diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,12026 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.9992701795358343,
+  "eval_steps": 500,
+  "global_step": 1712,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0011677127426653043,
+      "grad_norm": 0.27489905649053786,
+      "learning_rate": 1.1627906976744187e-07,
+      "loss": 0.631,
+      "step": 1
+    },
+    {
+      "epoch": 0.0023354254853306085,
+      "grad_norm": 0.3002081158101649,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 0.6812,
+      "step": 2
+    },
+    {
+      "epoch": 0.003503138227995913,
+      "grad_norm": 0.30123323572030125,
+      "learning_rate": 3.488372093023256e-07,
+      "loss": 0.6678,
+      "step": 3
+    },
+    {
+      "epoch": 0.004670850970661217,
+      "grad_norm": 0.3026509859717456,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 0.6703,
+      "step": 4
+    },
+    {
+      "epoch": 0.005838563713326522,
+      "grad_norm": 0.30132536116859593,
+      "learning_rate": 5.813953488372094e-07,
+      "loss": 0.6978,
+      "step": 5
+    },
+    {
+      "epoch": 0.007006276455991826,
+      "grad_norm": 0.294683882656284,
+      "learning_rate": 6.976744186046513e-07,
+      "loss": 0.6572,
+      "step": 6
+    },
+    {
+      "epoch": 0.00817398919865713,
+      "grad_norm": 0.305042408470332,
+      "learning_rate": 8.139534883720931e-07,
+      "loss": 0.6958,
+      "step": 7
+    },
+    {
+      "epoch": 0.009341701941322434,
+      "grad_norm": 0.2813949136159455,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 0.6734,
+      "step": 8
+    },
+    {
+      "epoch": 0.01050941468398774,
+      "grad_norm": 0.2891804825962349,
+      "learning_rate": 1.0465116279069768e-06,
+      "loss": 0.6766,
+      "step": 9
+    },
+    {
+      "epoch": 0.011677127426653043,
+      "grad_norm": 0.2600379768191449,
+      "learning_rate": 1.1627906976744188e-06,
+      "loss": 0.6502,
+      "step": 10
+    },
+    {
+      "epoch": 0.012844840169318347,
+      "grad_norm": 0.24258637824258455,
+      "learning_rate": 1.2790697674418605e-06,
+      "loss": 0.6485,
+      "step": 11
+    },
+    {
+      "epoch": 0.014012552911983653,
+      "grad_norm": 0.2527957872051737,
+      "learning_rate": 1.3953488372093025e-06,
+      "loss": 0.6536,
+      "step": 12
+    },
+    {
+      "epoch": 0.015180265654648957,
+      "grad_norm": 0.22147371373934885,
+      "learning_rate": 1.5116279069767443e-06,
+      "loss": 0.6312,
+      "step": 13
+    },
+    {
+      "epoch": 0.01634797839731426,
+      "grad_norm": 0.17981721298479086,
+      "learning_rate": 1.6279069767441862e-06,
+      "loss": 0.6692,
+      "step": 14
+    },
+    {
+      "epoch": 0.017515691139979566,
+      "grad_norm": 0.16875090165700557,
+      "learning_rate": 1.7441860465116282e-06,
+      "loss": 0.6351,
+      "step": 15
+    },
+    {
+      "epoch": 0.018683403882644868,
+      "grad_norm": 0.1489352384212706,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 0.6147,
+      "step": 16
+    },
+    {
+      "epoch": 0.019851116625310174,
+      "grad_norm": 0.14891944582759548,
+      "learning_rate": 1.976744186046512e-06,
+      "loss": 0.5939,
+      "step": 17
+    },
+    {
+      "epoch": 0.02101882936797548,
+      "grad_norm": 0.13597403739651417,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 0.5637,
+      "step": 18
+    },
+    {
+      "epoch": 0.02218654211064078,
+      "grad_norm": 0.1682171093110703,
+      "learning_rate": 2.2093023255813954e-06,
+      "loss": 0.5743,
+      "step": 19
+    },
+    {
+      "epoch": 0.023354254853306087,
+      "grad_norm": 0.22048955166068598,
+      "learning_rate": 2.3255813953488376e-06,
+      "loss": 0.5799,
+      "step": 20
+    },
+    {
+      "epoch": 0.024521967595971392,
+      "grad_norm": 0.2512676698549623,
+      "learning_rate": 2.4418604651162793e-06,
+      "loss": 0.6103,
+      "step": 21
+    },
+    {
+      "epoch": 0.025689680338636695,
+      "grad_norm": 0.2326511363321868,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 0.6048,
+      "step": 22
+    },
+    {
+      "epoch": 0.026857393081302,
+      "grad_norm": 0.17394073669804522,
+      "learning_rate": 2.674418604651163e-06,
+      "loss": 0.5454,
+      "step": 23
+    },
+    {
+      "epoch": 0.028025105823967306,
+      "grad_norm": 0.1545885757011556,
+      "learning_rate": 2.790697674418605e-06,
+      "loss": 0.5652,
+      "step": 24
+    },
+    {
+      "epoch": 0.029192818566632608,
+      "grad_norm": 0.1317183484673638,
+      "learning_rate": 2.9069767441860468e-06,
+      "loss": 0.5434,
+      "step": 25
+    },
+    {
+      "epoch": 0.030360531309297913,
+      "grad_norm": 0.12985660447351066,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 0.5372,
+      "step": 26
+    },
+    {
+      "epoch": 0.031528244051963215,
+      "grad_norm": 0.14013103073276897,
+      "learning_rate": 3.1395348837209307e-06,
+      "loss": 0.538,
+      "step": 27
+    },
+    {
+      "epoch": 0.03269595679462852,
+      "grad_norm": 0.1423346396983896,
+      "learning_rate": 3.2558139534883724e-06,
+      "loss": 0.5353,
+      "step": 28
+    },
+    {
+      "epoch": 0.033863669537293827,
+      "grad_norm": 0.14305875045372807,
+      "learning_rate": 3.372093023255814e-06,
+      "loss": 0.5181,
+      "step": 29
+    },
+    {
+      "epoch": 0.03503138227995913,
+      "grad_norm": 0.14414789108179643,
+      "learning_rate": 3.4883720930232564e-06,
+      "loss": 0.5205,
+      "step": 30
+    },
+    {
+      "epoch": 0.03619909502262444,
+      "grad_norm": 0.1444695727383064,
+      "learning_rate": 3.6046511627906977e-06,
+      "loss": 0.5422,
+      "step": 31
+    },
+    {
+      "epoch": 0.037366807765289736,
+      "grad_norm": 0.13270530407146588,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 0.5346,
+      "step": 32
+    },
+    {
+      "epoch": 0.03853452050795504,
+      "grad_norm": 0.11924962867391276,
+      "learning_rate": 3.837209302325582e-06,
+      "loss": 0.5264,
+      "step": 33
+    },
+    {
+      "epoch": 0.03970223325062035,
+      "grad_norm": 0.1254669739349559,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 0.5325,
+      "step": 34
+    },
+    {
+      "epoch": 0.04086994599328565,
+      "grad_norm": 0.11588881071519164,
+      "learning_rate": 4.0697674418604655e-06,
+      "loss": 0.5204,
+      "step": 35
+    },
+    {
+      "epoch": 0.04203765873595096,
+      "grad_norm": 0.11612741536743643,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 0.5323,
+      "step": 36
+    },
+    {
+      "epoch": 0.04320537147861626,
+      "grad_norm": 0.11277544470894416,
+      "learning_rate": 4.302325581395349e-06,
+      "loss": 0.5371,
+      "step": 37
+    },
+    {
+      "epoch": 0.04437308422128156,
+      "grad_norm": 0.11658226409363799,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 0.5323,
+      "step": 38
+    },
+    {
+      "epoch": 0.04554079696394687,
+      "grad_norm": 0.11339302601868977,
+      "learning_rate": 4.5348837209302326e-06,
+      "loss": 0.5089,
+      "step": 39
+    },
+    {
+      "epoch": 0.046708509706612174,
+      "grad_norm": 0.11106477703435136,
+      "learning_rate": 4.651162790697675e-06,
+      "loss": 0.5009,
+      "step": 40
+    },
+    {
+      "epoch": 0.04787622244927748,
+      "grad_norm": 0.10280695467528929,
+      "learning_rate": 4.767441860465117e-06,
+      "loss": 0.5146,
+      "step": 41
+    },
+    {
+      "epoch": 0.049043935191942785,
+      "grad_norm": 0.09838395364592965,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 0.517,
+      "step": 42
+    },
+    {
+      "epoch": 0.050211647934608084,
+      "grad_norm": 0.11528460300602575,
+      "learning_rate": 5e-06,
+      "loss": 0.5008,
+      "step": 43
+    },
+    {
+      "epoch": 0.05137936067727339,
+      "grad_norm": 0.11343300457122962,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 0.523,
+      "step": 44
+    },
+    {
+      "epoch": 0.052547073419938695,
+      "grad_norm": 0.11608938732902359,
+      "learning_rate": 5.232558139534885e-06,
+      "loss": 0.5355,
+      "step": 45
+    },
+    {
+      "epoch": 0.053714786162604,
+      "grad_norm": 0.09884240623054323,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 0.5028,
+      "step": 46
+    },
+    {
+      "epoch": 0.054882498905269306,
+      "grad_norm": 0.1001086141318868,
+      "learning_rate": 5.465116279069767e-06,
+      "loss": 0.4977,
+      "step": 47
+    },
+    {
+      "epoch": 0.05605021164793461,
+      "grad_norm": 0.09566641743297358,
+      "learning_rate": 5.58139534883721e-06,
+      "loss": 0.5042,
+      "step": 48
+    },
+    {
+      "epoch": 0.05721792439059991,
+      "grad_norm": 0.10364526295176196,
+      "learning_rate": 5.697674418604652e-06,
+      "loss": 0.4975,
+      "step": 49
+    },
+    {
+      "epoch": 0.058385637133265215,
+      "grad_norm": 0.0990673149132162,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 0.469,
+      "step": 50
+    },
+    {
+      "epoch": 0.05955334987593052,
+      "grad_norm": 0.09340356671463913,
+      "learning_rate": 5.930232558139536e-06,
+      "loss": 0.4764,
+      "step": 51
+    },
+    {
+      "epoch": 0.06072106261859583,
+      "grad_norm": 0.09410495022122617,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 0.4559,
+      "step": 52
+    },
+    {
+      "epoch": 0.06188877536126113,
+      "grad_norm": 0.09581826339031228,
+      "learning_rate": 6.162790697674419e-06,
+      "loss": 0.4827,
+      "step": 53
+    },
+    {
+      "epoch": 0.06305648810392643,
+      "grad_norm": 0.09879803731776882,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 0.4847,
+      "step": 54
+    },
+    {
+      "epoch": 0.06422420084659174,
+      "grad_norm": 0.09553677635956338,
+      "learning_rate": 6.395348837209303e-06,
+      "loss": 0.4568,
+      "step": 55
+    },
+    {
+      "epoch": 0.06539191358925704,
+      "grad_norm": 0.09537021737049738,
+      "learning_rate": 6.511627906976745e-06,
+      "loss": 0.4892,
+      "step": 56
+    },
+    {
+      "epoch": 0.06655962633192235,
+      "grad_norm": 0.09420990012243977,
+      "learning_rate": 6.627906976744186e-06,
+      "loss": 0.4512,
+      "step": 57
+    },
+    {
+      "epoch": 0.06772733907458765,
+      "grad_norm": 0.10041083370194115,
+      "learning_rate": 6.744186046511628e-06,
+      "loss": 0.472,
+      "step": 58
+    },
+    {
+      "epoch": 0.06889505181725296,
+      "grad_norm": 0.09394990707163217,
+      "learning_rate": 6.86046511627907e-06,
+      "loss": 0.45,
+      "step": 59
+    },
+    {
+      "epoch": 0.07006276455991826,
+      "grad_norm": 0.0970464273965833,
+      "learning_rate": 6.976744186046513e-06,
+      "loss": 0.4442,
+      "step": 60
+    },
+    {
+      "epoch": 0.07123047730258357,
+      "grad_norm": 0.09369474005080067,
+      "learning_rate": 7.0930232558139545e-06,
+      "loss": 0.4687,
+      "step": 61
+    },
+    {
+      "epoch": 0.07239819004524888,
+      "grad_norm": 0.09621367014500805,
+      "learning_rate": 7.209302325581395e-06,
+      "loss": 0.4799,
+      "step": 62
+    },
+    {
+      "epoch": 0.07356590278791417,
+      "grad_norm": 0.09878100583636068,
+      "learning_rate": 7.325581395348837e-06,
+      "loss": 0.4652,
+      "step": 63
+    },
+    {
+      "epoch": 0.07473361553057947,
+      "grad_norm": 0.09003758494196815,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 0.4665,
+      "step": 64
+    },
+    {
+      "epoch": 0.07590132827324478,
+      "grad_norm": 0.0973879400329809,
+      "learning_rate": 7.5581395348837215e-06,
+      "loss": 0.4635,
+      "step": 65
+    },
+    {
+      "epoch": 0.07706904101591008,
+      "grad_norm": 0.08975823517688108,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 0.4298,
+      "step": 66
+    },
+    {
+      "epoch": 0.07823675375857539,
+      "grad_norm": 0.08779449241675477,
+      "learning_rate": 7.790697674418605e-06,
+      "loss": 0.4404,
+      "step": 67
+    },
+    {
+      "epoch": 0.0794044665012407,
+      "grad_norm": 0.09894574408114423,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 0.4602,
+      "step": 68
+    },
+    {
+      "epoch": 0.080572179243906,
+      "grad_norm": 0.09713895385738276,
+      "learning_rate": 8.023255813953488e-06,
+      "loss": 0.4905,
+      "step": 69
+    },
+    {
+      "epoch": 0.0817398919865713,
+      "grad_norm": 0.09366233594929364,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 0.4832,
+      "step": 70
+    },
+    {
+      "epoch": 0.08290760472923661,
+      "grad_norm": 0.09087391160967755,
+      "learning_rate": 8.255813953488374e-06,
+      "loss": 0.4436,
+      "step": 71
+    },
+    {
+      "epoch": 0.08407531747190192,
+      "grad_norm": 0.09750853544820579,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 0.4672,
+      "step": 72
+    },
+    {
+      "epoch": 0.08524303021456722,
+      "grad_norm": 0.0942327123850502,
+      "learning_rate": 8.488372093023256e-06,
+      "loss": 0.4475,
+      "step": 73
+    },
+    {
+      "epoch": 0.08641074295723251,
+      "grad_norm": 0.09207783593915118,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 0.4189,
+      "step": 74
+    },
+    {
+      "epoch": 0.08757845569989782,
+      "grad_norm": 0.09600196999330328,
+      "learning_rate": 8.72093023255814e-06,
+      "loss": 0.444,
+      "step": 75
+    },
+    {
+      "epoch": 0.08874616844256313,
+      "grad_norm": 0.09931819707718452,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 0.4644,
+      "step": 76
+    },
+    {
+      "epoch": 0.08991388118522843,
+      "grad_norm": 0.10192856859444618,
+      "learning_rate": 8.953488372093024e-06,
+      "loss": 0.4556,
+      "step": 77
+    },
+    {
+      "epoch": 0.09108159392789374,
+      "grad_norm": 0.09888001941576431,
+      "learning_rate": 9.069767441860465e-06,
+      "loss": 0.4418,
+      "step": 78
+    },
+    {
+      "epoch": 0.09224930667055904,
+      "grad_norm": 0.09578560178157478,
+      "learning_rate": 9.186046511627908e-06,
+      "loss": 0.4391,
+      "step": 79
+    },
+    {
+      "epoch": 0.09341701941322435,
+      "grad_norm": 0.0973639381488298,
+      "learning_rate": 9.30232558139535e-06,
+      "loss": 0.4428,
+      "step": 80
+    },
+    {
+      "epoch": 0.09458473215588965,
+      "grad_norm": 0.10111860187732441,
+      "learning_rate": 9.418604651162791e-06,
+      "loss": 0.454,
+      "step": 81
+    },
+    {
+      "epoch": 0.09575244489855496,
+      "grad_norm": 0.09818228333593186,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 0.4348,
+      "step": 82
+    },
+    {
+      "epoch": 0.09692015764122026,
+      "grad_norm": 0.09643310718833838,
+      "learning_rate": 9.651162790697676e-06,
+      "loss": 0.4234,
+      "step": 83
+    },
+    {
+      "epoch": 0.09808787038388557,
+      "grad_norm": 0.0928146656143695,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 0.4344,
+      "step": 84
+    },
+    {
+      "epoch": 0.09925558312655088,
+      "grad_norm": 0.09848808838050511,
+      "learning_rate": 9.883720930232558e-06,
+      "loss": 0.4262,
+      "step": 85
+    },
+    {
+      "epoch": 0.10042329586921617,
+      "grad_norm": 0.09747673981955855,
+      "learning_rate": 1e-05,
+      "loss": 0.4375,
+      "step": 86
+    },
+    {
+      "epoch": 0.10159100861188147,
+      "grad_norm": 0.10552144171132115,
+      "learning_rate": 1.0116279069767443e-05,
+      "loss": 0.4776,
+      "step": 87
+    },
+    {
+      "epoch": 0.10275872135454678,
+      "grad_norm": 0.10003258728020441,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 0.4479,
+      "step": 88
+    },
+    {
+      "epoch": 0.10392643409721208,
+      "grad_norm": 0.10589150846191288,
+      "learning_rate": 1.0348837209302327e-05,
+      "loss": 0.473,
+      "step": 89
+    },
+    {
+      "epoch": 0.10509414683987739,
+      "grad_norm": 0.09539877426500698,
+      "learning_rate": 1.046511627906977e-05,
+      "loss": 0.4427,
+      "step": 90
+    },
+    {
+      "epoch": 0.1062618595825427,
+      "grad_norm": 0.10406562263646357,
+      "learning_rate": 1.058139534883721e-05,
+      "loss": 0.4746,
+      "step": 91
+    },
+    {
+      "epoch": 0.107429572325208,
+      "grad_norm": 0.09279493646767503,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 0.4278,
+      "step": 92
+    },
+    {
+      "epoch": 0.1085972850678733,
+      "grad_norm": 0.0913964913701737,
+      "learning_rate": 1.0813953488372094e-05,
+      "loss": 0.4264,
+      "step": 93
+    },
+    {
+      "epoch": 0.10976499781053861,
+      "grad_norm": 0.09545957291163429,
+      "learning_rate": 1.0930232558139535e-05,
+      "loss": 0.4546,
+      "step": 94
+    },
+    {
+      "epoch": 0.11093271055320392,
+      "grad_norm": 0.09970766962453986,
+      "learning_rate": 1.1046511627906977e-05,
+      "loss": 0.4419,
+      "step": 95
+    },
+    {
+      "epoch": 0.11210042329586922,
+      "grad_norm": 0.10432622085615005,
+      "learning_rate": 1.116279069767442e-05,
+      "loss": 0.4443,
+      "step": 96
+    },
+    {
+      "epoch": 0.11326813603853451,
+      "grad_norm": 0.09797100637498217,
+      "learning_rate": 1.1279069767441861e-05,
+      "loss": 0.4524,
+      "step": 97
+    },
+    {
+      "epoch": 0.11443584878119982,
+      "grad_norm": 0.09644245191577892,
+      "learning_rate": 1.1395348837209304e-05,
+      "loss": 0.446,
+      "step": 98
+    },
+    {
+      "epoch": 0.11560356152386513,
+      "grad_norm": 0.09764097742514231,
+      "learning_rate": 1.1511627906976746e-05,
+      "loss": 0.4198,
+      "step": 99
+    },
+    {
+      "epoch": 0.11677127426653043,
+      "grad_norm": 0.09674354198963808,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 0.4431,
+      "step": 100
+    },
+    {
+      "epoch": 0.11793898700919574,
+      "grad_norm": 0.09624286743610701,
+      "learning_rate": 1.174418604651163e-05,
+      "loss": 0.4337,
+      "step": 101
+    },
+    {
+      "epoch": 0.11910669975186104,
+      "grad_norm": 0.09449582338868262,
+      "learning_rate": 1.1860465116279072e-05,
+      "loss": 0.4478,
+      "step": 102
+    },
+    {
+      "epoch": 0.12027441249452635,
+      "grad_norm": 0.09674608034883651,
+      "learning_rate": 1.1976744186046511e-05,
+      "loss": 0.4479,
+      "step": 103
+    },
+    {
+      "epoch": 0.12144212523719165,
+      "grad_norm": 0.1005221754976316,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 0.4337,
+      "step": 104
+    },
+    {
+      "epoch": 0.12260983797985696,
+      "grad_norm": 0.098155003246203,
+      "learning_rate": 1.2209302325581397e-05,
+      "loss": 0.4447,
+      "step": 105
+    },
+    {
+      "epoch": 0.12377755072252226,
+      "grad_norm": 0.09193234247006324,
+      "learning_rate": 1.2325581395348838e-05,
+      "loss": 0.4159,
+      "step": 106
+    },
+    {
+      "epoch": 0.12494526346518757,
+      "grad_norm": 0.10544468744974621,
+      "learning_rate": 1.244186046511628e-05,
+      "loss": 0.4351,
+      "step": 107
+    },
+    {
+      "epoch": 0.12611297620785286,
+      "grad_norm": 0.10283312702294119,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 0.4481,
+      "step": 108
+    },
+    {
+      "epoch": 0.12728068895051817,
+      "grad_norm": 0.09682165364990364,
+      "learning_rate": 1.2674418604651164e-05,
+      "loss": 0.4367,
+      "step": 109
+    },
+    {
+      "epoch": 0.12844840169318347,
+      "grad_norm": 0.09713815959635383,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 0.4263,
+      "step": 110
+    },
+    {
+      "epoch": 0.12961611443584878,
+      "grad_norm": 0.08974739075992023,
+      "learning_rate": 1.2906976744186049e-05,
+      "loss": 0.4368,
+      "step": 111
+    },
+    {
+      "epoch": 0.13078382717851408,
+      "grad_norm": 0.09660438859734574,
+      "learning_rate": 1.302325581395349e-05,
+      "loss": 0.4434,
+      "step": 112
+    },
+    {
+      "epoch": 0.1319515399211794,
+      "grad_norm": 0.10534146467381685,
+      "learning_rate": 1.313953488372093e-05,
+      "loss": 0.445,
+      "step": 113
+    },
+    {
+      "epoch": 0.1331192526638447,
+      "grad_norm": 0.0990783619992381,
+      "learning_rate": 1.3255813953488372e-05,
+      "loss": 0.436,
+      "step": 114
+    },
+    {
+      "epoch": 0.13428696540651,
+      "grad_norm": 0.10424539047183187,
+      "learning_rate": 1.3372093023255814e-05,
+      "loss": 0.4699,
+      "step": 115
+    },
+    {
+      "epoch": 0.1354546781491753,
+      "grad_norm": 0.10159924717419655,
+      "learning_rate": 1.3488372093023257e-05,
+      "loss": 0.4251,
+      "step": 116
+    },
+    {
+      "epoch": 0.1366223908918406,
+      "grad_norm": 0.0986296971994857,
+      "learning_rate": 1.36046511627907e-05,
+      "loss": 0.4461,
+      "step": 117
+    },
+    {
+      "epoch": 0.13779010363450592,
+      "grad_norm": 0.11030616394396812,
+      "learning_rate": 1.372093023255814e-05,
+      "loss": 0.4368,
+      "step": 118
+    },
+    {
+      "epoch": 0.13895781637717122,
+      "grad_norm": 0.10620516678475647,
+      "learning_rate": 1.3837209302325583e-05,
+      "loss": 0.4192,
+      "step": 119
+    },
+    {
+      "epoch": 0.14012552911983653,
+      "grad_norm": 0.09906834642313171,
+      "learning_rate": 1.3953488372093025e-05,
+      "loss": 0.4449,
+      "step": 120
+    },
+    {
+      "epoch": 0.14129324186250183,
+      "grad_norm": 0.1039848685789375,
+      "learning_rate": 1.4069767441860466e-05,
+      "loss": 0.4453,
+      "step": 121
+    },
+    {
+      "epoch": 0.14246095460516714,
+      "grad_norm": 0.10469370044549552,
+      "learning_rate": 1.4186046511627909e-05,
+      "loss": 0.4404,
+      "step": 122
+    },
+    {
+      "epoch": 0.14362866734783244,
+      "grad_norm": 0.10130041523750288,
+      "learning_rate": 1.4302325581395352e-05,
+      "loss": 0.4142,
+      "step": 123
+    },
+    {
+      "epoch": 0.14479638009049775,
+      "grad_norm": 0.11110027771249799,
+      "learning_rate": 1.441860465116279e-05,
+      "loss": 0.4441,
+      "step": 124
+    },
+    {
+      "epoch": 0.14596409283316303,
+      "grad_norm": 0.10154109292155558,
+      "learning_rate": 1.4534883720930233e-05,
+      "loss": 0.4658,
+      "step": 125
+    },
+    {
+      "epoch": 0.14713180557582833,
+      "grad_norm": 0.11333352717260238,
+      "learning_rate": 1.4651162790697674e-05,
+      "loss": 0.443,
+      "step": 126
+    },
+    {
+      "epoch": 0.14829951831849364,
+      "grad_norm": 0.1064927682637495,
+      "learning_rate": 1.4767441860465117e-05,
+      "loss": 0.4545,
+      "step": 127
+    },
+    {
+      "epoch": 0.14946723106115894,
+      "grad_norm": 0.09866877090414088,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 0.4311,
+      "step": 128
+    },
+    {
+      "epoch": 0.15063494380382425,
+      "grad_norm": 0.11226465612406128,
+      "learning_rate": 1.5000000000000002e-05,
+      "loss": 0.4381,
+      "step": 129
+    },
+    {
+      "epoch": 0.15180265654648956,
+      "grad_norm": 0.11177095941474945,
+      "learning_rate": 1.5116279069767443e-05,
+      "loss": 0.436,
+      "step": 130
+    },
+    {
+      "epoch": 0.15297036928915486,
+      "grad_norm": 0.09333418281065686,
+      "learning_rate": 1.5232558139534886e-05,
+      "loss": 0.4124,
+      "step": 131
+    },
+    {
+      "epoch": 0.15413808203182017,
+      "grad_norm": 0.1067375115142491,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 0.4598,
+      "step": 132
+    },
+    {
+      "epoch": 0.15530579477448547,
+      "grad_norm": 0.10694070997152025,
+      "learning_rate": 1.546511627906977e-05,
+      "loss": 0.4286,
+      "step": 133
+    },
+    {
+      "epoch": 0.15647350751715078,
+      "grad_norm": 0.10621764251187082,
+      "learning_rate": 1.558139534883721e-05,
+      "loss": 0.4443,
+      "step": 134
+    },
+    {
+      "epoch": 0.15764122025981608,
+      "grad_norm": 0.10855297385360052,
+      "learning_rate": 1.569767441860465e-05,
+      "loss": 0.4705,
+      "step": 135
+    },
+    {
+      "epoch": 0.1588089330024814,
+      "grad_norm": 0.10650028577652582,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 0.4348,
+      "step": 136
+    },
+    {
+      "epoch": 0.1599766457451467,
+      "grad_norm": 0.10920991734815626,
+      "learning_rate": 1.5930232558139536e-05,
+      "loss": 0.411,
+      "step": 137
+    },
+    {
+      "epoch": 0.161144358487812,
+      "grad_norm": 0.10186573494914933,
+      "learning_rate": 1.6046511627906977e-05,
+      "loss": 0.447,
+      "step": 138
+    },
+    {
+      "epoch": 0.1623120712304773,
+      "grad_norm": 0.09976129415551556,
+      "learning_rate": 1.616279069767442e-05,
+      "loss": 0.4399,
+      "step": 139
+    },
+    {
+      "epoch": 0.1634797839731426,
+      "grad_norm": 0.10079858822890186,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 0.4214,
+      "step": 140
+    },
+    {
+      "epoch": 0.16464749671580792,
+      "grad_norm": 0.10026942907474212,
+      "learning_rate": 1.6395348837209303e-05,
+      "loss": 0.4238,
+      "step": 141
+    },
+    {
+      "epoch": 0.16581520945847322,
+      "grad_norm": 0.1028581428507134,
+      "learning_rate": 1.6511627906976747e-05,
+      "loss": 0.4471,
+      "step": 142
+    },
+    {
+      "epoch": 0.16698292220113853,
+      "grad_norm": 0.09093941041384292,
+      "learning_rate": 1.6627906976744188e-05,
+      "loss": 0.4404,
+      "step": 143
+    },
+    {
+      "epoch": 0.16815063494380383,
+      "grad_norm": 0.09919462220097262,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 0.4176,
+      "step": 144
+    },
+    {
+      "epoch": 0.16931834768646914,
+      "grad_norm": 0.09954492034605865,
+      "learning_rate": 1.6860465116279073e-05,
+      "loss": 0.4177,
+      "step": 145
+    },
+    {
+      "epoch": 0.17048606042913445,
+      "grad_norm": 0.09074251170562353,
+      "learning_rate": 1.697674418604651e-05,
+      "loss": 0.4327,
+      "step": 146
+    },
+    {
+      "epoch": 0.17165377317179975,
+      "grad_norm": 0.09732991341861935,
+      "learning_rate": 1.7093023255813955e-05,
+      "loss": 0.4304,
+      "step": 147
+    },
+    {
+      "epoch": 0.17282148591446503,
+      "grad_norm": 0.09429411714933719,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 0.4175,
+      "step": 148
+    },
+    {
+      "epoch": 0.17398919865713033,
+      "grad_norm": 0.09479678341363165,
+      "learning_rate": 1.7325581395348837e-05,
+      "loss": 0.412,
+      "step": 149
+    },
+    {
+      "epoch": 0.17515691139979564,
+      "grad_norm": 0.09255974169017858,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 0.4255,
+      "step": 150
+    },
+    {
+      "epoch": 0.17632462414246095,
+      "grad_norm": 0.09485466511118304,
+      "learning_rate": 1.7558139534883722e-05,
+      "loss": 0.4277,
+      "step": 151
+    },
+    {
+      "epoch": 0.17749233688512625,
+      "grad_norm": 0.1002950623051089,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 0.4494,
+      "step": 152
+    },
+    {
+      "epoch": 0.17866004962779156,
+      "grad_norm": 0.10222797710307821,
+      "learning_rate": 1.7790697674418608e-05,
+      "loss": 0.4453,
+      "step": 153
+    },
+    {
+      "epoch": 0.17982776237045686,
+      "grad_norm": 0.10106869583748171,
+      "learning_rate": 1.790697674418605e-05,
+      "loss": 0.4314,
+      "step": 154
+    },
+    {
+      "epoch": 0.18099547511312217,
+      "grad_norm": 0.10132745622185374,
+      "learning_rate": 1.802325581395349e-05,
+      "loss": 0.4161,
+      "step": 155
+    },
+    {
+      "epoch": 0.18216318785578747,
+      "grad_norm": 0.09808649363686031,
+      "learning_rate": 1.813953488372093e-05,
+      "loss": 0.4291,
+      "step": 156
+    },
+    {
+      "epoch": 0.18333090059845278,
+      "grad_norm": 0.10407683369487718,
+      "learning_rate": 1.825581395348837e-05,
+      "loss": 0.4482,
+      "step": 157
+    },
+    {
+      "epoch": 0.18449861334111808,
+      "grad_norm": 0.10381895568762513,
+      "learning_rate": 1.8372093023255815e-05,
+      "loss": 0.4516,
+      "step": 158
+    },
+    {
+      "epoch": 0.1856663260837834,
+      "grad_norm": 0.10279893804615373,
+      "learning_rate": 1.8488372093023256e-05,
+      "loss": 0.4381,
+      "step": 159
+    },
+    {
+      "epoch": 0.1868340388264487,
+      "grad_norm": 0.11149927178693692,
+      "learning_rate": 1.86046511627907e-05,
+      "loss": 0.4352,
+      "step": 160
+    },
+    {
+      "epoch": 0.188001751569114,
+      "grad_norm": 0.0968988039324055,
+      "learning_rate": 1.872093023255814e-05,
+      "loss": 0.4129,
+      "step": 161
+    },
+    {
+      "epoch": 0.1891694643117793,
+      "grad_norm": 0.09985219430310796,
+      "learning_rate": 1.8837209302325582e-05,
+      "loss": 0.4135,
+      "step": 162
+    },
+    {
+      "epoch": 0.1903371770544446,
+      "grad_norm": 0.10671578061758787,
+      "learning_rate": 1.8953488372093027e-05,
+      "loss": 0.4575,
+      "step": 163
+    },
+    {
+      "epoch": 0.19150488979710992,
+      "grad_norm": 0.09609715606866423,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 0.4375,
+      "step": 164
+    },
+    {
+      "epoch": 0.19267260253977522,
+      "grad_norm": 0.11823431761063469,
+      "learning_rate": 1.918604651162791e-05,
+      "loss": 0.4472,
+      "step": 165
+    },
+    {
+      "epoch": 0.19384031528244053,
+      "grad_norm": 0.10242249314362076,
+      "learning_rate": 1.9302325581395353e-05,
+      "loss": 0.4338,
+      "step": 166
+    },
+    {
+      "epoch": 0.19500802802510583,
+      "grad_norm": 0.11524647701771196,
+      "learning_rate": 1.941860465116279e-05,
+      "loss": 0.4468,
+      "step": 167
+    },
+    {
+      "epoch": 0.19617574076777114,
+      "grad_norm": 0.1091127679766114,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 0.427,
+      "step": 168
+    },
+    {
+      "epoch": 0.19734345351043645,
+      "grad_norm": 0.10257518847123905,
+      "learning_rate": 1.9651162790697676e-05,
+      "loss": 0.4121,
+      "step": 169
+    },
+    {
+      "epoch": 0.19851116625310175,
+      "grad_norm": 0.10628468265607431,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 0.447,
+      "step": 170
+    },
+    {
+      "epoch": 0.19967887899576703,
+      "grad_norm": 0.11853888107651372,
+      "learning_rate": 1.988372093023256e-05,
+      "loss": 0.4273,
+      "step": 171
+    },
+    {
+      "epoch": 0.20084659173843233,
+      "grad_norm": 0.09974423178515199,
+      "learning_rate": 2e-05,
+      "loss": 0.4378,
+      "step": 172
+    },
+    {
+      "epoch": 0.20201430448109764,
+      "grad_norm": 0.1034195474480665,
+      "learning_rate": 1.9999979192104532e-05,
+      "loss": 0.4246,
+      "step": 173
+    },
+    {
+      "epoch": 0.20318201722376295,
+      "grad_norm": 0.1040167970813603,
+      "learning_rate": 1.9999916768504724e-05,
+      "loss": 0.4491,
+      "step": 174
+    },
+    {
+      "epoch": 0.20434972996642825,
+      "grad_norm": 0.10724936751758436,
+      "learning_rate": 1.9999812729460353e-05,
+      "loss": 0.4682,
+      "step": 175
+    },
+    {
+      "epoch": 0.20551744270909356,
+      "grad_norm": 0.09913566325501685,
+      "learning_rate": 1.9999667075404385e-05,
+      "loss": 0.4305,
+      "step": 176
+    },
+    {
+      "epoch": 0.20668515545175886,
+      "grad_norm": 0.10432510832237837,
+      "learning_rate": 1.9999479806942976e-05,
+      "loss": 0.4453,
+      "step": 177
+    },
+    {
+      "epoch": 0.20785286819442417,
+      "grad_norm": 0.10696733478265621,
+      "learning_rate": 1.9999250924855458e-05,
+      "loss": 0.4535,
+      "step": 178
+    },
+    {
+      "epoch": 0.20902058093708947,
+      "grad_norm": 0.09589172725379992,
+      "learning_rate": 1.9998980430094333e-05,
+      "loss": 0.4213,
+      "step": 179
+    },
+    {
+      "epoch": 0.21018829367975478,
+      "grad_norm": 0.09036057842101787,
+      "learning_rate": 1.9998668323785298e-05,
+      "loss": 0.4042,
+      "step": 180
+    },
+    {
+      "epoch": 0.21135600642242008,
+      "grad_norm": 0.10196337296499106,
+      "learning_rate": 1.9998314607227203e-05,
+      "loss": 0.4295,
+      "step": 181
+    },
+    {
+      "epoch": 0.2125237191650854,
+      "grad_norm": 0.09782861605007927,
+      "learning_rate": 1.9997919281892066e-05,
+      "loss": 0.4188,
+      "step": 182
+    },
+    {
+      "epoch": 0.2136914319077507,
+      "grad_norm": 0.09437384014374631,
+      "learning_rate": 1.999748234942507e-05,
+      "loss": 0.435,
+      "step": 183
+    },
+    {
+      "epoch": 0.214859144650416,
+      "grad_norm": 0.10102324295097412,
+      "learning_rate": 1.9997003811644534e-05,
+      "loss": 0.4397,
+      "step": 184
+    },
+    {
+      "epoch": 0.2160268573930813,
+      "grad_norm": 0.0980705802427771,
+      "learning_rate": 1.999648367054194e-05,
+      "loss": 0.4285,
+      "step": 185
+    },
+    {
+      "epoch": 0.2171945701357466,
+      "grad_norm": 0.0995167402248456,
+      "learning_rate": 1.9995921928281893e-05,
+      "loss": 0.4199,
+      "step": 186
+    },
+    {
+      "epoch": 0.21836228287841192,
+      "grad_norm": 0.09654464394763666,
+      "learning_rate": 1.9995318587202132e-05,
+      "loss": 0.4347,
+      "step": 187
+    },
+    {
+      "epoch": 0.21952999562107722,
+      "grad_norm": 0.10710377909952767,
+      "learning_rate": 1.99946736498135e-05,
+      "loss": 0.4342,
+      "step": 188
+    },
+    {
+      "epoch": 0.22069770836374253,
+      "grad_norm": 0.09406833205093618,
+      "learning_rate": 1.999398711879996e-05,
+      "loss": 0.4375,
+      "step": 189
+    },
+    {
+      "epoch": 0.22186542110640783,
+      "grad_norm": 0.10505901683598359,
+      "learning_rate": 1.9993258997018564e-05,
+      "loss": 0.4324,
+      "step": 190
+    },
+    {
+      "epoch": 0.22303313384907314,
+      "grad_norm": 0.0919673312707725,
+      "learning_rate": 1.9992489287499454e-05,
+      "loss": 0.4033,
+      "step": 191
+    },
+    {
+      "epoch": 0.22420084659173845,
+      "grad_norm": 0.09881665041753462,
+      "learning_rate": 1.9991677993445832e-05,
+      "loss": 0.4149,
+      "step": 192
+    },
+    {
+      "epoch": 0.22536855933440372,
+      "grad_norm": 0.11198312162760661,
+      "learning_rate": 1.9990825118233958e-05,
+      "loss": 0.4137,
+      "step": 193
+    },
+    {
+      "epoch": 0.22653627207706903,
+      "grad_norm": 0.09890108645846746,
+      "learning_rate": 1.9989930665413148e-05,
+      "loss": 0.4212,
+      "step": 194
+    },
+    {
+      "epoch": 0.22770398481973433,
+      "grad_norm": 0.11495155483119726,
+      "learning_rate": 1.9988994638705734e-05,
+      "loss": 0.4191,
+      "step": 195
+    },
+    {
+      "epoch": 0.22887169756239964,
+      "grad_norm": 0.09682670197457316,
+      "learning_rate": 1.9988017042007066e-05,
+      "loss": 0.4055,
+      "step": 196
+    },
+    {
+      "epoch": 0.23003941030506495,
+      "grad_norm": 0.11074186776595134,
+      "learning_rate": 1.9986997879385488e-05,
+      "loss": 0.4027,
+      "step": 197
+    },
+    {
+      "epoch": 0.23120712304773025,
+      "grad_norm": 0.09854430220798126,
+      "learning_rate": 1.998593715508233e-05,
+      "loss": 0.4331,
+      "step": 198
+    },
+    {
+      "epoch": 0.23237483579039556,
+      "grad_norm": 0.09976192319601673,
+      "learning_rate": 1.9984834873511873e-05,
+      "loss": 0.4239,
+      "step": 199
+    },
+    {
+      "epoch": 0.23354254853306086,
+      "grad_norm": 0.10274387780991374,
+      "learning_rate": 1.9983691039261358e-05,
+      "loss": 0.429,
+      "step": 200
+    },
+    {
+      "epoch": 0.23471026127572617,
+      "grad_norm": 0.09350927098917027,
+      "learning_rate": 1.9982505657090936e-05,
+      "loss": 0.4422,
+      "step": 201
+    },
+    {
+      "epoch": 0.23587797401839147,
+      "grad_norm": 0.09411221193136099,
+      "learning_rate": 1.998127873193367e-05,
+      "loss": 0.4324,
+      "step": 202
+    },
+    {
+      "epoch": 0.23704568676105678,
+      "grad_norm": 0.09952623334530343,
+      "learning_rate": 1.9980010268895507e-05,
+      "loss": 0.4378,
+      "step": 203
+    },
+    {
+      "epoch": 0.23821339950372208,
+      "grad_norm": 0.10417306190277315,
+      "learning_rate": 1.9978700273255254e-05,
+      "loss": 0.4619,
+      "step": 204
+    },
+    {
+      "epoch": 0.2393811122463874,
+      "grad_norm": 0.08718371553057434,
+      "learning_rate": 1.997734875046456e-05,
+      "loss": 0.4185,
+      "step": 205
+    },
+    {
+      "epoch": 0.2405488249890527,
+      "grad_norm": 0.10246081057630183,
+      "learning_rate": 1.9975955706147902e-05,
+      "loss": 0.4269,
+      "step": 206
+    },
+    {
+      "epoch": 0.241716537731718,
+      "grad_norm": 0.09428747171094515,
+      "learning_rate": 1.9974521146102535e-05,
+      "loss": 0.4203,
+      "step": 207
+    },
+    {
+      "epoch": 0.2428842504743833,
+      "grad_norm": 0.09171382096220415,
+      "learning_rate": 1.99730450762985e-05,
+      "loss": 0.4117,
+      "step": 208
+    },
+    {
+      "epoch": 0.2440519632170486,
+      "grad_norm": 0.10566633091901284,
+      "learning_rate": 1.9971527502878576e-05,
+      "loss": 0.4379,
+      "step": 209
+    },
+    {
+      "epoch": 0.24521967595971392,
+      "grad_norm": 0.09678518666816144,
+      "learning_rate": 1.9969968432158266e-05,
+      "loss": 0.4243,
+      "step": 210
+    },
+    {
+      "epoch": 0.24638738870237922,
+      "grad_norm": 0.09450836315840508,
+      "learning_rate": 1.9968367870625762e-05,
+      "loss": 0.4271,
+      "step": 211
+    },
+    {
+      "epoch": 0.24755510144504453,
+      "grad_norm": 0.09276220442719804,
+      "learning_rate": 1.9966725824941933e-05,
+      "loss": 0.4333,
+      "step": 212
+    },
+    {
+      "epoch": 0.24872281418770983,
+      "grad_norm": 0.114413208833807,
+      "learning_rate": 1.996504230194028e-05,
+      "loss": 0.4119,
+      "step": 213
+    },
+    {
+      "epoch": 0.24989052693037514,
+      "grad_norm": 0.1012175889654461,
+      "learning_rate": 1.9963317308626916e-05,
+      "loss": 0.4225,
+      "step": 214
+    },
+    {
+      "epoch": 0.2510582396730404,
+      "grad_norm": 0.11054928564581837,
+      "learning_rate": 1.9961550852180535e-05,
+      "loss": 0.4126,
+      "step": 215
+    },
+    {
+      "epoch": 0.2522259524157057,
