diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,5776 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 1638,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003663003663003663,
+      "grad_norm": 2.6953821182250977,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 2.6832668781280518,
+      "step": 2
+    },
+    {
+      "epoch": 0.007326007326007326,
+      "grad_norm": 0.3516251742839813,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.6667841672897339,
+      "step": 4
+    },
+    {
+      "epoch": 0.01098901098901099,
+      "grad_norm": 0.37944722175598145,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.8819491863250732,
+      "step": 6
+    },
+    {
+      "epoch": 0.014652014652014652,
+      "grad_norm": 0.41177836060523987,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 2.0589606761932373,
+      "step": 8
+    },
+    {
+      "epoch": 0.018315018315018316,
+      "grad_norm": 0.226902037858963,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 2.2131433486938477,
+      "step": 10
+    },
+    {
+      "epoch": 0.02197802197802198,
+      "grad_norm": 0.6165574193000793,
+      "learning_rate": 4.4e-06,
+      "loss": 2.016043186187744,
+      "step": 12
+    },
+    {
+      "epoch": 0.02564102564102564,
+      "grad_norm": 0.41696810722351074,
+      "learning_rate": 5.2e-06,
+      "loss": 1.7563908100128174,
+      "step": 14
+    },
+    {
+      "epoch": 0.029304029304029304,
+      "grad_norm": 0.20804435014724731,
+      "learning_rate": 6e-06,
+      "loss": 1.8174551725387573,
+      "step": 16
+    },
+    {
+      "epoch": 0.03296703296703297,
+      "grad_norm": 0.6122412085533142,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.750252366065979,
+      "step": 18
+    },
+    {
+      "epoch": 0.03663003663003663,
+      "grad_norm": 0.1671883761882782,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.7453185319900513,
+      "step": 20
+    },
+    {
+      "epoch": 0.040293040293040296,
+      "grad_norm": 0.6752980947494507,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.116586685180664,
+      "step": 22
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 0.3505832552909851,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.2873910665512085,
+      "step": 24
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 0.13198299705982208,
+      "learning_rate": 1e-05,
+      "loss": 1.4626007080078125,
+      "step": 26
+    },
+    {
+      "epoch": 0.05128205128205128,
+      "grad_norm": 1.5736981630325317,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 1.4963958263397217,
+      "step": 28
+    },
+    {
+      "epoch": 0.054945054945054944,
+      "grad_norm": 0.40237972140312195,
+      "learning_rate": 1.16e-05,
+      "loss": 1.1777423620224,
+      "step": 30
+    },
+    {
+      "epoch": 0.05860805860805861,
+      "grad_norm": 0.6054655909538269,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.109989047050476,
+      "step": 32
+    },
+    {
+      "epoch": 0.06227106227106227,
+      "grad_norm": 0.07327596098184586,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 1.2445521354675293,
+      "step": 34
+    },
+    {
+      "epoch": 0.06593406593406594,
+      "grad_norm": 0.13201776146888733,
+      "learning_rate": 1.4e-05,
+      "loss": 1.6131020784378052,
+      "step": 36
+    },
+    {
+      "epoch": 0.0695970695970696,
+      "grad_norm": 0.24574972689151764,
+      "learning_rate": 1.48e-05,
+      "loss": 1.3740808963775635,
+      "step": 38
+    },
+    {
+      "epoch": 0.07326007326007326,
+      "grad_norm": 0.20254714787006378,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 1.4367315769195557,
+      "step": 40
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 0.1146855428814888,
+      "learning_rate": 1.64e-05,
+      "loss": 1.1236861944198608,
+      "step": 42
+    },
+    {
+      "epoch": 0.08058608058608059,
+      "grad_norm": 0.44093406200408936,
+      "learning_rate": 1.72e-05,
+      "loss": 1.5332335233688354,
+      "step": 44
+    },
+    {
+      "epoch": 0.08424908424908426,
+      "grad_norm": 0.11041630059480667,
+      "learning_rate": 1.8e-05,
+      "loss": 1.6215733289718628,
+      "step": 46
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 0.6309869289398193,
+      "learning_rate": 1.88e-05,
+      "loss": 1.0613853931427002,
+      "step": 48
+    },
+    {
+      "epoch": 0.09157509157509157,
+      "grad_norm": 0.07759591192007065,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.5116233825683594,
+      "step": 50
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 0.2021571397781372,
+      "learning_rate": 1.999998238790087e-05,
+      "loss": 0.7548582553863525,
+      "step": 52
+    },
+    {
+      "epoch": 0.0989010989010989,
+      "grad_norm": 0.448813259601593,
+      "learning_rate": 1.999984149152137e-05,
+      "loss": 0.9764630198478699,
+      "step": 54
+    },
+    {
+      "epoch": 0.10256410256410256,
+      "grad_norm": 1.4041872024536133,
+      "learning_rate": 1.999955970096814e-05,
+      "loss": 1.34801185131073,
+      "step": 56
+    },
+    {
+      "epoch": 0.10622710622710622,
+      "grad_norm": 0.12317265570163727,
+      "learning_rate": 1.9999137020652663e-05,
+      "loss": 1.1960068941116333,
+      "step": 58
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 0.10297997295856476,
+      "learning_rate": 1.999857345719207e-05,
+      "loss": 1.411942481994629,
+      "step": 60
+    },
+    {
+      "epoch": 0.11355311355311355,
+      "grad_norm": 0.21834629774093628,
+      "learning_rate": 1.9997869019409047e-05,
+      "loss": 1.4279382228851318,
+      "step": 62
+    },
+    {
+      "epoch": 0.11721611721611722,
+      "grad_norm": 0.12077513337135315,
+      "learning_rate": 1.9997023718331707e-05,
+      "loss": 1.3897336721420288,
+      "step": 64
+    },
+    {
+      "epoch": 0.12087912087912088,
+      "grad_norm": 0.26337432861328125,
+      "learning_rate": 1.9996037567193388e-05,
+      "loss": 1.3532465696334839,
+      "step": 66
+    },
+    {
+      "epoch": 0.12454212454212454,
+      "grad_norm": 0.1620335876941681,
+      "learning_rate": 1.9994910581432466e-05,
+      "loss": 1.3219798803329468,
+      "step": 68
+    },
+    {
+      "epoch": 0.1282051282051282,
+      "grad_norm": 0.4770932197570801,
+      "learning_rate": 1.9993642778692116e-05,
+      "loss": 1.0979112386703491,
+      "step": 70
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 0.11338527500629425,
+      "learning_rate": 1.999223417882002e-05,
+      "loss": 1.4318253993988037,
+      "step": 72
+    },
+    {
+      "epoch": 0.13553113553113552,
+      "grad_norm": 0.1759554147720337,
+      "learning_rate": 1.9990684803868068e-05,
+      "loss": 1.5916600227355957,
+      "step": 74
+    },
+    {
+      "epoch": 0.1391941391941392,
+      "grad_norm": 0.06032824143767357,
+      "learning_rate": 1.9988994678092007e-05,
+      "loss": 1.0843533277511597,
+      "step": 76
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 0.23505143821239471,
+      "learning_rate": 1.9987163827951077e-05,
+      "loss": 1.4368324279785156,
+      "step": 78
+    },
+    {
+      "epoch": 0.14652014652014653,
+      "grad_norm": 0.2501824200153351,
+      "learning_rate": 1.998519228210756e-05,
+      "loss": 1.5794371366500854,
+      "step": 80
+    },
+    {
+      "epoch": 0.15018315018315018,
+      "grad_norm": 0.16015635430812836,
+      "learning_rate": 1.998308007142638e-05,
+      "loss": 1.2044681310653687,
+      "step": 82
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 0.10342324525117874,
+      "learning_rate": 1.9980827228974575e-05,
+      "loss": 1.3625484704971313,
+      "step": 84
+    },
+    {
+      "epoch": 0.1575091575091575,
+      "grad_norm": 0.3542490005493164,
+      "learning_rate": 1.997843379002081e-05,
+      "loss": 1.497490406036377,
+      "step": 86
+    },
+    {
+      "epoch": 0.16117216117216118,
+      "grad_norm": 0.24796238541603088,
+      "learning_rate": 1.9975899792034824e-05,
+      "loss": 0.7302997708320618,
+      "step": 88
+    },
+    {
+      "epoch": 0.16483516483516483,
+      "grad_norm": 0.12096573412418365,
+      "learning_rate": 1.9973225274686804e-05,
+      "loss": 0.8656339645385742,
+      "step": 90
+    },
+    {
+      "epoch": 0.1684981684981685,
+      "grad_norm": 0.12484420835971832,
+      "learning_rate": 1.9970410279846816e-05,
+      "loss": 1.3151899576187134,
+      "step": 92
+    },
+    {
+      "epoch": 0.17216117216117216,
+      "grad_norm": 0.23127628862857819,
+      "learning_rate": 1.9967454851584132e-05,
+      "loss": 1.341341495513916,
+      "step": 94
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 0.12265295535326004,
+      "learning_rate": 1.996435903616651e-05,
+      "loss": 1.2242141962051392,
+      "step": 96
+    },
+    {
+      "epoch": 0.1794871794871795,
+      "grad_norm": 0.11659730225801468,
+      "learning_rate": 1.9961122882059523e-05,
+      "loss": 1.3279354572296143,
+      "step": 98
+    },
+    {
+      "epoch": 0.18315018315018314,
+      "grad_norm": 0.1897869110107422,
+      "learning_rate": 1.9957746439925748e-05,
+      "loss": 1.177512764930725,
+      "step": 100
+    },
+    {
+      "epoch": 0.18681318681318682,
+      "grad_norm": 0.1538306474685669,
+      "learning_rate": 1.9954229762624016e-05,
+      "loss": 1.2517194747924805,
+      "step": 102
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 0.1801384538412094,
+      "learning_rate": 1.995057290520855e-05,
+      "loss": 0.9078481793403625,
+      "step": 104
+    },
+    {
+      "epoch": 0.19413919413919414,
+      "grad_norm": 0.2679578959941864,
+      "learning_rate": 1.9946775924928132e-05,
+      "loss": 1.091381311416626,
+      "step": 106
+    },
+    {
+      "epoch": 0.1978021978021978,
+      "grad_norm": 0.08728894591331482,
+      "learning_rate": 1.9942838881225183e-05,
+      "loss": 1.313234806060791,
+      "step": 108
+    },
+    {
+      "epoch": 0.20146520146520147,
+      "grad_norm": 0.17695370316505432,
+      "learning_rate": 1.9938761835734842e-05,
+      "loss": 1.167786717414856,
+      "step": 110
+    },
+    {
+      "epoch": 0.20512820512820512,
+      "grad_norm": 0.14857837557792664,
+      "learning_rate": 1.9934544852284013e-05,
+      "loss": 1.225075602531433,
+      "step": 112
+    },
+    {
+      "epoch": 0.2087912087912088,
+      "grad_norm": 0.3321186602115631,
+      "learning_rate": 1.9930187996890347e-05,
+      "loss": 0.6301332712173462,
+      "step": 114
+    },
+    {
+      "epoch": 0.21245421245421245,
+      "grad_norm": 0.23063647747039795,
+      "learning_rate": 1.992569133776121e-05,
+      "loss": 1.2985056638717651,
+      "step": 116
+    },
+    {
+      "epoch": 0.21611721611721613,
+      "grad_norm": 0.2007586658000946,
+      "learning_rate": 1.992105494529264e-05,
+      "loss": 1.29304039478302,
+      "step": 118
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 0.18116247653961182,
+      "learning_rate": 1.99162788920682e-05,
+      "loss": 1.363271713256836,
+      "step": 120
+    },
+    {
+      "epoch": 0.22344322344322345,
+      "grad_norm": 0.5378410220146179,
+      "learning_rate": 1.9911363252857887e-05,
+      "loss": 1.2927430868148804,
+      "step": 122
+    },
+    {
+      "epoch": 0.2271062271062271,
+      "grad_norm": 0.4190959632396698,
+      "learning_rate": 1.990630810461694e-05,
+      "loss": 1.0202268362045288,
+      "step": 124
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 0.3055667281150818,
+      "learning_rate": 1.990111352648463e-05,
+      "loss": 0.894250750541687,
+      "step": 126
+    },
+    {
+      "epoch": 0.23443223443223443,
+      "grad_norm": 0.6129741668701172,
+      "learning_rate": 1.9895779599783033e-05,
+      "loss": 1.1953431367874146,
+      "step": 128
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 0.4913721978664398,
+      "learning_rate": 1.989030640801576e-05,
+      "loss": 1.2949491739273071,
+      "step": 130
+    },
+    {
+      "epoch": 0.24175824175824176,
+      "grad_norm": 0.1244189590215683,
+      "learning_rate": 1.9884694036866624e-05,
+      "loss": 1.394888997077942,
+      "step": 132
+    },
+    {
+      "epoch": 0.2454212454212454,
+      "grad_norm": 0.2685690224170685,
+      "learning_rate": 1.9878942574198334e-05,
+      "loss": 1.289100170135498,
+      "step": 134
+    },
+    {
+      "epoch": 0.2490842490842491,
+      "grad_norm": 0.6625173091888428,
+      "learning_rate": 1.9873052110051094e-05,
+      "loss": 1.2693977355957031,
+      "step": 136
+    },
+    {
+      "epoch": 0.25274725274725274,
+      "grad_norm": 0.18069927394390106,
+      "learning_rate": 1.9867022736641205e-05,
+      "loss": 1.0891296863555908,
+      "step": 138
+    },
+    {
+      "epoch": 0.2564102564102564,
+      "grad_norm": 0.17798098921775818,
+      "learning_rate": 1.9860854548359615e-05,
+      "loss": 1.2750816345214844,
+      "step": 140
+    },
+    {
+      "epoch": 0.2600732600732601,
+      "grad_norm": 0.3319156765937805,
+      "learning_rate": 1.9854547641770446e-05,
+      "loss": 1.287410855293274,
+      "step": 142
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 0.4369660019874573,
+      "learning_rate": 1.9848102115609483e-05,
+      "loss": 1.2549277544021606,
+      "step": 144
+    },
+    {
+      "epoch": 0.2673992673992674,
+      "grad_norm": 0.2410343736410141,
+      "learning_rate": 1.9841518070782615e-05,
+      "loss": 1.4124380350112915,
+      "step": 146
+    },
+    {
+      "epoch": 0.27106227106227104,
+      "grad_norm": 0.15733124315738678,
+      "learning_rate": 1.983479561036429e-05,
+      "loss": 1.3137692213058472,
+      "step": 148
+    },
+    {
+      "epoch": 0.27472527472527475,
+      "grad_norm": 0.07597765326499939,
+      "learning_rate": 1.982793483959585e-05,
+      "loss": 0.9285884499549866,
+      "step": 150
+    },
+    {
+      "epoch": 0.2783882783882784,
+      "grad_norm": 0.07245276868343353,
+      "learning_rate": 1.9820935865883924e-05,
+      "loss": 0.6362338662147522,
+      "step": 152
+    },
+    {
+      "epoch": 0.28205128205128205,
+      "grad_norm": 0.2137291133403778,
+      "learning_rate": 1.981379879879874e-05,
+      "loss": 1.0962413549423218,
+      "step": 154
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 0.11381687223911285,
+      "learning_rate": 1.9806523750072385e-05,
+      "loss": 1.3381097316741943,
+      "step": 156
+    },
+    {
+      "epoch": 0.2893772893772894,
+      "grad_norm": 0.12019956111907959,
+      "learning_rate": 1.9799110833597093e-05,
+      "loss": 1.2722944021224976,
+      "step": 158
+    },
+    {
+      "epoch": 0.29304029304029305,
+      "grad_norm": 0.10780342668294907,
+      "learning_rate": 1.9791560165423433e-05,
+      "loss": 0.9358208179473877,
+      "step": 160
+    },
+    {
+      "epoch": 0.2967032967032967,
+      "grad_norm": 0.23072299361228943,
+      "learning_rate": 1.9783871863758503e-05,
+      "loss": 1.5278652906417847,
+      "step": 162
+    },
+    {
+      "epoch": 0.30036630036630035,
+      "grad_norm": 0.1233779788017273,
+      "learning_rate": 1.9776046048964082e-05,
+      "loss": 1.045573115348816,
+      "step": 164
+    },
+    {
+      "epoch": 0.304029304029304,
+      "grad_norm": 1.4242370128631592,
+      "learning_rate": 1.9768082843554737e-05,
+      "loss": 1.3891277313232422,
+      "step": 166
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 0.1399042010307312,
+      "learning_rate": 1.9759982372195918e-05,
+      "loss": 1.1296571493148804,
+      "step": 168
+    },
+    {
+      "epoch": 0.31135531135531136,
+      "grad_norm": 0.1742742955684662,
+      "learning_rate": 1.9751744761701984e-05,
+      "loss": 1.2520029544830322,
+      "step": 170
+    },
+    {
+      "epoch": 0.315018315018315,
+      "grad_norm": 0.28852102160453796,
+      "learning_rate": 1.9743370141034248e-05,
+      "loss": 1.0042704343795776,
+      "step": 172
+    },
+    {
+      "epoch": 0.31868131868131866,
+      "grad_norm": 0.1639695167541504,
+      "learning_rate": 1.973485864129894e-05,
+      "loss": 0.8751227855682373,
+      "step": 174
+    },
+    {
+      "epoch": 0.32234432234432236,
+      "grad_norm": 0.12699364125728607,
+      "learning_rate": 1.9726210395745148e-05,
+      "loss": 1.3895294666290283,
+      "step": 176
+    },
+    {
+      "epoch": 0.326007326007326,
+      "grad_norm": 0.23989446461200714,
+      "learning_rate": 1.971742553976275e-05,
+      "loss": 0.924419105052948,
+      "step": 178
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 0.19326151907444,
+      "learning_rate": 1.9708504210880284e-05,
+      "loss": 1.484243631362915,
+      "step": 180
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 0.314147412776947,
+      "learning_rate": 1.969944654876279e-05,
+      "loss": 0.9775104522705078,
+      "step": 182
+    },
+    {
+      "epoch": 0.336996336996337,
+      "grad_norm": 0.2284594476222992,
+      "learning_rate": 1.9690252695209636e-05,
+      "loss": 1.2366245985031128,
+      "step": 184
+    },
+    {
+      "epoch": 0.34065934065934067,
+      "grad_norm": 0.9983108639717102,
+      "learning_rate": 1.9680922794152294e-05,
+      "loss": 1.3184081315994263,
+      "step": 186
+    },
+    {
+      "epoch": 0.3443223443223443,
+      "grad_norm": 0.1190100610256195,
+      "learning_rate": 1.9671456991652072e-05,
+      "loss": 1.1691346168518066,
+      "step": 188
+    },
+    {
+      "epoch": 0.34798534798534797,
+      "grad_norm": 0.16565515100955963,
+      "learning_rate": 1.9661855435897858e-05,
+      "loss": 1.2617851495742798,
+      "step": 190
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 0.35118862986564636,
+      "learning_rate": 1.9652118277203767e-05,
+      "loss": 1.1071832180023193,
+      "step": 192
+    },
+    {
+      "epoch": 0.3553113553113553,
+      "grad_norm": 0.1040225550532341,
+      "learning_rate": 1.9642245668006814e-05,
+      "loss": 1.2461092472076416,
+      "step": 194
+    },
+    {
+      "epoch": 0.358974358974359,
+      "grad_norm": 0.1812414824962616,
+      "learning_rate": 1.963223776286451e-05,
+      "loss": 1.2525054216384888,
+      "step": 196
+    },
+    {
+      "epoch": 0.3626373626373626,
+      "grad_norm": 0.19630086421966553,
+      "learning_rate": 1.9622094718452448e-05,
+      "loss": 0.8609032034873962,
+      "step": 198
+    },
+    {
+      "epoch": 0.3663003663003663,
+      "grad_norm": 0.2004232257604599,
+      "learning_rate": 1.9611816693561858e-05,
+      "loss": 1.0185203552246094,
+      "step": 200
+    },
+    {
+      "epoch": 0.36996336996337,
+      "grad_norm": 0.2183266133069992,
+      "learning_rate": 1.96014038490971e-05,
+      "loss": 1.4265942573547363,
+      "step": 202
+    },
+    {
+      "epoch": 0.37362637362637363,
+      "grad_norm": 0.47130584716796875,
+      "learning_rate": 1.9590856348073182e-05,
+      "loss": 1.204400897026062,
+      "step": 204
+    },
+    {
+      "epoch": 0.3772893772893773,
+      "grad_norm": 0.5094867944717407,
+      "learning_rate": 1.9580174355613168e-05,
+      "loss": 0.7685510516166687,
+      "step": 206
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 0.18307189643383026,
+      "learning_rate": 1.9569358038945617e-05,
+      "loss": 1.135070562362671,
+      "step": 208
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 0.3276787996292114,
