
+    {
+      "epoch": 2.6766620870915716,
+      "grad_norm": 0.47868079245627504,
+      "learning_rate": 6.307955939367449e-07,
+      "loss": 0.2636,
+      "step": 5602
+    },
+    {
+      "epoch": 2.677139955797145,
+      "grad_norm": 0.43232776082930185,
+      "learning_rate": 6.289559534171353e-07,
+      "loss": 0.2682,
+      "step": 5603
+    },
+    {
+      "epoch": 2.677617824502718,
+      "grad_norm": 0.44295068309351476,
+      "learning_rate": 6.27118912259842e-07,
+      "loss": 0.2532,
+      "step": 5604
+    },
+    {
+      "epoch": 2.678095693208291,
+      "grad_norm": 0.43112065814472633,
+      "learning_rate": 6.252844709744255e-07,
+      "loss": 0.2847,
+      "step": 5605
+    },
+    {
+      "epoch": 2.678573561913864,
+      "grad_norm": 0.44589686576732196,
+      "learning_rate": 6.234526300697308e-07,
+      "loss": 0.2586,
+      "step": 5606
+    },
+    {
+      "epoch": 2.6790514306194373,
+      "grad_norm": 0.46936007169833993,
+      "learning_rate": 6.216233900538782e-07,
+      "loss": 0.262,
+      "step": 5607
+    },
+    {
+      "epoch": 2.6795292993250106,
+      "grad_norm": 0.45133407836159795,
+      "learning_rate": 6.197967514342628e-07,
+      "loss": 0.2667,
+      "step": 5608
+    },
+    {
+      "epoch": 2.6800071680305835,
+      "grad_norm": 0.4988501134193684,
+      "learning_rate": 6.179727147175663e-07,
+      "loss": 0.2666,
+      "step": 5609
+    },
+    {
+      "epoch": 2.680485036736157,
+      "grad_norm": 0.43159011005236786,
+      "learning_rate": 6.161512804097436e-07,
+      "loss": 0.2696,
+      "step": 5610
+    },
+    {
+      "epoch": 2.6809629054417297,
+      "grad_norm": 0.4392692032042625,
+      "learning_rate": 6.143324490160252e-07,
+      "loss": 0.2564,
+      "step": 5611
+    },
+    {
+      "epoch": 2.681440774147303,
+      "grad_norm": 0.5262468308872646,
+      "learning_rate": 6.125162210409263e-07,
+      "loss": 0.2881,
+      "step": 5612
+    },
+    {
+      "epoch": 2.6819186428528763,
+      "grad_norm": 0.4429786878821737,
+      "learning_rate": 6.107025969882363e-07,
+      "loss": 0.2745,
+      "step": 5613
+    },
+    {
+      "epoch": 2.682396511558449,
+      "grad_norm": 0.4290197449049834,
+      "learning_rate": 6.088915773610194e-07,
+      "loss": 0.2735,
+      "step": 5614
+    },
+    {
+      "epoch": 2.6828743802640225,
+      "grad_norm": 0.44057175272421367,
+      "learning_rate": 6.070831626616236e-07,
+      "loss": 0.2694,
+      "step": 5615
+    },
+    {
+      "epoch": 2.6833522489695953,
+      "grad_norm": 0.46685871466805356,
+      "learning_rate": 6.052773533916712e-07,
+      "loss": 0.2663,
+      "step": 5616
+    },
+    {
+      "epoch": 2.6838301176751687,
+      "grad_norm": 0.44112906840605176,
+      "learning_rate": 6.034741500520591e-07,
+      "loss": 0.2731,
+      "step": 5617
+    },
+    {
+      "epoch": 2.684307986380742,
+      "grad_norm": 0.4480978993764805,
+      "learning_rate": 6.016735531429674e-07,
+      "loss": 0.2638,
+      "step": 5618
+    },
+    {
+      "epoch": 2.6847858550863153,
+      "grad_norm": 0.4757023777345707,
+      "learning_rate": 5.998755631638486e-07,
+      "loss": 0.2625,
+      "step": 5619
+    },
+    {
+      "epoch": 2.685263723791888,
+      "grad_norm": 0.43772130701917733,
+      "learning_rate": 5.980801806134318e-07,
+      "loss": 0.2628,
+      "step": 5620
+    },
+    {
+      "epoch": 2.6857415924974615,
+      "grad_norm": 0.4408707212239334,
+      "learning_rate": 5.962874059897273e-07,
+      "loss": 0.259,
+      "step": 5621
+    },
+    {
+      "epoch": 2.6862194612030343,
+      "grad_norm": 0.44875603266564296,
+      "learning_rate": 5.944972397900173e-07,
+      "loss": 0.252,
+      "step": 5622
+    },
+    {
+      "epoch": 2.6866973299086077,
+      "grad_norm": 0.7243707606230397,
+      "learning_rate": 5.927096825108614e-07,
+      "loss": 0.2644,
+      "step": 5623
+    },
+    {
+      "epoch": 2.687175198614181,
+      "grad_norm": 0.4437660770119595,
+      "learning_rate": 5.909247346480973e-07,
+      "loss": 0.2584,
+      "step": 5624
+    },
+    {
+      "epoch": 2.687653067319754,
+      "grad_norm": 0.4428332177909088,
+      "learning_rate": 5.891423966968413e-07,
+      "loss": 0.2785,
+      "step": 5625
+    },
+    {
+      "epoch": 2.688130936025327,
+      "grad_norm": 0.5234803476776241,
+      "learning_rate": 5.873626691514789e-07,
+      "loss": 0.249,
+      "step": 5626
+    },
+    {
+      "epoch": 2.6886088047309,
+      "grad_norm": 0.47236649678503034,
+      "learning_rate": 5.855855525056742e-07,
+      "loss": 0.2761,
+      "step": 5627
+    },
+    {
+      "epoch": 2.6890866734364733,
+      "grad_norm": 0.45396864953408195,
+      "learning_rate": 5.838110472523728e-07,
+      "loss": 0.2694,
+      "step": 5628
+    },
+    {
+      "epoch": 2.6895645421420467,
+      "grad_norm": 0.44956507553613845,
+      "learning_rate": 5.820391538837866e-07,
+      "loss": 0.2608,
+      "step": 5629
+    },
+    {
+      "epoch": 2.6900424108476195,
+      "grad_norm": 0.4401864408494058,
+      "learning_rate": 5.80269872891408e-07,
+      "loss": 0.2655,
+      "step": 5630
+    },
+    {
+      "epoch": 2.690520279553193,
+      "grad_norm": 0.4886914370387613,
+      "learning_rate": 5.785032047660077e-07,
+      "loss": 0.2549,
+      "step": 5631
+    },
+    {
+      "epoch": 2.6909981482587657,
+      "grad_norm": 0.44296980817168874,
+      "learning_rate": 5.76739149997625e-07,
+      "loss": 0.247,
+      "step": 5632
+    },
+    {
+      "epoch": 2.691476016964339,
+      "grad_norm": 0.44550057207332994,
+      "learning_rate": 5.749777090755781e-07,
+      "loss": 0.2618,
+      "step": 5633
+    },
+    {
+      "epoch": 2.6919538856699123,
+      "grad_norm": 0.43268666197260663,
+      "learning_rate": 5.73218882488461e-07,
+      "loss": 0.2688,
+      "step": 5634
+    },
+    {
+      "epoch": 2.692431754375485,
+      "grad_norm": 0.4461412386675045,
+      "learning_rate": 5.714626707241411e-07,
+      "loss": 0.2718,
+      "step": 5635
+    },
+    {
+      "epoch": 2.6929096230810585,
+      "grad_norm": 0.4558906789502105,
+      "learning_rate": 5.697090742697576e-07,
+      "loss": 0.2792,
+      "step": 5636
+    },
+    {
+      "epoch": 2.6933874917866314,
+      "grad_norm": 0.4744314781330036,
+      "learning_rate": 5.679580936117312e-07,
+      "loss": 0.2591,
+      "step": 5637
+    },
+    {
+      "epoch": 2.6938653604922047,
+      "grad_norm": 0.5192870350309746,
+      "learning_rate": 5.662097292357505e-07,
+      "loss": 0.2734,
+      "step": 5638
+    },
+    {
+      "epoch": 2.694343229197778,
+      "grad_norm": 0.4435569578566844,
+      "learning_rate": 5.644639816267817e-07,
+      "loss": 0.2666,
+      "step": 5639
+    },
+    {
+      "epoch": 2.694821097903351,
+      "grad_norm": 0.43856696864544303,
+      "learning_rate": 5.627208512690641e-07,
+      "loss": 0.2705,
+      "step": 5640
+    },
+    {
+      "epoch": 2.695298966608924,
+      "grad_norm": 0.45553085802947646,
+      "learning_rate": 5.609803386461133e-07,
+      "loss": 0.2656,
+      "step": 5641
+    },
+    {
+      "epoch": 2.695776835314497,
+      "grad_norm": 0.4666048945211158,
+      "learning_rate": 5.59242444240713e-07,
+      "loss": 0.2738,
+      "step": 5642
+    },
+    {
+      "epoch": 2.6962547040200704,
+      "grad_norm": 0.43113345766045935,
+      "learning_rate": 5.575071685349276e-07,
+      "loss": 0.2572,
+      "step": 5643
+    },
+    {
+      "epoch": 2.6967325727256437,
+      "grad_norm": 1.125132833480618,
+      "learning_rate": 5.55774512010091e-07,
+      "loss": 0.2558,
+      "step": 5644
+    },
+    {
+      "epoch": 2.697210441431217,
+      "grad_norm": 0.608140367809618,
+      "learning_rate": 5.5404447514681e-07,
+      "loss": 0.2737,
+      "step": 5645
+    },
+    {
+      "epoch": 2.69768831013679,
+      "grad_norm": 0.4533877869576038,
+      "learning_rate": 5.523170584249704e-07,
+      "loss": 0.2498,
+      "step": 5646
+    },
+    {
+      "epoch": 2.698166178842363,
+      "grad_norm": 0.4415778575924575,
+      "learning_rate": 5.505922623237237e-07,
+      "loss": 0.2639,
+      "step": 5647
+    },
+    {
+      "epoch": 2.698644047547936,
+      "grad_norm": 0.42774906759972836,
+      "learning_rate": 5.488700873214969e-07,
+      "loss": 0.2692,
+      "step": 5648
+    },
+    {
+      "epoch": 2.6991219162535094,
+      "grad_norm": 0.5330417311187563,
+      "learning_rate": 5.471505338959948e-07,
+      "loss": 0.2707,
+      "step": 5649
+    },
+    {
+      "epoch": 2.6995997849590827,
+      "grad_norm": 0.5461645932501131,
+      "learning_rate": 5.45433602524188e-07,
+      "loss": 0.2588,
+      "step": 5650
+    },
+    {
+      "epoch": 2.7000776536646556,
+      "grad_norm": 0.4242076160737116,
+      "learning_rate": 5.437192936823243e-07,
+      "loss": 0.2651,
+      "step": 5651
+    },
+    {
+      "epoch": 2.700555522370229,
+      "grad_norm": 0.4676634194490461,
+      "learning_rate": 5.420076078459236e-07,
+      "loss": 0.27,
+      "step": 5652
+    },
+    {
+      "epoch": 2.7010333910758018,
+      "grad_norm": 0.43263645783838944,
+      "learning_rate": 5.402985454897758e-07,
+      "loss": 0.2589,
+      "step": 5653
+    },
+    {
+      "epoch": 2.701511259781375,
+      "grad_norm": 0.45382249392362367,
+      "learning_rate": 5.385921070879441e-07,
+      "loss": 0.2487,
+      "step": 5654
+    },
+    {
+      "epoch": 2.7019891284869484,
+      "grad_norm": 0.49441942637541253,
+      "learning_rate": 5.368882931137675e-07,
+      "loss": 0.2565,
+      "step": 5655
+    },
+    {
+      "epoch": 2.7024669971925213,
+      "grad_norm": 0.4405834049292135,
+      "learning_rate": 5.351871040398515e-07,
+      "loss": 0.2744,
+      "step": 5656
+    },
+    {
+      "epoch": 2.7029448658980946,
+      "grad_norm": 0.5559524037795076,
+      "learning_rate": 5.33488540338074e-07,
+      "loss": 0.2583,
+      "step": 5657
+    },
+    {
+      "epoch": 2.7034227346036674,
+      "grad_norm": 0.43260600511156877,
+      "learning_rate": 5.317926024795906e-07,
+      "loss": 0.2572,
+      "step": 5658
+    },
+    {
+      "epoch": 2.7039006033092408,
+      "grad_norm": 0.4394492413785682,
+      "learning_rate": 5.300992909348234e-07,
+      "loss": 0.2545,
+      "step": 5659
+    },
+    {
+      "epoch": 2.704378472014814,
+      "grad_norm": 0.4362896925210051,
+      "learning_rate": 5.284086061734672e-07,
+      "loss": 0.2733,
+      "step": 5660
+    },
+    {
+      "epoch": 2.704856340720387,
+      "grad_norm": 0.5093789250306926,
+      "learning_rate": 5.267205486644866e-07,
+      "loss": 0.2701,
+      "step": 5661
+    },
+    {
+      "epoch": 2.7053342094259603,
+      "grad_norm": 0.48183431324608156,
+      "learning_rate": 5.250351188761204e-07,
+      "loss": 0.2541,
+      "step": 5662
+    },
+    {
+      "epoch": 2.705812078131533,
+      "grad_norm": 0.4409302418627837,
+      "learning_rate": 5.23352317275877e-07,
+      "loss": 0.2602,
+      "step": 5663
+    },
+    {
+      "epoch": 2.7062899468371064,
+      "grad_norm": 0.46481073306379156,
+      "learning_rate": 5.21672144330535e-07,
+      "loss": 0.2668,
+      "step": 5664
+    },
+    {
+      "epoch": 2.7067678155426798,
+      "grad_norm": 0.573145605718102,
+      "learning_rate": 5.199946005061462e-07,
+      "loss": 0.248,
+      "step": 5665
+    },
+    {
+      "epoch": 2.7072456842482526,
+      "grad_norm": 0.4458113575270228,
+      "learning_rate": 5.183196862680307e-07,
+      "loss": 0.2495,
+      "step": 5666
+    },
+    {
+      "epoch": 2.707723552953826,
+      "grad_norm": 0.566881416643556,
+      "learning_rate": 5.166474020807788e-07,
+      "loss": 0.2699,
+      "step": 5667
+    },
+    {
+      "epoch": 2.708201421659399,
+      "grad_norm": 0.49148904309220365,
+      "learning_rate": 5.149777484082552e-07,
+      "loss": 0.2686,
+      "step": 5668
+    },
+    {
+      "epoch": 2.708679290364972,
+      "grad_norm": 0.4396165691920815,
+      "learning_rate": 5.133107257135917e-07,
+      "loss": 0.2695,
+      "step": 5669
+    },
+    {
+      "epoch": 2.7091571590705454,
+      "grad_norm": 0.4733614560975121,
+      "learning_rate": 5.116463344591893e-07,
+      "loss": 0.2687,
+      "step": 5670
+    },
+    {
+      "epoch": 2.7096350277761188,
+      "grad_norm": 0.4756203545300059,
+      "learning_rate": 5.099845751067234e-07,
+      "loss": 0.2652,
+      "step": 5671
+    },
+    {
+      "epoch": 2.7101128964816916,
+      "grad_norm": 0.4877805042234815,
+      "learning_rate": 5.083254481171352e-07,
+      "loss": 0.2786,
+      "step": 5672
+    },
+    {
+      "epoch": 2.710590765187265,
+      "grad_norm": 0.43232222654764546,
+      "learning_rate": 5.066689539506353e-07,
+      "loss": 0.2713,
+      "step": 5673
+    },
+    {
+      "epoch": 2.711068633892838,
+      "grad_norm": 0.44816531622174133,
+      "learning_rate": 5.050150930667108e-07,
+      "loss": 0.2755,
+      "step": 5674
+    },
+    {
+      "epoch": 2.711546502598411,
+      "grad_norm": 0.4627071189494935,
+      "learning_rate": 5.033638659241102e-07,
+      "loss": 0.2574,
+      "step": 5675
+    },
+    {
+      "epoch": 2.7120243713039844,
+      "grad_norm": 0.433698330766837,
+      "learning_rate": 5.017152729808539e-07,
+      "loss": 0.2659,
+      "step": 5676
+    },
+    {
+      "epoch": 2.7125022400095573,
+      "grad_norm": 0.4537219642809981,
+      "learning_rate": 5.000693146942359e-07,
+      "loss": 0.2628,
+      "step": 5677
+    },
+    {
+      "epoch": 2.7129801087151306,
+      "grad_norm": 0.7128807971801155,
+      "learning_rate": 4.984259915208134e-07,
+      "loss": 0.2826,
+      "step": 5678
+    },
+    {
+      "epoch": 2.7134579774207035,
+      "grad_norm": 0.5197321996631703,
+      "learning_rate": 4.96785303916415e-07,
+      "loss": 0.2838,
+      "step": 5679
+    },
+    {
+      "epoch": 2.713935846126277,
+      "grad_norm": 0.43144572298262895,
+      "learning_rate": 4.951472523361401e-07,
+      "loss": 0.2493,
+      "step": 5680
+    },
+    {
+      "epoch": 2.71441371483185,
+      "grad_norm": 0.47874783163343665,
+      "learning_rate": 4.935118372343561e-07,
+      "loss": 0.2728,
+      "step": 5681
+    },
+    {
+      "epoch": 2.714891583537423,
+      "grad_norm": 0.4506171864407369,
+      "learning_rate": 4.918790590646938e-07,
+      "loss": 0.24,
+      "step": 5682
+    },
+    {
+      "epoch": 2.7153694522429963,
+      "grad_norm": 0.4365104968626452,
+      "learning_rate": 4.90248918280063e-07,
+      "loss": 0.2807,
+      "step": 5683
+    },
+    {
+      "epoch": 2.715847320948569,
+      "grad_norm": 0.43550955708683503,
+      "learning_rate": 4.88621415332633e-07,
+      "loss": 0.2549,
+      "step": 5684
+    },
+    {
+      "epoch": 2.7163251896541425,
+      "grad_norm": 0.43640296493089487,
+      "learning_rate": 4.869965506738416e-07,
+      "loss": 0.2524,
+      "step": 5685
+    },
+    {
+      "epoch": 2.716803058359716,
+      "grad_norm": 0.466624117863505,
+      "learning_rate": 4.85374324754404e-07,
+      "loss": 0.2825,
+      "step": 5686
+    },
+    {
+      "epoch": 2.7172809270652887,