+      "grad_norm": 0.11525756150518936,
+      "learning_rate": 1.9959742939952393e-05,
+      "loss": 0.4136,
+      "step": 216
+    },
+    {
+      "epoch": 0.25339366515837103,
+      "grad_norm": 0.09376897239847337,
+      "learning_rate": 1.9957893579466252e-05,
+      "loss": 0.4126,
+      "step": 217
+    },
+    {
+      "epoch": 0.25456137790103633,
+      "grad_norm": 0.11790506312873196,
+      "learning_rate": 1.995600277841837e-05,
+      "loss": 0.4351,
+      "step": 218
+    },
+    {
+      "epoch": 0.25572909064370164,
+      "grad_norm": 0.10341675963955912,
+      "learning_rate": 1.9954070544677476e-05,
+      "loss": 0.4452,
+      "step": 219
+    },
+    {
+      "epoch": 0.25689680338636695,
+      "grad_norm": 0.10411133743842631,
+      "learning_rate": 1.995209688628471e-05,
+      "loss": 0.4065,
+      "step": 220
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 0.10033257072820242,
+      "learning_rate": 1.9950081811453598e-05,
+      "loss": 0.414,
+      "step": 221
+    },
+    {
+      "epoch": 0.25923222887169756,
+      "grad_norm": 0.08691452182962699,
+      "learning_rate": 1.994802532857004e-05,
+      "loss": 0.3998,
+      "step": 222
+    },
+    {
+      "epoch": 0.26039994161436286,
+      "grad_norm": 0.09636318718636133,
+      "learning_rate": 1.9945927446192256e-05,
+      "loss": 0.4373,
+      "step": 223
+    },
+    {
+      "epoch": 0.26156765435702817,
+      "grad_norm": 0.09447063144935004,
+      "learning_rate": 1.9943788173050743e-05,
+      "loss": 0.4415,
+      "step": 224
+    },
+    {
+      "epoch": 0.2627353670996935,
+      "grad_norm": 0.09376541222576347,
+      "learning_rate": 1.994160751804826e-05,
+      "loss": 0.4215,
+      "step": 225
+    },
+    {
+      "epoch": 0.2639030798423588,
+      "grad_norm": 0.1033307062425339,
+      "learning_rate": 1.9939385490259773e-05,
+      "loss": 0.4369,
+      "step": 226
+    },
+    {
+      "epoch": 0.2650707925850241,
+      "grad_norm": 0.09211204059440765,
+      "learning_rate": 1.9937122098932428e-05,
+      "loss": 0.4209,
+      "step": 227
+    },
+    {
+      "epoch": 0.2662385053276894,
+      "grad_norm": 0.10440152985557515,
+      "learning_rate": 1.99348173534855e-05,
+      "loss": 0.4037,
+      "step": 228
+    },
+    {
+      "epoch": 0.2674062180703547,
+      "grad_norm": 0.09493809638082916,
+      "learning_rate": 1.9932471263510385e-05,
+      "loss": 0.4145,
+      "step": 229
+    },
+    {
+      "epoch": 0.26857393081302,
+      "grad_norm": 0.08892030818452207,
+      "learning_rate": 1.9930083838770504e-05,
+      "loss": 0.4256,
+      "step": 230
+    },
+    {
+      "epoch": 0.2697416435556853,
+      "grad_norm": 0.09924011463778395,
+      "learning_rate": 1.9927655089201327e-05,
+      "loss": 0.4333,
+      "step": 231
+    },
+    {
+      "epoch": 0.2709093562983506,
+      "grad_norm": 0.09043328995836368,
+      "learning_rate": 1.992518502491028e-05,
+      "loss": 0.4263,
+      "step": 232
+    },
+    {
+      "epoch": 0.2720770690410159,
+      "grad_norm": 0.09563693176313726,
+      "learning_rate": 1.9922673656176733e-05,
+      "loss": 0.4445,
+      "step": 233
+    },
+    {
+      "epoch": 0.2732447817836812,
+      "grad_norm": 0.09076954228129776,
+      "learning_rate": 1.9920120993451947e-05,
+      "loss": 0.4116,
+      "step": 234
+    },
+    {
+      "epoch": 0.27441249452634653,
+      "grad_norm": 0.08623563366814366,
+      "learning_rate": 1.991752704735903e-05,
+      "loss": 0.4145,
+      "step": 235
+    },
+    {
+      "epoch": 0.27558020726901183,
+      "grad_norm": 0.09084827398831406,
+      "learning_rate": 1.991489182869289e-05,
+      "loss": 0.4231,
+      "step": 236
+    },
+    {
+      "epoch": 0.27674792001167714,
+      "grad_norm": 0.08559186909800579,
+      "learning_rate": 1.99122153484202e-05,
+      "loss": 0.4096,
+      "step": 237
+    },
+    {
+      "epoch": 0.27791563275434245,
+      "grad_norm": 0.09500638066827784,
+      "learning_rate": 1.990949761767935e-05,
+      "loss": 0.4291,
+      "step": 238
+    },
+    {
+      "epoch": 0.27908334549700775,
+      "grad_norm": 0.08737780078828943,
+      "learning_rate": 1.9906738647780384e-05,
+      "loss": 0.4249,
+      "step": 239
+    },
+    {
+      "epoch": 0.28025105823967306,
+      "grad_norm": 0.10038682117069785,
+      "learning_rate": 1.9903938450204972e-05,
+      "loss": 0.4306,
+      "step": 240
+    },
+    {
+      "epoch": 0.28141877098233836,
+      "grad_norm": 0.09104988582005992,
+      "learning_rate": 1.990109703660637e-05,
+      "loss": 0.4174,
+      "step": 241
+    },
+    {
+      "epoch": 0.28258648372500367,
+      "grad_norm": 0.09769230132589866,
+      "learning_rate": 1.989821441880933e-05,
+      "loss": 0.4095,
+      "step": 242
+    },
+    {
+      "epoch": 0.283754196467669,
+      "grad_norm": 0.09555401949833497,
+      "learning_rate": 1.9895290608810103e-05,
+      "loss": 0.4266,
+      "step": 243
+    },
+    {
+      "epoch": 0.2849219092103343,
+      "grad_norm": 0.09195031825030202,
+      "learning_rate": 1.9892325618776353e-05,
+      "loss": 0.3942,
+      "step": 244
+    },
+    {
+      "epoch": 0.2860896219529996,
+      "grad_norm": 0.09261769595570847,
+      "learning_rate": 1.988931946104712e-05,
+      "loss": 0.4142,
+      "step": 245
+    },
+    {
+      "epoch": 0.2872573346956649,
+      "grad_norm": 0.09149294688936342,
+      "learning_rate": 1.988627214813277e-05,
+      "loss": 0.4171,
+      "step": 246
+    },
+    {
+      "epoch": 0.2884250474383302,
+      "grad_norm": 0.10122879635586773,
+      "learning_rate": 1.9883183692714935e-05,
+      "loss": 0.4327,
+      "step": 247
+    },
+    {
+      "epoch": 0.2895927601809955,
+      "grad_norm": 0.09032805386773264,
+      "learning_rate": 1.9880054107646467e-05,
+      "loss": 0.4202,
+      "step": 248
+    },
+    {
+      "epoch": 0.2907604729236608,
+      "grad_norm": 0.09070672548952627,
+      "learning_rate": 1.9876883405951378e-05,
+      "loss": 0.4417,
+      "step": 249
+    },
+    {
+      "epoch": 0.29192818566632606,
+      "grad_norm": 0.09341819997292193,
+      "learning_rate": 1.98736716008248e-05,
+      "loss": 0.4097,
+      "step": 250
+    },
+    {
+      "epoch": 0.29309589840899136,
+      "grad_norm": 0.09798950425485974,
+      "learning_rate": 1.987041870563291e-05,
+      "loss": 0.4211,
+      "step": 251
+    },
+    {
+      "epoch": 0.29426361115165667,
+      "grad_norm": 0.09590881882948156,
+      "learning_rate": 1.986712473391289e-05,
+      "loss": 0.4276,
+      "step": 252
+    },
+    {
+      "epoch": 0.295431323894322,
+      "grad_norm": 0.0923259495897974,
+      "learning_rate": 1.9863789699372863e-05,
+      "loss": 0.421,
+      "step": 253
+    },
+    {
+      "epoch": 0.2965990366369873,
+      "grad_norm": 0.08939598107219032,
+      "learning_rate": 1.986041361589184e-05,
+      "loss": 0.4069,
+      "step": 254
+    },
+    {
+      "epoch": 0.2977667493796526,
+      "grad_norm": 0.09816587094023935,
+      "learning_rate": 1.985699649751966e-05,
+      "loss": 0.4331,
+      "step": 255
+    },
+    {
+      "epoch": 0.2989344621223179,
+      "grad_norm": 0.08902788561809195,
+      "learning_rate": 1.9853538358476933e-05,
+      "loss": 0.3953,
+      "step": 256
+    },
+    {
+      "epoch": 0.3001021748649832,
+      "grad_norm": 0.09304853488843542,
+      "learning_rate": 1.9850039213154972e-05,
+      "loss": 0.4096,
+      "step": 257
+    },
+    {
+      "epoch": 0.3012698876076485,
+      "grad_norm": 0.10026319299288657,
+      "learning_rate": 1.984649907611575e-05,
+      "loss": 0.4217,
+      "step": 258
+    },
+    {
+      "epoch": 0.3024376003503138,
+      "grad_norm": 0.0939562542849814,
+      "learning_rate": 1.984291796209183e-05,
+      "loss": 0.4317,
+      "step": 259
+    },
+    {
+      "epoch": 0.3036053130929791,
+      "grad_norm": 0.1011282469869353,
+      "learning_rate": 1.98392958859863e-05,
+      "loss": 0.4261,
+      "step": 260
+    },
+    {
+      "epoch": 0.3047730258356444,
+      "grad_norm": 0.0888725054734187,
+      "learning_rate": 1.9835632862872712e-05,
+      "loss": 0.4288,
+      "step": 261
+    },
+    {
+      "epoch": 0.3059407385783097,
+      "grad_norm": 0.10146863841937535,
+      "learning_rate": 1.9831928907995032e-05,
+      "loss": 0.4256,
+      "step": 262
+    },
+    {
+      "epoch": 0.30710845132097503,
+      "grad_norm": 0.09319936998531873,
+      "learning_rate": 1.9828184036767556e-05,
+      "loss": 0.4244,
+      "step": 263
+    },
+    {
+      "epoch": 0.30827616406364033,
+      "grad_norm": 0.08510113584023328,
+      "learning_rate": 1.982439826477487e-05,
+      "loss": 0.3903,
+      "step": 264
+    },
+    {
+      "epoch": 0.30944387680630564,
+      "grad_norm": 0.0892576097387713,
+      "learning_rate": 1.9820571607771756e-05,
+      "loss": 0.399,
+      "step": 265
+    },
+    {
+      "epoch": 0.31061158954897095,
+      "grad_norm": 0.09367313894127703,
+      "learning_rate": 1.9816704081683152e-05,
+      "loss": 0.4028,
+      "step": 266
+    },
+    {
+      "epoch": 0.31177930229163625,
+      "grad_norm": 0.09865049374253693,
+      "learning_rate": 1.9812795702604073e-05,
+      "loss": 0.4114,
+      "step": 267
+    },
+    {
+      "epoch": 0.31294701503430156,
+      "grad_norm": 0.09641341847591313,
+      "learning_rate": 1.980884648679955e-05,
+      "loss": 0.3993,
+      "step": 268
+    },
+    {
+      "epoch": 0.31411472777696686,
+      "grad_norm": 0.09809084245207524,
+      "learning_rate": 1.980485645070455e-05,
+      "loss": 0.4036,
+      "step": 269
+    },
+    {
+      "epoch": 0.31528244051963217,
+      "grad_norm": 0.10073006912048482,
+      "learning_rate": 1.9800825610923937e-05,
+      "loss": 0.4202,
+      "step": 270
+    },
+    {
+      "epoch": 0.3164501532622975,
+      "grad_norm": 0.09502297304581203,
+      "learning_rate": 1.9796753984232357e-05,
+      "loss": 0.4166,
+      "step": 271
+    },
+    {
+      "epoch": 0.3176178660049628,
+      "grad_norm": 0.09363057626532845,
+      "learning_rate": 1.9792641587574212e-05,
+      "loss": 0.4234,
+      "step": 272
+    },
+    {
+      "epoch": 0.3187855787476281,
+      "grad_norm": 0.08658272678358302,
+      "learning_rate": 1.9788488438063566e-05,
+      "loss": 0.4105,
+      "step": 273
+    },
+    {
+      "epoch": 0.3199532914902934,
+      "grad_norm": 0.106071655162013,
+      "learning_rate": 1.9784294552984077e-05,
+      "loss": 0.4214,
+      "step": 274
+    },
+    {
+      "epoch": 0.3211210042329587,
+      "grad_norm": 0.0842952190722254,
+      "learning_rate": 1.9780059949788935e-05,
+      "loss": 0.4142,
+      "step": 275
+    },
+    {
+      "epoch": 0.322288716975624,
+      "grad_norm": 0.10749777188028763,
+      "learning_rate": 1.9775784646100768e-05,
+      "loss": 0.4017,
+      "step": 276
+    },
+    {
+      "epoch": 0.3234564297182893,
+      "grad_norm": 0.09455129762973902,
+      "learning_rate": 1.9771468659711595e-05,
+      "loss": 0.43,
+      "step": 277
+    },
+    {
+      "epoch": 0.3246241424609546,
+      "grad_norm": 0.09664664766763484,
+      "learning_rate": 1.9767112008582736e-05,
+      "loss": 0.4331,
+      "step": 278
+    },
+    {
+      "epoch": 0.3257918552036199,
+      "grad_norm": 0.09574474512222941,
+      "learning_rate": 1.9762714710844738e-05,
+      "loss": 0.3968,
+      "step": 279
+    },
+    {
+      "epoch": 0.3269595679462852,
+      "grad_norm": 0.08658733597688241,
+      "learning_rate": 1.97582767847973e-05,
+      "loss": 0.4125,
+      "step": 280
+    },
+    {
+      "epoch": 0.32812728068895053,
+      "grad_norm": 0.11893712110876606,
+      "learning_rate": 1.975379824890921e-05,
+      "loss": 0.4233,
+      "step": 281
+    },
+    {
+      "epoch": 0.32929499343161583,
+      "grad_norm": 0.08860108749523976,
+      "learning_rate": 1.9749279121818235e-05,
+      "loss": 0.3835,
+      "step": 282
+    },
+    {
+      "epoch": 0.33046270617428114,
+      "grad_norm": 0.09578694907304526,
+      "learning_rate": 1.9744719422331098e-05,
+      "loss": 0.4034,
+      "step": 283
+    },
+    {
+      "epoch": 0.33163041891694645,
+      "grad_norm": 0.100960350759047,
+      "learning_rate": 1.9740119169423337e-05,
+      "loss": 0.4214,
+      "step": 284
+    },
+    {
+      "epoch": 0.33279813165961175,
+      "grad_norm": 0.08761685739656637,
+      "learning_rate": 1.973547838223927e-05,
+      "loss": 0.4129,
+      "step": 285
+    },
+    {
+      "epoch": 0.33396584440227706,
+      "grad_norm": 0.0952078617312561,
+      "learning_rate": 1.9730797080091904e-05,
+      "loss": 0.423,
+      "step": 286
+    },
+    {
+      "epoch": 0.33513355714494236,
+      "grad_norm": 0.09308852043962536,
+      "learning_rate": 1.9726075282462847e-05,
+      "loss": 0.4162,
+      "step": 287
+    },
+    {
+      "epoch": 0.33630126988760767,
+      "grad_norm": 0.09213701770336757,
+      "learning_rate": 1.9721313009002228e-05,
+      "loss": 0.4248,
+      "step": 288
+    },
+    {
+      "epoch": 0.337468982630273,
+      "grad_norm": 0.09744039405008077,
+      "learning_rate": 1.9716510279528632e-05,
+      "loss": 0.4224,
+      "step": 289
+    },
+    {
+      "epoch": 0.3386366953729383,
+      "grad_norm": 0.08936185904370962,
+      "learning_rate": 1.971166711402899e-05,
+      "loss": 0.416,
+      "step": 290
+    },
+    {
+      "epoch": 0.3398044081156036,
+      "grad_norm": 0.09679521125505719,
+      "learning_rate": 1.9706783532658528e-05,
+      "loss": 0.4099,
+      "step": 291
+    },
+    {
+      "epoch": 0.3409721208582689,
+      "grad_norm": 0.08985344006144058,
+      "learning_rate": 1.9701859555740647e-05,
+      "loss": 0.4092,
+      "step": 292
+    },
+    {
+      "epoch": 0.3421398336009342,
+      "grad_norm": 0.09307235753443979,
+      "learning_rate": 1.969689520376687e-05,
+      "loss": 0.429,
+      "step": 293
+    },
+    {
+      "epoch": 0.3433075463435995,
+      "grad_norm": 0.09669787557004315,
+      "learning_rate": 1.969189049739674e-05,
+      "loss": 0.4086,
+      "step": 294
+    },
+    {
+      "epoch": 0.34447525908626475,
+      "grad_norm": 0.0843582521550282,
+      "learning_rate": 1.968684545745774e-05,
+      "loss": 0.4145,
+      "step": 295
+    },
+    {
+      "epoch": 0.34564297182893006,
+      "grad_norm": 0.08972996927693197,
+      "learning_rate": 1.9681760104945203e-05,
+      "loss": 0.4049,
+      "step": 296
+    },
+    {
+      "epoch": 0.34681068457159536,
+      "grad_norm": 0.08525586677837806,
+      "learning_rate": 1.9676634461022222e-05,
+      "loss": 0.368,
+      "step": 297
+    },
+    {
+      "epoch": 0.34797839731426067,
+      "grad_norm": 0.09333245628992055,
+      "learning_rate": 1.9671468547019575e-05,
+      "loss": 0.4066,
+      "step": 298
+    },
+    {
+      "epoch": 0.349146110056926,
+      "grad_norm": 0.09150099582342118,
+      "learning_rate": 1.966626238443561e-05,
+      "loss": 0.4158,
+      "step": 299
+    },
+    {
+      "epoch": 0.3503138227995913,
+      "grad_norm": 0.09038445342524738,
+      "learning_rate": 1.9661015994936204e-05,
+      "loss": 0.409,
+      "step": 300
+    },
+    {
+      "epoch": 0.3514815355422566,
+      "grad_norm": 0.0925940917584899,
+      "learning_rate": 1.9655729400354607e-05,
+      "loss": 0.4356,
+      "step": 301
+    },
+    {
+      "epoch": 0.3526492482849219,
+      "grad_norm": 0.08709107227861636,
+      "learning_rate": 1.96504026226914e-05,
+      "loss": 0.4028,
+      "step": 302
+    },
+    {
+      "epoch": 0.3538169610275872,
+      "grad_norm": 0.08922179299953607,
+      "learning_rate": 1.9645035684114398e-05,
+      "loss": 0.4141,
+      "step": 303
+    },
+    {
+      "epoch": 0.3549846737702525,
+      "grad_norm": 0.09025089733275102,
+      "learning_rate": 1.9639628606958535e-05,
+      "loss": 0.4137,
+      "step": 304
+    },
+    {
+      "epoch": 0.3561523865129178,
+      "grad_norm": 0.09096086392747053,
+      "learning_rate": 1.963418141372579e-05,
+      "loss": 0.4007,
+      "step": 305
+    },
+    {
+      "epoch": 0.3573200992555831,
+      "grad_norm": 0.09408592329902848,
+      "learning_rate": 1.9628694127085094e-05,
+      "loss": 0.3947,
+      "step": 306
+    },
+    {
+      "epoch": 0.3584878119982484,
+      "grad_norm": 0.09237307142729788,
+      "learning_rate": 1.9623166769872216e-05,
+      "loss": 0.4202,
+      "step": 307
+    },
+    {
+      "epoch": 0.3596555247409137,
+      "grad_norm": 0.0964039513110765,
+      "learning_rate": 1.9617599365089693e-05,
+      "loss": 0.4157,
+      "step": 308
+    },
+    {
+      "epoch": 0.36082323748357903,
+      "grad_norm": 0.09945369547460135,
+      "learning_rate": 1.9611991935906727e-05,
+      "loss": 0.4132,
+      "step": 309
+    },
+    {
+      "epoch": 0.36199095022624433,
+      "grad_norm": 0.08836607306429471,
+      "learning_rate": 1.9606344505659072e-05,
+      "loss": 0.3996,
+      "step": 310
+    },
+    {
+      "epoch": 0.36315866296890964,
+      "grad_norm": 0.09951889123609499,
+      "learning_rate": 1.9600657097848957e-05,
+      "loss": 0.41,
+      "step": 311
+    },
+    {
+      "epoch": 0.36432637571157495,
+      "grad_norm": 0.08971541348741108,
+      "learning_rate": 1.9594929736144978e-05,
+      "loss": 0.3837,
+      "step": 312
+    },
+    {
+      "epoch": 0.36549408845424025,
+      "grad_norm": 0.10125118232813089,
+      "learning_rate": 1.9589162444382e-05,
+      "loss": 0.4069,
+      "step": 313
+    },
+    {
+      "epoch": 0.36666180119690556,
+      "grad_norm": 0.10486899608643994,
+      "learning_rate": 1.9583355246561076e-05,
+      "loss": 0.4165,
+      "step": 314
+    },
+    {
+      "epoch": 0.36782951393957086,
+      "grad_norm": 0.0879412062374762,
+      "learning_rate": 1.9577508166849308e-05,
+      "loss": 0.3959,
+      "step": 315
+    },
+    {
+      "epoch": 0.36899722668223617,
+      "grad_norm": 0.1096678998177288,
+      "learning_rate": 1.957162122957978e-05,
+      "loss": 0.4183,
+      "step": 316
+    },
+    {
+      "epoch": 0.3701649394249015,
+      "grad_norm": 0.09142567142372546,
+      "learning_rate": 1.9565694459251457e-05,
+      "loss": 0.4028,
+      "step": 317
+    },
+    {
+      "epoch": 0.3713326521675668,
+      "grad_norm": 0.08734191605748313,
+      "learning_rate": 1.9559727880529057e-05,
+      "loss": 0.4026,
+      "step": 318
+    },
+    {
+      "epoch": 0.3725003649102321,
+      "grad_norm": 0.096293660696712,
+      "learning_rate": 1.955372151824297e-05,
+      "loss": 0.3991,
+      "step": 319
+    },
+    {
+      "epoch": 0.3736680776528974,
+      "grad_norm": 0.08650590243510535,
+      "learning_rate": 1.9547675397389144e-05,
+      "loss": 0.3996,
+      "step": 320
+    },
+    {
+      "epoch": 0.3748357903955627,
+      "grad_norm": 0.0899458181899923,
+      "learning_rate": 1.954158954312899e-05,
+      "loss": 0.3935,
+      "step": 321
+    },
+    {
+      "epoch": 0.376003503138228,
+      "grad_norm": 0.09001200049017398,
+      "learning_rate": 1.9535463980789277e-05,
+      "loss": 0.3901,
+      "step": 322
+    },
+    {
+      "epoch": 0.3771712158808933,
+      "grad_norm": 0.08398045970119282,
+      "learning_rate": 1.9529298735862015e-05,
+      "loss": 0.404,
+      "step": 323
+    },
+    {
+      "epoch": 0.3783389286235586,
+      "grad_norm": 0.08748668603384174,
+      "learning_rate": 1.9523093834004358e-05,
+      "loss": 0.3917,
+      "step": 324
+    },
+    {
+      "epoch": 0.3795066413662239,
+      "grad_norm": 0.0960429057953384,
+      "learning_rate": 1.951684930103849e-05,
+      "loss": 0.4039,
+      "step": 325
+    },
+    {
+      "epoch": 0.3806743541088892,
+      "grad_norm": 0.09290841311548105,
+      "learning_rate": 1.9510565162951538e-05,
+      "loss": 0.417,
+      "step": 326
+    },
+    {
+      "epoch": 0.38184206685155453,
+      "grad_norm": 0.0920388762671519,
+      "learning_rate": 1.9504241445895434e-05,
+      "loss": 0.4097,
+      "step": 327
+    },
+    {
+      "epoch": 0.38300977959421983,
+      "grad_norm": 0.09274242023371436,
+      "learning_rate": 1.949787817618683e-05,
+      "loss": 0.4006,
+      "step": 328
+    },
+    {
+      "epoch": 0.38417749233688514,
+      "grad_norm": 0.08723067737527367,
+      "learning_rate": 1.9491475380306975e-05,
+      "loss": 0.4107,
+      "step": 329
+    },
+    {
+      "epoch": 0.38534520507955045,
+      "grad_norm": 0.09932299039021772,
+      "learning_rate": 1.9485033084901606e-05,
+      "loss": 0.4055,
+      "step": 330
+    },
+    {
+      "epoch": 0.38651291782221575,
+      "grad_norm": 0.08858960755792726,
+      "learning_rate": 1.9478551316780855e-05,
+      "loss": 0.4157,
+      "step": 331
+    },
+    {
+      "epoch": 0.38768063056488106,
+      "grad_norm": 0.09450380666714202,
+      "learning_rate": 1.9472030102919102e-05,
+      "loss": 0.4142,
+      "step": 332
+    },
+    {
+      "epoch": 0.38884834330754636,
+      "grad_norm": 0.08506075776873054,
+      "learning_rate": 1.94654694704549e-05,
+      "loss": 0.4132,
+      "step": 333
+    },
+    {
+      "epoch": 0.39001605605021167,
+      "grad_norm": 0.08322662732428716,
+      "learning_rate": 1.945886944669084e-05,
+      "loss": 0.407,
+      "step": 334
+    },
+    {
+      "epoch": 0.391183768792877,
+      "grad_norm": 0.08369097452329556,
+      "learning_rate": 1.9452230059093442e-05,
+      "loss": 0.3981,
+      "step": 335
+    },
+    {
+      "epoch": 0.3923514815355423,
+      "grad_norm": 0.08975252075996117,
+      "learning_rate": 1.944555133529304e-05,
+      "loss": 0.3784,
+      "step": 336
+    },
+    {
+      "epoch": 0.3935191942782076,
+      "grad_norm": 0.08349603952049707,
+      "learning_rate": 1.9438833303083677e-05,
+      "loss": 0.3988,
+      "step": 337
+    },
+    {
+      "epoch": 0.3946869070208729,
+      "grad_norm": 0.09024144412975443,
+      "learning_rate": 1.943207599042297e-05,
+      "loss": 0.4256,
+      "step": 338
+    },
+    {
+      "epoch": 0.3958546197635382,
+      "grad_norm": 0.08444028120596131,
+      "learning_rate": 1.9425279425432016e-05,
+      "loss": 0.3897,
+      "step": 339
+    },
+    {
+      "epoch": 0.3970223325062035,
+      "grad_norm": 0.085469103509899,
+      "learning_rate": 1.941844363639525e-05,
+      "loss": 0.4193,
+      "step": 340
+    },
+    {
+      "epoch": 0.39819004524886875,
+      "grad_norm": 0.08939213033260533,
+      "learning_rate": 1.941156865176035e-05,
+      "loss": 0.4238,
+      "step": 341
+    },
+    {
+      "epoch": 0.39935775799153406,
+      "grad_norm": 0.08818020046627022,
+      "learning_rate": 1.9404654500138117e-05,
+      "loss": 0.4117,
+      "step": 342
+    },
+    {
+      "epoch": 0.40052547073419936,
+      "grad_norm": 0.09474320863334597,
+      "learning_rate": 1.939770121030233e-05,
+      "loss": 0.4264,
+      "step": 343
+    },
+    {
+      "epoch": 0.40169318347686467,
+      "grad_norm": 0.09326290671743713,
+      "learning_rate": 1.939070881118966e-05,
+      "loss": 0.4197,
+      "step": 344
+    },
+    {
+      "epoch": 0.40286089621953,
+      "grad_norm": 0.08948613531809213,
+      "learning_rate": 1.938367733189953e-05,
+      "loss": 0.4272,
+      "step": 345
+    },
+    {
+      "epoch": 0.4040286089621953,
+      "grad_norm": 0.08814575464145187,
+      "learning_rate": 1.937660680169399e-05,
+      "loss": 0.3952,
+      "step": 346
+    },
+    {
+      "epoch": 0.4051963217048606,
+      "grad_norm": 0.08385699196454079,
+      "learning_rate": 1.936949724999762e-05,
+      "loss": 0.4006,
+      "step": 347
+    },
+    {
+      "epoch": 0.4063640344475259,
+      "grad_norm": 0.09056105618167236,
+      "learning_rate": 1.9362348706397374e-05,
+      "loss": 0.4158,
+      "step": 348
+    },
+    {
+      "epoch": 0.4075317471901912,
+      "grad_norm": 0.08426123047233607,
+      "learning_rate": 1.9355161200642485e-05,
+      "loss": 0.3885,
+      "step": 349
+    },
+    {
+      "epoch": 0.4086994599328565,
+      "grad_norm": 0.08514301383903707,
+      "learning_rate": 1.9347934762644326e-05,
+      "loss": 0.4217,
+      "step": 350
+    },
+    {
+      "epoch": 0.4098671726755218,
+      "grad_norm": 0.08406047519796939,
+      "learning_rate": 1.934066942247629e-05,
+      "loss": 0.3913,
+      "step": 351
+    },
+    {
+      "epoch": 0.4110348854181871,
+      "grad_norm": 0.07889149013338989,
+      "learning_rate": 1.9333365210373668e-05,
+      "loss": 0.3906,
+      "step": 352
+    },
+    {
+      "epoch": 0.4122025981608524,
+      "grad_norm": 0.09530569927648494,
+      "learning_rate": 1.9326022156733514e-05,
+      "loss": 0.411,
+      "step": 353
+    },
+    {
+      "epoch": 0.4133703109035177,
+      "grad_norm": 0.08215979646415135,
+      "learning_rate": 1.9318640292114526e-05,
+      "loss": 0.4004,
+      "step": 354
+    },
+    {
+      "epoch": 0.41453802364618303,
+      "grad_norm": 0.0846000380587616,
+      "learning_rate": 1.9311219647236914e-05,
+      "loss": 0.3918,
+      "step": 355
+    },
+    {
+      "epoch": 0.41570573638884833,
+      "grad_norm": 0.09063476020508658,
+      "learning_rate": 1.9303760252982287e-05,
+      "loss": 0.3987,
+      "step": 356
+    },
+    {
+      "epoch": 0.41687344913151364,
+      "grad_norm": 0.08535328867487002,
+      "learning_rate": 1.9296262140393498e-05,
+      "loss": 0.3961,
+      "step": 357
+    },
+    {
+      "epoch": 0.41804116187417895,
+      "grad_norm": 0.08729627581794826,
+      "learning_rate": 1.9288725340674536e-05,
+      "loss": 0.4158,
+      "step": 358
+    },
+    {
+      "epoch": 0.41920887461684425,
+      "grad_norm": 0.09431695392066299,
+      "learning_rate": 1.928114988519039e-05,
+      "loss": 0.4165,
+      "step": 359
+    },
+    {
+      "epoch": 0.42037658735950956,
+      "grad_norm": 0.08528870923355873,
+      "learning_rate": 1.927353580546692e-05,
+      "loss": 0.3712,
+      "step": 360
+    },
+    {
+      "epoch": 0.42154430010217486,
+      "grad_norm": 0.09294511141157431,
+      "learning_rate": 1.9265883133190715e-05,
+      "loss": 0.4032,
+      "step": 361
+    },
+    {
+      "epoch": 0.42271201284484017,
+      "grad_norm": 0.0943567933076398,
+      "learning_rate": 1.925819190020898e-05,
+      "loss": 0.4162,
+      "step": 362
+    },
+    {
+      "epoch": 0.4238797255875055,
+      "grad_norm": 0.0886423542602759,
+      "learning_rate": 1.925046213852939e-05,
+      "loss": 0.4049,
+      "step": 363
+    },
+    {
+      "epoch": 0.4250474383301708,
+      "grad_norm": 0.1032117946278424,
+      "learning_rate": 1.9242693880319962e-05,
+      "loss": 0.4145,
+      "step": 364
+    },
+    {
+      "epoch": 0.4262151510728361,
+      "grad_norm": 0.08824895152971705,
+      "learning_rate": 1.923488715790891e-05,
+      "loss": 0.4266,
+      "step": 365
+    },
+    {
+      "epoch": 0.4273828638155014,
+      "grad_norm": 0.09098314873463258,
+      "learning_rate": 1.922704200378453e-05,
+      "loss": 0.3865,
+      "step": 366
+    },
+    {
+      "epoch": 0.4285505765581667,
+      "grad_norm": 0.08631972784322572,
+      "learning_rate": 1.921915845059505e-05,
+      "loss": 0.3905,
+      "step": 367
+    },
+    {
+      "epoch": 0.429718289300832,
+      "grad_norm": 0.08688876884783896,
+      "learning_rate": 1.92112365311485e-05,
+      "loss": 0.4225,
+      "step": 368
+    },
+    {
+      "epoch": 0.4308860020434973,
+      "grad_norm": 0.0894244687504853,
+      "learning_rate": 1.920327627841258e-05,
+      "loss": 0.4082,
+      "step": 369
+    },
+    {
+      "epoch": 0.4320537147861626,
+      "grad_norm": 0.09250120368904122,
+      "learning_rate": 1.919527772551451e-05,
+      "loss": 0.4142,
+      "step": 370
+    },
+    {
+      "epoch": 0.4332214275288279,
+      "grad_norm": 0.08921540471631043,
+      "learning_rate": 1.9187240905740893e-05,
+      "loss": 0.4044,
+      "step": 371
+    },
+    {
+      "epoch": 0.4343891402714932,
+      "grad_norm": 0.09663897463909352,
+      "learning_rate": 1.9179165852537596e-05,
+      "loss": 0.403,
+      "step": 372
+    },
+    {
+      "epoch": 0.43555685301415853,
+      "grad_norm": 0.0896033677985113,
+      "learning_rate": 1.917105259950959e-05,
+      "loss": 0.3989,
+      "step": 373
+    },
+    {
+      "epoch": 0.43672456575682383,
+      "grad_norm": 0.09427725110408793,
+      "learning_rate": 1.916290118042082e-05,
+      "loss": 0.3913,
+      "step": 374
+    },
+    {
+      "epoch": 0.43789227849948914,
+      "grad_norm": 0.09197162755532229,
+      "learning_rate": 1.9154711629194062e-05,
+      "loss": 0.4151,
+      "step": 375
+    },
+    {
+      "epoch": 0.43905999124215445,
+      "grad_norm": 0.09116206289687599,
+      "learning_rate": 1.914648397991078e-05,
+      "loss": 0.4028,
+      "step": 376
+    },
+    {
+      "epoch": 0.44022770398481975,
+      "grad_norm": 0.0850929363796077,
+      "learning_rate": 1.913821826681099e-05,
+      "loss": 0.4042,
+      "step": 377
+    },
+    {
+      "epoch": 0.44139541672748506,
+      "grad_norm": 0.08857221900024716,
+      "learning_rate": 1.9129914524293102e-05,
+      "loss": 0.4152,
+      "step": 378
+    },
+    {
+      "epoch": 0.44256312947015036,
+      "grad_norm": 0.09106689766584893,
+      "learning_rate": 1.9121572786913808e-05,
+      "loss": 0.4082,
+      "step": 379
+    },
+    {
+      "epoch": 0.44373084221281567,
+      "grad_norm": 0.08125001672293124,
+      "learning_rate": 1.91131930893879e-05,
+      "loss": 0.4057,
+      "step": 380
+    },
+    {
+      "epoch": 0.444898554955481,
+      "grad_norm": 0.08668386198419266,
+      "learning_rate": 1.9104775466588162e-05,
+      "loss": 0.4053,
+      "step": 381
+    },
+    {
+      "epoch": 0.4460662676981463,
+      "grad_norm": 0.0811045219575392,
+      "learning_rate": 1.9096319953545186e-05,
+      "loss": 0.3995,
+      "step": 382
+    },
+    {
+      "epoch": 0.4472339804408116,
+      "grad_norm": 0.08277990610814832,
+      "learning_rate": 1.9087826585447264e-05,
+      "loss": 0.3926,
+      "step": 383
+    },
+    {
+      "epoch": 0.4484016931834769,
+      "grad_norm": 0.08162754001944633,
+      "learning_rate": 1.9079295397640215e-05,
+      "loss": 0.3914,
+      "step": 384
+    },
+    {
+      "epoch": 0.4495694059261422,
+      "grad_norm": 0.08965249319155863,
+      "learning_rate": 1.9070726425627257e-05,
+      "loss": 0.3918,
+      "step": 385
+    },
+    {
+      "epoch": 0.45073711866880745,
+      "grad_norm": 0.08075224415564856,
+      "learning_rate": 1.9062119705068843e-05,
+      "loss": 0.4146,
+      "step": 386
+    },
+    {
+      "epoch": 0.45190483141147275,
+      "grad_norm": 0.0817317415885034,
+      "learning_rate": 1.9053475271782523e-05,
+      "loss": 0.3867,
+      "step": 387
+    },
+    {
+      "epoch": 0.45307254415413806,
+      "grad_norm": 0.08393552038758467,
+      "learning_rate": 1.9044793161742783e-05,
+      "loss": 0.414,
+      "step": 388
+    },
+    {
+      "epoch": 0.45424025689680336,
+      "grad_norm": 0.08422501822437604,
+      "learning_rate": 1.9036073411080917e-05,
+      "loss": 0.41,
+      "step": 389
+    },
+    {
+      "epoch": 0.45540796963946867,
+      "grad_norm": 0.0855711365597262,
+      "learning_rate": 1.902731605608486e-05,
+      "loss": 0.3671,
+      "step": 390
+    },
+    {
+      "epoch": 0.456575682382134,
+      "grad_norm": 0.09180747994251921,
+      "learning_rate": 1.901852113319903e-05,
+      "loss": 0.408,
+      "step": 391
+    },
+    {
+      "epoch": 0.4577433951247993,
+      "grad_norm": 0.08387540066564766,
+      "learning_rate": 1.900968867902419e-05,
+      "loss": 0.3803,
+      "step": 392
+    },
+    {
+      "epoch": 0.4589111078674646,
+      "grad_norm": 0.0928725523342326,
+      "learning_rate": 1.900081873031731e-05,
+      "loss": 0.4195,
+      "step": 393
+    },
+    {
+      "epoch": 0.4600788206101299,
+      "grad_norm": 0.08546644503632578,
+      "learning_rate": 1.899191132399138e-05,
+      "loss": 0.3733,
+      "step": 394
+    },
+    {
+      "epoch": 0.4612465333527952,
+      "grad_norm": 0.0916940989698139,
+      "learning_rate": 1.898296649711527e-05,
+      "loss": 0.4385,
+      "step": 395
+    },
+    {
+      "epoch": 0.4624142460954605,
+      "grad_norm": 0.0900761747704713,
+      "learning_rate": 1.8973984286913584e-05,
+      "loss": 0.3777,
+      "step": 396
+    },
+    {
+      "epoch": 0.4635819588381258,
+      "grad_norm": 0.07849243665961814,
+      "learning_rate": 1.896496473076651e-05,
+      "loss": 0.3921,
+      "step": 397
+    },
+    {
+      "epoch": 0.4647496715807911,
+      "grad_norm": 0.08919029447997913,
+      "learning_rate": 1.895590786620963e-05,
+      "loss": 0.4217,
+      "step": 398
+    },
+    {
+      "epoch": 0.4659173843234564,
+      "grad_norm": 0.08538716392653613,
+      "learning_rate": 1.8946813730933816e-05,
+      "loss": 0.4003,
+      "step": 399
+    },
+    {
+      "epoch": 0.4670850970661217,
+      "grad_norm": 0.08360281763931916,
+      "learning_rate": 1.8937682362785025e-05,
+      "loss": 0.4045,
+      "step": 400
+    },
+    {
+      "epoch": 0.46825280980878703,
+      "grad_norm": 0.08143545606484485,
+      "learning_rate": 1.8928513799764164e-05,
+      "loss": 0.4018,
+      "step": 401
+    },
+    {
+      "epoch": 0.46942052255145233,
+      "grad_norm": 0.08589121375634516,
+      "learning_rate": 1.891930808002694e-05,
+      "loss": 0.4045,
+      "step": 402
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.08349624618960612,
+      "learning_rate": 1.891006524188368e-05,
+      "loss": 0.3902,
+      "step": 403
+    },
+    {
+      "epoch": 0.47175594803678295,
+      "grad_norm": 0.0918939158979646,
+      "learning_rate": 1.890078532379919e-05,
+      "loss": 0.3994,
+      "step": 404
+    },
+    {
+      "epoch": 0.47292366077944825,
+      "grad_norm": 0.09076849240638607,
+      "learning_rate": 1.8891468364392577e-05,
+      "loss": 0.3862,
+      "step": 405
+    },
+    {
+      "epoch": 0.47409137352211356,
+      "grad_norm": 0.09126074987057312,
+      "learning_rate": 1.8882114402437105e-05,
+      "loss": 0.4061,
+      "step": 406
+    },
+    {
+      "epoch": 0.47525908626477886,
+      "grad_norm": 0.08990521571391341,
+      "learning_rate": 1.8872723476860033e-05,
+      "loss": 0.4028,
+      "step": 407
+    },
+    {
+      "epoch": 0.47642679900744417,
+      "grad_norm": 0.08538706078440066,
+      "learning_rate": 1.8863295626742438e-05,
+      "loss": 0.3836,
+      "step": 408
+    },
+    {
+      "epoch": 0.4775945117501095,
+      "grad_norm": 0.09402979164032187,
+      "learning_rate": 1.885383089131906e-05,
+      "loss": 0.3948,
+      "step": 409
+    },
+    {
+      "epoch": 0.4787622244927748,
+      "grad_norm": 0.08710373522414452,
+      "learning_rate": 1.8844329309978146e-05,
+      "loss": 0.3956,
+      "step": 410
+    },
+    {
+      "epoch": 0.4799299372354401,
+      "grad_norm": 0.08748652226198589,
+      "learning_rate": 1.8834790922261278e-05,
+      "loss": 0.4125,
+      "step": 411
+    },
+    {
+      "epoch": 0.4810976499781054,
+      "grad_norm": 0.09445892119689117,
+      "learning_rate": 1.8825215767863215e-05,
+      "loss": 0.4,
+      "step": 412
+    },
+    {
+      "epoch": 0.4822653627207707,
+      "grad_norm": 0.08433555212853534,
+      "learning_rate": 1.8815603886631715e-05,
+      "loss": 0.4004,
+      "step": 413
+    },
+    {
+      "epoch": 0.483433075463436,
+      "grad_norm": 0.08165322343166509,
+      "learning_rate": 1.880595531856738e-05,
+      "loss": 0.3923,
+      "step": 414
+    },
+    {
+      "epoch": 0.4846007882061013,
+      "grad_norm": 0.08816765160043341,
+      "learning_rate": 1.8796270103823494e-05,
+      "loss": 0.3958,
+      "step": 415
+    },
+    {
+      "epoch": 0.4857685009487666,
+      "grad_norm": 0.0825833229141437,
+      "learning_rate": 1.8786548282705847e-05,
+      "loss": 0.3918,
+      "step": 416
+    },
+    {
+      "epoch": 0.4869362136914319,
+      "grad_norm": 0.08781867611448836,
+      "learning_rate": 1.8776789895672557e-05,
+      "loss": 0.4305,
+      "step": 417
+    },
+    {
+      "epoch": 0.4881039264340972,
+      "grad_norm": 0.0783218816772852,
+      "learning_rate": 1.876699498333393e-05,
+      "loss": 0.3923,
+      "step": 418
+    },
+    {
+      "epoch": 0.48927163917676253,
+      "grad_norm": 0.07907671204832449,
+      "learning_rate": 1.8757163586452266e-05,
+      "loss": 0.3979,
+      "step": 419
+    },
+    {
+      "epoch": 0.49043935191942784,
+      "grad_norm": 0.08433337429265561,
+      "learning_rate": 1.8747295745941705e-05,
+      "loss": 0.4113,
+      "step": 420
+    },
+    {
+      "epoch": 0.49160706466209314,
+      "grad_norm": 0.08396657604691468,
+      "learning_rate": 1.8737391502868038e-05,
+      "loss": 0.3999,
+      "step": 421
+    },
+    {
+      "epoch": 0.49277477740475845,
+      "grad_norm": 0.08069499764000225,
+      "learning_rate": 1.8727450898448562e-05,
+      "loss": 0.3787,
+      "step": 422
+    },
+    {
+      "epoch": 0.49394249014742375,
+      "grad_norm": 0.09708554402957041,
+      "learning_rate": 1.8717473974051888e-05,
+      "loss": 0.3875,
+      "step": 423
+    },
+    {
+      "epoch": 0.49511020289008906,
+      "grad_norm": 0.08091008672413033,
+      "learning_rate": 1.8707460771197773e-05,
+      "loss": 0.4073,
+      "step": 424
+    },
+    {
+      "epoch": 0.49627791563275436,
+      "grad_norm": 0.08364418794003751,
+      "learning_rate": 1.8697411331556958e-05,
+      "loss": 0.3974,
+      "step": 425
+    },
+    {
+      "epoch": 0.49744562837541967,
+      "grad_norm": 0.09020059594052783,
+      "learning_rate": 1.8687325696950972e-05,
+      "loss": 0.4152,
+      "step": 426
+    },
+    {
+      "epoch": 0.498613341118085,
+      "grad_norm": 0.08195479412075683,
+      "learning_rate": 1.867720390935199e-05,
+      "loss": 0.3925,
+      "step": 427
+    },
+    {
+      "epoch": 0.4997810538607503,
+      "grad_norm": 0.09386944531071974,
+      "learning_rate": 1.8667046010882627e-05,
+      "loss": 0.4018,
+      "step": 428
+    },
+    {
+      "epoch": 0.5009487666034156,
+      "grad_norm": 0.08278023296648926,
+      "learning_rate": 1.865685204381578e-05,
+      "loss": 0.3838,
+      "step": 429
+    },
+    {
+      "epoch": 0.5021164793460808,