+      "learning_rate": 1.9558407567401945e-05,
+      "loss": 1.409148097038269,
+      "step": 210
+    },
+    {
+      "epoch": 0.3882783882783883,
+      "grad_norm": 0.2551312744617462,
+      "learning_rate": 1.9547323112413806e-05,
+      "loss": 1.0703577995300293,
+      "step": 212
+    },
+    {
+      "epoch": 0.39194139194139194,
+      "grad_norm": 0.4214024841785431,
+      "learning_rate": 1.9536104847510384e-05,
+      "loss": 1.1370527744293213,
+      "step": 214
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 0.18482302129268646,
+      "learning_rate": 1.9524752948315677e-05,
+      "loss": 1.2194429636001587,
+      "step": 216
+    },
+    {
+      "epoch": 0.3992673992673993,
+      "grad_norm": 0.1274556815624237,
+      "learning_rate": 1.9513267592545752e-05,
+      "loss": 1.2507331371307373,
+      "step": 218
+    },
+    {
+      "epoch": 0.40293040293040294,
+      "grad_norm": 0.11079567670822144,
+      "learning_rate": 1.9501648960005964e-05,
+      "loss": 0.6284658908843994,
+      "step": 220
+    },
+    {
+      "epoch": 0.4065934065934066,
+      "grad_norm": 0.1624489277601242,
+      "learning_rate": 1.948989723258815e-05,
+      "loss": 1.3324811458587646,
+      "step": 222
+    },
+    {
+      "epoch": 0.41025641025641024,
+      "grad_norm": 0.2038378268480301,
+      "learning_rate": 1.9478012594267757e-05,
+      "loss": 1.0770931243896484,
+      "step": 224
+    },
+    {
+      "epoch": 0.4139194139194139,
+      "grad_norm": 0.18007361888885498,
+      "learning_rate": 1.946599523110099e-05,
+      "loss": 1.2338424921035767,
+      "step": 226
+    },
+    {
+      "epoch": 0.4175824175824176,
+      "grad_norm": 0.24560625851154327,
+      "learning_rate": 1.945384533122187e-05,
+      "loss": 1.3044496774673462,
+      "step": 228
+    },
+    {
+      "epoch": 0.42124542124542125,
+      "grad_norm": 0.2072787582874298,
+      "learning_rate": 1.9441563084839324e-05,
+      "loss": 1.233733057975769,
+      "step": 230
+    },
+    {
+      "epoch": 0.4249084249084249,
+      "grad_norm": 0.39441704750061035,
+      "learning_rate": 1.942914868423417e-05,
+      "loss": 0.997710645198822,
+      "step": 232
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 0.30978092551231384,
+      "learning_rate": 1.941660232375614e-05,
+      "loss": 1.4850120544433594,
+      "step": 234
+    },
+    {
+      "epoch": 0.43223443223443225,
+      "grad_norm": 0.4644772708415985,
+      "learning_rate": 1.9403924199820813e-05,
+      "loss": 1.0170940160751343,
+      "step": 236
+    },
+    {
+      "epoch": 0.4358974358974359,
+      "grad_norm": 0.147545725107193,
+      "learning_rate": 1.9391114510906546e-05,
+      "loss": 1.0712804794311523,
+      "step": 238
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 0.23715418577194214,
+      "learning_rate": 1.937817345755138e-05,
+      "loss": 0.9054774641990662,
+      "step": 240
+    },
+    {
+      "epoch": 0.4432234432234432,
+      "grad_norm": 0.20748640596866608,
+      "learning_rate": 1.9365101242349883e-05,
+      "loss": 0.8780495524406433,
+      "step": 242
+    },
+    {
+      "epoch": 0.4468864468864469,
+      "grad_norm": 0.1142963096499443,
+      "learning_rate": 1.9351898069949985e-05,
+      "loss": 0.5809678435325623,
+      "step": 244
+    },
+    {
+      "epoch": 0.45054945054945056,
+      "grad_norm": 0.16483286023139954,
+      "learning_rate": 1.9338564147049785e-05,
+      "loss": 1.2516872882843018,
+      "step": 246
+    },
+    {
+      "epoch": 0.4542124542124542,
+      "grad_norm": 0.31650102138519287,
+      "learning_rate": 1.9325099682394296e-05,
+      "loss": 0.8757045269012451,
+      "step": 248
+    },
+    {
+      "epoch": 0.45787545787545786,
+      "grad_norm": 0.5607954263687134,
+      "learning_rate": 1.9311504886772183e-05,
+      "loss": 1.2788811922073364,
+      "step": 250
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 0.23190627992153168,
+      "learning_rate": 1.929777997301248e-05,
+      "loss": 1.1750297546386719,
+      "step": 252
+    },
+    {
+      "epoch": 0.4652014652014652,
+      "grad_norm": 0.18342120945453644,
+      "learning_rate": 1.9283925155981228e-05,
+      "loss": 0.9642306566238403,
+      "step": 254
+    },
+    {
+      "epoch": 0.46886446886446886,
+      "grad_norm": 0.20935122668743134,
+      "learning_rate": 1.9269940652578143e-05,
+      "loss": 1.2557034492492676,
+      "step": 256
+    },
+    {
+      "epoch": 0.4725274725274725,
+      "grad_norm": 0.26344409584999084,
+      "learning_rate": 1.9255826681733194e-05,
+      "loss": 1.2821006774902344,
+      "step": 258
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 0.08390780538320541,
+      "learning_rate": 1.924158346440319e-05,
+      "loss": 0.7510356307029724,
+      "step": 260
+    },
+    {
+      "epoch": 0.47985347985347987,
+      "grad_norm": 0.13334789872169495,
+      "learning_rate": 1.9227211223568317e-05,
+      "loss": 1.1425384283065796,
+      "step": 262
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 0.292468786239624,
+      "learning_rate": 1.9212710184228654e-05,
+      "loss": 1.2262020111083984,
+      "step": 264
+    },
+    {
+      "epoch": 0.48717948717948717,
+      "grad_norm": 0.25990208983421326,
+      "learning_rate": 1.9198080573400634e-05,
+      "loss": 1.493544578552246,
+      "step": 266
+    },
+    {
+      "epoch": 0.4908424908424908,
+      "grad_norm": 1.0916262865066528,
+      "learning_rate": 1.9183322620113505e-05,
+      "loss": 0.8011191487312317,
+      "step": 268
+    },
+    {
+      "epoch": 0.4945054945054945,
+      "grad_norm": 0.18405258655548096,
+      "learning_rate": 1.916843655540574e-05,
+      "loss": 1.1977633237838745,
+      "step": 270
+    },
+    {
+      "epoch": 0.4981684981684982,
+      "grad_norm": 0.808682918548584,
+      "learning_rate": 1.915342261232142e-05,
+      "loss": 0.8853522539138794,
+      "step": 272
+    },
+    {
+      "epoch": 0.5018315018315018,
+      "grad_norm": 0.27525171637535095,
+      "learning_rate": 1.913828102590659e-05,
+      "loss": 1.246770977973938,
+      "step": 274
+    },
+    {
+      "epoch": 0.5054945054945055,
+      "grad_norm": 0.2584117352962494,
+      "learning_rate": 1.9123012033205564e-05,
+      "loss": 0.8163619041442871,
+      "step": 276
+    },
+    {
+      "epoch": 0.5091575091575091,
+      "grad_norm": 0.24883966147899628,
+      "learning_rate": 1.9107615873257234e-05,
+      "loss": 0.8779458403587341,
+      "step": 278
+    },
+    {
+      "epoch": 0.5128205128205128,
+      "grad_norm": 0.3886781930923462,
+      "learning_rate": 1.909209278709131e-05,
+      "loss": 1.2527726888656616,
+      "step": 280
+    },
+    {
+      "epoch": 0.5164835164835165,
+      "grad_norm": 0.29291361570358276,
+      "learning_rate": 1.9076443017724568e-05,
+      "loss": 1.2405891418457031,
+      "step": 282
+    },
+    {
+      "epoch": 0.5201465201465202,
+      "grad_norm": 0.21082301437854767,
+      "learning_rate": 1.9060666810157025e-05,
+      "loss": 1.2477757930755615,
+      "step": 284
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 0.9164182543754578,
+      "learning_rate": 1.9044764411368106e-05,
+      "loss": 1.0075207948684692,
+      "step": 286
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 0.19336040318012238,
+      "learning_rate": 1.9028736070312796e-05,
+      "loss": 1.2511680126190186,
+      "step": 288
+    },
+    {
+      "epoch": 0.5311355311355311,
+      "grad_norm": 0.3687734305858612,
+      "learning_rate": 1.9012582037917713e-05,
+      "loss": 1.2166516780853271,
+      "step": 290
+    },
+    {
+      "epoch": 0.5347985347985348,
+      "grad_norm": 0.13847005367279053,
+      "learning_rate": 1.8996302567077217e-05,
+      "loss": 0.7319579720497131,
+      "step": 292
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 0.03533535450696945,
+      "learning_rate": 1.897989791264941e-05,
+      "loss": 0.947640061378479,
+      "step": 294
+    },
+    {
+      "epoch": 0.5421245421245421,
+      "grad_norm": 0.15593798458576202,
+      "learning_rate": 1.8963368331452172e-05,
+      "loss": 1.0241395235061646,
+      "step": 296
+    },
+    {
+      "epoch": 0.5457875457875457,
+      "grad_norm": 0.6071200370788574,
+      "learning_rate": 1.8946714082259145e-05,
+      "loss": 1.2927855253219604,
+      "step": 298
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 0.22545884549617767,
+      "learning_rate": 1.8929935425795655e-05,
+      "loss": 1.2249903678894043,
+      "step": 300
+    },
+    {
+      "epoch": 0.5531135531135531,
+      "grad_norm": 0.2020682394504547,
+      "learning_rate": 1.8913032624734657e-05,
+      "loss": 1.1927720308303833,
+      "step": 302
+    },
+    {
+      "epoch": 0.5567765567765568,
+      "grad_norm": 0.15021179616451263,
+      "learning_rate": 1.8896005943692614e-05,
+      "loss": 0.9835414886474609,
+      "step": 304
+    },
+    {
+      "epoch": 0.5604395604395604,
+      "grad_norm": 0.20427605509757996,
+      "learning_rate": 1.8878855649225346e-05,
+      "loss": 0.9726413488388062,
+      "step": 306
+    },
+    {
+      "epoch": 0.5641025641025641,
+      "grad_norm": 0.3385741412639618,
+      "learning_rate": 1.8861582009823868e-05,
+      "loss": 1.4145900011062622,
+      "step": 308
+    },
+    {
+      "epoch": 0.5677655677655677,
+      "grad_norm": 1.1410894393920898,
+      "learning_rate": 1.884418529591018e-05,
+      "loss": 0.9968710541725159,
+      "step": 310
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 0.13543498516082764,
+      "learning_rate": 1.882666577983304e-05,
+      "loss": 1.2110992670059204,
+      "step": 312
+    },
+    {
+      "epoch": 0.575091575091575,
+      "grad_norm": 0.8556830883026123,
+      "learning_rate": 1.8809023735863693e-05,
+      "loss": 1.139275074005127,
+      "step": 314
+    },
+    {
+      "epoch": 0.5787545787545788,
+      "grad_norm": 0.20495963096618652,
+      "learning_rate": 1.879125944019158e-05,
+      "loss": 1.282318353652954,
+      "step": 316
+    },
+    {
+      "epoch": 0.5824175824175825,
+      "grad_norm": 0.3021373748779297,
+      "learning_rate": 1.8773373170920022e-05,
+      "loss": 1.1240531206130981,
+      "step": 318
+    },
+    {
+      "epoch": 0.5860805860805861,
+      "grad_norm": 0.15304189920425415,
+      "learning_rate": 1.875536520806185e-05,
+      "loss": 1.3365286588668823,
+      "step": 320
+    },
+    {
+      "epoch": 0.5897435897435898,
+      "grad_norm": 0.35537189245224,
+      "learning_rate": 1.8737235833535033e-05,
+      "loss": 1.522202491760254,
+      "step": 322
+    },
+    {
+      "epoch": 0.5934065934065934,
+      "grad_norm": 0.18663744628429413,
+      "learning_rate": 1.871898533115827e-05,
+      "loss": 1.2638378143310547,
+      "step": 324
+    },
+    {
+      "epoch": 0.5970695970695971,
+      "grad_norm": 0.38809436559677124,
+      "learning_rate": 1.870061398664653e-05,
+      "loss": 1.3586820363998413,
+      "step": 326
+    },
+    {
+      "epoch": 0.6007326007326007,
+      "grad_norm": 0.17811237275600433,
+      "learning_rate": 1.868212208760658e-05,
+      "loss": 1.2239891290664673,
+      "step": 328
+    },
+    {
+      "epoch": 0.6043956043956044,
+      "grad_norm": 0.21789351105690002,
+      "learning_rate": 1.8663509923532514e-05,
+      "loss": 1.1124786138534546,
+      "step": 330
+    },
+    {
+      "epoch": 0.608058608058608,
+      "grad_norm": 0.1282075196504593,
+      "learning_rate": 1.8644777785801175e-05,
+      "loss": 1.1825826168060303,
+      "step": 332
+    },
+    {
+      "epoch": 0.6117216117216118,
+      "grad_norm": 0.13316233456134796,
+      "learning_rate": 1.862592596766763e-05,
+      "loss": 1.2802841663360596,
+      "step": 334
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 0.8452423214912415,
+      "learning_rate": 1.8606954764260556e-05,
+      "loss": 0.9047472476959229,
+      "step": 336
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 1.1467108726501465,
+      "learning_rate": 1.8587864472577632e-05,
+      "loss": 1.2362874746322632,
+      "step": 338
+    },
+    {
+      "epoch": 0.6227106227106227,
+      "grad_norm": 0.36012572050094604,
+      "learning_rate": 1.8568655391480882e-05,
+      "loss": 1.2298355102539062,
+      "step": 340
+    },
+    {
+      "epoch": 0.6263736263736264,
+      "grad_norm": 0.402875155210495,
+      "learning_rate": 1.8549327821692008e-05,
+      "loss": 0.5815396904945374,
+      "step": 342
+    },
+    {
+      "epoch": 0.63003663003663,
+      "grad_norm": 0.15672507882118225,
+      "learning_rate": 1.852988206578767e-05,
+      "loss": 1.4353272914886475,
+      "step": 344
+    },
+    {
+      "epoch": 0.6336996336996337,
+      "grad_norm": 0.48711514472961426,
+      "learning_rate": 1.851031842819475e-05,
+      "loss": 0.6951726675033569,
+      "step": 346
+    },
+    {
+      "epoch": 0.6373626373626373,
+      "grad_norm": 0.18968786299228668,
+      "learning_rate": 1.849063721518559e-05,
+      "loss": 1.1487343311309814,
+      "step": 348
+    },
+    {
+      "epoch": 0.6410256410256411,
+      "grad_norm": 0.28493592143058777,
+      "learning_rate": 1.8470838734873205e-05,
+      "loss": 0.8871045112609863,
+      "step": 350
+    },
+    {
+      "epoch": 0.6446886446886447,
+      "grad_norm": 0.1983412355184555,
+      "learning_rate": 1.8450923297206446e-05,
+      "loss": 0.9253445267677307,
+      "step": 352
+    },
+    {
+      "epoch": 0.6483516483516484,
+      "grad_norm": 0.37879106402397156,
+      "learning_rate": 1.8430891213965146e-05,
+      "loss": 0.9548962712287903,
+      "step": 354
+    },
+    {
+      "epoch": 0.652014652014652,
+      "grad_norm": 0.3444487750530243,
+      "learning_rate": 1.8410742798755255e-05,
+      "loss": 1.1662979125976562,
+      "step": 356
+    },
+    {
+      "epoch": 0.6556776556776557,
+      "grad_norm": 0.3182324171066284,
+      "learning_rate": 1.8390478367003922e-05,
+      "loss": 1.1501445770263672,
+      "step": 358
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 0.3570864796638489,
+      "learning_rate": 1.8370098235954553e-05,
+      "loss": 0.6915228366851807,
+      "step": 360
+    },
+    {
+      "epoch": 0.663003663003663,
+      "grad_norm": 0.25218188762664795,
+      "learning_rate": 1.834960272466184e-05,
+      "loss": 0.9469904899597168,
+      "step": 362
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 0.2885875105857849,
+      "learning_rate": 1.832899215398679e-05,
+      "loss": 0.9307279586791992,
+      "step": 364
+    },
+    {
+      "epoch": 0.6703296703296703,
+      "grad_norm": 0.3014400005340576,
+      "learning_rate": 1.8308266846591673e-05,
+      "loss": 1.1860599517822266,
+      "step": 366
+    },
+    {
+      "epoch": 0.673992673992674,
+      "grad_norm": 0.22383049130439758,
+      "learning_rate": 1.828742712693499e-05,
+      "loss": 1.05683434009552,
+      "step": 368
+    },
+    {
+      "epoch": 0.6776556776556777,
+      "grad_norm": 0.26446428894996643,
+      "learning_rate": 1.8266473321266385e-05,
+      "loss": 1.0885175466537476,
+      "step": 370
+    },
+    {
+      "epoch": 0.6813186813186813,
+      "grad_norm": 0.511090874671936,
+      "learning_rate": 1.824540575762154e-05,
+      "loss": 1.189785122871399,
+      "step": 372
+    },
+    {
+      "epoch": 0.684981684981685,
+      "grad_norm": 0.3425454795360565,
+      "learning_rate": 1.8224224765817033e-05,
+      "loss": 1.1994352340698242,
+      "step": 374
+    },
+    {
+      "epoch": 0.6886446886446886,
+      "grad_norm": 0.2739914655685425,
+      "learning_rate": 1.820293067744519e-05,
+      "loss": 0.9016158580780029,
+      "step": 376
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 0.42973825335502625,
+      "learning_rate": 1.8181523825868882e-05,
+      "loss": 0.8308103084564209,
+      "step": 378
+    },
+    {
+      "epoch": 0.6959706959706959,
+      "grad_norm": 0.6218528747558594,
+      "learning_rate": 1.816000454621631e-05,
+      "loss": 1.0427895784378052,
+      "step": 380
+    },
+    {
+      "epoch": 0.6996336996336996,
+      "grad_norm": 0.14869005978107452,
+      "learning_rate": 1.8138373175375744e-05,
+      "loss": 0.9768989086151123,
+      "step": 382
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 0.23566092550754547,
+      "learning_rate": 1.8116630051990283e-05,
+      "loss": 1.1809189319610596,
+      "step": 384
+    },
+    {
+      "epoch": 0.706959706959707,
+      "grad_norm": 1.1532787084579468,
+      "learning_rate": 1.8094775516452522e-05,
+      "loss": 1.0936170816421509,
+      "step": 386
+    },
+    {
+      "epoch": 0.7106227106227107,
+      "grad_norm": 0.19187268614768982,
+      "learning_rate": 1.807280991089923e-05,
+      "loss": 0.8833699226379395,
+      "step": 388
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 0.15410183370113373,
+      "learning_rate": 1.8050733579206005e-05,
+      "loss": 1.0913102626800537,
+      "step": 390
+    },
+    {
+      "epoch": 0.717948717948718,
+      "grad_norm": 0.42997273802757263,
+      "learning_rate": 1.8028546866981875e-05,
+      "loss": 1.175827980041504,
+      "step": 392
+    },
+    {
+      "epoch": 0.7216117216117216,
+      "grad_norm": 0.49263566732406616,
+      "learning_rate": 1.8006250121563903e-05,
+      "loss": 1.118180513381958,
+      "step": 394
+    },
+    {
+      "epoch": 0.7252747252747253,
+      "grad_norm": 0.24720455706119537,
+      "learning_rate": 1.798384369201174e-05,
+      "loss": 1.2381089925765991,
+      "step": 396
+    },
+    {
+      "epoch": 0.7289377289377289,
+      "grad_norm": 0.2054165154695511,
+      "learning_rate": 1.796132792910216e-05,
+      "loss": 0.9162780046463013,
+      "step": 398
+    },
+    {
+      "epoch": 0.7326007326007326,
+      "grad_norm": 0.2705683410167694,
+      "learning_rate": 1.7938703185323575e-05,
+      "loss": 0.8435118198394775,
+      "step": 400
+    },
+    {
+      "epoch": 0.7362637362637363,
+      "grad_norm": 0.18964652717113495,
+      "learning_rate": 1.7915969814870508e-05,
+      "loss": 1.2435356378555298,
+      "step": 402
+    },
+    {
+      "epoch": 0.73992673992674,
+      "grad_norm": 0.10768619179725647,
+      "learning_rate": 1.789312817363805e-05,
+      "loss": 0.8427106738090515,
+      "step": 404
+    },
+    {
+      "epoch": 0.7435897435897436,
+      "grad_norm": 0.6611467599868774,
+      "learning_rate": 1.7870178619216304e-05,
+      "loss": 1.0203189849853516,
+      "step": 406
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 0.6479913592338562,
+      "learning_rate": 1.784712151088476e-05,
+      "loss": 1.0223674774169922,
+      "step": 408
+    },
+    {
+      "epoch": 0.7509157509157509,
+      "grad_norm": 0.5998079776763916,
+      "learning_rate": 1.782395720960669e-05,
+      "loss": 0.8774959444999695,
+      "step": 410
+    },
+    {
+      "epoch": 0.7545787545787546,
+      "grad_norm": 0.31854361295700073,
+      "learning_rate": 1.780068607802349e-05,
+      "loss": 1.1714166402816772,
+      "step": 412
+    },
+    {
+      "epoch": 0.7582417582417582,
+      "grad_norm": 0.4224836528301239,
+      "learning_rate": 1.7777308480449006e-05,
+      "loss": 1.010443091392517,
+      "step": 414
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 0.16301847994327545,
+      "learning_rate": 1.7753824782863827e-05,
+      "loss": 1.2712624073028564,