+      "grad_norm": 0.4280281243864085,
+      "learning_rate": 4.837547380242924e-07,
+      "loss": 0.2718,
+      "step": 5687
+    },
+    {
+      "epoch": 2.717758795770862,
+      "grad_norm": 0.48966203018348153,
+      "learning_rate": 4.821377909327518e-07,
+      "loss": 0.2496,
+      "step": 5688
+    },
+    {
+      "epoch": 2.718236664476435,
+      "grad_norm": 0.43749253431245166,
+      "learning_rate": 4.805234839282968e-07,
+      "loss": 0.2738,
+      "step": 5689
+    },
+    {
+      "epoch": 2.718714533182008,
+      "grad_norm": 0.457351266972444,
+      "learning_rate": 4.789118174587071e-07,
+      "loss": 0.2867,
+      "step": 5690
+    },
+    {
+      "epoch": 2.7191924018875815,
+      "grad_norm": 0.4324107322349869,
+      "learning_rate": 4.773027919710272e-07,
+      "loss": 0.2698,
+      "step": 5691
+    },
+    {
+      "epoch": 2.719670270593155,
+      "grad_norm": 0.4372967375001939,
+      "learning_rate": 4.756964079115778e-07,
+      "loss": 0.2454,
+      "step": 5692
+    },
+    {
+      "epoch": 2.7201481392987277,
+      "grad_norm": 0.44502655752254144,
+      "learning_rate": 4.740926657259393e-07,
+      "loss": 0.27,
+      "step": 5693
+    },
+    {
+      "epoch": 2.7206260080043005,
+      "grad_norm": 0.4773080512151135,
+      "learning_rate": 4.7249156585895904e-07,
+      "loss": 0.273,
+      "step": 5694
+    },
+    {
+      "epoch": 2.721103876709874,
+      "grad_norm": 0.43589564759727806,
+      "learning_rate": 4.7089310875475856e-07,
+      "loss": 0.2702,
+      "step": 5695
+    },
+    {
+      "epoch": 2.721581745415447,
+      "grad_norm": 0.435891297678135,
+      "learning_rate": 4.692972948567187e-07,
+      "loss": 0.2721,
+      "step": 5696
+    },
+    {
+      "epoch": 2.7220596141210205,
+      "grad_norm": 0.4556795601807038,
+      "learning_rate": 4.677041246074887e-07,
+      "loss": 0.268,
+      "step": 5697
+    },
+    {
+      "epoch": 2.7225374828265934,
+      "grad_norm": 0.43816241757476965,
+      "learning_rate": 4.661135984489895e-07,
+      "loss": 0.2822,
+      "step": 5698
+    },
+    {
+      "epoch": 2.7230153515321667,
+      "grad_norm": 0.45203308436452894,
+      "learning_rate": 4.645257168224038e-07,
+      "loss": 0.2646,
+      "step": 5699
+    },
+    {
+      "epoch": 2.7234932202377395,
+      "grad_norm": 0.43051792277467377,
+      "learning_rate": 4.6294048016817917e-07,
+      "loss": 0.2576,
+      "step": 5700
+    },
+    {
+      "epoch": 2.723971088943313,
+      "grad_norm": 0.4279847606367255,
+      "learning_rate": 4.6135788892603615e-07,
+      "loss": 0.2505,
+      "step": 5701
+    },
+    {
+      "epoch": 2.724448957648886,
+      "grad_norm": 0.43586951533967105,
+      "learning_rate": 4.5977794353495584e-07,
+      "loss": 0.2667,
+      "step": 5702
+    },
+    {
+      "epoch": 2.724926826354459,
+      "grad_norm": 0.4436479201943215,
+      "learning_rate": 4.582006444331866e-07,
+      "loss": 0.2584,
+      "step": 5703
+    },
+    {
+      "epoch": 2.7254046950600324,
+      "grad_norm": 0.46315434030840746,
+      "learning_rate": 4.56625992058245e-07,
+      "loss": 0.276,
+      "step": 5704
+    },
+    {
+      "epoch": 2.7258825637656052,
+      "grad_norm": 0.46920092210161374,
+      "learning_rate": 4.550539868469106e-07,
+      "loss": 0.2829,
+      "step": 5705
+    },
+    {
+      "epoch": 2.7263604324711785,
+      "grad_norm": 0.49465749907516765,
+      "learning_rate": 4.5348462923523017e-07,
+      "loss": 0.2528,
+      "step": 5706
+    },
+    {
+      "epoch": 2.726838301176752,
+      "grad_norm": 0.5070139510002986,
+      "learning_rate": 4.519179196585166e-07,
+      "loss": 0.2578,
+      "step": 5707
+    },
+    {
+      "epoch": 2.7273161698823247,
+      "grad_norm": 0.4548552543610518,
+      "learning_rate": 4.5035385855134674e-07,
+      "loss": 0.2647,
+      "step": 5708
+    },
+    {
+      "epoch": 2.727794038587898,
+      "grad_norm": 0.46825179549102325,
+      "learning_rate": 4.4879244634756125e-07,
+      "loss": 0.2892,
+      "step": 5709
+    },
+    {
+      "epoch": 2.728271907293471,
+      "grad_norm": 0.42982413711336054,
+      "learning_rate": 4.4723368348027375e-07,
+      "loss": 0.2394,
+      "step": 5710
+    },
+    {
+      "epoch": 2.7287497759990442,
+      "grad_norm": 0.5254320930661824,
+      "learning_rate": 4.4567757038185387e-07,
+      "loss": 0.2501,
+      "step": 5711
+    },
+    {
+      "epoch": 2.7292276447046175,
+      "grad_norm": 0.43831906460766556,
+      "learning_rate": 4.4412410748393973e-07,
+      "loss": 0.2721,
+      "step": 5712
+    },
+    {
+      "epoch": 2.7297055134101904,
+      "grad_norm": 0.44601508670964723,
+      "learning_rate": 4.4257329521743554e-07,
+      "loss": 0.2687,
+      "step": 5713
+    },
+    {
+      "epoch": 2.7301833821157637,
+      "grad_norm": 0.4244997363062748,
+      "learning_rate": 4.4102513401251047e-07,
+      "loss": 0.2593,
+      "step": 5714
+    },
+    {
+      "epoch": 2.7306612508213366,
+      "grad_norm": 0.43283927918633053,
+      "learning_rate": 4.394796242985933e-07,
+      "loss": 0.2397,
+      "step": 5715
+    },
+    {
+      "epoch": 2.73113911952691,
+      "grad_norm": 0.4532562420987665,
+      "learning_rate": 4.3793676650438545e-07,
+      "loss": 0.2622,
+      "step": 5716
+    },
+    {
+      "epoch": 2.7316169882324832,
+      "grad_norm": 0.4699173683543495,
+      "learning_rate": 4.363965610578469e-07,
+      "loss": 0.2618,
+      "step": 5717
+    },
+    {
+      "epoch": 2.7320948569380565,
+      "grad_norm": 0.44230687360379534,
+      "learning_rate": 4.348590083862025e-07,
+      "loss": 0.2777,
+      "step": 5718
+    },
+    {
+      "epoch": 2.7325727256436294,
+      "grad_norm": 0.42842808714224573,
+      "learning_rate": 4.3332410891594346e-07,
+      "loss": 0.2648,
+      "step": 5719
+    },
+    {
+      "epoch": 2.7330505943492027,
+      "grad_norm": 0.4448453155274851,
+      "learning_rate": 4.317918630728235e-07,
+      "loss": 0.2507,
+      "step": 5720
+    },
+    {
+      "epoch": 2.7335284630547756,
+      "grad_norm": 0.45643300317990476,
+      "learning_rate": 4.302622712818594e-07,
+      "loss": 0.2588,
+      "step": 5721
+    },
+    {
+      "epoch": 2.734006331760349,
+      "grad_norm": 0.4357425686343285,
+      "learning_rate": 4.28735333967335e-07,
+      "loss": 0.2761,
+      "step": 5722
+    },
+    {
+      "epoch": 2.734484200465922,
+      "grad_norm": 0.4304752980285673,
+      "learning_rate": 4.2721105155279496e-07,
+      "loss": 0.2722,
+      "step": 5723
+    },
+    {
+      "epoch": 2.734962069171495,
+      "grad_norm": 0.43442802885881027,
+      "learning_rate": 4.2568942446104657e-07,
+      "loss": 0.2603,
+      "step": 5724
+    },
+    {
+      "epoch": 2.7354399378770684,
+      "grad_norm": 0.47160630821995014,
+      "learning_rate": 4.241704531141633e-07,
+      "loss": 0.2688,
+      "step": 5725
+    },
+    {
+      "epoch": 2.7359178065826413,
+      "grad_norm": 0.4501959887057499,
+      "learning_rate": 4.2265413793348363e-07,
+      "loss": 0.2668,
+      "step": 5726
+    },
+    {
+      "epoch": 2.7363956752882146,
+      "grad_norm": 0.4175734384990899,
+      "learning_rate": 4.2114047933960453e-07,
+      "loss": 0.2606,
+      "step": 5727
+    },
+    {
+      "epoch": 2.736873543993788,
+      "grad_norm": 0.45694768736227065,
+      "learning_rate": 4.196294777523868e-07,
+      "loss": 0.2674,
+      "step": 5728
+    },
+    {
+      "epoch": 2.7373514126993608,
+      "grad_norm": 0.4673946349847667,
+      "learning_rate": 4.181211335909585e-07,
+      "loss": 0.279,
+      "step": 5729
+    },
+    {
+      "epoch": 2.737829281404934,
+      "grad_norm": 0.44503813090400046,
+      "learning_rate": 4.166154472737061e-07,
+      "loss": 0.2631,
+      "step": 5730
+    },
+    {
+      "epoch": 2.738307150110507,
+      "grad_norm": 0.4340503304353276,
+      "learning_rate": 4.151124192182798e-07,
+      "loss": 0.2603,
+      "step": 5731
+    },
+    {
+      "epoch": 2.7387850188160803,
+      "grad_norm": 0.48847952518429966,
+      "learning_rate": 4.136120498415952e-07,
+      "loss": 0.2624,
+      "step": 5732
+    },
+    {
+      "epoch": 2.7392628875216536,
+      "grad_norm": 0.5059660565626302,
+      "learning_rate": 4.1211433955982707e-07,
+      "loss": 0.2595,
+      "step": 5733
+    },
+    {
+      "epoch": 2.7397407562272265,
+      "grad_norm": 0.46797353640218137,
+      "learning_rate": 4.1061928878841193e-07,
+      "loss": 0.237,
+      "step": 5734
+    },
+    {
+      "epoch": 2.7402186249327998,
+      "grad_norm": 0.45219798246778975,
+      "learning_rate": 4.091268979420537e-07,
+      "loss": 0.2592,
+      "step": 5735
+    },
+    {
+      "epoch": 2.7406964936383726,
+      "grad_norm": 0.4447422926180344,
+      "learning_rate": 4.0763716743471346e-07,
+      "loss": 0.2664,
+      "step": 5736
+    },
+    {
+      "epoch": 2.741174362343946,
+      "grad_norm": 0.666905018898308,
+      "learning_rate": 4.061500976796162e-07,
+      "loss": 0.2797,
+      "step": 5737
+    },
+    {
+      "epoch": 2.7416522310495193,
+      "grad_norm": 0.4393683214525845,
+      "learning_rate": 4.0466568908925087e-07,
+      "loss": 0.2648,
+      "step": 5738
+    },
+    {
+      "epoch": 2.742130099755092,
+      "grad_norm": 0.4571492436490682,
+      "learning_rate": 4.031839420753636e-07,
+      "loss": 0.2714,
+      "step": 5739
+    },
+    {
+      "epoch": 2.7426079684606655,
+      "grad_norm": 0.4227610872002852,
+      "learning_rate": 4.0170485704896453e-07,
+      "loss": 0.2658,
+      "step": 5740
+    },
+    {
+      "epoch": 2.7430858371662383,
+      "grad_norm": 0.4436196880313655,
+      "learning_rate": 4.002284344203289e-07,
+      "loss": 0.272,
+      "step": 5741
+    },
+    {
+      "epoch": 2.7435637058718116,
+      "grad_norm": 0.5126894803406438,
+      "learning_rate": 3.987546745989879e-07,
+      "loss": 0.2627,
+      "step": 5742
+    },
+    {
+      "epoch": 2.744041574577385,
+      "grad_norm": 0.5064687140304711,
+      "learning_rate": 3.9728357799373675e-07,
+      "loss": 0.2745,
+      "step": 5743
+    },
+    {
+      "epoch": 2.7445194432829583,
+      "grad_norm": 0.5135425690883777,
+      "learning_rate": 3.958151450126324e-07,
+      "loss": 0.2371,
+      "step": 5744
+    },
+    {
+      "epoch": 2.744997311988531,
+      "grad_norm": 0.45220414835144906,
+      "learning_rate": 3.943493760629924e-07,
+      "loss": 0.2614,
+      "step": 5745
+    },
+    {
+      "epoch": 2.7454751806941045,
+      "grad_norm": 0.43240096665660527,
+      "learning_rate": 3.928862715513937e-07,
+      "loss": 0.2611,
+      "step": 5746
+    },
+    {
+      "epoch": 2.7459530493996773,
+      "grad_norm": 0.43526686209905135,
+      "learning_rate": 3.914258318836772e-07,
+      "loss": 0.2631,
+      "step": 5747
+    },
+    {
+      "epoch": 2.7464309181052506,
+      "grad_norm": 0.47983339284389653,
+      "learning_rate": 3.8996805746494336e-07,
+      "loss": 0.2501,
+      "step": 5748
+    },
+    {
+      "epoch": 2.746908786810824,
+      "grad_norm": 0.48766854570737883,
+      "learning_rate": 3.885129486995498e-07,
+      "loss": 0.2514,
+      "step": 5749
+    },
+    {
+      "epoch": 2.747386655516397,
+      "grad_norm": 0.4342714380611848,
+      "learning_rate": 3.8706050599112363e-07,
+      "loss": 0.2641,
+      "step": 5750
+    },
+    {
+      "epoch": 2.74786452422197,
+      "grad_norm": 0.4493794754550567,
+      "learning_rate": 3.8561072974254267e-07,
+      "loss": 0.2584,
+      "step": 5751
+    },
+    {
+      "epoch": 2.748342392927543,
+      "grad_norm": 0.5181439015969199,
+      "learning_rate": 3.8416362035594847e-07,
+      "loss": 0.2698,
+      "step": 5752
+    },
+    {
+      "epoch": 2.7488202616331163,
+      "grad_norm": 0.6154335078519316,
+      "learning_rate": 3.827191782327477e-07,
+      "loss": 0.2791,
+      "step": 5753
+    },
+    {
+      "epoch": 2.7492981303386896,
+      "grad_norm": 0.43729212520698946,
+      "learning_rate": 3.812774037736011e-07,
+      "loss": 0.2656,
+      "step": 5754
+    },
+    {
+      "epoch": 2.7497759990442625,
+      "grad_norm": 0.45813669988799843,
+      "learning_rate": 3.798382973784298e-07,
+      "loss": 0.2571,
+      "step": 5755
+    },
+    {
+      "epoch": 2.750253867749836,
+      "grad_norm": 0.4464533961762353,
+      "learning_rate": 3.7840185944641894e-07,
+      "loss": 0.2612,
+      "step": 5756
+    },
+    {
+      "epoch": 2.7507317364554087,
+      "grad_norm": 0.4320135449118599,
+      "learning_rate": 3.769680903760109e-07,
+      "loss": 0.2753,
+      "step": 5757
+    },
+    {
+      "epoch": 2.751209605160982,
+      "grad_norm": 0.4585589926522614,
+      "learning_rate": 3.7553699056490536e-07,
+      "loss": 0.2537,
+      "step": 5758
+    },
+    {
+      "epoch": 2.7516874738665553,
+      "grad_norm": 1.3126054515046628,
+      "learning_rate": 3.7410856041006694e-07,
+      "loss": 0.2696,
+      "step": 5759
+    },
+    {
+      "epoch": 2.752165342572128,
+      "grad_norm": 0.48621630808612154,
+      "learning_rate": 3.7268280030771655e-07,
+      "loss": 0.2472,
+      "step": 5760
+    },
+    {
+      "epoch": 2.7526432112777015,
+      "grad_norm": 0.4490574992798825,
+      "learning_rate": 3.712597106533344e-07,
+      "loss": 0.2572,
+      "step": 5761
+    },
+    {
+      "epoch": 2.7531210799832744,
+      "grad_norm": 0.4203205745967875,
+      "learning_rate": 3.698392918416593e-07,
+      "loss": 0.2539,
+      "step": 5762
+    },
+    {
+      "epoch": 2.7535989486888477,
+      "grad_norm": 0.4656550654291942,
+      "learning_rate": 3.684215442666927e-07,
+      "loss": 0.2693,
+      "step": 5763
+    },
+    {
+      "epoch": 2.754076817394421,
+      "grad_norm": 0.45928373313115284,
+      "learning_rate": 3.670064683216912e-07,
+      "loss": 0.2512,
+      "step": 5764
+    },
+    {
+      "epoch": 2.754554686099994,
+      "grad_norm": 0.45661432116739664,
+      "learning_rate": 3.655940643991718e-07,
+      "loss": 0.2477,
+      "step": 5765
+    },
+    {
+      "epoch": 2.755032554805567,
+      "grad_norm": 0.4323656387442569,
+      "learning_rate": 3.641843328909123e-07,
+      "loss": 0.2485,
+      "step": 5766
+    },
+    {
+      "epoch": 2.75551042351114,
+      "grad_norm": 0.4525029456321567,
+      "learning_rate": 3.6277727418794537e-07,
+      "loss": 0.2465,
+      "step": 5767
+    },
+    {
+      "epoch": 2.7559882922167134,
+      "grad_norm": 0.45847229829882746,
+      "learning_rate": 3.613728886805634e-07,
+      "loss": 0.2678,
+      "step": 5768
+    },
+    {
+      "epoch": 2.7564661609222867,
+      "grad_norm": 0.4354223231831446,
+      "learning_rate": 3.599711767583214e-07,
+      "loss": 0.2592,
+      "step": 5769
+    },
+    {
+      "epoch": 2.75694402962786,
+      "grad_norm": 0.45784981558251925,
+      "learning_rate": 3.585721388100283e-07,
+      "loss": 0.2722,
+      "step": 5770
+    },
+    {
+      "epoch": 2.757421898333433,
+      "grad_norm": 0.7302615178327121,
+      "learning_rate": 3.5717577522375037e-07,
+      "loss": 0.2564,
+      "step": 5771