+      "grad_norm": 0.08954267224894312,
+      "learning_rate": 1.8646622050574456e-05,
+      "loss": 0.4041,
+      "step": 430
+    },
+    {
+      "epoch": 0.5032841920887462,
+      "grad_norm": 0.09240851068431329,
+      "learning_rate": 1.863635607373157e-05,
+      "loss": 0.4043,
+      "step": 431
+    },
+    {
+      "epoch": 0.5044519048314114,
+      "grad_norm": 0.08188952398616833,
+      "learning_rate": 1.8626054156009807e-05,
+      "loss": 0.3898,
+      "step": 432
+    },
+    {
+      "epoch": 0.5056196175740768,
+      "grad_norm": 0.08424139755096265,
+      "learning_rate": 1.8615716340281408e-05,
+      "loss": 0.3972,
+      "step": 433
+    },
+    {
+      "epoch": 0.5067873303167421,
+      "grad_norm": 0.09070828229874914,
+      "learning_rate": 1.8605342669568013e-05,
+      "loss": 0.3889,
+      "step": 434
+    },
+    {
+      "epoch": 0.5079550430594074,
+      "grad_norm": 0.08131329038869861,
+      "learning_rate": 1.8594933187040468e-05,
+      "loss": 0.4044,
+      "step": 435
+    },
+    {
+      "epoch": 0.5091227558020727,
+      "grad_norm": 0.09006946312146576,
+      "learning_rate": 1.8584487936018663e-05,
+      "loss": 0.4022,
+      "step": 436
+    },
+    {
+      "epoch": 0.510290468544738,
+      "grad_norm": 0.08820373009708539,
+      "learning_rate": 1.8574006959971335e-05,
+      "loss": 0.4049,
+      "step": 437
+    },
+    {
+      "epoch": 0.5114581812874033,
+      "grad_norm": 0.08211421896846621,
+      "learning_rate": 1.856349030251589e-05,
+      "loss": 0.3908,
+      "step": 438
+    },
+    {
+      "epoch": 0.5126258940300686,
+      "grad_norm": 0.0878779263170069,
+      "learning_rate": 1.8552938007418237e-05,
+      "loss": 0.3924,
+      "step": 439
+    },
+    {
+      "epoch": 0.5137936067727339,
+      "grad_norm": 0.08386236282264553,
+      "learning_rate": 1.8542350118592585e-05,
+      "loss": 0.3894,
+      "step": 440
+    },
+    {
+      "epoch": 0.5149613195153993,
+      "grad_norm": 0.08783105135229523,
+      "learning_rate": 1.8531726680101268e-05,
+      "loss": 0.3938,
+      "step": 441
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 0.08465110631770349,
+      "learning_rate": 1.8521067736154567e-05,
+      "loss": 0.3898,
+      "step": 442
+    },
+    {
+      "epoch": 0.5172967450007299,
+      "grad_norm": 0.08447566287015808,
+      "learning_rate": 1.851037333111052e-05,
+      "loss": 0.4066,
+      "step": 443
+    },
+    {
+      "epoch": 0.5184644577433951,
+      "grad_norm": 0.0915117680595366,
+      "learning_rate": 1.8499643509474738e-05,
+      "loss": 0.3784,
+      "step": 444
+    },
+    {
+      "epoch": 0.5196321704860605,
+      "grad_norm": 0.0881780613779701,
+      "learning_rate": 1.8488878315900228e-05,
+      "loss": 0.3947,
+      "step": 445
+    },
+    {
+      "epoch": 0.5207998832287257,
+      "grad_norm": 0.08967304558473868,
+      "learning_rate": 1.8478077795187188e-05,
+      "loss": 0.3903,
+      "step": 446
+    },
+    {
+      "epoch": 0.5219675959713911,
+      "grad_norm": 0.10129759027664205,
+      "learning_rate": 1.8467241992282842e-05,
+      "loss": 0.4188,
+      "step": 447
+    },
+    {
+      "epoch": 0.5231353087140563,
+      "grad_norm": 0.0828618847180962,
+      "learning_rate": 1.8456370952281243e-05,
+      "loss": 0.3988,
+      "step": 448
+    },
+    {
+      "epoch": 0.5243030214567217,
+      "grad_norm": 0.09019776841293548,
+      "learning_rate": 1.844546472042308e-05,
+      "loss": 0.4026,
+      "step": 449
+    },
+    {
+      "epoch": 0.525470734199387,
+      "grad_norm": 0.08560781186202324,
+      "learning_rate": 1.84345233420955e-05,
+      "loss": 0.3848,
+      "step": 450
+    },
+    {
+      "epoch": 0.5266384469420523,
+      "grad_norm": 0.09446965333610492,
+      "learning_rate": 1.8423546862831917e-05,
+      "loss": 0.4329,
+      "step": 451
+    },
+    {
+      "epoch": 0.5278061596847176,
+      "grad_norm": 0.08623687753412126,
+      "learning_rate": 1.8412535328311813e-05,
+      "loss": 0.3967,
+      "step": 452
+    },
+    {
+      "epoch": 0.5289738724273829,
+      "grad_norm": 0.07846293315661834,
+      "learning_rate": 1.8401488784360567e-05,
+      "loss": 0.377,
+      "step": 453
+    },
+    {
+      "epoch": 0.5301415851700482,
+      "grad_norm": 0.08289001475991109,
+      "learning_rate": 1.8390407276949235e-05,
+      "loss": 0.3829,
+      "step": 454
+    },
+    {
+      "epoch": 0.5313092979127134,
+      "grad_norm": 0.08867240380108897,
+      "learning_rate": 1.8379290852194395e-05,
+      "loss": 0.4091,
+      "step": 455
+    },
+    {
+      "epoch": 0.5324770106553788,
+      "grad_norm": 0.08248685091210971,
+      "learning_rate": 1.836813955635793e-05,
+      "loss": 0.3976,
+      "step": 456
+    },
+    {
+      "epoch": 0.533644723398044,
+      "grad_norm": 0.08325986221614923,
+      "learning_rate": 1.835695343584683e-05,
+      "loss": 0.39,
+      "step": 457
+    },
+    {
+      "epoch": 0.5348124361407094,
+      "grad_norm": 0.08065998308005909,
+      "learning_rate": 1.834573253721303e-05,
+      "loss": 0.3732,
+      "step": 458
+    },
+    {
+      "epoch": 0.5359801488833746,
+      "grad_norm": 0.09502593730007995,
+      "learning_rate": 1.8334476907153177e-05,
+      "loss": 0.4024,
+      "step": 459
+    },
+    {
+      "epoch": 0.53714786162604,
+      "grad_norm": 0.08508443898804512,
+      "learning_rate": 1.8323186592508474e-05,
+      "loss": 0.4138,
+      "step": 460
+    },
+    {
+      "epoch": 0.5383155743687053,
+      "grad_norm": 0.08583850229262162,
+      "learning_rate": 1.8311861640264456e-05,
+      "loss": 0.4018,
+      "step": 461
+    },
+    {
+      "epoch": 0.5394832871113706,
+      "grad_norm": 0.08683924039234107,
+      "learning_rate": 1.8300502097550807e-05,
+      "loss": 0.407,
+      "step": 462
+    },
+    {
+      "epoch": 0.5406509998540359,
+      "grad_norm": 0.08102012569666316,
+      "learning_rate": 1.8289108011641163e-05,
+      "loss": 0.3838,
+      "step": 463
+    },
+    {
+      "epoch": 0.5418187125967012,
+      "grad_norm": 0.08837402630907959,
+      "learning_rate": 1.8277679429952913e-05,
+      "loss": 0.387,
+      "step": 464
+    },
+    {
+      "epoch": 0.5429864253393665,
+      "grad_norm": 0.08364987886790341,
+      "learning_rate": 1.8266216400047002e-05,
+      "loss": 0.3919,
+      "step": 465
+    },
+    {
+      "epoch": 0.5441541380820318,
+      "grad_norm": 0.09226532263315484,
+      "learning_rate": 1.825471896962774e-05,
+      "loss": 0.3847,
+      "step": 466
+    },
+    {
+      "epoch": 0.5453218508246971,
+      "grad_norm": 0.08402683814369119,
+      "learning_rate": 1.8243187186542594e-05,
+      "loss": 0.3805,
+      "step": 467
+    },
+    {
+      "epoch": 0.5464895635673624,
+      "grad_norm": 0.08785917675007802,
+      "learning_rate": 1.8231621098781983e-05,
+      "loss": 0.4134,
+      "step": 468
+    },
+    {
+      "epoch": 0.5476572763100277,
+      "grad_norm": 0.08359072453863335,
+      "learning_rate": 1.8220020754479104e-05,
+      "loss": 0.3897,
+      "step": 469
+    },
+    {
+      "epoch": 0.5488249890526931,
+      "grad_norm": 0.08297901517680654,
+      "learning_rate": 1.8208386201909702e-05,
+      "loss": 0.3699,
+      "step": 470
+    },
+    {
+      "epoch": 0.5499927017953583,
+      "grad_norm": 0.08779947058915638,
+      "learning_rate": 1.8196717489491885e-05,
+      "loss": 0.4088,
+      "step": 471
+    },
+    {
+      "epoch": 0.5511604145380237,
+      "grad_norm": 0.09566889180388209,
+      "learning_rate": 1.8185014665785936e-05,
+      "loss": 0.3897,
+      "step": 472
+    },
+    {
+      "epoch": 0.5523281272806889,
+      "grad_norm": 0.09014022326150611,
+      "learning_rate": 1.817327777949407e-05,
+      "loss": 0.4014,
+      "step": 473
+    },
+    {
+      "epoch": 0.5534958400233543,
+      "grad_norm": 0.08489370061456808,
+      "learning_rate": 1.8161506879460274e-05,
+      "loss": 0.3932,
+      "step": 474
+    },
+    {
+      "epoch": 0.5546635527660195,
+      "grad_norm": 0.08469786876457724,
+      "learning_rate": 1.8149702014670074e-05,
+      "loss": 0.3853,
+      "step": 475
+    },
+    {
+      "epoch": 0.5558312655086849,
+      "grad_norm": 0.08331229684827948,
+      "learning_rate": 1.8137863234250346e-05,
+      "loss": 0.4025,
+      "step": 476
+    },
+    {
+      "epoch": 0.5569989782513501,
+      "grad_norm": 0.07743416989064332,
+      "learning_rate": 1.8125990587469125e-05,
+      "loss": 0.3853,
+      "step": 477
+    },
+    {
+      "epoch": 0.5581666909940155,
+      "grad_norm": 0.09616350243474774,
+      "learning_rate": 1.8114084123735355e-05,
+      "loss": 0.423,
+      "step": 478
+    },
+    {
+      "epoch": 0.5593344037366808,
+      "grad_norm": 0.08984881270114987,
+      "learning_rate": 1.8102143892598734e-05,
+      "loss": 0.4012,
+      "step": 479
+    },
+    {
+      "epoch": 0.5605021164793461,
+      "grad_norm": 0.08633927979787359,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 0.3984,
+      "step": 480
+    },
+    {
+      "epoch": 0.5616698292220114,
+      "grad_norm": 0.08978747818215024,
+      "learning_rate": 1.8078162327018115e-05,
+      "loss": 0.3718,
+      "step": 481
+    },
+    {
+      "epoch": 0.5628375419646767,
+      "grad_norm": 0.08779388489871119,
+      "learning_rate": 1.8066121092375303e-05,
+      "loss": 0.3795,
+      "step": 482
+    },
+    {
+      "epoch": 0.564005254707342,
+      "grad_norm": 0.09238106783688134,
+      "learning_rate": 1.8054046289931585e-05,
+      "loss": 0.4053,
+      "step": 483
+    },
+    {
+      "epoch": 0.5651729674500073,
+      "grad_norm": 0.09546231615997165,
+      "learning_rate": 1.8041937969937206e-05,
+      "loss": 0.4019,
+      "step": 484
+    },
+    {
+      "epoch": 0.5663406801926726,
+      "grad_norm": 0.08162357510112475,
+      "learning_rate": 1.80297961827819e-05,
+      "loss": 0.4092,
+      "step": 485
+    },
+    {
+      "epoch": 0.567508392935338,
+      "grad_norm": 0.08235804406982142,
+      "learning_rate": 1.8017620978994676e-05,
+      "loss": 0.4023,
+      "step": 486
+    },
+    {
+      "epoch": 0.5686761056780032,
+      "grad_norm": 0.08619249117608531,
+      "learning_rate": 1.8005412409243604e-05,
+      "loss": 0.3768,
+      "step": 487
+    },
+    {
+      "epoch": 0.5698438184206686,
+      "grad_norm": 0.08635833994282247,
+      "learning_rate": 1.7993170524335614e-05,
+      "loss": 0.4079,
+      "step": 488
+    },
+    {
+      "epoch": 0.5710115311633338,
+      "grad_norm": 0.08137388000462911,
+      "learning_rate": 1.7980895375216282e-05,
+      "loss": 0.3854,
+      "step": 489
+    },
+    {
+      "epoch": 0.5721792439059992,
+      "grad_norm": 0.10218494656911778,
+      "learning_rate": 1.7968587012969606e-05,
+      "loss": 0.4327,
+      "step": 490
+    },
+    {
+      "epoch": 0.5733469566486644,
+      "grad_norm": 0.07960225820480414,
+      "learning_rate": 1.795624548881781e-05,
+      "loss": 0.3766,
+      "step": 491
+    },
+    {
+      "epoch": 0.5745146693913298,
+      "grad_norm": 0.08505416166262397,
+      "learning_rate": 1.7943870854121126e-05,
+      "loss": 0.3893,
+      "step": 492
+    },
+    {
+      "epoch": 0.575682382133995,
+      "grad_norm": 0.08114429262487463,
+      "learning_rate": 1.7931463160377575e-05,
+      "loss": 0.3813,
+      "step": 493
+    },
+    {
+      "epoch": 0.5768500948766604,
+      "grad_norm": 0.07909828707775607,
+      "learning_rate": 1.7919022459222754e-05,
+      "loss": 0.3736,
+      "step": 494
+    },
+    {
+      "epoch": 0.5780178076193256,
+      "grad_norm": 0.0830518317827686,
+      "learning_rate": 1.7906548802429622e-05,
+      "loss": 0.3815,
+      "step": 495
+    },
+    {
+      "epoch": 0.579185520361991,
+      "grad_norm": 0.08428958135726784,
+      "learning_rate": 1.7894042241908293e-05,
+      "loss": 0.4028,
+      "step": 496
+    },
+    {
+      "epoch": 0.5803532331046563,
+      "grad_norm": 0.08119461499191251,
+      "learning_rate": 1.7881502829705808e-05,
+      "loss": 0.3898,
+      "step": 497
+    },
+    {
+      "epoch": 0.5815209458473216,
+      "grad_norm": 0.08270698643490845,
+      "learning_rate": 1.786893061800592e-05,
+      "loss": 0.3839,
+      "step": 498
+    },
+    {
+      "epoch": 0.5826886585899869,
+      "grad_norm": 0.09274062043933143,
+      "learning_rate": 1.785632565912888e-05,
+      "loss": 0.4023,
+      "step": 499
+    },
+    {
+      "epoch": 0.5838563713326521,
+      "grad_norm": 0.07939640429255715,
+      "learning_rate": 1.7843688005531227e-05,
+      "loss": 0.4056,
+      "step": 500
+    },
+    {
+      "epoch": 0.5850240840753175,
+      "grad_norm": 0.0817626084875038,
+      "learning_rate": 1.7831017709805555e-05,
+      "loss": 0.4099,
+      "step": 501
+    },
+    {
+      "epoch": 0.5861917968179827,
+      "grad_norm": 0.08390628228807545,
+      "learning_rate": 1.78183148246803e-05,
+      "loss": 0.3981,
+      "step": 502
+    },
+    {
+      "epoch": 0.5873595095606481,
+      "grad_norm": 0.08297778716417335,
+      "learning_rate": 1.7805579403019523e-05,
+      "loss": 0.3926,
+      "step": 503
+    },
+    {
+      "epoch": 0.5885272223033133,
+      "grad_norm": 0.07944792601598234,
+      "learning_rate": 1.779281149782269e-05,
+      "loss": 0.3702,
+      "step": 504
+    },
+    {
+      "epoch": 0.5896949350459787,
+      "grad_norm": 0.080791609997002,
+      "learning_rate": 1.7780011162224444e-05,
+      "loss": 0.4009,
+      "step": 505
+    },
+    {
+      "epoch": 0.590862647788644,
+      "grad_norm": 0.08215720254761476,
+      "learning_rate": 1.7767178449494403e-05,
+      "loss": 0.3879,
+      "step": 506
+    },
+    {
+      "epoch": 0.5920303605313093,
+      "grad_norm": 0.08404568551613353,
+      "learning_rate": 1.7754313413036907e-05,
+      "loss": 0.4101,
+      "step": 507
+    },
+    {
+      "epoch": 0.5931980732739746,
+      "grad_norm": 0.08462349156240151,
+      "learning_rate": 1.7741416106390828e-05,
+      "loss": 0.4065,
+      "step": 508
+    },
+    {
+      "epoch": 0.5943657860166399,
+      "grad_norm": 0.08291898940184876,
+      "learning_rate": 1.7728486583229323e-05,
+      "loss": 0.386,
+      "step": 509
+    },
+    {
+      "epoch": 0.5955334987593052,
+      "grad_norm": 0.08027803858445824,
+      "learning_rate": 1.771552489735963e-05,
+      "loss": 0.3694,
+      "step": 510
+    },
+    {
+      "epoch": 0.5967012115019705,
+      "grad_norm": 0.08535420625227937,
+      "learning_rate": 1.770253110272283e-05,
+      "loss": 0.3966,
+      "step": 511
+    },
+    {
+      "epoch": 0.5978689242446358,
+      "grad_norm": 0.08211028031138275,
+      "learning_rate": 1.768950525339362e-05,
+      "loss": 0.3938,
+      "step": 512
+    },
+    {
+      "epoch": 0.5990366369873011,
+      "grad_norm": 0.08284465334582522,
+      "learning_rate": 1.7676447403580114e-05,
+      "loss": 0.3854,
+      "step": 513
+    },
+    {
+      "epoch": 0.6002043497299664,
+      "grad_norm": 0.08320112149040973,
+      "learning_rate": 1.7663357607623577e-05,
+      "loss": 0.3937,
+      "step": 514
+    },
+    {
+      "epoch": 0.6013720624726318,
+      "grad_norm": 0.0873262652978579,
+      "learning_rate": 1.7650235919998234e-05,
+      "loss": 0.3792,
+      "step": 515
+    },
+    {
+      "epoch": 0.602539775215297,
+      "grad_norm": 0.07910255424719605,
+      "learning_rate": 1.7637082395311024e-05,
+      "loss": 0.3834,
+      "step": 516
+    },
+    {
+      "epoch": 0.6037074879579624,
+      "grad_norm": 0.0772990142828684,
+      "learning_rate": 1.7623897088301387e-05,
+      "loss": 0.3614,
+      "step": 517
+    },
+    {
+      "epoch": 0.6048752007006276,
+      "grad_norm": 0.0831116084910655,
+      "learning_rate": 1.7610680053841008e-05,
+      "loss": 0.4148,
+      "step": 518
+    },
+    {
+      "epoch": 0.606042913443293,
+      "grad_norm": 0.08860497680675289,
+      "learning_rate": 1.7597431346933633e-05,
+      "loss": 0.3812,
+      "step": 519
+    },
+    {
+      "epoch": 0.6072106261859582,
+      "grad_norm": 0.08036779666256168,
+      "learning_rate": 1.75841510227148e-05,
+      "loss": 0.3614,
+      "step": 520
+    },
+    {
+      "epoch": 0.6083783389286236,
+      "grad_norm": 0.0806736006210393,
+      "learning_rate": 1.757083913645163e-05,
+      "loss": 0.3832,
+      "step": 521
+    },
+    {
+      "epoch": 0.6095460516712888,
+      "grad_norm": 0.08734638395531652,
+      "learning_rate": 1.7557495743542586e-05,
+      "loss": 0.3928,
+      "step": 522
+    },
+    {
+      "epoch": 0.6107137644139542,
+      "grad_norm": 0.08429579211468886,
+      "learning_rate": 1.7544120899517255e-05,
+      "loss": 0.3954,
+      "step": 523
+    },
+    {
+      "epoch": 0.6118814771566194,
+      "grad_norm": 0.08071340535799106,
+      "learning_rate": 1.7530714660036112e-05,
+      "loss": 0.4253,
+      "step": 524
+    },
+    {
+      "epoch": 0.6130491898992848,
+      "grad_norm": 0.08226658873617282,
+      "learning_rate": 1.7517277080890277e-05,
+      "loss": 0.395,
+      "step": 525
+    },
+    {
+      "epoch": 0.6142169026419501,
+      "grad_norm": 0.08481686014021247,
+      "learning_rate": 1.7503808218001305e-05,
+      "loss": 0.3982,
+      "step": 526
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.0858626907658107,
+      "learning_rate": 1.7490308127420928e-05,
+      "loss": 0.4143,
+      "step": 527
+    },
+    {
+      "epoch": 0.6165523281272807,
+      "grad_norm": 0.08515454254250178,
+      "learning_rate": 1.7476776865330847e-05,
+      "loss": 0.3701,
+      "step": 528
+    },
+    {
+      "epoch": 0.617720040869946,
+      "grad_norm": 0.07921531414171196,
+      "learning_rate": 1.7463214488042472e-05,
+      "loss": 0.382,
+      "step": 529
+    },
+    {
+      "epoch": 0.6188877536126113,
+      "grad_norm": 0.08294683358619157,
+      "learning_rate": 1.7449621051996714e-05,
+      "loss": 0.3819,
+      "step": 530
+    },
+    {
+      "epoch": 0.6200554663552766,
+      "grad_norm": 0.08649807158531218,
+      "learning_rate": 1.743599661376373e-05,
+      "loss": 0.3802,
+      "step": 531
+    },
+    {
+      "epoch": 0.6212231790979419,
+      "grad_norm": 0.08573167512096355,
+      "learning_rate": 1.74223412300427e-05,
+      "loss": 0.3876,
+      "step": 532
+    },
+    {
+      "epoch": 0.6223908918406073,
+      "grad_norm": 0.07994693251934189,
+      "learning_rate": 1.7408654957661585e-05,
+      "loss": 0.3932,
+      "step": 533
+    },
+    {
+      "epoch": 0.6235586045832725,
+      "grad_norm": 0.1496404888750872,
+      "learning_rate": 1.739493785357688e-05,
+      "loss": 0.345,
+      "step": 534
+    },
+    {
+      "epoch": 0.6247263173259379,
+      "grad_norm": 0.08814521708492051,
+      "learning_rate": 1.738118997487341e-05,
+      "loss": 0.3782,
+      "step": 535
+    },
+    {
+      "epoch": 0.6258940300686031,
+      "grad_norm": 0.0787936646718142,
+      "learning_rate": 1.736741137876405e-05,
+      "loss": 0.3906,
+      "step": 536
+    },
+    {
+      "epoch": 0.6270617428112685,
+      "grad_norm": 0.09000307098393195,
+      "learning_rate": 1.7353602122589528e-05,
+      "loss": 0.4087,
+      "step": 537
+    },
+    {
+      "epoch": 0.6282294555539337,
+      "grad_norm": 0.08329332462387837,
+      "learning_rate": 1.7339762263818144e-05,
+      "loss": 0.3937,
+      "step": 538
+    },
+    {
+      "epoch": 0.6293971682965991,
+      "grad_norm": 0.0892472866940275,
+      "learning_rate": 1.7325891860045577e-05,
+      "loss": 0.3926,
+      "step": 539
+    },
+    {
+      "epoch": 0.6305648810392643,
+      "grad_norm": 0.09006241564685731,
+      "learning_rate": 1.7311990968994598e-05,
+      "loss": 0.4008,
+      "step": 540
+    },
+    {
+      "epoch": 0.6317325937819297,
+      "grad_norm": 0.08794113433272713,
+      "learning_rate": 1.7298059648514876e-05,
+      "loss": 0.3885,
+      "step": 541
+    },
+    {
+      "epoch": 0.632900306524595,
+      "grad_norm": 0.09983950116612306,
+      "learning_rate": 1.7284097956582694e-05,
+      "loss": 0.3791,
+      "step": 542
+    },
+    {
+      "epoch": 0.6340680192672603,
+      "grad_norm": 0.08238014337059973,
+      "learning_rate": 1.727010595130074e-05,
+      "loss": 0.3756,
+      "step": 543
+    },
+    {
+      "epoch": 0.6352357320099256,
+      "grad_norm": 0.08610461830026582,
+      "learning_rate": 1.725608369089785e-05,
+      "loss": 0.3779,
+      "step": 544
+    },
+    {
+      "epoch": 0.6364034447525908,
+      "grad_norm": 0.0968517870823535,
+      "learning_rate": 1.724203123372877e-05,
+      "loss": 0.3864,
+      "step": 545
+    },
+    {
+      "epoch": 0.6375711574952562,
+      "grad_norm": 0.08482729727415886,
+      "learning_rate": 1.7227948638273918e-05,
+      "loss": 0.3839,
+      "step": 546
+    },
+    {
+      "epoch": 0.6387388702379214,
+      "grad_norm": 0.08228621493477564,
+      "learning_rate": 1.721383596313912e-05,
+      "loss": 0.3761,
+      "step": 547
+    },
+    {
+      "epoch": 0.6399065829805868,
+      "grad_norm": 0.09216931022893587,
+      "learning_rate": 1.7199693267055392e-05,
+      "loss": 0.357,
+      "step": 548
+    },
+    {
+      "epoch": 0.641074295723252,
+      "grad_norm": 0.08645101205189916,
+      "learning_rate": 1.7185520608878685e-05,
+      "loss": 0.3816,
+      "step": 549
+    },
+    {
+      "epoch": 0.6422420084659174,
+      "grad_norm": 0.0777445756791792,
+      "learning_rate": 1.7171318047589637e-05,
+      "loss": 0.3838,
+      "step": 550
+    },
+    {
+      "epoch": 0.6434097212085826,
+      "grad_norm": 0.08595666144718332,
+      "learning_rate": 1.7157085642293328e-05,
+      "loss": 0.3862,
+      "step": 551
+    },
+    {
+      "epoch": 0.644577433951248,
+      "grad_norm": 0.08031201739642516,
+      "learning_rate": 1.7142823452219036e-05,
+      "loss": 0.3869,
+      "step": 552
+    },
+    {
+      "epoch": 0.6457451466939133,
+      "grad_norm": 0.08274060541077213,
+      "learning_rate": 1.712853153672e-05,
+      "loss": 0.3711,
+      "step": 553
+    },
+    {
+      "epoch": 0.6469128594365786,
+      "grad_norm": 0.07853951155097824,
+      "learning_rate": 1.7114209955273154e-05,
+      "loss": 0.3869,
+      "step": 554
+    },
+    {
+      "epoch": 0.6480805721792439,
+      "grad_norm": 0.08695214790477541,
+      "learning_rate": 1.7099858767478888e-05,
+      "loss": 0.4029,
+      "step": 555
+    },
+    {
+      "epoch": 0.6492482849219092,
+      "grad_norm": 0.08193749480035666,
+      "learning_rate": 1.7085478033060808e-05,
+      "loss": 0.3907,
+      "step": 556
+    },
+    {
+      "epoch": 0.6504159976645745,
+      "grad_norm": 0.07975865895756419,
+      "learning_rate": 1.7071067811865477e-05,
+      "loss": 0.3826,
+      "step": 557
+    },
+    {
+      "epoch": 0.6515837104072398,
+      "grad_norm": 0.08175585817781615,
+      "learning_rate": 1.705662816386217e-05,
+      "loss": 0.4015,
+      "step": 558
+    },
+    {
+      "epoch": 0.6527514231499051,
+      "grad_norm": 0.0813549598219114,
+      "learning_rate": 1.7042159149142626e-05,
+      "loss": 0.3843,
+      "step": 559
+    },
+    {
+      "epoch": 0.6539191358925704,
+      "grad_norm": 0.08047591821226387,
+      "learning_rate": 1.7027660827920798e-05,
+      "loss": 0.4068,
+      "step": 560
+    },
+    {
+      "epoch": 0.6550868486352357,
+      "grad_norm": 0.08056448710268416,
+      "learning_rate": 1.7013133260532584e-05,
+      "loss": 0.371,
+      "step": 561
+    },
+    {
+      "epoch": 0.6562545613779011,
+      "grad_norm": 0.07661086990205856,
+      "learning_rate": 1.699857650743562e-05,
+      "loss": 0.393,
+      "step": 562
+    },
+    {
+      "epoch": 0.6574222741205663,
+      "grad_norm": 0.07715630249291607,
+      "learning_rate": 1.6983990629208973e-05,
+      "loss": 0.3673,
+      "step": 563
+    },
+    {
+      "epoch": 0.6585899868632317,
+      "grad_norm": 0.08187400479494157,
+      "learning_rate": 1.696937568655294e-05,
+      "loss": 0.3945,
+      "step": 564
+    },
+    {
+      "epoch": 0.6597576996058969,
+      "grad_norm": 0.07850239038445646,
+      "learning_rate": 1.6954731740288744e-05,
+      "loss": 0.3762,
+      "step": 565
+    },
+    {
+      "epoch": 0.6609254123485623,
+      "grad_norm": 0.07301852380754381,
+      "learning_rate": 1.6940058851358344e-05,
+      "loss": 0.3688,
+      "step": 566
+    },
+    {
+      "epoch": 0.6620931250912275,
+      "grad_norm": 0.08259802083512213,
+      "learning_rate": 1.692535708082412e-05,
+      "loss": 0.4018,
+      "step": 567
+    },
+    {
+      "epoch": 0.6632608378338929,
+      "grad_norm": 0.08238710278327345,
+      "learning_rate": 1.691062648986865e-05,
+      "loss": 0.3786,
+      "step": 568
+    },
+    {
+      "epoch": 0.6644285505765581,
+      "grad_norm": 0.07731254148715028,
+      "learning_rate": 1.6895867139794455e-05,
+      "loss": 0.3746,
+      "step": 569
+    },
+    {
+      "epoch": 0.6655962633192235,
+      "grad_norm": 0.07952316987821033,
+      "learning_rate": 1.6881079092023742e-05,
+      "loss": 0.3932,
+      "step": 570
+    },
+    {
+      "epoch": 0.6667639760618888,
+      "grad_norm": 0.07901262951671865,
+      "learning_rate": 1.6866262408098134e-05,
+      "loss": 0.4016,
+      "step": 571
+    },
+    {
+      "epoch": 0.6679316888045541,
+      "grad_norm": 0.07597724290401711,
+      "learning_rate": 1.6851417149678442e-05,
+      "loss": 0.3603,
+      "step": 572
+    },
+    {
+      "epoch": 0.6690994015472194,
+      "grad_norm": 0.07671062763216194,
+      "learning_rate": 1.683654337854438e-05,
+      "loss": 0.3621,
+      "step": 573
+    },
+    {
+      "epoch": 0.6702671142898847,
+      "grad_norm": 0.07890637110612449,
+      "learning_rate": 1.682164115659432e-05,
+      "loss": 0.3844,
+      "step": 574
+    },
+    {
+      "epoch": 0.67143482703255,
+      "grad_norm": 0.07790488817141691,
+      "learning_rate": 1.6806710545845036e-05,
+      "loss": 0.3844,
+      "step": 575
+    },
+    {
+      "epoch": 0.6726025397752153,
+      "grad_norm": 0.07825206249180414,
+      "learning_rate": 1.679175160843145e-05,
+      "loss": 0.4013,
+      "step": 576
+    },
+    {
+      "epoch": 0.6737702525178806,
+      "grad_norm": 0.0777856119638522,
+      "learning_rate": 1.6776764406606362e-05,
+      "loss": 0.3771,
+      "step": 577
+    },
+    {
+      "epoch": 0.674937965260546,
+      "grad_norm": 0.07822960386073255,
+      "learning_rate": 1.6761749002740195e-05,
+      "loss": 0.3848,
+      "step": 578
+    },
+    {
+      "epoch": 0.6761056780032112,
+      "grad_norm": 0.07817797969857385,
+      "learning_rate": 1.6746705459320746e-05,
+      "loss": 0.3875,
+      "step": 579
+    },
+    {
+      "epoch": 0.6772733907458766,
+      "grad_norm": 0.08372265997730491,
+      "learning_rate": 1.6731633838952905e-05,
+      "loss": 0.3751,
+      "step": 580
+    },
+    {
+      "epoch": 0.6784411034885418,
+      "grad_norm": 0.0770730454366788,
+      "learning_rate": 1.6716534204358416e-05,
+      "loss": 0.3833,
+      "step": 581
+    },
+    {
+      "epoch": 0.6796088162312072,
+      "grad_norm": 0.08135664513777449,
+      "learning_rate": 1.67014066183756e-05,
+      "loss": 0.397,
+      "step": 582
+    },
+    {
+      "epoch": 0.6807765289738724,
+      "grad_norm": 0.07884837757255037,
+      "learning_rate": 1.6686251143959096e-05,
+      "loss": 0.3692,
+      "step": 583
+    },
+    {
+      "epoch": 0.6819442417165378,
+      "grad_norm": 0.07597716044597828,
+      "learning_rate": 1.6671067844179625e-05,
+      "loss": 0.3919,
+      "step": 584
+    },
+    {
+      "epoch": 0.683111954459203,
+      "grad_norm": 0.08039100076654557,
+      "learning_rate": 1.6655856782223682e-05,
+      "loss": 0.3781,
+      "step": 585
+    },
+    {
+      "epoch": 0.6842796672018684,
+      "grad_norm": 0.08058059971074993,
+      "learning_rate": 1.6640618021393305e-05,
+      "loss": 0.3851,
+      "step": 586
+    },
+    {
+      "epoch": 0.6854473799445336,
+      "grad_norm": 0.08016869121339747,
+      "learning_rate": 1.6625351625105796e-05,
+      "loss": 0.3807,
+      "step": 587
+    },
+    {
+      "epoch": 0.686615092687199,
+      "grad_norm": 0.09039401345482465,
+      "learning_rate": 1.6610057656893483e-05,
+      "loss": 0.3981,
+      "step": 588
+    },
+    {
+      "epoch": 0.6877828054298643,
+      "grad_norm": 0.08125119276361589,
+      "learning_rate": 1.6594736180403412e-05,
+      "loss": 0.3743,
+      "step": 589
+    },
+    {
+      "epoch": 0.6889505181725295,
+      "grad_norm": 0.08470958289641675,
+      "learning_rate": 1.657938725939713e-05,
+      "loss": 0.4047,
+      "step": 590
+    },
+    {
+      "epoch": 0.6901182309151949,
+      "grad_norm": 0.08978606367632151,
+      "learning_rate": 1.6564010957750375e-05,
+      "loss": 0.3835,
+      "step": 591
+    },
+    {
+      "epoch": 0.6912859436578601,
+      "grad_norm": 0.07827536527494881,
+      "learning_rate": 1.6548607339452853e-05,
+      "loss": 0.3786,
+      "step": 592
+    },
+    {
+      "epoch": 0.6924536564005255,
+      "grad_norm": 0.08470511741297303,
+      "learning_rate": 1.653317646860793e-05,
+      "loss": 0.3794,
+      "step": 593
+    },
+    {
+      "epoch": 0.6936213691431907,
+      "grad_norm": 0.08292019574958598,
+      "learning_rate": 1.6517718409432405e-05,
+      "loss": 0.3699,
+      "step": 594
+    },
+    {
+      "epoch": 0.6947890818858561,
+      "grad_norm": 0.07970565070125725,
+      "learning_rate": 1.650223322625621e-05,
+      "loss": 0.3884,
+      "step": 595
+    },
+    {
+      "epoch": 0.6959567946285213,
+      "grad_norm": 0.0829974053607285,
+      "learning_rate": 1.6486720983522156e-05,
+      "loss": 0.3774,
+      "step": 596
+    },
+    {
+      "epoch": 0.6971245073711867,
+      "grad_norm": 0.0747305766792251,
+      "learning_rate": 1.6471181745785673e-05,
+      "loss": 0.3928,
+      "step": 597
+    },
+    {
+      "epoch": 0.698292220113852,
+      "grad_norm": 0.08038045827804523,
+      "learning_rate": 1.645561557771453e-05,
+      "loss": 0.3624,
+      "step": 598
+    },
+    {
+      "epoch": 0.6994599328565173,
+      "grad_norm": 0.07562025255066594,
+      "learning_rate": 1.6440022544088553e-05,
+      "loss": 0.3371,
+      "step": 599
+    },
+    {
+      "epoch": 0.7006276455991826,
+      "grad_norm": 0.07521368246187433,
+      "learning_rate": 1.6424402709799404e-05,
+      "loss": 0.3833,
+      "step": 600
+    },
+    {
+      "epoch": 0.7017953583418479,
+      "grad_norm": 0.07641612620716239,
+      "learning_rate": 1.6408756139850243e-05,
+      "loss": 0.3781,
+      "step": 601
+    },
+    {
+      "epoch": 0.7029630710845132,
+      "grad_norm": 0.07587902479083829,
+      "learning_rate": 1.6393082899355516e-05,
+      "loss": 0.3923,
+      "step": 602
+    },
+    {
+      "epoch": 0.7041307838271785,
+      "grad_norm": 0.07876274061365084,
+      "learning_rate": 1.637738305354065e-05,
+      "loss": 0.3897,
+      "step": 603
+    },
+    {
+      "epoch": 0.7052984965698438,
+      "grad_norm": 0.07431240983481445,
+      "learning_rate": 1.63616566677418e-05,
+      "loss": 0.3845,
+      "step": 604
+    },
+    {
+      "epoch": 0.7064662093125091,
+      "grad_norm": 0.07745047289110622,
+      "learning_rate": 1.634590380740556e-05,
+      "loss": 0.3901,
+      "step": 605
+    },
+    {
+      "epoch": 0.7076339220551744,
+      "grad_norm": 0.07846235195601511,
+      "learning_rate": 1.6330124538088705e-05,
+      "loss": 0.3954,
+      "step": 606
+    },
+    {
+      "epoch": 0.7088016347978398,
+      "grad_norm": 0.07679355819259684,
+      "learning_rate": 1.631431892545791e-05,
+      "loss": 0.3774,
+      "step": 607
+    },
+    {
+      "epoch": 0.709969347540505,
+      "grad_norm": 0.08266946113424178,
+      "learning_rate": 1.629848703528949e-05,
+      "loss": 0.3686,
+      "step": 608
+    },
+    {
+      "epoch": 0.7111370602831704,
+      "grad_norm": 0.08649060107459645,
+      "learning_rate": 1.6282628933469097e-05,
+      "loss": 0.3772,
+      "step": 609
+    },
+    {
+      "epoch": 0.7123047730258356,
+      "grad_norm": 0.07913972246428395,
+      "learning_rate": 1.6266744685991487e-05,
+      "loss": 0.3745,
+      "step": 610
+    },
+    {
+      "epoch": 0.713472485768501,
+      "grad_norm": 0.08384802681073406,
+      "learning_rate": 1.6250834358960203e-05,
+      "loss": 0.3956,
+      "step": 611
+    },
+    {
+      "epoch": 0.7146401985111662,
+      "grad_norm": 0.07813523751338723,
+      "learning_rate": 1.6234898018587336e-05,
+      "loss": 0.3743,
+      "step": 612
+    },
+    {
+      "epoch": 0.7158079112538316,
+      "grad_norm": 0.08164067650408241,
+      "learning_rate": 1.6218935731193223e-05,
+      "loss": 0.3808,
+      "step": 613
+    },
+    {
+      "epoch": 0.7169756239964968,
+      "grad_norm": 0.08344931817552902,
+      "learning_rate": 1.620294756320619e-05,
+      "loss": 0.3643,
+      "step": 614
+    },
+    {
+      "epoch": 0.7181433367391622,
+      "grad_norm": 0.07844342300355021,
+      "learning_rate": 1.6186933581162254e-05,
+      "loss": 0.3835,
+      "step": 615
+    },
+    {
+      "epoch": 0.7193110494818274,
+      "grad_norm": 0.07398600522215758,
+      "learning_rate": 1.6170893851704875e-05,
+      "loss": 0.3952,
+      "step": 616
+    },
+    {
+      "epoch": 0.7204787622244928,
+      "grad_norm": 0.07914196558545777,
+      "learning_rate": 1.6154828441584655e-05,
+      "loss": 0.3893,
+      "step": 617
+    },
+    {
+      "epoch": 0.7216464749671581,
+      "grad_norm": 0.07586819115847923,
+      "learning_rate": 1.613873741765907e-05,
+      "loss": 0.4011,
+      "step": 618
+    },
+    {
+      "epoch": 0.7228141877098234,
+      "grad_norm": 0.07564493511358818,
+      "learning_rate": 1.6122620846892182e-05,
+      "loss": 0.3719,
+      "step": 619
+    },
+    {
+      "epoch": 0.7239819004524887,
+      "grad_norm": 0.07796492236533885,
+      "learning_rate": 1.6106478796354382e-05,
+      "loss": 0.3882,
+      "step": 620
+    },
+    {
+      "epoch": 0.725149613195154,
+      "grad_norm": 0.07869061662299254,
+      "learning_rate": 1.6090311333222088e-05,
+      "loss": 0.3911,
+      "step": 621
+    },
+    {
+      "epoch": 0.7263173259378193,
+      "grad_norm": 0.08001749302910599,
+      "learning_rate": 1.607411852477748e-05,
+      "loss": 0.3771,
+      "step": 622
+    },
+    {
+      "epoch": 0.7274850386804846,
+      "grad_norm": 0.08029632158826357,
+      "learning_rate": 1.60579004384082e-05,
+      "loss": 0.3873,
+      "step": 623
+    },
+    {
+      "epoch": 0.7286527514231499,
+      "grad_norm": 0.08112628822690886,
+      "learning_rate": 1.604165714160711e-05,
+      "loss": 0.3909,
+      "step": 624
+    },
+    {
+      "epoch": 0.7298204641658153,
+      "grad_norm": 0.08337786691963645,
+      "learning_rate": 1.6025388701971963e-05,
+      "loss": 0.3945,
+      "step": 625
+    },
+    {
+      "epoch": 0.7309881769084805,
+      "grad_norm": 0.08119653129350403,
+      "learning_rate": 1.600909518720517e-05,
+      "loss": 0.3643,
+      "step": 626
+    },
+    {
+      "epoch": 0.7321558896511459,
+      "grad_norm": 0.07782684463427958,
+      "learning_rate": 1.599277666511347e-05,
+      "loss": 0.3883,
+      "step": 627
+    },
+    {
+      "epoch": 0.7333236023938111,
+      "grad_norm": 0.07661682490250644,
+      "learning_rate": 1.597643320360769e-05,
+      "loss": 0.3843,
+      "step": 628
+    },
+    {
+      "epoch": 0.7344913151364765,
+      "grad_norm": 0.0756773492133668,
+      "learning_rate": 1.5960064870702434e-05,
+      "loss": 0.3749,
+      "step": 629
+    },
+    {
+      "epoch": 0.7356590278791417,
+      "grad_norm": 0.0759730466697951,
+      "learning_rate": 1.594367173451582e-05,
+      "loss": 0.3934,
+      "step": 630
+    },
+    {
+      "epoch": 0.7368267406218071,
+      "grad_norm": 0.07926779276503533,
+      "learning_rate": 1.5927253863269174e-05,
+      "loss": 0.3741,
+      "step": 631
+    },
+    {
+      "epoch": 0.7379944533644723,
+      "grad_norm": 0.07884218993714424,
+      "learning_rate": 1.5910811325286768e-05,
+      "loss": 0.4017,
+      "step": 632
+    },
+    {
+      "epoch": 0.7391621661071377,
+      "grad_norm": 0.07616364776029626,
+      "learning_rate": 1.5894344188995522e-05,
+      "loss": 0.3695,
+      "step": 633
+    },
+    {
+      "epoch": 0.740329878849803,
+      "grad_norm": 0.0762876060931229,
+      "learning_rate": 1.5877852522924733e-05,
+      "loss": 0.361,
+      "step": 634
+    },
+    {
+      "epoch": 0.7414975915924683,
+      "grad_norm": 0.07683755932202181,
+      "learning_rate": 1.5861336395705765e-05,
+      "loss": 0.3804,
+      "step": 635
+    },
+    {
+      "epoch": 0.7426653043351336,
+      "grad_norm": 0.07816323671220654,
+      "learning_rate": 1.58447958760718e-05,
+      "loss": 0.3926,
+      "step": 636
+    },
+    {
+      "epoch": 0.7438330170777988,
+      "grad_norm": 0.07833281435572936,
+      "learning_rate": 1.5828231032857503e-05,
+      "loss": 0.3692,
+      "step": 637
+    },
+    {
+      "epoch": 0.7450007298204642,
+      "grad_norm": 0.08654737243582108,
+      "learning_rate": 1.581164193499879e-05,
+      "loss": 0.3858,
+      "step": 638
+    },
+    {
+      "epoch": 0.7461684425631294,
+      "grad_norm": 0.07897494462443486,
+      "learning_rate": 1.5795028651532502e-05,
+      "loss": 0.3795,
+      "step": 639
+    },
+    {
+      "epoch": 0.7473361553057948,
+      "grad_norm": 0.07798182728581524,
+      "learning_rate": 1.577839125159613e-05,
+      "loss": 0.3877,
+      "step": 640
+    },
+    {
+      "epoch": 0.74850386804846,
+      "grad_norm": 0.0786099931651542,
+      "learning_rate": 1.576172980442753e-05,
+      "loss": 0.3736,
+      "step": 641
+    },
+    {
+      "epoch": 0.7496715807911254,
+      "grad_norm": 0.07599874395775832,
+      "learning_rate": 1.5745044379364637e-05,
+      "loss": 0.3669,
+      "step": 642
+    },
+    {
+      "epoch": 0.7508392935337906,
+      "grad_norm": 0.07715402450484536,
+      "learning_rate": 1.572833504584516e-05,
+      "loss": 0.3932,
+      "step": 643
+    },
+    {
+      "epoch": 0.752007006276456,
+      "grad_norm": 0.08015785246568063,
+      "learning_rate": 1.5711601873406314e-05,
+      "loss": 0.3891,