+      "step": 416
+    },
+    {
+      "epoch": 0.7655677655677655,
+      "grad_norm": 0.1754055619239807,
+      "learning_rate": 1.773023535290956e-05,
+      "loss": 0.6752095222473145,
+      "step": 418
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 0.53148353099823,
+      "learning_rate": 1.7706540559883066e-05,
+      "loss": 1.2383452653884888,
+      "step": 420
+    },
+    {
+      "epoch": 0.7728937728937729,
+      "grad_norm": 0.21067863702774048,
+      "learning_rate": 1.7682740774730688e-05,
+      "loss": 0.9969592094421387,
+      "step": 422
+    },
+    {
+      "epoch": 0.7765567765567766,
+      "grad_norm": 0.286111444234848,
+      "learning_rate": 1.7658836370042443e-05,
+      "loss": 0.49340909719467163,
+      "step": 424
+    },
+    {
+      "epoch": 0.7802197802197802,
+      "grad_norm": 0.48605793714523315,
+      "learning_rate": 1.7634827720046178e-05,
+      "loss": 0.8091546893119812,
+      "step": 426
+    },
+    {
+      "epoch": 0.7838827838827839,
+      "grad_norm": 0.5086367130279541,
+      "learning_rate": 1.7610715200601727e-05,
+      "loss": 1.0666477680206299,
+      "step": 428
+    },
+    {
+      "epoch": 0.7875457875457875,
+      "grad_norm": 0.3185122311115265,
+      "learning_rate": 1.7586499189195016e-05,
+      "loss": 1.2098157405853271,
+      "step": 430
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 0.23432797193527222,
+      "learning_rate": 1.7562180064932158e-05,
+      "loss": 1.291565179824829,
+      "step": 432
+    },
+    {
+      "epoch": 0.7948717948717948,
+      "grad_norm": 0.19452722370624542,
+      "learning_rate": 1.7537758208533516e-05,
+      "loss": 0.8848323225975037,
+      "step": 434
+    },
+    {
+      "epoch": 0.7985347985347986,
+      "grad_norm": 0.24718429148197174,
+      "learning_rate": 1.7513234002327738e-05,
+      "loss": 0.9670489430427551,
+      "step": 436
+    },
+    {
+      "epoch": 0.8021978021978022,
+      "grad_norm": 1.831937313079834,
+      "learning_rate": 1.748860783024579e-05,
+      "loss": 0.8808807134628296,
+      "step": 438
+    },
+    {
+      "epoch": 0.8058608058608059,
+      "grad_norm": 0.23156145215034485,
+      "learning_rate": 1.746388007781492e-05,
+      "loss": 1.3112839460372925,
+      "step": 440
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 0.1498938649892807,
+      "learning_rate": 1.7439051132152644e-05,
+      "loss": 1.2047144174575806,
+      "step": 442
+    },
+    {
+      "epoch": 0.8131868131868132,
+      "grad_norm": 0.21056711673736572,
+      "learning_rate": 1.741412138196067e-05,
+      "loss": 1.2016336917877197,
+      "step": 444
+    },
+    {
+      "epoch": 0.8168498168498168,
+      "grad_norm": 0.5614007115364075,
+      "learning_rate": 1.738909121751882e-05,
+      "loss": 1.227624773979187,
+      "step": 446
+    },
+    {
+      "epoch": 0.8205128205128205,
+      "grad_norm": 0.5933297872543335,
+      "learning_rate": 1.736396103067893e-05,
+      "loss": 1.2266818284988403,
+      "step": 448
+    },
+    {
+      "epoch": 0.8241758241758241,
+      "grad_norm": 0.2088741809129715,
+      "learning_rate": 1.7338731214858688e-05,
+      "loss": 1.3620444536209106,
+      "step": 450
+    },
+    {
+      "epoch": 0.8278388278388278,
+      "grad_norm": 0.30190014839172363,
+      "learning_rate": 1.7313402165035504e-05,
+      "loss": 0.9975396990776062,
+      "step": 452
+    },
+    {
+      "epoch": 0.8315018315018315,
+      "grad_norm": 0.07316415756940842,
+      "learning_rate": 1.728797427774031e-05,
+      "loss": 0.49027442932128906,
+      "step": 454
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 0.08831637352705002,
+      "learning_rate": 1.7262447951051366e-05,
+      "loss": 0.9005019664764404,
+      "step": 456
+    },
+    {
+      "epoch": 0.8388278388278388,
+      "grad_norm": 0.23266173899173737,
+      "learning_rate": 1.7236823584587995e-05,
+      "loss": 0.8408886790275574,
+      "step": 458
+    },
+    {
+      "epoch": 0.8424908424908425,
+      "grad_norm": 0.7009871006011963,
+      "learning_rate": 1.7211101579504382e-05,
+      "loss": 1.0286332368850708,
+      "step": 460
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 0.4962588846683502,
+      "learning_rate": 1.7185282338483243e-05,
+      "loss": 1.2270429134368896,
+      "step": 462
+    },
+    {
+      "epoch": 0.8498168498168498,
+      "grad_norm": 0.3354407548904419,
+      "learning_rate": 1.7159366265729537e-05,
+      "loss": 1.1797465085983276,
+      "step": 464
+    },
+    {
+      "epoch": 0.8534798534798534,
+      "grad_norm": 0.07982466369867325,
+      "learning_rate": 1.713335376696416e-05,
+      "loss": 1.201880693435669,
+      "step": 466
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 0.6047555208206177,
+      "learning_rate": 1.7107245249417556e-05,
+      "loss": 0.8821382522583008,
+      "step": 468
+    },
+    {
+      "epoch": 0.8608058608058609,
+      "grad_norm": 0.7960102558135986,
+      "learning_rate": 1.7081041121823375e-05,
+      "loss": 0.9151496887207031,
+      "step": 470
+    },
+    {
+      "epoch": 0.8644688644688645,
+      "grad_norm": 0.1812516152858734,
+      "learning_rate": 1.705474179441205e-05,
+      "loss": 1.1692742109298706,
+      "step": 472
+    },
+    {
+      "epoch": 0.8681318681318682,
+      "grad_norm": 0.09980528056621552,
+      "learning_rate": 1.7028347678904388e-05,
+      "loss": 0.8694528937339783,
+      "step": 474
+    },
+    {
+      "epoch": 0.8717948717948718,
+      "grad_norm": 0.34480732679367065,
+      "learning_rate": 1.700185918850512e-05,
+      "loss": 1.0994765758514404,
+      "step": 476
+    },
+    {
+      "epoch": 0.8754578754578755,
+      "grad_norm": 0.35163024067878723,
+      "learning_rate": 1.6975276737896443e-05,
+      "loss": 1.0502619743347168,
+      "step": 478
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 0.42012107372283936,
+      "learning_rate": 1.69486007432315e-05,
+      "loss": 1.0646753311157227,
+      "step": 480
+    },
+    {
+      "epoch": 0.8827838827838828,
+      "grad_norm": 0.39207834005355835,
+      "learning_rate": 1.6921831622127905e-05,
+      "loss": 1.1882476806640625,
+      "step": 482
+    },
+    {
+      "epoch": 0.8864468864468864,
+      "grad_norm": 0.34004685282707214,
+      "learning_rate": 1.6894969793661163e-05,
+      "loss": 1.263951063156128,
+      "step": 484
+    },
+    {
+      "epoch": 0.8901098901098901,
+      "grad_norm": 0.34993913769721985,
+      "learning_rate": 1.686801567835814e-05,
+      "loss": 0.902937650680542,
+      "step": 486
+    },
+    {
+      "epoch": 0.8937728937728938,
+      "grad_norm": 0.5401197075843811,
+      "learning_rate": 1.6840969698190467e-05,
+      "loss": 1.1535415649414062,
+      "step": 488
+    },
+    {
+      "epoch": 0.8974358974358975,
+      "grad_norm": 0.23730972409248352,
+      "learning_rate": 1.6813832276567942e-05,
+      "loss": 1.1057976484298706,
+      "step": 490
+    },
+    {
+      "epoch": 0.9010989010989011,
+      "grad_norm": 0.32215645909309387,
+      "learning_rate": 1.6786603838331894e-05,
+      "loss": 1.048466444015503,
+      "step": 492
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 0.22649593651294708,
+      "learning_rate": 1.6759284809748522e-05,
+      "loss": 0.5831862092018127,
+      "step": 494
+    },
+    {
+      "epoch": 0.9084249084249084,
+      "grad_norm": 0.2719734311103821,
+      "learning_rate": 1.673187561850225e-05,
+      "loss": 1.2790480852127075,
+      "step": 496
+    },
+    {
+      "epoch": 0.9120879120879121,
+      "grad_norm": 0.11034264415502548,
+      "learning_rate": 1.6704376693689003e-05,
+      "loss": 1.1298277378082275,
+      "step": 498
+    },
+    {
+      "epoch": 0.9157509157509157,
+      "grad_norm": 0.21742600202560425,
+      "learning_rate": 1.6676788465809506e-05,
+      "loss": 0.8114557862281799,
+      "step": 500
+    },
+    {
+      "epoch": 0.9194139194139194,
+      "grad_norm": 0.05262337997555733,
+      "learning_rate": 1.6649111366762552e-05,
+      "loss": 0.8611850738525391,
+      "step": 502
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 0.3120154142379761,
+      "learning_rate": 1.66213458298382e-05,
+      "loss": 0.9393185973167419,
+      "step": 504
+    },
+    {
+      "epoch": 0.9267399267399268,
+      "grad_norm": 0.21198345720767975,
+      "learning_rate": 1.659349228971105e-05,
+      "loss": 0.8434366583824158,
+      "step": 506
+    },
+    {
+      "epoch": 0.9304029304029304,
+      "grad_norm": 0.2922159731388092,
+      "learning_rate": 1.6565551182433382e-05,
+      "loss": 1.1591928005218506,
+      "step": 508
+    },
+    {
+      "epoch": 0.9340659340659341,
+      "grad_norm": 0.273187518119812,
+      "learning_rate": 1.6537522945428386e-05,
+      "loss": 1.2172907590866089,
+      "step": 510
+    },
+    {
+      "epoch": 0.9377289377289377,
+      "grad_norm": 0.3839268088340759,
+      "learning_rate": 1.6509408017483258e-05,
+      "loss": 1.1717743873596191,
+      "step": 512
+    },
+    {
+      "epoch": 0.9413919413919414,
+      "grad_norm": 0.1463819146156311,
+      "learning_rate": 1.6481206838742362e-05,
+      "loss": 0.9574457406997681,
+      "step": 514
+    },
+    {
+      "epoch": 0.945054945054945,
+      "grad_norm": 0.18058021366596222,
+      "learning_rate": 1.645291985070034e-05,
+      "loss": 1.1745246648788452,
+      "step": 516
+    },
+    {
+      "epoch": 0.9487179487179487,
+      "grad_norm": 0.13750861585140228,
+      "learning_rate": 1.64245474961952e-05,
+      "loss": 1.2134760618209839,
+      "step": 518
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 0.22304584085941315,
+      "learning_rate": 1.639609021940136e-05,
+      "loss": 1.2141666412353516,
+      "step": 520
+    },
+    {
+      "epoch": 0.9560439560439561,
+      "grad_norm": 0.17950168251991272,
+      "learning_rate": 1.6367548465822723e-05,
+      "loss": 0.8784098625183105,
+      "step": 522
+    },
+    {
+      "epoch": 0.9597069597069597,
+      "grad_norm": 0.3971985876560211,
+      "learning_rate": 1.6338922682285697e-05,
+      "loss": 1.0205121040344238,
+      "step": 524
+    },
+    {
+      "epoch": 0.9633699633699634,
+      "grad_norm": 0.21712158620357513,
+      "learning_rate": 1.6310213316932187e-05,
+      "loss": 0.9632675051689148,
+      "step": 526
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 0.28583937883377075,
+      "learning_rate": 1.6281420819212578e-05,
+      "loss": 0.647495448589325,
+      "step": 528
+    },
+    {
+      "epoch": 0.9706959706959707,
+      "grad_norm": 0.16611649096012115,
+      "learning_rate": 1.6252545639878728e-05,
+      "loss": 0.9034907817840576,
+      "step": 530
+    },
+    {
+      "epoch": 0.9743589743589743,
+      "grad_norm": 0.5935060977935791,
+      "learning_rate": 1.6223588230976874e-05,
+      "loss": 1.3511052131652832,
+      "step": 532
+    },
+    {
+      "epoch": 0.978021978021978,
+      "grad_norm": 0.13882240653038025,
+      "learning_rate": 1.6194549045840582e-05,
+      "loss": 0.6054647564888,
+      "step": 534
+    },
+    {
+      "epoch": 0.9816849816849816,
+      "grad_norm": 0.41965755820274353,
+      "learning_rate": 1.616542853908363e-05,
+      "loss": 0.8549352288246155,
+      "step": 536
+    },
+    {
+      "epoch": 0.9853479853479854,
+      "grad_norm": 0.10902281850576401,
+      "learning_rate": 1.6136227166592912e-05,
+      "loss": 0.7982230186462402,
+      "step": 538
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 0.26821738481521606,
+      "learning_rate": 1.6106945385521286e-05,
+      "loss": 1.1226874589920044,
+      "step": 540
+    },
+    {
+      "epoch": 0.9926739926739927,
+      "grad_norm": 0.6362835764884949,
+      "learning_rate": 1.6077583654280416e-05,
+      "loss": 1.169198989868164,
+      "step": 542
+    },
+    {
+      "epoch": 0.9963369963369964,
+      "grad_norm": 0.5359551906585693,
+      "learning_rate": 1.60481424325336e-05,
+      "loss": 1.181970238685608,
+      "step": 544
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.19026517868041992,
+      "learning_rate": 1.6018622181188594e-05,
+      "loss": 1.3597557544708252,
+      "step": 546
+    },
+    {
+      "epoch": 1.0036630036630036,
+      "grad_norm": 0.2745984196662903,
+      "learning_rate": 1.598902336239035e-05,
+      "loss": 0.9415228366851807,
+      "step": 548
+    },
+    {
+      "epoch": 1.0073260073260073,
+      "grad_norm": 0.5006294846534729,
+      "learning_rate": 1.595934643951382e-05,
+      "loss": 1.180120825767517,
+      "step": 550
+    },
+    {
+      "epoch": 1.010989010989011,
+      "grad_norm": 0.08472568541765213,
+      "learning_rate": 1.5929591877156694e-05,
+      "loss": 0.6712950468063354,
+      "step": 552
+    },
+    {
+      "epoch": 1.0146520146520146,
+      "grad_norm": 0.1828070878982544,
+      "learning_rate": 1.5899760141132115e-05,
+      "loss": 1.1309884786605835,
+      "step": 554
+    },
+    {
+      "epoch": 1.0183150183150182,
+      "grad_norm": 0.20885473489761353,
+      "learning_rate": 1.58698516984614e-05,
+      "loss": 1.025935411453247,
+      "step": 556
+    },
+    {
+      "epoch": 1.021978021978022,
+      "grad_norm": 0.15055039525032043,
+      "learning_rate": 1.583986701736672e-05,
+      "loss": 1.2158845663070679,
+      "step": 558
+    },
+    {
+      "epoch": 1.0256410256410255,
+      "grad_norm": 0.10766030102968216,
+      "learning_rate": 1.5809806567263767e-05,
+      "loss": 0.9240056276321411,
+      "step": 560
+    },
+    {
+      "epoch": 1.0293040293040292,
+      "grad_norm": 0.19933700561523438,
+      "learning_rate": 1.577967081875442e-05,
+      "loss": 1.1643272638320923,
+      "step": 562
+    },
+    {
+      "epoch": 1.032967032967033,
+      "grad_norm": 0.5722149014472961,
+      "learning_rate": 1.574946024361936e-05,
+      "loss": 1.150964617729187,
+      "step": 564
+    },
+    {
+      "epoch": 1.0366300366300367,
+      "grad_norm": 0.2388359010219574,
+      "learning_rate": 1.5719175314810706e-05,
+      "loss": 1.0588456392288208,
+      "step": 566
+    },
+    {
+      "epoch": 1.0402930402930404,
+      "grad_norm": 0.2551656663417816,
+      "learning_rate": 1.568881650644458e-05,
+      "loss": 1.045701265335083,
+      "step": 568
+    },
+    {
+      "epoch": 1.043956043956044,
+      "grad_norm": 0.25093504786491394,
+      "learning_rate": 1.565838429379371e-05,
+      "loss": 1.1056462526321411,
+      "step": 570
+    },
+    {
+      "epoch": 1.0476190476190477,
+      "grad_norm": 0.18721139430999756,
+      "learning_rate": 1.5627879153279986e-05,
+      "loss": 1.2383893728256226,
+      "step": 572
+    },
+    {
+      "epoch": 1.0512820512820513,
+      "grad_norm": 0.2834416627883911,
+      "learning_rate": 1.559730156246699e-05,
+      "loss": 1.0736546516418457,
+      "step": 574
+    },
+    {
+      "epoch": 1.054945054945055,
+      "grad_norm": 0.2321881204843521,
+      "learning_rate": 1.5566652000052533e-05,
+      "loss": 1.24500572681427,
+      "step": 576
+    },
+    {
+      "epoch": 1.0586080586080586,
+      "grad_norm": 0.23740430176258087,
+      "learning_rate": 1.553593094586115e-05,
+      "loss": 1.220283031463623,
+      "step": 578
+    },
+    {
+      "epoch": 1.0622710622710623,
+      "grad_norm": 0.24411214888095856,
+      "learning_rate": 1.5505138880836595e-05,
+      "loss": 1.2158234119415283,
+      "step": 580
+    },
+    {
+      "epoch": 1.065934065934066,
+      "grad_norm": 0.26935407519340515,
+      "learning_rate": 1.5474276287034305e-05,
+      "loss": 0.9861366152763367,
+      "step": 582
+    },
+    {
+      "epoch": 1.0695970695970696,
+      "grad_norm": 0.34071117639541626,
+      "learning_rate": 1.544334364761387e-05,
+      "loss": 1.146247386932373,
+      "step": 584
+    },
+    {
+      "epoch": 1.0732600732600732,
+      "grad_norm": 0.1204061508178711,
+      "learning_rate": 1.541234144683144e-05,
+      "loss": 1.0828087329864502,
+      "step": 586
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 0.24158774316310883,
+      "learning_rate": 1.5381270170032173e-05,
+      "loss": 0.8217217922210693,
+      "step": 588
+    },
+    {
+      "epoch": 1.0805860805860805,
+      "grad_norm": 0.3989861309528351,
+      "learning_rate": 1.5350130303642625e-05,
+      "loss": 1.1990596055984497,
+      "step": 590
+    },
+    {
+      "epoch": 1.0842490842490842,
+      "grad_norm": 0.09942612051963806,
+      "learning_rate": 1.5318922335163128e-05,
+      "loss": 1.0216456651687622,
+      "step": 592
+    },
+    {
+      "epoch": 1.0879120879120878,
+      "grad_norm": 0.2642340362071991,
+      "learning_rate": 1.5287646753160174e-05,
+      "loss": 0.9809384942054749,
+      "step": 594
+    },
+    {
+      "epoch": 1.0915750915750915,
+      "grad_norm": 0.33687108755111694,
+      "learning_rate": 1.5256304047258739e-05,
+      "loss": 1.0079659223556519,
+      "step": 596
+    },
+    {
+      "epoch": 1.0952380952380953,
+      "grad_norm": 0.44691038131713867,
+      "learning_rate": 1.522489470813466e-05,
+      "loss": 1.1001455783843994,
+      "step": 598
+    },
+    {
+      "epoch": 1.098901098901099,
+      "grad_norm": 1.3902361392974854,
+      "learning_rate": 1.5193419227506913e-05,
+      "loss": 1.1559927463531494,
+      "step": 600
+    },
+    {
+      "epoch": 1.1025641025641026,
+      "grad_norm": 0.5698482394218445,
+      "learning_rate": 1.5161878098129937e-05,
+      "loss": 0.8155554533004761,
+      "step": 602
+    },
+    {
+      "epoch": 1.1062271062271063,
+      "grad_norm": 0.33973053097724915,
+      "learning_rate": 1.5130271813785908e-05,
+      "loss": 0.9548834562301636,
+      "step": 604
+    },
+    {
+      "epoch": 1.10989010989011,
+      "grad_norm": 0.21200524270534515,
+      "learning_rate": 1.509860086927703e-05,
+      "loss": 0.8916766047477722,
+      "step": 606
+    },
+    {
+      "epoch": 1.1135531135531136,
+      "grad_norm": 0.9633941650390625,
+      "learning_rate": 1.5066865760417757e-05,
+      "loss": 1.2829712629318237,
+      "step": 608
+    },
+    {
+      "epoch": 1.1172161172161172,
+      "grad_norm": 0.3105233907699585,
+      "learning_rate": 1.5035066984027053e-05,
+      "loss": 0.7248431444168091,
+      "step": 610
+    },
+    {
+      "epoch": 1.120879120879121,
+      "grad_norm": 0.8827745318412781,
+      "learning_rate": 1.5003205037920616e-05,
+      "loss": 1.1643881797790527,
+      "step": 612
+    },
+    {
+      "epoch": 1.1245421245421245,
+      "grad_norm": 0.19762255251407623,
+      "learning_rate": 1.497128042090307e-05,
+      "loss": 0.9884153604507446,
+      "step": 614
+    },
+    {
+      "epoch": 1.1282051282051282,
+      "grad_norm": 0.11015181988477707,
+      "learning_rate": 1.493929363276017e-05,
+      "loss": 1.1548609733581543,
+      "step": 616
+    },
+    {
+      "epoch": 1.1318681318681318,
+      "grad_norm": 0.43714115023612976,
+      "learning_rate": 1.4907245174250957e-05,
+      "loss": 0.3965158462524414,
+      "step": 618
+    },
+    {
+      "epoch": 1.1355311355311355,
+      "grad_norm": 0.4128834307193756,
+      "learning_rate": 1.4875135547099953e-05,
+      "loss": 1.0857808589935303,
+      "step": 620
+    },
+    {
+      "epoch": 1.1391941391941391,
+      "grad_norm": 0.36989882588386536,
+      "learning_rate": 1.484296525398927e-05,
+      "loss": 0.7546265125274658,