+    },
+    {
+      "epoch": 2.757899767039006,
+      "grad_norm": 0.43796173111069236,
+      "learning_rate": 3.557820863868167e-07,
+      "loss": 0.2622,
+      "step": 5772
+    },
+    {
+      "epoch": 2.758377635744579,
+      "grad_norm": 0.462073755028673,
+      "learning_rate": 3.543910726858113e-07,
+      "loss": 0.2728,
+      "step": 5773
+    },
+    {
+      "epoch": 2.7588555044501524,
+      "grad_norm": 0.5233980811044266,
+      "learning_rate": 3.5300273450657564e-07,
+      "loss": 0.2708,
+      "step": 5774
+    },
+    {
+      "epoch": 2.7593333731557257,
+      "grad_norm": 0.5993407016424378,
+      "learning_rate": 3.516170722342127e-07,
+      "loss": 0.2667,
+      "step": 5775
+    },
+    {
+      "epoch": 2.7598112418612986,
+      "grad_norm": 0.425632476030798,
+      "learning_rate": 3.5023408625307844e-07,
+      "loss": 0.2765,
+      "step": 5776
+    },
+    {
+      "epoch": 2.760289110566872,
+      "grad_norm": 0.4264176424109537,
+      "learning_rate": 3.488537769467892e-07,
+      "loss": 0.2528,
+      "step": 5777
+    },
+    {
+      "epoch": 2.7607669792724447,
+      "grad_norm": 0.4322253085066871,
+      "learning_rate": 3.4747614469822e-07,
+      "loss": 0.271,
+      "step": 5778
+    },
+    {
+      "epoch": 2.761244847978018,
+      "grad_norm": 0.44062241649420975,
+      "learning_rate": 3.461011898895017e-07,
+      "loss": 0.2597,
+      "step": 5779
+    },
+    {
+      "epoch": 2.7617227166835914,
+      "grad_norm": 0.4288740826814237,
+      "learning_rate": 3.4472891290201927e-07,
+      "loss": 0.2659,
+      "step": 5780
+    },
+    {
+      "epoch": 2.7622005853891642,
+      "grad_norm": 0.441917086276448,
+      "learning_rate": 3.4335931411642153e-07,
+      "loss": 0.2716,
+      "step": 5781
+    },
+    {
+      "epoch": 2.7626784540947376,
+      "grad_norm": 0.7227881237165457,
+      "learning_rate": 3.419923939126102e-07,
+      "loss": 0.2642,
+      "step": 5782
+    },
+    {
+      "epoch": 2.7631563228003104,
+      "grad_norm": 0.4262654766844021,
+      "learning_rate": 3.4062815266974304e-07,
+      "loss": 0.2533,
+      "step": 5783
+    },
+    {
+      "epoch": 2.7636341915058837,
+      "grad_norm": 0.44298432828730744,
+      "learning_rate": 3.3926659076623846e-07,
+      "loss": 0.278,
+      "step": 5784
+    },
+    {
+      "epoch": 2.764112060211457,
+      "grad_norm": 0.43947410330039516,
+      "learning_rate": 3.3790770857976995e-07,
+      "loss": 0.2623,
+      "step": 5785
+    },
+    {
+      "epoch": 2.76458992891703,
+      "grad_norm": 0.42339510096523475,
+      "learning_rate": 3.3655150648726485e-07,
+      "loss": 0.279,
+      "step": 5786
+    },
+    {
+      "epoch": 2.7650677976226032,
+      "grad_norm": 0.45899672855736107,
+      "learning_rate": 3.351979848649134e-07,
+      "loss": 0.257,
+      "step": 5787
+    },
+    {
+      "epoch": 2.765545666328176,
+      "grad_norm": 0.45322359078395513,
+      "learning_rate": 3.3384714408815745e-07,
+      "loss": 0.2552,
+      "step": 5788
+    },
+    {
+      "epoch": 2.7660235350337494,
+      "grad_norm": 0.42446680964479644,
+      "learning_rate": 3.324989845316928e-07,
+      "loss": 0.2699,
+      "step": 5789
+    },
+    {
+      "epoch": 2.7665014037393227,
+      "grad_norm": 0.42230616756573375,
+      "learning_rate": 3.3115350656948043e-07,
+      "loss": 0.2662,
+      "step": 5790
+    },
+    {
+      "epoch": 2.7669792724448956,
+      "grad_norm": 0.5214966776605962,
+      "learning_rate": 3.298107105747295e-07,
+      "loss": 0.2722,
+      "step": 5791
+    },
+    {
+      "epoch": 2.767457141150469,
+      "grad_norm": 0.4648415035471319,
+      "learning_rate": 3.2847059691990644e-07,
+      "loss": 0.2555,
+      "step": 5792
+    },
+    {
+      "epoch": 2.767935009856042,
+      "grad_norm": 0.4484544390935882,
+      "learning_rate": 3.271331659767385e-07,
+      "loss": 0.2464,
+      "step": 5793
+    },
+    {
+      "epoch": 2.768412878561615,
+      "grad_norm": 0.5447610391648413,
+      "learning_rate": 3.257984181162044e-07,
+      "loss": 0.2698,
+      "step": 5794
+    },
+    {
+      "epoch": 2.7688907472671884,
+      "grad_norm": 0.4602729737342174,
+      "learning_rate": 3.2446635370853686e-07,
+      "loss": 0.2563,
+      "step": 5795
+    },
+    {
+      "epoch": 2.7693686159727617,
+      "grad_norm": 0.4937625519242985,
+      "learning_rate": 3.2313697312323143e-07,
+      "loss": 0.2561,
+      "step": 5796
+    },
+    {
+      "epoch": 2.7698464846783346,
+      "grad_norm": 0.47243649147806527,
+      "learning_rate": 3.218102767290332e-07,
+      "loss": 0.2559,
+      "step": 5797
+    },
+    {
+      "epoch": 2.770324353383908,
+      "grad_norm": 0.5039359434080682,
+      "learning_rate": 3.204862648939422e-07,
+      "loss": 0.2622,
+      "step": 5798
+    },
+    {
+      "epoch": 2.770802222089481,
+      "grad_norm": 0.438956565568596,
+      "learning_rate": 3.19164937985218e-07,
+      "loss": 0.2659,
+      "step": 5799
+    },
+    {
+      "epoch": 2.771280090795054,
+      "grad_norm": 0.4635475451765875,
+      "learning_rate": 3.1784629636937404e-07,
+      "loss": 0.2584,
+      "step": 5800
+    },
+    {
+      "epoch": 2.7717579595006274,
+      "grad_norm": 0.4439760668552996,
+      "learning_rate": 3.1653034041217555e-07,
+      "loss": 0.2537,
+      "step": 5801
+    },
+    {
+      "epoch": 2.7722358282062003,
+      "grad_norm": 0.42748355624347,
+      "learning_rate": 3.1521707047864836e-07,
+      "loss": 0.266,
+      "step": 5802
+    },
+    {
+      "epoch": 2.7727136969117736,
+      "grad_norm": 0.45627260236223416,
+      "learning_rate": 3.139064869330699e-07,
+      "loss": 0.2684,
+      "step": 5803
+    },
+    {
+      "epoch": 2.7731915656173465,
+      "grad_norm": 0.45339200404546826,
+      "learning_rate": 3.125985901389694e-07,
+      "loss": 0.2669,
+      "step": 5804
+    },
+    {
+      "epoch": 2.77366943432292,
+      "grad_norm": 0.45868520533855495,
+      "learning_rate": 3.1129338045914004e-07,
+      "loss": 0.2461,
+      "step": 5805
+    },
+    {
+      "epoch": 2.774147303028493,
+      "grad_norm": 0.4882764629386277,
+      "learning_rate": 3.099908582556199e-07,
+      "loss": 0.2725,
+      "step": 5806
+    },
+    {
+      "epoch": 2.774625171734066,
+      "grad_norm": 0.46500740654242584,
+      "learning_rate": 3.0869102388970673e-07,
+      "loss": 0.2506,
+      "step": 5807
+    },
+    {
+      "epoch": 2.7751030404396393,
+      "grad_norm": 0.452079842828752,
+      "learning_rate": 3.0739387772195205e-07,
+      "loss": 0.2631,
+      "step": 5808
+    },
+    {
+      "epoch": 2.775580909145212,
+      "grad_norm": 0.4381477199767286,
+      "learning_rate": 3.0609942011216144e-07,
+      "loss": 0.2754,
+      "step": 5809
+    },
+    {
+      "epoch": 2.7760587778507855,
+      "grad_norm": 0.4408816665348534,
+      "learning_rate": 3.0480765141939316e-07,
+      "loss": 0.2743,
+      "step": 5810
+    },
+    {
+      "epoch": 2.776536646556359,
+      "grad_norm": 0.46086546867548916,
+      "learning_rate": 3.035185720019629e-07,
+      "loss": 0.2704,
+      "step": 5811
+    },
+    {
+      "epoch": 2.7770145152619317,
+      "grad_norm": 0.44332016111656897,
+      "learning_rate": 3.022321822174379e-07,
+      "loss": 0.2682,
+      "step": 5812
+    },
+    {
+      "epoch": 2.777492383967505,
+      "grad_norm": 0.4492380713862776,
+      "learning_rate": 3.0094848242263943e-07,
+      "loss": 0.2673,
+      "step": 5813
+    },
+    {
+      "epoch": 2.777970252673078,
+      "grad_norm": 0.43400035663218267,
+      "learning_rate": 2.9966747297364375e-07,
+      "loss": 0.261,
+      "step": 5814
+    },
+    {
+      "epoch": 2.778448121378651,
+      "grad_norm": 0.44189833048994365,
+      "learning_rate": 2.9838915422578e-07,
+      "loss": 0.2563,
+      "step": 5815
+    },
+    {
+      "epoch": 2.7789259900842245,
+      "grad_norm": 0.4644677539843851,
+      "learning_rate": 2.9711352653363115e-07,
+      "loss": 0.2566,
+      "step": 5816
+    },
+    {
+      "epoch": 2.7794038587897973,
+      "grad_norm": 0.4475255303780797,
+      "learning_rate": 2.9584059025103415e-07,
+      "loss": 0.2636,
+      "step": 5817
+    },
+    {
+      "epoch": 2.7798817274953707,
+      "grad_norm": 0.443274368634649,
+      "learning_rate": 2.9457034573108e-07,
+      "loss": 0.2544,
+      "step": 5818
+    },
+    {
+      "epoch": 2.7803595962009435,
+      "grad_norm": 0.4450610430568845,
+      "learning_rate": 2.933027933261101e-07,
+      "loss": 0.2714,
+      "step": 5819
+    },
+    {
+      "epoch": 2.780837464906517,
+      "grad_norm": 0.42837406809925016,
+      "learning_rate": 2.920379333877221e-07,
+      "loss": 0.2525,
+      "step": 5820
+    },
+    {
+      "epoch": 2.78131533361209,
+      "grad_norm": 0.45240402207297176,
+      "learning_rate": 2.907757662667665e-07,
+      "loss": 0.2589,
+      "step": 5821
+    },
+    {
+      "epoch": 2.7817932023176635,
+      "grad_norm": 0.46379312040070836,
+      "learning_rate": 2.8951629231334434e-07,
+      "loss": 0.2472,
+      "step": 5822
+    },
+    {
+      "epoch": 2.7822710710232363,
+      "grad_norm": 0.4412306348002121,
+      "learning_rate": 2.8825951187681387e-07,
+      "loss": 0.276,
+      "step": 5823
+    },
+    {
+      "epoch": 2.7827489397288097,
+      "grad_norm": 0.4603946932137324,
+      "learning_rate": 2.87005425305783e-07,
+      "loss": 0.2722,
+      "step": 5824
+    },
+    {
+      "epoch": 2.7832268084343825,
+      "grad_norm": 0.45987250970095805,
+      "learning_rate": 2.8575403294811123e-07,
+      "loss": 0.2744,
+      "step": 5825
+    },
+    {
+      "epoch": 2.783704677139956,
+      "grad_norm": 0.43880186402447113,
+      "learning_rate": 2.845053351509142e-07,
+      "loss": 0.2384,
+      "step": 5826
+    },
+    {
+      "epoch": 2.784182545845529,
+      "grad_norm": 0.45035442317121277,
+      "learning_rate": 2.8325933226056033e-07,
+      "loss": 0.2759,
+      "step": 5827
+    },
+    {
+      "epoch": 2.784660414551102,
+      "grad_norm": 0.4669047993561521,
+      "learning_rate": 2.8201602462266775e-07,
+      "loss": 0.2575,
+      "step": 5828
+    },
+    {
+      "epoch": 2.7851382832566753,
+      "grad_norm": 0.5303829700742767,
+      "learning_rate": 2.8077541258210607e-07,
+      "loss": 0.2591,
+      "step": 5829
+    },
+    {
+      "epoch": 2.785616151962248,
+      "grad_norm": 0.6146920950572496,
+      "learning_rate": 2.795374964830022e-07,
+      "loss": 0.2634,
+      "step": 5830
+    },
+    {
+      "epoch": 2.7860940206678215,
+      "grad_norm": 0.4537712641615584,
+      "learning_rate": 2.7830227666872933e-07,
+      "loss": 0.2754,
+      "step": 5831
+    },
+    {
+      "epoch": 2.786571889373395,
+      "grad_norm": 0.4379198918938658,
+      "learning_rate": 2.770697534819178e-07,
+      "loss": 0.2787,
+      "step": 5832
+    },
+    {
+      "epoch": 2.7870497580789677,
+      "grad_norm": 0.5202726893687436,
+      "learning_rate": 2.758399272644474e-07,
+      "loss": 0.284,
+      "step": 5833
+    },
+    {
+      "epoch": 2.787527626784541,
+      "grad_norm": 0.45661668436412567,
+      "learning_rate": 2.746127983574498e-07,
+      "loss": 0.2649,
+      "step": 5834
+    },
+    {
+      "epoch": 2.788005495490114,
+      "grad_norm": 0.4649814028316174,
+      "learning_rate": 2.733883671013082e-07,
+      "loss": 0.2505,
+      "step": 5835
+    },
+    {
+      "epoch": 2.788483364195687,
+      "grad_norm": 0.5661711250377324,
+      "learning_rate": 2.721666338356599e-07,
+      "loss": 0.2538,
+      "step": 5836
+    },
+    {
+      "epoch": 2.7889612329012605,
+      "grad_norm": 0.4777341668360208,
+      "learning_rate": 2.709475988993915e-07,
+      "loss": 0.2666,
+      "step": 5837
+    },
+    {
+      "epoch": 2.7894391016068334,
+      "grad_norm": 0.47234324125709043,
+      "learning_rate": 2.6973126263064143e-07,
+      "loss": 0.2677,
+      "step": 5838
+    },
+    {
+      "epoch": 2.7899169703124067,
+      "grad_norm": 0.4313068768665777,
+      "learning_rate": 2.685176253667998e-07,
+      "loss": 0.2648,
+      "step": 5839
+    },
+    {
+      "epoch": 2.7903948390179796,
+      "grad_norm": 0.44321493365822373,
+      "learning_rate": 2.673066874445096e-07,
+      "loss": 0.2479,
+      "step": 5840
+    },
+    {
+      "epoch": 2.790872707723553,
+      "grad_norm": 0.47299846939177753,
+      "learning_rate": 2.66098449199661e-07,
+      "loss": 0.256,
+      "step": 5841
+    },
+    {
+      "epoch": 2.791350576429126,
+      "grad_norm": 0.44183881798639174,
+      "learning_rate": 2.648929109674003e-07,
+      "loss": 0.2559,
+      "step": 5842
+    },
+    {
+      "epoch": 2.791828445134699,
+      "grad_norm": 0.48856394471690884,
+      "learning_rate": 2.6369007308212233e-07,
+      "loss": 0.2528,
+      "step": 5843
+    },
+    {
+      "epoch": 2.7923063138402724,
+      "grad_norm": 0.42875317220473846,
+      "learning_rate": 2.6248993587747017e-07,
+      "loss": 0.2632,
+      "step": 5844
+    },
+    {
+      "epoch": 2.7927841825458453,
+      "grad_norm": 0.439006550935603,
+      "learning_rate": 2.612924996863453e-07,
+      "loss": 0.275,
+      "step": 5845
+    },
+    {
+      "epoch": 2.7932620512514186,
+      "grad_norm": 0.502986151150374,
+      "learning_rate": 2.600977648408931e-07,
+      "loss": 0.2728,
+      "step": 5846
+    },
+    {
+      "epoch": 2.793739919956992,
+      "grad_norm": 0.502221402752791,
+      "learning_rate": 2.5890573167251076e-07,
+      "loss": 0.2759,
+      "step": 5847
+    },
+    {
+      "epoch": 2.794217788662565,
+      "grad_norm": 0.46120467386193476,
+      "learning_rate": 2.5771640051184933e-07,
+      "loss": 0.2901,
+      "step": 5848
+    },
+    {
+      "epoch": 2.794695657368138,
+      "grad_norm": 0.46525631074461465,
+      "learning_rate": 2.565297716888082e-07,
+      "loss": 0.2537,
+      "step": 5849
+    },
+    {
+      "epoch": 2.7951735260737114,
+      "grad_norm": 0.5367360951991326,
+      "learning_rate": 2.5534584553253526e-07,
+      "loss": 0.2737,
+      "step": 5850
+    },
+    {
+      "epoch": 2.7956513947792843,
+      "grad_norm": 0.477928653419597,
+      "learning_rate": 2.5416462237143224e-07,
+      "loss": 0.2723,
+      "step": 5851
+    },
+    {
+      "epoch": 2.7961292634848576,
+      "grad_norm": 0.43779540429422825,
+      "learning_rate": 2.5298610253315037e-07,
+      "loss": 0.2492,
+      "step": 5852
+    },
+    {
+      "epoch": 2.796607132190431,
+      "grad_norm": 0.4473417651730253,
+      "learning_rate": 2.5181028634458704e-07,
+      "loss": 0.2591,
+      "step": 5853
+    },
+    {
+      "epoch": 2.7970850008960038,
+      "grad_norm": 0.4237788400316185,
+      "learning_rate": 2.5063717413189695e-07,
+      "loss": 0.2713,
+      "step": 5854
+    },
+    {
+      "epoch": 2.797562869601577,
+      "grad_norm": 0.46121222455074623,
+      "learning_rate": 2.494667662204797e-07,
+      "loss": 0.2587,
+      "step": 5855
+    },
+    {
+      "epoch": 2.79804073830715,
+      "grad_norm": 0.44337477946493636,
+      "learning_rate": 2.482990629349824e-07,
+      "loss": 0.277,