+      "step": 644
+    },
+    {
+      "epoch": 0.7531747190191213,
+      "grad_norm": 0.07755857590010232,
+      "learning_rate": 1.569484493168452e-05,
+      "loss": 0.3913,
+      "step": 645
+    },
+    {
+      "epoch": 0.7543424317617866,
+      "grad_norm": 0.082547848671012,
+      "learning_rate": 1.5678064290415122e-05,
+      "loss": 0.3926,
+      "step": 646
+    },
+    {
+      "epoch": 0.7555101445044519,
+      "grad_norm": 0.08274049036157854,
+      "learning_rate": 1.5661260019432077e-05,
+      "loss": 0.4049,
+      "step": 647
+    },
+    {
+      "epoch": 0.7566778572471172,
+      "grad_norm": 0.08218219698683112,
+      "learning_rate": 1.5644432188667695e-05,
+      "loss": 0.3795,
+      "step": 648
+    },
+    {
+      "epoch": 0.7578455699897825,
+      "grad_norm": 0.07635170465336136,
+      "learning_rate": 1.5627580868152316e-05,
+      "loss": 0.3805,
+      "step": 649
+    },
+    {
+      "epoch": 0.7590132827324478,
+      "grad_norm": 0.08214129261317375,
+      "learning_rate": 1.5610706128014054e-05,
+      "loss": 0.3836,
+      "step": 650
+    },
+    {
+      "epoch": 0.7601809954751131,
+      "grad_norm": 0.0784095196251521,
+      "learning_rate": 1.5593808038478465e-05,
+      "loss": 0.3645,
+      "step": 651
+    },
+    {
+      "epoch": 0.7613487082177784,
+      "grad_norm": 0.07636062590004973,
+      "learning_rate": 1.5576886669868297e-05,
+      "loss": 0.3976,
+      "step": 652
+    },
+    {
+      "epoch": 0.7625164209604437,
+      "grad_norm": 0.08315370704787003,
+      "learning_rate": 1.5559942092603152e-05,
+      "loss": 0.3734,
+      "step": 653
+    },
+    {
+      "epoch": 0.7636841337031091,
+      "grad_norm": 0.08055405643989327,
+      "learning_rate": 1.554297437719923e-05,
+      "loss": 0.3786,
+      "step": 654
+    },
+    {
+      "epoch": 0.7648518464457743,
+      "grad_norm": 0.0794816540743545,
+      "learning_rate": 1.5525983594269026e-05,
+      "loss": 0.3876,
+      "step": 655
+    },
+    {
+      "epoch": 0.7660195591884397,
+      "grad_norm": 0.07581385566690736,
+      "learning_rate": 1.5508969814521026e-05,
+      "loss": 0.3787,
+      "step": 656
+    },
+    {
+      "epoch": 0.7671872719311049,
+      "grad_norm": 0.08195991903284303,
+      "learning_rate": 1.5491933108759417e-05,
+      "loss": 0.3983,
+      "step": 657
+    },
+    {
+      "epoch": 0.7683549846737703,
+      "grad_norm": 0.07250714767046053,
+      "learning_rate": 1.5474873547883803e-05,
+      "loss": 0.3626,
+      "step": 658
+    },
+    {
+      "epoch": 0.7695226974164355,
+      "grad_norm": 0.08405666864248548,
+      "learning_rate": 1.545779120288889e-05,
+      "loss": 0.3854,
+      "step": 659
+    },
+    {
+      "epoch": 0.7706904101591009,
+      "grad_norm": 0.0769794862519204,
+      "learning_rate": 1.5440686144864207e-05,
+      "loss": 0.39,
+      "step": 660
+    },
+    {
+      "epoch": 0.7718581229017661,
+      "grad_norm": 0.0805412530109551,
+      "learning_rate": 1.5423558444993808e-05,
+      "loss": 0.3622,
+      "step": 661
+    },
+    {
+      "epoch": 0.7730258356444315,
+      "grad_norm": 0.07893749227000672,
+      "learning_rate": 1.5406408174555978e-05,
+      "loss": 0.3688,
+      "step": 662
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 0.0755455373786893,
+      "learning_rate": 1.5389235404922912e-05,
+      "loss": 0.3869,
+      "step": 663
+    },
+    {
+      "epoch": 0.7753612611297621,
+      "grad_norm": 0.08315679750405869,
+      "learning_rate": 1.5372040207560457e-05,
+      "loss": 0.3926,
+      "step": 664
+    },
+    {
+      "epoch": 0.7765289738724274,
+      "grad_norm": 0.08021710958630009,
+      "learning_rate": 1.5354822654027787e-05,
+      "loss": 0.3862,
+      "step": 665
+    },
+    {
+      "epoch": 0.7776966866150927,
+      "grad_norm": 0.07878698889108533,
+      "learning_rate": 1.5337582815977102e-05,
+      "loss": 0.376,
+      "step": 666
+    },
+    {
+      "epoch": 0.778864399357758,
+      "grad_norm": 0.0801482152928564,
+      "learning_rate": 1.5320320765153367e-05,
+      "loss": 0.3699,
+      "step": 667
+    },
+    {
+      "epoch": 0.7800321121004233,
+      "grad_norm": 0.0836464173822777,
+      "learning_rate": 1.5303036573393964e-05,
+      "loss": 0.3782,
+      "step": 668
+    },
+    {
+      "epoch": 0.7811998248430886,
+      "grad_norm": 0.08252169435339383,
+      "learning_rate": 1.528573031262842e-05,
+      "loss": 0.3835,
+      "step": 669
+    },
+    {
+      "epoch": 0.782367537585754,
+      "grad_norm": 0.08061446994305019,
+      "learning_rate": 1.5268402054878116e-05,
+      "loss": 0.3678,
+      "step": 670
+    },
+    {
+      "epoch": 0.7835352503284192,
+      "grad_norm": 0.08284481119815125,
+      "learning_rate": 1.5251051872255963e-05,
+      "loss": 0.3971,
+      "step": 671
+    },
+    {
+      "epoch": 0.7847029630710846,
+      "grad_norm": 0.0788874079861648,
+      "learning_rate": 1.5233679836966122e-05,
+      "loss": 0.3814,
+      "step": 672
+    },
+    {
+      "epoch": 0.7858706758137498,
+      "grad_norm": 0.07773604876193624,
+      "learning_rate": 1.5216286021303685e-05,
+      "loss": 0.3858,
+      "step": 673
+    },
+    {
+      "epoch": 0.7870383885564152,
+      "grad_norm": 0.0791955498752232,
+      "learning_rate": 1.5198870497654397e-05,
+      "loss": 0.3866,
+      "step": 674
+    },
+    {
+      "epoch": 0.7882061012990804,
+      "grad_norm": 0.0798999748411578,
+      "learning_rate": 1.5181433338494337e-05,
+      "loss": 0.375,
+      "step": 675
+    },
+    {
+      "epoch": 0.7893738140417458,
+      "grad_norm": 0.07616956765172526,
+      "learning_rate": 1.5163974616389621e-05,
+      "loss": 0.3749,
+      "step": 676
+    },
+    {
+      "epoch": 0.790541526784411,
+      "grad_norm": 0.07765194112276463,
+      "learning_rate": 1.5146494403996101e-05,
+      "loss": 0.3987,
+      "step": 677
+    },
+    {
+      "epoch": 0.7917092395270764,
+      "grad_norm": 0.0811134283893946,
+      "learning_rate": 1.5128992774059063e-05,
+      "loss": 0.3726,
+      "step": 678
+    },
+    {
+      "epoch": 0.7928769522697416,
+      "grad_norm": 0.0775861196227279,
+      "learning_rate": 1.5111469799412927e-05,
+      "loss": 0.3738,
+      "step": 679
+    },
+    {
+      "epoch": 0.794044665012407,
+      "grad_norm": 0.078057247810283,
+      "learning_rate": 1.5093925552980934e-05,
+      "loss": 0.3921,
+      "step": 680
+    },
+    {
+      "epoch": 0.7952123777550723,
+      "grad_norm": 0.08241859560355469,
+      "learning_rate": 1.5076360107774855e-05,
+      "loss": 0.3791,
+      "step": 681
+    },
+    {
+      "epoch": 0.7963800904977375,
+      "grad_norm": 0.07216388956585741,
+      "learning_rate": 1.5058773536894685e-05,
+      "loss": 0.3841,
+      "step": 682
+    },
+    {
+      "epoch": 0.7975478032404029,
+      "grad_norm": 0.08063907524488093,
+      "learning_rate": 1.504116591352832e-05,
+      "loss": 0.3659,
+      "step": 683
+    },
+    {
+      "epoch": 0.7987155159830681,
+      "grad_norm": 0.0801965320821325,
+      "learning_rate": 1.5023537310951284e-05,
+      "loss": 0.3755,
+      "step": 684
+    },
+    {
+      "epoch": 0.7998832287257335,
+      "grad_norm": 0.08567531125356652,
+      "learning_rate": 1.5005887802526398e-05,
+      "loss": 0.4106,
+      "step": 685
+    },
+    {
+      "epoch": 0.8010509414683987,
+      "grad_norm": 0.0796021579824488,
+      "learning_rate": 1.4988217461703488e-05,
+      "loss": 0.377,
+      "step": 686
+    },
+    {
+      "epoch": 0.8022186542110641,
+      "grad_norm": 0.08250301800984143,
+      "learning_rate": 1.4970526362019077e-05,
+      "loss": 0.3821,
+      "step": 687
+    },
+    {
+      "epoch": 0.8033863669537293,
+      "grad_norm": 0.07812639184787751,
+      "learning_rate": 1.4952814577096073e-05,
+      "loss": 0.3684,
+      "step": 688
+    },
+    {
+      "epoch": 0.8045540796963947,
+      "grad_norm": 0.07945074509613084,
+      "learning_rate": 1.493508218064347e-05,
+      "loss": 0.3756,
+      "step": 689
+    },
+    {
+      "epoch": 0.80572179243906,
+      "grad_norm": 0.08567708460138454,
+      "learning_rate": 1.491732924645604e-05,
+      "loss": 0.4059,
+      "step": 690
+    },
+    {
+      "epoch": 0.8068895051817253,
+      "grad_norm": 0.07574325811308835,
+      "learning_rate": 1.489955584841402e-05,
+      "loss": 0.3898,
+      "step": 691
+    },
+    {
+      "epoch": 0.8080572179243906,
+      "grad_norm": 0.08017380384477853,
+      "learning_rate": 1.4881762060482814e-05,
+      "loss": 0.3852,
+      "step": 692
+    },
+    {
+      "epoch": 0.8092249306670559,
+      "grad_norm": 0.07887113989919158,
+      "learning_rate": 1.4863947956712679e-05,
+      "loss": 0.3725,
+      "step": 693
+    },
+    {
+      "epoch": 0.8103926434097212,
+      "grad_norm": 0.07092090134136286,
+      "learning_rate": 1.4846113611238413e-05,
+      "loss": 0.3529,
+      "step": 694
+    },
+    {
+      "epoch": 0.8115603561523865,
+      "grad_norm": 0.081654484740652,
+      "learning_rate": 1.4828259098279059e-05,
+      "loss": 0.3917,
+      "step": 695
+    },
+    {
+      "epoch": 0.8127280688950518,
+      "grad_norm": 0.07494824281433919,
+      "learning_rate": 1.4810384492137582e-05,
+      "loss": 0.3736,
+      "step": 696
+    },
+    {
+      "epoch": 0.8138957816377171,
+      "grad_norm": 0.07929825822420254,
+      "learning_rate": 1.479248986720057e-05,
+      "loss": 0.3782,
+      "step": 697
+    },
+    {
+      "epoch": 0.8150634943803824,
+      "grad_norm": 0.08075278818822018,
+      "learning_rate": 1.4774575297937921e-05,
+      "loss": 0.3699,
+      "step": 698
+    },
+    {
+      "epoch": 0.8162312071230478,
+      "grad_norm": 0.0783063318083679,
+      "learning_rate": 1.4756640858902531e-05,
+      "loss": 0.3883,
+      "step": 699
+    },
+    {
+      "epoch": 0.817398919865713,
+      "grad_norm": 0.07429156153403146,
+      "learning_rate": 1.4738686624729987e-05,
+      "loss": 0.3919,
+      "step": 700
+    },
+    {
+      "epoch": 0.8185666326083784,
+      "grad_norm": 0.08140792369531477,
+      "learning_rate": 1.4720712670138255e-05,
+      "loss": 0.3691,
+      "step": 701
+    },
+    {
+      "epoch": 0.8197343453510436,
+      "grad_norm": 0.07543051423451785,
+      "learning_rate": 1.4702719069927368e-05,
+      "loss": 0.3703,
+      "step": 702
+    },
+    {
+      "epoch": 0.820902058093709,
+      "grad_norm": 0.0755732356128336,
+      "learning_rate": 1.4684705898979119e-05,
+      "loss": 0.3757,
+      "step": 703
+    },
+    {
+      "epoch": 0.8220697708363742,
+      "grad_norm": 0.0712993026386215,
+      "learning_rate": 1.4666673232256738e-05,
+      "loss": 0.365,
+      "step": 704
+    },
+    {
+      "epoch": 0.8232374835790396,
+      "grad_norm": 0.07882860100603642,
+      "learning_rate": 1.46486211448046e-05,
+      "loss": 0.3902,
+      "step": 705
+    },
+    {
+      "epoch": 0.8244051963217048,
+      "grad_norm": 0.07439905909384101,
+      "learning_rate": 1.4630549711747888e-05,
+      "loss": 0.3654,
+      "step": 706
+    },
+    {
+      "epoch": 0.8255729090643702,
+      "grad_norm": 0.07914686185195155,
+      "learning_rate": 1.4612459008292307e-05,
+      "loss": 0.3717,
+      "step": 707
+    },
+    {
+      "epoch": 0.8267406218070354,
+      "grad_norm": 0.08272308835634432,
+      "learning_rate": 1.4594349109723744e-05,
+      "loss": 0.3685,
+      "step": 708
+    },
+    {
+      "epoch": 0.8279083345497008,
+      "grad_norm": 0.07530449460481148,
+      "learning_rate": 1.4576220091407981e-05,
+      "loss": 0.379,
+      "step": 709
+    },
+    {
+      "epoch": 0.8290760472923661,
+      "grad_norm": 0.08178490479419573,
+      "learning_rate": 1.4558072028790356e-05,
+      "loss": 0.3785,
+      "step": 710
+    },
+    {
+      "epoch": 0.8302437600350314,
+      "grad_norm": 0.07478480088763145,
+      "learning_rate": 1.4539904997395468e-05,
+      "loss": 0.3531,
+      "step": 711
+    },
+    {
+      "epoch": 0.8314114727776967,
+      "grad_norm": 0.07690442945011591,
+      "learning_rate": 1.4521719072826858e-05,
+      "loss": 0.3788,
+      "step": 712
+    },
+    {
+      "epoch": 0.832579185520362,
+      "grad_norm": 0.07584739393828409,
+      "learning_rate": 1.4503514330766689e-05,
+      "loss": 0.3793,
+      "step": 713
+    },
+    {
+      "epoch": 0.8337468982630273,
+      "grad_norm": 0.07817216517985535,
+      "learning_rate": 1.4485290846975431e-05,
+      "loss": 0.3808,
+      "step": 714
+    },
+    {
+      "epoch": 0.8349146110056926,
+      "grad_norm": 0.07475193700035321,
+      "learning_rate": 1.4467048697291554e-05,
+      "loss": 0.3713,
+      "step": 715
+    },
+    {
+      "epoch": 0.8360823237483579,
+      "grad_norm": 0.07496765409106003,
+      "learning_rate": 1.4448787957631209e-05,
+      "loss": 0.3636,
+      "step": 716
+    },
+    {
+      "epoch": 0.8372500364910233,
+      "grad_norm": 0.0789335687092063,
+      "learning_rate": 1.4430508703987914e-05,
+      "loss": 0.3605,
+      "step": 717
+    },
+    {
+      "epoch": 0.8384177492336885,
+      "grad_norm": 0.08168487022861051,
+      "learning_rate": 1.4412211012432213e-05,
+      "loss": 0.401,
+      "step": 718
+    },
+    {
+      "epoch": 0.8395854619763539,
+      "grad_norm": 0.07479792475615664,
+      "learning_rate": 1.4393894959111415e-05,
+      "loss": 0.3753,
+      "step": 719
+    },
+    {
+      "epoch": 0.8407531747190191,
+      "grad_norm": 0.0750177034821023,
+      "learning_rate": 1.4375560620249209e-05,
+      "loss": 0.3717,
+      "step": 720
+    },
+    {
+      "epoch": 0.8419208874616845,
+      "grad_norm": 0.08177298925442884,
+      "learning_rate": 1.4357208072145405e-05,
+      "loss": 0.3929,
+      "step": 721
+    },
+    {
+      "epoch": 0.8430886002043497,
+      "grad_norm": 0.08040290254029245,
+      "learning_rate": 1.4338837391175582e-05,
+      "loss": 0.3857,
+      "step": 722
+    },
+    {
+      "epoch": 0.8442563129470151,
+      "grad_norm": 0.07455009307883982,
+      "learning_rate": 1.4320448653790782e-05,
+      "loss": 0.3674,
+      "step": 723
+    },
+    {
+      "epoch": 0.8454240256896803,
+      "grad_norm": 0.08611010070155466,
+      "learning_rate": 1.430204193651719e-05,
+      "loss": 0.3797,
+      "step": 724
+    },
+    {
+      "epoch": 0.8465917384323457,
+      "grad_norm": 0.0781509491821336,
+      "learning_rate": 1.4283617315955815e-05,
+      "loss": 0.3679,
+      "step": 725
+    },
+    {
+      "epoch": 0.847759451175011,
+      "grad_norm": 0.07794521315691431,
+      "learning_rate": 1.4265174868782172e-05,
+      "loss": 0.3944,
+      "step": 726
+    },
+    {
+      "epoch": 0.8489271639176762,
+      "grad_norm": 0.08717419882297162,
+      "learning_rate": 1.4246714671745964e-05,
+      "loss": 0.3778,
+      "step": 727
+    },
+    {
+      "epoch": 0.8500948766603416,
+      "grad_norm": 0.07771317658808101,
+      "learning_rate": 1.4228236801670762e-05,
+      "loss": 0.3707,
+      "step": 728
+    },
+    {
+      "epoch": 0.8512625894030068,
+      "grad_norm": 0.08053833950506085,
+      "learning_rate": 1.4209741335453688e-05,
+      "loss": 0.3869,
+      "step": 729
+    },
+    {
+      "epoch": 0.8524303021456722,
+      "grad_norm": 0.08585738096831833,
+      "learning_rate": 1.4191228350065079e-05,
+      "loss": 0.3896,
+      "step": 730
+    },
+    {
+      "epoch": 0.8535980148883374,
+      "grad_norm": 0.07823061370043644,
+      "learning_rate": 1.417269792254819e-05,
+      "loss": 0.396,
+      "step": 731
+    },
+    {
+      "epoch": 0.8547657276310028,
+      "grad_norm": 0.07671992248508282,
+      "learning_rate": 1.4154150130018867e-05,
+      "loss": 0.3855,
+      "step": 732
+    },
+    {
+      "epoch": 0.855933440373668,
+      "grad_norm": 0.07657472885483273,
+      "learning_rate": 1.4135585049665207e-05,
+      "loss": 0.3677,
+      "step": 733
+    },
+    {
+      "epoch": 0.8571011531163334,
+      "grad_norm": 0.07494914918728009,
+      "learning_rate": 1.4117002758747268e-05,
+      "loss": 0.3946,
+      "step": 734
+    },
+    {
+      "epoch": 0.8582688658589986,
+      "grad_norm": 0.0763186259678258,
+      "learning_rate": 1.409840333459672e-05,
+      "loss": 0.3669,
+      "step": 735
+    },
+    {
+      "epoch": 0.859436578601664,
+      "grad_norm": 0.08251195962184903,
+      "learning_rate": 1.4079786854616537e-05,
+      "loss": 0.3686,
+      "step": 736
+    },
+    {
+      "epoch": 0.8606042913443293,
+      "grad_norm": 0.07759026018439226,
+      "learning_rate": 1.4061153396280673e-05,
+      "loss": 0.3624,
+      "step": 737
+    },
+    {
+      "epoch": 0.8617720040869946,
+      "grad_norm": 0.07154286984522339,
+      "learning_rate": 1.4042503037133736e-05,
+      "loss": 0.3704,
+      "step": 738
+    },
+    {
+      "epoch": 0.8629397168296599,
+      "grad_norm": 0.08124277727264502,
+      "learning_rate": 1.4023835854790682e-05,
+      "loss": 0.3759,
+      "step": 739
+    },
+    {
+      "epoch": 0.8641074295723252,
+      "grad_norm": 0.07896977403405496,
+      "learning_rate": 1.400515192693645e-05,
+      "loss": 0.3549,
+      "step": 740
+    },
+    {
+      "epoch": 0.8652751423149905,
+      "grad_norm": 0.07812841976272523,
+      "learning_rate": 1.3986451331325697e-05,
+      "loss": 0.3772,
+      "step": 741
+    },
+    {
+      "epoch": 0.8664428550576558,
+      "grad_norm": 0.08397982004643613,
+      "learning_rate": 1.3967734145782425e-05,
+      "loss": 0.3553,
+      "step": 742
+    },
+    {
+      "epoch": 0.8676105678003211,
+      "grad_norm": 0.08280263234648232,
+      "learning_rate": 1.3949000448199686e-05,
+      "loss": 0.3799,
+      "step": 743
+    },
+    {
+      "epoch": 0.8687782805429864,
+      "grad_norm": 0.08702297821442,
+      "learning_rate": 1.3930250316539237e-05,
+      "loss": 0.3628,
+      "step": 744
+    },
+    {
+      "epoch": 0.8699459932856517,
+      "grad_norm": 0.08802802180439938,
+      "learning_rate": 1.391148382883124e-05,
+      "loss": 0.3956,
+      "step": 745
+    },
+    {
+      "epoch": 0.8711137060283171,
+      "grad_norm": 0.07548043589699228,
+      "learning_rate": 1.3892701063173917e-05,
+      "loss": 0.3752,
+      "step": 746
+    },
+    {
+      "epoch": 0.8722814187709823,
+      "grad_norm": 0.07537899556666493,
+      "learning_rate": 1.3873902097733231e-05,
+      "loss": 0.3737,
+      "step": 747
+    },
+    {
+      "epoch": 0.8734491315136477,
+      "grad_norm": 0.08446161353277726,
+      "learning_rate": 1.3855087010742563e-05,
+      "loss": 0.3816,
+      "step": 748
+    },
+    {
+      "epoch": 0.8746168442563129,
+      "grad_norm": 0.07501492418503952,
+      "learning_rate": 1.3836255880502389e-05,
+      "loss": 0.3589,
+      "step": 749
+    },
+    {
+      "epoch": 0.8757845569989783,
+      "grad_norm": 0.07731590291146685,
+      "learning_rate": 1.3817408785379944e-05,
+      "loss": 0.3679,
+      "step": 750
+    },
+    {
+      "epoch": 0.8769522697416435,
+      "grad_norm": 0.08058624430912953,
+      "learning_rate": 1.379854580380891e-05,
+      "loss": 0.3845,
+      "step": 751
+    },
+    {
+      "epoch": 0.8781199824843089,
+      "grad_norm": 0.07864274531647339,
+      "learning_rate": 1.3779667014289067e-05,
+      "loss": 0.3731,
+      "step": 752
+    },
+    {
+      "epoch": 0.8792876952269741,
+      "grad_norm": 0.08003159207111005,
+      "learning_rate": 1.3760772495385998e-05,
+      "loss": 0.361,
+      "step": 753
+    },
+    {
+      "epoch": 0.8804554079696395,
+      "grad_norm": 0.0773208830966861,
+      "learning_rate": 1.3741862325730739e-05,
+      "loss": 0.3745,
+      "step": 754
+    },
+    {
+      "epoch": 0.8816231207123048,
+      "grad_norm": 0.07470570321644134,
+      "learning_rate": 1.3722936584019453e-05,
+      "loss": 0.3499,
+      "step": 755
+    },
+    {
+      "epoch": 0.8827908334549701,
+      "grad_norm": 0.08126476995046047,
+      "learning_rate": 1.3703995349013113e-05,
+      "loss": 0.3728,
+      "step": 756
+    },
+    {
+      "epoch": 0.8839585461976354,
+      "grad_norm": 0.07601013104969213,
+      "learning_rate": 1.3685038699537165e-05,
+      "loss": 0.3407,
+      "step": 757
+    },
+    {
+      "epoch": 0.8851262589403007,
+      "grad_norm": 0.07623370959947391,
+      "learning_rate": 1.3666066714481205e-05,
+      "loss": 0.3603,
+      "step": 758
+    },
+    {
+      "epoch": 0.886293971682966,
+      "grad_norm": 0.07303434999504319,
+      "learning_rate": 1.3647079472798652e-05,
+      "loss": 0.3669,
+      "step": 759
+    },
+    {
+      "epoch": 0.8874616844256313,
+      "grad_norm": 0.07386171392276678,
+      "learning_rate": 1.362807705350641e-05,
+      "loss": 0.3505,
+      "step": 760
+    },
+    {
+      "epoch": 0.8886293971682966,
+      "grad_norm": 0.07307053468230346,
+      "learning_rate": 1.3609059535684557e-05,
+      "loss": 0.3475,
+      "step": 761
+    },
+    {
+      "epoch": 0.889797109910962,
+      "grad_norm": 0.07457663946603955,
+      "learning_rate": 1.3590026998475987e-05,
+      "loss": 0.3577,
+      "step": 762
+    },
+    {
+      "epoch": 0.8909648226536272,
+      "grad_norm": 0.07335790289788151,
+      "learning_rate": 1.3570979521086118e-05,
+      "loss": 0.35,
+      "step": 763
+    },
+    {
+      "epoch": 0.8921325353962926,
+      "grad_norm": 0.08012302507133773,
+      "learning_rate": 1.355191718278253e-05,
+      "loss": 0.3764,
+      "step": 764
+    },
+    {
+      "epoch": 0.8933002481389578,
+      "grad_norm": 0.07876640308899285,
+      "learning_rate": 1.3532840062894654e-05,
+      "loss": 0.3664,
+      "step": 765
+    },
+    {
+      "epoch": 0.8944679608816232,
+      "grad_norm": 0.07962910073328938,
+      "learning_rate": 1.3513748240813429e-05,
+      "loss": 0.3773,
+      "step": 766
+    },
+    {
+      "epoch": 0.8956356736242884,
+      "grad_norm": 0.07196602257302229,
+      "learning_rate": 1.3494641795990986e-05,
+      "loss": 0.358,
+      "step": 767
+    },
+    {
+      "epoch": 0.8968033863669538,
+      "grad_norm": 0.06893304486104167,
+      "learning_rate": 1.3475520807940303e-05,
+      "loss": 0.3619,
+      "step": 768
+    },
+    {
+      "epoch": 0.897971099109619,
+      "grad_norm": 0.08062104772932854,
+      "learning_rate": 1.345638535623489e-05,
+      "loss": 0.3674,
+      "step": 769
+    },
+    {
+      "epoch": 0.8991388118522844,
+      "grad_norm": 0.077168291060146,
+      "learning_rate": 1.3437235520508434e-05,
+      "loss": 0.3641,
+      "step": 770
+    },
+    {
+      "epoch": 0.9003065245949496,
+      "grad_norm": 0.07573594632721393,
+      "learning_rate": 1.3418071380454501e-05,
+      "loss": 0.3595,
+      "step": 771
+    },
+    {
+      "epoch": 0.9014742373376149,
+      "grad_norm": 0.07502859077693709,
+      "learning_rate": 1.3398893015826166e-05,
+      "loss": 0.35,
+      "step": 772
+    },
+    {
+      "epoch": 0.9026419500802803,
+      "grad_norm": 0.08020449313375975,
+      "learning_rate": 1.337970050643572e-05,
+      "loss": 0.4074,
+      "step": 773
+    },
+    {
+      "epoch": 0.9038096628229455,
+      "grad_norm": 0.07699401113801972,
+      "learning_rate": 1.3360493932154301e-05,
+      "loss": 0.3799,
+      "step": 774
+    },
+    {
+      "epoch": 0.9049773755656109,
+      "grad_norm": 0.07809725936061165,
+      "learning_rate": 1.3341273372911591e-05,
+      "loss": 0.3832,
+      "step": 775
+    },
+    {
+      "epoch": 0.9061450883082761,
+      "grad_norm": 0.08181779775755989,
+      "learning_rate": 1.3322038908695466e-05,
+      "loss": 0.3785,
+      "step": 776
+    },
+    {
+      "epoch": 0.9073128010509415,
+      "grad_norm": 0.0746739642498148,
+      "learning_rate": 1.3302790619551673e-05,
+      "loss": 0.3887,
+      "step": 777
+    },
+    {
+      "epoch": 0.9084805137936067,
+      "grad_norm": 0.07532842120639169,
+      "learning_rate": 1.3283528585583484e-05,
+      "loss": 0.3818,
+      "step": 778
+    },
+    {
+      "epoch": 0.9096482265362721,
+      "grad_norm": 0.07597128303593385,
+      "learning_rate": 1.326425288695138e-05,
+      "loss": 0.3778,
+      "step": 779
+    },
+    {
+      "epoch": 0.9108159392789373,
+      "grad_norm": 0.07793089765367432,
+      "learning_rate": 1.3244963603872707e-05,
+      "loss": 0.3734,
+      "step": 780
+    },
+    {
+      "epoch": 0.9119836520216027,
+      "grad_norm": 0.07672989590199356,
+      "learning_rate": 1.3225660816621342e-05,
+      "loss": 0.388,
+      "step": 781
+    },
+    {
+      "epoch": 0.913151364764268,
+      "grad_norm": 0.07660575204600395,
+      "learning_rate": 1.3206344605527355e-05,
+      "loss": 0.3878,
+      "step": 782
+    },
+    {
+      "epoch": 0.9143190775069333,
+      "grad_norm": 0.07519098676692171,
+      "learning_rate": 1.3187015050976696e-05,
+      "loss": 0.3601,
+      "step": 783
+    },
+    {
+      "epoch": 0.9154867902495986,
+      "grad_norm": 0.0769579026836288,
+      "learning_rate": 1.3167672233410826e-05,
+      "loss": 0.3748,
+      "step": 784
+    },
+    {
+      "epoch": 0.9166545029922639,
+      "grad_norm": 0.07241008934501528,
+      "learning_rate": 1.3148316233326414e-05,
+      "loss": 0.368,
+      "step": 785
+    },
+    {
+      "epoch": 0.9178222157349292,
+      "grad_norm": 0.07850763980785358,
+      "learning_rate": 1.312894713127499e-05,
+      "loss": 0.3929,
+      "step": 786
+    },
+    {
+      "epoch": 0.9189899284775945,
+      "grad_norm": 0.07454027444060184,
+      "learning_rate": 1.3109565007862597e-05,
+      "loss": 0.3761,
+      "step": 787
+    },
+    {
+      "epoch": 0.9201576412202598,
+      "grad_norm": 0.07547792136803365,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 0.3717,
+      "step": 788
+    },
+    {
+      "epoch": 0.9213253539629251,
+      "grad_norm": 0.07059274768568762,
+      "learning_rate": 1.3070762019649722e-05,
+      "loss": 0.3631,
+      "step": 789
+    },
+    {
+      "epoch": 0.9224930667055904,
+      "grad_norm": 0.07765894911710981,
+      "learning_rate": 1.3051341316330946e-05,
+      "loss": 0.3665,
+      "step": 790
+    },
+    {
+      "epoch": 0.9236607794482558,
+      "grad_norm": 0.07672385766792943,
+      "learning_rate": 1.303190791461394e-05,
+      "loss": 0.3673,
+      "step": 791
+    },
+    {
+      "epoch": 0.924828492190921,
+      "grad_norm": 0.0750013077731108,
+      "learning_rate": 1.3012461895372343e-05,
+      "loss": 0.3541,
+      "step": 792
+    },
+    {
+      "epoch": 0.9259962049335864,
+      "grad_norm": 0.07956355565619551,
+      "learning_rate": 1.2993003339532304e-05,
+      "loss": 0.3777,
+      "step": 793
+    },
+    {
+      "epoch": 0.9271639176762516,
+      "grad_norm": 0.07638821859675929,
+      "learning_rate": 1.2973532328072138e-05,
+      "loss": 0.3567,
+      "step": 794
+    },
+    {
+      "epoch": 0.928331630418917,
+      "grad_norm": 0.07563338702930678,
+      "learning_rate": 1.2954048942022002e-05,
+      "loss": 0.348,
+      "step": 795
+    },
+    {
+      "epoch": 0.9294993431615822,
+      "grad_norm": 0.08005002737595618,
+      "learning_rate": 1.2934553262463548e-05,
+      "loss": 0.4047,
+      "step": 796
+    },
+    {
+      "epoch": 0.9306670559042476,
+      "grad_norm": 0.07480335954140425,
+      "learning_rate": 1.2915045370529586e-05,
+      "loss": 0.3589,
+      "step": 797
+    },
+    {
+      "epoch": 0.9318347686469128,
+      "grad_norm": 0.07751617447951605,
+      "learning_rate": 1.2895525347403757e-05,
+      "loss": 0.3768,
+      "step": 798
+    },
+    {
+      "epoch": 0.9330024813895782,
+      "grad_norm": 0.07610488615087871,
+      "learning_rate": 1.2875993274320173e-05,
+      "loss": 0.3585,
+      "step": 799
+    },
+    {
+      "epoch": 0.9341701941322434,
+      "grad_norm": 0.0755755427637437,
+      "learning_rate": 1.285644923256311e-05,
+      "loss": 0.3989,
+      "step": 800
+    },
+    {
+      "epoch": 0.9353379068749088,
+      "grad_norm": 0.08006814867141315,
+      "learning_rate": 1.2836893303466636e-05,
+      "loss": 0.4004,
+      "step": 801
+    },
+    {
+      "epoch": 0.9365056196175741,
+      "grad_norm": 0.0777672752960404,
+      "learning_rate": 1.2817325568414299e-05,
+      "loss": 0.3946,
+      "step": 802
+    },
+    {
+      "epoch": 0.9376733323602394,
+      "grad_norm": 0.07603554460878112,
+      "learning_rate": 1.2797746108838777e-05,
+      "loss": 0.375,
+      "step": 803
+    },
+    {
+      "epoch": 0.9388410451029047,
+      "grad_norm": 0.07542537807935788,
+      "learning_rate": 1.277815500622154e-05,
+      "loss": 0.3799,
+      "step": 804
+    },
+    {
+      "epoch": 0.94000875784557,
+      "grad_norm": 0.07581824232729469,
+      "learning_rate": 1.275855234209251e-05,
+      "loss": 0.386,
+      "step": 805
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.07425375663101144,
+      "learning_rate": 1.2738938198029725e-05,
+      "loss": 0.3549,
+      "step": 806
+    },
+    {
+      "epoch": 0.9423441833309006,
+      "grad_norm": 0.07769811844456231,
+      "learning_rate": 1.2719312655658995e-05,
+      "loss": 0.36,
+      "step": 807
+    },
+    {
+      "epoch": 0.9435118960735659,
+      "grad_norm": 0.0710275309761252,
+      "learning_rate": 1.269967579665357e-05,
+      "loss": 0.3649,
+      "step": 808
+    },
+    {
+      "epoch": 0.9446796088162313,
+      "grad_norm": 0.08211239218248492,
+      "learning_rate": 1.2680027702733791e-05,
+      "loss": 0.3761,
+      "step": 809
+    },
+    {
+      "epoch": 0.9458473215588965,
+      "grad_norm": 0.07617398323333258,
+      "learning_rate": 1.2660368455666752e-05,
+      "loss": 0.3654,
+      "step": 810
+    },
+    {
+      "epoch": 0.9470150343015619,
+      "grad_norm": 0.07642803727076203,
+      "learning_rate": 1.2640698137265967e-05,
+      "loss": 0.3809,
+      "step": 811
+    },
+    {
+      "epoch": 0.9481827470442271,
+      "grad_norm": 0.08172859274179377,
+      "learning_rate": 1.2621016829391022e-05,
+      "loss": 0.3915,
+      "step": 812
+    },
+    {
+      "epoch": 0.9493504597868925,
+      "grad_norm": 0.08052988638111096,
+      "learning_rate": 1.2601324613947239e-05,
+      "loss": 0.3706,
+      "step": 813
+    },
+    {
+      "epoch": 0.9505181725295577,
+      "grad_norm": 0.07970986534297027,
+      "learning_rate": 1.2581621572885324e-05,
+      "loss": 0.3898,
+      "step": 814
+    },
+    {
+      "epoch": 0.9516858852722231,
+      "grad_norm": 0.07669640563141726,
+      "learning_rate": 1.2561907788201042e-05,
+      "loss": 0.386,
+      "step": 815
+    },
+    {
+      "epoch": 0.9528535980148883,
+      "grad_norm": 0.07806075534105351,
+      "learning_rate": 1.2542183341934873e-05,
+      "loss": 0.3652,
+      "step": 816
+    },
+    {
+      "epoch": 0.9540213107575536,
+      "grad_norm": 0.07864711749442883,
+      "learning_rate": 1.2522448316171651e-05,
+      "loss": 0.3616,
+      "step": 817
+    },
+    {
+      "epoch": 0.955189023500219,
+      "grad_norm": 0.0732367666618531,
+      "learning_rate": 1.2502702793040257e-05,
+      "loss": 0.3472,
+      "step": 818
+    },
+    {
+      "epoch": 0.9563567362428842,
+      "grad_norm": 0.07758023757916682,
+      "learning_rate": 1.2482946854713238e-05,
+      "loss": 0.3756,
+      "step": 819
+    },
+    {
+      "epoch": 0.9575244489855496,
+      "grad_norm": 0.07831879373041077,
+      "learning_rate": 1.24631805834065e-05,
+      "loss": 0.3688,
+      "step": 820
+    },
+    {
+      "epoch": 0.9586921617282148,
+      "grad_norm": 0.0764586847551535,
+      "learning_rate": 1.2443404061378941e-05,
+      "loss": 0.3541,
+      "step": 821
+    },
+    {
+      "epoch": 0.9598598744708802,
+      "grad_norm": 0.07581819469490918,
+      "learning_rate": 1.2423617370932127e-05,
+      "loss": 0.3802,
+      "step": 822
+    },
+    {
+      "epoch": 0.9610275872135454,
+      "grad_norm": 0.08540929086484622,
+      "learning_rate": 1.2403820594409926e-05,
+      "loss": 0.3683,
+      "step": 823
+    },
+    {
+      "epoch": 0.9621952999562108,
+      "grad_norm": 0.06991002784860403,
+      "learning_rate": 1.2384013814198197e-05,
+      "loss": 0.3476,
+      "step": 824
+    },
+    {
+      "epoch": 0.963363012698876,
+      "grad_norm": 0.07638143994262912,
+      "learning_rate": 1.2364197112724421e-05,
+      "loss": 0.3781,
+      "step": 825
+    },
+    {
+      "epoch": 0.9645307254415414,
+      "grad_norm": 0.07280987255521489,
+      "learning_rate": 1.2344370572457368e-05,
+      "loss": 0.3412,
+      "step": 826
+    },
+    {
+      "epoch": 0.9656984381842066,
+      "grad_norm": 0.07761668175987112,
+      "learning_rate": 1.2324534275906747e-05,
+      "loss": 0.3638,
+      "step": 827
+    },
+    {
+      "epoch": 0.966866150926872,
+      "grad_norm": 0.07780034755564047,
+      "learning_rate": 1.2304688305622889e-05,
+      "loss": 0.3615,
+      "step": 828
+    },
+    {
+      "epoch": 0.9680338636695373,
+      "grad_norm": 0.08417221260673953,
+      "learning_rate": 1.2284832744196355e-05,
+      "loss": 0.3592,
+      "step": 829
+    },
+    {
+      "epoch": 0.9692015764122026,
+      "grad_norm": 0.07462167278011304,
+      "learning_rate": 1.2264967674257647e-05,
+      "loss": 0.3716,
+      "step": 830
+    },
+    {
+      "epoch": 0.9703692891548679,
+      "grad_norm": 0.08246158041273449,
+      "learning_rate": 1.2245093178476813e-05,
+      "loss": 0.3751,
+      "step": 831
+    },
+    {
+      "epoch": 0.9715370018975332,
+      "grad_norm": 0.08287184703438338,
+      "learning_rate": 1.2225209339563144e-05,
+      "loss": 0.3727,
+      "step": 832
+    },
+    {
+      "epoch": 0.9727047146401985,
+      "grad_norm": 0.07182430772999325,
+      "learning_rate": 1.2205316240264812e-05,
+      "loss": 0.3645,
+      "step": 833
+    },
+    {
+      "epoch": 0.9738724273828638,
+      "grad_norm": 0.07520968983400671,
+      "learning_rate": 1.218541396336852e-05,
+      "loss": 0.3771,
+      "step": 834
+    },
+    {
+      "epoch": 0.9750401401255291,
+      "grad_norm": 0.08020977315378439,
+      "learning_rate": 1.2165502591699166e-05,
+      "loss": 0.3653,
+      "step": 835
+    },
+    {
+      "epoch": 0.9762078528681944,
+      "grad_norm": 0.07392786418005536,
+      "learning_rate": 1.2145582208119497e-05,
+      "loss": 0.3597,
+      "step": 836
+    },
+    {
+      "epoch": 0.9773755656108597,
+      "grad_norm": 0.07994435491085636,
+      "learning_rate": 1.2125652895529766e-05,
+      "loss": 0.3718,
+      "step": 837
+    },
+    {
+      "epoch": 0.9785432783535251,
+      "grad_norm": 0.08015957616561484,
+      "learning_rate": 1.2105714736867391e-05,
+      "loss": 0.3717,
+      "step": 838
+    },
+    {
+      "epoch": 0.9797109910961903,
+      "grad_norm": 0.0748741865772379,
+      "learning_rate": 1.2085767815106583e-05,
+      "loss": 0.3692,
+      "step": 839
+    },
+    {
+      "epoch": 0.9808787038388557,
+      "grad_norm": 0.07684827334796537,
+      "learning_rate": 1.2065812213258051e-05,
+      "loss": 0.3548,
+      "step": 840
+    },
+    {
+      "epoch": 0.9820464165815209,
+      "grad_norm": 0.07599453758843067,
+      "learning_rate": 1.2045848014368591e-05,
+      "loss": 0.388,
+      "step": 841
+    },
+    {
+      "epoch": 0.9832141293241863,
+      "grad_norm": 0.07491162332442328,
+      "learning_rate": 1.2025875301520811e-05,
+      "loss": 0.3655,
+      "step": 842
+    },
+    {
+      "epoch": 0.9843818420668515,
+      "grad_norm": 0.07780363402310586,
+      "learning_rate": 1.200589415783273e-05,
+      "loss": 0.3722,
+      "step": 843
+    },
+    {
+      "epoch": 0.9855495548095169,
+      "grad_norm": 0.0809281160483852,
+      "learning_rate": 1.1985904666457455e-05,
+      "loss": 0.3675,
+      "step": 844
+    },
+    {
+      "epoch": 0.9867172675521821,
+      "grad_norm": 0.07420257695367186,
+      "learning_rate": 1.196590691058284e-05,
+      "loss": 0.3553,
+      "step": 845
+    },
+    {
+      "epoch": 0.9878849802948475,
+      "grad_norm": 0.07941911344791804,
+      "learning_rate": 1.1945900973431127e-05,
+      "loss": 0.372,
+      "step": 846
+    },
+    {
+      "epoch": 0.9890526930375128,
+      "grad_norm": 0.07768966407764083,
+      "learning_rate": 1.1925886938258606e-05,
+      "loss": 0.375,
+      "step": 847
+    },
+    {
+      "epoch": 0.9902204057801781,
+      "grad_norm": 0.07047316650368235,
+      "learning_rate": 1.1905864888355264e-05,
+      "loss": 0.3503,
+      "step": 848
+    },
+    {
+      "epoch": 0.9913881185228434,
+      "grad_norm": 0.08274505901332616,
+      "learning_rate": 1.1885834907044444e-05,
+      "loss": 0.364,
+      "step": 849
+    },
+    {
+      "epoch": 0.9925558312655087,
+      "grad_norm": 0.07956441073279254,
+      "learning_rate": 1.1865797077682509e-05,
+      "loss": 0.3701,
+      "step": 850
+    },
+    {
+      "epoch": 0.993723544008174,
+      "grad_norm": 0.07739316194164023,
+      "learning_rate": 1.1845751483658454e-05,
+      "loss": 0.3616,
+      "step": 851
+    },
+    {
+      "epoch": 0.9948912567508393,
+      "grad_norm": 0.08409801662073191,
+      "learning_rate": 1.182569820839362e-05,
+      "loss": 0.3608,
+      "step": 852
+    },
+    {
+      "epoch": 0.9960589694935046,
+      "grad_norm": 0.07177392897066465,
+      "learning_rate": 1.1805637335341285e-05,
+      "loss": 0.351,
+      "step": 853
+    },
+    {
+      "epoch": 0.99722668223617,
+      "grad_norm": 0.07596539110028662,
+      "learning_rate": 1.1785568947986368e-05,
+      "loss": 0.3658,
+      "step": 854
+    },
+    {
+      "epoch": 0.9983943949788352,
+      "grad_norm": 0.07989069309022584,
+      "learning_rate": 1.1765493129845047e-05,
+      "loss": 0.3542,
+      "step": 855
+    },
+    {
+      "epoch": 0.9995621077215006,
+      "grad_norm": 0.07654122031638624,
+      "learning_rate": 1.1745409964464425e-05,
+      "loss": 0.3602,
+      "step": 856
+    },
+    {
+      "epoch": 1.0008757845569989,
+      "grad_norm": 0.16517292142897697,
+      "learning_rate": 1.1725319535422189e-05,
+      "loss": 0.5973,
+      "step": 857
+    },
+    {
+      "epoch": 1.0020434972996644,
+      "grad_norm": 0.09343084156558717,
+      "learning_rate": 1.170522192632624e-05,
+      "loss": 0.2939,
+      "step": 858
+    },
+    {
+      "epoch": 1.0032112100423296,