+      "step": 622
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.32148852944374084,
+      "learning_rate": 1.4810734798550769e-05,
+      "loss": 1.0892126560211182,
+      "step": 624
+    },
+    {
+      "epoch": 1.1465201465201464,
+      "grad_norm": 0.3738671541213989,
+      "learning_rate": 1.4778444685358147e-05,
+      "loss": 1.3703358173370361,
+      "step": 626
+    },
+    {
+      "epoch": 1.15018315018315,
+      "grad_norm": 0.4861363172531128,
+      "learning_rate": 1.4746095419919075e-05,
+      "loss": 0.4386288821697235,
+      "step": 628
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 0.04527602717280388,
+      "learning_rate": 1.4713687508667251e-05,
+      "loss": 1.1120142936706543,
+      "step": 630
+    },
+    {
+      "epoch": 1.1575091575091574,
+      "grad_norm": 0.2540167570114136,
+      "learning_rate": 1.4681221458954484e-05,
+      "loss": 1.0848397016525269,
+      "step": 632
+    },
+    {
+      "epoch": 1.1611721611721613,
+      "grad_norm": 0.40409815311431885,
+      "learning_rate": 1.4648697779042754e-05,
+      "loss": 0.852532148361206,
+      "step": 634
+    },
+    {
+      "epoch": 1.164835164835165,
+      "grad_norm": 0.35430774092674255,
+      "learning_rate": 1.461611697809625e-05,
+      "loss": 0.9712074995040894,
+      "step": 636
+    },
+    {
+      "epoch": 1.1684981684981686,
+      "grad_norm": 0.3518703281879425,
+      "learning_rate": 1.4583479566173401e-05,
+      "loss": 1.1836236715316772,
+      "step": 638
+    },
+    {
+      "epoch": 1.1721611721611722,
+      "grad_norm": 0.08637535572052002,
+      "learning_rate": 1.4550786054218902e-05,
+      "loss": 0.7413387894630432,
+      "step": 640
+    },
+    {
+      "epoch": 1.1758241758241759,
+      "grad_norm": 0.25933629274368286,
+      "learning_rate": 1.4518036954055685e-05,
+      "loss": 1.1189838647842407,
+      "step": 642
+    },
+    {
+      "epoch": 1.1794871794871795,
+      "grad_norm": 0.32789698243141174,
+      "learning_rate": 1.4485232778376945e-05,
+      "loss": 1.0686718225479126,
+      "step": 644
+    },
+    {
+      "epoch": 1.1831501831501832,
+      "grad_norm": 0.17057186365127563,
+      "learning_rate": 1.4452374040738078e-05,
+      "loss": 0.7557857632637024,
+      "step": 646
+    },
+    {
+      "epoch": 1.1868131868131868,
+      "grad_norm": 0.3353375494480133,
+      "learning_rate": 1.4419461255548666e-05,
+      "loss": 0.8876537084579468,
+      "step": 648
+    },
+    {
+      "epoch": 1.1904761904761905,
+      "grad_norm": 0.22652913630008698,
+      "learning_rate": 1.4386494938064417e-05,
+      "loss": 1.1409285068511963,
+      "step": 650
+    },
+    {
+      "epoch": 1.1941391941391941,
+      "grad_norm": 0.1531185358762741,
+      "learning_rate": 1.4353475604379093e-05,
+      "loss": 0.671623170375824,
+      "step": 652
+    },
+    {
+      "epoch": 1.1978021978021978,
+      "grad_norm": 0.4093690812587738,
+      "learning_rate": 1.4320403771416438e-05,
+      "loss": 1.3042771816253662,
+      "step": 654
+    },
+    {
+      "epoch": 1.2014652014652014,
+      "grad_norm": 1.0563490390777588,
+      "learning_rate": 1.4287279956922076e-05,
+      "loss": 0.8223202228546143,
+      "step": 656
+    },
+    {
+      "epoch": 1.205128205128205,
+      "grad_norm": 0.4100596606731415,
+      "learning_rate": 1.4254104679455416e-05,
+      "loss": 0.74940425157547,
+      "step": 658
+    },
+    {
+      "epoch": 1.2087912087912087,
+      "grad_norm": 0.6579332947731018,
+      "learning_rate": 1.4220878458381523e-05,
+      "loss": 1.1517488956451416,
+      "step": 660
+    },
+    {
+      "epoch": 1.2124542124542124,
+      "grad_norm": 0.9731626510620117,
+      "learning_rate": 1.418760181386301e-05,
+      "loss": 1.026989459991455,
+      "step": 662
+    },
+    {
+      "epoch": 1.2161172161172162,
+      "grad_norm": 0.30989789962768555,
+      "learning_rate": 1.4154275266851856e-05,
+      "loss": 0.8894543051719666,
+      "step": 664
+    },
+    {
+      "epoch": 1.2197802197802199,
+      "grad_norm": 0.23721382021903992,
+      "learning_rate": 1.4120899339081291e-05,
+      "loss": 0.8070454001426697,
+      "step": 666
+    },
+    {
+      "epoch": 1.2234432234432235,
+      "grad_norm": 0.2819421887397766,
+      "learning_rate": 1.4087474553057599e-05,
+      "loss": 0.8887211084365845,
+      "step": 668
+    },
+    {
+      "epoch": 1.2271062271062272,
+      "grad_norm": 0.27606913447380066,
+      "learning_rate": 1.405400143205195e-05,
+      "loss": 0.675251305103302,
+      "step": 670
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 0.23686662316322327,
+      "learning_rate": 1.4020480500092217e-05,
+      "loss": 1.177452802658081,
+      "step": 672
+    },
+    {
+      "epoch": 1.2344322344322345,
+      "grad_norm": 0.32350587844848633,
+      "learning_rate": 1.3986912281954745e-05,
+      "loss": 1.1530712842941284,
+      "step": 674
+    },
+    {
+      "epoch": 1.2380952380952381,
+      "grad_norm": 0.1811443716287613,
+      "learning_rate": 1.3953297303156174e-05,
+      "loss": 1.1759916543960571,
+      "step": 676
+    },
+    {
+      "epoch": 1.2417582417582418,
+      "grad_norm": 0.19516615569591522,
+      "learning_rate": 1.391963608994517e-05,
+      "loss": 0.7716677784919739,
+      "step": 678
+    },
+    {
+      "epoch": 1.2454212454212454,
+      "grad_norm": 0.19349494576454163,
+      "learning_rate": 1.3885929169294218e-05,
+      "loss": 0.8281112313270569,
+      "step": 680
+    },
+    {
+      "epoch": 1.249084249084249,
+      "grad_norm": 0.9416590929031372,
+      "learning_rate": 1.3852177068891364e-05,
+      "loss": 1.1924660205841064,
+      "step": 682
+    },
+    {
+      "epoch": 1.2527472527472527,
+      "grad_norm": 0.26466265320777893,
+      "learning_rate": 1.3818380317131946e-05,
+      "loss": 1.1589021682739258,
+      "step": 684
+    },
+    {
+      "epoch": 1.2564102564102564,
+      "grad_norm": 0.24540755152702332,
+      "learning_rate": 1.3784539443110323e-05,
+      "loss": 0.8185974359512329,
+      "step": 686
+    },
+    {
+      "epoch": 1.26007326007326,
+      "grad_norm": 0.23551802337169647,
+      "learning_rate": 1.375065497661161e-05,
+      "loss": 1.111383318901062,
+      "step": 688
+    },
+    {
+      "epoch": 1.2637362637362637,
+      "grad_norm": 0.6621056199073792,
+      "learning_rate": 1.3716727448103356e-05,
+      "loss": 1.1371651887893677,
+      "step": 690
+    },
+    {
+      "epoch": 1.2673992673992673,
+      "grad_norm": 0.16852295398712158,
+      "learning_rate": 1.3682757388727261e-05,
+      "loss": 1.1790460348129272,
+      "step": 692
+    },
+    {
+      "epoch": 1.271062271062271,
+      "grad_norm": 0.7090398073196411,
+      "learning_rate": 1.3648745330290848e-05,
+      "loss": 0.9768311381340027,
+      "step": 694
+    },
+    {
+      "epoch": 1.2747252747252746,
+      "grad_norm": 0.5497918128967285,
+      "learning_rate": 1.361469180525916e-05,
+      "loss": 1.0126756429672241,
+      "step": 696
+    },
+    {
+      "epoch": 1.2783882783882783,
+      "grad_norm": 0.31008267402648926,
+      "learning_rate": 1.358059734674638e-05,
+      "loss": 0.9194446802139282,
+      "step": 698
+    },
+    {
+      "epoch": 1.282051282051282,
+      "grad_norm": 0.10667116194963455,
+      "learning_rate": 1.3546462488507532e-05,
+      "loss": 0.6427603960037231,
+      "step": 700
+    },
+    {
+      "epoch": 1.2857142857142856,
+      "grad_norm": 0.2069520652294159,
+      "learning_rate": 1.3512287764930102e-05,
+      "loss": 0.5255513787269592,
+      "step": 702
+    },
+    {
+      "epoch": 1.2893772893772895,
+      "grad_norm": 0.4747946262359619,
+      "learning_rate": 1.347807371102567e-05,
+      "loss": 1.2734118700027466,
+      "step": 704
+    },
+    {
+      "epoch": 1.293040293040293,
+      "grad_norm": 1.2546977996826172,
+      "learning_rate": 1.3443820862421542e-05,
+      "loss": 1.0028886795043945,
+      "step": 706
+    },
+    {
+      "epoch": 1.2967032967032968,
+      "grad_norm": 0.09716126322746277,
+      "learning_rate": 1.3409529755352361e-05,
+      "loss": 0.9534561634063721,
+      "step": 708
+    },
+    {
+      "epoch": 1.3003663003663004,
+      "grad_norm": 0.16677714884281158,
+      "learning_rate": 1.3375200926651719e-05,
+      "loss": 0.5677421689033508,
+      "step": 710
+    },
+    {
+      "epoch": 1.304029304029304,
+      "grad_norm": 0.5648819208145142,
+      "learning_rate": 1.3340834913743742e-05,
+      "loss": 0.8859728574752808,
+      "step": 712
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 0.3466564118862152,
+      "learning_rate": 1.3306432254634676e-05,
+      "loss": 1.2071309089660645,
+      "step": 714
+    },
+    {
+      "epoch": 1.3113553113553114,
+      "grad_norm": 0.34816622734069824,
+      "learning_rate": 1.3271993487904485e-05,
+      "loss": 1.1303997039794922,
+      "step": 716
+    },
+    {
+      "epoch": 1.315018315018315,
+      "grad_norm": 0.17138248682022095,
+      "learning_rate": 1.3237519152698392e-05,
+      "loss": 0.9440135359764099,
+      "step": 718
+    },
+    {
+      "epoch": 1.3186813186813187,
+      "grad_norm": 0.2666979432106018,
+      "learning_rate": 1.3203009788718454e-05,
+      "loss": 0.9132227301597595,
+      "step": 720
+    },
+    {
+      "epoch": 1.3223443223443223,
+      "grad_norm": 0.3015995919704437,
+      "learning_rate": 1.3168465936215114e-05,
+      "loss": 0.9173396229743958,
+      "step": 722
+    },
+    {
+      "epoch": 1.326007326007326,
+      "grad_norm": 0.3288915157318115,
+      "learning_rate": 1.3133888135978733e-05,
+      "loss": 1.2012741565704346,
+      "step": 724
+    },
+    {
+      "epoch": 1.3296703296703296,
+      "grad_norm": 0.1542222797870636,
+      "learning_rate": 1.3099276929331132e-05,
+      "loss": 1.10904061794281,
+      "step": 726
+    },
+    {
+      "epoch": 1.3333333333333333,
+      "grad_norm": 0.22464217245578766,
+      "learning_rate": 1.3064632858117123e-05,
+      "loss": 1.1415292024612427,
+      "step": 728
+    },
+    {
+      "epoch": 1.3369963369963371,
+      "grad_norm": 0.5490389466285706,
+      "learning_rate": 1.3029956464696006e-05,
+      "loss": 0.7423273324966431,
+      "step": 730
+    },
+    {
+      "epoch": 1.3406593406593408,
+      "grad_norm": 0.4176447093486786,
+      "learning_rate": 1.2995248291933099e-05,
+      "loss": 0.8297606110572815,
+      "step": 732
+    },
+    {
+      "epoch": 1.3443223443223444,
+      "grad_norm": 0.05192068964242935,
+      "learning_rate": 1.296050888319123e-05,
+      "loss": 0.6149548292160034,
+      "step": 734
+    },
+    {
+      "epoch": 1.347985347985348,
+      "grad_norm": 0.8712805509567261,
+      "learning_rate": 1.2925738782322232e-05,
+      "loss": 1.1270866394042969,
+      "step": 736
+    },
+    {
+      "epoch": 1.3516483516483517,
+      "grad_norm": 0.20594623684883118,
+      "learning_rate": 1.2890938533658429e-05,
+      "loss": 0.7438766360282898,
+      "step": 738
+    },
+    {
+      "epoch": 1.3553113553113554,
+      "grad_norm": 0.23629705607891083,
+      "learning_rate": 1.2856108682004116e-05,
+      "loss": 1.00837242603302,
+      "step": 740
+    },
+    {
+      "epoch": 1.358974358974359,
+      "grad_norm": 1.347622036933899,
+      "learning_rate": 1.282124977262702e-05,
+      "loss": 0.6100921630859375,
+      "step": 742
+    },
+    {
+      "epoch": 1.3626373626373627,
+      "grad_norm": 0.1946464329957962,
+      "learning_rate": 1.2786362351249785e-05,
+      "loss": 1.2331260442733765,
+      "step": 744
+    },
+    {
+      "epoch": 1.3663003663003663,
+      "grad_norm": 0.2741944193840027,
+      "learning_rate": 1.2751446964041405e-05,
+      "loss": 1.0520378351211548,
+      "step": 746
+    },
+    {
+      "epoch": 1.36996336996337,
+      "grad_norm": 0.18286797404289246,
+      "learning_rate": 1.2716504157608693e-05,
+      "loss": 1.065680742263794,
+      "step": 748
+    },
+    {
+      "epoch": 1.3736263736263736,
+      "grad_norm": 0.47365298867225647,
+      "learning_rate": 1.2681534478987703e-05,
+      "loss": 0.7774507403373718,
+      "step": 750
+    },
+    {
+      "epoch": 1.3772893772893773,
+      "grad_norm": 0.19642506539821625,
+      "learning_rate": 1.264653847563519e-05,
+      "loss": 1.1186778545379639,
+      "step": 752
+    },
+    {
+      "epoch": 1.380952380952381,
+      "grad_norm": 0.12549656629562378,
+      "learning_rate": 1.2611516695420023e-05,
+      "loss": 0.963532567024231,
+      "step": 754
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 0.2738596796989441,
+      "learning_rate": 1.2576469686614608e-05,
+      "loss": 1.3037950992584229,
+      "step": 756
+    },
+    {
+      "epoch": 1.3882783882783882,
+      "grad_norm": 0.265229731798172,
+      "learning_rate": 1.2541397997886317e-05,
+      "loss": 1.2012255191802979,
+      "step": 758
+    },
+    {
+      "epoch": 1.3919413919413919,
+      "grad_norm": 0.11555524170398712,
+      "learning_rate": 1.2506302178288887e-05,
+      "loss": 1.145890712738037,
+      "step": 760
+    },
+    {
+      "epoch": 1.3956043956043955,
+      "grad_norm": 0.33266228437423706,
+      "learning_rate": 1.2471182777253832e-05,
+      "loss": 1.140316367149353,
+      "step": 762
+    },
+    {
+      "epoch": 1.3992673992673992,
+      "grad_norm": 0.371504545211792,
+      "learning_rate": 1.2436040344581824e-05,
+      "loss": 0.6883344650268555,
+      "step": 764
+    },
+    {
+      "epoch": 1.4029304029304028,
+      "grad_norm": 0.26427412033081055,
+      "learning_rate": 1.2400875430434119e-05,
+      "loss": 0.8899670243263245,
+      "step": 766
+    },
+    {
+      "epoch": 1.4065934065934065,
+      "grad_norm": 1.1619846820831299,
+      "learning_rate": 1.236568858532391e-05,
+      "loss": 0.8952260613441467,
+      "step": 768
+    },
+    {
+      "epoch": 1.4102564102564101,
+      "grad_norm": 0.22171390056610107,
+      "learning_rate": 1.2330480360107728e-05,
+      "loss": 1.182429552078247,
+      "step": 770
+    },
+    {
+      "epoch": 1.4139194139194138,
+      "grad_norm": 0.25736895203590393,
+      "learning_rate": 1.2295251305976818e-05,
+      "loss": 1.2108275890350342,
+      "step": 772
+    },
+    {
+      "epoch": 1.4175824175824177,
+      "grad_norm": 0.40875253081321716,
+      "learning_rate": 1.2260001974448504e-05,
+      "loss": 1.021303415298462,
+      "step": 774
+    },
+    {
+      "epoch": 1.4212454212454213,
+      "grad_norm": 0.13845305144786835,
+      "learning_rate": 1.222473291735754e-05,
+      "loss": 1.154242753982544,
+      "step": 776
+    },
+    {
+      "epoch": 1.424908424908425,
+      "grad_norm": 0.18540990352630615,
+      "learning_rate": 1.218944468684752e-05,
+      "loss": 0.8286201357841492,
+      "step": 778
+    },
+    {
+      "epoch": 1.4285714285714286,
+      "grad_norm": 0.18653152883052826,
+      "learning_rate": 1.215413783536217e-05,
+      "loss": 1.3623868227005005,
+      "step": 780
+    },
+    {
+      "epoch": 1.4322344322344323,
+      "grad_norm": 0.531924307346344,
+      "learning_rate": 1.2118812915636744e-05,
+      "loss": 1.2334803342819214,
+      "step": 782
+    },
+    {
+      "epoch": 1.435897435897436,
+      "grad_norm": 3.9369657039642334,
+      "learning_rate": 1.2083470480689363e-05,
+      "loss": 1.1573415994644165,
+      "step": 784
+    },
+    {
+      "epoch": 1.4395604395604396,
+      "grad_norm": 0.710831880569458,
+      "learning_rate": 1.2048111083812342e-05,
+      "loss": 0.9501171112060547,
+      "step": 786
+    },
+    {
+      "epoch": 1.4432234432234432,
+      "grad_norm": 0.5587278008460999,
+      "learning_rate": 1.2012735278563546e-05,
+      "loss": 1.135019063949585,
+      "step": 788
+    },
+    {
+      "epoch": 1.4468864468864469,
+      "grad_norm": 0.06678500026464462,
+      "learning_rate": 1.1977343618757702e-05,
+      "loss": 0.7230968475341797,
+      "step": 790
+    },
+    {
+      "epoch": 1.4505494505494505,
+      "grad_norm": 0.3107512295246124,
+      "learning_rate": 1.1941936658457769e-05,
+      "loss": 0.9185922741889954,
+      "step": 792
+    },
+    {
+      "epoch": 1.4542124542124542,
+      "grad_norm": 0.6749280691146851,
+      "learning_rate": 1.1906514951966208e-05,
+      "loss": 0.8022876977920532,
+      "step": 794
+    },
+    {
+      "epoch": 1.4578754578754578,
+      "grad_norm": 0.17742304503917694,
+      "learning_rate": 1.1871079053816357e-05,
+      "loss": 1.1445965766906738,
+      "step": 796
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 0.25644049048423767,
+      "learning_rate": 1.1835629518763714e-05,
+      "loss": 0.792290985584259,
+      "step": 798
+    },
+    {
+      "epoch": 1.4652014652014653,
+      "grad_norm": 0.275553822517395,
+      "learning_rate": 1.1800166901777272e-05,
+      "loss": 1.024118423461914,
+      "step": 800
+    },
+    {
+      "epoch": 1.468864468864469,
+      "grad_norm": 0.7942085266113281,
+      "learning_rate": 1.1764691758030825e-05,
+      "loss": 1.5601838827133179,
+      "step": 802
+    },
+    {
+      "epoch": 1.4725274725274726,
+      "grad_norm": 0.20729446411132812,
+      "learning_rate": 1.1729204642894265e-05,
+      "loss": 1.0184231996536255,
+      "step": 804
+    },
+    {
+      "epoch": 1.4761904761904763,
+      "grad_norm": 0.6443558931350708,
+      "learning_rate": 1.1693706111924912e-05,
+      "loss": 1.1878175735473633,
+      "step": 806
+    },
+    {
+      "epoch": 1.47985347985348,
+      "grad_norm": 0.24344100058078766,
+      "learning_rate": 1.1658196720858794e-05,
+      "loss": 1.1712795495986938,
+      "step": 808
+    },
+    {
+      "epoch": 1.4835164835164836,
+      "grad_norm": 0.5350912809371948,
+      "learning_rate": 1.1622677025601966e-05,
+      "loss": 0.998128354549408,
+      "step": 810
+    },
+    {
+      "epoch": 1.4871794871794872,
+      "grad_norm": 0.22139771282672882,
+      "learning_rate": 1.1587147582221776e-05,
+      "loss": 0.8508384823799133,
+      "step": 812
+    },
+    {
+      "epoch": 1.4908424908424909,
+      "grad_norm": 0.23050732910633087,
+      "learning_rate": 1.1551608946938208e-05,
+      "loss": 1.2045676708221436,
+      "step": 814
+    },
+    {
+      "epoch": 1.4945054945054945,
+      "grad_norm": 0.22233106195926666,
+      "learning_rate": 1.1516061676115124e-05,
+      "loss": 1.1350584030151367,
+      "step": 816
+    },
+    {
+      "epoch": 1.4981684981684982,
+      "grad_norm": 0.09867780655622482,
+      "learning_rate": 1.1480506326251595e-05,
+      "loss": 0.4009166657924652,
+      "step": 818
+    },
+    {
+      "epoch": 1.5018315018315018,
+      "grad_norm": 0.19808466732501984,
+      "learning_rate": 1.1444943453973155e-05,
+      "loss": 1.1480183601379395,
+      "step": 820
+    },
+    {
+      "epoch": 1.5054945054945055,
+      "grad_norm": 0.2406286746263504,
+      "learning_rate": 1.1409373616023111e-05,
+      "loss": 0.9088360071182251,
+      "step": 822
+    },
+    {
+      "epoch": 1.5091575091575091,
+      "grad_norm": 0.45124879479408264,
+      "learning_rate": 1.1373797369253818e-05,
+      "loss": 0.8226873874664307,
+      "step": 824
+    },
+    {
+      "epoch": 1.5128205128205128,
+      "grad_norm": 0.2533392310142517,
+      "learning_rate": 1.1338215270617967e-05,
+      "loss": 1.046976089477539,
+      "step": 826