+      "step": 5856
+    },
+    {
+      "epoch": 2.7985186070127233,
+      "grad_norm": 0.4594080279967684,
+      "learning_rate": 2.471340645993103e-07,
+      "loss": 0.2668,
+      "step": 5857
+    },
+    {
+      "epoch": 2.7989964757182966,
+      "grad_norm": 0.4400931256728512,
+      "learning_rate": 2.4597177153661056e-07,
+      "loss": 0.2598,
+      "step": 5858
+    },
+    {
+      "epoch": 2.7994743444238694,
+      "grad_norm": 0.4426802092154914,
+      "learning_rate": 2.4481218406928297e-07,
+      "loss": 0.2763,
+      "step": 5859
+    },
+    {
+      "epoch": 2.7999522131294428,
+      "grad_norm": 0.4353913914636723,
+      "learning_rate": 2.436553025189758e-07,
+      "loss": 0.2573,
+      "step": 5860
+    },
+    {
+      "epoch": 2.8004300818350156,
+      "grad_norm": 0.5177432480573205,
+      "learning_rate": 2.4250112720659024e-07,
+      "loss": 0.2689,
+      "step": 5861
+    },
+    {
+      "epoch": 2.800907950540589,
+      "grad_norm": 0.5709604532387811,
+      "learning_rate": 2.413496584522723e-07,
+      "loss": 0.2526,
+      "step": 5862
+    },
+    {
+      "epoch": 2.8013858192461623,
+      "grad_norm": 0.45801417663423666,
+      "learning_rate": 2.402008965754199e-07,
+      "loss": 0.2537,
+      "step": 5863
+    },
+    {
+      "epoch": 2.801863687951735,
+      "grad_norm": 0.4707325970556673,
+      "learning_rate": 2.3905484189467807e-07,
+      "loss": 0.2759,
+      "step": 5864
+    },
+    {
+      "epoch": 2.8023415566573084,
+      "grad_norm": 0.4486900242136094,
+      "learning_rate": 2.3791149472794373e-07,
+      "loss": 0.257,
+      "step": 5865
+    },
+    {
+      "epoch": 2.8028194253628813,
+      "grad_norm": 0.45965714254275614,
+      "learning_rate": 2.3677085539235977e-07,
+      "loss": 0.2709,
+      "step": 5866
+    },
+    {
+      "epoch": 2.8032972940684546,
+      "grad_norm": 0.43566747125808025,
+      "learning_rate": 2.3563292420432094e-07,
+      "loss": 0.274,
+      "step": 5867
+    },
+    {
+      "epoch": 2.803775162774028,
+      "grad_norm": 0.4541848106577836,
+      "learning_rate": 2.3449770147946804e-07,
+      "loss": 0.2683,
+      "step": 5868
+    },
+    {
+      "epoch": 2.8042530314796013,
+      "grad_norm": 0.8078669068971982,
+      "learning_rate": 2.3336518753269144e-07,
+      "loss": 0.2577,
+      "step": 5869
+    },
+    {
+      "epoch": 2.804730900185174,
+      "grad_norm": 0.42680658838886604,
+      "learning_rate": 2.3223538267813317e-07,
+      "loss": 0.2434,
+      "step": 5870
+    },
+    {
+      "epoch": 2.805208768890747,
+      "grad_norm": 0.4358025311911329,
+      "learning_rate": 2.3110828722917812e-07,
+      "loss": 0.2625,
+      "step": 5871
+    },
+    {
+      "epoch": 2.8056866375963203,
+      "grad_norm": 0.45786374708678146,
+      "learning_rate": 2.2998390149846395e-07,
+      "loss": 0.268,
+      "step": 5872
+    },
+    {
+      "epoch": 2.8061645063018936,
+      "grad_norm": 0.44034045702230934,
+      "learning_rate": 2.2886222579787565e-07,
+      "loss": 0.2593,
+      "step": 5873
+    },
+    {
+      "epoch": 2.806642375007467,
+      "grad_norm": 0.629034529435717,
+      "learning_rate": 2.2774326043854656e-07,
+      "loss": 0.2797,
+      "step": 5874
+    },
+    {
+      "epoch": 2.80712024371304,
+      "grad_norm": 0.4663816214402025,
+      "learning_rate": 2.2662700573085505e-07,
+      "loss": 0.2589,
+      "step": 5875
+    },
+    {
+      "epoch": 2.807598112418613,
+      "grad_norm": 0.4520335805323787,
+      "learning_rate": 2.255134619844357e-07,
+      "loss": 0.2693,
+      "step": 5876
+    },
+    {
+      "epoch": 2.808075981124186,
+      "grad_norm": 0.45421012246090176,
+      "learning_rate": 2.2440262950816138e-07,
+      "loss": 0.2486,
+      "step": 5877
+    },
+    {
+      "epoch": 2.8085538498297593,
+      "grad_norm": 0.44872612791769856,
+      "learning_rate": 2.23294508610159e-07,
+      "loss": 0.2772,
+      "step": 5878
+    },
+    {
+      "epoch": 2.8090317185353326,
+      "grad_norm": 0.4471553331629049,
+      "learning_rate": 2.2218909959780265e-07,
+      "loss": 0.2622,
+      "step": 5879
+    },
+    {
+      "epoch": 2.8095095872409055,
+      "grad_norm": 0.4326277574229714,
+      "learning_rate": 2.2108640277771153e-07,
+      "loss": 0.241,
+      "step": 5880
+    },
+    {
+      "epoch": 2.809987455946479,
+      "grad_norm": 0.4447810782306284,
+      "learning_rate": 2.1998641845575542e-07,
+      "loss": 0.2596,
+      "step": 5881
+    },
+    {
+      "epoch": 2.8104653246520517,
+      "grad_norm": 0.4438607779450878,
+      "learning_rate": 2.1888914693705132e-07,
+      "loss": 0.2781,
+      "step": 5882
+    },
+    {
+      "epoch": 2.810943193357625,
+      "grad_norm": 0.4515855882029372,
+      "learning_rate": 2.1779458852596136e-07,
+      "loss": 0.26,
+      "step": 5883
+    },
+    {
+      "epoch": 2.8114210620631983,
+      "grad_norm": 0.46722738200529673,
+      "learning_rate": 2.167027435260971e-07,
+      "loss": 0.263,
+      "step": 5884
+    },
+    {
+      "epoch": 2.811898930768771,
+      "grad_norm": 0.8295117627891755,
+      "learning_rate": 2.156136122403174e-07,
+      "loss": 0.2524,
+      "step": 5885
+    },
+    {
+      "epoch": 2.8123767994743445,
+      "grad_norm": 0.4993459060161254,
+      "learning_rate": 2.1452719497072839e-07,
+      "loss": 0.2702,
+      "step": 5886
+    },
+    {
+      "epoch": 2.8128546681799174,
+      "grad_norm": 0.4414682776865743,
+      "learning_rate": 2.1344349201868232e-07,
+      "loss": 0.2771,
+      "step": 5887
+    },
+    {
+      "epoch": 2.8133325368854907,
+      "grad_norm": 0.44309446914027134,
+      "learning_rate": 2.1236250368477985e-07,
+      "loss": 0.2827,
+      "step": 5888
+    },
+    {
+      "epoch": 2.813810405591064,
+      "grad_norm": 0.4392176207768714,
+      "learning_rate": 2.1128423026886892e-07,
+      "loss": 0.2721,
+      "step": 5889
+    },
+    {
+      "epoch": 2.814288274296637,
+      "grad_norm": 0.4418095009271383,
+      "learning_rate": 2.1020867207004026e-07,
+      "loss": 0.2673,
+      "step": 5890
+    },
+    {
+      "epoch": 2.81476614300221,
+      "grad_norm": 0.4422044960477242,
+      "learning_rate": 2.0913582938663855e-07,
+      "loss": 0.2686,
+      "step": 5891
+    },
+    {
+      "epoch": 2.815244011707783,
+      "grad_norm": 0.4697009809230947,
+      "learning_rate": 2.0806570251625023e-07,
+      "loss": 0.2533,
+      "step": 5892
+    },
+    {
+      "epoch": 2.8157218804133564,
+      "grad_norm": 0.4507224779670188,
+      "learning_rate": 2.0699829175570785e-07,
+      "loss": 0.2519,
+      "step": 5893
+    },
+    {
+      "epoch": 2.8161997491189297,
+      "grad_norm": 0.4528792818717239,
+      "learning_rate": 2.0593359740109452e-07,
+      "loss": 0.2701,
+      "step": 5894
+    },
+    {
+      "epoch": 2.816677617824503,
+      "grad_norm": 0.4460463818994321,
+      "learning_rate": 2.048716197477374e-07,
+      "loss": 0.2599,
+      "step": 5895
+    },
+    {
+      "epoch": 2.817155486530076,
+      "grad_norm": 0.49958157563151245,
+      "learning_rate": 2.038123590902086e-07,
+      "loss": 0.2625,
+      "step": 5896
+    },
+    {
+      "epoch": 2.8176333552356487,
+      "grad_norm": 0.5406609588682755,
+      "learning_rate": 2.0275581572233083e-07,
+      "loss": 0.2459,
+      "step": 5897
+    },
+    {
+      "epoch": 2.818111223941222,
+      "grad_norm": 0.4618527016308968,
+      "learning_rate": 2.017019899371686e-07,
+      "loss": 0.2725,
+      "step": 5898
+    },
+    {
+      "epoch": 2.8185890926467954,
+      "grad_norm": 0.43028097781936997,
+      "learning_rate": 2.0065088202703587e-07,
+      "loss": 0.2742,
+      "step": 5899
+    },
+    {
+      "epoch": 2.8190669613523687,
+      "grad_norm": 0.4615866120409888,
+      "learning_rate": 1.996024922834905e-07,
+      "loss": 0.2488,
+      "step": 5900
+    },
+    {
+      "epoch": 2.8195448300579415,
+      "grad_norm": 0.4243814410711499,
+      "learning_rate": 1.9855682099733876e-07,
+      "loss": 0.251,
+      "step": 5901
+    },
+    {
+      "epoch": 2.820022698763515,
+      "grad_norm": 0.42845665939883043,
+      "learning_rate": 1.9751386845862864e-07,
+      "loss": 0.2627,
+      "step": 5902
+    },
+    {
+      "epoch": 2.8205005674690877,
+      "grad_norm": 0.44017495021305203,
+      "learning_rate": 1.9647363495665983e-07,
+      "loss": 0.2529,
+      "step": 5903
+    },
+    {
+      "epoch": 2.820978436174661,
+      "grad_norm": 0.4587273313715358,
+      "learning_rate": 1.9543612077997376e-07,
+      "loss": 0.2563,
+      "step": 5904
+    },
+    {
+      "epoch": 2.8214563048802344,
+      "grad_norm": 0.47195775641267157,
+      "learning_rate": 1.9440132621635687e-07,
+      "loss": 0.2654,
+      "step": 5905
+    },
+    {
+      "epoch": 2.8219341735858072,
+      "grad_norm": 0.43800418779605343,
+      "learning_rate": 1.9336925155284514e-07,
+      "loss": 0.2698,
+      "step": 5906
+    },
+    {
+      "epoch": 2.8224120422913805,
+      "grad_norm": 0.5249547147069987,
+      "learning_rate": 1.9233989707571732e-07,
+      "loss": 0.2738,
+      "step": 5907
+    },
+    {
+      "epoch": 2.8228899109969534,
+      "grad_norm": 0.4479165992631786,
+      "learning_rate": 1.9131326307049724e-07,
+      "loss": 0.2583,
+      "step": 5908
+    },
+    {
+      "epoch": 2.8233677797025267,
+      "grad_norm": 0.4367008126895965,
+      "learning_rate": 1.9028934982195602e-07,
+      "loss": 0.2483,
+      "step": 5909
+    },
+    {
+      "epoch": 2.8238456484081,
+      "grad_norm": 0.4453324046755028,
+      "learning_rate": 1.8926815761410867e-07,
+      "loss": 0.2595,
+      "step": 5910
+    },
+    {
+      "epoch": 2.824323517113673,
+      "grad_norm": 0.4515610654582882,
+      "learning_rate": 1.8824968673021525e-07,
+      "loss": 0.2633,
+      "step": 5911
+    },
+    {
+      "epoch": 2.8248013858192462,
+      "grad_norm": 0.44115993805926473,
+      "learning_rate": 1.872339374527843e-07,
+      "loss": 0.2536,
+      "step": 5912
+    },
+    {
+      "epoch": 2.825279254524819,
+      "grad_norm": 0.4228775353271241,
+      "learning_rate": 1.8622091006356368e-07,
+      "loss": 0.2481,
+      "step": 5913
+    },
+    {
+      "epoch": 2.8257571232303924,
+      "grad_norm": 0.44436167133939714,
+      "learning_rate": 1.852106048435498e-07,
+      "loss": 0.2657,
+      "step": 5914
+    },
+    {
+      "epoch": 2.8262349919359657,
+      "grad_norm": 0.4406708131002175,
+      "learning_rate": 1.8420302207298623e-07,
+      "loss": 0.2604,
+      "step": 5915
+    },
+    {
+      "epoch": 2.8267128606415386,
+      "grad_norm": 0.4654540147793532,
+      "learning_rate": 1.831981620313561e-07,
+      "loss": 0.2753,
+      "step": 5916
+    },
+    {
+      "epoch": 2.827190729347112,
+      "grad_norm": 0.4371867269522205,
+      "learning_rate": 1.8219602499738863e-07,
+      "loss": 0.2584,
+      "step": 5917
+    },
+    {
+      "epoch": 2.827668598052685,
+      "grad_norm": 0.4341168805289483,
+      "learning_rate": 1.8119661124906262e-07,
+      "loss": 0.2697,
+      "step": 5918
+    },
+    {
+      "epoch": 2.828146466758258,
+      "grad_norm": 0.5332475340009322,
+      "learning_rate": 1.801999210635952e-07,
+      "loss": 0.2544,
+      "step": 5919
+    },
+    {
+      "epoch": 2.8286243354638314,
+      "grad_norm": 0.4415760206731807,
+      "learning_rate": 1.792059547174507e-07,
+      "loss": 0.2608,
+      "step": 5920
+    },
+    {
+      "epoch": 2.8291022041694047,
+      "grad_norm": 0.4539316651950751,
+      "learning_rate": 1.7821471248633982e-07,
+      "loss": 0.2545,
+      "step": 5921
+    },
+    {
+      "epoch": 2.8295800728749776,
+      "grad_norm": 0.4269206432587368,
+      "learning_rate": 1.7722619464521363e-07,
+      "loss": 0.254,
+      "step": 5922
+    },
+    {
+      "epoch": 2.830057941580551,
+      "grad_norm": 0.43673477313105225,
+      "learning_rate": 1.762404014682706e-07,
+      "loss": 0.2667,
+      "step": 5923
+    },
+    {
+      "epoch": 2.8305358102861238,
+      "grad_norm": 0.4414874551970224,
+      "learning_rate": 1.75257333228952e-07,
+      "loss": 0.2696,
+      "step": 5924
+    },
+    {
+      "epoch": 2.831013678991697,
+      "grad_norm": 0.4497054490741176,
+      "learning_rate": 1.7427699019994415e-07,
+      "loss": 0.2676,
+      "step": 5925
+    },
+    {
+      "epoch": 2.8314915476972704,
+      "grad_norm": 0.4908756734479855,
+      "learning_rate": 1.7329937265317508e-07,
+      "loss": 0.2771,
+      "step": 5926
+    },
+    {
+      "epoch": 2.8319694164028433,
+      "grad_norm": 0.4454065328919432,
+      "learning_rate": 1.7232448085982012e-07,
+      "loss": 0.2597,
+      "step": 5927
+    },
+    {
+      "epoch": 2.8324472851084166,
+      "grad_norm": 0.42499929417919846,
+      "learning_rate": 1.713523150902985e-07,
+      "loss": 0.2689,
+      "step": 5928
+    },
+    {
+      "epoch": 2.8329251538139895,
+      "grad_norm": 0.48284867449405694,
+      "learning_rate": 1.7038287561426892e-07,
+      "loss": 0.2686,
+      "step": 5929
+    },
+    {
+      "epoch": 2.8334030225195628,
+      "grad_norm": 0.44411974690243017,
+      "learning_rate": 1.6941616270063854e-07,
+      "loss": 0.2877,
+      "step": 5930
+    },
+    {
+      "epoch": 2.833880891225136,
+      "grad_norm": 0.442865545746602,
+      "learning_rate": 1.684521766175562e-07,
+      "loss": 0.2575,
+      "step": 5931
+    },
+    {
+      "epoch": 2.834358759930709,
+      "grad_norm": 0.5103520942204934,
+      "learning_rate": 1.6749091763241464e-07,
+      "loss": 0.2606,
+      "step": 5932
+    },
+    {
+      "epoch": 2.8348366286362823,
+      "grad_norm": 0.447650390063257,
+      "learning_rate": 1.665323860118495e-07,
+      "loss": 0.2705,
+      "step": 5933
+    },
+    {
+      "epoch": 2.835314497341855,
+      "grad_norm": 0.44565808230621423,
+      "learning_rate": 1.6557658202174254e-07,
+      "loss": 0.2703,
+      "step": 5934
+    },
+    {
+      "epoch": 2.8357923660474285,
+      "grad_norm": 0.43861667920681485,
+      "learning_rate": 1.6462350592721498e-07,
+      "loss": 0.2447,
+      "step": 5935
+    },
+    {
+      "epoch": 2.8362702347530018,
+      "grad_norm": 0.4633523191534991,
+      "learning_rate": 1.6367315799263206e-07,
+      "loss": 0.2713,
+      "step": 5936
+    },
+    {
+      "epoch": 2.8367481034585746,
+      "grad_norm": 0.4531309552740638,
+      "learning_rate": 1.6272553848160733e-07,
+      "loss": 0.2649,
+      "step": 5937
+    },
+    {
+      "epoch": 2.837225972164148,
+      "grad_norm": 0.4621397423597151,
+      "learning_rate": 1.6178064765699052e-07,
+      "loss": 0.2752,
+      "step": 5938
+    },
+    {
+      "epoch": 2.837703840869721,
+      "grad_norm": 0.4415497248019033,
+      "learning_rate": 1.6083848578087868e-07,
+      "loss": 0.2527,
+      "step": 5939
+    },
+    {
+      "epoch": 2.838181709575294,
+      "grad_norm": 0.44367203604390565,
+      "learning_rate": 1.5989905311461274e-07,
+      "loss": 0.2677,
+      "step": 5940
+    },
+    {
+      "epoch": 2.8386595782808675,
+      "grad_norm": 0.5809887272446366,
+      "learning_rate": 1.5896234991877202e-07,