+      "grad_norm": 0.09039991848384979,
+      "learning_rate": 1.1685117220814373e-05,
+      "loss": 0.274,
+      "step": 859
+    },
+    {
+      "epoch": 1.0043789227849949,
+      "grad_norm": 0.08399901999828686,
+      "learning_rate": 1.1665005502553912e-05,
+      "loss": 0.2849,
+      "step": 860
+    },
+    {
+      "epoch": 1.00554663552766,
+      "grad_norm": 0.0873128560561207,
+      "learning_rate": 1.164488685524136e-05,
+      "loss": 0.2833,
+      "step": 861
+    },
+    {
+      "epoch": 1.0067143482703256,
+      "grad_norm": 0.09480591388656644,
+      "learning_rate": 1.1624761362602061e-05,
+      "loss": 0.2795,
+      "step": 862
+    },
+    {
+      "epoch": 1.0078820610129908,
+      "grad_norm": 0.0891393021865796,
+      "learning_rate": 1.160462910838984e-05,
+      "loss": 0.2816,
+      "step": 863
+    },
+    {
+      "epoch": 1.009049773755656,
+      "grad_norm": 0.09944336070918125,
+      "learning_rate": 1.1584490176386671e-05,
+      "loss": 0.298,
+      "step": 864
+    },
+    {
+      "epoch": 1.0102174864983213,
+      "grad_norm": 0.0822750183005768,
+      "learning_rate": 1.156434465040231e-05,
+      "loss": 0.2814,
+      "step": 865
+    },
+    {
+      "epoch": 1.0113851992409868,
+      "grad_norm": 0.08569179133843166,
+      "learning_rate": 1.1544192614273957e-05,
+      "loss": 0.2767,
+      "step": 866
+    },
+    {
+      "epoch": 1.012552911983652,
+      "grad_norm": 0.08438763669781942,
+      "learning_rate": 1.1524034151865904e-05,
+      "loss": 0.2797,
+      "step": 867
+    },
+    {
+      "epoch": 1.0137206247263173,
+      "grad_norm": 0.08905719178603445,
+      "learning_rate": 1.1503869347069186e-05,
+      "loss": 0.2927,
+      "step": 868
+    },
+    {
+      "epoch": 1.0148883374689825,
+      "grad_norm": 0.08232358044206647,
+      "learning_rate": 1.1483698283801233e-05,
+      "loss": 0.2938,
+      "step": 869
+    },
+    {
+      "epoch": 1.016056050211648,
+      "grad_norm": 0.08043952046008332,
+      "learning_rate": 1.1463521046005524e-05,
+      "loss": 0.295,
+      "step": 870
+    },
+    {
+      "epoch": 1.0172237629543133,
+      "grad_norm": 0.08336126364750034,
+      "learning_rate": 1.1443337717651221e-05,
+      "loss": 0.2748,
+      "step": 871
+    },
+    {
+      "epoch": 1.0183914756969785,
+      "grad_norm": 0.08277262922119626,
+      "learning_rate": 1.1423148382732854e-05,
+      "loss": 0.2584,
+      "step": 872
+    },
+    {
+      "epoch": 1.0195591884396438,
+      "grad_norm": 0.08413849107243834,
+      "learning_rate": 1.1402953125269927e-05,
+      "loss": 0.2861,
+      "step": 873
+    },
+    {
+      "epoch": 1.0207269011823092,
+      "grad_norm": 0.08614464414986499,
+      "learning_rate": 1.1382752029306606e-05,
+      "loss": 0.2855,
+      "step": 874
+    },
+    {
+      "epoch": 1.0218946139249745,
+      "grad_norm": 0.09137694790853189,
+      "learning_rate": 1.1362545178911346e-05,
+      "loss": 0.3029,
+      "step": 875
+    },
+    {
+      "epoch": 1.0230623266676397,
+      "grad_norm": 0.07392806680383539,
+      "learning_rate": 1.1342332658176556e-05,
+      "loss": 0.2481,
+      "step": 876
+    },
+    {
+      "epoch": 1.024230039410305,
+      "grad_norm": 0.08337217738228136,
+      "learning_rate": 1.132211455121824e-05,
+      "loss": 0.2885,
+      "step": 877
+    },
+    {
+      "epoch": 1.0253977521529705,
+      "grad_norm": 0.08463729229383211,
+      "learning_rate": 1.1301890942175649e-05,
+      "loss": 0.2963,
+      "step": 878
+    },
+    {
+      "epoch": 1.0265654648956357,
+      "grad_norm": 0.08040166875151934,
+      "learning_rate": 1.1281661915210931e-05,
+      "loss": 0.2933,
+      "step": 879
+    },
+    {
+      "epoch": 1.027733177638301,
+      "grad_norm": 0.09621760243167839,
+      "learning_rate": 1.126142755450878e-05,
+      "loss": 0.2761,
+      "step": 880
+    },
+    {
+      "epoch": 1.0289008903809662,
+      "grad_norm": 0.08515421201685731,
+      "learning_rate": 1.1241187944276088e-05,
+      "loss": 0.2849,
+      "step": 881
+    },
+    {
+      "epoch": 1.0300686031236317,
+      "grad_norm": 0.07488236099251518,
+      "learning_rate": 1.12209431687416e-05,
+      "loss": 0.2598,
+      "step": 882
+    },
+    {
+      "epoch": 1.031236315866297,
+      "grad_norm": 0.08694858370764072,
+      "learning_rate": 1.1200693312155545e-05,
+      "loss": 0.2967,
+      "step": 883
+    },
+    {
+      "epoch": 1.0324040286089622,
+      "grad_norm": 0.08036911231299018,
+      "learning_rate": 1.1180438458789305e-05,
+      "loss": 0.2784,
+      "step": 884
+    },
+    {
+      "epoch": 1.0335717413516274,
+      "grad_norm": 0.07869950005957105,
+      "learning_rate": 1.1160178692935054e-05,
+      "loss": 0.2907,
+      "step": 885
+    },
+    {
+      "epoch": 1.034739454094293,
+      "grad_norm": 0.08344238990977787,
+      "learning_rate": 1.1139914098905406e-05,
+      "loss": 0.2674,
+      "step": 886
+    },
+    {
+      "epoch": 1.0359071668369582,
+      "grad_norm": 0.08865145954606783,
+      "learning_rate": 1.1119644761033079e-05,
+      "loss": 0.3018,
+      "step": 887
+    },
+    {
+      "epoch": 1.0370748795796234,
+      "grad_norm": 0.08273819882632716,
+      "learning_rate": 1.1099370763670523e-05,
+      "loss": 0.3037,
+      "step": 888
+    },
+    {
+      "epoch": 1.0382425923222887,
+      "grad_norm": 0.07819121943619413,
+      "learning_rate": 1.107909219118958e-05,
+      "loss": 0.2802,
+      "step": 889
+    },
+    {
+      "epoch": 1.039410305064954,
+      "grad_norm": 0.08013520772620775,
+      "learning_rate": 1.1058809127981133e-05,
+      "loss": 0.2802,
+      "step": 890
+    },
+    {
+      "epoch": 1.0405780178076194,
+      "grad_norm": 0.0914781916112486,
+      "learning_rate": 1.1038521658454753e-05,
+      "loss": 0.3015,
+      "step": 891
+    },
+    {
+      "epoch": 1.0417457305502846,
+      "grad_norm": 0.08153694933106341,
+      "learning_rate": 1.1018229867038358e-05,
+      "loss": 0.2689,
+      "step": 892
+    },
+    {
+      "epoch": 1.0429134432929499,
+      "grad_norm": 0.08015258385134942,
+      "learning_rate": 1.0997933838177828e-05,
+      "loss": 0.2886,
+      "step": 893
+    },
+    {
+      "epoch": 1.0440811560356151,
+      "grad_norm": 0.08198038969460766,
+      "learning_rate": 1.0977633656336706e-05,
+      "loss": 0.2801,
+      "step": 894
+    },
+    {
+      "epoch": 1.0452488687782806,
+      "grad_norm": 0.0760255192710916,
+      "learning_rate": 1.0957329405995794e-05,
+      "loss": 0.2563,
+      "step": 895
+    },
+    {
+      "epoch": 1.0464165815209459,
+      "grad_norm": 0.08135222629376372,
+      "learning_rate": 1.0937021171652842e-05,
+      "loss": 0.2875,
+      "step": 896
+    },
+    {
+      "epoch": 1.047584294263611,
+      "grad_norm": 0.07834618259453457,
+      "learning_rate": 1.0916709037822172e-05,
+      "loss": 0.2591,
+      "step": 897
+    },
+    {
+      "epoch": 1.0487520070062764,
+      "grad_norm": 0.0799749320000738,
+      "learning_rate": 1.0896393089034336e-05,
+      "loss": 0.277,
+      "step": 898
+    },
+    {
+      "epoch": 1.0499197197489418,
+      "grad_norm": 0.07858453723602436,
+      "learning_rate": 1.0876073409835762e-05,
+      "loss": 0.2674,
+      "step": 899
+    },
+    {
+      "epoch": 1.051087432491607,
+      "grad_norm": 0.08475275233217454,
+      "learning_rate": 1.08557500847884e-05,
+      "loss": 0.2775,
+      "step": 900
+    },
+    {
+      "epoch": 1.0522551452342723,
+      "grad_norm": 0.07897523616077559,
+      "learning_rate": 1.0835423198469373e-05,
+      "loss": 0.2681,
+      "step": 901
+    },
+    {
+      "epoch": 1.0534228579769376,
+      "grad_norm": 0.08370398238528254,
+      "learning_rate": 1.0815092835470633e-05,
+      "loss": 0.2948,
+      "step": 902
+    },
+    {
+      "epoch": 1.054590570719603,
+      "grad_norm": 0.08179263330417788,
+      "learning_rate": 1.0794759080398587e-05,
+      "loss": 0.2846,
+      "step": 903
+    },
+    {
+      "epoch": 1.0557582834622683,
+      "grad_norm": 0.08333176727688445,
+      "learning_rate": 1.077442201787377e-05,
+      "loss": 0.2688,
+      "step": 904
+    },
+    {
+      "epoch": 1.0569259962049335,
+      "grad_norm": 0.07736412053451988,
+      "learning_rate": 1.0754081732530473e-05,
+      "loss": 0.2791,
+      "step": 905
+    },
+    {
+      "epoch": 1.0580937089475988,
+      "grad_norm": 0.07705885965192447,
+      "learning_rate": 1.0733738309016402e-05,
+      "loss": 0.2663,
+      "step": 906
+    },
+    {
+      "epoch": 1.0592614216902643,
+      "grad_norm": 0.07470827491174549,
+      "learning_rate": 1.0713391831992324e-05,
+      "loss": 0.2715,
+      "step": 907
+    },
+    {
+      "epoch": 1.0604291344329295,
+      "grad_norm": 0.08134758306297472,
+      "learning_rate": 1.0693042386131713e-05,
+      "loss": 0.2828,
+      "step": 908
+    },
+    {
+      "epoch": 1.0615968471755948,
+      "grad_norm": 0.0828966336199791,
+      "learning_rate": 1.0672690056120398e-05,
+      "loss": 0.2867,
+      "step": 909
+    },
+    {
+      "epoch": 1.06276455991826,
+      "grad_norm": 0.07633044462336469,
+      "learning_rate": 1.0652334926656209e-05,
+      "loss": 0.2561,
+      "step": 910
+    },
+    {
+      "epoch": 1.0639322726609255,
+      "grad_norm": 0.08581370227706839,
+      "learning_rate": 1.0631977082448626e-05,
+      "loss": 0.296,
+      "step": 911
+    },
+    {
+      "epoch": 1.0650999854035907,
+      "grad_norm": 0.07833028260132129,
+      "learning_rate": 1.0611616608218429e-05,
+      "loss": 0.2682,
+      "step": 912
+    },
+    {
+      "epoch": 1.066267698146256,
+      "grad_norm": 0.07911123754484767,
+      "learning_rate": 1.059125358869734e-05,
+      "loss": 0.2642,
+      "step": 913
+    },
+    {
+      "epoch": 1.0674354108889212,
+      "grad_norm": 0.0832393765513744,
+      "learning_rate": 1.0570888108627682e-05,
+      "loss": 0.2876,
+      "step": 914
+    },
+    {
+      "epoch": 1.0686031236315867,
+      "grad_norm": 0.08221309774220456,
+      "learning_rate": 1.0550520252762002e-05,
+      "loss": 0.2953,
+      "step": 915
+    },
+    {
+      "epoch": 1.069770836374252,
+      "grad_norm": 0.0829068310854862,
+      "learning_rate": 1.0530150105862748e-05,
+      "loss": 0.2986,
+      "step": 916
+    },
+    {
+      "epoch": 1.0709385491169172,
+      "grad_norm": 0.08225001079249061,
+      "learning_rate": 1.05097777527019e-05,
+      "loss": 0.2777,
+      "step": 917
+    },
+    {
+      "epoch": 1.0721062618595825,
+      "grad_norm": 0.080958559988261,
+      "learning_rate": 1.0489403278060613e-05,
+      "loss": 0.2587,
+      "step": 918
+    },
+    {
+      "epoch": 1.073273974602248,
+      "grad_norm": 0.07799433034081676,
+      "learning_rate": 1.0469026766728876e-05,
+      "loss": 0.2799,
+      "step": 919
+    },
+    {
+      "epoch": 1.0744416873449132,
+      "grad_norm": 0.08399749129184007,
+      "learning_rate": 1.044864830350515e-05,
+      "loss": 0.269,
+      "step": 920
+    },
+    {
+      "epoch": 1.0756094000875784,
+      "grad_norm": 0.07674405414387737,
+      "learning_rate": 1.0428267973196027e-05,
+      "loss": 0.2633,
+      "step": 921
+    },
+    {
+      "epoch": 1.0767771128302437,
+      "grad_norm": 0.08140683349217916,
+      "learning_rate": 1.0407885860615859e-05,
+      "loss": 0.3014,
+      "step": 922
+    },
+    {
+      "epoch": 1.0779448255729092,
+      "grad_norm": 0.08283032180442632,
+      "learning_rate": 1.0387502050586418e-05,
+      "loss": 0.2737,
+      "step": 923
+    },
+    {
+      "epoch": 1.0791125383155744,
+      "grad_norm": 0.07723300906216603,
+      "learning_rate": 1.0367116627936549e-05,
+      "loss": 0.2765,
+      "step": 924
+    },
+    {
+      "epoch": 1.0802802510582397,
+      "grad_norm": 0.08252304557766704,
+      "learning_rate": 1.0346729677501793e-05,
+      "loss": 0.2844,
+      "step": 925
+    },
+    {
+      "epoch": 1.081447963800905,
+      "grad_norm": 0.08058927953103147,
+      "learning_rate": 1.0326341284124061e-05,
+      "loss": 0.2811,
+      "step": 926
+    },
+    {
+      "epoch": 1.0826156765435704,
+      "grad_norm": 0.08139300603015542,
+      "learning_rate": 1.0305951532651262e-05,
+      "loss": 0.2785,
+      "step": 927
+    },
+    {
+      "epoch": 1.0837833892862356,
+      "grad_norm": 0.07775859275047513,
+      "learning_rate": 1.0285560507936962e-05,
+      "loss": 0.2729,
+      "step": 928
+    },
+    {
+      "epoch": 1.0849511020289009,
+      "grad_norm": 0.08438260314212347,
+      "learning_rate": 1.0265168294840025e-05,
+      "loss": 0.2841,
+      "step": 929
+    },
+    {
+      "epoch": 1.0861188147715661,
+      "grad_norm": 0.07823354755817875,
+      "learning_rate": 1.0244774978224254e-05,
+      "loss": 0.272,
+      "step": 930
+    },
+    {
+      "epoch": 1.0872865275142316,
+      "grad_norm": 0.08295812549893493,
+      "learning_rate": 1.0224380642958052e-05,
+      "loss": 0.2722,
+      "step": 931
+    },
+    {
+      "epoch": 1.0884542402568969,
+      "grad_norm": 0.09056460745414549,
+      "learning_rate": 1.0203985373914056e-05,
+      "loss": 0.2913,
+      "step": 932
+    },
+    {
+      "epoch": 1.089621952999562,
+      "grad_norm": 0.07727197332583013,
+      "learning_rate": 1.018358925596879e-05,
+      "loss": 0.2727,
+      "step": 933
+    },
+    {
+      "epoch": 1.0907896657422274,
+      "grad_norm": 0.07970485016869089,
+      "learning_rate": 1.016319237400232e-05,
+      "loss": 0.2812,
+      "step": 934
+    },
+    {
+      "epoch": 1.0919573784848926,
+      "grad_norm": 0.07795805426338806,
+      "learning_rate": 1.0142794812897874e-05,
+      "loss": 0.2681,
+      "step": 935
+    },
+    {
+      "epoch": 1.093125091227558,
+      "grad_norm": 0.08313968843654403,
+      "learning_rate": 1.0122396657541522e-05,
+      "loss": 0.2673,
+      "step": 936
+    },
+    {
+      "epoch": 1.0942928039702233,
+      "grad_norm": 0.08406632728297002,
+      "learning_rate": 1.0101997992821798e-05,
+      "loss": 0.2813,
+      "step": 937
+    },
+    {
+      "epoch": 1.0954605167128886,
+      "grad_norm": 0.08349429341386375,
+      "learning_rate": 1.0081598903629359e-05,
+      "loss": 0.2839,
+      "step": 938
+    },
+    {
+      "epoch": 1.096628229455554,
+      "grad_norm": 0.08198872459314788,
+      "learning_rate": 1.0061199474856628e-05,
+      "loss": 0.2691,
+      "step": 939
+    },
+    {
+      "epoch": 1.0977959421982193,
+      "grad_norm": 0.08210462318157938,
+      "learning_rate": 1.0040799791397444e-05,
+      "loss": 0.2945,
+      "step": 940
+    },
+    {
+      "epoch": 1.0989636549408845,
+      "grad_norm": 0.08172153178017141,
+      "learning_rate": 1.0020399938146701e-05,
+      "loss": 0.2938,
+      "step": 941
+    },
+    {
+      "epoch": 1.1001313676835498,
+      "grad_norm": 0.09562504871471336,
+      "learning_rate": 1e-05,
+      "loss": 0.2774,
+      "step": 942
+    },
+    {
+      "epoch": 1.101299080426215,
+      "grad_norm": 0.07950210914849173,
+      "learning_rate": 9.979600061853302e-06,
+      "loss": 0.2664,
+      "step": 943
+    },
+    {
+      "epoch": 1.1024667931688805,
+      "grad_norm": 0.0784110405335003,
+      "learning_rate": 9.95920020860256e-06,
+      "loss": 0.268,
+      "step": 944
+    },
+    {
+      "epoch": 1.1036345059115458,
+      "grad_norm": 0.08071004199160792,
+      "learning_rate": 9.938800525143374e-06,
+      "loss": 0.2683,
+      "step": 945
+    },
+    {
+      "epoch": 1.104802218654211,
+      "grad_norm": 0.08443974893231895,
+      "learning_rate": 9.918401096370644e-06,
+      "loss": 0.2835,
+      "step": 946
+    },
+    {
+      "epoch": 1.1059699313968763,
+      "grad_norm": 0.08072267048755444,
+      "learning_rate": 9.898002007178204e-06,
+      "loss": 0.2579,
+      "step": 947
+    },
+    {
+      "epoch": 1.1071376441395417,
+      "grad_norm": 0.08259958207390311,
+      "learning_rate": 9.877603342458483e-06,
+      "loss": 0.2729,
+      "step": 948
+    },
+    {
+      "epoch": 1.108305356882207,
+      "grad_norm": 0.08109987333700891,
+      "learning_rate": 9.85720518710213e-06,
+      "loss": 0.2782,
+      "step": 949
+    },
+    {
+      "epoch": 1.1094730696248722,
+      "grad_norm": 0.08516207120472324,
+      "learning_rate": 9.836807625997683e-06,
+      "loss": 0.2735,
+      "step": 950
+    },
+    {
+      "epoch": 1.1106407823675375,
+      "grad_norm": 0.08208543580647606,
+      "learning_rate": 9.81641074403121e-06,
+      "loss": 0.2697,
+      "step": 951
+    },
+    {
+      "epoch": 1.111808495110203,
+      "grad_norm": 0.0807460578597101,
+      "learning_rate": 9.79601462608595e-06,
+      "loss": 0.309,
+      "step": 952
+    },
+    {
+      "epoch": 1.1129762078528682,
+      "grad_norm": 0.08569538212210046,
+      "learning_rate": 9.775619357041952e-06,
+      "loss": 0.2859,
+      "step": 953
+    },
+    {
+      "epoch": 1.1141439205955335,
+      "grad_norm": 0.0815669062110687,
+      "learning_rate": 9.75522502177575e-06,
+      "loss": 0.2784,
+      "step": 954
+    },
+    {
+      "epoch": 1.1153116333381987,
+      "grad_norm": 0.08206633688349131,
+      "learning_rate": 9.734831705159975e-06,
+      "loss": 0.271,
+      "step": 955
+    },
+    {
+      "epoch": 1.1164793460808642,
+      "grad_norm": 0.08342378646595387,
+      "learning_rate": 9.71443949206304e-06,
+      "loss": 0.3006,
+      "step": 956
+    },
+    {
+      "epoch": 1.1176470588235294,
+      "grad_norm": 0.07593773161166709,
+      "learning_rate": 9.69404846734874e-06,
+      "loss": 0.2783,
+      "step": 957
+    },
+    {
+      "epoch": 1.1188147715661947,
+      "grad_norm": 0.07693777256476621,
+      "learning_rate": 9.67365871587594e-06,
+      "loss": 0.2594,
+      "step": 958
+    },
+    {
+      "epoch": 1.11998248430886,
+      "grad_norm": 0.08075550402319591,
+      "learning_rate": 9.653270322498209e-06,
+      "loss": 0.2811,
+      "step": 959
+    },
+    {
+      "epoch": 1.1211501970515254,
+      "grad_norm": 0.08039914128155738,
+      "learning_rate": 9.632883372063458e-06,
+      "loss": 0.2732,
+      "step": 960
+    },
+    {
+      "epoch": 1.1223179097941907,
+      "grad_norm": 0.07437019721228857,
+      "learning_rate": 9.612497949413586e-06,
+      "loss": 0.2689,
+      "step": 961
+    },
+    {
+      "epoch": 1.123485622536856,
+      "grad_norm": 0.08481936247950216,
+      "learning_rate": 9.592114139384146e-06,
+      "loss": 0.2793,
+      "step": 962
+    },
+    {
+      "epoch": 1.1246533352795212,
+      "grad_norm": 0.08228250496212267,
+      "learning_rate": 9.571732026803978e-06,
+      "loss": 0.2892,
+      "step": 963
+    },
+    {
+      "epoch": 1.1258210480221866,
+      "grad_norm": 0.07900994146648677,
+      "learning_rate": 9.551351696494854e-06,
+      "loss": 0.274,
+      "step": 964
+    },
+    {
+      "epoch": 1.1269887607648519,
+      "grad_norm": 0.07838416628938394,
+      "learning_rate": 9.530973233271127e-06,
+      "loss": 0.2627,
+      "step": 965
+    },
+    {
+      "epoch": 1.1281564735075171,
+      "grad_norm": 0.08113935875804923,
+      "learning_rate": 9.51059672193939e-06,
+      "loss": 0.2717,
+      "step": 966
+    },
+    {
+      "epoch": 1.1293241862501824,
+      "grad_norm": 0.08120939933012636,
+      "learning_rate": 9.4902222472981e-06,
+      "loss": 0.2801,
+      "step": 967
+    },
+    {
+      "epoch": 1.1304918989928479,
+      "grad_norm": 0.08053551941043195,
+      "learning_rate": 9.469849894137254e-06,
+      "loss": 0.2831,
+      "step": 968
+    },
+    {
+      "epoch": 1.131659611735513,
+      "grad_norm": 0.07971007564720493,
+      "learning_rate": 9.449479747238001e-06,
+      "loss": 0.2924,
+      "step": 969
+    },
+    {
+      "epoch": 1.1328273244781784,
+      "grad_norm": 0.07920684021218934,
+      "learning_rate": 9.42911189137232e-06,
+      "loss": 0.2957,
+      "step": 970
+    },
+    {
+      "epoch": 1.1339950372208436,
+      "grad_norm": 0.08137826968374537,
+      "learning_rate": 9.408746411302661e-06,
+      "loss": 0.2763,
+      "step": 971
+    },
+    {
+      "epoch": 1.1351627499635089,
+      "grad_norm": 0.08090088832172371,
+      "learning_rate": 9.388383391781576e-06,
+      "loss": 0.2809,
+      "step": 972
+    },
+    {
+      "epoch": 1.1363304627061743,
+      "grad_norm": 0.07850495741218193,
+      "learning_rate": 9.368022917551377e-06,
+      "loss": 0.282,
+      "step": 973
+    },
+    {
+      "epoch": 1.1374981754488396,
+      "grad_norm": 0.08009702913003008,
+      "learning_rate": 9.347665073343795e-06,
+      "loss": 0.284,
+      "step": 974
+    },
+    {
+      "epoch": 1.1386658881915048,
+      "grad_norm": 0.08587030541890489,
+      "learning_rate": 9.327309943879604e-06,
+      "loss": 0.2867,
+      "step": 975
+    },
+    {
+      "epoch": 1.1398336009341703,
+      "grad_norm": 0.07647982863928011,
+      "learning_rate": 9.306957613868292e-06,
+      "loss": 0.2561,
+      "step": 976
+    },
+    {
+      "epoch": 1.1410013136768355,
+      "grad_norm": 0.07914875824166742,
+      "learning_rate": 9.286608168007678e-06,
+      "loss": 0.2764,
+      "step": 977
+    },
+    {
+      "epoch": 1.1421690264195008,
+      "grad_norm": 0.08061397808985103,
+      "learning_rate": 9.266261690983601e-06,
+      "loss": 0.2631,
+      "step": 978
+    },
+    {
+      "epoch": 1.143336739162166,
+      "grad_norm": 0.08291425573597028,
+      "learning_rate": 9.245918267469528e-06,
+      "loss": 0.2748,
+      "step": 979
+    },
+    {
+      "epoch": 1.1445044519048313,
+      "grad_norm": 0.08057517819015265,
+      "learning_rate": 9.225577982126234e-06,
+      "loss": 0.2775,
+      "step": 980
+    },
+    {
+      "epoch": 1.1456721646474968,
+      "grad_norm": 0.08112344957303487,
+      "learning_rate": 9.205240919601417e-06,
+      "loss": 0.2744,
+      "step": 981
+    },
+    {
+      "epoch": 1.146839877390162,
+      "grad_norm": 0.0775943590658758,
+      "learning_rate": 9.184907164529369e-06,
+      "loss": 0.274,
+      "step": 982
+    },
+    {
+      "epoch": 1.1480075901328273,
+      "grad_norm": 0.08176658316393533,
+      "learning_rate": 9.164576801530628e-06,
+      "loss": 0.2925,
+      "step": 983
+    },
+    {
+      "epoch": 1.1491753028754927,
+      "grad_norm": 0.07883860539775449,
+      "learning_rate": 9.144249915211605e-06,
+      "loss": 0.2775,
+      "step": 984
+    },
+    {
+      "epoch": 1.150343015618158,
+      "grad_norm": 0.08338836379348273,
+      "learning_rate": 9.123926590164241e-06,
+      "loss": 0.2764,
+      "step": 985
+    },
+    {
+      "epoch": 1.1515107283608232,
+      "grad_norm": 0.07465442480926174,
+      "learning_rate": 9.103606910965666e-06,
+      "loss": 0.2681,
+      "step": 986
+    },
+    {
+      "epoch": 1.1526784411034885,
+      "grad_norm": 0.07707708031753953,
+      "learning_rate": 9.083290962177828e-06,
+      "loss": 0.2638,
+      "step": 987
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.0778282472586912,
+      "learning_rate": 9.06297882834716e-06,
+      "loss": 0.2794,
+      "step": 988
+    },
+    {
+      "epoch": 1.1550138665888192,
+      "grad_norm": 0.0781441880067255,
+      "learning_rate": 9.04267059400421e-06,
+      "loss": 0.2747,
+      "step": 989
+    },
+    {
+      "epoch": 1.1561815793314845,
+      "grad_norm": 0.07801947545236715,
+      "learning_rate": 9.022366343663299e-06,
+      "loss": 0.2729,
+      "step": 990
+    },
+    {
+      "epoch": 1.1573492920741497,
+      "grad_norm": 0.08030440834700672,
+      "learning_rate": 9.002066161822174e-06,
+      "loss": 0.279,
+      "step": 991
+    },
+    {
+      "epoch": 1.1585170048168152,
+      "grad_norm": 0.08365742079263765,
+      "learning_rate": 8.981770132961649e-06,
+      "loss": 0.275,
+      "step": 992
+    },
+    {
+      "epoch": 1.1596847175594804,
+      "grad_norm": 0.0781805469621553,
+      "learning_rate": 8.961478341545248e-06,
+      "loss": 0.2823,
+      "step": 993
+    },
+    {
+      "epoch": 1.1608524303021457,
+      "grad_norm": 0.08559490841023729,
+      "learning_rate": 8.94119087201887e-06,
+      "loss": 0.2906,
+      "step": 994
+    },
+    {
+      "epoch": 1.162020143044811,
+      "grad_norm": 0.08161137099657678,
+      "learning_rate": 8.920907808810423e-06,
+      "loss": 0.2696,
+      "step": 995
+    },
+    {
+      "epoch": 1.1631878557874762,
+      "grad_norm": 0.07695757798410034,
+      "learning_rate": 8.900629236329482e-06,
+      "loss": 0.2643,
+      "step": 996
+    },
+    {
+      "epoch": 1.1643555685301417,
+      "grad_norm": 0.07809513074575625,
+      "learning_rate": 8.880355238966923e-06,
+      "loss": 0.2925,
+      "step": 997
+    },
+    {
+      "epoch": 1.165523281272807,
+      "grad_norm": 0.08379663757752805,
+      "learning_rate": 8.860085901094595e-06,
+      "loss": 0.2721,
+      "step": 998
+    },
+    {
+      "epoch": 1.1666909940154722,
+      "grad_norm": 0.07543794301079866,
+      "learning_rate": 8.839821307064948e-06,
+      "loss": 0.2617,
+      "step": 999
+    },
+    {
+      "epoch": 1.1678587067581374,
+      "grad_norm": 0.08423719903504832,
+      "learning_rate": 8.819561541210698e-06,
+      "loss": 0.2874,
+      "step": 1000
+    },
+    {
+      "epoch": 1.1690264195008029,
+      "grad_norm": 0.08569440713729129,
+      "learning_rate": 8.799306687844459e-06,
+      "loss": 0.2775,
+      "step": 1001
+    },
+    {
+      "epoch": 1.1701941322434681,
+      "grad_norm": 0.07493197820006275,
+      "learning_rate": 8.779056831258402e-06,
+      "loss": 0.2623,
+      "step": 1002
+    },
+    {
+      "epoch": 1.1713618449861334,
+      "grad_norm": 0.08354860221700293,
+      "learning_rate": 8.758812055723912e-06,
+      "loss": 0.2898,
+      "step": 1003
+    },
+    {
+      "epoch": 1.1725295577287986,
+      "grad_norm": 0.07788456837147728,
+      "learning_rate": 8.738572445491225e-06,
+      "loss": 0.2587,
+      "step": 1004
+    },
+    {
+      "epoch": 1.173697270471464,
+      "grad_norm": 0.08101723872220169,
+      "learning_rate": 8.718338084789074e-06,
+      "loss": 0.2868,
+      "step": 1005
+    },
+    {
+      "epoch": 1.1748649832141294,
+      "grad_norm": 0.08412878003638012,
+      "learning_rate": 8.698109057824354e-06,
+      "loss": 0.284,
+      "step": 1006
+    },
+    {
+      "epoch": 1.1760326959567946,
+      "grad_norm": 0.0826473645928135,
+      "learning_rate": 8.67788544878176e-06,
+      "loss": 0.2717,
+      "step": 1007
+    },
+    {
+      "epoch": 1.1772004086994599,
+      "grad_norm": 0.08020358076230268,
+      "learning_rate": 8.657667341823449e-06,
+      "loss": 0.2762,
+      "step": 1008
+    },
+    {
+      "epoch": 1.1783681214421253,
+      "grad_norm": 0.07626942628845558,
+      "learning_rate": 8.637454821088657e-06,
+      "loss": 0.2637,
+      "step": 1009
+    },
+    {
+      "epoch": 1.1795358341847906,
+      "grad_norm": 0.07434651256416426,
+      "learning_rate": 8.6172479706934e-06,
+      "loss": 0.2603,
+      "step": 1010
+    },
+    {
+      "epoch": 1.1807035469274558,
+      "grad_norm": 0.0814707675369328,
+      "learning_rate": 8.597046874730074e-06,
+      "loss": 0.2639,
+      "step": 1011
+    },
+    {
+      "epoch": 1.181871259670121,
+      "grad_norm": 0.080501885405812,
+      "learning_rate": 8.576851617267151e-06,
+      "loss": 0.2898,
+      "step": 1012
+    },
+    {
+      "epoch": 1.1830389724127865,
+      "grad_norm": 0.08406491578986007,
+      "learning_rate": 8.556662282348782e-06,
+      "loss": 0.2944,
+      "step": 1013
+    },
+    {
+      "epoch": 1.1842066851554518,
+      "grad_norm": 0.07673596160305597,
+      "learning_rate": 8.53647895399448e-06,
+      "loss": 0.2651,
+      "step": 1014
+    },
+    {
+      "epoch": 1.185374397898117,
+      "grad_norm": 0.07598376004588657,
+      "learning_rate": 8.516301716198769e-06,
+      "loss": 0.2824,
+      "step": 1015
+    },
+    {
+      "epoch": 1.1865421106407823,
+      "grad_norm": 0.08003925504579347,
+      "learning_rate": 8.496130652930818e-06,
+      "loss": 0.2791,
+      "step": 1016
+    },
+    {
+      "epoch": 1.1877098233834475,
+      "grad_norm": 0.08087027248944775,
+      "learning_rate": 8.4759658481341e-06,
+      "loss": 0.2862,
+      "step": 1017
+    },
+    {
+      "epoch": 1.188877536126113,
+      "grad_norm": 0.0790661673741903,
+      "learning_rate": 8.455807385726047e-06,
+      "loss": 0.281,
+      "step": 1018
+    },
+    {
+      "epoch": 1.1900452488687783,
+      "grad_norm": 0.07386206124236583,
+      "learning_rate": 8.43565534959769e-06,
+      "loss": 0.2618,
+      "step": 1019
+    },
+    {
+      "epoch": 1.1912129616114435,
+      "grad_norm": 0.08327663386213299,
+      "learning_rate": 8.415509823613332e-06,
+      "loss": 0.2861,
+      "step": 1020
+    },
+    {
+      "epoch": 1.192380674354109,
+      "grad_norm": 0.07714949392581731,
+      "learning_rate": 8.395370891610162e-06,
+      "loss": 0.2888,
+      "step": 1021
+    },
+    {
+      "epoch": 1.1935483870967742,
+      "grad_norm": 0.0748906532153506,
+      "learning_rate": 8.375238637397942e-06,
+      "loss": 0.25,
+      "step": 1022
+    },
+    {
+      "epoch": 1.1947160998394395,
+      "grad_norm": 0.0789008532248992,
+      "learning_rate": 8.355113144758642e-06,
+      "loss": 0.2749,
+      "step": 1023
+    },
+    {
+      "epoch": 1.1958838125821047,
+      "grad_norm": 0.07990417609337108,
+      "learning_rate": 8.334994497446091e-06,
+      "loss": 0.2775,
+      "step": 1024
+    },
+    {
+      "epoch": 1.19705152532477,
+      "grad_norm": 0.07668941186712465,
+      "learning_rate": 8.31488277918563e-06,
+      "loss": 0.2619,
+      "step": 1025
+    },
+    {
+      "epoch": 1.1982192380674355,
+      "grad_norm": 0.07618539140622843,
+      "learning_rate": 8.294778073673762e-06,
+      "loss": 0.2479,
+      "step": 1026
+    },
+    {
+      "epoch": 1.1993869508101007,
+      "grad_norm": 0.07907294244471036,
+      "learning_rate": 8.274680464577816e-06,
+      "loss": 0.2748,
+      "step": 1027
+    },
+    {
+      "epoch": 1.200554663552766,
+      "grad_norm": 0.08080930894676777,
+      "learning_rate": 8.25459003553558e-06,
+      "loss": 0.2819,
+      "step": 1028
+    },
+    {
+      "epoch": 1.2017223762954314,
+      "grad_norm": 0.08494190048887319,
+      "learning_rate": 8.234506870154956e-06,
+      "loss": 0.2891,
+      "step": 1029
+    },
+    {
+      "epoch": 1.2028900890380967,
+      "grad_norm": 0.0764389720470024,
+      "learning_rate": 8.214431052013636e-06,
+      "loss": 0.2703,
+      "step": 1030
+    },
+    {
+      "epoch": 1.204057801780762,
+      "grad_norm": 0.07789634161747819,
+      "learning_rate": 8.194362664658715e-06,
+      "loss": 0.269,
+      "step": 1031
+    },
+    {
+      "epoch": 1.2052255145234272,
+      "grad_norm": 0.08006360543571898,
+      "learning_rate": 8.174301791606384e-06,
+      "loss": 0.2693,
+      "step": 1032
+    },
+    {
+      "epoch": 1.2063932272660924,
+      "grad_norm": 0.09741892366829225,
+      "learning_rate": 8.154248516341547e-06,
+      "loss": 0.2774,
+      "step": 1033
+    },
+    {
+      "epoch": 1.207560940008758,
+      "grad_norm": 0.08062972601341313,
+      "learning_rate": 8.134202922317496e-06,
+      "loss": 0.2751,
+      "step": 1034
+    },
+    {
+      "epoch": 1.2087286527514232,
+      "grad_norm": 0.08334318688173589,
+      "learning_rate": 8.114165092955556e-06,
+      "loss": 0.2772,
+      "step": 1035
+    },
+    {
+      "epoch": 1.2098963654940884,
+      "grad_norm": 0.08037497211699068,
+      "learning_rate": 8.094135111644741e-06,
+      "loss": 0.2859,
+      "step": 1036
+    },
+    {
+      "epoch": 1.2110640782367539,
+      "grad_norm": 0.07904553097925633,
+      "learning_rate": 8.074113061741397e-06,
+      "loss": 0.2724,
+      "step": 1037
+    },
+    {
+      "epoch": 1.2122317909794191,
+      "grad_norm": 0.0795808656813931,
+      "learning_rate": 8.054099026568874e-06,
+      "loss": 0.2601,
+      "step": 1038
+    },
+    {
+      "epoch": 1.2133995037220844,
+      "grad_norm": 0.08326089055067643,
+      "learning_rate": 8.03409308941716e-06,
+      "loss": 0.2799,
+      "step": 1039
+    },
+    {
+      "epoch": 1.2145672164647496,
+      "grad_norm": 0.08063248035880013,
+      "learning_rate": 8.014095333542548e-06,
+      "loss": 0.2905,
+      "step": 1040
+    },
+    {
+      "epoch": 1.2157349292074149,
+      "grad_norm": 0.08024486632063314,
+      "learning_rate": 7.994105842167274e-06,
+      "loss": 0.2904,
+      "step": 1041
+    },
+    {
+      "epoch": 1.2169026419500804,
+      "grad_norm": 0.07915332249303049,
+      "learning_rate": 7.974124698479192e-06,
+      "loss": 0.2596,
+      "step": 1042
+    },
+    {
+      "epoch": 1.2180703546927456,
+      "grad_norm": 0.08246987668611584,
+      "learning_rate": 7.954151985631409e-06,
+      "loss": 0.2747,
+      "step": 1043
+    },
+    {
+      "epoch": 1.2192380674354109,
+      "grad_norm": 0.0814932252080588,
+      "learning_rate": 7.934187786741956e-06,
+      "loss": 0.2754,
+      "step": 1044
+    },
+    {
+      "epoch": 1.220405780178076,
+      "grad_norm": 0.08330155290607977,
+      "learning_rate": 7.914232184893418e-06,
+      "loss": 0.2881,
+      "step": 1045
+    },
+    {
+      "epoch": 1.2215734929207416,
+      "grad_norm": 0.08259194625738267,
+      "learning_rate": 7.894285263132612e-06,
+      "loss": 0.2844,
+      "step": 1046
+    },
+    {
+      "epoch": 1.2227412056634068,
+      "grad_norm": 0.08187094551002078,
+      "learning_rate": 7.874347104470234e-06,
+      "loss": 0.2912,
+      "step": 1047
+    },
+    {
+      "epoch": 1.223908918406072,
+      "grad_norm": 0.08438380028099325,
+      "learning_rate": 7.854417791880508e-06,
+      "loss": 0.2843,
+      "step": 1048
+    },
+    {
+      "epoch": 1.2250766311487373,
+      "grad_norm": 0.07243368364570395,
+      "learning_rate": 7.834497408300839e-06,
+      "loss": 0.2653,
+      "step": 1049
+    },
+    {
+      "epoch": 1.2262443438914028,
+      "grad_norm": 0.07895538849723772,
+      "learning_rate": 7.814586036631485e-06,
+      "loss": 0.2717,
+      "step": 1050
+    },
+    {
+      "epoch": 1.227412056634068,
+      "grad_norm": 0.07702724140287973,
+      "learning_rate": 7.79468375973519e-06,
+      "loss": 0.2708,
+      "step": 1051
+    },
+    {
+      "epoch": 1.2285797693767333,
+      "grad_norm": 0.07698786186296315,
+      "learning_rate": 7.774790660436857e-06,
+      "loss": 0.2869,
+      "step": 1052
+    },
+    {
+      "epoch": 1.2297474821193985,
+      "grad_norm": 0.07662828407039395,
+      "learning_rate": 7.75490682152319e-06,
+      "loss": 0.2647,
+      "step": 1053
+    },
+    {
+      "epoch": 1.230915194862064,
+      "grad_norm": 0.07763008117014991,
+      "learning_rate": 7.735032325742355e-06,
+      "loss": 0.2538,
+      "step": 1054
+    },
+    {
+      "epoch": 1.2320829076047293,
+      "grad_norm": 0.07983742338818058,
+      "learning_rate": 7.715167255803643e-06,
+      "loss": 0.2634,
+      "step": 1055
+    },
+    {
+      "epoch": 1.2332506203473945,
+      "grad_norm": 0.0868417849770032,
+      "learning_rate": 7.695311694377116e-06,
+      "loss": 0.2868,
+      "step": 1056
+    },
+    {
+      "epoch": 1.2344183330900598,
+      "grad_norm": 0.08480968287078687,
+      "learning_rate": 7.675465724093254e-06,
+      "loss": 0.2922,
+      "step": 1057
+    },
+    {
+      "epoch": 1.2355860458327252,
+      "grad_norm": 0.08329316423799644,
+      "learning_rate": 7.655629427542636e-06,
+      "loss": 0.2825,
+      "step": 1058
+    },
+    {
+      "epoch": 1.2367537585753905,
+      "grad_norm": 0.085753863608901,
+      "learning_rate": 7.635802887275582e-06,
+      "loss": 0.2752,
+      "step": 1059
+    },
+    {
+      "epoch": 1.2379214713180557,
+      "grad_norm": 0.07886908735682098,
+      "learning_rate": 7.615986185801807e-06,
+      "loss": 0.2622,
+      "step": 1060
+    },
+    {
+      "epoch": 1.239089184060721,
+      "grad_norm": 0.07617134340513744,
+      "learning_rate": 7.596179405590076e-06,
+      "loss": 0.2588,
+      "step": 1061
+    },
+    {
+      "epoch": 1.2402568968033862,
+      "grad_norm": 0.07757648335835945,
+      "learning_rate": 7.576382629067877e-06,
+      "loss": 0.2564,
+      "step": 1062
+    },
+    {
+      "epoch": 1.2414246095460517,
+      "grad_norm": 0.0818095148993952,
+      "learning_rate": 7.556595938621058e-06,
+      "loss": 0.2548,
+      "step": 1063
+    },
+    {
+      "epoch": 1.242592322288717,
+      "grad_norm": 0.08333806428822678,
+      "learning_rate": 7.536819416593504e-06,
+      "loss": 0.2713,
+      "step": 1064
+    },
+    {
+      "epoch": 1.2437600350313822,
+      "grad_norm": 0.08354258345789843,
+      "learning_rate": 7.517053145286766e-06,
+      "loss": 0.2804,
+      "step": 1065
+    },
+    {
+      "epoch": 1.2449277477740477,
+      "grad_norm": 0.08419248652210201,
+      "learning_rate": 7.497297206959746e-06,
+      "loss": 0.2715,
+      "step": 1066
+    },
+    {
+      "epoch": 1.246095460516713,
+      "grad_norm": 0.08927010945782225,
+      "learning_rate": 7.47755168382835e-06,
+      "loss": 0.281,
+      "step": 1067
+    },
+    {
+      "epoch": 1.2472631732593782,
+      "grad_norm": 0.07984650767614575,
+      "learning_rate": 7.4578166580651335e-06,
+      "loss": 0.2731,
+      "step": 1068
+    },
+    {
+      "epoch": 1.2484308860020434,
+      "grad_norm": 0.08450138939694145,
+      "learning_rate": 7.438092211798959e-06,
+      "loss": 0.271,
+      "step": 1069
+    },
+    {
+      "epoch": 1.2495985987447087,
+      "grad_norm": 0.08376884222497256,
+      "learning_rate": 7.4183784271146796e-06,
+      "loss": 0.2755,
+      "step": 1070
+    },
+    {
+      "epoch": 1.2507663114873742,
+      "grad_norm": 0.08386174985918632,
+      "learning_rate": 7.398675386052767e-06,
+      "loss": 0.2809,
+      "step": 1071
+    },
+    {
+      "epoch": 1.2519340242300394,
+      "grad_norm": 0.0806098415740781,
+      "learning_rate": 7.378983170608982e-06,
+      "loss": 0.275,
+      "step": 1072
+    },
+    {
+      "epoch": 1.2531017369727047,
+      "grad_norm": 0.08216422637564584,
+      "learning_rate": 7.359301862734034e-06,
+      "loss": 0.2721,
+      "step": 1073
+    },
+    {