+    },
+    {
+      "epoch": 1.5164835164835164,
+      "grad_norm": 0.9905543327331543,
+      "learning_rate": 1.130262787715985e-05,
+      "loss": 1.0114374160766602,
+      "step": 828
+    },
+    {
+      "epoch": 1.52014652014652,
+      "grad_norm": 0.423899382352829,
+      "learning_rate": 1.1267035746006658e-05,
+      "loss": 1.003442406654358,
+      "step": 830
+    },
+    {
+      "epoch": 1.5238095238095237,
+      "grad_norm": 0.5581493377685547,
+      "learning_rate": 1.1231439434359755e-05,
+      "loss": 1.2469056844711304,
+      "step": 832
+    },
+    {
+      "epoch": 1.5274725274725274,
+      "grad_norm": 0.25180450081825256,
+      "learning_rate": 1.119583949948594e-05,
+      "loss": 0.8143158555030823,
+      "step": 834
+    },
+    {
+      "epoch": 1.531135531135531,
+      "grad_norm": 0.8289534449577332,
+      "learning_rate": 1.1160236498708742e-05,
+      "loss": 1.1500300168991089,
+      "step": 836
+    },
+    {
+      "epoch": 1.5347985347985347,
+      "grad_norm": 0.6642610430717468,
+      "learning_rate": 1.112463098939969e-05,
+      "loss": 1.2071179151535034,
+      "step": 838
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 0.41747331619262695,
+      "learning_rate": 1.1089023528969576e-05,
+      "loss": 1.0496968030929565,
+      "step": 840
+    },
+    {
+      "epoch": 1.542124542124542,
+      "grad_norm": 0.20446696877479553,
+      "learning_rate": 1.1053414674859741e-05,
+      "loss": 0.8783029913902283,
+      "step": 842
+    },
+    {
+      "epoch": 1.5457875457875456,
+      "grad_norm": 0.3679927885532379,
+      "learning_rate": 1.1017804984533351e-05,
+      "loss": 1.0342862606048584,
+      "step": 844
+    },
+    {
+      "epoch": 1.5494505494505495,
+      "grad_norm": 0.16902367770671844,
+      "learning_rate": 1.0982195015466652e-05,
+      "loss": 1.1826444864273071,
+      "step": 846
+    },
+    {
+      "epoch": 1.5531135531135531,
+      "grad_norm": 1.2264890670776367,
+      "learning_rate": 1.0946585325140261e-05,
+      "loss": 0.5550611615180969,
+      "step": 848
+    },
+    {
+      "epoch": 1.5567765567765568,
+      "grad_norm": 0.5985056757926941,
+      "learning_rate": 1.0910976471030428e-05,
+      "loss": 1.0341655015945435,
+      "step": 850
+    },
+    {
+      "epoch": 1.5604395604395604,
+      "grad_norm": 0.19666093587875366,
+      "learning_rate": 1.0875369010600317e-05,
+      "loss": 1.1766899824142456,
+      "step": 852
+    },
+    {
+      "epoch": 1.564102564102564,
+      "grad_norm": 0.6608942151069641,
+      "learning_rate": 1.083976350129126e-05,
+      "loss": 0.8616961240768433,
+      "step": 854
+    },
+    {
+      "epoch": 1.5677655677655677,
+      "grad_norm": 0.24679680168628693,
+      "learning_rate": 1.0804160500514062e-05,
+      "loss": 0.809736430644989,
+      "step": 856
+    },
+    {
+      "epoch": 1.5714285714285714,
+      "grad_norm": 0.9515239596366882,
+      "learning_rate": 1.0768560565640252e-05,
+      "loss": 0.9475131630897522,
+      "step": 858
+    },
+    {
+      "epoch": 1.575091575091575,
+      "grad_norm": 0.8901296257972717,
+      "learning_rate": 1.0732964253993343e-05,
+      "loss": 0.7505902051925659,
+      "step": 860
+    },
+    {
+      "epoch": 1.578754578754579,
+      "grad_norm": 0.2663470506668091,
+      "learning_rate": 1.0697372122840156e-05,
+      "loss": 1.1575169563293457,
+      "step": 862
+    },
+    {
+      "epoch": 1.5824175824175826,
+      "grad_norm": 1.596577525138855,
+      "learning_rate": 1.0661784729382036e-05,
+      "loss": 0.9081068634986877,
+      "step": 864
+    },
+    {
+      "epoch": 1.5860805860805862,
+      "grad_norm": 0.19321048259735107,
+      "learning_rate": 1.0626202630746183e-05,
+      "loss": 1.1882286071777344,
+      "step": 866
+    },
+    {
+      "epoch": 1.5897435897435899,
+      "grad_norm": 0.2287406027317047,
+      "learning_rate": 1.0590626383976894e-05,
+      "loss": 1.1828335523605347,
+      "step": 868
+    },
+    {
+      "epoch": 1.5934065934065935,
+      "grad_norm": 0.8435899019241333,
+      "learning_rate": 1.055505654602685e-05,
+      "loss": 0.587890625,
+      "step": 870
+    },
+    {
+      "epoch": 1.5970695970695972,
+      "grad_norm": 0.290306955575943,
+      "learning_rate": 1.0519493673748406e-05,
+      "loss": 1.1838449239730835,
+      "step": 872
+    },
+    {
+      "epoch": 1.6007326007326008,
+      "grad_norm": 0.3988237679004669,
+      "learning_rate": 1.0483938323884879e-05,
+      "loss": 1.066159963607788,
+      "step": 874
+    },
+    {
+      "epoch": 1.6043956043956045,
+      "grad_norm": 0.3236004114151001,
+      "learning_rate": 1.0448391053061795e-05,
+      "loss": 0.7672710418701172,
+      "step": 876
+    },
+    {
+      "epoch": 1.6080586080586081,
+      "grad_norm": 0.41797590255737305,
+      "learning_rate": 1.0412852417778225e-05,
+      "loss": 0.994400143623352,
+      "step": 878
+    },
+    {
+      "epoch": 1.6117216117216118,
+      "grad_norm": 0.3434767425060272,
+      "learning_rate": 1.037732297439804e-05,
+      "loss": 0.9163567423820496,
+      "step": 880
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 0.4138431251049042,
+      "learning_rate": 1.034180327914121e-05,
+      "loss": 1.1349215507507324,
+      "step": 882
+    },
+    {
+      "epoch": 1.619047619047619,
+      "grad_norm": 1.8183345794677734,
+      "learning_rate": 1.030629388807509e-05,
+      "loss": 0.945977509021759,
+      "step": 884
+    },
+    {
+      "epoch": 1.6227106227106227,
+      "grad_norm": 0.15837231278419495,
+      "learning_rate": 1.0270795357105738e-05,
+      "loss": 1.1418633460998535,
+      "step": 886
+    },
+    {
+      "epoch": 1.6263736263736264,
+      "grad_norm": 0.18444959819316864,
+      "learning_rate": 1.023530824196918e-05,
+      "loss": 0.8405629396438599,
+      "step": 888
+    },
+    {
+      "epoch": 1.63003663003663,
+      "grad_norm": 0.7059314250946045,
+      "learning_rate": 1.019983309822273e-05,
+      "loss": 1.2206584215164185,
+      "step": 890
+    },
+    {
+      "epoch": 1.6336996336996337,
+      "grad_norm": 5.118255615234375,
+      "learning_rate": 1.0164370481236292e-05,
+      "loss": 0.9084610939025879,
+      "step": 892
+    },
+    {
+      "epoch": 1.6373626373626373,
+      "grad_norm": 0.31705963611602783,
+      "learning_rate": 1.0128920946183646e-05,
+      "loss": 1.1889030933380127,
+      "step": 894
+    },
+    {
+      "epoch": 1.641025641025641,
+      "grad_norm": 0.30195608735084534,
+      "learning_rate": 1.0093485048033798e-05,
+      "loss": 0.7536079287528992,
+      "step": 896
+    },
+    {
+      "epoch": 1.6446886446886446,
+      "grad_norm": 0.5159720778465271,
+      "learning_rate": 1.0058063341542238e-05,
+      "loss": 1.1412334442138672,
+      "step": 898
+    },
+    {
+      "epoch": 1.6483516483516483,
+      "grad_norm": 0.37968581914901733,
+      "learning_rate": 1.0022656381242297e-05,
+      "loss": 0.8290563821792603,
+      "step": 900
+    },
+    {
+      "epoch": 1.652014652014652,
+      "grad_norm": 0.21596157550811768,
+      "learning_rate": 9.98726472143646e-06,
+      "loss": 0.8873069286346436,
+      "step": 902
+    },
+    {
+      "epoch": 1.6556776556776556,
+      "grad_norm": 0.17106634378433228,
+      "learning_rate": 9.951888916187662e-06,
+      "loss": 1.0214476585388184,
+      "step": 904
+    },
+    {
+      "epoch": 1.6593406593406592,
+      "grad_norm": 0.31868913769721985,
+      "learning_rate": 9.916529519310638e-06,
+      "loss": 1.139350414276123,
+      "step": 906
+    },
+    {
+      "epoch": 1.6630036630036629,
+      "grad_norm": 0.19221165776252747,
+      "learning_rate": 9.881187084363257e-06,
+      "loss": 0.7617915868759155,
+      "step": 908
+    },
+    {
+      "epoch": 1.6666666666666665,
+      "grad_norm": 12.817173957824707,
+      "learning_rate": 9.845862164637834e-06,
+      "loss": 1.206006646156311,
+      "step": 910
+    },
+    {
+      "epoch": 1.6703296703296702,
+      "grad_norm": 0.4153216481208801,
+      "learning_rate": 9.810555313152486e-06,
+      "loss": 1.2200528383255005,
+      "step": 912
+    },
+    {
+      "epoch": 1.673992673992674,
+      "grad_norm": 0.33920755982398987,
+      "learning_rate": 9.775267082642461e-06,
+      "loss": 1.0979692935943604,
+      "step": 914
+    },
+    {
+      "epoch": 1.6776556776556777,
+      "grad_norm": 0.4049103260040283,
+      "learning_rate": 9.7399980255515e-06,
+      "loss": 1.2116343975067139,
+      "step": 916
+    },
+    {
+      "epoch": 1.6813186813186813,
+      "grad_norm": 0.2901177704334259,
+      "learning_rate": 9.704748694023183e-06,
+      "loss": 0.8435489535331726,
+      "step": 918
+    },
+    {
+      "epoch": 1.684981684981685,
+      "grad_norm": 0.29318225383758545,
+      "learning_rate": 9.669519639892275e-06,
+      "loss": 1.2188955545425415,
+      "step": 920
+    },
+    {
+      "epoch": 1.6886446886446886,
+      "grad_norm": 0.13588818907737732,
+      "learning_rate": 9.634311414676096e-06,
+      "loss": 1.0209497213363647,
+      "step": 922
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 0.16762422025203705,
+      "learning_rate": 9.599124569565887e-06,
+      "loss": 0.851481556892395,
+      "step": 924
+    },
+    {
+      "epoch": 1.695970695970696,
+      "grad_norm": 0.2845093309879303,
+      "learning_rate": 9.56395965541818e-06,
+      "loss": 0.9773625135421753,
+      "step": 926
+    },
+    {
+      "epoch": 1.6996336996336996,
+      "grad_norm": 0.2550683617591858,
+      "learning_rate": 9.528817222746171e-06,
+      "loss": 0.8091166615486145,
+      "step": 928
+    },
+    {
+      "epoch": 1.7032967032967035,
+      "grad_norm": 0.030724653974175453,
+      "learning_rate": 9.493697821711116e-06,
+      "loss": 0.8957905173301697,
+      "step": 930
+    },
+    {
+      "epoch": 1.7069597069597071,
+      "grad_norm": 0.6097201704978943,
+      "learning_rate": 9.458602002113684e-06,
+      "loss": 0.936105489730835,
+      "step": 932
+    },
+    {
+      "epoch": 1.7106227106227108,
+      "grad_norm": 0.16642826795578003,
+      "learning_rate": 9.423530313385395e-06,
+      "loss": 1.3963714838027954,
+      "step": 934
+    },
+    {
+      "epoch": 1.7142857142857144,
+      "grad_norm": 0.23555639386177063,
+      "learning_rate": 9.388483304579983e-06,
+      "loss": 1.195056676864624,
+      "step": 936
+    },
+    {
+      "epoch": 1.717948717948718,
+      "grad_norm": 0.2510616183280945,
+      "learning_rate": 9.353461524364814e-06,
+      "loss": 0.48157253861427307,
+      "step": 938
+    },
+    {
+      "epoch": 1.7216117216117217,
+      "grad_norm": 0.1841297298669815,
+      "learning_rate": 9.318465521012298e-06,
+      "loss": 0.5561771988868713,
+      "step": 940
+    },
+    {
+      "epoch": 1.7252747252747254,
+      "grad_norm": 0.24431364238262177,
+      "learning_rate": 9.283495842391313e-06,
+      "loss": 1.127332329750061,
+      "step": 942
+    },
+    {
+      "epoch": 1.728937728937729,
+      "grad_norm": 0.34080448746681213,
+      "learning_rate": 9.248553035958596e-06,
+      "loss": 0.931099534034729,
+      "step": 944
+    },
+    {
+      "epoch": 1.7326007326007327,
+      "grad_norm": 0.25586462020874023,
+      "learning_rate": 9.213637648750217e-06,
+      "loss": 1.1583143472671509,
+      "step": 946
+    },
+    {
+      "epoch": 1.7362637362637363,
+      "grad_norm": 0.17189502716064453,
+      "learning_rate": 9.178750227372983e-06,
+      "loss": 1.1040928363800049,
+      "step": 948
+    },
+    {
+      "epoch": 1.73992673992674,
+      "grad_norm": 0.21041589975357056,
+      "learning_rate": 9.143891317995888e-06,
+      "loss": 0.9976288676261902,
+      "step": 950
+    },
+    {
+      "epoch": 1.7435897435897436,
+      "grad_norm": 0.2621840238571167,
+      "learning_rate": 9.109061466341576e-06,
+      "loss": 0.9492763876914978,
+      "step": 952
+    },
+    {
+      "epoch": 1.7472527472527473,
+      "grad_norm": 1.197185754776001,
+      "learning_rate": 9.074261217677771e-06,
+      "loss": 1.2083392143249512,
+      "step": 954
+    },
+    {
+      "epoch": 1.750915750915751,
+      "grad_norm": 0.19799767434597015,
+      "learning_rate": 9.039491116808773e-06,
+      "loss": 0.7902243137359619,
+      "step": 956
+    },
+    {
+      "epoch": 1.7545787545787546,
+      "grad_norm": 0.4546964466571808,
+      "learning_rate": 9.004751708066906e-06,
+      "loss": 1.2029194831848145,
+      "step": 958
+    },
+    {
+      "epoch": 1.7582417582417582,
+      "grad_norm": 0.13221463561058044,
+      "learning_rate": 8.970043535303999e-06,
+      "loss": 0.5475955009460449,
+      "step": 960
+    },
+    {
+      "epoch": 1.7619047619047619,
+      "grad_norm": 0.378535658121109,
+      "learning_rate": 8.93536714188288e-06,
+      "loss": 0.9881755709648132,
+      "step": 962
+    },
+    {
+      "epoch": 1.7655677655677655,
+      "grad_norm": 0.2644976079463959,
+      "learning_rate": 8.900723070668869e-06,
+      "loss": 1.0287450551986694,
+      "step": 964
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 0.6430861353874207,
+      "learning_rate": 8.86611186402127e-06,
+      "loss": 1.1860679388046265,
+      "step": 966
+    },
+    {
+      "epoch": 1.7728937728937728,
+      "grad_norm": 0.4337344467639923,
+      "learning_rate": 8.831534063784891e-06,
+      "loss": 0.5775477886199951,
+      "step": 968
+    },
+    {
+      "epoch": 1.7765567765567765,
+      "grad_norm": 0.27584108710289,
+      "learning_rate": 8.796990211281549e-06,
+      "loss": 0.8390186429023743,
+      "step": 970
+    },
+    {
+      "epoch": 1.7802197802197801,
+      "grad_norm": 0.7176196575164795,
+      "learning_rate": 8.76248084730161e-06,
+      "loss": 0.9147000908851624,
+      "step": 972
+    },
+    {
+      "epoch": 1.7838827838827838,
+      "grad_norm": 0.29480162262916565,
+      "learning_rate": 8.728006512095517e-06,
+      "loss": 1.221879243850708,
+      "step": 974
+    },
+    {
+      "epoch": 1.7875457875457874,
+      "grad_norm": 0.2008034586906433,
+      "learning_rate": 8.693567745365325e-06,
+      "loss": 1.1880409717559814,
+      "step": 976
+    },
+    {
+      "epoch": 1.791208791208791,
+      "grad_norm": 0.2755773961544037,
+      "learning_rate": 8.659165086256263e-06,
+      "loss": 0.9174471497535706,
+      "step": 978
+    },
+    {
+      "epoch": 1.7948717948717947,
+      "grad_norm": 0.23839768767356873,
+      "learning_rate": 8.624799073348282e-06,
+      "loss": 0.9608420729637146,
+      "step": 980
+    },
+    {
+      "epoch": 1.7985347985347986,
+      "grad_norm": 0.33076974749565125,
+      "learning_rate": 8.590470244647643e-06,
+      "loss": 1.1425825357437134,
+      "step": 982
+    },
+    {
+      "epoch": 1.8021978021978022,
+      "grad_norm": 1.4269599914550781,
+      "learning_rate": 8.556179137578461e-06,
+      "loss": 1.1003776788711548,
+      "step": 984
+    },
+    {
+      "epoch": 1.8058608058608059,
+      "grad_norm": 0.10697717219591141,
+      "learning_rate": 8.521926288974336e-06,
+      "loss": 0.6553662419319153,
+      "step": 986
+    },
+    {
+      "epoch": 1.8095238095238095,
+      "grad_norm": 0.325648695230484,
+      "learning_rate": 8.487712235069901e-06,
+      "loss": 0.8113435506820679,
+      "step": 988
+    },
+    {
+      "epoch": 1.8131868131868132,
+      "grad_norm": 0.7604925632476807,
+      "learning_rate": 8.453537511492469e-06,
+      "loss": 0.7590834498405457,
+      "step": 990
+    },
+    {
+      "epoch": 1.8168498168498168,
+      "grad_norm": 0.5670424699783325,
+      "learning_rate": 8.419402653253623e-06,
+      "loss": 0.7830116748809814,
+      "step": 992
+    },
+    {
+      "epoch": 1.8205128205128205,
+      "grad_norm": 0.2287733256816864,
+      "learning_rate": 8.385308194740846e-06,
+      "loss": 0.8508743047714233,
+      "step": 994
+    },
+    {
+      "epoch": 1.8241758241758241,
+      "grad_norm": 0.3089318573474884,
+      "learning_rate": 8.35125466970915e-06,
+      "loss": 1.1281416416168213,
+      "step": 996
+    },
+    {
+      "epoch": 1.8278388278388278,
+      "grad_norm": 0.19302226603031158,
+      "learning_rate": 8.317242611272745e-06,
+      "loss": 0.8560003042221069,
+      "step": 998
+    },
+    {
+      "epoch": 1.8315018315018317,
+      "grad_norm": 0.21388359367847443,
+      "learning_rate": 8.283272551896649e-06,
+      "loss": 1.1172125339508057,
+      "step": 1000
+    },
+    {
+      "epoch": 1.8351648351648353,
+      "grad_norm": 0.4454153776168823,
+      "learning_rate": 8.249345023388393e-06,
+      "loss": 1.1452471017837524,
+      "step": 1002
+    },
+    {
+      "epoch": 1.838827838827839,
+      "grad_norm": 0.4535501301288605,
+      "learning_rate": 8.21546055688968e-06,
+      "loss": 1.1414506435394287,
+      "step": 1004
+    },
+    {
+      "epoch": 1.8424908424908426,
+      "grad_norm": 0.3102174699306488,
+      "learning_rate": 8.181619682868059e-06,
+      "loss": 1.1409966945648193,
+      "step": 1006
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 0.29411637783050537,
+      "learning_rate": 8.147822931108638e-06,
+      "loss": 0.7928659915924072,
+      "step": 1008
+    },
+    {
+      "epoch": 1.84981684981685,
+      "grad_norm": 0.39852795004844666,
+      "learning_rate": 8.114070830705785e-06,
+      "loss": 1.103369116783142,
+      "step": 1010
+    },
+    {
+      "epoch": 1.8534798534798536,
+      "grad_norm": 0.10598917305469513,
+      "learning_rate": 8.080363910054833e-06,
+      "loss": 0.765097439289093,
+      "step": 1012
+    },
+    {
+      "epoch": 1.8571428571428572,
+      "grad_norm": 0.25210511684417725,
+      "learning_rate": 8.04670269684383e-06,
+      "loss": 1.1578898429870605,
+      "step": 1014
+    },
+    {
+      "epoch": 1.8608058608058609,
+      "grad_norm": 0.25960707664489746,
+      "learning_rate": 8.013087718045256e-06,
+      "loss": 1.1436736583709717,
+      "step": 1016
+    },
+    {
+      "epoch": 1.8644688644688645,
+      "grad_norm": 0.8823764324188232,
+      "learning_rate": 7.979519499907786e-06,
+      "loss": 1.224177598953247,
+      "step": 1018
+    },
+    {
+      "epoch": 1.8681318681318682,
+      "grad_norm": 0.031707342714071274,
+      "learning_rate": 7.945998567948052e-06,
+      "loss": 0.9655585289001465,
+      "step": 1020
+    },
+    {
+      "epoch": 1.8717948717948718,
+      "grad_norm": 0.4221316874027252,
+      "learning_rate": 7.912525446942406e-06,
+      "loss": 1.1163830757141113,
+      "step": 1022
+    },
+    {
+      "epoch": 1.8754578754578755,
+      "grad_norm": 0.4291134774684906,
+      "learning_rate": 7.879100660918713e-06,
+      "loss": 0.5574125051498413,
+      "step": 1024
+    },
+    {
+      "epoch": 1.879120879120879,
+      "grad_norm": 0.27034541964530945,
+      "learning_rate": 7.845724733148149e-06,
+      "loss": 1.154983401298523,
+      "step": 1026
+    },
+    {
+      "epoch": 1.8827838827838828,
+      "grad_norm": 0.6794618964195251,
+      "learning_rate": 7.812398186136994e-06,
+      "loss": 0.9258843064308167,
+      "step": 1028
+    },
+    {
+      "epoch": 1.8864468864468864,
+      "grad_norm": 0.2529967129230499,
+      "learning_rate": 7.779121541618478e-06,
+      "loss": 1.1542246341705322,
+      "step": 1030
+    },
+    {
+      "epoch": 1.89010989010989,
+      "grad_norm": 0.2132919579744339,