+      "loss": 0.2713,
+      "step": 5941
+    },
+    {
+      "epoch": 2.8391374469864403,
+      "grad_norm": 0.4303037195655415,
+      "learning_rate": 1.5802837645318203e-07,
+      "loss": 0.2455,
+      "step": 5942
+    },
+    {
+      "epoch": 2.8396153156920136,
+      "grad_norm": 0.47424596948568043,
+      "learning_rate": 1.570971329769111e-07,
+      "loss": 0.2524,
+      "step": 5943
+    },
+    {
+      "epoch": 2.8400931843975865,
+      "grad_norm": 0.4178692383364968,
+      "learning_rate": 1.5616861974827036e-07,
+      "loss": 0.2526,
+      "step": 5944
+    },
+    {
+      "epoch": 2.84057105310316,
+      "grad_norm": 0.46166989830773797,
+      "learning_rate": 1.5524283702481158e-07,
+      "loss": 0.271,
+      "step": 5945
+    },
+    {
+      "epoch": 2.841048921808733,
+      "grad_norm": 0.4440783366258334,
+      "learning_rate": 1.5431978506333155e-07,
+      "loss": 0.2633,
+      "step": 5946
+    },
+    {
+      "epoch": 2.8415267905143065,
+      "grad_norm": 0.441365885091842,
+      "learning_rate": 1.5339946411986885e-07,
+      "loss": 0.2528,
+      "step": 5947
+    },
+    {
+      "epoch": 2.8420046592198793,
+      "grad_norm": 0.4737324334331007,
+      "learning_rate": 1.5248187444970252e-07,
+      "loss": 0.2599,
+      "step": 5948
+    },
+    {
+      "epoch": 2.8424825279254526,
+      "grad_norm": 0.48811117143680444,
+      "learning_rate": 1.5156701630735792e-07,
+      "loss": 0.2453,
+      "step": 5949
+    },
+    {
+      "epoch": 2.8429603966310255,
+      "grad_norm": 0.43188605131922725,
+      "learning_rate": 1.5065488994659983e-07,
+      "loss": 0.2669,
+      "step": 5950
+    },
+    {
+      "epoch": 2.843438265336599,
+      "grad_norm": 0.434588565968877,
+      "learning_rate": 1.497454956204347e-07,
+      "loss": 0.2621,
+      "step": 5951
+    },
+    {
+      "epoch": 2.843916134042172,
+      "grad_norm": 0.45504777884303677,
+      "learning_rate": 1.4883883358111418e-07,
+      "loss": 0.2599,
+      "step": 5952
+    },
+    {
+      "epoch": 2.844394002747745,
+      "grad_norm": 0.43178757210896046,
+      "learning_rate": 1.4793490408013033e-07,
+      "loss": 0.2617,
+      "step": 5953
+    },
+    {
+      "epoch": 2.8448718714533183,
+      "grad_norm": 0.45654767138484836,
+      "learning_rate": 1.4703370736821487e-07,
+      "loss": 0.268,
+      "step": 5954
+    },
+    {
+      "epoch": 2.845349740158891,
+      "grad_norm": 0.48119573262488585,
+      "learning_rate": 1.461352436953478e-07,
+      "loss": 0.2622,
+      "step": 5955
+    },
+    {
+      "epoch": 2.8458276088644645,
+      "grad_norm": 0.4351617846746113,
+      "learning_rate": 1.4523951331074426e-07,
+      "loss": 0.2601,
+      "step": 5956
+    },
+    {
+      "epoch": 2.846305477570038,
+      "grad_norm": 0.45633355644087814,
+      "learning_rate": 1.4434651646286325e-07,
+      "loss": 0.2653,
+      "step": 5957
+    },
+    {
+      "epoch": 2.8467833462756107,
+      "grad_norm": 0.4736965437562393,
+      "learning_rate": 1.4345625339940994e-07,
+      "loss": 0.2781,
+      "step": 5958
+    },
+    {
+      "epoch": 2.847261214981184,
+      "grad_norm": 0.46355248896208456,
+      "learning_rate": 1.4256872436732461e-07,
+      "loss": 0.2591,
+      "step": 5959
+    },
+    {
+      "epoch": 2.847739083686757,
+      "grad_norm": 0.4283329171153488,
+      "learning_rate": 1.4168392961279254e-07,
+      "loss": 0.256,
+      "step": 5960
+    },
+    {
+      "epoch": 2.84821695239233,
+      "grad_norm": 0.46933751036840643,
+      "learning_rate": 1.4080186938124074e-07,
+      "loss": 0.264,
+      "step": 5961
+    },
+    {
+      "epoch": 2.8486948210979035,
+      "grad_norm": 0.4421456212297339,
+      "learning_rate": 1.3992254391733794e-07,
+      "loss": 0.2435,
+      "step": 5962
+    },
+    {
+      "epoch": 2.8491726898034764,
+      "grad_norm": 0.43361352767650047,
+      "learning_rate": 1.390459534649924e-07,
+      "loss": 0.2628,
+      "step": 5963
+    },
+    {
+      "epoch": 2.8496505585090497,
+      "grad_norm": 0.43095231476269297,
+      "learning_rate": 1.38172098267354e-07,
+      "loss": 0.2607,
+      "step": 5964
+    },
+    {
+      "epoch": 2.8501284272146226,
+      "grad_norm": 0.4703203261155736,
+      "learning_rate": 1.3730097856681668e-07,
+      "loss": 0.2504,
+      "step": 5965
+    },
+    {
+      "epoch": 2.850606295920196,
+      "grad_norm": 0.4376878982675997,
+      "learning_rate": 1.364325946050138e-07,
+      "loss": 0.2799,
+      "step": 5966
+    },
+    {
+      "epoch": 2.851084164625769,
+      "grad_norm": 0.4391097408323728,
+      "learning_rate": 1.355669466228171e-07,
+      "loss": 0.2491,
+      "step": 5967
+    },
+    {
+      "epoch": 2.851562033331342,
+      "grad_norm": 0.5148238033140623,
+      "learning_rate": 1.3470403486034566e-07,
+      "loss": 0.268,
+      "step": 5968
+    },
+    {
+      "epoch": 2.8520399020369154,
+      "grad_norm": 0.44170548346421123,
+      "learning_rate": 1.3384385955695355e-07,
+      "loss": 0.2603,
+      "step": 5969
+    },
+    {
+      "epoch": 2.8525177707424882,
+      "grad_norm": 0.46229977499534447,
+      "learning_rate": 1.329864209512377e-07,
+      "loss": 0.2643,
+      "step": 5970
+    },
+    {
+      "epoch": 2.8529956394480616,
+      "grad_norm": 0.4703215322101558,
+      "learning_rate": 1.3213171928103785e-07,
+      "loss": 0.2613,
+      "step": 5971
+    },
+    {
+      "epoch": 2.853473508153635,
+      "grad_norm": 0.4977139206239717,
+      "learning_rate": 1.3127975478343435e-07,
+      "loss": 0.259,
+      "step": 5972
+    },
+    {
+      "epoch": 2.853951376859208,
+      "grad_norm": 0.44329852164883693,
+      "learning_rate": 1.3043052769474375e-07,
+      "loss": 0.25,
+      "step": 5973
+    },
+    {
+      "epoch": 2.854429245564781,
+      "grad_norm": 0.44218052941146163,
+      "learning_rate": 1.2958403825052978e-07,
+      "loss": 0.254,
+      "step": 5974
+    },
+    {
+      "epoch": 2.8549071142703544,
+      "grad_norm": 0.4466772904005583,
+      "learning_rate": 1.2874028668559247e-07,
+      "loss": 0.2651,
+      "step": 5975
+    },
+    {
+      "epoch": 2.8553849829759272,
+      "grad_norm": 0.43862131247970004,
+      "learning_rate": 1.2789927323397232e-07,
+      "loss": 0.2602,
+      "step": 5976
+    },
+    {
+      "epoch": 2.8558628516815006,
+      "grad_norm": 0.45565339829882245,
+      "learning_rate": 1.270609981289539e-07,
+      "loss": 0.2699,
+      "step": 5977
+    },
+    {
+      "epoch": 2.856340720387074,
+      "grad_norm": 0.45795122735762345,
+      "learning_rate": 1.2622546160305894e-07,
+      "loss": 0.2498,
+      "step": 5978
+    },
+    {
+      "epoch": 2.8568185890926467,
+      "grad_norm": 0.43648336997823883,
+      "learning_rate": 1.2539266388804981e-07,
+      "loss": 0.2627,
+      "step": 5979
+    },
+    {
+      "epoch": 2.85729645779822,
+      "grad_norm": 0.4637171759508475,
+      "learning_rate": 1.245626052149318e-07,
+      "loss": 0.2535,
+      "step": 5980
+    },
+    {
+      "epoch": 2.857774326503793,
+      "grad_norm": 0.45562841177816166,
+      "learning_rate": 1.2373528581394733e-07,
+      "loss": 0.2462,
+      "step": 5981
+    },
+    {
+      "epoch": 2.8582521952093662,
+      "grad_norm": 0.4450586940009108,
+      "learning_rate": 1.2291070591457842e-07,
+      "loss": 0.2788,
+      "step": 5982
+    },
+    {
+      "epoch": 2.8587300639149396,
+      "grad_norm": 0.4386721155226896,
+      "learning_rate": 1.2208886574555323e-07,
+      "loss": 0.2725,
+      "step": 5983
+    },
+    {
+      "epoch": 2.8592079326205124,
+      "grad_norm": 0.42956994998421677,
+      "learning_rate": 1.212697655348316e-07,
+      "loss": 0.266,
+      "step": 5984
+    },
+    {
+      "epoch": 2.8596858013260857,
+      "grad_norm": 0.44622397158469107,
+      "learning_rate": 1.2045340550961958e-07,
+      "loss": 0.2606,
+      "step": 5985
+    },
+    {
+      "epoch": 2.8601636700316586,
+      "grad_norm": 0.8670959918097046,
+      "learning_rate": 1.196397858963616e-07,
+      "loss": 0.275,
+      "step": 5986
+    },
+    {
+      "epoch": 2.860641538737232,
+      "grad_norm": 0.42188794447231875,
+      "learning_rate": 1.1882890692073933e-07,
+      "loss": 0.2709,
+      "step": 5987
+    },
+    {
+      "epoch": 2.8611194074428052,
+      "grad_norm": 0.45699426887028993,
+      "learning_rate": 1.1802076880767732e-07,
+      "loss": 0.261,
+      "step": 5988
+    },
+    {
+      "epoch": 2.861597276148378,
+      "grad_norm": 0.6013209724994314,
+      "learning_rate": 1.1721537178133958e-07,
+      "loss": 0.2665,
+      "step": 5989
+    },
+    {
+      "epoch": 2.8620751448539514,
+      "grad_norm": 0.5033962646495204,
+      "learning_rate": 1.164127160651285e-07,
+      "loss": 0.2886,
+      "step": 5990
+    },
+    {
+      "epoch": 2.8625530135595243,
+      "grad_norm": 0.4387868638958641,
+      "learning_rate": 1.15612801881686e-07,
+      "loss": 0.2631,
+      "step": 5991
+    },
+    {
+      "epoch": 2.8630308822650976,
+      "grad_norm": 0.4470342389927088,
+      "learning_rate": 1.148156294528946e-07,
+      "loss": 0.2748,
+      "step": 5992
+    },
+    {
+      "epoch": 2.863508750970671,
+      "grad_norm": 0.44782132948777004,
+      "learning_rate": 1.1402119899987629e-07,
+      "loss": 0.2482,
+      "step": 5993
+    },
+    {
+      "epoch": 2.863986619676244,
+      "grad_norm": 0.4309117184393657,
+      "learning_rate": 1.1322951074299149e-07,
+      "loss": 0.2616,
+      "step": 5994
+    },
+    {
+      "epoch": 2.864464488381817,
+      "grad_norm": 0.45641574770069654,
+      "learning_rate": 1.1244056490184008e-07,
+      "loss": 0.2741,
+      "step": 5995
+    },
+    {
+      "epoch": 2.86494235708739,
+      "grad_norm": 0.4795660239874649,
+      "learning_rate": 1.1165436169526366e-07,
+      "loss": 0.2679,
+      "step": 5996
+    },
+    {
+      "epoch": 2.8654202257929633,
+      "grad_norm": 0.5323760308177075,
+      "learning_rate": 1.1087090134134005e-07,
+      "loss": 0.2589,
+      "step": 5997
+    },
+    {
+      "epoch": 2.8658980944985366,
+      "grad_norm": 0.504237551040962,
+      "learning_rate": 1.1009018405738536e-07,
+      "loss": 0.2625,
+      "step": 5998
+    },
+    {
+      "epoch": 2.86637596320411,
+      "grad_norm": 0.4402136999052143,
+      "learning_rate": 1.0931221005996084e-07,
+      "loss": 0.2755,
+      "step": 5999
+    },
+    {
+      "epoch": 2.866853831909683,
+      "grad_norm": 0.4491859848447432,
+      "learning_rate": 1.0853697956485942e-07,
+      "loss": 0.245,
+      "step": 6000
+    },
+    {
+      "epoch": 2.867331700615256,
+      "grad_norm": 0.45231341758240234,
+      "learning_rate": 1.0776449278711686e-07,
+      "loss": 0.2491,
+      "step": 6001
+    },
+    {
+      "epoch": 2.867809569320829,
+      "grad_norm": 0.4369763191287368,
+      "learning_rate": 1.0699474994100845e-07,
+      "loss": 0.2558,
+      "step": 6002
+    },
+    {
+      "epoch": 2.8682874380264023,
+      "grad_norm": 0.5358207159433925,
+      "learning_rate": 1.0622775124004669e-07,
+      "loss": 0.2668,
+      "step": 6003
+    },
+    {
+      "epoch": 2.8687653067319756,
+      "grad_norm": 0.471953453207217,
+      "learning_rate": 1.0546349689698365e-07,
+      "loss": 0.2685,
+      "step": 6004
+    },
+    {
+      "epoch": 2.8692431754375485,
+      "grad_norm": 0.44877487345300576,
+      "learning_rate": 1.0470198712381086e-07,
+      "loss": 0.2629,
+      "step": 6005
+    },
+    {
+      "epoch": 2.869721044143122,
+      "grad_norm": 0.43639707706363223,
+      "learning_rate": 1.03943222131756e-07,
+      "loss": 0.2643,
+      "step": 6006
+    },
+    {
+      "epoch": 2.8701989128486947,
+      "grad_norm": 0.43048550859297546,
+      "learning_rate": 1.0318720213128741e-07,
+      "loss": 0.2682,
+      "step": 6007
+    },
+    {
+      "epoch": 2.870676781554268,
+      "grad_norm": 0.4366175477899045,
+      "learning_rate": 1.0243392733211289e-07,
+      "loss": 0.2576,
+      "step": 6008
+    },
+    {
+      "epoch": 2.8711546502598413,
+      "grad_norm": 0.4420799299343235,
+      "learning_rate": 1.0168339794317638e-07,
+      "loss": 0.2664,
+      "step": 6009
+    },
+    {
+      "epoch": 2.871632518965414,
+      "grad_norm": 0.43731535232122004,
+      "learning_rate": 1.009356141726614e-07,
+      "loss": 0.2485,
+      "step": 6010
+    },
+    {
+      "epoch": 2.8721103876709875,
+      "grad_norm": 0.45731961228738927,
+      "learning_rate": 1.00190576227992e-07,
+      "loss": 0.2713,
+      "step": 6011
+    },
+    {
+      "epoch": 2.8725882563765603,
+      "grad_norm": 0.4357113263887429,
+      "learning_rate": 9.94482843158262e-08,
+      "loss": 0.2542,
+      "step": 6012
+    },
+    {
+      "epoch": 2.8730661250821337,
+      "grad_norm": 0.4222045104321912,
+      "learning_rate": 9.870873864206376e-08,
+      "loss": 0.2551,
+      "step": 6013
+    },
+    {
+      "epoch": 2.873543993787707,
+      "grad_norm": 0.4337713135885946,
+      "learning_rate": 9.797193941184169e-08,
+      "loss": 0.2286,
+      "step": 6014
+    },
+    {
+      "epoch": 2.87402186249328,
+      "grad_norm": 0.521268220454681,
+      "learning_rate": 9.723788682953539e-08,
+      "loss": 0.274,
+      "step": 6015
+    },
+    {
+      "epoch": 2.874499731198853,
+      "grad_norm": 0.4312558999875928,
+      "learning_rate": 9.650658109875533e-08,
+      "loss": 0.2431,
+      "step": 6016
+    },
+    {
+      "epoch": 2.874977599904426,
+      "grad_norm": 0.43570324284996176,
+      "learning_rate": 9.5778022422357e-08,
+      "loss": 0.2506,
+      "step": 6017
+    },
+    {
+      "epoch": 2.8754554686099993,
+      "grad_norm": 0.43684524986259893,
+      "learning_rate": 9.505221100242767e-08,
+      "loss": 0.2554,
+      "step": 6018
+    },
+    {
+      "epoch": 2.8759333373155727,
+      "grad_norm": 0.45826436214172095,
+      "learning_rate": 9.432914704029406e-08,
+      "loss": 0.2606,
+      "step": 6019
+    },
+    {
+      "epoch": 2.8764112060211455,
+      "grad_norm": 0.4296388690654118,
+      "learning_rate": 9.360883073652238e-08,
+      "loss": 0.2677,
+      "step": 6020
+    },
+    {
+      "epoch": 2.876889074726719,
+      "grad_norm": 1.0983884559688328,
+      "learning_rate": 9.289126229091505e-08,
+      "loss": 0.2423,
+      "step": 6021
+    },
+    {
+      "epoch": 2.8773669434322917,
+      "grad_norm": 0.4443043294613616,
+      "learning_rate": 9.217644190251285e-08,
+      "loss": 0.2727,
+      "step": 6022
+    },
+    {
+      "epoch": 2.877844812137865,
+      "grad_norm": 0.45206181728205963,
+      "learning_rate": 9.146436976959605e-08,
+      "loss": 0.2572,
+      "step": 6023
+    },
+    {
+      "epoch": 2.8783226808434383,
+      "grad_norm": 0.45123890210077205,
+      "learning_rate": 9.075504608967889e-08,
+      "loss": 0.266,
+      "step": 6024
+    },
+    {
+      "epoch": 2.8788005495490117,
+      "grad_norm": 0.4575185175136139,
+      "learning_rate": 9.004847105951509e-08,
+      "loss": 0.2639,
+      "step": 6025
+    },
+    {
+      "epoch": 2.8792784182545845,
+      "grad_norm": 0.48693714644668074,