+      "epoch": 1.2542694497153701,
+      "grad_norm": 0.08260955403311829,
+      "learning_rate": 7.33963154433325e-06,
+      "loss": 0.2756,
+      "step": 1074
+    },
+    {
+      "epoch": 1.2554371624580354,
+      "grad_norm": 0.08016003853725238,
+      "learning_rate": 7.319972297266215e-06,
+      "loss": 0.2828,
+      "step": 1075
+    },
+    {
+      "epoch": 1.2566048752007006,
+      "grad_norm": 0.07965272205510701,
+      "learning_rate": 7.3003242033464314e-06,
+      "loss": 0.2709,
+      "step": 1076
+    },
+    {
+      "epoch": 1.2577725879433659,
+      "grad_norm": 0.0781004977141375,
+      "learning_rate": 7.280687344341006e-06,
+      "loss": 0.2585,
+      "step": 1077
+    },
+    {
+      "epoch": 1.2589403006860311,
+      "grad_norm": 0.08098251212479755,
+      "learning_rate": 7.2610618019702776e-06,
+      "loss": 0.2689,
+      "step": 1078
+    },
+    {
+      "epoch": 1.2601080134286966,
+      "grad_norm": 0.076513111414102,
+      "learning_rate": 7.241447657907494e-06,
+      "loss": 0.2583,
+      "step": 1079
+    },
+    {
+      "epoch": 1.2612757261713619,
+      "grad_norm": 0.07621357997830668,
+      "learning_rate": 7.221844993778464e-06,
+      "loss": 0.2617,
+      "step": 1080
+    },
+    {
+      "epoch": 1.262443438914027,
+      "grad_norm": 0.0774296712257869,
+      "learning_rate": 7.202253891161225e-06,
+      "loss": 0.2596,
+      "step": 1081
+    },
+    {
+      "epoch": 1.2636111516566926,
+      "grad_norm": 0.07935248855464203,
+      "learning_rate": 7.182674431585703e-06,
+      "loss": 0.2788,
+      "step": 1082
+    },
+    {
+      "epoch": 1.2647788643993578,
+      "grad_norm": 0.07825514277099219,
+      "learning_rate": 7.1631066965333705e-06,
+      "loss": 0.2611,
+      "step": 1083
+    },
+    {
+      "epoch": 1.265946577142023,
+      "grad_norm": 0.07854297650646869,
+      "learning_rate": 7.143550767436894e-06,
+      "loss": 0.26,
+      "step": 1084
+    },
+    {
+      "epoch": 1.2671142898846883,
+      "grad_norm": 0.07940161385242879,
+      "learning_rate": 7.124006725679828e-06,
+      "loss": 0.2692,
+      "step": 1085
+    },
+    {
+      "epoch": 1.2682820026273536,
+      "grad_norm": 0.07925514645687337,
+      "learning_rate": 7.1044746525962446e-06,
+      "loss": 0.2667,
+      "step": 1086
+    },
+    {
+      "epoch": 1.269449715370019,
+      "grad_norm": 0.07814962373250532,
+      "learning_rate": 7.084954629470417e-06,
+      "loss": 0.266,
+      "step": 1087
+    },
+    {
+      "epoch": 1.2706174281126843,
+      "grad_norm": 0.08193719478513752,
+      "learning_rate": 7.065446737536455e-06,
+      "loss": 0.2582,
+      "step": 1088
+    },
+    {
+      "epoch": 1.2717851408553495,
+      "grad_norm": 0.08060990954698471,
+      "learning_rate": 7.045951057978001e-06,
+      "loss": 0.267,
+      "step": 1089
+    },
+    {
+      "epoch": 1.272952853598015,
+      "grad_norm": 0.07713724537862279,
+      "learning_rate": 7.026467671927863e-06,
+      "loss": 0.2736,
+      "step": 1090
+    },
+    {
+      "epoch": 1.2741205663406803,
+      "grad_norm": 0.08342934409623246,
+      "learning_rate": 7.0069966604676996e-06,
+      "loss": 0.2921,
+      "step": 1091
+    },
+    {
+      "epoch": 1.2752882790833455,
+      "grad_norm": 0.07680877584897408,
+      "learning_rate": 6.9875381046276605e-06,
+      "loss": 0.2592,
+      "step": 1092
+    },
+    {
+      "epoch": 1.2764559918260108,
+      "grad_norm": 0.08042092555178937,
+      "learning_rate": 6.968092085386063e-06,
+      "loss": 0.2638,
+      "step": 1093
+    },
+    {
+      "epoch": 1.277623704568676,
+      "grad_norm": 0.08150644702869746,
+      "learning_rate": 6.948658683669056e-06,
+      "loss": 0.2936,
+      "step": 1094
+    },
+    {
+      "epoch": 1.2787914173113415,
+      "grad_norm": 0.0818423783038204,
+      "learning_rate": 6.929237980350283e-06,
+      "loss": 0.2714,
+      "step": 1095
+    },
+    {
+      "epoch": 1.2799591300540067,
+      "grad_norm": 0.07656920890413661,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 0.273,
+      "step": 1096
+    },
+    {
+      "epoch": 1.281126842796672,
+      "grad_norm": 0.0760517895248664,
+      "learning_rate": 6.8904349921374065e-06,
+      "loss": 0.286,
+      "step": 1097
+    },
+    {
+      "epoch": 1.2822945555393372,
+      "grad_norm": 0.0796067296601819,
+      "learning_rate": 6.871052868725012e-06,
+      "loss": 0.2719,
+      "step": 1098
+    },
+    {
+      "epoch": 1.2834622682820025,
+      "grad_norm": 0.08039989735896282,
+      "learning_rate": 6.851683766673587e-06,
+      "loss": 0.2682,
+      "step": 1099
+    },
+    {
+      "epoch": 1.284629981024668,
+      "grad_norm": 0.07724338091088408,
+      "learning_rate": 6.832327766589177e-06,
+      "loss": 0.2706,
+      "step": 1100
+    },
+    {
+      "epoch": 1.2857976937673332,
+      "grad_norm": 0.08105156403485583,
+      "learning_rate": 6.812984949023307e-06,
+      "loss": 0.2773,
+      "step": 1101
+    },
+    {
+      "epoch": 1.2869654065099985,
+      "grad_norm": 0.07858344073735259,
+      "learning_rate": 6.793655394472645e-06,
+      "loss": 0.2636,
+      "step": 1102
+    },
+    {
+      "epoch": 1.288133119252664,
+      "grad_norm": 0.08208782906644843,
+      "learning_rate": 6.774339183378663e-06,
+      "loss": 0.2821,
+      "step": 1103
+    },
+    {
+      "epoch": 1.2893008319953292,
+      "grad_norm": 0.0850003781469753,
+      "learning_rate": 6.755036396127297e-06,
+      "loss": 0.2793,
+      "step": 1104
+    },
+    {
+      "epoch": 1.2904685447379944,
+      "grad_norm": 0.07715932037299873,
+      "learning_rate": 6.73574711304862e-06,
+      "loss": 0.2754,
+      "step": 1105
+    },
+    {
+      "epoch": 1.2916362574806597,
+      "grad_norm": 0.07899790498043929,
+      "learning_rate": 6.716471414416519e-06,
+      "loss": 0.2746,
+      "step": 1106
+    },
+    {
+      "epoch": 1.292803970223325,
+      "grad_norm": 0.08247397366576847,
+      "learning_rate": 6.697209380448333e-06,
+      "loss": 0.2643,
+      "step": 1107
+    },
+    {
+      "epoch": 1.2939716829659904,
+      "grad_norm": 0.08459414607892492,
+      "learning_rate": 6.6779610913045344e-06,
+      "loss": 0.2679,
+      "step": 1108
+    },
+    {
+      "epoch": 1.2951393957086557,
+      "grad_norm": 0.08190684329130002,
+      "learning_rate": 6.6587266270884095e-06,
+      "loss": 0.2723,
+      "step": 1109
+    },
+    {
+      "epoch": 1.296307108451321,
+      "grad_norm": 0.07843294005827077,
+      "learning_rate": 6.639506067845698e-06,
+      "loss": 0.2694,
+      "step": 1110
+    },
+    {
+      "epoch": 1.2974748211939864,
+      "grad_norm": 0.08016611228098161,
+      "learning_rate": 6.620299493564283e-06,
+      "loss": 0.2574,
+      "step": 1111
+    },
+    {
+      "epoch": 1.2986425339366516,
+      "grad_norm": 0.08480717097442429,
+      "learning_rate": 6.601106984173835e-06,
+      "loss": 0.2766,
+      "step": 1112
+    },
+    {
+      "epoch": 1.2998102466793169,
+      "grad_norm": 0.08017697699328992,
+      "learning_rate": 6.581928619545501e-06,
+      "loss": 0.2429,
+      "step": 1113
+    },
+    {
+      "epoch": 1.3009779594219821,
+      "grad_norm": 0.0836600049344475,
+      "learning_rate": 6.562764479491566e-06,
+      "loss": 0.2927,
+      "step": 1114
+    },
+    {
+      "epoch": 1.3021456721646474,
+      "grad_norm": 0.07629170764523045,
+      "learning_rate": 6.543614643765116e-06,
+      "loss": 0.2551,
+      "step": 1115
+    },
+    {
+      "epoch": 1.3033133849073129,
+      "grad_norm": 0.07861034815269324,
+      "learning_rate": 6.524479192059699e-06,
+      "loss": 0.2714,
+      "step": 1116
+    },
+    {
+      "epoch": 1.304481097649978,
+      "grad_norm": 0.08055861316480498,
+      "learning_rate": 6.505358204009018e-06,
+      "loss": 0.2743,
+      "step": 1117
+    },
+    {
+      "epoch": 1.3056488103926434,
+      "grad_norm": 0.081575870071612,
+      "learning_rate": 6.486251759186573e-06,
+      "loss": 0.2701,
+      "step": 1118
+    },
+    {
+      "epoch": 1.3068165231353088,
+      "grad_norm": 0.08560564003163043,
+      "learning_rate": 6.467159937105352e-06,
+      "loss": 0.301,
+      "step": 1119
+    },
+    {
+      "epoch": 1.307984235877974,
+      "grad_norm": 0.08159609786850255,
+      "learning_rate": 6.4480828172174714e-06,
+      "loss": 0.2734,
+      "step": 1120
+    },
+    {
+      "epoch": 1.3091519486206393,
+      "grad_norm": 0.08150193939102428,
+      "learning_rate": 6.429020478913885e-06,
+      "loss": 0.2694,
+      "step": 1121
+    },
+    {
+      "epoch": 1.3103196613633046,
+      "grad_norm": 0.08373102111792718,
+      "learning_rate": 6.409973001524012e-06,
+      "loss": 0.288,
+      "step": 1122
+    },
+    {
+      "epoch": 1.3114873741059698,
+      "grad_norm": 0.08258863711797637,
+      "learning_rate": 6.390940464315448e-06,
+      "loss": 0.2709,
+      "step": 1123
+    },
+    {
+      "epoch": 1.3126550868486353,
+      "grad_norm": 0.08348137798554175,
+      "learning_rate": 6.3719229464935915e-06,
+      "loss": 0.275,
+      "step": 1124
+    },
+    {
+      "epoch": 1.3138227995913005,
+      "grad_norm": 0.07833514047542045,
+      "learning_rate": 6.352920527201349e-06,
+      "loss": 0.2585,
+      "step": 1125
+    },
+    {
+      "epoch": 1.3149905123339658,
+      "grad_norm": 0.07930536615532793,
+      "learning_rate": 6.333933285518797e-06,
+      "loss": 0.268,
+      "step": 1126
+    },
+    {
+      "epoch": 1.3161582250766313,
+      "grad_norm": 0.0805589853983907,
+      "learning_rate": 6.314961300462841e-06,
+      "loss": 0.2699,
+      "step": 1127
+    },
+    {
+      "epoch": 1.3173259378192965,
+      "grad_norm": 0.0793180349120997,
+      "learning_rate": 6.29600465098689e-06,
+      "loss": 0.2602,
+      "step": 1128
+    },
+    {
+      "epoch": 1.3184936505619618,
+      "grad_norm": 0.07664422646887044,
+      "learning_rate": 6.277063415980549e-06,
+      "loss": 0.2666,
+      "step": 1129
+    },
+    {
+      "epoch": 1.319661363304627,
+      "grad_norm": 0.08253273207218506,
+      "learning_rate": 6.258137674269262e-06,
+      "loss": 0.274,
+      "step": 1130
+    },
+    {
+      "epoch": 1.3208290760472923,
+      "grad_norm": 0.08024537750629436,
+      "learning_rate": 6.239227504614004e-06,
+      "loss": 0.2741,
+      "step": 1131
+    },
+    {
+      "epoch": 1.3219967887899577,
+      "grad_norm": 0.0803343252692184,
+      "learning_rate": 6.220332985710936e-06,
+      "loss": 0.272,
+      "step": 1132
+    },
+    {
+      "epoch": 1.323164501532623,
+      "grad_norm": 0.08329667522706533,
+      "learning_rate": 6.201454196191095e-06,
+      "loss": 0.2703,
+      "step": 1133
+    },
+    {
+      "epoch": 1.3243322142752882,
+      "grad_norm": 0.08261926075969993,
+      "learning_rate": 6.182591214620057e-06,
+      "loss": 0.2722,
+      "step": 1134
+    },
+    {
+      "epoch": 1.3254999270179537,
+      "grad_norm": 0.07645899960075225,
+      "learning_rate": 6.1637441194976145e-06,
+      "loss": 0.2534,
+      "step": 1135
+    },
+    {
+      "epoch": 1.326667639760619,
+      "grad_norm": 0.07901132006163175,
+      "learning_rate": 6.144912989257441e-06,
+      "loss": 0.2532,
+      "step": 1136
+    },
+    {
+      "epoch": 1.3278353525032842,
+      "grad_norm": 0.08339994929479366,
+      "learning_rate": 6.126097902266772e-06,
+      "loss": 0.2877,
+      "step": 1137
+    },
+    {
+      "epoch": 1.3290030652459495,
+      "grad_norm": 0.08096689005571037,
+      "learning_rate": 6.107298936826086e-06,
+      "loss": 0.2723,
+      "step": 1138
+    },
+    {
+      "epoch": 1.3301707779886147,
+      "grad_norm": 0.07795372763316186,
+      "learning_rate": 6.088516171168763e-06,
+      "loss": 0.271,
+      "step": 1139
+    },
+    {
+      "epoch": 1.3313384907312802,
+      "grad_norm": 0.0805117599592193,
+      "learning_rate": 6.069749683460765e-06,
+      "loss": 0.2739,
+      "step": 1140
+    },
+    {
+      "epoch": 1.3325062034739454,
+      "grad_norm": 0.07790860166980448,
+      "learning_rate": 6.050999551800318e-06,
+      "loss": 0.2547,
+      "step": 1141
+    },
+    {
+      "epoch": 1.3336739162166107,
+      "grad_norm": 0.07877282166707496,
+      "learning_rate": 6.032265854217574e-06,
+      "loss": 0.2696,
+      "step": 1142
+    },
+    {
+      "epoch": 1.334841628959276,
+      "grad_norm": 0.08533417583683628,
+      "learning_rate": 6.013548668674305e-06,
+      "loss": 0.2799,
+      "step": 1143
+    },
+    {
+      "epoch": 1.3360093417019412,
+      "grad_norm": 0.07945614986915076,
+      "learning_rate": 5.994848073063552e-06,
+      "loss": 0.2724,
+      "step": 1144
+    },
+    {
+      "epoch": 1.3371770544446067,
+      "grad_norm": 0.0836577934092827,
+      "learning_rate": 5.9761641452093225e-06,
+      "loss": 0.2992,
+      "step": 1145
+    },
+    {
+      "epoch": 1.338344767187272,
+      "grad_norm": 0.07996822338376348,
+      "learning_rate": 5.957496962866263e-06,
+      "loss": 0.2787,
+      "step": 1146
+    },
+    {
+      "epoch": 1.3395124799299372,
+      "grad_norm": 0.07599143610111771,
+      "learning_rate": 5.938846603719333e-06,
+      "loss": 0.2609,
+      "step": 1147
+    },
+    {
+      "epoch": 1.3406801926726026,
+      "grad_norm": 0.07789766899374329,
+      "learning_rate": 5.9202131453834664e-06,
+      "loss": 0.2608,
+      "step": 1148
+    },
+    {
+      "epoch": 1.3418479054152679,
+      "grad_norm": 0.07919250882558741,
+      "learning_rate": 5.901596665403284e-06,
+      "loss": 0.2662,
+      "step": 1149
+    },
+    {
+      "epoch": 1.3430156181579331,
+      "grad_norm": 0.07676325607928641,
+      "learning_rate": 5.8829972412527325e-06,
+      "loss": 0.2627,
+      "step": 1150
+    },
+    {
+      "epoch": 1.3441833309005984,
+      "grad_norm": 0.07992204659835077,
+      "learning_rate": 5.864414950334796e-06,
+      "loss": 0.2973,
+      "step": 1151
+    },
+    {
+      "epoch": 1.3453510436432636,
+      "grad_norm": 0.07727685971770347,
+      "learning_rate": 5.845849869981137e-06,
+      "loss": 0.2658,
+      "step": 1152
+    },
+    {
+      "epoch": 1.346518756385929,
+      "grad_norm": 0.08068880781983256,
+      "learning_rate": 5.827302077451813e-06,
+      "loss": 0.309,
+      "step": 1153
+    },
+    {
+      "epoch": 1.3476864691285944,
+      "grad_norm": 0.08063308684081251,
+      "learning_rate": 5.808771649934924e-06,
+      "loss": 0.2927,
+      "step": 1154
+    },
+    {
+      "epoch": 1.3488541818712596,
+      "grad_norm": 0.08306988990919738,
+      "learning_rate": 5.790258664546318e-06,
+      "loss": 0.2761,
+      "step": 1155
+    },
+    {
+      "epoch": 1.350021894613925,
+      "grad_norm": 0.08395625707166714,
+      "learning_rate": 5.7717631983292375e-06,
+      "loss": 0.2866,
+      "step": 1156
+    },
+    {
+      "epoch": 1.3511896073565903,
+      "grad_norm": 0.07993611916684124,
+      "learning_rate": 5.753285328254039e-06,
+      "loss": 0.2672,
+      "step": 1157
+    },
+    {
+      "epoch": 1.3523573200992556,
+      "grad_norm": 0.07793082342638116,
+      "learning_rate": 5.73482513121783e-06,
+      "loss": 0.26,
+      "step": 1158
+    },
+    {
+      "epoch": 1.3535250328419208,
+      "grad_norm": 0.07890986523489135,
+      "learning_rate": 5.716382684044191e-06,
+      "loss": 0.2734,
+      "step": 1159
+    },
+    {
+      "epoch": 1.354692745584586,
+      "grad_norm": 0.08567493100767498,
+      "learning_rate": 5.6979580634828125e-06,
+      "loss": 0.3005,
+      "step": 1160
+    },
+    {
+      "epoch": 1.3558604583272515,
+      "grad_norm": 0.07551187840047584,
+      "learning_rate": 5.679551346209218e-06,
+      "loss": 0.2501,
+      "step": 1161
+    },
+    {
+      "epoch": 1.3570281710699168,
+      "grad_norm": 0.07616023128663583,
+      "learning_rate": 5.66116260882442e-06,
+      "loss": 0.2752,
+      "step": 1162
+    },
+    {
+      "epoch": 1.358195883812582,
+      "grad_norm": 0.0733543340354056,
+      "learning_rate": 5.642791927854599e-06,
+      "loss": 0.2534,
+      "step": 1163
+    },
+    {
+      "epoch": 1.3593635965552475,
+      "grad_norm": 0.08356776130802876,
+      "learning_rate": 5.6244393797507944e-06,
+      "loss": 0.264,
+      "step": 1164
+    },
+    {
+      "epoch": 1.3605313092979128,
+      "grad_norm": 0.08281552325150608,
+      "learning_rate": 5.60610504088859e-06,
+      "loss": 0.2758,
+      "step": 1165
+    },
+    {
+      "epoch": 1.361699022040578,
+      "grad_norm": 0.07948430307289199,
+      "learning_rate": 5.587788987567785e-06,
+      "loss": 0.2707,
+      "step": 1166
+    },
+    {
+      "epoch": 1.3628667347832433,
+      "grad_norm": 0.08113946372699823,
+      "learning_rate": 5.569491296012095e-06,
+      "loss": 0.2867,
+      "step": 1167
+    },
+    {
+      "epoch": 1.3640344475259085,
+      "grad_norm": 0.07653624580511537,
+      "learning_rate": 5.551212042368792e-06,
+      "loss": 0.2552,
+      "step": 1168
+    },
+    {
+      "epoch": 1.365202160268574,
+      "grad_norm": 0.07445160639644861,
+      "learning_rate": 5.532951302708447e-06,
+      "loss": 0.2694,
+      "step": 1169
+    },
+    {
+      "epoch": 1.3663698730112392,
+      "grad_norm": 0.08136504481152963,
+      "learning_rate": 5.514709153024571e-06,
+      "loss": 0.2709,
+      "step": 1170
+    },
+    {
+      "epoch": 1.3675375857539045,
+      "grad_norm": 0.08256352402110746,
+      "learning_rate": 5.496485669233315e-06,
+      "loss": 0.2775,
+      "step": 1171
+    },
+    {
+      "epoch": 1.36870529849657,
+      "grad_norm": 0.07865392035346942,
+      "learning_rate": 5.478280927173145e-06,
+      "loss": 0.2383,
+      "step": 1172
+    },
+    {
+      "epoch": 1.3698730112392352,
+      "grad_norm": 0.0849835453643583,
+      "learning_rate": 5.460095002604533e-06,
+      "loss": 0.2745,
+      "step": 1173
+    },
+    {
+      "epoch": 1.3710407239819005,
+      "grad_norm": 0.07776509661779822,
+      "learning_rate": 5.441927971209644e-06,
+      "loss": 0.2622,
+      "step": 1174
+    },
+    {
+      "epoch": 1.3722084367245657,
+      "grad_norm": 0.07537194321466595,
+      "learning_rate": 5.423779908592022e-06,
+      "loss": 0.2637,
+      "step": 1175
+    },
+    {
+      "epoch": 1.373376149467231,
+      "grad_norm": 0.07648349043669173,
+      "learning_rate": 5.405650890276255e-06,
+      "loss": 0.2681,
+      "step": 1176
+    },
+    {
+      "epoch": 1.3745438622098964,
+      "grad_norm": 0.07814792208639848,
+      "learning_rate": 5.387540991707697e-06,
+      "loss": 0.268,
+      "step": 1177
+    },
+    {
+      "epoch": 1.3757115749525617,
+      "grad_norm": 0.07828776000560031,
+      "learning_rate": 5.369450288252113e-06,
+      "loss": 0.2646,
+      "step": 1178
+    },
+    {
+      "epoch": 1.376879287695227,
+      "grad_norm": 0.08008323740940324,
+      "learning_rate": 5.351378855195405e-06,
+      "loss": 0.2597,
+      "step": 1179
+    },
+    {
+      "epoch": 1.3780470004378924,
+      "grad_norm": 0.07619955079635542,
+      "learning_rate": 5.333326767743263e-06,
+      "loss": 0.2551,
+      "step": 1180
+    },
+    {
+      "epoch": 1.3792147131805577,
+      "grad_norm": 0.07977278828629672,
+      "learning_rate": 5.315294101020882e-06,
+      "loss": 0.2854,
+      "step": 1181
+    },
+    {
+      "epoch": 1.380382425923223,
+      "grad_norm": 0.0792971340192185,
+      "learning_rate": 5.297280930072632e-06,
+      "loss": 0.2625,
+      "step": 1182
+    },
+    {
+      "epoch": 1.3815501386658882,
+      "grad_norm": 0.07847328809806553,
+      "learning_rate": 5.279287329861749e-06,
+      "loss": 0.2724,
+      "step": 1183
+    },
+    {
+      "epoch": 1.3827178514085534,
+      "grad_norm": 0.0736938942860433,
+      "learning_rate": 5.2613133752700145e-06,
+      "loss": 0.2482,
+      "step": 1184
+    },
+    {
+      "epoch": 1.3838855641512189,
+      "grad_norm": 0.08697928983561644,
+      "learning_rate": 5.243359141097469e-06,
+      "loss": 0.2857,
+      "step": 1185
+    },
+    {
+      "epoch": 1.3850532768938841,
+      "grad_norm": 0.08029362471925712,
+      "learning_rate": 5.225424702062082e-06,
+      "loss": 0.2751,
+      "step": 1186
+    },
+    {
+      "epoch": 1.3862209896365494,
+      "grad_norm": 0.0772852833373016,
+      "learning_rate": 5.207510132799436e-06,
+      "loss": 0.2832,
+      "step": 1187
+    },
+    {
+      "epoch": 1.3873887023792146,
+      "grad_norm": 0.07910870894787105,
+      "learning_rate": 5.1896155078624225e-06,
+      "loss": 0.2652,
+      "step": 1188
+    },
+    {
+      "epoch": 1.3885564151218799,
+      "grad_norm": 0.08005795422067913,
+      "learning_rate": 5.171740901720943e-06,
+      "loss": 0.2809,
+      "step": 1189
+    },
+    {
+      "epoch": 1.3897241278645454,
+      "grad_norm": 0.08188021763932364,
+      "learning_rate": 5.153886388761586e-06,
+      "loss": 0.2646,
+      "step": 1190
+    },
+    {
+      "epoch": 1.3908918406072106,
+      "grad_norm": 0.07634842502542961,
+      "learning_rate": 5.136052043287324e-06,
+      "loss": 0.2528,
+      "step": 1191
+    },
+    {
+      "epoch": 1.3920595533498759,
+      "grad_norm": 0.07589538521530355,
+      "learning_rate": 5.11823793951719e-06,
+      "loss": 0.2575,
+      "step": 1192
+    },
+    {
+      "epoch": 1.3932272660925413,
+      "grad_norm": 0.0805296094764102,
+      "learning_rate": 5.100444151585982e-06,
+      "loss": 0.2603,
+      "step": 1193
+    },
+    {
+      "epoch": 1.3943949788352066,
+      "grad_norm": 0.07761960609909402,
+      "learning_rate": 5.082670753543961e-06,
+      "loss": 0.2599,
+      "step": 1194
+    },
+    {
+      "epoch": 1.3955626915778718,
+      "grad_norm": 0.079144359541128,
+      "learning_rate": 5.064917819356532e-06,
+      "loss": 0.2583,
+      "step": 1195
+    },
+    {
+      "epoch": 1.396730404320537,
+      "grad_norm": 0.08512709631402766,
+      "learning_rate": 5.0471854229039286e-06,
+      "loss": 0.2611,
+      "step": 1196
+    },
+    {
+      "epoch": 1.3978981170632023,
+      "grad_norm": 0.08088631161690457,
+      "learning_rate": 5.029473637980926e-06,
+      "loss": 0.2792,
+      "step": 1197
+    },
+    {
+      "epoch": 1.3990658298058678,
+      "grad_norm": 0.07992030180841313,
+      "learning_rate": 5.011782538296512e-06,
+      "loss": 0.2694,
+      "step": 1198
+    },
+    {
+      "epoch": 1.400233542548533,
+      "grad_norm": 0.0840365320080537,
+      "learning_rate": 4.994112197473605e-06,
+      "loss": 0.2774,
+      "step": 1199
+    },
+    {
+      "epoch": 1.4014012552911983,
+      "grad_norm": 0.07985596042687791,
+      "learning_rate": 4.976462689048718e-06,
+      "loss": 0.2524,
+      "step": 1200
+    },
+    {
+      "epoch": 1.4025689680338638,
+      "grad_norm": 0.07931090954197853,
+      "learning_rate": 4.958834086471683e-06,
+      "loss": 0.2638,
+      "step": 1201
+    },
+    {
+      "epoch": 1.403736680776529,
+      "grad_norm": 0.07820684465962696,
+      "learning_rate": 4.941226463105322e-06,
+      "loss": 0.2624,
+      "step": 1202
+    },
+    {
+      "epoch": 1.4049043935191943,
+      "grad_norm": 0.08175188156784262,
+      "learning_rate": 4.923639892225147e-06,
+      "loss": 0.2844,
+      "step": 1203
+    },
+    {
+      "epoch": 1.4060721062618595,
+      "grad_norm": 0.07728509159517827,
+      "learning_rate": 4.906074447019068e-06,
+      "loss": 0.2716,
+      "step": 1204
+    },
+    {
+      "epoch": 1.4072398190045248,
+      "grad_norm": 0.07405519434221318,
+      "learning_rate": 4.888530200587075e-06,
+      "loss": 0.2489,
+      "step": 1205
+    },
+    {
+      "epoch": 1.4084075317471902,
+      "grad_norm": 0.07755369466614201,
+      "learning_rate": 4.87100722594094e-06,
+      "loss": 0.2773,
+      "step": 1206
+    },
+    {
+      "epoch": 1.4095752444898555,
+      "grad_norm": 0.07989074088973884,
+      "learning_rate": 4.8535055960039046e-06,
+      "loss": 0.2784,
+      "step": 1207
+    },
+    {
+      "epoch": 1.4107429572325207,
+      "grad_norm": 0.0788935589776889,
+      "learning_rate": 4.836025383610382e-06,
+      "loss": 0.2684,
+      "step": 1208
+    },
+    {
+      "epoch": 1.4119106699751862,
+      "grad_norm": 0.07648079270079874,
+      "learning_rate": 4.818566661505663e-06,
+      "loss": 0.2551,
+      "step": 1209
+    },
+    {
+      "epoch": 1.4130783827178515,
+      "grad_norm": 0.07798116120298156,
+      "learning_rate": 4.801129502345605e-06,
+      "loss": 0.2745,
+      "step": 1210
+    },
+    {
+      "epoch": 1.4142460954605167,
+      "grad_norm": 0.076503866296083,
+      "learning_rate": 4.783713978696315e-06,
+      "loss": 0.264,
+      "step": 1211
+    },
+    {
+      "epoch": 1.415413808203182,
+      "grad_norm": 0.07734816358236511,
+      "learning_rate": 4.766320163033882e-06,
+      "loss": 0.2676,
+      "step": 1212
+    },
+    {
+      "epoch": 1.4165815209458472,
+      "grad_norm": 0.07694343776607515,
+      "learning_rate": 4.748948127744037e-06,
+      "loss": 0.2524,
+      "step": 1213
+    },
+    {
+      "epoch": 1.4177492336885127,
+      "grad_norm": 0.08101944697952569,
+      "learning_rate": 4.7315979451218866e-06,
+      "loss": 0.2744,
+      "step": 1214
+    },
+    {
+      "epoch": 1.418916946431178,
+      "grad_norm": 0.0801269111579197,
+      "learning_rate": 4.714269687371581e-06,
+      "loss": 0.2647,
+      "step": 1215
+    },
+    {
+      "epoch": 1.4200846591738432,
+      "grad_norm": 0.07953778184970071,
+      "learning_rate": 4.696963426606041e-06,
+      "loss": 0.262,
+      "step": 1216
+    },
+    {
+      "epoch": 1.4212523719165087,
+      "grad_norm": 0.0831927907536156,
+      "learning_rate": 4.679679234846636e-06,
+      "loss": 0.2816,
+      "step": 1217
+    },
+    {
+      "epoch": 1.422420084659174,
+      "grad_norm": 0.07942767120772899,
+      "learning_rate": 4.6624171840229005e-06,
+      "loss": 0.2783,
+      "step": 1218
+    },
+    {
+      "epoch": 1.4235877974018392,
+      "grad_norm": 0.07756529871067964,
+      "learning_rate": 4.645177345972218e-06,
+      "loss": 0.2636,
+      "step": 1219
+    },
+    {
+      "epoch": 1.4247555101445044,
+      "grad_norm": 0.07857272686904918,
+      "learning_rate": 4.6279597924395434e-06,
+      "loss": 0.2595,
+      "step": 1220
+    },
+    {
+      "epoch": 1.4259232228871697,
+      "grad_norm": 0.08414206879289153,
+      "learning_rate": 4.61076459507709e-06,
+      "loss": 0.2725,
+      "step": 1221
+    },
+    {
+      "epoch": 1.4270909356298351,
+      "grad_norm": 0.07603649809030508,
+      "learning_rate": 4.593591825444028e-06,
+      "loss": 0.2668,
+      "step": 1222
+    },
+    {
+      "epoch": 1.4282586483725004,
+      "grad_norm": 0.08704109368354832,
+      "learning_rate": 4.576441555006193e-06,
+      "loss": 0.2659,
+      "step": 1223
+    },
+    {
+      "epoch": 1.4294263611151656,
+      "grad_norm": 0.08033482670318451,
+      "learning_rate": 4.559313855135795e-06,
+      "loss": 0.2536,
+      "step": 1224
+    },
+    {
+      "epoch": 1.430594073857831,
+      "grad_norm": 0.07824713875071342,
+      "learning_rate": 4.542208797111113e-06,
+      "loss": 0.2586,
+      "step": 1225
+    },
+    {
+      "epoch": 1.4317617866004964,
+      "grad_norm": 0.08554111011278187,
+      "learning_rate": 4.525126452116201e-06,
+      "loss": 0.2732,
+      "step": 1226
+    },
+    {
+      "epoch": 1.4329294993431616,
+      "grad_norm": 0.08128608008949294,
+      "learning_rate": 4.508066891240586e-06,
+      "loss": 0.2686,
+      "step": 1227
+    },
+    {
+      "epoch": 1.4340972120858269,
+      "grad_norm": 0.07568038711948781,
+      "learning_rate": 4.491030185478976e-06,
+      "loss": 0.2521,
+      "step": 1228
+    },
+    {
+      "epoch": 1.435264924828492,
+      "grad_norm": 0.08247777461704438,
+      "learning_rate": 4.474016405730973e-06,
+      "loss": 0.2739,
+      "step": 1229
+    },
+    {
+      "epoch": 1.4364326375711576,
+      "grad_norm": 0.08197298061941852,
+      "learning_rate": 4.457025622800771e-06,
+      "loss": 0.2617,
+      "step": 1230
+    },
+    {
+      "epoch": 1.4376003503138228,
+      "grad_norm": 0.08508547490135471,
+      "learning_rate": 4.440057907396854e-06,
+      "loss": 0.2776,
+      "step": 1231
+    },
+    {
+      "epoch": 1.438768063056488,
+      "grad_norm": 0.0773544432377798,
+      "learning_rate": 4.423113330131708e-06,
+      "loss": 0.2644,
+      "step": 1232
+    },
+    {
+      "epoch": 1.4399357757991533,
+      "grad_norm": 0.0802679994222836,
+      "learning_rate": 4.406191961521533e-06,
+      "loss": 0.253,
+      "step": 1233
+    },
+    {
+      "epoch": 1.4411034885418186,
+      "grad_norm": 0.07978746375175322,
+      "learning_rate": 4.389293871985949e-06,
+      "loss": 0.2743,
+      "step": 1234
+    },
+    {
+      "epoch": 1.442271201284484,
+      "grad_norm": 0.08141814050314564,
+      "learning_rate": 4.372419131847684e-06,
+      "loss": 0.275,
+      "step": 1235
+    },
+    {
+      "epoch": 1.4434389140271493,
+      "grad_norm": 0.08438685483576053,
+      "learning_rate": 4.355567811332311e-06,
+      "loss": 0.2539,
+      "step": 1236
+    },
+    {
+      "epoch": 1.4446066267698146,
+      "grad_norm": 0.08996661076773448,
+      "learning_rate": 4.338739980567926e-06,
+      "loss": 0.27,
+      "step": 1237
+    },
+    {
+      "epoch": 1.44577433951248,
+      "grad_norm": 0.07785124188790704,
+      "learning_rate": 4.321935709584884e-06,
+      "loss": 0.2695,
+      "step": 1238
+    },
+    {
+      "epoch": 1.4469420522551453,
+      "grad_norm": 0.0798680963392215,
+      "learning_rate": 4.305155068315481e-06,
+      "loss": 0.2652,
+      "step": 1239
+    },
+    {
+      "epoch": 1.4481097649978105,
+      "grad_norm": 0.08932525010240631,
+      "learning_rate": 4.2883981265936884e-06,
+      "loss": 0.2837,
+      "step": 1240
+    },
+    {
+      "epoch": 1.4492774777404758,
+      "grad_norm": 0.0839601959196822,
+      "learning_rate": 4.271664954154845e-06,
+      "loss": 0.2772,
+      "step": 1241
+    },
+    {
+      "epoch": 1.450445190483141,
+      "grad_norm": 0.08212802574304408,
+      "learning_rate": 4.254955620635371e-06,
+      "loss": 0.2648,
+      "step": 1242
+    },
+    {
+      "epoch": 1.4516129032258065,
+      "grad_norm": 0.07779111159590947,
+      "learning_rate": 4.2382701955724724e-06,
+      "loss": 0.2514,
+      "step": 1243
+    },
+    {
+      "epoch": 1.4527806159684717,
+      "grad_norm": 0.07685475034593989,
+      "learning_rate": 4.221608748403872e-06,
+      "loss": 0.2591,
+      "step": 1244
+    },
+    {
+      "epoch": 1.453948328711137,
+      "grad_norm": 0.08254453069374255,
+      "learning_rate": 4.204971348467498e-06,
+      "loss": 0.2542,
+      "step": 1245
+    },
+    {
+      "epoch": 1.4551160414538025,
+      "grad_norm": 0.08281089722387855,
+      "learning_rate": 4.188358065001215e-06,
+      "loss": 0.2629,
+      "step": 1246
+    },
+    {
+      "epoch": 1.4562837541964677,
+      "grad_norm": 0.08518132088850595,
+      "learning_rate": 4.1717689671425e-06,
+      "loss": 0.2716,
+      "step": 1247
+    },
+    {
+      "epoch": 1.457451466939133,
+      "grad_norm": 0.07766952294314886,
+      "learning_rate": 4.155204123928205e-06,
+      "loss": 0.2528,
+      "step": 1248
+    },
+    {
+      "epoch": 1.4586191796817982,
+      "grad_norm": 0.08021912732855258,
+      "learning_rate": 4.138663604294233e-06,
+      "loss": 0.2677,
+      "step": 1249
+    },
+    {
+      "epoch": 1.4597868924244635,
+      "grad_norm": 0.07539662880436784,
+      "learning_rate": 4.12214747707527e-06,
+      "loss": 0.2584,
+      "step": 1250
+    },
+    {
+      "epoch": 1.460954605167129,
+      "grad_norm": 0.07905228325090419,
+      "learning_rate": 4.105655811004481e-06,
+      "loss": 0.2603,
+      "step": 1251
+    },
+    {
+      "epoch": 1.4621223179097942,
+      "grad_norm": 0.07638986703177673,
+      "learning_rate": 4.0891886747132356e-06,
+      "loss": 0.2593,
+      "step": 1252
+    },
+    {
+      "epoch": 1.4632900306524594,
+      "grad_norm": 0.07768459627684542,
+      "learning_rate": 4.072746136730828e-06,
+      "loss": 0.266,
+      "step": 1253
+    },
+    {
+      "epoch": 1.464457743395125,
+      "grad_norm": 0.07528934725243515,
+      "learning_rate": 4.056328265484184e-06,
+      "loss": 0.2731,
+      "step": 1254
+    },
+    {
+      "epoch": 1.4656254561377902,
+      "grad_norm": 0.07876685847921226,
+      "learning_rate": 4.039935129297565e-06,
+      "loss": 0.2599,
+      "step": 1255
+    },
+    {
+      "epoch": 1.4667931688804554,
+      "grad_norm": 0.08336030061528776,
+      "learning_rate": 4.023566796392313e-06,
+      "loss": 0.2887,
+      "step": 1256
+    },
+    {
+      "epoch": 1.4679608816231207,
+      "grad_norm": 0.08363369385669653,
+      "learning_rate": 4.007223334886531e-06,
+      "loss": 0.2604,
+      "step": 1257
+    },
+    {
+      "epoch": 1.469128594365786,
+      "grad_norm": 0.09348375523045091,
+      "learning_rate": 3.990904812794835e-06,
+      "loss": 0.2915,
+      "step": 1258
+    },
+    {
+      "epoch": 1.4702963071084514,
+      "grad_norm": 0.0800804281368496,
+      "learning_rate": 3.974611298028037e-06,
+      "loss": 0.2835,
+      "step": 1259
+    },
+    {
+      "epoch": 1.4714640198511166,
+      "grad_norm": 0.0786753548544992,
+      "learning_rate": 3.958342858392893e-06,
+      "loss": 0.2745,
+      "step": 1260
+    },
+    {
+      "epoch": 1.4726317325937819,
+      "grad_norm": 0.07620199092031878,
+      "learning_rate": 3.942099561591802e-06,
+      "loss": 0.2614,
+      "step": 1261
+    },
+    {
+      "epoch": 1.4737994453364474,
+      "grad_norm": 0.08058845902465511,
+      "learning_rate": 3.925881475222528e-06,
+      "loss": 0.2568,
+      "step": 1262
+    },
+    {
+      "epoch": 1.4749671580791126,
+      "grad_norm": 0.07590446077057654,
+      "learning_rate": 3.909688666777914e-06,
+      "loss": 0.2544,
+      "step": 1263
+    },
+    {
+      "epoch": 1.4761348708217779,
+      "grad_norm": 0.08027425644154586,
+      "learning_rate": 3.893521203645618e-06,
+      "loss": 0.2688,
+      "step": 1264
+    },
+    {
+      "epoch": 1.477302583564443,
+      "grad_norm": 0.08056329154986873,
+      "learning_rate": 3.877379153107821e-06,
+      "loss": 0.281,
+      "step": 1265
+    },
+    {
+      "epoch": 1.4784702963071084,
+      "grad_norm": 0.07826607641945059,
+      "learning_rate": 3.8612625823409365e-06,
+      "loss": 0.2639,
+      "step": 1266
+    },
+    {
+      "epoch": 1.4796380090497738,
+      "grad_norm": 0.08408757457091125,
+      "learning_rate": 3.845171558415348e-06,
+      "loss": 0.257,
+      "step": 1267
+    },
+    {
+      "epoch": 1.480805721792439,
+      "grad_norm": 0.08272146443456467,
+      "learning_rate": 3.829106148295127e-06,
+      "loss": 0.2702,
+      "step": 1268
+    },
+    {
+      "epoch": 1.4819734345351043,
+      "grad_norm": 0.07836314520153707,
+      "learning_rate": 3.8130664188377464e-06,
+      "loss": 0.2623,
+      "step": 1269
+    },
+    {
+      "epoch": 1.4831411472777698,
+      "grad_norm": 0.07750146884465708,
+      "learning_rate": 3.7970524367938146e-06,
+      "loss": 0.2604,
+      "step": 1270
+    },
+    {
+      "epoch": 1.484308860020435,
+      "grad_norm": 0.08291633846365706,
+      "learning_rate": 3.78106426880678e-06,
+      "loss": 0.2664,
+      "step": 1271
+    },
+    {
+      "epoch": 1.4854765727631003,
+      "grad_norm": 0.08140350589772292,
+      "learning_rate": 3.7651019814126656e-06,
+      "loss": 0.2712,
+      "step": 1272
+    },
+    {
+      "epoch": 1.4866442855057656,
+      "grad_norm": 0.085658016847854,
+      "learning_rate": 3.749165641039796e-06,
+      "loss": 0.2693,
+      "step": 1273
+    },
+    {
+      "epoch": 1.4878119982484308,
+      "grad_norm": 0.07275168163064365,
+      "learning_rate": 3.733255314008516e-06,
+      "loss": 0.2401,
+      "step": 1274
+    },
+    {
+      "epoch": 1.4889797109910963,
+      "grad_norm": 0.07867285260601095,
+      "learning_rate": 3.7173710665309017e-06,
+      "loss": 0.2889,
+      "step": 1275
+    },
+    {
+      "epoch": 1.4901474237337615,
+      "grad_norm": 0.07460707356900328,
+      "learning_rate": 3.701512964710513e-06,
+      "loss": 0.2413,
+      "step": 1276
+    },
+    {
+      "epoch": 1.4913151364764268,
+      "grad_norm": 0.08382215348918909,
+      "learning_rate": 3.685681074542089e-06,
+      "loss": 0.2482,
+      "step": 1277
+    },
+    {
+      "epoch": 1.492482849219092,
+      "grad_norm": 0.08679173284648142,
+      "learning_rate": 3.6698754619112974e-06,
+      "loss": 0.2751,
+      "step": 1278
+    },
+    {
+      "epoch": 1.4936505619617573,
+      "grad_norm": 0.07745258891975401,
+      "learning_rate": 3.654096192594442e-06,
+      "loss": 0.2557,
+      "step": 1279
+    },
+    {
+      "epoch": 1.4948182747044227,
+      "grad_norm": 0.07113041452910802,
+      "learning_rate": 3.6383433322582028e-06,
+      "loss": 0.2407,
+      "step": 1280
+    },
+    {
+      "epoch": 1.495985987447088,
+      "grad_norm": 0.07822743609506634,
+      "learning_rate": 3.62261694645935e-06,
+      "loss": 0.2577,
+      "step": 1281
+    },
+    {
+      "epoch": 1.4971537001897532,
+      "grad_norm": 0.08125347274331288,
+      "learning_rate": 3.6069171006444882e-06,
+      "loss": 0.2639,
+      "step": 1282
+    },
+    {
+      "epoch": 1.4983214129324187,
+      "grad_norm": 0.08238892170609849,
+      "learning_rate": 3.591243860149759e-06,
+      "loss": 0.2515,
+      "step": 1283
+    },
+    {
+      "epoch": 1.499489125675084,
+      "grad_norm": 0.0840317835423618,
+      "learning_rate": 3.575597290200599e-06,
+      "loss": 0.2693,
+      "step": 1284
+    },
+    {
+      "epoch": 1.5006568384177492,
+      "grad_norm": 0.07640809297334533,
+      "learning_rate": 3.5599774559114475e-06,
+      "loss": 0.2559,
+      "step": 1285
+    },
+    {
+      "epoch": 1.5018245511604147,
+      "grad_norm": 0.0746326260041171,
+      "learning_rate": 3.544384422285477e-06,
+      "loss": 0.2567,
+      "step": 1286
+    },
+    {
+      "epoch": 1.5029922639030797,
+      "grad_norm": 0.0775960099688879,
+      "learning_rate": 3.528818254214329e-06,
+      "loss": 0.2632,
+      "step": 1287
+    },
+    {