+      "learning_rate": 7.74589532054459e-06,
+      "loss": 0.9495857954025269,
+      "step": 1032
+    },
+    {
+      "epoch": 1.8937728937728937,
+      "grad_norm": 1.0942879915237427,
+      "learning_rate": 7.712720043077929e-06,
+      "loss": 0.7942955493927002,
+      "step": 1034
+    },
+    {
+      "epoch": 1.8974358974358974,
+      "grad_norm": 0.38435038924217224,
+      "learning_rate": 7.679596228583563e-06,
+      "loss": 1.1871325969696045,
+      "step": 1036
+    },
+    {
+      "epoch": 1.901098901098901,
+      "grad_norm": 0.26236408948898315,
+      "learning_rate": 7.646524395620908e-06,
+      "loss": 1.1567392349243164,
+      "step": 1038
+    },
+    {
+      "epoch": 1.9047619047619047,
+      "grad_norm": 0.7066712975502014,
+      "learning_rate": 7.613505061935584e-06,
+      "loss": 1.227746844291687,
+      "step": 1040
+    },
+    {
+      "epoch": 1.9084249084249083,
+      "grad_norm": 0.363916277885437,
+      "learning_rate": 7.580538744451336e-06,
+      "loss": 0.6182924509048462,
+      "step": 1042
+    },
+    {
+      "epoch": 1.912087912087912,
+      "grad_norm": 0.4058935046195984,
+      "learning_rate": 7.547625959261928e-06,
+      "loss": 0.8744593858718872,
+      "step": 1044
+    },
+    {
+      "epoch": 1.9157509157509156,
+      "grad_norm": 0.15468518435955048,
+      "learning_rate": 7.5147672216230605e-06,
+      "loss": 1.1123764514923096,
+      "step": 1046
+    },
+    {
+      "epoch": 1.9194139194139193,
+      "grad_norm": 0.3142243027687073,
+      "learning_rate": 7.481963045944318e-06,
+      "loss": 0.4596446752548218,
+      "step": 1048
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 1.057568907737732,
+      "learning_rate": 7.449213945781102e-06,
+      "loss": 0.8759109973907471,
+      "step": 1050
+    },
+    {
+      "epoch": 1.9267399267399268,
+      "grad_norm": 0.2798309028148651,
+      "learning_rate": 7.416520433826599e-06,
+      "loss": 0.8125253319740295,
+      "step": 1052
+    },
+    {
+      "epoch": 1.9304029304029304,
+      "grad_norm": 0.3273449242115021,
+      "learning_rate": 7.383883021903755e-06,
+      "loss": 1.1237201690673828,
+      "step": 1054
+    },
+    {
+      "epoch": 1.934065934065934,
+      "grad_norm": 1.727766990661621,
+      "learning_rate": 7.351302220957251e-06,
+      "loss": 0.7471660375595093,
+      "step": 1056
+    },
+    {
+      "epoch": 1.9377289377289377,
+      "grad_norm": 0.21788397431373596,
+      "learning_rate": 7.318778541045517e-06,
+      "loss": 0.9874610304832458,
+      "step": 1058
+    },
+    {
+      "epoch": 1.9413919413919414,
+      "grad_norm": 0.3261561393737793,
+      "learning_rate": 7.286312491332754e-06,
+      "loss": 1.0792856216430664,
+      "step": 1060
+    },
+    {
+      "epoch": 1.945054945054945,
+      "grad_norm": 0.19054535031318665,
+      "learning_rate": 7.253904580080926e-06,
+      "loss": 0.8402732014656067,
+      "step": 1062
+    },
+    {
+      "epoch": 1.9487179487179487,
+      "grad_norm": 0.28284960985183716,
+      "learning_rate": 7.221555314641853e-06,
+      "loss": 0.8096812963485718,
+      "step": 1064
+    },
+    {
+      "epoch": 1.9523809523809523,
+      "grad_norm": 0.4529983699321747,
+      "learning_rate": 7.18926520144924e-06,
+      "loss": 1.15970778465271,
+      "step": 1066
+    },
+    {
+      "epoch": 1.9560439560439562,
+      "grad_norm": 0.2033073455095291,
+      "learning_rate": 7.1570347460107335e-06,
+      "loss": 1.1832486391067505,
+      "step": 1068
+    },
+    {
+      "epoch": 1.9597069597069599,
+      "grad_norm": 0.09281979501247406,
+      "learning_rate": 7.124864452900049e-06,
+      "loss": 0.7031117081642151,
+      "step": 1070
+    },
+    {
+      "epoch": 1.9633699633699635,
+      "grad_norm": 0.20894154906272888,
+      "learning_rate": 7.0927548257490465e-06,
+      "loss": 0.8534321188926697,
+      "step": 1072
+    },
+    {
+      "epoch": 1.9670329670329672,
+      "grad_norm": 0.292833149433136,
+      "learning_rate": 7.060706367239836e-06,
+      "loss": 1.1420583724975586,
+      "step": 1074
+    },
+    {
+      "epoch": 1.9706959706959708,
+      "grad_norm": 0.16723449528217316,
+      "learning_rate": 7.028719579096932e-06,
+      "loss": 1.12359619140625,
+      "step": 1076
+    },
+    {
+      "epoch": 1.9743589743589745,
+      "grad_norm": 0.36365193128585815,
+      "learning_rate": 6.9967949620793854e-06,
+      "loss": 0.999005138874054,
+      "step": 1078
+    },
+    {
+      "epoch": 1.978021978021978,
+      "grad_norm": 0.2431362122297287,
+      "learning_rate": 6.964933015972947e-06,
+      "loss": 1.0200340747833252,
+      "step": 1080
+    },
+    {
+      "epoch": 1.9816849816849818,
+      "grad_norm": 0.23444706201553345,
+      "learning_rate": 6.933134239582246e-06,
+      "loss": 1.1037390232086182,
+      "step": 1082
+    },
+    {
+      "epoch": 1.9853479853479854,
+      "grad_norm": 0.13071784377098083,
+      "learning_rate": 6.9013991307229745e-06,
+      "loss": 0.7239289879798889,
+      "step": 1084
+    },
+    {
+      "epoch": 1.989010989010989,
+      "grad_norm": 1.6581590175628662,
+      "learning_rate": 6.869728186214093e-06,
+      "loss": 0.9764227271080017,
+      "step": 1086
+    },
+    {
+      "epoch": 1.9926739926739927,
+      "grad_norm": 0.40288543701171875,
+      "learning_rate": 6.8381219018700675e-06,
+      "loss": 0.976794421672821,
+      "step": 1088
+    },
+    {
+      "epoch": 1.9963369963369964,
+      "grad_norm": 0.14982423186302185,
+      "learning_rate": 6.806580772493088e-06,
+      "loss": 0.9331335425376892,
+      "step": 1090
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 0.9287575483322144,
+      "learning_rate": 6.775105291865343e-06,
+      "loss": 1.042527198791504,
+      "step": 1092
+    },
+    {
+      "epoch": 2.0036630036630036,
+      "grad_norm": 0.3157676160335541,
+      "learning_rate": 6.743695952741265e-06,
+      "loss": 1.0824307203292847,
+      "step": 1094
+    },
+    {
+      "epoch": 2.0073260073260073,
+      "grad_norm": 0.2867039144039154,
+      "learning_rate": 6.71235324683983e-06,
+      "loss": 1.1392470598220825,
+      "step": 1096
+    },
+    {
+      "epoch": 2.010989010989011,
+      "grad_norm": 0.24429775774478912,
+      "learning_rate": 6.681077664836872e-06,
+      "loss": 1.0826945304870605,
+      "step": 1098
+    },
+    {
+      "epoch": 2.0146520146520146,
+      "grad_norm": 0.49506935477256775,
+      "learning_rate": 6.649869696357381e-06,
+      "loss": 1.162758469581604,
+      "step": 1100
+    },
+    {
+      "epoch": 2.0183150183150182,
+      "grad_norm": 0.2504410743713379,
+      "learning_rate": 6.6187298299678295e-06,
+      "loss": 0.8383739590644836,
+      "step": 1102
+    },
+    {
+      "epoch": 2.021978021978022,
+      "grad_norm": 0.2852114140987396,
+      "learning_rate": 6.587658553168563e-06,
+      "loss": 1.142665147781372,
+      "step": 1104
+    },
+    {
+      "epoch": 2.0256410256410255,
+      "grad_norm": 0.21823060512542725,
+      "learning_rate": 6.556656352386135e-06,
+      "loss": 0.7671647667884827,
+      "step": 1106
+    },
+    {
+      "epoch": 2.029304029304029,
+      "grad_norm": 0.23471537232398987,
+      "learning_rate": 6.525723712965698e-06,
+      "loss": 1.1795700788497925,
+      "step": 1108
+    },
+    {
+      "epoch": 2.032967032967033,
+      "grad_norm": 0.2844531536102295,
+      "learning_rate": 6.494861119163412e-06,
+      "loss": 0.8791682720184326,
+      "step": 1110
+    },
+    {
+      "epoch": 2.0366300366300365,
+      "grad_norm": 0.09012774378061295,
+      "learning_rate": 6.464069054138853e-06,
+      "loss": 0.6179495453834534,
+      "step": 1112
+    },
+    {
+      "epoch": 2.04029304029304,
+      "grad_norm": 0.4640715718269348,
+      "learning_rate": 6.433347999947468e-06,
+      "loss": 0.8358191251754761,
+      "step": 1114
+    },
+    {
+      "epoch": 2.043956043956044,
+      "grad_norm": 0.4230431318283081,
+      "learning_rate": 6.402698437533012e-06,
+      "loss": 1.157958984375,
+      "step": 1116
+    },
+    {
+      "epoch": 2.0476190476190474,
+      "grad_norm": 0.2907547950744629,
+      "learning_rate": 6.372120846720018e-06,
+      "loss": 1.117270827293396,
+      "step": 1118
+    },
+    {
+      "epoch": 2.051282051282051,
+      "grad_norm": 0.406525582075119,
+      "learning_rate": 6.341615706206292e-06,
+      "loss": 0.840583324432373,
+      "step": 1120
+    },
+    {
+      "epoch": 2.0549450549450547,
+      "grad_norm": 3.0575461387634277,
+      "learning_rate": 6.311183493555426e-06,
+      "loss": 1.3155573606491089,
+      "step": 1122
+    },
+    {
+      "epoch": 2.0586080586080584,
+      "grad_norm": 0.2396925538778305,
+      "learning_rate": 6.280824685189296e-06,
+      "loss": 1.136577844619751,
+      "step": 1124
+    },
+    {
+      "epoch": 2.062271062271062,
+      "grad_norm": 0.2372613549232483,
+      "learning_rate": 6.25053975638064e-06,
+      "loss": 0.761581540107727,
+      "step": 1126
+    },
+    {
+      "epoch": 2.065934065934066,
+      "grad_norm": 0.1678687483072281,
+      "learning_rate": 6.220329181245585e-06,
+      "loss": 1.104315161705017,
+      "step": 1128
+    },
+    {
+      "epoch": 2.06959706959707,
+      "grad_norm": 0.24123162031173706,
+      "learning_rate": 6.1901934327362355e-06,
+      "loss": 1.0985783338546753,
+      "step": 1130
+    },
+    {
+      "epoch": 2.0732600732600734,
+      "grad_norm": 0.2733916640281677,
+      "learning_rate": 6.16013298263328e-06,
+      "loss": 1.13059401512146,
+      "step": 1132
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 0.2685484290122986,
+      "learning_rate": 6.130148301538601e-06,
+      "loss": 1.1149473190307617,
+      "step": 1134
+    },
+    {
+      "epoch": 2.0805860805860807,
+      "grad_norm": 0.2145625352859497,
+      "learning_rate": 6.100239858867887e-06,
+      "loss": 0.7262951731681824,
+      "step": 1136
+    },
+    {
+      "epoch": 2.0842490842490844,
+      "grad_norm": 0.21125853061676025,
+      "learning_rate": 6.070408122843311e-06,
+      "loss": 1.1171889305114746,
+      "step": 1138
+    },
+    {
+      "epoch": 2.087912087912088,
+      "grad_norm": 0.6227608323097229,
+      "learning_rate": 6.040653560486183e-06,
+      "loss": 1.0216211080551147,
+      "step": 1140
+    },
+    {
+      "epoch": 2.0915750915750917,
+      "grad_norm": 1.5662765502929688,
+      "learning_rate": 6.010976637609653e-06,
+      "loss": 1.1097993850708008,
+      "step": 1142
+    },
+    {
+      "epoch": 2.0952380952380953,
+      "grad_norm": 0.17495885491371155,
+      "learning_rate": 5.9813778188114125e-06,
+      "loss": 0.38371744751930237,
+      "step": 1144
+    },
+    {
+      "epoch": 2.098901098901099,
+      "grad_norm": 0.5771632790565491,
+      "learning_rate": 5.951857567466401e-06,
+      "loss": 0.9106112718582153,
+      "step": 1146
+    },
+    {
+      "epoch": 2.1025641025641026,
+      "grad_norm": 0.33344414830207825,
+      "learning_rate": 5.922416345719588e-06,
+      "loss": 0.8066908717155457,
+      "step": 1148
+    },
+    {
+      "epoch": 2.1062271062271063,
+      "grad_norm": 0.22791945934295654,
+      "learning_rate": 5.893054614478718e-06,
+      "loss": 0.8162147998809814,
+      "step": 1150
+    },
+    {
+      "epoch": 2.10989010989011,
+      "grad_norm": 0.2723308801651001,
+      "learning_rate": 5.8637728334070905e-06,
+      "loss": 0.9493057131767273,
+      "step": 1152
+    },
+    {
+      "epoch": 2.1135531135531136,
+      "grad_norm": 0.5839288830757141,
+      "learning_rate": 5.834571460916371e-06,
+      "loss": 0.7895359992980957,
+      "step": 1154
+    },
+    {
+      "epoch": 2.1172161172161172,
+      "grad_norm": 0.22610190510749817,
+      "learning_rate": 5.805450954159422e-06,
+      "loss": 1.0748239755630493,
+      "step": 1156
+    },
+    {
+      "epoch": 2.120879120879121,
+      "grad_norm": 0.5260972380638123,
+      "learning_rate": 5.776411769023127e-06,
+      "loss": 1.0262991189956665,
+      "step": 1158
+    },
+    {
+      "epoch": 2.1245421245421245,
+      "grad_norm": 1.0613799095153809,
+      "learning_rate": 5.747454360121274e-06,
+      "loss": 0.8698439002037048,
+      "step": 1160
+    },
+    {
+      "epoch": 2.128205128205128,
+      "grad_norm": 0.7695472240447998,
+      "learning_rate": 5.718579180787425e-06,
+      "loss": 0.7719033360481262,
+      "step": 1162
+    },
+    {
+      "epoch": 2.131868131868132,
+      "grad_norm": 0.3090405762195587,
+      "learning_rate": 5.689786683067817e-06,
+      "loss": 0.9184219837188721,
+      "step": 1164
+    },
+    {
+      "epoch": 2.1355311355311355,
+      "grad_norm": 0.3468276262283325,
+      "learning_rate": 5.661077317714303e-06,
+      "loss": 0.4223445951938629,
+      "step": 1166
+    },
+    {
+      "epoch": 2.139194139194139,
+      "grad_norm": 0.24201928079128265,
+      "learning_rate": 5.632451534177276e-06,
+      "loss": 0.4232040047645569,
+      "step": 1168
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.6462178826332092,
+      "learning_rate": 5.603909780598644e-06,
+      "loss": 0.8980230689048767,
+      "step": 1170
+    },
+    {
+      "epoch": 2.1465201465201464,
+      "grad_norm": 0.6586897969245911,
+      "learning_rate": 5.575452503804805e-06,
+      "loss": 1.1271127462387085,
+      "step": 1172
+    },
+    {
+      "epoch": 2.15018315018315,
+      "grad_norm": 0.5097822546958923,
+      "learning_rate": 5.5470801492996605e-06,
+      "loss": 1.2748925685882568,
+      "step": 1174
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 0.3521871864795685,
+      "learning_rate": 5.518793161257641e-06,
+      "loss": 0.7676212191581726,
+      "step": 1176
+    },
+    {
+      "epoch": 2.1575091575091574,
+      "grad_norm": 0.15629425644874573,
+      "learning_rate": 5.490591982516749e-06,
+      "loss": 1.1415168046951294,
+      "step": 1178
+    },
+    {
+      "epoch": 2.161172161172161,
+      "grad_norm": 0.1941349059343338,
+      "learning_rate": 5.462477054571617e-06,
+      "loss": 1.12305748462677,
+      "step": 1180
+    },
+    {
+      "epoch": 2.1648351648351647,
+      "grad_norm": 0.22985997796058655,
+      "learning_rate": 5.4344488175666154e-06,
+      "loss": 1.1313296556472778,
+      "step": 1182
+    },
+    {
+      "epoch": 2.1684981684981683,
+      "grad_norm": 0.15692557394504547,
+      "learning_rate": 5.406507710288955e-06,
+      "loss": 1.1444305181503296,
+      "step": 1184
+    },
+    {
+      "epoch": 2.172161172161172,
+      "grad_norm": 0.247311532497406,
+      "learning_rate": 5.378654170161805e-06,
+      "loss": 0.31945711374282837,
+      "step": 1186
+    },
+    {
+      "epoch": 2.1758241758241756,
+      "grad_norm": 1.0448226928710938,
+      "learning_rate": 5.3508886332374534e-06,
+      "loss": 1.2239130735397339,
+      "step": 1188
+    },
+    {
+      "epoch": 2.1794871794871793,
+      "grad_norm": 1.5233474969863892,
+      "learning_rate": 5.323211534190496e-06,
+      "loss": 0.7491469979286194,
+      "step": 1190
+    },
+    {
+      "epoch": 2.183150183150183,
+      "grad_norm": 0.4820270538330078,
+      "learning_rate": 5.295623306310999e-06,
+      "loss": 0.9774181246757507,
+      "step": 1192
+    },
+    {
+      "epoch": 2.186813186813187,
+      "grad_norm": 0.19126620888710022,
+      "learning_rate": 5.268124381497755e-06,
+      "loss": 0.8192515969276428,
+      "step": 1194
+    },
+    {
+      "epoch": 2.1904761904761907,
+      "grad_norm": 0.45134642720222473,
+      "learning_rate": 5.240715190251484e-06,
+      "loss": 0.8925463557243347,
+      "step": 1196
+    },
+    {
+      "epoch": 2.1941391941391943,
+      "grad_norm": 0.4911046624183655,
+      "learning_rate": 5.213396161668111e-06,
+      "loss": 1.1126856803894043,
+      "step": 1198
+    },
+    {
+      "epoch": 2.197802197802198,
+      "grad_norm": 3.0808489322662354,
+      "learning_rate": 5.186167723432061e-06,
+      "loss": 0.7509762048721313,
+      "step": 1200
+    },
+    {
+      "epoch": 2.2014652014652016,
+      "grad_norm": 0.27347448468208313,
+      "learning_rate": 5.159030301809534e-06,
+      "loss": 1.1802153587341309,
+      "step": 1202
+    },
+    {
+      "epoch": 2.2051282051282053,
+      "grad_norm": 0.2847099006175995,
+      "learning_rate": 5.131984321641865e-06,
+      "loss": 1.129975438117981,
+      "step": 1204
+    },
+    {
+      "epoch": 2.208791208791209,
+      "grad_norm": 0.4671761393547058,
+      "learning_rate": 5.105030206338843e-06,
+      "loss": 0.4967349171638489,
+      "step": 1206
+    },
+    {
+      "epoch": 2.2124542124542126,
+      "grad_norm": 0.3511367738246918,
+      "learning_rate": 5.0781683778720965e-06,
+      "loss": 0.7303721308708191,
+      "step": 1208
+    },
+    {
+      "epoch": 2.2161172161172162,
+      "grad_norm": 0.5738789439201355,
+      "learning_rate": 5.051399256768498e-06,
+      "loss": 0.6005207896232605,
+      "step": 1210
+    },
+    {
+      "epoch": 2.21978021978022,
+      "grad_norm": 0.14039315283298492,
+      "learning_rate": 5.024723262103559e-06,
+      "loss": 1.0828591585159302,
+      "step": 1212
+    },
+    {
+      "epoch": 2.2234432234432235,
+      "grad_norm": 0.6400618553161621,
+      "learning_rate": 4.998140811494881e-06,
+      "loss": 0.8194615840911865,
+      "step": 1214
+    },
+    {
+      "epoch": 2.227106227106227,
+      "grad_norm": 0.4215098023414612,
+      "learning_rate": 4.971652321095614e-06,
+      "loss": 0.8995423913002014,
+      "step": 1216
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 0.35582196712493896,
+      "learning_rate": 4.945258205587955e-06,
+      "loss": 1.1104472875595093,
+      "step": 1218
+    },
+    {
+      "epoch": 2.2344322344322345,
+      "grad_norm": 0.3917670249938965,
+      "learning_rate": 4.918958878176628e-06,
+      "loss": 1.117361068725586,
+      "step": 1220
+    },
+    {
+      "epoch": 2.238095238095238,
+      "grad_norm": 0.25310394167900085,
+      "learning_rate": 4.8927547505824465e-06,
+      "loss": 1.1276395320892334,
+      "step": 1222
+    },
+    {
+      "epoch": 2.241758241758242,
+      "grad_norm": 0.3766920566558838,
+      "learning_rate": 4.866646233035845e-06,
+      "loss": 0.5073704123497009,
+      "step": 1224
+    },
+    {
+      "epoch": 2.2454212454212454,
+      "grad_norm": 0.36118075251579285,
+      "learning_rate": 4.840633734270464e-06,
+      "loss": 1.083837866783142,
+      "step": 1226
+    },
+    {
+      "epoch": 2.249084249084249,
+      "grad_norm": 0.9737128019332886,
+      "learning_rate": 4.814717661516762e-06,
+      "loss": 0.6141456365585327,
+      "step": 1228
+    },
+    {
+      "epoch": 2.2527472527472527,
+      "grad_norm": 0.060193657875061035,
+      "learning_rate": 4.788898420495622e-06,
+      "loss": 0.5997035503387451,
+      "step": 1230
+    },
+    {
+      "epoch": 2.2564102564102564,
+      "grad_norm": 0.1481347531080246,
+      "learning_rate": 4.763176415412006e-06,
+      "loss": 0.6170958876609802,
+      "step": 1232
+    },
+    {
+      "epoch": 2.26007326007326,
+      "grad_norm": 2.355868339538574,
+      "learning_rate": 4.7375520489486395e-06,
+      "loss": 1.1160298585891724,
+      "step": 1234
+    },
+    {
+      "epoch": 2.2637362637362637,
+      "grad_norm": 0.468891978263855,
+      "learning_rate": 4.71202572225969e-06,