+      "learning_rate": 8.934464487509786e-08,
+      "loss": 0.2693,
+      "step": 6026
+    },
+    {
+      "epoch": 2.879756286960158,
+      "grad_norm": 0.4425901100192589,
+      "learning_rate": 8.86435677316544e-08,
+      "loss": 0.2525,
+      "step": 6027
+    },
+    {
+      "epoch": 2.8802341556657307,
+      "grad_norm": 0.44877326238514714,
+      "learning_rate": 8.794523982365134e-08,
+      "loss": 0.2641,
+      "step": 6028
+    },
+    {
+      "epoch": 2.880712024371304,
+      "grad_norm": 0.4285119487781865,
+      "learning_rate": 8.724966134479374e-08,
+      "loss": 0.2534,
+      "step": 6029
+    },
+    {
+      "epoch": 2.8811898930768773,
+      "grad_norm": 0.46436342305670525,
+      "learning_rate": 8.655683248802282e-08,
+      "loss": 0.2821,
+      "step": 6030
+    },
+    {
+      "epoch": 2.88166776178245,
+      "grad_norm": 0.44312616547475814,
+      "learning_rate": 8.586675344551599e-08,
+      "loss": 0.2731,
+      "step": 6031
+    },
+    {
+      "epoch": 2.8821456304880235,
+      "grad_norm": 0.43442743265027634,
+      "learning_rate": 8.517942440868898e-08,
+      "loss": 0.2506,
+      "step": 6032
+    },
+    {
+      "epoch": 2.8826234991935964,
+      "grad_norm": 0.4308115226635295,
+      "learning_rate": 8.449484556819598e-08,
+      "loss": 0.2701,
+      "step": 6033
+    },
+    {
+      "epoch": 2.8831013678991697,
+      "grad_norm": 0.4460495905460458,
+      "learning_rate": 8.38130171139262e-08,
+      "loss": 0.2551,
+      "step": 6034
+    },
+    {
+      "epoch": 2.883579236604743,
+      "grad_norm": 0.5389940715079629,
+      "learning_rate": 8.313393923500613e-08,
+      "loss": 0.2657,
+      "step": 6035
+    },
+    {
+      "epoch": 2.884057105310316,
+      "grad_norm": 0.44293129901679096,
+      "learning_rate": 8.245761211980174e-08,
+      "loss": 0.242,
+      "step": 6036
+    },
+    {
+      "epoch": 2.884534974015889,
+      "grad_norm": 0.47983678443629796,
+      "learning_rate": 8.1784035955913e-08,
+      "loss": 0.2629,
+      "step": 6037
+    },
+    {
+      "epoch": 2.885012842721462,
+      "grad_norm": 0.4318505645360447,
+      "learning_rate": 8.11132109301782e-08,
+      "loss": 0.2722,
+      "step": 6038
+    },
+    {
+      "epoch": 2.8854907114270354,
+      "grad_norm": 0.43438969618864703,
+      "learning_rate": 8.044513722867298e-08,
+      "loss": 0.2696,
+      "step": 6039
+    },
+    {
+      "epoch": 2.8859685801326087,
+      "grad_norm": 0.4569375021145621,
+      "learning_rate": 7.977981503670795e-08,
+      "loss": 0.2851,
+      "step": 6040
+    },
+    {
+      "epoch": 2.8864464488381816,
+      "grad_norm": 0.4479908637746793,
+      "learning_rate": 7.911724453883329e-08,
+      "loss": 0.2706,
+      "step": 6041
+    },
+    {
+      "epoch": 2.886924317543755,
+      "grad_norm": 0.4326153900898914,
+      "learning_rate": 7.845742591883309e-08,
+      "loss": 0.2551,
+      "step": 6042
+    },
+    {
+      "epoch": 2.8874021862493278,
+      "grad_norm": 0.4357317158977019,
+      "learning_rate": 7.780035935972985e-08,
+      "loss": 0.2599,
+      "step": 6043
+    },
+    {
+      "epoch": 2.887880054954901,
+      "grad_norm": 0.47807706881714124,
+      "learning_rate": 7.714604504378332e-08,
+      "loss": 0.2515,
+      "step": 6044
+    },
+    {
+      "epoch": 2.8883579236604744,
+      "grad_norm": 0.48906636149481963,
+      "learning_rate": 7.64944831524872e-08,
+      "loss": 0.2628,
+      "step": 6045
+    },
+    {
+      "epoch": 2.8888357923660477,
+      "grad_norm": 0.48643435510610833,
+      "learning_rate": 7.584567386657248e-08,
+      "loss": 0.2653,
+      "step": 6046
+    },
+    {
+      "epoch": 2.8893136610716206,
+      "grad_norm": 0.4656245737115965,
+      "learning_rate": 7.519961736601078e-08,
+      "loss": 0.272,
+      "step": 6047
+    },
+    {
+      "epoch": 2.8897915297771934,
+      "grad_norm": 0.44143454689073186,
+      "learning_rate": 7.455631383000428e-08,
+      "loss": 0.2603,
+      "step": 6048
+    },
+    {
+      "epoch": 2.8902693984827668,
+      "grad_norm": 0.622134137502078,
+      "learning_rate": 7.391576343699359e-08,
+      "loss": 0.2628,
+      "step": 6049
+    },
+    {
+      "epoch": 2.89074726718834,
+      "grad_norm": 0.4276530571655008,
+      "learning_rate": 7.327796636465767e-08,
+      "loss": 0.2689,
+      "step": 6050
+    },
+    {
+      "epoch": 2.8912251358939134,
+      "grad_norm": 0.4382139339105153,
+      "learning_rate": 7.264292278990947e-08,
+      "loss": 0.251,
+      "step": 6051
+    },
+    {
+      "epoch": 2.8917030045994863,
+      "grad_norm": 0.47876504229735695,
+      "learning_rate": 7.201063288889809e-08,
+      "loss": 0.2814,
+      "step": 6052
+    },
+    {
+      "epoch": 2.8921808733050596,
+      "grad_norm": 0.4448938723916756,
+      "learning_rate": 7.138109683701211e-08,
+      "loss": 0.2654,
+      "step": 6053
+    },
+    {
+      "epoch": 2.8926587420106324,
+      "grad_norm": 0.4453495110101877,
+      "learning_rate": 7.075431480887074e-08,
+      "loss": 0.2682,
+      "step": 6054
+    },
+    {
+      "epoch": 2.8931366107162058,
+      "grad_norm": 0.4498675594265008,
+      "learning_rate": 7.01302869783338e-08,
+      "loss": 0.2618,
+      "step": 6055
+    },
+    {
+      "epoch": 2.893614479421779,
+      "grad_norm": 0.44182308271744547,
+      "learning_rate": 6.950901351849504e-08,
+      "loss": 0.2738,
+      "step": 6056
+    },
+    {
+      "epoch": 2.894092348127352,
+      "grad_norm": 0.4297806934525359,
+      "learning_rate": 6.889049460168662e-08,
+      "loss": 0.265,
+      "step": 6057
+    },
+    {
+      "epoch": 2.8945702168329253,
+      "grad_norm": 0.4270875134609304,
+      "learning_rate": 6.827473039947131e-08,
+      "loss": 0.2455,
+      "step": 6058
+    },
+    {
+      "epoch": 2.895048085538498,
+      "grad_norm": 0.4386255563204099,
+      "learning_rate": 6.766172108265356e-08,
+      "loss": 0.272,
+      "step": 6059
+    },
+    {
+      "epoch": 2.8955259542440714,
+      "grad_norm": 0.44212270253634706,
+      "learning_rate": 6.705146682127184e-08,
+      "loss": 0.2708,
+      "step": 6060
+    },
+    {
+      "epoch": 2.8960038229496448,
+      "grad_norm": 0.4827942892265252,
+      "learning_rate": 6.64439677845985e-08,
+      "loss": 0.2766,
+      "step": 6061
+    },
+    {
+      "epoch": 2.8964816916552176,
+      "grad_norm": 0.4419630496584052,
+      "learning_rate": 6.583922414114318e-08,
+      "loss": 0.2609,
+      "step": 6062
+    },
+    {
+      "epoch": 2.896959560360791,
+      "grad_norm": 0.4750351230744057,
+      "learning_rate": 6.523723605865174e-08,
+      "loss": 0.2763,
+      "step": 6063
+    },
+    {
+      "epoch": 2.897437429066364,
+      "grad_norm": 0.47028028478371914,
+      "learning_rate": 6.4638003704105e-08,
+      "loss": 0.2661,
+      "step": 6064
+    },
+    {
+      "epoch": 2.897915297771937,
+      "grad_norm": 0.4625973271853386,
+      "learning_rate": 6.404152724371892e-08,
+      "loss": 0.2702,
+      "step": 6065
+    },
+    {
+      "epoch": 2.8983931664775104,
+      "grad_norm": 0.7072555980090101,
+      "learning_rate": 6.344780684294671e-08,
+      "loss": 0.2669,
+      "step": 6066
+    },
+    {
+      "epoch": 2.8988710351830833,
+      "grad_norm": 0.4482616347126552,
+      "learning_rate": 6.28568426664744e-08,
+      "loss": 0.2533,
+      "step": 6067
+    },
+    {
+      "epoch": 2.8993489038886566,
+      "grad_norm": 0.4303610137833005,
+      "learning_rate": 6.226863487822532e-08,
+      "loss": 0.2702,
+      "step": 6068
+    },
+    {
+      "epoch": 2.8998267725942295,
+      "grad_norm": 0.4385548652125175,
+      "learning_rate": 6.168318364135895e-08,
+      "loss": 0.2586,
+      "step": 6069
+    },
+    {
+      "epoch": 2.900304641299803,
+      "grad_norm": 0.46216493168944667,
+      "learning_rate": 6.110048911826871e-08,
+      "loss": 0.2605,
+      "step": 6070
+    },
+    {
+      "epoch": 2.900782510005376,
+      "grad_norm": 0.42831561535218843,
+      "learning_rate": 6.05205514705831e-08,
+      "loss": 0.2689,
+      "step": 6071
+    },
+    {
+      "epoch": 2.9012603787109494,
+      "grad_norm": 0.42968742796774145,
+      "learning_rate": 5.99433708591679e-08,
+      "loss": 0.2601,
+      "step": 6072
+    },
+    {
+      "epoch": 2.9017382474165223,
+      "grad_norm": 1.0032255802524332,
+      "learning_rate": 5.9368947444121695e-08,
+      "loss": 0.2851,
+      "step": 6073
+    },
+    {
+      "epoch": 2.902216116122095,
+      "grad_norm": 0.44746843283921595,
+      "learning_rate": 5.879728138477925e-08,
+      "loss": 0.2645,
+      "step": 6074
+    },
+    {
+      "epoch": 2.9026939848276685,
+      "grad_norm": 0.45839779776483,
+      "learning_rate": 5.82283728397115e-08,
+      "loss": 0.2928,
+      "step": 6075
+    },
+    {
+      "epoch": 2.903171853533242,
+      "grad_norm": 0.4426768127258644,
+      "learning_rate": 5.766222196672333e-08,
+      "loss": 0.2529,
+      "step": 6076
+    },
+    {
+      "epoch": 2.903649722238815,
+      "grad_norm": 0.4306421815201952,
+      "learning_rate": 5.709882892285468e-08,
+      "loss": 0.2596,
+      "step": 6077
+    },
+    {
+      "epoch": 2.904127590944388,
+      "grad_norm": 0.4339299174215774,
+      "learning_rate": 5.653819386438164e-08,
+      "loss": 0.2543,
+      "step": 6078
+    },
+    {
+      "epoch": 2.9046054596499613,
+      "grad_norm": 0.46448596548019094,
+      "learning_rate": 5.598031694681316e-08,
+      "loss": 0.2698,
+      "step": 6079
+    },
+    {
+      "epoch": 2.905083328355534,
+      "grad_norm": 0.4319137355146327,
+      "learning_rate": 5.542519832489546e-08,
+      "loss": 0.2696,
+      "step": 6080
+    },
+    {
+      "epoch": 2.9055611970611075,
+      "grad_norm": 0.5684354169120892,
+      "learning_rate": 5.487283815260869e-08,
+      "loss": 0.2694,
+      "step": 6081
+    },
+    {
+      "epoch": 2.906039065766681,
+      "grad_norm": 0.4360547156677553,
+      "learning_rate": 5.432323658316807e-08,
+      "loss": 0.2629,
+      "step": 6082
+    },
+    {
+      "epoch": 2.9065169344722537,
+      "grad_norm": 0.4342812482531317,
+      "learning_rate": 5.3776393769021664e-08,
+      "loss": 0.271,
+      "step": 6083
+    },
+    {
+      "epoch": 2.906994803177827,
+      "grad_norm": 0.46003291233940047,
+      "learning_rate": 5.3232309861857015e-08,
+      "loss": 0.2489,
+      "step": 6084
+    },
+    {
+      "epoch": 2.9074726718834,
+      "grad_norm": 0.4325963268001337,
+      "learning_rate": 5.269098501259007e-08,
+      "loss": 0.2794,
+      "step": 6085
+    },
+    {
+      "epoch": 2.907950540588973,
+      "grad_norm": 0.4637871151293466,
+      "learning_rate": 5.2152419371376276e-08,
+      "loss": 0.2726,
+      "step": 6086
+    },
+    {
+      "epoch": 2.9084284092945465,
+      "grad_norm": 0.45775569434566626,
+      "learning_rate": 5.161661308760613e-08,
+      "loss": 0.2769,
+      "step": 6087
+    },
+    {
+      "epoch": 2.9089062780001194,
+      "grad_norm": 0.5008124956659363,
+      "learning_rate": 5.108356630989963e-08,
+      "loss": 0.2598,
+      "step": 6088
+    },
+    {
+      "epoch": 2.9093841467056927,
+      "grad_norm": 0.4407017585102997,
+      "learning_rate": 5.05532791861163e-08,
+      "loss": 0.2704,
+      "step": 6089
+    },
+    {
+      "epoch": 2.9098620154112655,
+      "grad_norm": 0.461333224542017,
+      "learning_rate": 5.002575186334735e-08,
+      "loss": 0.2817,
+      "step": 6090
+    },
+    {
+      "epoch": 2.910339884116839,
+      "grad_norm": 0.44289468727956904,
+      "learning_rate": 4.950098448792129e-08,
+      "loss": 0.2656,
+      "step": 6091
+    },
+    {
+      "epoch": 2.910817752822412,
+      "grad_norm": 0.47285680491757126,
+      "learning_rate": 4.897897720539835e-08,
+      "loss": 0.2561,
+      "step": 6092
+    },
+    {
+      "epoch": 2.911295621527985,
+      "grad_norm": 0.4770909008792775,
+      "learning_rate": 4.8459730160573814e-08,
+      "loss": 0.2543,
+      "step": 6093
+    },
+    {
+      "epoch": 2.9117734902335584,
+      "grad_norm": 0.4711159546492226,
+      "learning_rate": 4.794324349747803e-08,
+      "loss": 0.263,
+      "step": 6094
+    },
+    {
+      "epoch": 2.9122513589391312,
+      "grad_norm": 0.4302650665937928,
+      "learning_rate": 4.742951735937418e-08,
+      "loss": 0.263,
+      "step": 6095
+    },
+    {
+      "epoch": 2.9127292276447045,
+      "grad_norm": 0.44234241950445463,
+      "learning_rate": 4.691855188876271e-08,
+      "loss": 0.2591,
+      "step": 6096
+    },
+    {
+      "epoch": 2.913207096350278,
+      "grad_norm": 0.43773577729487084,
+      "learning_rate": 4.641034722737581e-08,
+      "loss": 0.2743,
+      "step": 6097
+    },
+    {
+      "epoch": 2.913684965055851,
+      "grad_norm": 0.4482633004011929,
+      "learning_rate": 4.590490351618071e-08,
+      "loss": 0.2542,
+      "step": 6098
+    },
+    {
+      "epoch": 2.914162833761424,
+      "grad_norm": 0.45388506640042264,
+      "learning_rate": 4.5402220895377494e-08,
+      "loss": 0.2608,
+      "step": 6099
+    },
+    {
+      "epoch": 2.9146407024669974,
+      "grad_norm": 0.4677634390084611,
+      "learning_rate": 4.490229950440239e-08,
+      "loss": 0.267,
+      "step": 6100
+    },
+    {
+      "epoch": 2.9151185711725702,
+      "grad_norm": 0.43322907347723366,
+      "learning_rate": 4.4405139481924485e-08,
+      "loss": 0.2852,
+      "step": 6101
+    },
+    {
+      "epoch": 2.9155964398781435,
+      "grad_norm": 0.43253137272245884,
+      "learning_rate": 4.39107409658468e-08,
+      "loss": 0.2523,
+      "step": 6102
+    },
+    {
+      "epoch": 2.916074308583717,
+      "grad_norm": 0.8312520936498295,
+      "learning_rate": 4.341910409330741e-08,
+      "loss": 0.2635,
+      "step": 6103
+    },
+    {
+      "epoch": 2.9165521772892897,
+      "grad_norm": 0.6559893173522333,
+      "learning_rate": 4.293022900067723e-08,
+      "loss": 0.2503,
+      "step": 6104
+    },
+    {
+      "epoch": 2.917030045994863,
+      "grad_norm": 0.4843023216915645,
+      "learning_rate": 4.2444115823562226e-08,
+      "loss": 0.2675,
+      "step": 6105
+    },
+    {
+      "epoch": 2.917507914700436,
+      "grad_norm": 0.43442410136217113,
+      "learning_rate": 4.196076469680122e-08,
+      "loss": 0.2843,
+      "step": 6106
+    },
+    {
+      "epoch": 2.9179857834060092,
+      "grad_norm": 0.43098967320599624,
+      "learning_rate": 4.148017575446695e-08,
+      "loss": 0.2564,
+      "step": 6107
+    },
+    {
+      "epoch": 2.9184636521115825,
+      "grad_norm": 0.4679928302022156,
+      "learning_rate": 4.100234912986611e-08,
+      "loss": 0.2544,
+      "step": 6108
+    },
+    {
+      "epoch": 2.9189415208171554,
+      "grad_norm": 0.4269404926785668,
+      "learning_rate": 4.052728495554159e-08,
+      "loss": 0.2624,
+      "step": 6109
+    },
+    {
+      "epoch": 2.9194193895227287,
+      "grad_norm": 0.46370533071441733,
+      "learning_rate": 4.005498336326463e-08,
+      "loss": 0.2847,
+      "step": 6110
+    },
+    {
+      "epoch": 2.9198972582283016,
+      "grad_norm": 0.4586911857247521,