+      "epoch": 1.5041599766457452,
+      "grad_norm": 0.08505098675011394,
+      "learning_rate": 3.513279016477844e-06,
+      "loss": 0.2833,
+      "step": 1288
+    },
+    {
+      "epoch": 1.5053276893884104,
+      "grad_norm": 0.0797442174034154,
+      "learning_rate": 3.4977667737437905e-06,
+      "loss": 0.2562,
+      "step": 1289
+    },
+    {
+      "epoch": 1.5064954021310757,
+      "grad_norm": 0.07871076571611205,
+      "learning_rate": 3.4822815905675955e-06,
+      "loss": 0.2632,
+      "step": 1290
+    },
+    {
+      "epoch": 1.5076631148737412,
+      "grad_norm": 0.07808429085184933,
+      "learning_rate": 3.4668235313920707e-06,
+      "loss": 0.2687,
+      "step": 1291
+    },
+    {
+      "epoch": 1.5088308276164064,
+      "grad_norm": 0.07837093471048523,
+      "learning_rate": 3.4513926605471504e-06,
+      "loss": 0.266,
+      "step": 1292
+    },
+    {
+      "epoch": 1.5099985403590717,
+      "grad_norm": 0.0749420142417114,
+      "learning_rate": 3.4359890422496245e-06,
+      "loss": 0.2563,
+      "step": 1293
+    },
+    {
+      "epoch": 1.5111662531017371,
+      "grad_norm": 0.0809012592981458,
+      "learning_rate": 3.4206127406028744e-06,
+      "loss": 0.2626,
+      "step": 1294
+    },
+    {
+      "epoch": 1.5123339658444022,
+      "grad_norm": 0.08161682288997191,
+      "learning_rate": 3.4052638195965924e-06,
+      "loss": 0.2629,
+      "step": 1295
+    },
+    {
+      "epoch": 1.5135016785870676,
+      "grad_norm": 0.07690937445744435,
+      "learning_rate": 3.3899423431065215e-06,
+      "loss": 0.2611,
+      "step": 1296
+    },
+    {
+      "epoch": 1.5146693913297329,
+      "grad_norm": 0.07943997394608077,
+      "learning_rate": 3.374648374894204e-06,
+      "loss": 0.2601,
+      "step": 1297
+    },
+    {
+      "epoch": 1.5158371040723981,
+      "grad_norm": 0.08572967469813945,
+      "learning_rate": 3.3593819786067005e-06,
+      "loss": 0.2738,
+      "step": 1298
+    },
+    {
+      "epoch": 1.5170048168150636,
+      "grad_norm": 0.07335110407034547,
+      "learning_rate": 3.344143217776319e-06,
+      "loss": 0.254,
+      "step": 1299
+    },
+    {
+      "epoch": 1.5181725295577286,
+      "grad_norm": 0.07714360505391435,
+      "learning_rate": 3.3289321558203767e-06,
+      "loss": 0.265,
+      "step": 1300
+    },
+    {
+      "epoch": 1.519340242300394,
+      "grad_norm": 0.08324755920931051,
+      "learning_rate": 3.3137488560409027e-06,
+      "loss": 0.2683,
+      "step": 1301
+    },
+    {
+      "epoch": 1.5205079550430594,
+      "grad_norm": 0.07862678727410853,
+      "learning_rate": 3.2985933816244065e-06,
+      "loss": 0.2538,
+      "step": 1302
+    },
+    {
+      "epoch": 1.5216756677857246,
+      "grad_norm": 0.07927134663218857,
+      "learning_rate": 3.283465795641588e-06,
+      "loss": 0.2696,
+      "step": 1303
+    },
+    {
+      "epoch": 1.52284338052839,
+      "grad_norm": 0.07992711628857654,
+      "learning_rate": 3.268366161047096e-06,
+      "loss": 0.2557,
+      "step": 1304
+    },
+    {
+      "epoch": 1.5240110932710553,
+      "grad_norm": 0.07855447706284988,
+      "learning_rate": 3.2532945406792573e-06,
+      "loss": 0.2538,
+      "step": 1305
+    },
+    {
+      "epoch": 1.5251788060137206,
+      "grad_norm": 0.07681470175527068,
+      "learning_rate": 3.2382509972598087e-06,
+      "loss": 0.2525,
+      "step": 1306
+    },
+    {
+      "epoch": 1.526346518756386,
+      "grad_norm": 0.08041659728177199,
+      "learning_rate": 3.223235593393642e-06,
+      "loss": 0.2588,
+      "step": 1307
+    },
+    {
+      "epoch": 1.527514231499051,
+      "grad_norm": 0.08815748432640158,
+      "learning_rate": 3.2082483915685526e-06,
+      "loss": 0.2921,
+      "step": 1308
+    },
+    {
+      "epoch": 1.5286819442417166,
+      "grad_norm": 0.07905695791408421,
+      "learning_rate": 3.1932894541549643e-06,
+      "loss": 0.2717,
+      "step": 1309
+    },
+    {
+      "epoch": 1.5298496569843818,
+      "grad_norm": 0.08010188364759564,
+      "learning_rate": 3.1783588434056845e-06,
+      "loss": 0.2708,
+      "step": 1310
+    },
+    {
+      "epoch": 1.531017369727047,
+      "grad_norm": 0.07690225602490307,
+      "learning_rate": 3.163456621455625e-06,
+      "loss": 0.2591,
+      "step": 1311
+    },
+    {
+      "epoch": 1.5321850824697125,
+      "grad_norm": 0.0774141016874316,
+      "learning_rate": 3.1485828503215588e-06,
+      "loss": 0.274,
+      "step": 1312
+    },
+    {
+      "epoch": 1.5333527952123778,
+      "grad_norm": 0.07442339127653769,
+      "learning_rate": 3.133737591901864e-06,
+      "loss": 0.2514,
+      "step": 1313
+    },
+    {
+      "epoch": 1.534520507955043,
+      "grad_norm": 0.07903260578712436,
+      "learning_rate": 3.1189209079762606e-06,
+      "loss": 0.2689,
+      "step": 1314
+    },
+    {
+      "epoch": 1.5356882206977085,
+      "grad_norm": 0.07987133165506831,
+      "learning_rate": 3.104132860205549e-06,
+      "loss": 0.2818,
+      "step": 1315
+    },
+    {
+      "epoch": 1.5368559334403735,
+      "grad_norm": 0.07744923383905995,
+      "learning_rate": 3.089373510131354e-06,
+      "loss": 0.2645,
+      "step": 1316
+    },
+    {
+      "epoch": 1.538023646183039,
+      "grad_norm": 0.08007946492774028,
+      "learning_rate": 3.074642919175883e-06,
+      "loss": 0.2704,
+      "step": 1317
+    },
+    {
+      "epoch": 1.5391913589257042,
+      "grad_norm": 0.08453861351447903,
+      "learning_rate": 3.0599411486416587e-06,
+      "loss": 0.2867,
+      "step": 1318
+    },
+    {
+      "epoch": 1.5403590716683695,
+      "grad_norm": 0.07919138155487893,
+      "learning_rate": 3.045268259711256e-06,
+      "loss": 0.2593,
+      "step": 1319
+    },
+    {
+      "epoch": 1.541526784411035,
+      "grad_norm": 0.08149806894422167,
+      "learning_rate": 3.0306243134470668e-06,
+      "loss": 0.2802,
+      "step": 1320
+    },
+    {
+      "epoch": 1.5426944971537002,
+      "grad_norm": 0.0785173147947615,
+      "learning_rate": 3.0160093707910275e-06,
+      "loss": 0.2569,
+      "step": 1321
+    },
+    {
+      "epoch": 1.5438622098963655,
+      "grad_norm": 0.07867724491836121,
+      "learning_rate": 3.001423492564384e-06,
+      "loss": 0.2517,
+      "step": 1322
+    },
+    {
+      "epoch": 1.545029922639031,
+      "grad_norm": 0.07990539632115672,
+      "learning_rate": 2.9868667394674143e-06,
+      "loss": 0.2713,
+      "step": 1323
+    },
+    {
+      "epoch": 1.546197635381696,
+      "grad_norm": 0.07869574456453977,
+      "learning_rate": 2.972339172079204e-06,
+      "loss": 0.2772,
+      "step": 1324
+    },
+    {
+      "epoch": 1.5473653481243614,
+      "grad_norm": 0.07785010001846468,
+      "learning_rate": 2.957840850857373e-06,
+      "loss": 0.2501,
+      "step": 1325
+    },
+    {
+      "epoch": 1.5485330608670267,
+      "grad_norm": 0.08081102160255414,
+      "learning_rate": 2.9433718361378326e-06,
+      "loss": 0.2668,
+      "step": 1326
+    },
+    {
+      "epoch": 1.549700773609692,
+      "grad_norm": 0.08230556652983366,
+      "learning_rate": 2.9289321881345257e-06,
+      "loss": 0.2661,
+      "step": 1327
+    },
+    {
+      "epoch": 1.5508684863523574,
+      "grad_norm": 0.07562744585638968,
+      "learning_rate": 2.9145219669391944e-06,
+      "loss": 0.2435,
+      "step": 1328
+    },
+    {
+      "epoch": 1.5520361990950227,
+      "grad_norm": 0.07420220092980435,
+      "learning_rate": 2.9001412325211163e-06,
+      "loss": 0.2447,
+      "step": 1329
+    },
+    {
+      "epoch": 1.553203911837688,
+      "grad_norm": 0.07897561503597733,
+      "learning_rate": 2.885790044726853e-06,
+      "loss": 0.2677,
+      "step": 1330
+    },
+    {
+      "epoch": 1.5543716245803534,
+      "grad_norm": 0.07525804263994605,
+      "learning_rate": 2.871468463280003e-06,
+      "loss": 0.2478,
+      "step": 1331
+    },
+    {
+      "epoch": 1.5555393373230184,
+      "grad_norm": 0.07889563738055212,
+      "learning_rate": 2.8571765477809645e-06,
+      "loss": 0.247,
+      "step": 1332
+    },
+    {
+      "epoch": 1.5567070500656839,
+      "grad_norm": 0.07818313876127692,
+      "learning_rate": 2.8429143577066764e-06,
+      "loss": 0.2628,
+      "step": 1333
+    },
+    {
+      "epoch": 1.5578747628083491,
+      "grad_norm": 0.08184578464727632,
+      "learning_rate": 2.8286819524103657e-06,
+      "loss": 0.2751,
+      "step": 1334
+    },
+    {
+      "epoch": 1.5590424755510144,
+      "grad_norm": 0.08272476819971665,
+      "learning_rate": 2.8144793911213174e-06,
+      "loss": 0.2885,
+      "step": 1335
+    },
+    {
+      "epoch": 1.5602101882936799,
+      "grad_norm": 0.08102574319178926,
+      "learning_rate": 2.80030673294461e-06,
+      "loss": 0.2797,
+      "step": 1336
+    },
+    {
+      "epoch": 1.561377901036345,
+      "grad_norm": 0.08182694109159067,
+      "learning_rate": 2.7861640368608846e-06,
+      "loss": 0.2678,
+      "step": 1337
+    },
+    {
+      "epoch": 1.5625456137790104,
+      "grad_norm": 0.08135291674165994,
+      "learning_rate": 2.7720513617260857e-06,
+      "loss": 0.28,
+      "step": 1338
+    },
+    {
+      "epoch": 1.5637133265216758,
+      "grad_norm": 0.0762395403609643,
+      "learning_rate": 2.7579687662712285e-06,
+      "loss": 0.2468,
+      "step": 1339
+    },
+    {
+      "epoch": 1.5648810392643409,
+      "grad_norm": 0.07501806743888521,
+      "learning_rate": 2.7439163091021525e-06,
+      "loss": 0.2691,
+      "step": 1340
+    },
+    {
+      "epoch": 1.5660487520070063,
+      "grad_norm": 0.08099641476171478,
+      "learning_rate": 2.7298940486992654e-06,
+      "loss": 0.27,
+      "step": 1341
+    },
+    {
+      "epoch": 1.5672164647496716,
+      "grad_norm": 0.07426058727609583,
+      "learning_rate": 2.71590204341731e-06,
+      "loss": 0.2512,
+      "step": 1342
+    },
+    {
+      "epoch": 1.5683841774923368,
+      "grad_norm": 0.07897203108851987,
+      "learning_rate": 2.7019403514851274e-06,
+      "loss": 0.2576,
+      "step": 1343
+    },
+    {
+      "epoch": 1.5695518902350023,
+      "grad_norm": 0.0780157729738713,
+      "learning_rate": 2.688009031005403e-06,
+      "loss": 0.2788,
+      "step": 1344
+    },
+    {
+      "epoch": 1.5707196029776673,
+      "grad_norm": 0.0786597461216064,
+      "learning_rate": 2.674108139954429e-06,
+      "loss": 0.2745,
+      "step": 1345
+    },
+    {
+      "epoch": 1.5718873157203328,
+      "grad_norm": 0.08058117553455028,
+      "learning_rate": 2.6602377361818576e-06,
+      "loss": 0.2731,
+      "step": 1346
+    },
+    {
+      "epoch": 1.573055028462998,
+      "grad_norm": 0.08173134179524645,
+      "learning_rate": 2.646397877410476e-06,
+      "loss": 0.267,
+      "step": 1347
+    },
+    {
+      "epoch": 1.5742227412056633,
+      "grad_norm": 0.07863184501504081,
+      "learning_rate": 2.6325886212359496e-06,
+      "loss": 0.267,
+      "step": 1348
+    },
+    {
+      "epoch": 1.5753904539483288,
+      "grad_norm": 0.08054997733220864,
+      "learning_rate": 2.6188100251265947e-06,
+      "loss": 0.2691,
+      "step": 1349
+    },
+    {
+      "epoch": 1.576558166690994,
+      "grad_norm": 0.0747444303729358,
+      "learning_rate": 2.6050621464231243e-06,
+      "loss": 0.2657,
+      "step": 1350
+    },
+    {
+      "epoch": 1.5777258794336593,
+      "grad_norm": 0.0803647974003624,
+      "learning_rate": 2.591345042338421e-06,
+      "loss": 0.2645,
+      "step": 1351
+    },
+    {
+      "epoch": 1.5788935921763247,
+      "grad_norm": 0.07952710444690898,
+      "learning_rate": 2.5776587699573007e-06,
+      "loss": 0.2819,
+      "step": 1352
+    },
+    {
+      "epoch": 1.5800613049189898,
+      "grad_norm": 0.07930465054090932,
+      "learning_rate": 2.564003386236271e-06,
+      "loss": 0.2715,
+      "step": 1353
+    },
+    {
+      "epoch": 1.5812290176616552,
+      "grad_norm": 0.08010503425088382,
+      "learning_rate": 2.550378948003287e-06,
+      "loss": 0.2472,
+      "step": 1354
+    },
+    {
+      "epoch": 1.5823967304043205,
+      "grad_norm": 0.07483815628645234,
+      "learning_rate": 2.5367855119575314e-06,
+      "loss": 0.2616,
+      "step": 1355
+    },
+    {
+      "epoch": 1.5835644431469857,
+      "grad_norm": 0.08231485501739232,
+      "learning_rate": 2.523223134669157e-06,
+      "loss": 0.2944,
+      "step": 1356
+    },
+    {
+      "epoch": 1.5847321558896512,
+      "grad_norm": 0.07727655056251498,
+      "learning_rate": 2.5096918725790754e-06,
+      "loss": 0.2592,
+      "step": 1357
+    },
+    {
+      "epoch": 1.5858998686323165,
+      "grad_norm": 0.08415670820761577,
+      "learning_rate": 2.496191781998698e-06,
+      "loss": 0.2842,
+      "step": 1358
+    },
+    {
+      "epoch": 1.5870675813749817,
+      "grad_norm": 0.07415868894810655,
+      "learning_rate": 2.4827229191097236e-06,
+      "loss": 0.2371,
+      "step": 1359
+    },
+    {
+      "epoch": 1.5882352941176472,
+      "grad_norm": 0.07805833499088902,
+      "learning_rate": 2.469285339963892e-06,
+      "loss": 0.2409,
+      "step": 1360
+    },
+    {
+      "epoch": 1.5894030068603122,
+      "grad_norm": 0.0776157206705961,
+      "learning_rate": 2.4558791004827487e-06,
+      "loss": 0.2543,
+      "step": 1361
+    },
+    {
+      "epoch": 1.5905707196029777,
+      "grad_norm": 0.0771207940505542,
+      "learning_rate": 2.4425042564574186e-06,
+      "loss": 0.2692,
+      "step": 1362
+    },
+    {
+      "epoch": 1.591738432345643,
+      "grad_norm": 0.07991308578982041,
+      "learning_rate": 2.4291608635483733e-06,
+      "loss": 0.2855,
+      "step": 1363
+    },
+    {
+      "epoch": 1.5929061450883082,
+      "grad_norm": 0.07862542591242891,
+      "learning_rate": 2.4158489772852035e-06,
+      "loss": 0.2606,
+      "step": 1364
+    },
+    {
+      "epoch": 1.5940738578309737,
+      "grad_norm": 0.08060954509300897,
+      "learning_rate": 2.4025686530663716e-06,
+      "loss": 0.2714,
+      "step": 1365
+    },
+    {
+      "epoch": 1.595241570573639,
+      "grad_norm": 0.0819829749310937,
+      "learning_rate": 2.3893199461589943e-06,
+      "loss": 0.2837,
+      "step": 1366
+    },
+    {
+      "epoch": 1.5964092833163042,
+      "grad_norm": 0.07979124661313144,
+      "learning_rate": 2.376102911698618e-06,
+      "loss": 0.2775,
+      "step": 1367
+    },
+    {
+      "epoch": 1.5975769960589696,
+      "grad_norm": 0.08049777773562532,
+      "learning_rate": 2.3629176046889755e-06,
+      "loss": 0.2656,
+      "step": 1368
+    },
+    {
+      "epoch": 1.5987447088016347,
+      "grad_norm": 0.08088200557882774,
+      "learning_rate": 2.3497640800017687e-06,
+      "loss": 0.2698,
+      "step": 1369
+    },
+    {
+      "epoch": 1.5999124215443001,
+      "grad_norm": 0.07455936891735196,
+      "learning_rate": 2.3366423923764268e-06,
+      "loss": 0.2541,
+      "step": 1370
+    },
+    {
+      "epoch": 1.6010801342869654,
+      "grad_norm": 0.08229639524874141,
+      "learning_rate": 2.323552596419889e-06,
+      "loss": 0.2718,
+      "step": 1371
+    },
+    {
+      "epoch": 1.6022478470296306,
+      "grad_norm": 0.07881482004760194,
+      "learning_rate": 2.3104947466063785e-06,
+      "loss": 0.2644,
+      "step": 1372
+    },
+    {
+      "epoch": 1.603415559772296,
+      "grad_norm": 0.07382070855009282,
+      "learning_rate": 2.2974688972771743e-06,
+      "loss": 0.2504,
+      "step": 1373
+    },
+    {
+      "epoch": 1.6045832725149614,
+      "grad_norm": 0.08110745909600775,
+      "learning_rate": 2.284475102640371e-06,
+      "loss": 0.2561,
+      "step": 1374
+    },
+    {
+      "epoch": 1.6057509852576266,
+      "grad_norm": 0.07594890167834709,
+      "learning_rate": 2.27151341677068e-06,
+      "loss": 0.2454,
+      "step": 1375
+    },
+    {
+      "epoch": 1.606918698000292,
+      "grad_norm": 0.07262955354322789,
+      "learning_rate": 2.2585838936091753e-06,
+      "loss": 0.239,
+      "step": 1376
+    },
+    {
+      "epoch": 1.608086410742957,
+      "grad_norm": 0.07732855678256541,
+      "learning_rate": 2.2456865869630974e-06,
+      "loss": 0.2535,
+      "step": 1377
+    },
+    {
+      "epoch": 1.6092541234856226,
+      "grad_norm": 0.0780029184133591,
+      "learning_rate": 2.2328215505056004e-06,
+      "loss": 0.2566,
+      "step": 1378
+    },
+    {
+      "epoch": 1.6104218362282878,
+      "grad_norm": 0.08168552097824294,
+      "learning_rate": 2.2199888377755573e-06,
+      "loss": 0.2596,
+      "step": 1379
+    },
+    {
+      "epoch": 1.611589548970953,
+      "grad_norm": 0.0814986691901058,
+      "learning_rate": 2.207188502177313e-06,
+      "loss": 0.2574,
+      "step": 1380
+    },
+    {
+      "epoch": 1.6127572617136186,
+      "grad_norm": 0.08083684255176249,
+      "learning_rate": 2.194420596980481e-06,
+      "loss": 0.2644,
+      "step": 1381
+    },
+    {
+      "epoch": 1.6139249744562838,
+      "grad_norm": 0.08209395777960729,
+      "learning_rate": 2.1816851753197023e-06,
+      "loss": 0.265,
+      "step": 1382
+    },
+    {
+      "epoch": 1.615092687198949,
+      "grad_norm": 0.07776153149559802,
+      "learning_rate": 2.1689822901944456e-06,
+      "loss": 0.2564,
+      "step": 1383
+    },
+    {
+      "epoch": 1.6162603999416145,
+      "grad_norm": 0.07896565918769227,
+      "learning_rate": 2.156311994468774e-06,
+      "loss": 0.2537,
+      "step": 1384
+    },
+    {
+      "epoch": 1.6174281126842796,
+      "grad_norm": 0.07855682967240248,
+      "learning_rate": 2.1436743408711234e-06,
+      "loss": 0.2443,
+      "step": 1385
+    },
+    {
+      "epoch": 1.618595825426945,
+      "grad_norm": 0.07909889888605495,
+      "learning_rate": 2.1310693819940842e-06,
+      "loss": 0.251,
+      "step": 1386
+    },
+    {
+      "epoch": 1.6197635381696103,
+      "grad_norm": 0.0807148608411226,
+      "learning_rate": 2.1184971702941947e-06,
+      "loss": 0.2512,
+      "step": 1387
+    },
+    {
+      "epoch": 1.6209312509122755,
+      "grad_norm": 0.08279719243609422,
+      "learning_rate": 2.1059577580917067e-06,
+      "loss": 0.2679,
+      "step": 1388
+    },
+    {
+      "epoch": 1.622098963654941,
+      "grad_norm": 0.08216405298799125,
+      "learning_rate": 2.0934511975703795e-06,
+      "loss": 0.2655,
+      "step": 1389
+    },
+    {
+      "epoch": 1.623266676397606,
+      "grad_norm": 0.07786907423957151,
+      "learning_rate": 2.0809775407772505e-06,
+      "loss": 0.2511,
+      "step": 1390
+    },
+    {
+      "epoch": 1.6244343891402715,
+      "grad_norm": 0.07764456122845351,
+      "learning_rate": 2.068536839622427e-06,
+      "loss": 0.2472,
+      "step": 1391
+    },
+    {
+      "epoch": 1.6256021018829367,
+      "grad_norm": 0.0814837229805338,
+      "learning_rate": 2.0561291458788736e-06,
+      "loss": 0.2824,
+      "step": 1392
+    },
+    {
+      "epoch": 1.626769814625602,
+      "grad_norm": 0.08124288741191887,
+      "learning_rate": 2.043754511182191e-06,
+      "loss": 0.2545,
+      "step": 1393
+    },
+    {
+      "epoch": 1.6279375273682675,
+      "grad_norm": 0.07797873924081514,
+      "learning_rate": 2.031412987030398e-06,
+      "loss": 0.2648,
+      "step": 1394
+    },
+    {
+      "epoch": 1.6291052401109327,
+      "grad_norm": 0.07749694380415467,
+      "learning_rate": 2.01910462478372e-06,
+      "loss": 0.2573,
+      "step": 1395
+    },
+    {
+      "epoch": 1.630272952853598,
+      "grad_norm": 0.0801337993803082,
+      "learning_rate": 2.0068294756643846e-06,
+      "loss": 0.2527,
+      "step": 1396
+    },
+    {
+      "epoch": 1.6314406655962634,
+      "grad_norm": 0.08167650466644079,
+      "learning_rate": 1.994587590756397e-06,
+      "loss": 0.2781,
+      "step": 1397
+    },
+    {
+      "epoch": 1.6326083783389285,
+      "grad_norm": 0.08026650883304005,
+      "learning_rate": 1.982379021005325e-06,
+      "loss": 0.2621,
+      "step": 1398
+    },
+    {
+      "epoch": 1.633776091081594,
+      "grad_norm": 0.07724038965837339,
+      "learning_rate": 1.9702038172181016e-06,
+      "loss": 0.2492,
+      "step": 1399
+    },
+    {
+      "epoch": 1.6349438038242592,
+      "grad_norm": 0.07951981309037479,
+      "learning_rate": 1.958062030062795e-06,
+      "loss": 0.2711,
+      "step": 1400
+    },
+    {
+      "epoch": 1.6361115165669244,
+      "grad_norm": 0.0789666172057193,
+      "learning_rate": 1.94595371006842e-06,
+      "loss": 0.2554,
+      "step": 1401
+    },
+    {
+      "epoch": 1.63727922930959,
+      "grad_norm": 0.07942147664905624,
+      "learning_rate": 1.9338789076247e-06,
+      "loss": 0.2543,
+      "step": 1402
+    },
+    {
+      "epoch": 1.6384469420522552,
+      "grad_norm": 0.07796848501256445,
+      "learning_rate": 1.9218376729818842e-06,
+      "loss": 0.2591,
+      "step": 1403
+    },
+    {
+      "epoch": 1.6396146547949204,
+      "grad_norm": 0.07995059124566346,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.2547,
+      "step": 1404
+    },
+    {
+      "epoch": 1.6407823675375859,
+      "grad_norm": 0.08175728687348538,
+      "learning_rate": 1.8978561074012702e-06,
+      "loss": 0.2652,
+      "step": 1405
+    },
+    {
+      "epoch": 1.641950080280251,
+      "grad_norm": 0.08786850774524364,
+      "learning_rate": 1.8859158762646468e-06,
+      "loss": 0.2827,
+      "step": 1406
+    },
+    {
+      "epoch": 1.6431177930229164,
+      "grad_norm": 0.08815404435664587,
+      "learning_rate": 1.8740094125308773e-06,
+      "loss": 0.2703,
+      "step": 1407
+    },
+    {
+      "epoch": 1.6442855057655816,
+      "grad_norm": 0.0780110840740333,
+      "learning_rate": 1.8621367657496504e-06,
+      "loss": 0.2557,
+      "step": 1408
+    },
+    {
+      "epoch": 1.6454532185082469,
+      "grad_norm": 0.07675350242547717,
+      "learning_rate": 1.850297985329932e-06,
+      "loss": 0.2618,
+      "step": 1409
+    },
+    {
+      "epoch": 1.6466209312509124,
+      "grad_norm": 0.08177865400902078,
+      "learning_rate": 1.8384931205397305e-06,
+      "loss": 0.2697,
+      "step": 1410
+    },
+    {
+      "epoch": 1.6477886439935776,
+      "grad_norm": 0.07794440699768336,
+      "learning_rate": 1.826722220505931e-06,
+      "loss": 0.2397,
+      "step": 1411
+    },
+    {
+      "epoch": 1.6489563567362429,
+      "grad_norm": 0.07822615678116109,
+      "learning_rate": 1.8149853342140644e-06,
+      "loss": 0.2748,
+      "step": 1412
+    },
+    {
+      "epoch": 1.6501240694789083,
+      "grad_norm": 0.07571536600586848,
+      "learning_rate": 1.8032825105081141e-06,
+      "loss": 0.2498,
+      "step": 1413
+    },
+    {
+      "epoch": 1.6512917822215734,
+      "grad_norm": 0.07648584432097491,
+      "learning_rate": 1.7916137980903047e-06,
+      "loss": 0.2515,
+      "step": 1414
+    },
+    {
+      "epoch": 1.6524594949642388,
+      "grad_norm": 0.07648327237103313,
+      "learning_rate": 1.7799792455209019e-06,
+      "loss": 0.2571,
+      "step": 1415
+    },
+    {
+      "epoch": 1.653627207706904,
+      "grad_norm": 0.07621987554822612,
+      "learning_rate": 1.7683789012180196e-06,
+      "loss": 0.2606,
+      "step": 1416
+    },
+    {
+      "epoch": 1.6547949204495693,
+      "grad_norm": 0.0823442379523928,
+      "learning_rate": 1.7568128134574114e-06,
+      "loss": 0.2833,
+      "step": 1417
+    },
+    {
+      "epoch": 1.6559626331922348,
+      "grad_norm": 0.08284025896933331,
+      "learning_rate": 1.74528103037226e-06,
+      "loss": 0.2637,
+      "step": 1418
+    },
+    {
+      "epoch": 1.6571303459349,
+      "grad_norm": 0.0804996711085964,
+      "learning_rate": 1.7337835999529995e-06,
+      "loss": 0.2544,
+      "step": 1419
+    },
+    {
+      "epoch": 1.6582980586775653,
+      "grad_norm": 0.08173260565541558,
+      "learning_rate": 1.722320570047089e-06,
+      "loss": 0.2797,
+      "step": 1420
+    },
+    {
+      "epoch": 1.6594657714202308,
+      "grad_norm": 0.07927287093655697,
+      "learning_rate": 1.7108919883588404e-06,
+      "loss": 0.2629,
+      "step": 1421
+    },
+    {
+      "epoch": 1.6606334841628958,
+      "grad_norm": 0.08110951360486841,
+      "learning_rate": 1.6994979024491943e-06,
+      "loss": 0.2723,
+      "step": 1422
+    },
+    {
+      "epoch": 1.6618011969055613,
+      "grad_norm": 0.0867066293658335,
+      "learning_rate": 1.6881383597355438e-06,
+      "loss": 0.2906,
+      "step": 1423
+    },
+    {
+      "epoch": 1.6629689096482265,
+      "grad_norm": 0.0800296653988266,
+      "learning_rate": 1.6768134074915277e-06,
+      "loss": 0.2662,
+      "step": 1424
+    },
+    {
+      "epoch": 1.6641366223908918,
+      "grad_norm": 0.08002471583318056,
+      "learning_rate": 1.6655230928468257e-06,
+      "loss": 0.2636,
+      "step": 1425
+    },
+    {
+      "epoch": 1.6653043351335572,
+      "grad_norm": 0.08099337863966825,
+      "learning_rate": 1.6542674627869738e-06,
+      "loss": 0.2743,
+      "step": 1426
+    },
+    {
+      "epoch": 1.6664720478762225,
+      "grad_norm": 0.07924255406093328,
+      "learning_rate": 1.6430465641531701e-06,
+      "loss": 0.2469,
+      "step": 1427
+    },
+    {
+      "epoch": 1.6676397606188877,
+      "grad_norm": 0.07634508098142323,
+      "learning_rate": 1.6318604436420738e-06,
+      "loss": 0.2497,
+      "step": 1428
+    },
+    {
+      "epoch": 1.6688074733615532,
+      "grad_norm": 0.07754813918789646,
+      "learning_rate": 1.6207091478056059e-06,
+      "loss": 0.2614,
+      "step": 1429
+    },
+    {
+      "epoch": 1.6699751861042182,
+      "grad_norm": 0.07625132059114992,
+      "learning_rate": 1.609592723050767e-06,
+      "loss": 0.2477,
+      "step": 1430
+    },
+    {
+      "epoch": 1.6711428988468837,
+      "grad_norm": 0.07895406471733508,
+      "learning_rate": 1.5985112156394368e-06,
+      "loss": 0.2602,
+      "step": 1431
+    },
+    {
+      "epoch": 1.672310611589549,
+      "grad_norm": 0.08195562179077556,
+      "learning_rate": 1.587464671688187e-06,
+      "loss": 0.2695,
+      "step": 1432
+    },
+    {
+      "epoch": 1.6734783243322142,
+      "grad_norm": 0.07517102028477157,
+      "learning_rate": 1.5764531371680859e-06,
+      "loss": 0.2411,
+      "step": 1433
+    },
+    {
+      "epoch": 1.6746460370748797,
+      "grad_norm": 0.07906277594438223,
+      "learning_rate": 1.5654766579045033e-06,
+      "loss": 0.2676,
+      "step": 1434
+    },
+    {
+      "epoch": 1.6758137498175447,
+      "grad_norm": 0.08183936477424668,
+      "learning_rate": 1.5545352795769231e-06,
+      "loss": 0.2576,
+      "step": 1435
+    },
+    {
+      "epoch": 1.6769814625602102,
+      "grad_norm": 0.08502350820440116,
+      "learning_rate": 1.5436290477187589e-06,
+      "loss": 0.2754,
+      "step": 1436
+    },
+    {
+      "epoch": 1.6781491753028754,
+      "grad_norm": 0.07789396858917345,
+      "learning_rate": 1.5327580077171589e-06,
+      "loss": 0.2475,
+      "step": 1437
+    },
+    {
+      "epoch": 1.6793168880455407,
+      "grad_norm": 0.0818786223280055,
+      "learning_rate": 1.5219222048128125e-06,
+      "loss": 0.2763,
+      "step": 1438
+    },
+    {
+      "epoch": 1.6804846007882062,
+      "grad_norm": 0.07701279870054979,
+      "learning_rate": 1.5111216840997745e-06,
+      "loss": 0.2435,
+      "step": 1439
+    },
+    {
+      "epoch": 1.6816523135308714,
+      "grad_norm": 0.07459780298038836,
+      "learning_rate": 1.500356490525261e-06,
+      "loss": 0.246,
+      "step": 1440
+    },
+    {
+      "epoch": 1.6828200262735367,
+      "grad_norm": 0.07493427528641244,
+      "learning_rate": 1.4896266688894823e-06,
+      "loss": 0.2412,
+      "step": 1441
+    },
+    {
+      "epoch": 1.6839877390162021,
+      "grad_norm": 0.07886217156192031,
+      "learning_rate": 1.478932263845435e-06,
+      "loss": 0.2637,
+      "step": 1442
+    },
+    {
+      "epoch": 1.6851554517588672,
+      "grad_norm": 0.07560069807073111,
+      "learning_rate": 1.4682733198987353e-06,
+      "loss": 0.2521,
+      "step": 1443
+    },
+    {
+      "epoch": 1.6863231645015326,
+      "grad_norm": 0.07354499391082163,
+      "learning_rate": 1.457649881407417e-06,
+      "loss": 0.2642,
+      "step": 1444
+    },
+    {
+      "epoch": 1.6874908772441979,
+      "grad_norm": 0.07742638216577596,
+      "learning_rate": 1.4470619925817654e-06,
+      "loss": 0.2634,
+      "step": 1445
+    },
+    {
+      "epoch": 1.6886585899868631,
+      "grad_norm": 0.07783969257616558,
+      "learning_rate": 1.436509697484111e-06,
+      "loss": 0.2623,
+      "step": 1446
+    },
+    {
+      "epoch": 1.6898263027295286,
+      "grad_norm": 0.07653812970637325,
+      "learning_rate": 1.425993040028667e-06,
+      "loss": 0.2419,
+      "step": 1447
+    },
+    {
+      "epoch": 1.6909940154721939,
+      "grad_norm": 0.08092146695157554,
+      "learning_rate": 1.4155120639813392e-06,
+      "loss": 0.2721,
+      "step": 1448
+    },
+    {
+      "epoch": 1.692161728214859,
+      "grad_norm": 0.07604653055459228,
+      "learning_rate": 1.4050668129595346e-06,
+      "loss": 0.2462,
+      "step": 1449
+    },
+    {
+      "epoch": 1.6933294409575246,
+      "grad_norm": 0.0799527904158048,
+      "learning_rate": 1.39465733043199e-06,
+      "loss": 0.2634,
+      "step": 1450
+    },
+    {
+      "epoch": 1.6944971537001896,
+      "grad_norm": 0.07599001109728998,
+      "learning_rate": 1.3842836597185916e-06,
+      "loss": 0.2543,
+      "step": 1451
+    },
+    {
+      "epoch": 1.695664866442855,
+      "grad_norm": 0.09713457997090362,
+      "learning_rate": 1.373945843990192e-06,
+      "loss": 0.2677,
+      "step": 1452
+    },
+    {
+      "epoch": 1.6968325791855203,
+      "grad_norm": 0.07983226708833321,
+      "learning_rate": 1.3636439262684299e-06,
+      "loss": 0.2694,
+      "step": 1453
+    },
+    {
+      "epoch": 1.6980002919281856,
+      "grad_norm": 0.08129183917122117,
+      "learning_rate": 1.3533779494255483e-06,
+      "loss": 0.2663,
+      "step": 1454
+    },
+    {
+      "epoch": 1.699168004670851,
+      "grad_norm": 0.07722487241689717,
+      "learning_rate": 1.343147956184221e-06,
+      "loss": 0.2524,
+      "step": 1455
+    },
+    {
+      "epoch": 1.7003357174135163,
+      "grad_norm": 0.08081191697503175,
+      "learning_rate": 1.332953989117377e-06,
+      "loss": 0.2615,
+      "step": 1456
+    },
+    {
+      "epoch": 1.7015034301561816,
+      "grad_norm": 0.07493845234741875,
+      "learning_rate": 1.322796090648013e-06,
+      "loss": 0.2558,
+      "step": 1457
+    },
+    {
+      "epoch": 1.702671142898847,
+      "grad_norm": 0.07951796741060621,
+      "learning_rate": 1.3126743030490307e-06,
+      "loss": 0.2701,
+      "step": 1458
+    },
+    {
+      "epoch": 1.703838855641512,
+      "grad_norm": 0.07989596882274806,
+      "learning_rate": 1.3025886684430467e-06,
+      "loss": 0.2527,
+      "step": 1459
+    },
+    {
+      "epoch": 1.7050065683841775,
+      "grad_norm": 0.07936106901228837,
+      "learning_rate": 1.2925392288022299e-06,
+      "loss": 0.2502,
+      "step": 1460
+    },
+    {
+      "epoch": 1.7061742811268428,
+      "grad_norm": 0.07987683095114405,
+      "learning_rate": 1.2825260259481143e-06,
+      "loss": 0.2671,
+      "step": 1461
+    },
+    {
+      "epoch": 1.707341993869508,
+      "grad_norm": 0.07426042770264062,
+      "learning_rate": 1.272549101551438e-06,
+      "loss": 0.2326,
+      "step": 1462
+    },
+    {
+      "epoch": 1.7085097066121735,
+      "grad_norm": 0.0772549673220044,
+      "learning_rate": 1.2626084971319641e-06,
+      "loss": 0.2596,
+      "step": 1463
+    },
+    {
+      "epoch": 1.7096774193548387,
+      "grad_norm": 0.0785148960981749,
+      "learning_rate": 1.2527042540583e-06,
+      "loss": 0.2647,
+      "step": 1464
+    },
+    {
+      "epoch": 1.710845132097504,
+      "grad_norm": 0.08157378216882825,
+      "learning_rate": 1.2428364135477356e-06,
+      "loss": 0.2789,
+      "step": 1465
+    },
+    {
+      "epoch": 1.7120128448401695,
+      "grad_norm": 0.07798542110755398,
+      "learning_rate": 1.2330050166660712e-06,
+      "loss": 0.2532,
+      "step": 1466
+    },
+    {
+      "epoch": 1.7131805575828345,
+      "grad_norm": 0.0796007025793864,
+      "learning_rate": 1.2232101043274437e-06,
+      "loss": 0.2757,
+      "step": 1467
+    },
+    {
+      "epoch": 1.7143482703255,
+      "grad_norm": 0.07764754894379473,
+      "learning_rate": 1.2134517172941563e-06,
+      "loss": 0.2388,
+      "step": 1468
+    },
+    {
+      "epoch": 1.7155159830681652,
+      "grad_norm": 0.07790331559753141,
+      "learning_rate": 1.2037298961765065e-06,
+      "loss": 0.261,
+      "step": 1469
+    },
+    {
+      "epoch": 1.7166836958108305,
+      "grad_norm": 0.07657350134613017,
+      "learning_rate": 1.19404468143262e-06,
+      "loss": 0.257,
+      "step": 1470
+    },
+    {
+      "epoch": 1.717851408553496,
+      "grad_norm": 0.0759851125378571,
+      "learning_rate": 1.184396113368288e-06,
+      "loss": 0.2596,
+      "step": 1471
+    },
+    {
+      "epoch": 1.7190191212961612,
+      "grad_norm": 0.07823427556998627,
+      "learning_rate": 1.1747842321367886e-06,
+      "loss": 0.2469,
+      "step": 1472
+    },
+    {
+      "epoch": 1.7201868340388264,
+      "grad_norm": 0.08043479390303003,
+      "learning_rate": 1.1652090777387248e-06,
+      "loss": 0.2701,
+      "step": 1473
+    },
+    {
+      "epoch": 1.721354546781492,
+      "grad_norm": 0.07593573488875346,
+      "learning_rate": 1.1556706900218572e-06,
+      "loss": 0.2566,
+      "step": 1474
+    },
+    {
+      "epoch": 1.722522259524157,
+      "grad_norm": 0.08115834315197834,
+      "learning_rate": 1.1461691086809434e-06,
+      "loss": 0.2569,
+      "step": 1475
+    },
+    {
+      "epoch": 1.7236899722668224,
+      "grad_norm": 0.07762469806082506,
+      "learning_rate": 1.1367043732575666e-06,
+      "loss": 0.2565,
+      "step": 1476
+    },
+    {
+      "epoch": 1.7248576850094877,
+      "grad_norm": 0.08154758047718524,
+      "learning_rate": 1.1272765231399684e-06,
+      "loss": 0.2616,
+      "step": 1477
+    },
+    {
+      "epoch": 1.726025397752153,
+      "grad_norm": 0.0771850140525394,
+      "learning_rate": 1.1178855975628967e-06,
+      "loss": 0.2575,
+      "step": 1478
+    },
+    {
+      "epoch": 1.7271931104948184,
+      "grad_norm": 0.0759198945187522,
+      "learning_rate": 1.1085316356074272e-06,
+      "loss": 0.2383,
+      "step": 1479
+    },
+    {
+      "epoch": 1.7283608232374836,
+      "grad_norm": 0.0761863118653295,
+      "learning_rate": 1.099214676200816e-06,
+      "loss": 0.2513,
+      "step": 1480
+    },
+    {
+      "epoch": 1.7295285359801489,
+      "grad_norm": 0.07874928626068532,
+      "learning_rate": 1.0899347581163222e-06,
+      "loss": 0.2629,
+      "step": 1481
+    },
+    {
+      "epoch": 1.7306962487228141,
+      "grad_norm": 0.08052289425848017,
+      "learning_rate": 1.0806919199730614e-06,
+      "loss": 0.255,
+      "step": 1482
+    },
+    {
+      "epoch": 1.7318639614654794,
+      "grad_norm": 0.07919274332405193,
+      "learning_rate": 1.0714862002358385e-06,
+      "loss": 0.2631,
+      "step": 1483
+    },
+    {
+      "epoch": 1.7330316742081449,
+      "grad_norm": 0.0779284852108313,
+      "learning_rate": 1.0623176372149802e-06,
+      "loss": 0.26,
+      "step": 1484
+    },
+    {
+      "epoch": 1.73419938695081,
+      "grad_norm": 0.07660339622797761,
+      "learning_rate": 1.0531862690661865e-06,
+      "loss": 0.2386,
+      "step": 1485
+    },
+    {
+      "epoch": 1.7353670996934754,
+      "grad_norm": 0.07507470051848304,
+      "learning_rate": 1.0440921337903699e-06,
+      "loss": 0.2615,
+      "step": 1486
+    },
+    {
+      "epoch": 1.7365348124361408,
+      "grad_norm": 0.07965349298587879,
+      "learning_rate": 1.035035269233493e-06,
+      "loss": 0.2701,
+      "step": 1487
+    },
+    {
+      "epoch": 1.7377025251788059,
+      "grad_norm": 0.08038351272597795,
+      "learning_rate": 1.0260157130864178e-06,
+      "loss": 0.2558,
+      "step": 1488
+    },
+    {
+      "epoch": 1.7388702379214713,
+      "grad_norm": 0.08065193167454812,
+      "learning_rate": 1.0170335028847345e-06,
+      "loss": 0.2572,
+      "step": 1489
+    },
+    {
+      "epoch": 1.7400379506641366,
+      "grad_norm": 0.08097248826591374,
+      "learning_rate": 1.0080886760086227e-06,
+      "loss": 0.2766,
+      "step": 1490
+    },
+    {
+      "epoch": 1.7412056634068018,
+      "grad_norm": 0.0748457933004247,
+      "learning_rate": 9.991812696826887e-07,
+      "loss": 0.2401,
+      "step": 1491
+    },
+    {
+      "epoch": 1.7423733761494673,
+      "grad_norm": 0.09345219592974412,
+      "learning_rate": 9.903113209758098e-07,
+      "loss": 0.2594,
+      "step": 1492
+    },
+    {
+      "epoch": 1.7435410888921326,
+      "grad_norm": 0.07903294853501726,
+      "learning_rate": 9.814788668009766e-07,
+      "loss": 0.2524,
+      "step": 1493
+    },
+    {
+      "epoch": 1.7447088016347978,
+      "grad_norm": 0.0806733559258006,
+      "learning_rate": 9.726839439151447e-07,
+      "loss": 0.2677,
+      "step": 1494
+    },
+    {
+      "epoch": 1.7458765143774633,
+      "grad_norm": 0.08095910672015483,
+      "learning_rate": 9.63926588919083e-07,
+      "loss": 0.2725,
+      "step": 1495
+    },
+    {
+      "epoch": 1.7470442271201283,
+      "grad_norm": 0.07806711328665183,
+      "learning_rate": 9.55206838257219e-07,
+      "loss": 0.2642,
+      "step": 1496
+    },
+    {
+      "epoch": 1.7482119398627938,
+      "grad_norm": 0.07679311190264576,
+      "learning_rate": 9.465247282174806e-07,
+      "loss": 0.255,
+      "step": 1497
+    },
+    {
+      "epoch": 1.749379652605459,
+      "grad_norm": 0.07686143671448521,
+      "learning_rate": 9.378802949311583e-07,
+      "loss": 0.2531,
+      "step": 1498
+    },
+    {
+      "epoch": 1.7505473653481243,
+      "grad_norm": 0.08029864588648263,
+      "learning_rate": 9.292735743727432e-07,
+      "loss": 0.2623,
+      "step": 1499
+    },
+    {
+      "epoch": 1.7517150780907897,
+      "grad_norm": 0.08072448003328657,
+      "learning_rate": 9.207046023597866e-07,
+      "loss": 0.266,
+      "step": 1500
+    },
+    {
+      "epoch": 1.752882790833455,
+      "grad_norm": 0.07934296039711586,