+      "loss": 0.7722483277320862,
+      "step": 1236
+    },
+    {
+      "epoch": 2.2673992673992673,
+      "grad_norm": 0.17015230655670166,
+      "learning_rate": 4.686597834964499e-06,
+      "loss": 0.9469270706176758,
+      "step": 1238
+    },
+    {
+      "epoch": 2.271062271062271,
+      "grad_norm": 0.5269258618354797,
+      "learning_rate": 4.661268785141316e-06,
+      "loss": 0.8888347744941711,
+      "step": 1240
+    },
+    {
+      "epoch": 2.2747252747252746,
+      "grad_norm": 0.799193799495697,
+      "learning_rate": 4.636038969321073e-06,
+      "loss": 0.9122021198272705,
+      "step": 1242
+    },
+    {
+      "epoch": 2.2783882783882783,
+      "grad_norm": 0.239344522356987,
+      "learning_rate": 4.610908782481179e-06,
+      "loss": 1.1070375442504883,
+      "step": 1244
+    },
+    {
+      "epoch": 2.282051282051282,
+      "grad_norm": 0.10058314353227615,
+      "learning_rate": 4.5858786180393326e-06,
+      "loss": 0.7249715328216553,
+      "step": 1246
+    },
+    {
+      "epoch": 2.2857142857142856,
+      "grad_norm": 0.4845520853996277,
+      "learning_rate": 4.560948867847359e-06,
+      "loss": 0.7078754305839539,
+      "step": 1248
+    },
+    {
+      "epoch": 2.2893772893772892,
+      "grad_norm": 0.3414648175239563,
+      "learning_rate": 4.536119922185082e-06,
+      "loss": 1.100352168083191,
+      "step": 1250
+    },
+    {
+      "epoch": 2.293040293040293,
+      "grad_norm": 0.4015507400035858,
+      "learning_rate": 4.511392169754214e-06,
+      "loss": 0.8295157551765442,
+      "step": 1252
+    },
+    {
+      "epoch": 2.2967032967032965,
+      "grad_norm": 1.4144065380096436,
+      "learning_rate": 4.486765997672263e-06,
+      "loss": 1.1649739742279053,
+      "step": 1254
+    },
+    {
+      "epoch": 2.3003663003663,
+      "grad_norm": 0.19826774299144745,
+      "learning_rate": 4.46224179146649e-06,
+      "loss": 1.1785088777542114,
+      "step": 1256
+    },
+    {
+      "epoch": 2.304029304029304,
+      "grad_norm": 0.8159498572349548,
+      "learning_rate": 4.437819935067847e-06,
+      "loss": 0.7735372185707092,
+      "step": 1258
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 0.6313501596450806,
+      "learning_rate": 4.413500810804986e-06,
+      "loss": 1.1706268787384033,
+      "step": 1260
+    },
+    {
+      "epoch": 2.311355311355311,
+      "grad_norm": 0.29691141843795776,
+      "learning_rate": 4.389284799398276e-06,
+      "loss": 1.0271168947219849,
+      "step": 1262
+    },
+    {
+      "epoch": 2.315018315018315,
+      "grad_norm": 0.36938998103141785,
+      "learning_rate": 4.365172279953825e-06,
+      "loss": 0.8776956796646118,
+      "step": 1264
+    },
+    {
+      "epoch": 2.3186813186813184,
+      "grad_norm": 0.44165417551994324,
+      "learning_rate": 4.34116362995756e-06,
+      "loss": 0.7929419279098511,
+      "step": 1266
+    },
+    {
+      "epoch": 2.3223443223443225,
+      "grad_norm": 0.5072553157806396,
+      "learning_rate": 4.317259225269313e-06,
+      "loss": 0.8222364187240601,
+      "step": 1268
+    },
+    {
+      "epoch": 2.326007326007326,
+      "grad_norm": 0.37858933210372925,
+      "learning_rate": 4.293459440116935e-06,
+      "loss": 1.1856492757797241,
+      "step": 1270
+    },
+    {
+      "epoch": 2.32967032967033,
+      "grad_norm": 0.27582108974456787,
+      "learning_rate": 4.269764647090442e-06,
+      "loss": 1.0090421438217163,
+      "step": 1272
+    },
+    {
+      "epoch": 2.3333333333333335,
+      "grad_norm": 0.26256465911865234,
+      "learning_rate": 4.246175217136176e-06,
+      "loss": 0.6174538135528564,
+      "step": 1274
+    },
+    {
+      "epoch": 2.336996336996337,
+      "grad_norm": 0.2537238597869873,
+      "learning_rate": 4.2226915195509954e-06,
+      "loss": 1.0155659914016724,
+      "step": 1276
+    },
+    {
+      "epoch": 2.340659340659341,
+      "grad_norm": 0.8025590777397156,
+      "learning_rate": 4.199313921976511e-06,
+      "loss": 0.8254562616348267,
+      "step": 1278
+    },
+    {
+      "epoch": 2.3443223443223444,
+      "grad_norm": 0.2071162611246109,
+      "learning_rate": 4.176042790393313e-06,
+      "loss": 0.752740204334259,
+      "step": 1280
+    },
+    {
+      "epoch": 2.347985347985348,
+      "grad_norm": 0.381971538066864,
+      "learning_rate": 4.152878489115244e-06,
+      "loss": 0.6165816783905029,
+      "step": 1282
+    },
+    {
+      "epoch": 2.3516483516483517,
+      "grad_norm": 0.16895699501037598,
+      "learning_rate": 4.129821380783698e-06,
+      "loss": 1.092519760131836,
+      "step": 1284
+    },
+    {
+      "epoch": 2.3553113553113554,
+      "grad_norm": 2.325608491897583,
+      "learning_rate": 4.106871826361952e-06,
+      "loss": 0.7248493432998657,
+      "step": 1286
+    },
+    {
+      "epoch": 2.358974358974359,
+      "grad_norm": 0.22204965353012085,
+      "learning_rate": 4.084030185129495e-06,
+      "loss": 0.924081027507782,
+      "step": 1288
+    },
+    {
+      "epoch": 2.3626373626373627,
+      "grad_norm": 0.368304044008255,
+      "learning_rate": 4.061296814676429e-06,
+      "loss": 0.921589195728302,
+      "step": 1290
+    },
+    {
+      "epoch": 2.3663003663003663,
+      "grad_norm": 0.35086876153945923,
+      "learning_rate": 4.038672070897844e-06,
+      "loss": 0.7640153765678406,
+      "step": 1292
+    },
+    {
+      "epoch": 2.36996336996337,
+      "grad_norm": 0.29217588901519775,
+      "learning_rate": 4.016156307988262e-06,
+      "loss": 0.8899466395378113,
+      "step": 1294
+    },
+    {
+      "epoch": 2.3736263736263736,
+      "grad_norm": 0.3964160978794098,
+      "learning_rate": 3.9937498784361e-06,
+      "loss": 1.2294063568115234,
+      "step": 1296
+    },
+    {
+      "epoch": 2.3772893772893773,
+      "grad_norm": 0.282614141702652,
+      "learning_rate": 3.9714531330181275e-06,
+      "loss": 1.0998694896697998,
+      "step": 1298
+    },
+    {
+      "epoch": 2.380952380952381,
+      "grad_norm": 0.35640949010849,
+      "learning_rate": 3.949266420793999e-06,
+      "loss": 1.1518428325653076,
+      "step": 1300
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 0.8390282988548279,
+      "learning_rate": 3.9271900891007734e-06,
+      "loss": 0.8655717968940735,
+      "step": 1302
+    },
+    {
+      "epoch": 2.3882783882783882,
+      "grad_norm": 0.27162355184555054,
+      "learning_rate": 3.905224483547479e-06,
+      "loss": 0.4701066017150879,
+      "step": 1304
+    },
+    {
+      "epoch": 2.391941391941392,
+      "grad_norm": 0.5130133628845215,
+      "learning_rate": 3.883369948009714e-06,
+      "loss": 0.9348631501197815,
+      "step": 1306
+    },
+    {
+      "epoch": 2.3956043956043955,
+      "grad_norm": 0.19122500717639923,
+      "learning_rate": 3.861626824624258e-06,
+      "loss": 1.1146955490112305,
+      "step": 1308
+    },
+    {
+      "epoch": 2.399267399267399,
+      "grad_norm": 0.7361530065536499,
+      "learning_rate": 3.839995453783694e-06,
+      "loss": 0.514556348323822,
+      "step": 1310
+    },
+    {
+      "epoch": 2.402930402930403,
+      "grad_norm": 0.4885084629058838,
+      "learning_rate": 3.818476174131118e-06,
+      "loss": 1.121952772140503,
+      "step": 1312
+    },
+    {
+      "epoch": 2.4065934065934065,
+      "grad_norm": 0.2934611439704895,
+      "learning_rate": 3.7970693225548116e-06,
+      "loss": 0.8278881311416626,
+      "step": 1314
+    },
+    {
+      "epoch": 2.41025641025641,
+      "grad_norm": 0.24372442066669464,
+      "learning_rate": 3.7757752341829723e-06,
+      "loss": 1.1141726970672607,
+      "step": 1316
+    },
+    {
+      "epoch": 2.413919413919414,
+      "grad_norm": 0.7190600633621216,
+      "learning_rate": 3.754594242378466e-06,
+      "loss": 0.7716094255447388,
+      "step": 1318
+    },
+    {
+      "epoch": 2.4175824175824174,
+      "grad_norm": 0.14406920969486237,
+      "learning_rate": 3.7335266787336194e-06,
+      "loss": 0.7671050429344177,
+      "step": 1320
+    },
+    {
+      "epoch": 2.421245421245421,
+      "grad_norm": 0.32376912236213684,
+      "learning_rate": 3.712572873065012e-06,
+      "loss": 0.6145520210266113,
+      "step": 1322
+    },
+    {
+      "epoch": 2.4249084249084247,
+      "grad_norm": 0.44639283418655396,
+      "learning_rate": 3.69173315340833e-06,
+      "loss": 0.7646421194076538,
+      "step": 1324
+    },
+    {
+      "epoch": 2.4285714285714284,
+      "grad_norm": 0.21105606853961945,
+      "learning_rate": 3.6710078460132137e-06,
+      "loss": 0.7964082956314087,
+      "step": 1326
+    },
+    {
+      "epoch": 2.4322344322344325,
+      "grad_norm": 0.0819050669670105,
+      "learning_rate": 3.650397275338161e-06,
+      "loss": 0.6835886240005493,
+      "step": 1328
+    },
+    {
+      "epoch": 2.435897435897436,
+      "grad_norm": 0.21763509511947632,
+      "learning_rate": 3.6299017640454516e-06,
+      "loss": 0.7992286682128906,
+      "step": 1330
+    },
+    {
+      "epoch": 2.4395604395604398,
+      "grad_norm": 0.2989668846130371,
+      "learning_rate": 3.6095216329960786e-06,
+      "loss": 1.0545233488082886,
+      "step": 1332
+    },
+    {
+      "epoch": 2.4432234432234434,
+      "grad_norm": 0.5623311400413513,
+      "learning_rate": 3.5892572012447457e-06,
+      "loss": 0.7113739848136902,
+      "step": 1334
+    },
+    {
+      "epoch": 2.446886446886447,
+      "grad_norm": 0.372221440076828,
+      "learning_rate": 3.5691087860348577e-06,
+      "loss": 0.6705981492996216,
+      "step": 1336
+    },
+    {
+      "epoch": 2.4505494505494507,
+      "grad_norm": 0.29782795906066895,
+      "learning_rate": 3.549076702793557e-06,
+      "loss": 1.0251377820968628,
+      "step": 1338
+    },
+    {
+      "epoch": 2.4542124542124544,
+      "grad_norm": 0.12661497294902802,
+      "learning_rate": 3.529161265126795e-06,
+      "loss": 0.8721184730529785,
+      "step": 1340
+    },
+    {
+      "epoch": 2.457875457875458,
+      "grad_norm": 0.20114953815937042,
+      "learning_rate": 3.5093627848144128e-06,
+      "loss": 0.6586589813232422,
+      "step": 1342
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 0.22184857726097107,
+      "learning_rate": 3.4896815718052534e-06,
+      "loss": 0.8093105554580688,
+      "step": 1344
+    },
+    {
+      "epoch": 2.4652014652014653,
+      "grad_norm": 0.3594045639038086,
+      "learning_rate": 3.4701179342123313e-06,
+      "loss": 1.0655725002288818,
+      "step": 1346
+    },
+    {
+      "epoch": 2.468864468864469,
+      "grad_norm": 0.19054588675498962,
+      "learning_rate": 3.4506721783079925e-06,
+      "loss": 1.1080653667449951,
+      "step": 1348
+    },
+    {
+      "epoch": 2.4725274725274726,
+      "grad_norm": 0.2145404815673828,
+      "learning_rate": 3.4313446085191203e-06,
+      "loss": 0.7264165878295898,
+      "step": 1350
+    },
+    {
+      "epoch": 2.4761904761904763,
+      "grad_norm": 0.21821582317352295,
+      "learning_rate": 3.4121355274223727e-06,
+      "loss": 0.8345311284065247,
+      "step": 1352
+    },
+    {
+      "epoch": 2.47985347985348,
+      "grad_norm": 0.7348686456680298,
+      "learning_rate": 3.3930452357394473e-06,
+      "loss": 0.9193353056907654,
+      "step": 1354
+    },
+    {
+      "epoch": 2.4835164835164836,
+      "grad_norm": 4.1631364822387695,
+      "learning_rate": 3.3740740323323705e-06,
+      "loss": 0.9116562604904175,
+      "step": 1356
+    },
+    {
+      "epoch": 2.4871794871794872,
+      "grad_norm": 0.33455634117126465,
+      "learning_rate": 3.3552222141988257e-06,
+      "loss": 1.0841375589370728,
+      "step": 1358
+    },
+    {
+      "epoch": 2.490842490842491,
+      "grad_norm": 0.42869114875793457,
+      "learning_rate": 3.336490076467489e-06,
+      "loss": 0.8795775175094604,
+      "step": 1360
+    },
+    {
+      "epoch": 2.4945054945054945,
+      "grad_norm": 0.1992548406124115,
+      "learning_rate": 3.31787791239342e-06,
+      "loss": 0.8005448579788208,
+      "step": 1362
+    },
+    {
+      "epoch": 2.498168498168498,
+      "grad_norm": 0.4712922275066376,
+      "learning_rate": 3.2993860133534763e-06,
+      "loss": 0.78562992811203,
+      "step": 1364
+    },
+    {
+      "epoch": 2.501831501831502,
+      "grad_norm": 0.33841851353645325,
+      "learning_rate": 3.2810146688417304e-06,
+      "loss": 1.1035329103469849,
+      "step": 1366
+    },
+    {
+      "epoch": 2.5054945054945055,
+      "grad_norm": 0.44614377617836,
+      "learning_rate": 3.2627641664649666e-06,
+      "loss": 0.8065237998962402,
+      "step": 1368
+    },
+    {
+      "epoch": 2.509157509157509,
+      "grad_norm": 0.30702492594718933,
+      "learning_rate": 3.2446347919381533e-06,
+      "loss": 0.9479190707206726,
+      "step": 1370
+    },
+    {
+      "epoch": 2.5128205128205128,
+      "grad_norm": 0.13606101274490356,
+      "learning_rate": 3.226626829079979e-06,
+      "loss": 0.6367249488830566,
+      "step": 1372
+    },
+    {
+      "epoch": 2.5164835164835164,
+      "grad_norm": 0.3242248296737671,
+      "learning_rate": 3.2087405598084194e-06,
+      "loss": 0.7743354439735413,
+      "step": 1374
+    },
+    {
+      "epoch": 2.52014652014652,
+      "grad_norm": 0.9609295129776001,
+      "learning_rate": 3.1909762641363083e-06,
+      "loss": 0.9668674468994141,
+      "step": 1376
+    },
+    {
+      "epoch": 2.5238095238095237,
+      "grad_norm": 0.35177093744277954,
+      "learning_rate": 3.173334220166962e-06,
+      "loss": 0.7824784517288208,
+      "step": 1378
+    },
+    {
+      "epoch": 2.5274725274725274,
+      "grad_norm": 0.26057368516921997,
+      "learning_rate": 3.155814704089823e-06,
+      "loss": 0.8903582692146301,
+      "step": 1380
+    },
+    {
+      "epoch": 2.531135531135531,
+      "grad_norm": 0.06582798063755035,
+      "learning_rate": 3.1384179901761343e-06,
+      "loss": 0.8639101982116699,
+      "step": 1382
+    },
+    {
+      "epoch": 2.5347985347985347,
+      "grad_norm": 0.33872169256210327,
+      "learning_rate": 3.1211443507746546e-06,
+      "loss": 0.7889021039009094,
+      "step": 1384
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 0.21574373543262482,
+      "learning_rate": 3.1039940563073894e-06,
+      "loss": 1.0973457098007202,
+      "step": 1386
+    },
+    {
+      "epoch": 2.542124542124542,
+      "grad_norm": 0.23947708308696747,
+      "learning_rate": 3.0869673752653447e-06,
+      "loss": 0.755636990070343,
+      "step": 1388
+    },
+    {
+      "epoch": 2.5457875457875456,
+      "grad_norm": 0.2152792364358902,
+      "learning_rate": 3.0700645742043476e-06,
+      "loss": 0.8107748627662659,
+      "step": 1390
+    },
+    {
+      "epoch": 2.5494505494505493,
+      "grad_norm": 0.23220431804656982,
+      "learning_rate": 3.0532859177408587e-06,
+      "loss": 0.9791698455810547,
+      "step": 1392
+    },
+    {
+      "epoch": 2.553113553113553,
+      "grad_norm": 0.3370932638645172,
+      "learning_rate": 3.03663166854783e-06,
+      "loss": 0.7153705358505249,
+      "step": 1394
+    },
+    {
+      "epoch": 2.5567765567765566,
+      "grad_norm": 0.1939450353384018,
+      "learning_rate": 3.020102087350594e-06,
+      "loss": 0.3394162356853485,
+      "step": 1396
+    },
+    {
+      "epoch": 2.5604395604395602,
+      "grad_norm": 0.34136688709259033,
+      "learning_rate": 3.0036974329227862e-06,
+      "loss": 1.1137323379516602,
+      "step": 1398
+    },
+    {
+      "epoch": 2.564102564102564,
+      "grad_norm": 0.5828881859779358,
+      "learning_rate": 2.9874179620822856e-06,
+      "loss": 1.1080697774887085,
+      "step": 1400
+    },
+    {
+      "epoch": 2.5677655677655675,
+      "grad_norm": 0.3508465886116028,
+      "learning_rate": 2.971263929687207e-06,
+      "loss": 0.892469048500061,
+      "step": 1402
+    },
+    {
+      "epoch": 2.571428571428571,
+      "grad_norm": 0.27003780007362366,
+      "learning_rate": 2.9552355886318968e-06,
+      "loss": 1.1012053489685059,
+      "step": 1404
+    },
+    {
+      "epoch": 2.575091575091575,
+      "grad_norm": 4.884339809417725,
+      "learning_rate": 2.9393331898429777e-06,
+      "loss": 0.46851903200149536,
+      "step": 1406
+    },
+    {
+      "epoch": 2.578754578754579,
+      "grad_norm": 0.24890904128551483,
+      "learning_rate": 2.9235569822754317e-06,
+      "loss": 1.1480822563171387,
+      "step": 1408
+    },
+    {
+      "epoch": 2.5824175824175826,
+      "grad_norm": 0.41308149695396423,
+      "learning_rate": 2.9079072129086906e-06,
+      "loss": 0.8952036499977112,
+      "step": 1410
+    },
+    {
+      "epoch": 2.586080586080586,
+      "grad_norm": 0.7369239330291748,
+      "learning_rate": 2.89238412674277e-06,
+      "loss": 0.8408064246177673,
+      "step": 1412
+    },
+    {
+      "epoch": 2.58974358974359,
+      "grad_norm": 0.5877048969268799,
+      "learning_rate": 2.8769879667944393e-06,
+      "loss": 0.9505341053009033,
+      "step": 1414
+    },
+    {
+      "epoch": 2.5934065934065935,
+      "grad_norm": 0.2738566994667053,
+      "learning_rate": 2.8617189740934113e-06,
+      "loss": 1.0852067470550537,
+      "step": 1416
+    },
+    {
+      "epoch": 2.597069597069597,
+      "grad_norm": 0.7545573711395264,
+      "learning_rate": 2.8465773876785786e-06,
+      "loss": 0.7618016600608826,
+      "step": 1418
+    },
+    {
+      "epoch": 2.600732600732601,
+      "grad_norm": 0.4518103301525116,
+      "learning_rate": 2.8315634445942623e-06,
+      "loss": 1.0812193155288696,
+      "step": 1420
+    },
+    {
+      "epoch": 2.6043956043956045,
+      "grad_norm": 0.3092360496520996,
+      "learning_rate": 2.8166773798864978e-06,
+      "loss": 1.0327297449111938,
+      "step": 1422
+    },
+    {
+      "epoch": 2.608058608058608,
+      "grad_norm": 0.30283570289611816,
+      "learning_rate": 2.8019194265993683e-06,
+      "loss": 1.1834499835968018,
+      "step": 1424
+    },
+    {
+      "epoch": 2.6117216117216118,
+      "grad_norm": 0.4102247357368469,
+      "learning_rate": 2.787289815771348e-06,
+      "loss": 1.272240161895752,
+      "step": 1426
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 0.2674598693847656,
+      "learning_rate": 2.7727887764316835e-06,
+      "loss": 1.137374997138977,
+      "step": 1428
+    },
+    {
+      "epoch": 2.619047619047619,
+      "grad_norm": 0.8950290083885193,
+      "learning_rate": 2.758416535596812e-06,
+      "loss": 1.0797327756881714,
+      "step": 1430
+    },
+    {
+      "epoch": 2.6227106227106227,
+      "grad_norm": 0.3816538155078888,
+      "learning_rate": 2.744173318266809e-06,
+      "loss": 0.9714991450309753,
+      "step": 1432
+    },
+    {
+      "epoch": 2.6263736263736264,
+      "grad_norm": 0.43699485063552856,
+      "learning_rate": 2.7300593474218583e-06,
+      "loss": 0.9439188838005066,
+      "step": 1434
+    },
+    {
+      "epoch": 2.63003663003663,
+      "grad_norm": 0.805819571018219,
+      "learning_rate": 2.7160748440187736e-06,
+      "loss": 1.3011527061462402,
+      "step": 1436
+    },
+    {
+      "epoch": 2.6336996336996337,
+      "grad_norm": 2.7278735637664795,
+      "learning_rate": 2.702220026987525e-06,
+      "loss": 1.1346534490585327,
+      "step": 1438
+    },
+    {
+      "epoch": 2.6373626373626373,
+      "grad_norm": 0.21300676465034485,
+      "learning_rate": 2.6884951132278185e-06,