+      "learning_rate": 3.958544448404489e-08,
+      "loss": 0.2683,
+      "step": 6111
+    },
+    {
+      "epoch": 2.920375126933875,
+      "grad_norm": 0.4440943392862457,
+      "learning_rate": 3.911866844812484e-08,
+      "loss": 0.2576,
+      "step": 6112
+    },
+    {
+      "epoch": 2.9208529956394482,
+      "grad_norm": 0.4520138400943036,
+      "learning_rate": 3.865465538497981e-08,
+      "loss": 0.2718,
+      "step": 6113
+    },
+    {
+      "epoch": 2.921330864345021,
+      "grad_norm": 0.7797476182577636,
+      "learning_rate": 3.819340542331684e-08,
+      "loss": 0.2698,
+      "step": 6114
+    },
+    {
+      "epoch": 2.9218087330505944,
+      "grad_norm": 0.46668125602963445,
+      "learning_rate": 3.773491869108137e-08,
+      "loss": 0.2821,
+      "step": 6115
+    },
+    {
+      "epoch": 2.9222866017561673,
+      "grad_norm": 0.4473526269050483,
+      "learning_rate": 3.727919531544721e-08,
+      "loss": 0.2619,
+      "step": 6116
+    },
+    {
+      "epoch": 2.9227644704617406,
+      "grad_norm": 0.4465730584095882,
+      "learning_rate": 3.6826235422824375e-08,
+      "loss": 0.2719,
+      "step": 6117
+    },
+    {
+      "epoch": 2.923242339167314,
+      "grad_norm": 0.4405667747199723,
+      "learning_rate": 3.63760391388579e-08,
+      "loss": 0.2582,
+      "step": 6118
+    },
+    {
+      "epoch": 2.923720207872887,
+      "grad_norm": 0.43242706008898524,
+      "learning_rate": 3.5928606588422344e-08,
+      "loss": 0.2616,
+      "step": 6119
+    },
+    {
+      "epoch": 2.92419807657846,
+      "grad_norm": 0.43740281555793414,
+      "learning_rate": 3.548393789562732e-08,
+      "loss": 0.2542,
+      "step": 6120
+    },
+    {
+      "epoch": 2.924675945284033,
+      "grad_norm": 0.436622075631485,
+      "learning_rate": 3.504203318381749e-08,
+      "loss": 0.26,
+      "step": 6121
+    },
+    {
+      "epoch": 2.9251538139896063,
+      "grad_norm": 0.4323874580312323,
+      "learning_rate": 3.460289257556926e-08,
+      "loss": 0.2544,
+      "step": 6122
+    },
+    {
+      "epoch": 2.9256316826951796,
+      "grad_norm": 0.4449893196540946,
+      "learning_rate": 3.416651619269073e-08,
+      "loss": 0.2632,
+      "step": 6123
+    },
+    {
+      "epoch": 2.926109551400753,
+      "grad_norm": 0.4383521097185537,
+      "learning_rate": 3.373290415622732e-08,
+      "loss": 0.2594,
+      "step": 6124
+    },
+    {
+      "epoch": 2.9265874201063258,
+      "grad_norm": 0.4616607673944823,
+      "learning_rate": 3.3302056586453916e-08,
+      "loss": 0.2419,
+      "step": 6125
+    },
+    {
+      "epoch": 2.927065288811899,
+      "grad_norm": 0.43426940100872696,
+      "learning_rate": 3.287397360288047e-08,
+      "loss": 0.2517,
+      "step": 6126
+    },
+    {
+      "epoch": 2.927543157517472,
+      "grad_norm": 0.43983360686347156,
+      "learning_rate": 3.244865532424979e-08,
+      "loss": 0.2776,
+      "step": 6127
+    },
+    {
+      "epoch": 2.9280210262230453,
+      "grad_norm": 0.42574611069588764,
+      "learning_rate": 3.2026101868538605e-08,
+      "loss": 0.255,
+      "step": 6128
+    },
+    {
+      "epoch": 2.9284988949286186,
+      "grad_norm": 0.4331855136966984,
+      "learning_rate": 3.1606313352953166e-08,
+      "loss": 0.2561,
+      "step": 6129
+    },
+    {
+      "epoch": 2.9289767636341915,
+      "grad_norm": 0.4360937130460965,
+      "learning_rate": 3.118928989393699e-08,
+      "loss": 0.2666,
+      "step": 6130
+    },
+    {
+      "epoch": 2.9294546323397648,
+      "grad_norm": 0.4647708422391565,
+      "learning_rate": 3.077503160716533e-08,
+      "loss": 0.2639,
+      "step": 6131
+    },
+    {
+      "epoch": 2.9299325010453376,
+      "grad_norm": 0.43528666761458085,
+      "learning_rate": 3.0363538607546264e-08,
+      "loss": 0.2655,
+      "step": 6132
+    },
+    {
+      "epoch": 2.930410369750911,
+      "grad_norm": 0.4433119943756909,
+      "learning_rate": 2.99548110092196e-08,
+      "loss": 0.2583,
+      "step": 6133
+    },
+    {
+      "epoch": 2.9308882384564843,
+      "grad_norm": 0.440304631789597,
+      "learning_rate": 2.9548848925560204e-08,
+      "loss": 0.2647,
+      "step": 6134
+    },
+    {
+      "epoch": 2.931366107162057,
+      "grad_norm": 0.4259765178512256,
+      "learning_rate": 2.9145652469174666e-08,
+      "loss": 0.2695,
+      "step": 6135
+    },
+    {
+      "epoch": 2.9318439758676305,
+      "grad_norm": 0.4379793362905143,
+      "learning_rate": 2.874522175190242e-08,
+      "loss": 0.2571,
+      "step": 6136
+    },
+    {
+      "epoch": 2.9323218445732033,
+      "grad_norm": 0.4505313706517066,
+      "learning_rate": 2.8347556884814608e-08,
+      "loss": 0.2732,
+      "step": 6137
+    },
+    {
+      "epoch": 2.9327997132787766,
+      "grad_norm": 0.43130212488145003,
+      "learning_rate": 2.7952657978218557e-08,
+      "loss": 0.2715,
+      "step": 6138
+    },
+    {
+      "epoch": 2.93327758198435,
+      "grad_norm": 0.455093876629633,
+      "learning_rate": 2.7560525141651085e-08,
+      "loss": 0.2708,
+      "step": 6139
+    },
+    {
+      "epoch": 2.933755450689923,
+      "grad_norm": 0.542151458023748,
+      "learning_rate": 2.7171158483882963e-08,
+      "loss": 0.2667,
+      "step": 6140
+    },
+    {
+      "epoch": 2.934233319395496,
+      "grad_norm": 0.43987512532735656,
+      "learning_rate": 2.678455811291669e-08,
+      "loss": 0.2648,
+      "step": 6141
+    },
+    {
+      "epoch": 2.934711188101069,
+      "grad_norm": 0.4611481415195722,
+      "learning_rate": 2.640072413599093e-08,
+      "loss": 0.2523,
+      "step": 6142
+    },
+    {
+      "epoch": 2.9351890568066423,
+      "grad_norm": 0.4680505547187445,
+      "learning_rate": 2.6019656659572734e-08,
+      "loss": 0.2761,
+      "step": 6143
+    },
+    {
+      "epoch": 2.9356669255122156,
+      "grad_norm": 0.4347789880573274,
+      "learning_rate": 2.564135578936422e-08,
+      "loss": 0.2678,
+      "step": 6144
+    },
+    {
+      "epoch": 2.9361447942177885,
+      "grad_norm": 0.43779621846285605,
+      "learning_rate": 2.5265821630298116e-08,
+      "loss": 0.2703,
+      "step": 6145
+    },
+    {
+      "epoch": 2.936622662923362,
+      "grad_norm": 0.5080676465853426,
+      "learning_rate": 2.4893054286542207e-08,
+      "loss": 0.282,
+      "step": 6146
+    },
+    {
+      "epoch": 2.9371005316289347,
+      "grad_norm": 0.4240127151465205,
+      "learning_rate": 2.4523053861494894e-08,
+      "loss": 0.2666,
+      "step": 6147
+    },
+    {
+      "epoch": 2.937578400334508,
+      "grad_norm": 0.43808035306246806,
+      "learning_rate": 2.4155820457788525e-08,
+      "loss": 0.2559,
+      "step": 6148
+    },
+    {
+      "epoch": 2.9380562690400813,
+      "grad_norm": 0.4600124982961411,
+      "learning_rate": 2.3791354177286062e-08,
+      "loss": 0.2551,
+      "step": 6149
+    },
+    {
+      "epoch": 2.9385341377456546,
+      "grad_norm": 0.4808481604273592,
+      "learning_rate": 2.3429655121085525e-08,
+      "loss": 0.2572,
+      "step": 6150
+    },
+    {
+      "epoch": 2.9390120064512275,
+      "grad_norm": 0.4420542869404689,
+      "learning_rate": 2.3070723389514437e-08,
+      "loss": 0.2684,
+      "step": 6151
+    },
+    {
+      "epoch": 2.939489875156801,
+      "grad_norm": 0.42935263804364654,
+      "learning_rate": 2.2714559082134267e-08,
+      "loss": 0.2482,
+      "step": 6152
+    },
+    {
+      "epoch": 2.9399677438623737,
+      "grad_norm": 0.4476128760597761,
+      "learning_rate": 2.2361162297739327e-08,
+      "loss": 0.2692,
+      "step": 6153
+    },
+    {
+      "epoch": 2.940445612567947,
+      "grad_norm": 0.4687545989577253,
+      "learning_rate": 2.201053313435564e-08,
+      "loss": 0.2652,
+      "step": 6154
+    },
+    {
+      "epoch": 2.9409234812735203,
+      "grad_norm": 0.4447212940871088,
+      "learning_rate": 2.1662671689242076e-08,
+      "loss": 0.2644,
+      "step": 6155
+    },
+    {
+      "epoch": 2.941401349979093,
+      "grad_norm": 0.4315254295809012,
+      "learning_rate": 2.131757805888701e-08,
+      "loss": 0.2588,
+      "step": 6156
+    },
+    {
+      "epoch": 2.9418792186846665,
+      "grad_norm": 0.43460273368097063,
+      "learning_rate": 2.0975252339016095e-08,
+      "loss": 0.2647,
+      "step": 6157
+    },
+    {
+      "epoch": 2.9423570873902394,
+      "grad_norm": 0.4537080705014196,
+      "learning_rate": 2.0635694624582258e-08,
+      "loss": 0.2605,
+      "step": 6158
+    },
+    {
+      "epoch": 2.9428349560958127,
+      "grad_norm": 0.5945505658301509,
+      "learning_rate": 2.0298905009774606e-08,
+      "loss": 0.2566,
+      "step": 6159
+    },
+    {
+      "epoch": 2.943312824801386,
+      "grad_norm": 0.4221694390673157,
+      "learning_rate": 1.996488358801174e-08,
+      "loss": 0.2622,
+      "step": 6160
+    },
+    {
+      "epoch": 2.943790693506959,
+      "grad_norm": 0.4369946666917783,
+      "learning_rate": 1.9633630451945106e-08,
+      "loss": 0.2588,
+      "step": 6161
+    },
+    {
+      "epoch": 2.944268562212532,
+      "grad_norm": 0.4581796037391759,
+      "learning_rate": 1.9305145693457873e-08,
+      "loss": 0.262,
+      "step": 6162
+    },
+    {
+      "epoch": 2.944746430918105,
+      "grad_norm": 0.45812073938100284,
+      "learning_rate": 1.897942940366715e-08,
+      "loss": 0.2633,
+      "step": 6163
+    },
+    {
+      "epoch": 2.9452242996236784,
+      "grad_norm": 0.4342899841434601,
+      "learning_rate": 1.8656481672921778e-08,
+      "loss": 0.2524,
+      "step": 6164
+    },
+    {
+      "epoch": 2.9457021683292517,
+      "grad_norm": 0.5515670473317069,
+      "learning_rate": 1.8336302590798992e-08,
+      "loss": 0.2504,
+      "step": 6165
+    },
+    {
+      "epoch": 2.9461800370348246,
+      "grad_norm": 0.4650205257532432,
+      "learning_rate": 1.8018892246113307e-08,
+      "loss": 0.2555,
+      "step": 6166
+    },
+    {
+      "epoch": 2.946657905740398,
+      "grad_norm": 0.5280605407104664,
+      "learning_rate": 1.7704250726907623e-08,
+      "loss": 0.2598,
+      "step": 6167
+    },
+    {
+      "epoch": 2.9471357744459707,
+      "grad_norm": 0.4312322170959968,
+      "learning_rate": 1.7392378120457686e-08,
+      "loss": 0.2695,
+      "step": 6168
+    },
+    {
+      "epoch": 2.947613643151544,
+      "grad_norm": 0.43878655465975397,
+      "learning_rate": 1.7083274513272075e-08,
+      "loss": 0.2416,
+      "step": 6169
+    },
+    {
+      "epoch": 2.9480915118571174,
+      "grad_norm": 0.47724365855005124,
+      "learning_rate": 1.677693999109109e-08,
+      "loss": 0.2799,
+      "step": 6170
+    },
+    {
+      "epoch": 2.9485693805626902,
+      "grad_norm": 0.42611304191004884,
+      "learning_rate": 1.6473374638885653e-08,
+      "loss": 0.2656,
+      "step": 6171
+    },
+    {
+      "epoch": 2.9490472492682636,
+      "grad_norm": 0.4437090768992525,
+      "learning_rate": 1.6172578540859518e-08,
+      "loss": 0.2739,
+      "step": 6172
+    },
+    {
+      "epoch": 2.9495251179738364,
+      "grad_norm": 0.4993388134546404,
+      "learning_rate": 1.5874551780448168e-08,
+      "loss": 0.2611,
+      "step": 6173
+    },
+    {
+      "epoch": 2.9500029866794097,
+      "grad_norm": 0.5246384618817709,
+      "learning_rate": 1.5579294440319914e-08,
+      "loss": 0.26,
+      "step": 6174
+    },
+    {
+      "epoch": 2.950480855384983,
+      "grad_norm": 0.4396611447170219,
+      "learning_rate": 1.5286806602372583e-08,
+      "loss": 0.2662,
+      "step": 6175
+    },
+    {
+      "epoch": 2.9509587240905564,
+      "grad_norm": 0.43916399298231507,
+      "learning_rate": 1.4997088347737942e-08,
+      "loss": 0.2786,
+      "step": 6176
+    },
+    {
+      "epoch": 2.9514365927961292,
+      "grad_norm": 0.4416382489397697,
+      "learning_rate": 1.4710139756778374e-08,
+      "loss": 0.2701,
+      "step": 6177
+    },
+    {
+      "epoch": 2.9519144615017026,
+      "grad_norm": 0.44737409568336867,
+      "learning_rate": 1.4425960909087989e-08,
+      "loss": 0.27,
+      "step": 6178
+    },
+    {
+      "epoch": 2.9523923302072754,
+      "grad_norm": 0.44284766864103403,
+      "learning_rate": 1.414455188349484e-08,
+      "loss": 0.264,
+      "step": 6179
+    },
+    {
+      "epoch": 2.9528701989128487,
+      "grad_norm": 0.5614132268071157,
+      "learning_rate": 1.3865912758054267e-08,
+      "loss": 0.2693,
+      "step": 6180
+    },
+    {
+      "epoch": 2.953348067618422,
+      "grad_norm": 0.44645974337163896,
+      "learning_rate": 1.3590043610057779e-08,
+      "loss": 0.2856,
+      "step": 6181
+    },
+    {
+      "epoch": 2.953825936323995,
+      "grad_norm": 0.4384106162295247,
+      "learning_rate": 1.3316944516026386e-08,
+      "loss": 0.2532,
+      "step": 6182
+    },
+    {
+      "epoch": 2.9543038050295682,
+      "grad_norm": 0.43462058917441077,
+      "learning_rate": 1.3046615551711716e-08,
+      "loss": 0.2646,
+      "step": 6183
+    },
+    {
+      "epoch": 2.954781673735141,
+      "grad_norm": 0.5819832565699149,
+      "learning_rate": 1.2779056792099343e-08,
+      "loss": 0.2679,
+      "step": 6184
+    },
+    {
+      "epoch": 2.9552595424407144,
+      "grad_norm": 0.4460079501289947,
+      "learning_rate": 1.2514268311405452e-08,
+      "loss": 0.246,
+      "step": 6185
+    },
+    {
+      "epoch": 2.9557374111462877,
+      "grad_norm": 0.4356231894949297,
+      "learning_rate": 1.2252250183076852e-08,
+      "loss": 0.262,
+      "step": 6186
+    },
+    {
+      "epoch": 2.9562152798518606,
+      "grad_norm": 0.43337849139435436,
+      "learning_rate": 1.1993002479793181e-08,
+      "loss": 0.2683,
+      "step": 6187
+    },
+    {
+      "epoch": 2.956693148557434,
+      "grad_norm": 0.4622749840308141,
+      "learning_rate": 1.1736525273465805e-08,
+      "loss": 0.2842,
+      "step": 6188
+    },
+    {
+      "epoch": 2.957171017263007,
+      "grad_norm": 0.4502631856582929,
+      "learning_rate": 1.1482818635235592e-08,
+      "loss": 0.2788,
+      "step": 6189
+    },
+    {
+      "epoch": 2.95764888596858,
+      "grad_norm": 0.47079401887102384,
+      "learning_rate": 1.1231882635477364e-08,
+      "loss": 0.2614,
+      "step": 6190
+    },
+    {
+      "epoch": 2.9581267546741534,
+      "grad_norm": 0.4695321059725002,
+      "learning_rate": 1.0983717343796552e-08,
+      "loss": 0.2673,
+      "step": 6191
+    },
+    {
+      "epoch": 2.9586046233797263,
+      "grad_norm": 0.44118500556042844,
+      "learning_rate": 1.0738322829028092e-08,
+      "loss": 0.259,
+      "step": 6192
+    },
+    {
+      "epoch": 2.9590824920852996,
+      "grad_norm": 0.4277238079715053,
+      "learning_rate": 1.0495699159241979e-08,
+      "loss": 0.2523,
+      "step": 6193
+    },
+    {
+      "epoch": 2.9595603607908725,
+      "grad_norm": 0.4388918259690208,
+      "learning_rate": 1.0255846401737713e-08,
+      "loss": 0.2711,
+      "step": 6194
+    },
+    {
+      "epoch": 2.960038229496446,
+      "grad_norm": 0.43524845914682875,
+      "learning_rate": 1.0018764623045407e-08,
+      "loss": 0.2712,
+      "step": 6195
+    },
+    {
+      "epoch": 2.960516098202019,