+      "learning_rate": 9.121734145527394e-07,
+      "loss": 0.2731,
+      "step": 1501
+    },
+    {
+      "epoch": 1.7540505035761202,
+      "grad_norm": 0.07333230612330445,
+      "learning_rate": 9.036800464548157e-07,
+      "loss": 0.243,
+      "step": 1502
+    },
+    {
+      "epoch": 1.7552182163187857,
+      "grad_norm": 0.07939509929200192,
+      "learning_rate": 8.952245334118415e-07,
+      "loss": 0.2607,
+      "step": 1503
+    },
+    {
+      "epoch": 1.7563859290614507,
+      "grad_norm": 0.07837730804950148,
+      "learning_rate": 8.868069106121002e-07,
+      "loss": 0.2549,
+      "step": 1504
+    },
+    {
+      "epoch": 1.7575536418041162,
+      "grad_norm": 0.08160419419396278,
+      "learning_rate": 8.784272130861937e-07,
+      "loss": 0.2795,
+      "step": 1505
+    },
+    {
+      "epoch": 1.7587213545467815,
+      "grad_norm": 0.07688846341007273,
+      "learning_rate": 8.700854757068988e-07,
+      "loss": 0.2656,
+      "step": 1506
+    },
+    {
+      "epoch": 1.7598890672894467,
+      "grad_norm": 0.07998683474185457,
+      "learning_rate": 8.617817331890155e-07,
+      "loss": 0.2537,
+      "step": 1507
+    },
+    {
+      "epoch": 1.7610567800321122,
+      "grad_norm": 0.07975124126089718,
+      "learning_rate": 8.535160200892234e-07,
+      "loss": 0.2614,
+      "step": 1508
+    },
+    {
+      "epoch": 1.7622244927747774,
+      "grad_norm": 0.07369575611915384,
+      "learning_rate": 8.4528837080594e-07,
+      "loss": 0.2369,
+      "step": 1509
+    },
+    {
+      "epoch": 1.7633922055174427,
+      "grad_norm": 0.08098885798844098,
+      "learning_rate": 8.370988195791807e-07,
+      "loss": 0.2808,
+      "step": 1510
+    },
+    {
+      "epoch": 1.7645599182601082,
+      "grad_norm": 0.07796422594644435,
+      "learning_rate": 8.289474004904108e-07,
+      "loss": 0.2586,
+      "step": 1511
+    },
+    {
+      "epoch": 1.7657276310027732,
+      "grad_norm": 0.07544677071712035,
+      "learning_rate": 8.208341474624071e-07,
+      "loss": 0.2399,
+      "step": 1512
+    },
+    {
+      "epoch": 1.7668953437454387,
+      "grad_norm": 0.07520035168795429,
+      "learning_rate": 8.127590942591112e-07,
+      "loss": 0.2723,
+      "step": 1513
+    },
+    {
+      "epoch": 1.768063056488104,
+      "grad_norm": 0.07489298115916565,
+      "learning_rate": 8.047222744854943e-07,
+      "loss": 0.2588,
+      "step": 1514
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.0732397368120482,
+      "learning_rate": 7.96723721587419e-07,
+      "loss": 0.2396,
+      "step": 1515
+    },
+    {
+      "epoch": 1.7703984819734346,
+      "grad_norm": 0.07877178597374775,
+      "learning_rate": 7.887634688515e-07,
+      "loss": 0.2646,
+      "step": 1516
+    },
+    {
+      "epoch": 1.7715661947160999,
+      "grad_norm": 0.07176300591067772,
+      "learning_rate": 7.808415494049515e-07,
+      "loss": 0.2431,
+      "step": 1517
+    },
+    {
+      "epoch": 1.7727339074587651,
+      "grad_norm": 0.07545226763256507,
+      "learning_rate": 7.729579962154743e-07,
+      "loss": 0.2626,
+      "step": 1518
+    },
+    {
+      "epoch": 1.7739016202014306,
+      "grad_norm": 0.07386070456650222,
+      "learning_rate": 7.651128420910925e-07,
+      "loss": 0.2371,
+      "step": 1519
+    },
+    {
+      "epoch": 1.7750693329440956,
+      "grad_norm": 0.07990490518982167,
+      "learning_rate": 7.573061196800414e-07,
+      "loss": 0.2626,
+      "step": 1520
+    },
+    {
+      "epoch": 1.776237045686761,
+      "grad_norm": 0.07841334892856994,
+      "learning_rate": 7.495378614706095e-07,
+      "loss": 0.2633,
+      "step": 1521
+    },
+    {
+      "epoch": 1.7774047584294264,
+      "grad_norm": 0.07613744302270758,
+      "learning_rate": 7.418080997910204e-07,
+      "loss": 0.2508,
+      "step": 1522
+    },
+    {
+      "epoch": 1.7785724711720916,
+      "grad_norm": 0.07651058185420806,
+      "learning_rate": 7.341168668092857e-07,
+      "loss": 0.2562,
+      "step": 1523
+    },
+    {
+      "epoch": 1.779740183914757,
+      "grad_norm": 0.07872505486031017,
+      "learning_rate": 7.26464194533083e-07,
+      "loss": 0.2499,
+      "step": 1524
+    },
+    {
+      "epoch": 1.7809078966574223,
+      "grad_norm": 0.0758761796281371,
+      "learning_rate": 7.188501148096117e-07,
+      "loss": 0.2637,
+      "step": 1525
+    },
+    {
+      "epoch": 1.7820756094000876,
+      "grad_norm": 0.07084579005295237,
+      "learning_rate": 7.11274659325465e-07,
+      "loss": 0.2416,
+      "step": 1526
+    },
+    {
+      "epoch": 1.7832433221427528,
+      "grad_norm": 0.07974081327978304,
+      "learning_rate": 7.037378596065047e-07,
+      "loss": 0.2619,
+      "step": 1527
+    },
+    {
+      "epoch": 1.784411034885418,
+      "grad_norm": 0.07598278734294456,
+      "learning_rate": 6.962397470177162e-07,
+      "loss": 0.2592,
+      "step": 1528
+    },
+    {
+      "epoch": 1.7855787476280836,
+      "grad_norm": 0.07785644824968838,
+      "learning_rate": 6.887803527630865e-07,
+      "loss": 0.2577,
+      "step": 1529
+    },
+    {
+      "epoch": 1.7867464603707488,
+      "grad_norm": 0.07967712480953873,
+      "learning_rate": 6.813597078854772e-07,
+      "loss": 0.2823,
+      "step": 1530
+    },
+    {
+      "epoch": 1.787914173113414,
+      "grad_norm": 0.07290349970073395,
+      "learning_rate": 6.739778432664867e-07,
+      "loss": 0.2414,
+      "step": 1531
+    },
+    {
+      "epoch": 1.7890818858560795,
+      "grad_norm": 0.07247510165757806,
+      "learning_rate": 6.666347896263326e-07,
+      "loss": 0.2461,
+      "step": 1532
+    },
+    {
+      "epoch": 1.7902495985987446,
+      "grad_norm": 0.07635987014608486,
+      "learning_rate": 6.593305775237102e-07,
+      "loss": 0.2477,
+      "step": 1533
+    },
+    {
+      "epoch": 1.79141731134141,
+      "grad_norm": 0.07800903671004232,
+      "learning_rate": 6.520652373556746e-07,
+      "loss": 0.2578,
+      "step": 1534
+    },
+    {
+      "epoch": 1.7925850240840753,
+      "grad_norm": 0.08386231433521177,
+      "learning_rate": 6.448387993575155e-07,
+      "loss": 0.2653,
+      "step": 1535
+    },
+    {
+      "epoch": 1.7937527368267405,
+      "grad_norm": 0.07581371847099755,
+      "learning_rate": 6.37651293602628e-07,
+      "loss": 0.2519,
+      "step": 1536
+    },
+    {
+      "epoch": 1.794920449569406,
+      "grad_norm": 0.07828831424842875,
+      "learning_rate": 6.305027500023841e-07,
+      "loss": 0.2657,
+      "step": 1537
+    },
+    {
+      "epoch": 1.7960881623120712,
+      "grad_norm": 0.07939158803221771,
+      "learning_rate": 6.233931983060104e-07,
+      "loss": 0.2699,
+      "step": 1538
+    },
+    {
+      "epoch": 1.7972558750547365,
+      "grad_norm": 0.08169256348608685,
+      "learning_rate": 6.16322668100472e-07,
+      "loss": 0.2689,
+      "step": 1539
+    },
+    {
+      "epoch": 1.798423587797402,
+      "grad_norm": 0.07820549708942788,
+      "learning_rate": 6.092911888103404e-07,
+      "loss": 0.2448,
+      "step": 1540
+    },
+    {
+      "epoch": 1.799591300540067,
+      "grad_norm": 0.0794530190262452,
+      "learning_rate": 6.022987896976706e-07,
+      "loss": 0.2656,
+      "step": 1541
+    },
+    {
+      "epoch": 1.8007590132827325,
+      "grad_norm": 0.07703391394646772,
+      "learning_rate": 5.953454998618857e-07,
+      "loss": 0.2532,
+      "step": 1542
+    },
+    {
+      "epoch": 1.8019267260253977,
+      "grad_norm": 0.08029336626772833,
+      "learning_rate": 5.884313482396497e-07,
+      "loss": 0.2624,
+      "step": 1543
+    },
+    {
+      "epoch": 1.803094438768063,
+      "grad_norm": 0.07385751398294457,
+      "learning_rate": 5.815563636047539e-07,
+      "loss": 0.2439,
+      "step": 1544
+    },
+    {
+      "epoch": 1.8042621515107284,
+      "grad_norm": 0.07282880887025286,
+      "learning_rate": 5.747205745679885e-07,
+      "loss": 0.2452,
+      "step": 1545
+    },
+    {
+      "epoch": 1.8054298642533937,
+      "grad_norm": 0.07698082435073023,
+      "learning_rate": 5.6792400957703e-07,
+      "loss": 0.2518,
+      "step": 1546
+    },
+    {
+      "epoch": 1.806597576996059,
+      "grad_norm": 0.07949112943986474,
+      "learning_rate": 5.611666969163243e-07,
+      "loss": 0.2585,
+      "step": 1547
+    },
+    {
+      "epoch": 1.8077652897387244,
+      "grad_norm": 0.0833812828815662,
+      "learning_rate": 5.544486647069614e-07,
+      "loss": 0.27,
+      "step": 1548
+    },
+    {
+      "epoch": 1.8089330024813894,
+      "grad_norm": 0.08048399327818755,
+      "learning_rate": 5.477699409065607e-07,
+      "loss": 0.273,
+      "step": 1549
+    },
+    {
+      "epoch": 1.810100715224055,
+      "grad_norm": 0.07890038368486035,
+      "learning_rate": 5.411305533091604e-07,
+      "loss": 0.2586,
+      "step": 1550
+    },
+    {
+      "epoch": 1.8112684279667202,
+      "grad_norm": 0.07764545636350652,
+      "learning_rate": 5.345305295450997e-07,
+      "loss": 0.2579,
+      "step": 1551
+    },
+    {
+      "epoch": 1.8124361407093854,
+      "grad_norm": 0.08044680634987915,
+      "learning_rate": 5.279698970809011e-07,
+      "loss": 0.2664,
+      "step": 1552
+    },
+    {
+      "epoch": 1.8136038534520509,
+      "grad_norm": 0.08135300121440152,
+      "learning_rate": 5.21448683219149e-07,
+      "loss": 0.2661,
+      "step": 1553
+    },
+    {
+      "epoch": 1.8147715661947161,
+      "grad_norm": 0.07806127863594534,
+      "learning_rate": 5.149669150983938e-07,
+      "loss": 0.2613,
+      "step": 1554
+    },
+    {
+      "epoch": 1.8159392789373814,
+      "grad_norm": 0.08218811765284277,
+      "learning_rate": 5.085246196930271e-07,
+      "loss": 0.2537,
+      "step": 1555
+    },
+    {
+      "epoch": 1.8171069916800469,
+      "grad_norm": 0.07648609840384582,
+      "learning_rate": 5.02121823813172e-07,
+      "loss": 0.252,
+      "step": 1556
+    },
+    {
+      "epoch": 1.8182747044227119,
+      "grad_norm": 0.07574516022177001,
+      "learning_rate": 4.957585541045684e-07,
+      "loss": 0.2643,
+      "step": 1557
+    },
+    {
+      "epoch": 1.8194424171653774,
+      "grad_norm": 0.07664469027884323,
+      "learning_rate": 4.894348370484648e-07,
+      "loss": 0.2565,
+      "step": 1558
+    },
+    {
+      "epoch": 1.8206101299080426,
+      "grad_norm": 0.07138790864225504,
+      "learning_rate": 4.831506989615109e-07,
+      "loss": 0.2394,
+      "step": 1559
+    },
+    {
+      "epoch": 1.8217778426507079,
+      "grad_norm": 0.07884746886833019,
+      "learning_rate": 4.769061659956464e-07,
+      "loss": 0.2583,
+      "step": 1560
+    },
+    {
+      "epoch": 1.8229455553933733,
+      "grad_norm": 0.07771845949478473,
+      "learning_rate": 4.707012641379871e-07,
+      "loss": 0.2706,
+      "step": 1561
+    },
+    {
+      "epoch": 1.8241132681360386,
+      "grad_norm": 0.07993440858016668,
+      "learning_rate": 4.64536019210724e-07,
+      "loss": 0.2608,
+      "step": 1562
+    },
+    {
+      "epoch": 1.8252809808787038,
+      "grad_norm": 0.07347490252778902,
+      "learning_rate": 4.5841045687101017e-07,
+      "loss": 0.2633,
+      "step": 1563
+    },
+    {
+      "epoch": 1.8264486936213693,
+      "grad_norm": 0.07659255371537814,
+      "learning_rate": 4.5232460261085966e-07,
+      "loss": 0.2436,
+      "step": 1564
+    },
+    {
+      "epoch": 1.8276164063640343,
+      "grad_norm": 0.0785182241692157,
+      "learning_rate": 4.4627848175703315e-07,
+      "loss": 0.2544,
+      "step": 1565
+    },
+    {
+      "epoch": 1.8287841191066998,
+      "grad_norm": 0.0793060732298849,
+      "learning_rate": 4.402721194709436e-07,
+      "loss": 0.281,
+      "step": 1566
+    },
+    {
+      "epoch": 1.829951831849365,
+      "grad_norm": 0.07823504164754766,
+      "learning_rate": 4.34305540748543e-07,
+      "loss": 0.2577,
+      "step": 1567
+    },
+    {
+      "epoch": 1.8311195445920303,
+      "grad_norm": 0.0805906961505188,
+      "learning_rate": 4.2837877042021915e-07,
+      "loss": 0.2691,
+      "step": 1568
+    },
+    {
+      "epoch": 1.8322872573346958,
+      "grad_norm": 0.07108007502035381,
+      "learning_rate": 4.224918331506955e-07,
+      "loss": 0.2442,
+      "step": 1569
+    },
+    {
+      "epoch": 1.833454970077361,
+      "grad_norm": 0.0781186721440721,
+      "learning_rate": 4.1664475343892727e-07,
+      "loss": 0.2695,
+      "step": 1570
+    },
+    {
+      "epoch": 1.8346226828200263,
+      "grad_norm": 0.07579016303213373,
+      "learning_rate": 4.1083755561799997e-07,
+      "loss": 0.2528,
+      "step": 1571
+    },
+    {
+      "epoch": 1.8357903955626915,
+      "grad_norm": 0.07847261168127065,
+      "learning_rate": 4.0507026385502747e-07,
+      "loss": 0.2673,
+      "step": 1572
+    },
+    {
+      "epoch": 1.8369581083053568,
+      "grad_norm": 0.07562963650485842,
+      "learning_rate": 3.9934290215104754e-07,
+      "loss": 0.2618,
+      "step": 1573
+    },
+    {
+      "epoch": 1.8381258210480222,
+      "grad_norm": 0.0754134538949893,
+      "learning_rate": 3.9365549434092987e-07,
+      "loss": 0.2457,
+      "step": 1574
+    },
+    {
+      "epoch": 1.8392935337906875,
+      "grad_norm": 0.07246848526568568,
+      "learning_rate": 3.8800806409327263e-07,
+      "loss": 0.2366,
+      "step": 1575
+    },
+    {
+      "epoch": 1.8404612465333527,
+      "grad_norm": 0.07593608688100592,
+      "learning_rate": 3.82400634910306e-07,
+      "loss": 0.2543,
+      "step": 1576
+    },
+    {
+      "epoch": 1.8416289592760182,
+      "grad_norm": 0.07558256103432852,
+      "learning_rate": 3.768332301277866e-07,
+      "loss": 0.2551,
+      "step": 1577
+    },
+    {
+      "epoch": 1.8427966720186832,
+      "grad_norm": 0.07958308087260572,
+      "learning_rate": 3.7130587291490995e-07,
+      "loss": 0.2691,
+      "step": 1578
+    },
+    {
+      "epoch": 1.8439643847613487,
+      "grad_norm": 0.07745327818071505,
+      "learning_rate": 3.658185862742103e-07,
+      "loss": 0.2549,
+      "step": 1579
+    },
+    {
+      "epoch": 1.845132097504014,
+      "grad_norm": 0.0794579075198483,
+      "learning_rate": 3.603713930414676e-07,
+      "loss": 0.2766,
+      "step": 1580
+    },
+    {
+      "epoch": 1.8462998102466792,
+      "grad_norm": 0.0811117184453819,
+      "learning_rate": 3.549643158856042e-07,
+      "loss": 0.2679,
+      "step": 1581
+    },
+    {
+      "epoch": 1.8474675229893447,
+      "grad_norm": 0.07908268448234197,
+      "learning_rate": 3.4959737730860144e-07,
+      "loss": 0.2764,
+      "step": 1582
+    },
+    {
+      "epoch": 1.84863523573201,
+      "grad_norm": 0.0770454001538255,
+      "learning_rate": 3.4427059964539654e-07,
+      "loss": 0.258,
+      "step": 1583
+    },
+    {
+      "epoch": 1.8498029484746752,
+      "grad_norm": 0.07314005794660752,
+      "learning_rate": 3.3898400506379937e-07,
+      "loss": 0.2383,
+      "step": 1584
+    },
+    {
+      "epoch": 1.8509706612173407,
+      "grad_norm": 0.07451931154699065,
+      "learning_rate": 3.337376155643879e-07,
+      "loss": 0.2509,
+      "step": 1585
+    },
+    {
+      "epoch": 1.8521383739600057,
+      "grad_norm": 0.07393451058833735,
+      "learning_rate": 3.2853145298042954e-07,
+      "loss": 0.2311,
+      "step": 1586
+    },
+    {
+      "epoch": 1.8533060867026712,
+      "grad_norm": 0.07199259069656545,
+      "learning_rate": 3.2336553897778014e-07,
+      "loss": 0.2413,
+      "step": 1587
+    },
+    {
+      "epoch": 1.8544737994453364,
+      "grad_norm": 0.07958191904924056,
+      "learning_rate": 3.1823989505479934e-07,
+      "loss": 0.2712,
+      "step": 1588
+    },
+    {
+      "epoch": 1.8556415121880017,
+      "grad_norm": 0.0797691187652681,
+      "learning_rate": 3.1315454254226105e-07,
+      "loss": 0.2613,
+      "step": 1589
+    },
+    {
+      "epoch": 1.8568092249306671,
+      "grad_norm": 0.07652746488242397,
+      "learning_rate": 3.081095026032599e-07,
+      "loss": 0.2526,
+      "step": 1590
+    },
+    {
+      "epoch": 1.8579769376733324,
+      "grad_norm": 0.08180630535649164,
+      "learning_rate": 3.0310479623313125e-07,
+      "loss": 0.2756,
+      "step": 1591
+    },
+    {
+      "epoch": 1.8591446504159976,
+      "grad_norm": 0.07585024011426149,
+      "learning_rate": 2.9814044425935605e-07,
+      "loss": 0.2393,
+      "step": 1592
+    },
+    {
+      "epoch": 1.860312363158663,
+      "grad_norm": 0.07937997935520368,
+      "learning_rate": 2.93216467341475e-07,
+      "loss": 0.2616,
+      "step": 1593
+    },
+    {
+      "epoch": 1.8614800759013281,
+      "grad_norm": 0.07688302894556984,
+      "learning_rate": 2.8833288597100996e-07,
+      "loss": 0.251,
+      "step": 1594
+    },
+    {
+      "epoch": 1.8626477886439936,
+      "grad_norm": 0.0783899147038521,
+      "learning_rate": 2.8348972047137155e-07,
+      "loss": 0.2602,
+      "step": 1595
+    },
+    {
+      "epoch": 1.8638155013866589,
+      "grad_norm": 0.07765775196938386,
+      "learning_rate": 2.7868699099777295e-07,
+      "loss": 0.2452,
+      "step": 1596
+    },
+    {
+      "epoch": 1.864983214129324,
+      "grad_norm": 0.08514704300502161,
+      "learning_rate": 2.7392471753715623e-07,
+      "loss": 0.2861,
+      "step": 1597
+    },
+    {
+      "epoch": 1.8661509268719896,
+      "grad_norm": 0.07776993507461574,
+      "learning_rate": 2.69202919908097e-07,
+      "loss": 0.2523,
+      "step": 1598
+    },
+    {
+      "epoch": 1.8673186396146548,
+      "grad_norm": 0.07677177379717748,
+      "learning_rate": 2.645216177607313e-07,
+      "loss": 0.2534,
+      "step": 1599
+    },
+    {
+      "epoch": 1.86848635235732,
+      "grad_norm": 0.07999517309713652,
+      "learning_rate": 2.5988083057666534e-07,
+      "loss": 0.2721,
+      "step": 1600
+    },
+    {
+      "epoch": 1.8696540650999856,
+      "grad_norm": 0.07839536144027943,
+      "learning_rate": 2.552805776689049e-07,
+      "loss": 0.2435,
+      "step": 1601
+    },
+    {
+      "epoch": 1.8708217778426506,
+      "grad_norm": 0.07963267255339665,
+      "learning_rate": 2.507208781817638e-07,
+      "loss": 0.2768,
+      "step": 1602
+    },
+    {
+      "epoch": 1.871989490585316,
+      "grad_norm": 0.07686706762571728,
+      "learning_rate": 2.462017510907955e-07,
+      "loss": 0.2459,
+      "step": 1603
+    },
+    {
+      "epoch": 1.8731572033279813,
+      "grad_norm": 0.07930139274602299,
+      "learning_rate": 2.417232152027016e-07,
+      "loss": 0.2605,
+      "step": 1604
+    },
+    {
+      "epoch": 1.8743249160706466,
+      "grad_norm": 0.07999464890653225,
+      "learning_rate": 2.3728528915526462e-07,
+      "loss": 0.2619,
+      "step": 1605
+    },
+    {
+      "epoch": 1.875492628813312,
+      "grad_norm": 0.07728794951113374,
+      "learning_rate": 2.3288799141726548e-07,
+      "loss": 0.2617,
+      "step": 1606
+    },
+    {
+      "epoch": 1.8766603415559773,
+      "grad_norm": 0.07479892272675191,
+      "learning_rate": 2.2853134028840594e-07,
+      "loss": 0.2428,
+      "step": 1607
+    },
+    {
+      "epoch": 1.8778280542986425,
+      "grad_norm": 0.0820233032520148,
+      "learning_rate": 2.242153538992331e-07,
+      "loss": 0.2874,
+      "step": 1608
+    },
+    {
+      "epoch": 1.878995767041308,
+      "grad_norm": 0.081328111690936,
+      "learning_rate": 2.1994005021106713e-07,
+      "loss": 0.2688,
+      "step": 1609
+    },
+    {
+      "epoch": 1.880163479783973,
+      "grad_norm": 0.07826451648225927,
+      "learning_rate": 2.1570544701592254e-07,
+      "loss": 0.2532,
+      "step": 1610
+    },
+    {
+      "epoch": 1.8813311925266385,
+      "grad_norm": 0.07252962339610157,
+      "learning_rate": 2.1151156193643496e-07,
+      "loss": 0.2397,
+      "step": 1611
+    },
+    {
+      "epoch": 1.8824989052693037,
+      "grad_norm": 0.07673164136196615,
+      "learning_rate": 2.0735841242578992e-07,
+      "loss": 0.2575,
+      "step": 1612
+    },
+    {
+      "epoch": 1.883666618011969,
+      "grad_norm": 0.07818213737500586,
+      "learning_rate": 2.0324601576764525e-07,
+      "loss": 0.2515,
+      "step": 1613
+    },
+    {
+      "epoch": 1.8848343307546345,
+      "grad_norm": 0.07817913727612012,
+      "learning_rate": 1.9917438907606556e-07,
+      "loss": 0.2618,
+      "step": 1614
+    },
+    {
+      "epoch": 1.8860020434972997,
+      "grad_norm": 0.08499463378061385,
+      "learning_rate": 1.951435492954501e-07,
+      "loss": 0.2505,
+      "step": 1615
+    },
+    {
+      "epoch": 1.887169756239965,
+      "grad_norm": 0.07885672869869695,
+      "learning_rate": 1.9115351320045495e-07,
+      "loss": 0.2748,
+      "step": 1616
+    },
+    {
+      "epoch": 1.8883374689826302,
+      "grad_norm": 0.07563868662354395,
+      "learning_rate": 1.8720429739592983e-07,
+      "loss": 0.2554,
+      "step": 1617
+    },
+    {
+      "epoch": 1.8895051817252955,
+      "grad_norm": 0.07703854672872308,
+      "learning_rate": 1.8329591831685146e-07,
+      "loss": 0.2293,
+      "step": 1618
+    },
+    {
+      "epoch": 1.890672894467961,
+      "grad_norm": 0.07795718487406356,
+      "learning_rate": 1.7942839222824804e-07,
+      "loss": 0.2545,
+      "step": 1619
+    },
+    {
+      "epoch": 1.8918406072106262,
+      "grad_norm": 0.0739048102689836,
+      "learning_rate": 1.7560173522513268e-07,
+      "loss": 0.2542,
+      "step": 1620
+    },
+    {
+      "epoch": 1.8930083199532914,
+      "grad_norm": 0.08128990973293676,
+      "learning_rate": 1.7181596323244453e-07,
+      "loss": 0.2685,
+      "step": 1621
+    },
+    {
+      "epoch": 1.894176032695957,
+      "grad_norm": 0.07564788738108068,
+      "learning_rate": 1.6807109200496995e-07,
+      "loss": 0.2431,
+      "step": 1622
+    },
+    {
+      "epoch": 1.895343745438622,
+      "grad_norm": 0.07890428569034055,
+      "learning_rate": 1.6436713712729036e-07,
+      "loss": 0.2659,
+      "step": 1623
+    },
+    {
+      "epoch": 1.8965114581812874,
+      "grad_norm": 0.07514073514442632,
+      "learning_rate": 1.6070411401370335e-07,
+      "loss": 0.2581,
+      "step": 1624
+    },
+    {
+      "epoch": 1.8976791709239527,
+      "grad_norm": 0.07887006232213557,
+      "learning_rate": 1.5708203790817166e-07,
+      "loss": 0.2687,
+      "step": 1625
+    },
+    {
+      "epoch": 1.898846883666618,
+      "grad_norm": 0.07566626836756654,
+      "learning_rate": 1.5350092388425108e-07,
+      "loss": 0.2472,
+      "step": 1626
+    },
+    {
+      "epoch": 1.9000145964092834,
+      "grad_norm": 0.08001971585025656,
+      "learning_rate": 1.4996078684503145e-07,
+      "loss": 0.2668,
+      "step": 1627
+    },
+    {
+      "epoch": 1.9011823091519486,
+      "grad_norm": 0.07414001726963626,
+      "learning_rate": 1.464616415230702e-07,
+      "loss": 0.2422,
+      "step": 1628
+    },
+    {
+      "epoch": 1.9023500218946139,
+      "grad_norm": 0.080874398733993,
+      "learning_rate": 1.4300350248034002e-07,
+      "loss": 0.2714,
+      "step": 1629
+    },
+    {
+      "epoch": 1.9035177346372794,
+      "grad_norm": 0.07345760053156387,
+      "learning_rate": 1.3958638410815905e-07,
+      "loss": 0.2399,
+      "step": 1630
+    },
+    {
+      "epoch": 1.9046854473799444,
+      "grad_norm": 0.07273587834637274,
+      "learning_rate": 1.3621030062713759e-07,
+      "loss": 0.2414,
+      "step": 1631
+    },
+    {
+      "epoch": 1.9058531601226099,
+      "grad_norm": 0.0770397880570712,
+      "learning_rate": 1.3287526608711132e-07,
+      "loss": 0.2498,
+      "step": 1632
+    },
+    {
+      "epoch": 1.907020872865275,
+      "grad_norm": 0.07696361245879378,
+      "learning_rate": 1.2958129436709154e-07,
+      "loss": 0.2586,
+      "step": 1633
+    },
+    {
+      "epoch": 1.9081885856079404,
+      "grad_norm": 0.07645789973875446,
+      "learning_rate": 1.2632839917520179e-07,
+      "loss": 0.2554,
+      "step": 1634
+    },
+    {
+      "epoch": 1.9093562983506058,
+      "grad_norm": 0.07852733642228696,
+      "learning_rate": 1.231165940486234e-07,
+      "loss": 0.2636,
+      "step": 1635
+    },
+    {
+      "epoch": 1.910524011093271,
+      "grad_norm": 0.07859129678924007,
+      "learning_rate": 1.1994589235353682e-07,
+      "loss": 0.2527,
+      "step": 1636
+    },
+    {
+      "epoch": 1.9116917238359363,
+      "grad_norm": 0.07877824058937302,
+      "learning_rate": 1.1681630728506699e-07,
+      "loss": 0.2513,
+      "step": 1637
+    },
+    {
+      "epoch": 1.9128594365786018,
+      "grad_norm": 0.07983489067232094,
+      "learning_rate": 1.1372785186723135e-07,
+      "loss": 0.2551,
+      "step": 1638
+    },
+    {
+      "epoch": 1.9140271493212668,
+      "grad_norm": 0.07003899627460247,
+      "learning_rate": 1.1068053895288089e-07,
+      "loss": 0.2382,
+      "step": 1639
+    },
+    {
+      "epoch": 1.9151948620639323,
+      "grad_norm": 0.0804296118050923,
+      "learning_rate": 1.0767438122364914e-07,
+      "loss": 0.2498,
+      "step": 1640
+    },
+    {
+      "epoch": 1.9163625748065976,
+      "grad_norm": 0.07525440663336658,
+      "learning_rate": 1.0470939118989997e-07,
+      "loss": 0.252,
+      "step": 1641
+    },
+    {
+      "epoch": 1.9175302875492628,
+      "grad_norm": 0.07606006654408859,
+      "learning_rate": 1.0178558119067316e-07,
+      "loss": 0.2558,
+      "step": 1642
+    },
+    {
+      "epoch": 1.9186980002919283,
+      "grad_norm": 0.07932009868129135,
+      "learning_rate": 9.890296339363448e-08,
+      "loss": 0.2749,
+      "step": 1643
+    },
+    {
+      "epoch": 1.9198657130345935,
+      "grad_norm": 0.08187407282514407,
+      "learning_rate": 9.606154979502791e-08,
+      "loss": 0.2701,
+      "step": 1644
+    },
+    {
+      "epoch": 1.9210334257772588,
+      "grad_norm": 0.07579011476005312,
+      "learning_rate": 9.326135221961907e-08,
+      "loss": 0.2521,
+      "step": 1645
+    },
+    {
+      "epoch": 1.9222011385199242,
+      "grad_norm": 0.07940224616651698,
+      "learning_rate": 9.0502382320653e-08,
+      "loss": 0.2813,
+      "step": 1646
+    },
+    {
+      "epoch": 1.9233688512625893,
+      "grad_norm": 0.07856409711833774,
+      "learning_rate": 8.778465157979976e-08,
+      "loss": 0.2511,
+      "step": 1647
+    },
+    {
+      "epoch": 1.9245365640052547,
+      "grad_norm": 0.07828077185998099,
+      "learning_rate": 8.510817130711224e-08,
+      "loss": 0.2491,
+      "step": 1648
+    },
+    {
+      "epoch": 1.92570427674792,
+      "grad_norm": 0.07814190057386258,
+      "learning_rate": 8.247295264097288e-08,
+      "loss": 0.2607,
+      "step": 1649
+    },
+    {
+      "epoch": 1.9268719894905852,
+      "grad_norm": 0.07371518432137043,
+      "learning_rate": 7.987900654805481e-08,
+      "loss": 0.2484,
+      "step": 1650
+    },
+    {
+      "epoch": 1.9280397022332507,
+      "grad_norm": 0.07896113211826396,
+      "learning_rate": 7.732634382326965e-08,
+      "loss": 0.2591,
+      "step": 1651
+    },
+    {
+      "epoch": 1.929207414975916,
+      "grad_norm": 0.07749226933080794,
+      "learning_rate": 7.481497508972313e-08,
+      "loss": 0.2406,
+      "step": 1652
+    },
+    {
+      "epoch": 1.9303751277185812,
+      "grad_norm": 0.07747234188389834,
+      "learning_rate": 7.234491079867623e-08,
+      "loss": 0.2573,
+      "step": 1653
+    },
+    {
+      "epoch": 1.9315428404612467,
+      "grad_norm": 0.07653223135561892,
+      "learning_rate": 6.991616122949629e-08,
+      "loss": 0.2457,
+      "step": 1654
+    },
+    {
+      "epoch": 1.9327105532039117,
+      "grad_norm": 0.08010434188263514,
+      "learning_rate": 6.752873648961821e-08,
+      "loss": 0.2636,
+      "step": 1655
+    },
+    {
+      "epoch": 1.9338782659465772,
+      "grad_norm": 0.08412375275964458,
+      "learning_rate": 6.51826465144978e-08,
+      "loss": 0.2596,
+      "step": 1656
+    },
+    {
+      "epoch": 1.9350459786892424,
+      "grad_norm": 0.07645681655689134,
+      "learning_rate": 6.287790106757396e-08,
+      "loss": 0.2446,
+      "step": 1657
+    },
+    {
+      "epoch": 1.9362136914319077,
+      "grad_norm": 0.07563278414206911,
+      "learning_rate": 6.061450974022775e-08,
+      "loss": 0.2558,
+      "step": 1658
+    },
+    {
+      "epoch": 1.9373814041745732,
+      "grad_norm": 0.07325470411618562,
+      "learning_rate": 5.839248195174008e-08,
+      "loss": 0.2297,
+      "step": 1659
+    },
+    {
+      "epoch": 1.9385491169172384,
+      "grad_norm": 0.07688749382354473,
+      "learning_rate": 5.621182694925731e-08,
+      "loss": 0.2536,
+      "step": 1660
+    },
+    {
+      "epoch": 1.9397168296599037,
+      "grad_norm": 0.07890619129680801,
+      "learning_rate": 5.407255380774579e-08,
+      "loss": 0.2453,
+      "step": 1661
+    },
+    {
+      "epoch": 1.940884542402569,
+      "grad_norm": 0.07371858228673574,
+      "learning_rate": 5.19746714299596e-08,
+      "loss": 0.2431,
+      "step": 1662
+    },
+    {
+      "epoch": 1.9420522551452342,
+      "grad_norm": 0.07746842372449814,
+      "learning_rate": 4.991818854640396e-08,
+      "loss": 0.2565,
+      "step": 1663
+    },
+    {
+      "epoch": 1.9432199678878996,
+      "grad_norm": 0.08069843094357293,
+      "learning_rate": 4.79031137152941e-08,
+      "loss": 0.2613,
+      "step": 1664
+    },
+    {
+      "epoch": 1.9443876806305649,
+      "grad_norm": 0.07727197370738575,
+      "learning_rate": 4.592945532252424e-08,
+      "loss": 0.2668,
+      "step": 1665
+    },
+    {
+      "epoch": 1.9455553933732301,
+      "grad_norm": 0.07734228643641089,
+      "learning_rate": 4.399722158162867e-08,
+      "loss": 0.2576,
+      "step": 1666
+    },
+    {
+      "epoch": 1.9467231061158956,
+      "grad_norm": 0.08975151701731471,
+      "learning_rate": 4.210642053375069e-08,
+      "loss": 0.2603,
+      "step": 1667
+    },
+    {
+      "epoch": 1.9478908188585606,
+      "grad_norm": 0.0736629183823235,
+      "learning_rate": 4.025706004760932e-08,
+      "loss": 0.2309,
+      "step": 1668
+    },
+    {
+      "epoch": 1.949058531601226,
+      "grad_norm": 0.08239743016718891,
+      "learning_rate": 3.844914781946485e-08,
+      "loss": 0.2529,
+      "step": 1669
+    },
+    {
+      "epoch": 1.9502262443438914,
+      "grad_norm": 0.07704483895838904,
+      "learning_rate": 3.668269137308666e-08,
+      "loss": 0.2586,
+      "step": 1670
+    },
+    {
+      "epoch": 1.9513939570865566,
+      "grad_norm": 0.07587481031890778,
+      "learning_rate": 3.495769805972215e-08,
+      "loss": 0.2618,
+      "step": 1671
+    },
+    {
+      "epoch": 1.952561669829222,
+      "grad_norm": 0.07829397739232974,
+      "learning_rate": 3.327417505806785e-08,
+      "loss": 0.2517,
+      "step": 1672
+    },
+    {
+      "epoch": 1.9537293825718873,
+      "grad_norm": 0.08099718851749868,
+      "learning_rate": 3.163212937423832e-08,
+      "loss": 0.2822,
+      "step": 1673
+    },
+    {
+      "epoch": 1.9548970953145526,
+      "grad_norm": 0.07794870734123716,
+      "learning_rate": 3.003156784173511e-08,
+      "loss": 0.2619,
+      "step": 1674
+    },
+    {
+      "epoch": 1.956064808057218,
+      "grad_norm": 0.07641602320722825,
+      "learning_rate": 2.8472497121424524e-08,
+      "loss": 0.2656,
+      "step": 1675
+    },
+    {
+      "epoch": 1.957232520799883,
+      "grad_norm": 0.07576627502753103,
+      "learning_rate": 2.695492370149988e-08,
+      "loss": 0.2445,
+      "step": 1676
+    },
+    {
+      "epoch": 1.9584002335425486,
+      "grad_norm": 0.07702683149389916,
+      "learning_rate": 2.547885389746485e-08,
+      "loss": 0.2457,
+      "step": 1677
+    },
+    {
+      "epoch": 1.9595679462852138,
+      "grad_norm": 0.07394709950079212,
+      "learning_rate": 2.4044293852099055e-08,
+      "loss": 0.2555,
+      "step": 1678
+    },
+    {
+      "epoch": 1.960735659027879,
+      "grad_norm": 0.07767013198663524,
+      "learning_rate": 2.265124953543918e-08,
+      "loss": 0.2536,
+      "step": 1679
+    },
+    {
+      "epoch": 1.9619033717705445,
+      "grad_norm": 0.08278253717428045,
+      "learning_rate": 2.1299726744747896e-08,
+      "loss": 0.269,
+      "step": 1680
+    },
+    {
+      "epoch": 1.9630710845132098,
+      "grad_norm": 0.07468232453167403,
+      "learning_rate": 1.998973110449498e-08,
+      "loss": 0.2443,
+      "step": 1681
+    },
+    {
+      "epoch": 1.964238797255875,
+      "grad_norm": 0.08010236297987025,
+      "learning_rate": 1.872126806633068e-08,
+      "loss": 0.2753,
+      "step": 1682
+    },
+    {
+      "epoch": 1.9654065099985405,
+      "grad_norm": 0.07935094710516827,
+      "learning_rate": 1.749434290906571e-08,
+      "loss": 0.263,
+      "step": 1683
+    },
+    {
+      "epoch": 1.9665742227412055,
+      "grad_norm": 0.08213571508785227,
+      "learning_rate": 1.630896073864352e-08,
+      "loss": 0.2757,
+      "step": 1684
+    },
+    {
+      "epoch": 1.967741935483871,
+      "grad_norm": 0.07730293828175867,
+      "learning_rate": 1.5165126488126957e-08,
+      "loss": 0.2497,
+      "step": 1685
+    },
+    {
+      "epoch": 1.9689096482265362,
+      "grad_norm": 0.0761534333923635,
+      "learning_rate": 1.4062844917672735e-08,
+      "loss": 0.2691,
+      "step": 1686
+    },
+    {
+      "epoch": 1.9700773609692015,
+      "grad_norm": 0.08135662552189675,
+      "learning_rate": 1.300212061451367e-08,
+      "loss": 0.2627,
+      "step": 1687
+    },
+    {
+      "epoch": 1.971245073711867,
+      "grad_norm": 0.08641298025046816,
+      "learning_rate": 1.1982957992936472e-08,
+      "loss": 0.2614,
+      "step": 1688
+    },
+    {
+      "epoch": 1.9724127864545322,
+      "grad_norm": 0.077122933527249,
+      "learning_rate": 1.1005361294268435e-08,
+      "loss": 0.2574,
+      "step": 1689
+    },
+    {
+      "epoch": 1.9735804991971975,
+      "grad_norm": 0.07922004189903463,
+      "learning_rate": 1.0069334586854106e-08,
+      "loss": 0.257,
+      "step": 1690
+    },
+    {
+      "epoch": 1.974748211939863,
+      "grad_norm": 0.07998409956578675,
+      "learning_rate": 9.174881766043086e-09,
+      "loss": 0.2671,
+      "step": 1691
+    },
+    {
+      "epoch": 1.975915924682528,
+      "grad_norm": 0.075997074802238,
+      "learning_rate": 8.322006554171147e-09,
+      "loss": 0.2528,
+      "step": 1692
+    },
+    {
+      "epoch": 1.9770836374251934,
+      "grad_norm": 0.07680465975860558,
+      "learning_rate": 7.510712500546913e-09,
+      "loss": 0.2668,
+      "step": 1693
+    },
+    {
+      "epoch": 1.9782513501678587,
+      "grad_norm": 0.0796025907394313,
+      "learning_rate": 6.741002981435207e-09,
+      "loss": 0.2606,
+      "step": 1694
+    },
+    {
+      "epoch": 1.979419062910524,
+      "grad_norm": 0.07696695149115396,
+      "learning_rate": 6.0128812000415096e-09,
+      "loss": 0.2598,
+      "step": 1695
+    },
+    {
+      "epoch": 1.9805867756531894,
+      "grad_norm": 0.07317533864141566,
+      "learning_rate": 5.326350186503071e-09,
+      "loss": 0.2365,
+      "step": 1696
+    },
+    {
+      "epoch": 1.9817544883958547,
+      "grad_norm": 0.1286030906700583,
+      "learning_rate": 4.681412797872265e-09,
+      "loss": 0.2733,
+      "step": 1697
+    },
+    {
+      "epoch": 1.98292220113852,
+      "grad_norm": 0.0777051928672563,
+      "learning_rate": 4.0780717181077015e-09,
+      "loss": 0.2693,
+      "step": 1698
+    },
+    {
+      "epoch": 1.9840899138811854,
+      "grad_norm": 0.07267120001906875,
+      "learning_rate": 3.5163294580609076e-09,
+      "loss": 0.2395,
+      "step": 1699
+    },
+    {
+      "epoch": 1.9852576266238504,
+      "grad_norm": 0.08122901445852307,
+      "learning_rate": 2.9961883554674443e-09,
+      "loss": 0.26,
+      "step": 1700
+    },
+    {
+      "epoch": 1.9864253393665159,
+      "grad_norm": 0.07827133689241787,
+      "learning_rate": 2.5176505749346937e-09,
+      "loss": 0.2543,
+      "step": 1701
+    },
+    {
+      "epoch": 1.9875930521091811,
+      "grad_norm": 0.07908400147124423,
+      "learning_rate": 2.080718107935198e-09,
+      "loss": 0.2424,
+      "step": 1702
+    },
+    {
+      "epoch": 1.9887607648518464,
+      "grad_norm": 0.07843393631819837,
+      "learning_rate": 1.6853927727988883e-09,
+      "loss": 0.2612,
+      "step": 1703
+    },
+    {
+      "epoch": 1.9899284775945119,
+      "grad_norm": 0.08089341672220379,
+      "learning_rate": 1.3316762147030925e-09,
+      "loss": 0.269,
+      "step": 1704
+    },
+    {
+      "epoch": 1.991096190337177,
+      "grad_norm": 0.07567350639132964,
+      "learning_rate": 1.019569905666984e-09,
+      "loss": 0.2615,
+      "step": 1705
+    },
+    {
+      "epoch": 1.9922639030798424,
+      "grad_norm": 0.07857600454358737,
+      "learning_rate": 7.490751445449196e-10,
+      "loss": 0.2626,
+      "step": 1706
+    },
+    {
+      "epoch": 1.9934316158225076,
+      "grad_norm": 0.0783171274506891,
+      "learning_rate": 5.201930570242209e-10,
+      "loss": 0.2606,
+      "step": 1707
+    },
+    {
+      "epoch": 1.9945993285651729,
+      "grad_norm": 0.07706546401763702,
+      "learning_rate": 3.3292459561518055e-10,
+      "loss": 0.2536,
+      "step": 1708
+    },
+    {
+      "epoch": 1.9957670413078383,
+      "grad_norm": 0.07680542145536048,
+      "learning_rate": 1.872705396488428e-10,
+      "loss": 0.2602,
+      "step": 1709
+    },
+    {
+      "epoch": 1.9969347540505036,
+      "grad_norm": 0.0755685197915718,
+      "learning_rate": 8.323149527811325e-11,
+      "loss": 0.2562,
+      "step": 1710
+    },
+    {
+      "epoch": 1.9981024667931688,
+      "grad_norm": 0.07786633852868742,
+      "learning_rate": 2.080789546776707e-11,
+      "loss": 0.265,
+      "step": 1711
+    },
+    {
+      "epoch": 1.9992701795358343,
+      "grad_norm": 0.07121582636345448,
+      "learning_rate": 0.0,
+      "loss": 0.2448,
+      "step": 1712
+    },
+    {
+      "epoch": 1.9992701795358343,
+      "step": 1712,
+      "total_flos": 254681753763840.0,
+      "train_loss": 0.33850599720064445,
+      "train_runtime": 6372.8107,
+      "train_samples_per_second": 34.397,
+      "train_steps_per_second": 0.269
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 1712,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 254681753763840.0,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}