+      "loss": 1.095414161682129,
+      "step": 1440
+    },
+    {
+      "epoch": 2.641025641025641,
+      "grad_norm": 0.16259470582008362,
+      "learning_rate": 2.6749003176057092e-06,
+      "loss": 1.1217632293701172,
+      "step": 1442
+    },
+    {
+      "epoch": 2.6446886446886446,
+      "grad_norm": 0.4592875838279724,
+      "learning_rate": 2.6614358529502165e-06,
+      "loss": 1.0697576999664307,
+      "step": 1444
+    },
+    {
+      "epoch": 2.6483516483516483,
+      "grad_norm": 0.3081722557544708,
+      "learning_rate": 2.6481019300500166e-06,
+      "loss": 0.7924908995628357,
+      "step": 1446
+    },
+    {
+      "epoch": 2.652014652014652,
+      "grad_norm": 0.12425895035266876,
+      "learning_rate": 2.634898757650121e-06,
+      "loss": 0.8344003558158875,
+      "step": 1448
+    },
+    {
+      "epoch": 2.6556776556776556,
+      "grad_norm": 0.213938370347023,
+      "learning_rate": 2.6218265424486233e-06,
+      "loss": 1.090872049331665,
+      "step": 1450
+    },
+    {
+      "epoch": 2.659340659340659,
+      "grad_norm": 0.21575380861759186,
+      "learning_rate": 2.608885489093455e-06,
+      "loss": 0.7010395526885986,
+      "step": 1452
+    },
+    {
+      "epoch": 2.663003663003663,
+      "grad_norm": 0.07869178056716919,
+      "learning_rate": 2.5960758001791893e-06,
+      "loss": 0.6901416182518005,
+      "step": 1454
+    },
+    {
+      "epoch": 2.6666666666666665,
+      "grad_norm": 0.32918277382850647,
+      "learning_rate": 2.5833976762438605e-06,
+      "loss": 1.0698477029800415,
+      "step": 1456
+    },
+    {
+      "epoch": 2.67032967032967,
+      "grad_norm": 0.39367684721946716,
+      "learning_rate": 2.5708513157658295e-06,
+      "loss": 0.7514879107475281,
+      "step": 1458
+    },
+    {
+      "epoch": 2.6739926739926743,
+      "grad_norm": 0.07575741410255432,
+      "learning_rate": 2.5584369151606785e-06,
+      "loss": 0.7730086445808411,
+      "step": 1460
+    },
+    {
+      "epoch": 2.677655677655678,
+      "grad_norm": 0.2596700191497803,
+      "learning_rate": 2.5461546687781325e-06,
+      "loss": 0.7882156372070312,
+      "step": 1462
+    },
+    {
+      "epoch": 2.6813186813186816,
+      "grad_norm": 0.3725094199180603,
+      "learning_rate": 2.5340047688990142e-06,
+      "loss": 1.0619717836380005,
+      "step": 1464
+    },
+    {
+      "epoch": 2.684981684981685,
+      "grad_norm": 0.4336903989315033,
+      "learning_rate": 2.5219874057322453e-06,
+      "loss": 0.8744177222251892,
+      "step": 1466
+    },
+    {
+      "epoch": 2.688644688644689,
+      "grad_norm": 0.31862640380859375,
+      "learning_rate": 2.5101027674118523e-06,
+      "loss": 1.03647780418396,
+      "step": 1468
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 0.636555552482605,
+      "learning_rate": 2.4983510399940377e-06,
+      "loss": 0.6445450186729431,
+      "step": 1470
+    },
+    {
+      "epoch": 2.695970695970696,
+      "grad_norm": 0.5853780508041382,
+      "learning_rate": 2.4867324074542525e-06,
+      "loss": 0.7737827301025391,
+      "step": 1472
+    },
+    {
+      "epoch": 2.6996336996337,
+      "grad_norm": 0.12945802509784698,
+      "learning_rate": 2.4752470516843257e-06,
+      "loss": 0.43008971214294434,
+      "step": 1474
+    },
+    {
+      "epoch": 2.7032967032967035,
+      "grad_norm": 0.24216750264167786,
+      "learning_rate": 2.463895152489617e-06,
+      "loss": 1.1383622884750366,
+      "step": 1476
+    },
+    {
+      "epoch": 2.706959706959707,
+      "grad_norm": 0.29354915022850037,
+      "learning_rate": 2.4526768875861938e-06,
+      "loss": 0.6484386324882507,
+      "step": 1478
+    },
+    {
+      "epoch": 2.7106227106227108,
+      "grad_norm": 0.2161005586385727,
+      "learning_rate": 2.4415924325980575e-06,
+      "loss": 1.0878047943115234,
+      "step": 1480
+    },
+    {
+      "epoch": 2.7142857142857144,
+      "grad_norm": 0.2699625492095947,
+      "learning_rate": 2.4306419610543885e-06,
+      "loss": 1.0988802909851074,
+      "step": 1482
+    },
+    {
+      "epoch": 2.717948717948718,
+      "grad_norm": 0.20100824534893036,
+      "learning_rate": 2.4198256443868327e-06,
+      "loss": 1.0038988590240479,
+      "step": 1484
+    },
+    {
+      "epoch": 2.7216117216117217,
+      "grad_norm": 0.37425243854522705,
+      "learning_rate": 2.4091436519268167e-06,
+      "loss": 1.1373145580291748,
+      "step": 1486
+    },
+    {
+      "epoch": 2.7252747252747254,
+      "grad_norm": 0.8878154158592224,
+      "learning_rate": 2.3985961509028994e-06,
+      "loss": 1.1287609338760376,
+      "step": 1488
+    },
+    {
+      "epoch": 2.728937728937729,
+      "grad_norm": 0.36207541823387146,
+      "learning_rate": 2.3881833064381478e-06,
+      "loss": 0.5098522901535034,
+      "step": 1490
+    },
+    {
+      "epoch": 2.7326007326007327,
+      "grad_norm": 0.2365618795156479,
+      "learning_rate": 2.3779052815475553e-06,
+      "loss": 1.1206612586975098,
+      "step": 1492
+    },
+    {
+      "epoch": 2.7362637362637363,
+      "grad_norm": 0.5252121090888977,
+      "learning_rate": 2.3677622371354932e-06,
+      "loss": 0.6413739919662476,
+      "step": 1494
+    },
+    {
+      "epoch": 2.73992673992674,
+      "grad_norm": 0.575518012046814,
+      "learning_rate": 2.357754331993187e-06,
+      "loss": 1.2001720666885376,
+      "step": 1496
+    },
+    {
+      "epoch": 2.7435897435897436,
+      "grad_norm": 0.3244689404964447,
+      "learning_rate": 2.347881722796234e-06,
+      "loss": 1.0752826929092407,
+      "step": 1498
+    },
+    {
+      "epoch": 2.7472527472527473,
+      "grad_norm": 0.33936455845832825,
+      "learning_rate": 2.3381445641021445e-06,
+      "loss": 0.7569327354431152,
+      "step": 1500
+    },
+    {
+      "epoch": 2.750915750915751,
+      "grad_norm": 0.28682929277420044,
+      "learning_rate": 2.328543008347928e-06,
+      "loss": 1.0332515239715576,
+      "step": 1502
+    },
+    {
+      "epoch": 2.7545787545787546,
+      "grad_norm": 0.23232409358024597,
+      "learning_rate": 2.31907720584771e-06,
+      "loss": 1.1103003025054932,
+      "step": 1504
+    },
+    {
+      "epoch": 2.758241758241758,
+      "grad_norm": 0.2504388988018036,
+      "learning_rate": 2.3097473047903645e-06,
+      "loss": 1.1210606098175049,
+      "step": 1506
+    },
+    {
+      "epoch": 2.761904761904762,
+      "grad_norm": 0.05474567040801048,
+      "learning_rate": 2.3005534512372106e-06,
+      "loss": 0.9155768752098083,
+      "step": 1508
+    },
+    {
+      "epoch": 2.7655677655677655,
+      "grad_norm": 0.4082033932209015,
+      "learning_rate": 2.2914957891197182e-06,
+      "loss": 0.9890879988670349,
+      "step": 1510
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 0.2753402292728424,
+      "learning_rate": 2.2825744602372506e-06,
+      "loss": 0.9224788546562195,
+      "step": 1512
+    },
+    {
+      "epoch": 2.772893772893773,
+      "grad_norm": 0.3398146629333496,
+      "learning_rate": 2.2737896042548537e-06,
+      "loss": 1.0887320041656494,
+      "step": 1514
+    },
+    {
+      "epoch": 2.7765567765567765,
+      "grad_norm": 0.31702274084091187,
+      "learning_rate": 2.2651413587010634e-06,
+      "loss": 1.0131077766418457,
+      "step": 1516
+    },
+    {
+      "epoch": 2.78021978021978,
+      "grad_norm": 0.06860560923814774,
+      "learning_rate": 2.2566298589657546e-06,
+      "loss": 0.9779605865478516,
+      "step": 1518
+    },
+    {
+      "epoch": 2.7838827838827838,
+      "grad_norm": 0.3455958664417267,
+      "learning_rate": 2.2482552382980194e-06,
+      "loss": 0.5144427418708801,
+      "step": 1520
+    },
+    {
+      "epoch": 2.7875457875457874,
+      "grad_norm": 0.2682342231273651,
+      "learning_rate": 2.240017627804088e-06,
+      "loss": 0.7746493816375732,
+      "step": 1522
+    },
+    {
+      "epoch": 2.791208791208791,
+      "grad_norm": 0.24560944736003876,
+      "learning_rate": 2.231917156445265e-06,
+      "loss": 0.825928807258606,
+      "step": 1524
+    },
+    {
+      "epoch": 2.7948717948717947,
+      "grad_norm": 0.5874441862106323,
+      "learning_rate": 2.223953951035919e-06,
+      "loss": 1.140251874923706,
+      "step": 1526
+    },
+    {
+      "epoch": 2.7985347985347984,
+      "grad_norm": 0.2171962410211563,
+      "learning_rate": 2.216128136241497e-06,
+      "loss": 1.0953679084777832,
+      "step": 1528
+    },
+    {
+      "epoch": 2.802197802197802,
+      "grad_norm": 0.5042185187339783,
+      "learning_rate": 2.208439834576568e-06,
+      "loss": 1.093937635421753,
+      "step": 1530
+    },
+    {
+      "epoch": 2.8058608058608057,
+      "grad_norm": 0.2920451760292053,
+      "learning_rate": 2.200889166402908e-06,
+      "loss": 0.9886079430580139,
+      "step": 1532
+    },
+    {
+      "epoch": 2.8095238095238093,
+      "grad_norm": 0.5010365843772888,
+      "learning_rate": 2.193476249927617e-06,
+      "loss": 0.7637280821800232,
+      "step": 1534
+    },
+    {
+      "epoch": 2.813186813186813,
+      "grad_norm": 0.313466340303421,
+      "learning_rate": 2.1862012012012647e-06,
+      "loss": 1.2877520322799683,
+      "step": 1536
+    },
+    {
+      "epoch": 2.8168498168498166,
+      "grad_norm": 0.2607482373714447,
+      "learning_rate": 2.179064134116078e-06,
+      "loss": 0.9093769192695618,
+      "step": 1538
+    },
+    {
+      "epoch": 2.8205128205128203,
+      "grad_norm": 0.9507859349250793,
+      "learning_rate": 2.1720651604041543e-06,
+      "loss": 0.7832710146903992,
+      "step": 1540
+    },
+    {
+      "epoch": 2.824175824175824,
+      "grad_norm": 0.5445396900177002,
+      "learning_rate": 2.1652043896357132e-06,
+      "loss": 0.8669724464416504,
+      "step": 1542
+    },
+    {
+      "epoch": 2.8278388278388276,
+      "grad_norm": 0.2889232933521271,
+      "learning_rate": 2.1584819292173844e-06,
+      "loss": 1.0738338232040405,
+      "step": 1544
+    },
+    {
+      "epoch": 2.8315018315018317,
+      "grad_norm": 0.3314497172832489,
+      "learning_rate": 2.1518978843905204e-06,
+      "loss": 1.120241641998291,
+      "step": 1546
+    },
+    {
+      "epoch": 2.8351648351648353,
+      "grad_norm": 0.43042734265327454,
+      "learning_rate": 2.1454523582295567e-06,
+      "loss": 1.1897751092910767,
+      "step": 1548
+    },
+    {
+      "epoch": 2.838827838827839,
+      "grad_norm": 0.41529417037963867,
+      "learning_rate": 2.1391454516403876e-06,
+      "loss": 0.8347918391227722,
+      "step": 1550
+    },
+    {
+      "epoch": 2.8424908424908426,
+      "grad_norm": 0.06782124936580658,
+      "learning_rate": 2.1329772633587976e-06,
+      "loss": 0.5578287839889526,
+      "step": 1552
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 0.28071868419647217,
+      "learning_rate": 2.1269478899489068e-06,
+      "loss": 1.1331698894500732,
+      "step": 1554
+    },
+    {
+      "epoch": 2.84981684981685,
+      "grad_norm": 4.985344409942627,
+      "learning_rate": 2.1210574258016675e-06,
+      "loss": 0.910840630531311,
+      "step": 1556
+    },
+    {
+      "epoch": 2.8534798534798536,
+      "grad_norm": 5.596924781799316,
+      "learning_rate": 2.1153059631333785e-06,
+      "loss": 1.0615748167037964,
+      "step": 1558
+    },
+    {
+      "epoch": 2.857142857142857,
+      "grad_norm": 0.5638831853866577,
+      "learning_rate": 2.1096935919842434e-06,
+      "loss": 0.5931094288825989,
+      "step": 1560
+    },
+    {
+      "epoch": 2.860805860805861,
+      "grad_norm": 0.3857717514038086,
+      "learning_rate": 2.104220400216967e-06,
+      "loss": 0.7401585578918457,
+      "step": 1562
+    },
+    {
+      "epoch": 2.8644688644688645,
+      "grad_norm": 0.32050344347953796,
+      "learning_rate": 2.0988864735153724e-06,
+      "loss": 0.8747660517692566,
+      "step": 1564
+    },
+    {
+      "epoch": 2.868131868131868,
+      "grad_norm": 1.977245569229126,
+      "learning_rate": 2.0936918953830633e-06,
+      "loss": 0.6961140036582947,
+      "step": 1566
+    },
+    {
+      "epoch": 2.871794871794872,
+      "grad_norm": 0.33985045552253723,
+      "learning_rate": 2.088636747142114e-06,
+      "loss": 0.7156071662902832,
+      "step": 1568
+    },
+    {
+      "epoch": 2.8754578754578755,
+      "grad_norm": 0.1318127065896988,
+      "learning_rate": 2.083721107931803e-06,
+      "loss": 0.7040363550186157,
+      "step": 1570
+    },
+    {
+      "epoch": 2.879120879120879,
+      "grad_norm": 0.434373676776886,
+      "learning_rate": 2.0789450547073634e-06,
+      "loss": 0.6764128804206848,
+      "step": 1572
+    },
+    {
+      "epoch": 2.8827838827838828,
+      "grad_norm": 0.7847844958305359,
+      "learning_rate": 2.074308662238789e-06,
+      "loss": 0.9994201064109802,
+      "step": 1574
+    },
+    {
+      "epoch": 2.8864468864468864,
+      "grad_norm": 5.1912055015563965,
+      "learning_rate": 2.069812003109654e-06,
+      "loss": 0.9506621956825256,
+      "step": 1576
+    },
+    {
+      "epoch": 2.89010989010989,
+      "grad_norm": 0.22814835608005524,
+      "learning_rate": 2.0654551477159868e-06,
+      "loss": 0.8836541175842285,
+      "step": 1578
+    },
+    {
+      "epoch": 2.8937728937728937,
+      "grad_norm": 1.1357293128967285,
+      "learning_rate": 2.0612381642651584e-06,
+      "loss": 1.1715425252914429,
+      "step": 1580
+    },
+    {
+      "epoch": 2.8974358974358974,
+      "grad_norm": 0.6360121369361877,
+      "learning_rate": 2.057161118774821e-06,
+      "loss": 0.8418415784835815,
+      "step": 1582
+    },
+    {
+      "epoch": 2.901098901098901,
+      "grad_norm": 0.03927984833717346,
+      "learning_rate": 2.05322407507187e-06,
+      "loss": 0.7738023996353149,
+      "step": 1584
+    },
+    {
+      "epoch": 2.9047619047619047,
+      "grad_norm": 0.14784355461597443,
+      "learning_rate": 2.0494270947914507e-06,
+      "loss": 0.785838782787323,
+      "step": 1586
+    },
+    {
+      "epoch": 2.9084249084249083,
+      "grad_norm": 0.17373321950435638,
+      "learning_rate": 2.0457702373759864e-06,
+      "loss": 1.1310051679611206,
+      "step": 1588
+    },
+    {
+      "epoch": 2.912087912087912,
+      "grad_norm": 0.2693270444869995,
+      "learning_rate": 2.0422535600742526e-06,
+      "loss": 1.1932168006896973,
+      "step": 1590
+    },
+    {
+      "epoch": 2.9157509157509156,
+      "grad_norm": 0.29240870475769043,
+      "learning_rate": 2.03887711794048e-06,
+      "loss": 0.5143145322799683,
+      "step": 1592
+    },
+    {
+      "epoch": 2.9194139194139193,
+      "grad_norm": 0.2930733859539032,
+      "learning_rate": 2.0356409638334902e-06,
+      "loss": 1.1435730457305908,
+      "step": 1594
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 0.4619033634662628,
+      "learning_rate": 2.032545148415871e-06,
+      "loss": 0.7462179064750671,
+      "step": 1596
+    },
+    {
+      "epoch": 2.926739926739927,
+      "grad_norm": 0.5609884858131409,
+      "learning_rate": 2.0295897201531838e-06,
+      "loss": 1.1510316133499146,
+      "step": 1598
+    },
+    {
+      "epoch": 2.9304029304029307,
+      "grad_norm": 0.1557919830083847,
+      "learning_rate": 2.026774725313199e-06,
+      "loss": 0.7830919623374939,
+      "step": 1600
+    },
+    {
+      "epoch": 2.9340659340659343,
+      "grad_norm": 0.5669822692871094,
+      "learning_rate": 2.0241002079651803e-06,
+      "loss": 1.1249396800994873,
+      "step": 1602
+    },
+    {
+      "epoch": 2.937728937728938,
+      "grad_norm": 0.2051316201686859,
+      "learning_rate": 2.0215662099791874e-06,
+      "loss": 0.8561792969703674,
+      "step": 1604
+    },
+    {
+      "epoch": 2.9413919413919416,
+      "grad_norm": 0.621616005897522,
+      "learning_rate": 2.019172771025426e-06,
+      "loss": 1.1106517314910889,
+      "step": 1606
+    },
+    {
+      "epoch": 2.9450549450549453,
+      "grad_norm": 0.3406311869621277,
+      "learning_rate": 2.0169199285736234e-06,
+      "loss": 0.6978309154510498,
+      "step": 1608
+    },
+    {
+      "epoch": 2.948717948717949,
+      "grad_norm": 0.1934393346309662,
+      "learning_rate": 2.0148077178924412e-06,
+      "loss": 1.0018409490585327,
+      "step": 1610
+    },
+    {
+      "epoch": 2.9523809523809526,
+      "grad_norm": 0.5513986349105835,
+      "learning_rate": 2.0128361720489263e-06,
+      "loss": 0.868675708770752,
+      "step": 1612
+    },
+    {
+      "epoch": 2.956043956043956,
+      "grad_norm": 0.5031806826591492,
+      "learning_rate": 2.0110053219079927e-06,
+      "loss": 0.6836384534835815,
+      "step": 1614
+    },
+    {
+      "epoch": 2.95970695970696,
+      "grad_norm": 1.7091178894042969,
+      "learning_rate": 2.009315196131934e-06,
+      "loss": 0.8271127343177795,
+      "step": 1616
+    },
+    {
+      "epoch": 2.9633699633699635,
+      "grad_norm": 0.2841132581233978,
+      "learning_rate": 2.0077658211799823e-06,
+      "loss": 1.3688499927520752,
+      "step": 1618
+    },
+    {
+      "epoch": 2.967032967032967,
+      "grad_norm": 0.20573776960372925,
+      "learning_rate": 2.0063572213078856e-06,
+      "loss": 1.2264219522476196,
+      "step": 1620
+    },
+    {
+      "epoch": 2.970695970695971,
+      "grad_norm": 0.29216548800468445,
+      "learning_rate": 2.0050894185675354e-06,
+      "loss": 0.9183750152587891,
+      "step": 1622
+    },
+    {
+      "epoch": 2.9743589743589745,
+      "grad_norm": 0.47401827573776245,
+      "learning_rate": 2.0039624328066154e-06,
+      "loss": 0.7726737260818481,
+      "step": 1624
+    },
+    {
+      "epoch": 2.978021978021978,
+      "grad_norm": 0.17664581537246704,
+      "learning_rate": 2.0029762816682963e-06,
+      "loss": 0.8954548239707947,
+      "step": 1626
+    },
+    {
+      "epoch": 2.9816849816849818,
+      "grad_norm": 0.2111157476902008,
+      "learning_rate": 2.0021309805909546e-06,
+      "loss": 0.9459641575813293,
+      "step": 1628
+    },
+    {
+      "epoch": 2.9853479853479854,
+      "grad_norm": 0.19905273616313934,
+      "learning_rate": 2.001426542807935e-06,
+      "loss": 1.3968180418014526,
+      "step": 1630
+    },
+    {
+      "epoch": 2.989010989010989,
+      "grad_norm": 0.42986905574798584,
+      "learning_rate": 2.000862979347339e-06,
+      "loss": 1.0010343790054321,
+      "step": 1632
+    },
+    {
+      "epoch": 2.9926739926739927,
+      "grad_norm": 0.26831236481666565,
+      "learning_rate": 2.0004402990318574e-06,
+      "loss": 0.7905582189559937,
+      "step": 1634
+    },
+    {
+      "epoch": 2.9963369963369964,
+      "grad_norm": 0.35902222990989685,
+      "learning_rate": 2.000158508478629e-06,
+      "loss": 1.2652593851089478,
+      "step": 1636
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 0.369983971118927,
+      "learning_rate": 2.0000176120991345e-06,
+      "loss": 1.0647433996200562,
+      "step": 1638
+    },
+    {
+      "epoch": 3.0,
+      "step": 1638,
+      "total_flos": 8.4482141520606e+18,
+      "train_loss": 1.0226961683586548,
+      "train_runtime": 55136.4168,
+      "train_samples_per_second": 0.713,
+      "train_steps_per_second": 0.03
+    }
+  ],
+  "logging_steps": 2,
+  "max_steps": 1638,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 99999,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 8.4482141520606e+18,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}