+      "grad_norm": 0.43647257946844553,
+      "learning_rate": 9.784453888926903e-09,
+      "loss": 0.2615,
+      "step": 6196
+    },
+    {
+      "epoch": 2.960993966907592,
+      "grad_norm": 0.4421145755712329,
+      "learning_rate": 9.552914264376878e-09,
+      "loss": 0.2705,
+      "step": 6197
+    },
+    {
+      "epoch": 2.9614718356131653,
+      "grad_norm": 0.44265663704994035,
+      "learning_rate": 9.324145813619512e-09,
+      "loss": 0.2559,
+      "step": 6198
+    },
+    {
+      "epoch": 2.961949704318738,
+      "grad_norm": 0.49174281184358326,
+      "learning_rate": 9.098148600111822e-09,
+      "loss": 0.2795,
+      "step": 6199
+    },
+    {
+      "epoch": 2.9624275730243115,
+      "grad_norm": 0.4440331427084486,
+      "learning_rate": 8.874922686541442e-09,
+      "loss": 0.2811,
+      "step": 6200
+    },
+    {
+      "epoch": 2.962905441729885,
+      "grad_norm": 0.44320171963272337,
+      "learning_rate": 8.654468134826621e-09,
+      "loss": 0.2501,
+      "step": 6201
+    },
+    {
+      "epoch": 2.963383310435458,
+      "grad_norm": 0.44354434096176776,
+      "learning_rate": 8.436785006118442e-09,
+      "loss": 0.2701,
+      "step": 6202
+    },
+    {
+      "epoch": 2.963861179141031,
+      "grad_norm": 0.4389788144482154,
+      "learning_rate": 8.221873360798604e-09,
+      "loss": 0.2738,
+      "step": 6203
+    },
+    {
+      "epoch": 2.9643390478466043,
+      "grad_norm": 0.49592260296870216,
+      "learning_rate": 8.009733258478314e-09,
+      "loss": 0.268,
+      "step": 6204
+    },
+    {
+      "epoch": 2.964816916552177,
+      "grad_norm": 0.4400993468472191,
+      "learning_rate": 7.800364758002721e-09,
+      "loss": 0.2597,
+      "step": 6205
+    },
+    {
+      "epoch": 2.9652947852577505,
+      "grad_norm": 0.4309074500166779,
+      "learning_rate": 7.593767917445372e-09,
+      "loss": 0.264,
+      "step": 6206
+    },
+    {
+      "epoch": 2.965772653963324,
+      "grad_norm": 0.45179483541632487,
+      "learning_rate": 7.389942794114868e-09,
+      "loss": 0.2736,
+      "step": 6207
+    },
+    {
+      "epoch": 2.9662505226688967,
+      "grad_norm": 0.4220551564794903,
+      "learning_rate": 7.188889444548208e-09,
+      "loss": 0.2685,
+      "step": 6208
+    },
+    {
+      "epoch": 2.96672839137447,
+      "grad_norm": 0.4806287977254884,
+      "learning_rate": 6.990607924511894e-09,
+      "loss": 0.2508,
+      "step": 6209
+    },
+    {
+      "epoch": 2.967206260080043,
+      "grad_norm": 0.4320953184140779,
+      "learning_rate": 6.795098289008595e-09,
+      "loss": 0.2842,
+      "step": 6210
+    },
+    {
+      "epoch": 2.967684128785616,
+      "grad_norm": 0.43834356326205404,
+      "learning_rate": 6.602360592267154e-09,
+      "loss": 0.279,
+      "step": 6211
+    },
+    {
+      "epoch": 2.9681619974911895,
+      "grad_norm": 0.4481157300961235,
+      "learning_rate": 6.412394887750362e-09,
+      "loss": 0.2663,
+      "step": 6212
+    },
+    {
+      "epoch": 2.9686398661967623,
+      "grad_norm": 0.4295979402150286,
+      "learning_rate": 6.225201228151623e-09,
+      "loss": 0.2652,
+      "step": 6213
+    },
+    {
+      "epoch": 2.9691177349023357,
+      "grad_norm": 0.6031224976708806,
+      "learning_rate": 6.040779665394958e-09,
+      "loss": 0.2506,
+      "step": 6214
+    },
+    {
+      "epoch": 2.9695956036079085,
+      "grad_norm": 0.4315091977947095,
+      "learning_rate": 5.859130250636113e-09,
+      "loss": 0.2695,
+      "step": 6215
+    },
+    {
+      "epoch": 2.970073472313482,
+      "grad_norm": 0.6583738125813944,
+      "learning_rate": 5.680253034260341e-09,
+      "loss": 0.2548,
+      "step": 6216
+    },
+    {
+      "epoch": 2.970551341019055,
+      "grad_norm": 0.4378215901474907,
+      "learning_rate": 5.504148065885728e-09,
+      "loss": 0.2646,
+      "step": 6217
+    },
+    {
+      "epoch": 2.971029209724628,
+      "grad_norm": 0.4503142761167827,
+      "learning_rate": 5.330815394359867e-09,
+      "loss": 0.2473,
+      "step": 6218
+    },
+    {
+      "epoch": 2.9715070784302013,
+      "grad_norm": 0.4469746516278154,
+      "learning_rate": 5.160255067764297e-09,
+      "loss": 0.268,
+      "step": 6219
+    },
+    {
+      "epoch": 2.971984947135774,
+      "grad_norm": 0.4335561129582791,
+      "learning_rate": 4.992467133406731e-09,
+      "loss": 0.2516,
+      "step": 6220
+    },
+    {
+      "epoch": 2.9724628158413475,
+      "grad_norm": 0.4418315312058818,
+      "learning_rate": 4.827451637829938e-09,
+      "loss": 0.2604,
+      "step": 6221
+    },
+    {
+      "epoch": 2.972940684546921,
+      "grad_norm": 0.4494480907514312,
+      "learning_rate": 4.665208626807305e-09,
+      "loss": 0.254,
+      "step": 6222
+    },
+    {
+      "epoch": 2.9734185532524937,
+      "grad_norm": 0.4324550053316957,
+      "learning_rate": 4.50573814534061e-09,
+      "loss": 0.2436,
+      "step": 6223
+    },
+    {
+      "epoch": 2.973896421958067,
+      "grad_norm": 0.47336069337487435,
+      "learning_rate": 4.34904023766447e-09,
+      "loss": 0.2705,
+      "step": 6224
+    },
+    {
+      "epoch": 2.97437429066364,
+      "grad_norm": 0.4628580904559059,
+      "learning_rate": 4.195114947244117e-09,
+      "loss": 0.2687,
+      "step": 6225
+    },
+    {
+      "epoch": 2.974852159369213,
+      "grad_norm": 0.43605946474328056,
+      "learning_rate": 4.043962316775396e-09,
+      "loss": 0.2634,
+      "step": 6226
+    },
+    {
+      "epoch": 2.9753300280747865,
+      "grad_norm": 0.42009356204523896,
+      "learning_rate": 3.895582388186991e-09,
+      "loss": 0.2566,
+      "step": 6227
+    },
+    {
+      "epoch": 2.97580789678036,
+      "grad_norm": 0.44595200592177053,
+      "learning_rate": 3.749975202635981e-09,
+      "loss": 0.2609,
+      "step": 6228
+    },
+    {
+      "epoch": 2.9762857654859327,
+      "grad_norm": 0.45865874193509315,
+      "learning_rate": 3.607140800510056e-09,
+      "loss": 0.2625,
+      "step": 6229
+    },
+    {
+      "epoch": 2.976763634191506,
+      "grad_norm": 0.5227990370320047,
+      "learning_rate": 3.4670792214297476e-09,
+      "loss": 0.2808,
+      "step": 6230
+    },
+    {
+      "epoch": 2.977241502897079,
+      "grad_norm": 0.444601214169423,
+      "learning_rate": 3.3297905042462e-09,
+      "loss": 0.2686,
+      "step": 6231
+    },
+    {
+      "epoch": 2.977719371602652,
+      "grad_norm": 0.44325315886349526,
+      "learning_rate": 3.1952746870411723e-09,
+      "loss": 0.259,
+      "step": 6232
+    },
+    {
+      "epoch": 2.9781972403082255,
+      "grad_norm": 0.4263331644343002,
+      "learning_rate": 3.06353180712593e-09,
+      "loss": 0.2601,
+      "step": 6233
+    },
+    {
+      "epoch": 2.9786751090137984,
+      "grad_norm": 0.445388777913223,
+      "learning_rate": 2.9345619010434644e-09,
+      "loss": 0.2709,
+      "step": 6234
+    },
+    {
+      "epoch": 2.9791529777193717,
+      "grad_norm": 0.4290478297016977,
+      "learning_rate": 2.808365004569602e-09,
+      "loss": 0.2349,
+      "step": 6235
+    },
+    {
+      "epoch": 2.9796308464249446,
+      "grad_norm": 0.43300161485588196,
+      "learning_rate": 2.684941152706344e-09,
+      "loss": 0.254,
+      "step": 6236
+    },
+    {
+      "epoch": 2.980108715130518,
+      "grad_norm": 0.44745868192997307,
+      "learning_rate": 2.5642903796918583e-09,
+      "loss": 0.2586,
+      "step": 6237
+    },
+    {
+      "epoch": 2.980586583836091,
+      "grad_norm": 0.5017074416234102,
+      "learning_rate": 2.4464127189915975e-09,
+      "loss": 0.2563,
+      "step": 6238
+    },
+    {
+      "epoch": 2.981064452541664,
+      "grad_norm": 0.43904803227553557,
+      "learning_rate": 2.33130820330163e-09,
+      "loss": 0.2695,
+      "step": 6239
+    },
+    {
+      "epoch": 2.9815423212472374,
+      "grad_norm": 0.4428267317870827,
+      "learning_rate": 2.2189768645519693e-09,
+      "loss": 0.2783,
+      "step": 6240
+    },
+    {
+      "epoch": 2.9820201899528103,
+      "grad_norm": 0.44591251523595293,
+      "learning_rate": 2.109418733899915e-09,
+      "loss": 0.2721,
+      "step": 6241
+    },
+    {
+      "epoch": 2.9824980586583836,
+      "grad_norm": 0.4610652816160574,
+      "learning_rate": 2.0026338417344913e-09,
+      "loss": 0.2717,
+      "step": 6242
+    },
+    {
+      "epoch": 2.982975927363957,
+      "grad_norm": 0.4494233009141444,
+      "learning_rate": 1.898622217677559e-09,
+      "loss": 0.2636,
+      "step": 6243
+    },
+    {
+      "epoch": 2.9834537960695298,
+      "grad_norm": 0.4363101853354611,
+      "learning_rate": 1.7973838905793739e-09,
+      "loss": 0.2728,
+      "step": 6244
+    },
+    {
+      "epoch": 2.983931664775103,
+      "grad_norm": 0.4882459582895431,
+      "learning_rate": 1.6989188885219165e-09,
+      "loss": 0.2542,
+      "step": 6245
+    },
+    {
+      "epoch": 2.984409533480676,
+      "grad_norm": 0.44390851843666584,
+      "learning_rate": 1.6032272388166736e-09,
+      "loss": 0.2714,
+      "step": 6246
+    },
+    {
+      "epoch": 2.9848874021862493,
+      "grad_norm": 0.4429583465368995,
+      "learning_rate": 1.5103089680079674e-09,
+      "loss": 0.271,
+      "step": 6247
+    },
+    {
+      "epoch": 2.9853652708918226,
+      "grad_norm": 0.4302641158411793,
+      "learning_rate": 1.4201641018685152e-09,
+      "loss": 0.264,
+      "step": 6248
+    },
+    {
+      "epoch": 2.985843139597396,
+      "grad_norm": 0.4371029630970897,
+      "learning_rate": 1.3327926654049805e-09,
+      "loss": 0.2669,
+      "step": 6249
+    },
+    {
+      "epoch": 2.9863210083029688,
+      "grad_norm": 0.44144481030011334,
+      "learning_rate": 1.2481946828502011e-09,
+      "loss": 0.2598,
+      "step": 6250
+    },
+    {
+      "epoch": 2.9867988770085416,
+      "grad_norm": 0.5001018088911403,
+      "learning_rate": 1.1663701776709613e-09,
+      "loss": 0.2434,
+      "step": 6251
+    },
+    {
+      "epoch": 2.987276745714115,
+      "grad_norm": 0.430612900839865,
+      "learning_rate": 1.0873191725646604e-09,
+      "loss": 0.2798,
+      "step": 6252
+    },
+    {
+      "epoch": 2.9877546144196883,
+      "grad_norm": 0.45350891778339647,
+      "learning_rate": 1.0110416894593133e-09,
+      "loss": 0.2716,
+      "step": 6253
+    },
+    {
+      "epoch": 2.9882324831252616,
+      "grad_norm": 0.4337165441795163,
+      "learning_rate": 9.375377495102201e-10,
+      "loss": 0.254,
+      "step": 6254
+    },
+    {
+      "epoch": 2.9887103518308344,
+      "grad_norm": 0.4462339228646419,
+      "learning_rate": 8.668073731088467e-10,
+      "loss": 0.2599,
+      "step": 6255
+    },
+    {
+      "epoch": 2.9891882205364078,
+      "grad_norm": 0.45995272797479353,
+      "learning_rate": 7.988505798728341e-10,
+      "loss": 0.2673,
+      "step": 6256
+    },
+    {
+      "epoch": 2.9896660892419806,
+      "grad_norm": 0.43074554696329737,
+      "learning_rate": 7.33667388652659e-10,
+      "loss": 0.2644,
+      "step": 6257
+    },
+    {
+      "epoch": 2.990143957947554,
+      "grad_norm": 0.44281340320268414,
+      "learning_rate": 6.712578175294138e-10,
+      "loss": 0.2684,
+      "step": 6258
+    },
+    {
+      "epoch": 2.9906218266531273,
+      "grad_norm": 0.45094383291551116,
+      "learning_rate": 6.116218838148058e-10,
+      "loss": 0.2601,
+      "step": 6259
+    },
+    {
+      "epoch": 2.9910996953587,
+      "grad_norm": 0.4263274399474076,
+      "learning_rate": 5.547596040489378e-10,
+      "loss": 0.2628,
+      "step": 6260
+    },
+    {
+      "epoch": 2.9915775640642734,
+      "grad_norm": 0.44924007993713105,
+      "learning_rate": 5.006709940058585e-10,
+      "loss": 0.2646,
+      "step": 6261
+    },
+    {
+      "epoch": 2.9920554327698463,
+      "grad_norm": 0.4157503424555004,
+      "learning_rate": 4.4935606868912186e-10,
+      "loss": 0.2688,
+      "step": 6262
+    },
+    {
+      "epoch": 2.9925333014754196,
+      "grad_norm": 0.5530870038829894,
+      "learning_rate": 4.008148423306768e-10,
+      "loss": 0.261,
+      "step": 6263
+    },
+    {
+      "epoch": 2.993011170180993,
+      "grad_norm": 0.43360985730611284,
+      "learning_rate": 3.5504732839752867e-10,
+      "loss": 0.2605,
+      "step": 6264
+    },
+    {
+      "epoch": 2.993489038886566,
+      "grad_norm": 0.4305991915878331,
+      "learning_rate": 3.1205353958285724e-10,
+      "loss": 0.2565,
+      "step": 6265
+    },
+    {
+      "epoch": 2.993966907592139,
+      "grad_norm": 0.46592325654029976,
+      "learning_rate": 2.718334878137885e-10,
+      "loss": 0.2678,
+      "step": 6266
+    },
+    {
+      "epoch": 2.994444776297712,
+      "grad_norm": 0.44676400478930745,
+      "learning_rate": 2.3438718424473315e-10,
+      "loss": 0.2524,
+      "step": 6267
+    },
+    {
+      "epoch": 2.9949226450032853,
+      "grad_norm": 0.4574246654627526,
+      "learning_rate": 1.9971463926515833e-10,
+      "loss": 0.2602,
+      "step": 6268
+    },
+    {
+      "epoch": 2.9954005137088586,
+      "grad_norm": 0.45138653999661854,
+      "learning_rate": 1.6781586249070559e-10,
+      "loss": 0.2489,
+      "step": 6269
+    },
+    {
+      "epoch": 2.9958783824144315,
+      "grad_norm": 0.44396985604168704,
+      "learning_rate": 1.3869086276985243e-10,
+      "loss": 0.2679,
+      "step": 6270
+    },
+    {
+      "epoch": 2.996356251120005,
+      "grad_norm": 0.4756036796960099,
+      "learning_rate": 1.1233964818169185e-10,
+      "loss": 0.2624,
+      "step": 6271
+    },
+    {
+      "epoch": 2.9968341198255777,
+      "grad_norm": 0.46382667279220274,
+      "learning_rate": 8.876222603593221e-11,
+      "loss": 0.2646,
+      "step": 6272
+    },
+    {
+      "epoch": 2.997311988531151,
+      "grad_norm": 0.4785623678848972,
+      "learning_rate": 6.795860287178713e-11,
+      "loss": 0.2526,
+      "step": 6273
+    },
+    {
+      "epoch": 2.9977898572367243,
+      "grad_norm": 0.42404628861817545,
+      "learning_rate": 4.992878446019589e-11,
+      "loss": 0.2651,
+      "step": 6274
+    },
+    {
+      "epoch": 2.9982677259422976,
+      "grad_norm": 0.45160121782102897,
+      "learning_rate": 3.467277580271322e-11,
+      "loss": 0.2663,
+      "step": 6275
+    },
+    {
+      "epoch": 2.9987455946478705,
+      "grad_norm": 0.4354577405562673,
+      "learning_rate": 2.219058113039907e-11,
+      "loss": 0.2655,
+      "step": 6276
+    },
+    {
+      "epoch": 2.9992234633534434,
+      "grad_norm": 0.45460340307340963,
+      "learning_rate": 1.2482203904928824e-11,
+      "loss": 0.2514,
+      "step": 6277
+    },
+    {
+      "epoch": 2.9997013320590167,
+      "grad_norm": 0.47846568959514535,
+      "learning_rate": 5.547646820813768e-12,
+      "loss": 0.2667,
+      "step": 6278
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.5483148609921531,
+      "learning_rate": 1.3869118009601778e-12,
+      "loss": 0.2654,
+      "step": 6279
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 6279,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 800,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5031825133207552e+16,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}